Compare commits

..

1605 Commits

Author SHA1 Message Date
Stéphane Nicolet b4c239b625 Stockfish 10
Official release version of Stockfish 10.

This is also the 10th anniversary version of the Stockfish project, which
started exactly ten years ago! I wish to extend a huge thank you to
all contributors and authors in our amazing community :-)

Bench: 3939338
2018-11-29 15:45:26 +01:00
Stéphane Nicolet 622360ad66 Update list of authors
No functional change
2018-11-29 15:15:43 +01:00
Sebastian Buchwald 340e9ea509 Use emplace_back() in TB code
The patch was tested for correctness by running bench with and
without the change against current master, and the tablebase hit
numbers were found to be identical in both cases. See the pull
request comments for details:
https://github.com/official-stockfish/Stockfish/pull/1826

No functional change.
2018-11-29 15:01:54 +01:00
31m059 7b6fa353a3 Simplify casting extension
On November 16th, before the removal of the depth condition, I tried
revising castling extensions to only handle castling moves, rather than
moves that change castling rights generally. It appeared to be a slight
Elo gain at STC but insufficient to pass [0, 4] (+0.5 Elo), but what I
overlooked was that it made pos.can_castle(us) irrelevant and should
have been a simplification. Recent discussion with @Chess13234 and
Michael Chaly (@Vizvezdenec) inspired me to take a second look, and
the simplification continues to pass when rebased on the current master.

This replaces two conditions with one, because type_of(move) == CASTLING
implies pos.can_castle(Us), allowing us to remove the latter condition.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 110948 W: 24209 L: 24263 D: 62476
http://tests.stockfishchess.org/tests/view/5bf8f65c0ebc5902bced3a63

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 88283 W: 14681 L: 14668 D: 58934
http://tests.stockfishchess.org/tests/view/5bf994a60ebc5902bced4349

Bench: 3939338
2018-11-27 08:53:14 +01:00
Steinar H. Gunderson de7182f4ee Turn on MADV_RANDOM for Syzygy mmaps (on Unix-like builds)
When running on a cloud VM (n1-highcpu-96) with several NVMe SSDs and
some non-SSDs for tablebases, I noticed that the average SSD request size was
more than 256 kB. This doesn't make a lot of sense for Syzygy tablebases,
which have a block size of 32 bytes and very low locality.

Seemingly, the tablebase access patterns during probing make the OS,
at least Linux, think that readahead is advantageous; normally, it
gives up doing readahead if there are too many misses, but it doesn't,
perhaps due to the fairly high overall hit rates. (It seems the kernel cannot
distinguish between reading a block that was paged in because the userspace
wanted it explicitly, and one that was read as part of readahead.)

Setting MADV_RANDOM effectively turns off readahead, which causes
the request size to drop to 4 kB. In the aforemented cloud VM test,
this roughly tripled the amount of I/O requests that were able to go
through, while reducing the total traffic from 2.8 GB/sec to 56 MB/sec
(moving the bottleneck to the non-SSDs; it seems the SSDs could have
sustained many more requests).

Closes https://github.com/official-stockfish/Stockfish/pull/1829

No functional change.
2018-11-27 08:39:23 +01:00
Jörg Oster 6ab92d2e1c Qsearch simplification. (#1828)
Don't do an extra TT update in case of a fail-high,
but simply break off the moves loop and let the TT update
at the end of qsearch do this job.
Same workflow/logic as in our main search function now.

Tested for no regression to be on the safe side.
STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 30237 W: 6665 L: 6560 D: 17012
http://tests.stockfishchess.org/tests/view/5bf928e80ebc5902bced3f3a

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 51067 W: 8625 L: 8553 D: 33889
http://tests.stockfishchess.org/tests/view/5bf937180ebc5902bced3fdc

No functional change.
2018-11-25 11:27:40 +01:00
Vizvezdenec bb58bc215c Reintroduce tropism to kingdanger
Tropism in kingdanger was simplified away in this pull request #1821.
This patch reintroduces tropism in kingdanger with using quadratic scaling.

Passed STC http://tests.stockfishchess.org/tests/view/5bf7c1b10ebc5902bced1f8f
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 52803 W: 11835 L: 11442 D: 29526

Passed LTC http://tests.stockfishchess.org/tests/view/5bf816e90ebc5902bced24f1
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 17204 W: 2988 L: 2795 D: 11421

How do we continue from there?

I've recently tried to introduce tropism difference term in kingdanger which
passed STC 6 times but failed LTC all the time. Maybe using quadratic scaling
for it will also be helpful.

Bench 4041387
2018-11-24 02:14:18 +01:00
31m059 6e66e7aae2 Remove the tropism term from kingDanger
A recent LTC tuning session by @candirufish showed this term decreasing significantly. It appears that it can be removed altogether without significant Elo loss.

I also thank @GuardianRM, whose attempt to remove tropism from king danger inspired this one.

After this PR is merged, my next step will be to attempt to tune the coefficients of this new, simplified kingDanger calculation.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 12518 W: 2795 L: 2656 D: 7067
http://tests.stockfishchess.org/tests/view/5befadda0ebc595e0ae3a289

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 164771 W: 26463 L: 26566 D: 111742
http://tests.stockfishchess.org/tests/view/5befcca70ebc595e0ae3a343

LTC 2, rebased on Stockfish 10 beta:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 75226 W: 12563 L: 12529 D: 50134
http://tests.stockfishchess.org/tests/view/5bf2e8910ebc5902bcecb919

Bench: 3412071
2018-11-24 02:09:35 +01:00
Joost VandeVondele 4b88bea4fc Force time check on TB probe in search.
Because of aggressive time management and optimistic assumptions
about move overhead, it's still very easy to get Stockfish to forfeit
on time when we hit an endgame and have Syzygy EGTB on a spinning
drive. The latency from serving a few thousand EGTB probes (~10ms each),
of which there can currently be up to 4000 outstanding before a time
check, will easily overwhelm the default Move Overhead of 30ms.

This problem was first raised by Gian-Carlo Pascutto and some solutions
and improvements were discussed in the following pull requests:
https://github.com/official-stockfish/Stockfish/pull/1471
https://github.com/official-stockfish/Stockfish/pull/1623
https://github.com/official-stockfish/Stockfish/pull/1783

This patch is a minimal change proposed by Marco Costalba to lower
the impact of the bug. We now force a check of the clock right after
each tablebase read.

No functional change.
2018-11-20 08:00:19 +01:00
xoto10 3925750945 Bonus for restricting opponent's piece moves
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 51883 W: 11297 L: 10915 D: 29671
http://tests.stockfishchess.org/tests/view/5bf1e2ee0ebc595e0ae3cacd

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 15859 W: 2752 L: 2565 D: 10542
http://tests.stockfishchess.org/tests/view/5bf337980ebc5902bcecbf62

Notes:

(1) The bonus value has not been carefully tested, so it may be possible
to find slightly better values.

(2) Plan is to now try adding similar restriction for pawns. I wanted to
include that as part of this pull request, but I was advised to do it as
two separate pull requests. STC is currently running here, but may not add
enough value to pass green.

Bench: 3679086
2018-11-20 07:50:12 +01:00
Stéphane Nicolet cf5d683408 Stockfish 10-beta
Preparation commit for the upcoming Stockfish 10 version, giving a chance to catch last minute feature bugs and evaluation regression during the one-week code freeze period. Also changing the copyright dates to include 2019.

No functional change
2018-11-19 11:18:21 +01:00
SFisGOD 3f2ec5b3d5 Tweak Queen PSQT based on tuned values
STC: (Yellow)
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 63140 W: 13433 L: 13353 D: 36354
http://tests.stockfishchess.org/tests/view/5bed42c90ebc595e0ae37cf5

LTC: (Green)
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 47714 W: 7785 L: 7485 D: 32444
http://tests.stockfishchess.org/tests/view/5bec3b8c0ebc595e0ae36dec

Closes https://github.com/official-stockfish/Stockfish/pull/1816

Bench: 3717396
2018-11-19 10:50:55 +01:00
Kurt b92206305f Tune evaluation scores
STC:
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 84697 W: 18173 L: 18009 D: 48515
http://tests.stockfishchess.org/tests/view/5bea366f0ebc595e0ae34793

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 157625 W: 25533 L: 24893 D: 107199
http://tests.stockfishchess.org/tests/view/5be8b69e0ebc595e0ae33024

Personally, I feel like SF has been tuned to death recently and that we
need to step away from existing-parameter tunes for a bit and focus more
on new ideas. I don't really think there's much more ELO in these tunes
(for now). For me at least, this was the last existing-parameter tune I'll
be running for quite a while. Cheers!

Bench: 3572567
2018-11-19 10:42:46 +01:00
protonspring d2274e609c Remove BlockedStorm array
Apparently, only RANK_3 is relevant. This removes a look-up and the
BlockedStorm array, but adds another conditional.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 84340 W: 18054 L: 18054 D: 48232
http://tests.stockfishchess.org/tests/view/5bea10f40ebc595e0ae3457b

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 31874 W: 5135 L: 5032 D: 21707
http://tests.stockfishchess.org/tests/view/5beadb6a0ebc595e0ae35542

Closes https://github.com/official-stockfish/Stockfish/pull/1814

Bench: 3799443
2018-11-19 10:37:34 +01:00
VoyagerOne 4111f36f45 Simplify Castle Extension
Remove depth condition in castle extension, also don't extend if
Singular Extension and Check Extansion fail to extend.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 42070 W: 9118 L: 9036 D: 23916
http://tests.stockfishchess.org/tests/view/5be899cc0ebc595e0ae32f07

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 78278 W: 12490 L: 12458 D: 53330
http://tests.stockfishchess.org/tests/view/5be8ac420ebc595e0ae33010

Bench: 3611041
2018-11-19 10:28:55 +01:00
protonspring 0e508f30bb Code style in search.cpp
It does not appear to be not necessary or advantageous to
conditionally initialize kingRing[Us] or kingAttackersCount[Them],
so the 'else' can be removed.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 22873 W: 4923 L: 4804 D: 13146
http://tests.stockfishchess.org/tests/view/5be9a8270ebc595e0ae33c7e

No functional change
2018-11-19 10:16:07 +01:00
Nikolay Kostov 4350a66ffa Update a comment in the evaluate.cpp file to reflect recent change
No functional change
2018-11-19 10:10:47 +01:00
SFisGOD 8a9c298dee Rook PSQT Tuned
Failed STC (Yellow )
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 56302 W: 12007 L: 11953 D: 32342
http://tests.stockfishchess.org/tests/view/5be69d210ebc595e0ae3185b

Passed 1st LTC (Green)
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 8745 W: 1480 L: 1301 D: 5964
http://tests.stockfishchess.org/tests/view/5be682960ebc595e0ae31818

Failed 2nd LTC (Red)
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19398 W: 3040 L: 3133 D: 13225
http://tests.stockfishchess.org/tests/view/5be69b840ebc595e0ae31856

Passed 3rd LTC (Green)
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 107516 W: 17342 L: 16858 D: 73316
http://tests.stockfishchess.org/tests/view/5bea879a0ebc595e0ae34d80

Closes https://github.com/official-stockfish/Stockfish/pull/1809

How to continue from there?

The values in the rook table now look a bit strange for a human eye
and are hard to explain, maybe it would be nice to simplify them
by hand and see if we can pass another (clean) double green with a
more regular array.

Bench: 3188070
2018-11-19 10:02:31 +01:00
Vizvezdenec 2a7213f720 Change default contempt from 21 to 24 centipawns
To top the rating lists and get more interesting middle play, it
is a good habit to set the default contempt to the highest value
that does not regress against contempt=0. We recently decreased
PawnValueEg it is logical that to raise a little bit the default
higher contempt because of the following internal dependency in
line 334 of search.cpp :

````
int ct = int(Options["Contempt"]) * PawnValueEg / 100; // From centipawns
````

STC: contempt=24 passed non-regression vs contempt=0
http://tests.stockfishchess.org/tests/view/5bd6d7f80ebc595e0ae21e14

LTC: contempt=24 passed non-regression LTC vs contempt=0
http://tests.stockfishchess.org/tests/view/5bd6e0980ebc595e0ae21f07

On 2018-11-01, we also tested the effects of contempt=21 and contempt=24
against Stockfish 9, and the net result was neutral:

Contempt 21
ELO: 51.68 +-1.9 (95%) LOS: 100.0%
Total: 40000 W: 9487 L: 3581 D: 26932
http://tests.stockfishchess.org/tests/view/5bdb1a140ebc595e0ae2620a

Contempt 24
ELO: 52.21 +-2.0 (95%) LOS: 100.0%
Total: 40000 W: 9759 L: 3793 D: 26448
http://tests.stockfishchess.org/tests/view/5bdb1b680ebc595e0ae2620d

Bench: 3459874
2018-11-19 09:47:19 +01:00
Nooby b9f1c9bf3f Clear TableBase mappings in Search::clear()
This patch will make possible to free mapped TB files with "ucinewgame" command.

We wrote this patch specifically to address a problem that arose while
running Stockfish with 7-piece tablebases as a kibitzer at TCEC for
extended periods of time across multiple games. It was noted that after
some time, the NPS of the kibitzing Stockfish (which is usually 3x faster
than the Stockfish actually competing) would drop precipitously, eventually
falling to preposterously low numbers until restarted.

Their eval bot basically inputs FEN, go infinite, stop and loop, it probably
didn't do ucinewgame either. As time goes it gradually slowed down and OS
starts to use swap, this is not reasonable since the engine only uses 16GB
hash and the machine has 1TB physical RAM and does nothing else.

Author : noobpwnftw

Closes https://github.com/official-stockfish/Stockfish/pull/1790

No functional change.
2018-11-19 09:40:42 +01:00
protonspring 3cbb05b1b8 Replace the PassedDanger array by an equation
This equation seems to do as well as the current PassedDanger array.

Master values were: 3, 7, 11, 20
The new values given by the equation are: 3, 6, 11, 18

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 84301 W: 18155 L: 18156 D: 47990
http://tests.stockfishchess.org/tests/view/5bda03180ebc595e0ae2518e

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7940 W: 1358 L: 1217 D: 5365
http://tests.stockfishchess.org/tests/view/5bdc69880ebc595e0ae27d28

We stopped a LTC run after 70000 games:
LLR: 0.74 (-2.94,2.94) [0.00,4.00]
Total: 70257 W: 11319 L: 11064 D: 47874
http://tests.stockfishchess.org/tests/view/5bdca8420ebc595e0ae281a9

Bench: 3913185
2018-11-12 20:33:07 +01:00
mstembera 68209c9121 Remove redundant king square parameter
We don't need to pass the king square as an explicit parameter to the functions
king_safety() and do_king_safety() since we already pass in the position.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 69686 W: 14894 L: 14866 D: 39926
http://tests.stockfishchess.org/tests/view/5be84ac20ebc595e0ae3283c

No functional change.
2018-11-12 19:45:05 +01:00
31m059 30a905c95d Simplify tropism. (#1807)
We calculate tropism as a sum of two factors. The first is the number of squares in our kingFlank and Camp that are attacked by the enemy; the second is number of these squares that are attacked twice. Prior to this commit, we excluded squares we defended with pawns from this second value, but this appears unnecessary. (Doubly-attacked squares near our king are still dangerous.) The removal of this exclusion is a possible small Elo gain at STC (estimated +1.59) and almost exactly neutral at LTC (estimated +0.04).

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20942 W: 4550 L: 4427 D: 11965
http://tests.stockfishchess.org/tests/view/5be4e0ae0ebc595e0ae308a0

LTC:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 56941 W: 9172 L: 9108 D: 38661
http://tests.stockfishchess.org/tests/view/5be4ec340ebc595e0ae30938

Bench: 3813986
2018-11-11 22:14:28 +01:00
Stephane Nicolet 05aa34e00e Update list of top CPU contributors
Contributors with >10,000 CPU hours as of November 4, 2018. Thank you!

No functional change
2018-11-08 17:09:44 +01:00
SFisGOD cd732c080b Pawn and Piece Values Tuned at LTC
Failed STC
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 27487 W: 5846 L: 5903 D: 15738
http://tests.stockfishchess.org/tests/view/5be1d3190ebc595e0ae2e5b8

Passed 1st LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 38503 W: 6270 L: 5999 D: 26234
http://tests.stockfishchess.org/tests/view/5be1f5ef0ebc595e0ae2e750

Passed 2nd LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 34016 W: 5584 L: 5326 D: 23106
http://tests.stockfishchess.org/tests/view/5be2a1970ebc595e0ae2f1b4

This pull request lead to an interesting discussion about testing
methodology for Stockfish:
https://github.com/official-stockfish/Stockfish/pull/1804

Bench: 3647775
2018-11-08 16:34:10 +01:00
Joost VandeVondele df50ea5dc6 fixup 2018-11-08 16:20:23 +01:00
Joost VandeVondele 9315ba60e6 Extension for king moves changing castling rights
passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8463 W: 1919 L: 1747 D: 4797
http://tests.stockfishchess.org/tests/view/5be15d510ebc595e0ae2dec6

passed LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 142590 W: 23263 L: 22587 D: 96740
http://tests.stockfishchess.org/tests/view/5be1667b0ebc595e0ae2df2d

Bench: 3607243
2018-11-08 16:20:23 +01:00
Fabian Fichter a6fe035977 Simplify mobility danger
Check sign only after adding mobility danger term.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 9090 W: 2001 L: 1856 D: 5233
http://tests.stockfishchess.org/tests/view/5bdc5ee10ebc595e0ae27bc2

LTC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 123466 W: 19766 L: 19805 D: 83895
http://tests.stockfishchess.org/tests/view/5bdc678e0ebc595e0ae27cf3

bench: 3630207
2018-11-04 21:30:35 +01:00
Stéphane Nicolet 8bb7a73708 Rook tweaks in evaluation
Some small changes in evaluation to try to convince Stockfish to centralize
her rooks more in middle game and avoid trapping them in the corners. Joint
work by SFisGOD and snicolet.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 99826 W: 21895 L: 21341 D: 56590
http://tests.stockfishchess.org/tests/view/5bdc3e280ebc595e0ae277df

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21467 W: 3541 L: 3322 D: 14604
http://tests.stockfishchess.org/tests/view/5bdc9ff30ebc595e0ae28119

Bench: 3631608
2018-11-02 22:08:26 +01:00
Joost VandeVondele 3f1eb85a1c Fix issues from using adjustedDepth too broadly
The recently committed Fail-High patch (081af90805)
had a number of changes beyond adjusting the depth of search on fail high, with
some undesirable side effects.

1) Decreasing depth on PV output, confusing GUIs and players alike as described in
   issue #1787. The depth printed is anyway a convention, let's consider adjustedDepth
   an implementation detail, and continue to print rootDepth. Depth, nodes, time and
   move quality all increase as we compute more. (fixing this output has no effect on
   play).

2) Fixes go depth output (now based on rootDepth again, no effect on play), also
   reported in issue #1787

3) The depth lastBestDepth is used to compute how long a move is stable, a new move
   found during fail-high is incorrectly considered stable if based on adjustedDepth
   instead of rootDepth (this changes time management). Reverting this passed STC
   and LTC:

   STC
   LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
   Total: 82982 W: 17810 L: 17808 D: 47364
   http://tests.stockfishchess.org/tests/view/5bd391a80ebc595e0ae1e993

   LTC
   LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
   Total: 109083 W: 17602 L: 17619 D: 73862
   http://tests.stockfishchess.org/tests/view/5bd40c820ebc595e0ae1f1fb

4) In the thread voting scheme, the rank of the fail-high thread is now artificially
   low, incorrectly since the quality of the move is much better than what adjustedDepth
   suggests (e.g. if it takes 10 iterations to find VALUE_KNOWN_WIN, it has very low
   depth). Further evidence comes from a test that showed that the move of highest
   depth is not better than that of the last PV (which is potentially of much lower
   adjustedDepth).

   I.e. this test http://tests.stockfishchess.org/tests/view/5bd37a120ebc595e0ae1e7c3
   failed SPRT[0, 5]:

   LLR: -2.95 (-2.94,2.94) [0.00,5.00]
   Total: 10609 W: 2266 L: 2345 D: 5998

   In a running 5+0.05 th 8 test (more than 10000 games) a positive Elo estimate is
   shown (strong enough for a [-3,1], possibly not [0,4]):

   http://tests.stockfishchess.org/tests/view/5bd421be0ebc595e0ae1f315
   LLR: -0.13 (-2.94,2.94) [0.00,4.00]
   Total: 13644 W: 2573 L: 2532 D: 8539
   Elo	1.04 [-2.52,4.61] / LOS 71%

Thus, restore old behavior as a bugfix, keeping the core of the fail-high patch
idea as resolving scheme. This is non-functional for bench, but changes searches
via time management and in the threaded case.

Bench: 3556672
2018-11-01 16:00:56 +01:00
SFisGOD 4a0db9ea3c Combo
Combo of two parameter tweaks and tuned values for Queen and ThreatByKing.

STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 20180 W: 4439 L: 4198 D: 11543
http://tests.stockfishchess.org/tests/view/5bd7b8250ebc595e0ae22e97

LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 86312 W: 14106 L: 13685 D: 58521
http://tests.stockfishchess.org/tests/view/5bd803560ebc595e0ae23213

This combo consists of the following:

Queen Value (tuned values)
Iter: 72056, A: 5000, alpha 0.602000, gamma 0.101000, clipping old, rounding deterministic
param: QueenValueMg, best: 2528.91, start: 2528.00
param: QueenValueEg, best: 2687.12, start: 2698.00

ThreatByKing (tuned values)
Green STC (50.8k games)
http://tests.stockfishchess.org/tests/view/5bd1d5a00ebc595e0ae1cbec
LTC (I stopped this test at 71.2k games. It's likely yellow.)
http://tests.stockfishchess.org/tests/view/5bd263e70ebc595e0ae1d77e

WeakUnopposedPawn (tweak) by xoto (https://github.com/xoto10)
Green STC (102.8k games)
http://tests.stockfishchess.org/tests/view/5bd306bb0ebc595e0ae1e146
Yellow LTC (90.8k games)
http://tests.stockfishchess.org/tests/view/5bd3ea660ebc595e0ae1f16b

aspiTune1 (tweak) by vondele (https://github.com/vondele)
Green STC (125.9k games)
http://tests.stockfishchess.org/tests/view/5bd2ae100ebc595e0ae1dab0
Yellow LTC (107.9k games)
http://tests.stockfishchess.org/tests/view/5bd3eb700ebc595e0ae1f16f

Thank you @31m059 (Mark Tenzer) for helping me! Also, thank you very much
for recognizing my efforts. I genuinely appreciate it.

Bench: 3556672
2018-11-01 15:39:19 +01:00
Vizvezdenec 7a61368971 Tweak of knight PSQT and mobility bonuses
STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 16906 W: 3745 L: 3516 D: 9645
http://tests.stockfishchess.org/tests/view/5bd306a40ebc595e0ae1e144

LTC
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 62779 W: 10249 L: 9901 D: 42629
http://tests.stockfishchess.org/tests/view/5bd3188f0ebc595e0ae1e296

Bench 3166402
2018-10-27 09:23:11 +02:00
Guenther Demetz 081af90805 On main thread: reduce depth after fail high
This helps resolving consecutive FH's during aspiration more efficiently

STC:
http://tests.stockfishchess.org/tests/view/5bc857920ebc592439f85765
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 4992 W: 1134 L: 980 D: 2878 Elo +10.72 

LTC:
http://tests.stockfishchess.org/tests/view/5bc868050ebc592439f857ef
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8123 W: 1363 L: 1210 D: 5550 Elo +6.54

No-Regression test with 8 threads, tc=15+0.15:
http://tests.stockfishchess.org/tests/view/5bc874ca0ebc592439f85938
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 24740 W: 3977 L: 3863 D: 16900 Elo +1.60

This was a cooperation between me and Michael Stembera:
-me recognizing SF having problems with resolving FH's efficiently at
high depths, thus starting some tests based on consecutive FH's.
-mstembera picking up the idea with first success at STC & LTC (so full
credits to him!)
-me suggesting how to resolve the issues pinpointed by S.G on PR #1768
and finally restricting the logic to the main thread so that it don't
regresses at multi-thread.

bench: 3314347
2018-10-25 23:08:06 +02:00
Peter Zsifkovits bc3b148d57 NUMA for 9 threads or more
Enable numa machinery only for STRICTLY MORE than 8 threads. Reason for this
change is that nowadays SMP tests are always done with 8 threads. That is a
problem for multi-socket Windows machines running on fishtest.

No functional change
2018-10-25 23:03:25 +02:00
Günther Demetz 9fff272209 Revert Pull Request #1771, see issue #1785 (#1786)
no functional change

bench: 4274207
2018-10-23 18:04:30 +02:00
mstembera 542a2b39ed Small simplification in castling rights
There is no need for a special struct with a static member
to generate castling rights.

No functional change.
2018-10-21 08:15:04 +02:00
ElbertoOne 738a6dfd4c Simplify check extensions
Remove the !moveCountPruning condition for check extensions, which seems not necessary.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 22238 W: 4835 L: 4715 D: 12688
http://tests.stockfishchess.org/tests/view/5bb3241a0ebc592439f6d2ac

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 36593 W: 5898 L: 5802 D: 24893
http://tests.stockfishchess.org/tests/view/5bb34c220ebc592439f6d5dc

Bench: 4274207
2018-10-14 20:40:57 +02:00
Joost VandeVondele 97d2cc9a9c Randomize draw eval
The patch adds a small random component (+-1) to VALUE_DRAW for the evaluation
of draw positions (mostly 3folds). This random component is not static, but
potentially different for each visit of the node (hence derived from the node
counter). The effect is that in positions with many 3fold draw lines, different
lines are followed at each iteration. This keeps the search much more dynamic,
as opposed to being locked to one particular 3fold.

An example of a position where master suffers from 3fold-blindness and this patch
solves quickly is the famous TCEC game 53:

FEN: 3r2k1/pr6/1p3q1p/5R2/3P3p/8/5RP1/3Q2K1 b - - 0 51

master doesn't see that this is a lost position (draw eval up to depth 50) as
Qf6-e6 d4-d5 (found by patch at depth 23) leads to a loss.

The 3fold-blindness is more important at longer TC, the patch was yellow STC and
LTC, but passed VLTC:

STC
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 46328 W: 10048 L: 9953 D: 26327
http://tests.stockfishchess.org/tests/view/5b9c0ca20ebc592cf275f7c7

LTC
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 54663 W: 8938 L: 8846 D: 36879
http://tests.stockfishchess.org/tests/view/5b9ca1610ebc592cf27601d3

VLTC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 31789 W: 4512 L: 4284 D: 22993
http://tests.stockfishchess.org/tests/view/5b9d1a670ebc592cf276076d

Credit to @crossbr for pointing to this problem repeatedly, and giving the hint
that many draw lines are typical in those situations.

Bench: 4756639
2018-10-14 20:33:52 +02:00
Guenther Demetz cb0111d3db Correctly track down pv even in fail-high case
Currently we update (track up) the pv even in the fail high case.
However most times in such cases the pv in the ply below remains unset
because there we have value == alpha and so finally we see truncated
pv's (=just one move) in fail high cases.
Of course tracking down these pv's (+sending them to the gui) comes at a
certian cost, but no-regression tests passed:

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16300 W: 3556 L: 3424 D: 9320
http://tests.stockfishchess.org/tests/view/5b9b73500ebc592cf275ea92

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 202411 W: 32734 L: 32897 D: 136780
http://tests.stockfishchess.org/tests/view/5b9baed10ebc592cf275ef6d

N.B.: Digging also into qsearch was tried in another version but seemed
not to pass the tests. This means that we don't always will get a pv
until the very tips.

No functional change
2018-10-14 20:19:46 +02:00
Miguel Lahoz 0370077c37 Simplify evaluation of blockers_for_king
Currently, we have two evaluation terms which account for pinned pieces.
One is for all pinned pieces in kingDanger computation and another for
just pinned pawns in ThreatByRank. We can increase the relevant bonus
for kingDanger calculation and do away with the ThreatByRank, which
seems to just add more complexity.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 113353 W: 24299 L: 24356 D: 64698
http://tests.stockfishchess.org/tests/view/5ba348c20ebc592cf2766e61

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 96458 W: 15514 L: 15511 D: 65433
http://tests.stockfishchess.org/tests/view/5ba398830ebc592cf2767563

At 100k games, I thought it struggles a bit, but some related [0,4]
tests attempting individual tweaks seem to fail:

I tried directly tweaking ThreatByRank:
http://tests.stockfishchess.org/tests/view/5ba3c6300ebc592cf276791c
http://tests.stockfishchess.org/tests/view/5ba3c6190ebc592cf2767917

@Vizveznedec was also recently trying to tweak the same coeffecients
for kingDanger calculation:
http://tests.stockfishchess.org/tests/view/5ba2c7320ebc592cf27664b2
http://tests.stockfishchess.org/tests/view/5ba2c8220ebc592cf27664b8
http://tests.stockfishchess.org/tests/view/5ba2c7880ebc592cf27664b4
http://tests.stockfishchess.org/tests/view/5ba2c7ce0ebc592cf27664b6

Bench: 4648095
2018-10-14 20:15:16 +02:00
Joost VandeVondele d615f15fce small ttCapture simplification.
ttCapture can be assigned to only once outside of the main loop. The patch seems
functional at higher depths (seems possible in the case of non-legal TTmoves that
are captures).

passed STC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 23189 W: 5098 L: 4980 D: 13111
http://tests.stockfishchess.org/tests/view/5bb3822c0ebc592439f6d966

passed LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 10336 W: 1665 L: 1529 D: 7142
http://tests.stockfishchess.org/tests/view/5bb39a190ebc592439f6db8a

unchanged bench: 4312846
2018-10-14 20:10:47 +02:00
31m059 489357d7b2 Combo
This PR is a combination of two unrelated [0, 4] patches that appeared promising
but not quite strong enough to pass on their own. The combination initially failed
STC with a positive score after a long run, and the subsequent speculative LTC test
passed.

* tweak_threatOnQueen4 :

Increase the middlegame components of ThreatByMinor[QUEEN]
and ThreatByRook[QUEEN] by 15 each. Bryan's (@crossbr) analysis of CCC Bonus Game 10
inspired several tests on penalizing a queen with limited safe mobility. While
attempting to implement this idea, I noticed that when I did not include the queen's
current square in the calculations, the Elo gains seemed to vanish--and only then did
I have the idea to revisit ThreatByMinor[QUEEN] and ThreatByRook[QUEEN], adding a
corresponding value to each. Without Bryan's work, this test would never have been
submitted. I would also like to recognize the efforts and contributions of @SFisGOD,
who also vigorously worked on this idea.

* Use pure static eval for null move pruning :

This idea was directly re-purposed from a promising test by Jerry Donald Watson
(@jerrydonaldwatson) in August. It was also independently developed and tested by
Stefan Geschwentner (@locutus2) previously.
Thank you all!

STC (failed yellow):
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 83913 W: 17986 L: 17825 D: 48102
http://tests.stockfishchess.org/tests/view/5bbc59300ebc592439f76aa5

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 137198 W: 22351 L: 21772 D: 93075
http://tests.stockfishchess.org/tests/view/5bbce35f0ebc592439f77639

Bench: 4312846
2018-10-14 20:02:31 +02:00
Eduardo Caceres 8141bdd179 Fix two typos in comments
Note by snicolet: I use this non-functional change patch
as a pretext to correct the wrong bench number I introduced
in the message of the previous commit.

Bench: 4059356
2018-09-27 21:39:36 +02:00
Joost VandeVondele bbf9daa175 Remove essentially unused code
this was added recently as part of a larger commit, but only changes eval of positions at MAX_PLY depth a little. Can be safely removed:

passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 7424 W: 1640 L: 1492 D: 4292
http://tests.stockfishchess.org/html/live_elo.html?5ba3bcbe0ebc592cf27677ff

passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 73554 W: 12028 L: 11990 D: 49536
http://tests.stockfishchess.org/html/live_elo.html?5ba397ee0ebc592cf2767556

unchanged Bench: 4248710
2018-09-27 21:28:38 +02:00
protonspring 13d06edb84 Two simplifications in passed pawns evaluation
These two simplifications appear to be affecting and/or offsetting each other.
Neither can be removed independently, but in combination they pass -3,1.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 36391 W: 7888 L: 7795 D: 20708
http://tests.stockfishchess.org/tests/view/5b9bce410ebc592cf275f1b2

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19513 W: 3237 L: 3114 D: 13162
http://tests.stockfishchess.org/tests/view/5b9c0edf0ebc592cf275f80e

Closes https://github.com/official-stockfish/Stockfish/pull/1769

bench 4059356
2018-09-27 21:18:18 +02:00
Rocky640 49b1591505 Pawn PSQT Tuned
Tested against master "Tweak opposite color bishops endgame scaling"
using values from a 100K SPSA with ck=10

Passed STC
http://tests.stockfishchess.org/tests/view/5ba7fe7a0ebc592cf276b971
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 27717 W: 6052 L: 5782 D: 15883

Passed LTC
http://tests.stockfishchess.org/tests/view/5ba815790ebc592cf276bb6b
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 17486 W: 2919 L: 2712 D: 11855

bench: 4441247
2018-09-27 20:58:40 +02:00
Joost VandeVondele 33b2f6398c Remove unneeded branch
Storing unconditionally the current generation and bound is equivalent to master.
Part of the condition was added as a speed optimization in #429.
Here the branch is fully eliminated.

passed STC single-threaded:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 73515 W: 16378 L: 16359 D: 40778
http://tests.stockfishchess.org/tests/view/5b2fc38c0ebc5902b2e57fd5

passed STC multi-threaded:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 63725 W: 12916 L: 12874 D: 37935
http://tests.stockfishchess.org/tests/view/5b307b8f0ebc5902b2e5895f

The multithreaded test was run after a plausible suggestion by @mstembera that the effect of this could be larger with many cores. The result seems to indicate this doesn't really matter on the 8core architecture abundantly available on fishtest.

No functional change
2018-09-27 20:48:11 +02:00
Vizvezdenec 0fa957cf66 Tweak opposite colord bishops endgame scaling.
Make scale factor dependant on asymmetry of pawn structure.

STC http://tests.stockfishchess.org/tests/view/5b92a2a80ebc592cf2753dd4
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 31490 W: 6870 L: 6587 D: 18033

LTC http://tests.stockfishchess.org/tests/view/5b92f8170ebc592cf2754438
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 54928 W: 8988 L: 8653 D: 37287

This patch shows that SF can use some more complicated endgame heuristics to evaluate endgames better from the distance.

Closes https://github.com/official-stockfish/Stockfish/pull/1767

Bench: 4248710
2018-09-10 12:22:44 +02:00
ElbertoOne 4bef7aa5cd Parameter tweaks in PSQT and NMP
This patch is a combinaison of two parameters tweaks patches which
have failed as strong yellows at LTC recently, by Alain Savard (Rocky640)
and Fabian Fichter (ianfab):
  http://tests.stockfishchess.org/tests/view/5b8a71e60ebc592cf2749b1d
  http://tests.stockfishchess.org/tests/view/5b81ce3b0ebc5902bdbb6585

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 57200 W: 12392 L: 12008 D: 32800
http://tests.stockfishchess.org/tests/view/5b8d0a5a0ebc592cf274c48f

And LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 37215 W: 6233 L: 5962 D: 25020
http://tests.stockfishchess.org/tests/view/5b8d56090ebc592cf274cb53

Closes https://github.com/official-stockfish/Stockfish/pull/1764

Bench: 4136116

---------------

How to continue from there?

The null move reduction formula in line 769 of search.cpp is quite convoluted
and full of mysterious magic constants at the moment, it would certainly be
nice to simplify it and/or gain more Elo from it:

```
Depth R = (  (823 + 67 * depth / ONE_PLY) / 256
           + std::min(int(eval - beta) / 200, 3)) * ONE_PLY;
```
2018-09-04 10:43:02 +02:00
Stéphane Nicolet 767c4ad1fc Update list of authors
And also fix some spaces and formatting oddities in the code.

No functional change
2018-09-03 22:11:30 +02:00
Stéphane Nicolet 2bfaf45455 Re-introduce "keep pawns on both flanks"
Re-introduce the "keep pawns on both flanks" idea.

STC yellow:
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 93279 W: 20175 L: 19853 D: 53251
http://tests.stockfishchess.org/tests/view/5b8a00370ebc592cf274916a

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 11440 W: 1960 L: 1792 D: 7688
http://tests.stockfishchess.org/tests/view/5b8a329f0ebc592cf2749615

Closes https://github.com/official-stockfish/Stockfish/pull/1761

Bench: 4609645
2018-09-01 11:30:38 +02:00
Rocky640 f923dc0fe5 Long Diagonal Tweaks
a) Reduce PSQT values along the long diagonals on non-central squares
and increase the LongDiagonal bonus accordingly. The effect is to penalise
bishops on the long diagonal which can not "see" the 2 central squares.
The "good" bishops still have more or less the same bonus as current master.

b) For a bishop on a central square, because of the "| s" term in the code,
the LongDiagonalBonus was always given. So while being there, remove the "| s"
and compensate the central Bishop PSQT accordingly.

Passed STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 44498 W: 9658 L: 9323 D: 25517
http://tests.stockfishchess.org/tests/view/5b8992770ebc592cf2748942

Passed LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 63092 W: 10324 L: 9975 D: 42793
http://tests.stockfishchess.org/tests/view/5b89a17a0ebc592cf2748b59

Closes https://github.com/official-stockfish/Stockfish/pull/1760

bench: 4693901
2018-09-01 04:33:17 +02:00
protonspring e846a9306d Remove PawnsOnBothFlanks
It looks like PawnsOnBothFlanks can be removed from initiative().
A barrage of tests seem to confirm that the adjustment to -110
does not gain elo to offset any potential loss by removing
PawnsOnBothFlanks.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 22014 W: 4760 L: 4639 D: 12615
http://tests.stockfishchess.org/tests/view/5b7f50cc0ebc5902bdbb3a3e

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 40561 W: 6667 L: 6577 D: 27317
http://tests.stockfishchess.org/tests/view/5b801f9f0ebc5902bdbb4467

The barrage of 0,4 tests on the -136 value are in my ps_tunetests branch.
http://tests.stockfishchess.org/tests/user/protonspring

Closes https://github.com/official-stockfish/Stockfish/pull/1751

Bench: 4413173

-------------

How to continue from there?

The fact that endgames with all the pawns on only one flank are
drawish is a well-known chess idea, so it seems quite strange that
this can be removed so easily without losing Elo.

In the past there had been attempts to improve on PawnsOnBothFlanks
with similar concepts (for instance using the pawn span value), but
the tests were at best neutral. Maybe Stockfish is now mature enough
that these refined ideas would work to replace PawnsOnBothFlanks?
2018-08-29 02:49:10 +02:00
MJZ1977 10bb2e6cdb Fix bug with "excludedMove" for probcut
Bugfix: "excludedMove" has to be skipped in the probcut loop too.
If it is not skipped, the probcut can exit quickly with a wrong return
value corresponding to the excluded move. See the following forum
thread for a discussion:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/GGithf_VwSU

STC :
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17130 W: 3747 L: 3617 D: 9766
http://tests.stockfishchess.org/tests/view/5b8460c40ebc5902bdbb999a

LTC :
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 12387 W: 2064 L: 1930 D: 8393
http://tests.stockfishchess.org/tests/view/5b8466f90ebc5902bdbb9a21

To go further : it can be perhaps useful to tune the singular extension
search parameters.

Closes https://github.com/official-stockfish/Stockfish/pull/1754

Bench: 4308541
2018-08-29 02:28:09 +02:00
Steinar H. Gunderson 166bf90e41 Shrink the hash table of tablebases back to 4096 entries
There is no need to make this as large as 65536 just for the sake of the
single 7-man tablebase that happens to have the key 0xf9247fff. Idea for the
fix by Ronald de Man, who suggested simply to allow more buckets past the end.

We also implement Robin Hood hashing for the hash table, which takes the worst
-case search for full 7-man tablebases down from 68 to 11 probes (Also takes
the average probe length from 2.06 to 2.05). For a table with 8K entries, the
corresponding numbers would be worst-case from 9 to 4, with average from 1.30
to 1.29.

https://github.com/official-stockfish/Stockfish/pull/1747

No functional change
2018-08-29 02:00:20 +02:00
Ondrej Mosnacek 4aa091cf44 Refactor pure static eval code
This commit tries to make the new pure static eval code more readable by
splitting up the nested assignments into separate lines and making a few
more cosmetic tweaks.

No functional change.
2018-08-29 01:24:45 +02:00
protonspring 8a4821923a make DistanceRing more consistent
This is a non-functional change. By pre-incrementing minKingPawnDistance
instead of post-incrementing, we can remove this -1.

This also makes DistanceRing more consistent with the rest of stockfish
since it now holds an actual "distance" instead of a less natural distance-1.

In current master, PseudoAttacks[KING][ksq] == DistanceRingBB[ksq][0]
With this patch, it will be PseudoAttacks[KING][ksq] == DistanceRingBB[ksq][1]
ie squares at distance 1 from the king. This is more natural use of distance.

The current array size DistanceRingBB[SQUARE_NB][8] is still OK with the new
definition, because maximum distance between two squares on a chess board is
seven (for example Kh1 and a8).

No functional change.
2018-08-29 01:07:38 +02:00
Vizvezdenec 6307fd08e6 Tweak stat bonus formula
Tweak stat bonus formula on top of latest elo gain by @snicolet

STC
http://tests.stockfishchess.org/tests/view/5b830a810ebc5902bdbb7e9c
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 27797 W: 6113 L: 5842 D: 15842

LTC
http://tests.stockfishchess.org/tests/view/5b831f2c0ebc5902bdbb8038
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 13655 W: 2294 L: 2099 D: 9262

I think that more elo can be found in tweaks of this parameters so I plan
to further try some "hand-tuning", including increasing/decreasing ratio of
two constants and making bonus assimetric to 0. Thx to @AndyGrant for helping
with github and @jerrydonaldwatson for original idea.

Closes https://github.com/official-stockfish/Stockfish/pull/1748

Bench: 4172767
2018-08-29 00:53:31 +02:00
VoyagerOne 3ac3b68540 Don't modify Eval with search stats at ttHits
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 28344 W: 6148 L: 6040 D: 16156
http://tests.stockfishchess.org/tests/view/5b7d6b4e0ebc5902bdbb1914

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41084 W: 6769 L: 6680 D: 27635
http://tests.stockfishchess.org/tests/view/5b7d7f5b0ebc5902bdbb1b85

Bench: 4457440
2018-08-29 00:41:53 +02:00
Stefan Geschwentner 28543cddc6 Store only unchanged static evaluations in TT
A recent commit introduced a decrease of the static evaluation of
an inner node dependent on the previous stat score, which finally
was also stored in the transposition table. Now only the unchanged
static evaluation are stored there.

Remark:
For the case that a static evaluation can be retrieved from the
transposition table the value is now used unchanged. Another test
which also applies the modification in this case failed:
http://tests.stockfishchess.org/tests/view/5b7af6df0ebc5902bdbae2f6

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 6707 W: 1547 L: 1383 D: 3777
http://tests.stockfishchess.org/tests/view/5b7a92df0ebc5902bdbadcf3

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 36203 W: 6046 L: 5781 D: 24376
http://tests.stockfishchess.org/tests/view/5b7abaa10ebc5902bdbadfa9

Closes https://github.com/official-stockfish/Stockfish/pull/1742

Bench: 4457440
2018-08-20 21:52:29 +02:00
Stéphane Nicolet f3b8a69919 Use an affine formula to mix stats and eval
Follow-up for the previous patch: we use an affine formula to mix stats
and evaluation in search. The idea is to give a bonus if the previous
move of the opponent was historically bad, and a malus if the previous
move of the opponent was historically good.

More precisely, if x is the stat score of the previous move by the opponent,
we implement the following formulas to tweak the evaluation at an internal
node of the tree for our pruning decisions at this node:

if x = 0, use v' = eval(P)
if x > 0, use v' = eval(P) - 5 - x/1024
if x < 0, use v' = eval(P) + 5 - x/1024

For reference, the previous master had this simpler rule:

if x > 0, use v' = eval(P) - 10
if x <= 0, use v' = eval(P)

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 29322 W: 6359 L: 6088 D: 16875
http://tests.stockfishchess.org/tests/view/5b76a5980ebc5902bdba957f

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 30893 W: 5154 L: 4910 D: 20829
http://tests.stockfishchess.org/tests/view/5b76ca6d0ebc5902bdba9914

Closes https://github.com/official-stockfish/Stockfish/pull/1740

Bench: 4592766
2018-08-18 01:23:36 +02:00
VoyagerOne 96c3a1f2ec Mix search stats with evaluation
Mix search stats with evaluation: if the opponent's move has a good historyStat,
then decrease the evaluation of the internal node a bit for the pruning decisions
during search.

STC;
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 72083 W: 15683 L: 15203 D: 41197
http://tests.stockfishchess.org/tests/view/5b74c3ea0ebc5902bdba7d41

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 29104 W: 4867 L: 4630 D: 19607
http://tests.stockfishchess.org/tests/view/5b7565000ebc5902bdba851b

Closes https://github.com/official-stockfish/Stockfish/pull/1738

Bench: 4514101

-----------

How to continue from there?

• the use of the previous stat score can probably be simplified in lines 587 and 716
• we could try to use a continuous bonus based on the previous stat score, instead
  of just a fixed offset of -10 when the opponent previous move was good.

----------

Comments by Stefan Geschwentner:

Interesting idea. Because only the eval in search is tweak this should only
influence the eval and static eval used at inner nodes, and not on the return
search value (which comes in the end from quiescence search), except through
saving in TT followed by a TT cutoff.

So essentialy this effects diverse pruning/reduction parts -- eval and static
eval  are lowered for good opponent moves:

• tt cutoff (ttValue)
• improving (static eval)
• more razoring (eval)
• less futility pruning (eval)
• less null move pruning (eval + static eval) (but with little more depth)
• more probcut (static eval)
• more move futility pruning (static eval)
2018-08-17 11:40:29 +02:00
protonspring d0f09de2d2 Simplify king file dependancy in evaluate_shelter()
Remove the special value we used for the file of the king in the
evaluate_shelter() function, and compensate by tweaking some of
the ShelterStrength[] array values.

STC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 17069 W: 3782 L: 3652 D: 9635
http://tests.stockfishchess.org/tests/view/5b75eb0d0ebc5902bdba8f3d

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 42639 W: 6973 L: 6887 D: 28779
http://tests.stockfishchess.org/tests/view/5b75fd7f0ebc5902bdba906b

Closes https://github.com/official-stockfish/Stockfish/pull/1739

Bench: 4639508
2018-08-17 10:21:20 +02:00
Stéphane Nicolet 881cab2525 Double weight of capture history
We double in this patch the weight of the capture history table in the
local scoring of captures for move ordering.

The capture history table is indexed by the triplet (capturing piece,
capture square, captured piece) and gets information like "it seems to
have been historically good in that part of the search tree to capture
a pawn with a rook on g3, even if it seems to lose material", and affect
the normaly pure « Most Valuable Victim » ordering of captures.

Finished yellow at STC after 228842 games (posting a +1.36 Elo gain):
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 228842 W: 50894 L: 50152 D: 127796
http://tests.stockfishchess.org/tests/view/5b714bb00ebc5902bdba332d

Passed LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 43251 W: 7425 L: 7131 D: 28695
http://tests.stockfishchess.org/tests/view/5b71c7d40ebc5902bdba3e51

Thanks to user Vizvezdenec for running the LTC test.

Closes https://github.com/official-stockfish/Stockfish/pull/1736

Bench: 4272361
2018-08-14 10:12:31 +02:00
Alain SAVARD 4d22d3e52d Remove pawncount array in imbalance
This is a natural follow up to last commit where values on the
QuadraticOurs diagonal and some piece value deltas were changed.
@Stefano80 tried to simplify the newly introduced pawncount array
using QuadraticOurs[1][1] =52 and a -30 adjustment on pawn values

His STC [-3,1] was green
http://tests.stockfishchess.org/tests/view/5b707f5b0ebc5902bdba2745
but not his LTC[-3,1]
http://tests.stockfishchess.org/tests/view/5b7095700ebc5902bdba2a49

So I started a 80000 30+0.3 SPSA on the QuadraticOurs diagonal and
on the piece values using @Stefano80 start values.

SPSA gave the new values QuadraticOurs[1][1] =38 and a -33 on pawn
values (the other changes on QuadraticOurs were kept, but were not
ignificant according to this test
http://tests.stockfishchess.org/tests/view/5b710ccb0ebc5902bdba2f27)

STC
http://tests.stockfishchess.org/tests/view/5b710b220ebc5902bdba2f19
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 50902 W: 11214 L: 11150 D: 28538

LTC
http://tests.stockfishchess.org/tests/view/5b7124ef0ebc5902bdba3106
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 34271 W: 5852 L: 5753 D: 22666

Closes https://github.com/official-stockfish/Stockfish/pull/1735

bench: 4738555
2018-08-14 08:36:27 +02:00
GuardianRM 41cc4eb953 Non-linear bonus for pawn count
This patch introduces a non-linear bonus for pawns, along with some
(linear) corrections for the other pieces types.

The original values were obtained by a massive non-linear tuning of both
pawns and other pieces by GuardianRM, while Alain Savard and Chris Cain
later simplified the patch by observing that, apart from the pawn case, the
tuned corrections were in fact almost affine and could be incorporated in
our current code base via the piece values in types.h (offset) and the diagonal
of the quadratic matrix (slope). See discussion in PR#1725 :
https://github.com/official-stockfish/Stockfish/pull/1725

STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 42948 W: 9662 L: 9317 D: 23969
http://tests.stockfishchess.org/tests/view/5b6ff6e60ebc5902bdba1d87

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 19683 W: 3409 L: 3206 D: 13068
http://tests.stockfishchess.org/tests/view/5b702dbd0ebc5902bdba216b

How to continue from there?
- Maybe the non-linearity for the pawn value could be somewhat tempered
  again and a simpler linear correction for pawns would work?

Closes https://github.com/official-stockfish/Stockfish/pull/1734

Bench: 4681496
2018-08-12 18:40:11 +02:00
Stefano Cardanobile b5581b7779 Combo of several promising parameter tweaks
Combo of several tuning patches which finished yellow at LTC.

[STC](http://tests.stockfishchess.org/tests/view/5b6ead340ebc5902bdba14ce)
LR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 10668 W: 2445 L: 2239 D: 5984
Elo: 6.25 [1.76,10.69] (95%)

[LTC](http://tests.stockfishchess.org/tests/view/5b6eb50e0ebc5902bdba151f)
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 23761 W: 4155 L: 3923 D: 15683
Elo: 3.02 [0.29,5.67] (95%)

Original patches:
- [Piece values](http://tests.stockfishchess.org/tests/view/5b6d2cc00ebc5902bdba02d5) by Stefano Cardanobile
- [Stat bonus](http://tests.stockfishchess.org/tests/view/5b6adbc90ebc5902bdb9da73) by Stefan Geschwentner
- [Rook on pawn](http://tests.stockfishchess.org/tests/view/5b62a95b0ebc5902bdb961c0) by Mark Tenzer
- [Hanging bonus](http://tests.stockfishchess.org/tests/view/5b5d2fa00ebc5902bdb90855) by Ivan Ilvec
- [ss tweak](http://tests.stockfishchess.org/tests/view/5b58b7240ebc5902bdb89025) by miguel-l

Bench: 4694813
2018-08-12 10:09:30 +02:00
Jerry Donald Watson 348cd5ed74 Simple razoring: depth 1 only, no distinction between PV / NonPV
We simplify the razoring logic by applying it to all nodes at depth 1 only.
An added advantage is that only one razor margin is needed now, and we treat
PV and Non-PV nodes in the same manner.

How to continue?
- There may be some conditions in which depth 2 razoring is beneficial.
- We can see whether the razor margin can be tuned, perhaps even with a
  different value for PV nodes.
- Perhaps we can unify the treatment of PV and Non-PV nodes in other parts
  of the search as well.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 5474 W: 1281 L: 1127 D: 3066
http://tests.stockfishchess.org/tests/view/5b6de3b20ebc5902bdba0d1e

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 62670 W: 10749 L: 10697 D: 41224
http://tests.stockfishchess.org/tests/view/5b6dee340ebc5902bdba0eb0

In addition, we ran a fixed LTC test against a similar patch which also
passed SPRT [-3, 1]:

ELO: 0.23 +-2.1 (95%) LOS: 58.6%
Total: 36412 W: 6168 L: 6144 D: 24100
http://tests.stockfishchess.org/tests/view/5b6e83940ebc5902bdba1485

We are opting for this patch as the more logical and simple of the two,
and it appears to be no less strong. Thanks in particular to @DU-jdto
for input into this patch.

Bench: 4476945
2018-08-12 09:54:16 +02:00
Miguel Lahoz f1088c9822 Remove Condition For Passed Pawns
Currently, we do not consider pawns passed if there is another pawn of
the same color in front of them. It appears that this condition is not
necessary. The idea is that the doubled pawns are likely to be weak and
one of them will be likely captured anyway. On the other hand, if we do
somehow manage to promote a pawn, then the pawn behind it becomes passed
as well. In any case, the end result is we end up with an extra
potentially passed pawn. The current evaluation for passed pawns already
handles this case by also scaling down this effect.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 28291 W: 6287 L: 6178 D: 15826
http://tests.stockfishchess.org/tests/view/5b6c4b960ebc5902bdb9f256

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 30717 W: 5256 L: 5151 D: 20310
http://tests.stockfishchess.org/tests/view/5b6c82980ebc5902bdb9f863

Bench: 4938285
2018-08-10 06:16:29 +02:00
Stefan Geschwentner 198418ee67 LMR simplification
Unify the "quiet" and "non-quiet" reduction rules for use at any kind of moves.
The idea behind it was that both rules reduce at similiar cases in master:
one directly for late previous moves and the other indirectly by using a
bad stat score which is used for most move sorting and so approximates the
late move condition.

For captures/promotions the old rule was triggered in 25% but the new
rule only for 3% of all cases (so now more reductions are done, whereas
for quiet moves reductions keep the same level).

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 162327 W: 35976 L: 36134 D: 90217
http://tests.stockfishchess.org/tests/view/5b6a9a430ebc5902bdb9d5c1

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 29570 W: 5083 L: 4976 D: 19511
http://tests.stockfishchess.org/tests/view/5b6bc5d00ebc5902bdb9e9d6

Bench: 4526980
2018-08-09 14:45:35 +02:00
Stefano Cardanobile bd4d2b0576 First check threshold in space evaluation
Currently, we first calculate some bitboards at the top of Evaluation::space()
and then check whether we actually need them. Invert the ordering. Of course this
does not make a difference in current master because the constexpr bitboard
calculations are in fact done at compile time by any decent compiler, but I find
my version a bit healthier since it will always meet or exceed current implementation
even if we eventually change the spaceMask to something not contsexpr.

No functional change.
2018-08-08 17:58:41 +02:00
FauziAkram c569cf263d King Psqt Tuning
After a session of tuning for King Psqt I got some new values, which was later
tweaked manually by me Fauzi, to result in an Elo-gain patch which seems to scale
pretty well:

STC: LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 100653 W: 22550 L: 22314 D: 55789 [Yellow patch]

LTC: LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 147079 W: 25584 L: 24947 D: 96548 [Green Patch]

Bench: 4669050
2018-08-08 17:49:16 +02:00
Stefano Cardanobile d96c1c32a2 Introduce voting system for best move selection
Introduce voting system for best move selction in multi-threads mode.
Joint work with Stefan Geschwentner, based on ideas introduced by
Michael Stembera.

Moves are upvoted by every thread using the margin to the minimum score
across threads and the completed depth.

First thread voting for the winner move is selected as best thread.

Passed STC, LTC. A further LTC test with only 4 threads failed with positive
score. A LTC with 31 threads was stopped with LLR 0.77 after 25k games to
avoid use of excessive resources (equivalent to 1.5M STC games).

Similar ideas were proposed by Michael Stembera 2 years ago #507, #508.
This implementation seems simpler and more understandable, the results
slightly more promising.

Further possible work:

1) Tweak of the formula using for assigning votes.
2) Use a different baseline for the score dependent part: maximum score
or winning probability could make more sense.
3) Assign votes in `Thread::Search` as iterations are completed and use
voting results to stop search.
4) Select best thread as the threads voting for best move with the highest
completed depth or, alternatively, vote on PV moves.

Link to SPRT tests

[stopped LTC, 31 threads 20+0.02](http://tests.stockfishchess.org/tests/view/5b61dc090ebc5902bdb95192)
LLR: 0.77 (-2.94,2.94) [0.00,5.00]
Total: 25602 W: 3977 L: 3850 D: 17775
Elo: 1.70 [-0.68,4.07] (95%)

[passed LTC, 8 threads 20+0.02](http://tests.stockfishchess.org/tests/view/5b5df5180ebc5902bdb9162d)
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 44478 W: 7602 L: 7300 D: 29576
Elo: 1.92 [-0.29,3.94] (95%)

[failed LTC, 4 threads 20+0.02](http://tests.stockfishchess.org/tests/view/5b5f39ef0ebc5902bdb92792)
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 29922 W: 5286 L: 5285 D: 19351
Elo: 0.48 [-1.98,3.10] (95%)

[passed STC, 4 threads 5+0.05](http://tests.stockfishchess.org/tests/view/5b5dbf0f0ebc5902bdb9131c)
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 9108 W: 2033 L: 1858 D: 5217
Elo: 6.11 [1.26,10.89] (95%)

No functional change (in simple threat mode)
2018-08-08 17:34:12 +02:00
Marco Costalba 571f54b176 Improve Stats definition
Use operator const T&() instead of operator T() to avoid possible
costly hidden copies of non-scalar nested types.

Currently StatsEntry has a single member T, so assuming
sizeof(StatsEntry) == sizeof(T) it happens to work, but it's
better to use the size of the proper entry type in std::fill.
Note that current code works because std::array items are ensured
to be allocated in contiguous memory and there is no padding among
nested arrays. The latter condition does not seem to be strictly
enforced by the standard, so be careful here.

Finally use address-of operator instead of get() to fully hide the
wrapper class StatsEntry at calling sites. For completness add
the arrow operator too and simplify the C++ code a bit more.

Same binary code as previous master under the Clang compiler.

No functional change.
2018-08-01 12:40:12 +02:00
Marco Costalba fae57273b2 Small tweaks to recent code changes
As a note, current 2 LMR conditions on stat score
could be simplified in a single line:

r -= ((ss->statScore >= 0) - ((ss-1)->statScore >= 0)) * ONE_PLY;

We keep them splitted in 2 "if" statements because are easier
to (immediately) read.

No functional change.
2018-07-31 11:56:10 +02:00
noobpwnftw 9afa03b80e 7-pieces Syzygy tablebase support
This is the first patch teaching Stockfish how to use the 7-pieces
Syzygy tablebase currently calculated by Bujun Guo (@noobpwnftw) and
Ronald de Man (@syzygy1). The 7-pieces database are so big that they
required a change in the internal format of the files (technically,
some DTZ values are 16 bits long, so this had to be stored as wide
integers in the Huffman tree).

Here are the estimated file size for the 7-pieces Syzygy files,
compared to the 151G of the 6-pieces Syzygy:

```
7.1T    ./7men_testing/4v3_pawnful (ongoing, 120 of 325 sets remaining)
2.4T    ./7men_testing/4v3_pawnless
2.3T    ./7men_testing/5v2_pawnful
660G    ./7men_testing/5v2_pawnless
117G    ./7men_testing/6v1_pawnful
87G     ./7men_testing/6v1_pawnless
```
Some pointers to download or recalculate the tables:

Location of original files, by Bujun Guo:
ftp://ftp.chessdb.cn/pub/syzygy/

Mirrors:
http://tablebase.sesse.net/ (partial)
http://tablebase.lichess.ovh/tables/standard/7/

Generator code:
https://github.com/syzygy1/tb/

Closes https://github.com/official-stockfish/Stockfish/pull/1707

Bench: 5591925 (No functional change if SyzygyTB is not used)

----------------------

Comment by Leonardo Ljubičić (@DragonMist)

This is an amazing achievement, generating and being able to use 7 men syzygy
on the fly. Thank you for your efforts @noobpwnftw !! Looking forward how this
will work in real life, and expecting some trade off between gaining perfect
play and slow disc Access, but once the disc speed and space is not a problem,
I expect 7 men to yield something like 30 elo at least.

-----------------------

Comment by Michael Byrne (@MichaelB7)

This definitely has a bright future. I turned off the 50 move rule (ala ICCF
new rules) for the following position:  `[d]8/8/1b6/8/4N2r/1k6/7B/R1K5 w - - 0 1`
This position is a 451 ply win for white (sans the 50 move rule, this position
was identified by the generator as the longest cursed win for white in KRBN v KRB).

Now Stockfish finds it instantly (as it should), nice work 👊👍 .
```
dep score	    nodes	    time
  7	+132.79 	4339    	0:00.00	Rb1+ Kc4 Nd6+ Kc5 Bg1+ Kxd6 Rxb6+ Kc7 Be3 Rh2 Bd4
  6	+132.79 	1652    	0:00.00	Rb1+ Kc4 Nd2+ Kd5 Rxb6 Rxh2 Nf3 Rf2
  5	+132.79 	589      	0:00.00	Rb1+ Kc4 Rxb6 Rxh2 Nf6 Rh1+ Kb2
  4	+132.79 	308      	0:00.00	Rb1+ Kc4 Nd6+ Kc3 Rxb6 Rxh2
  3	+132.79 	88        	0:00.00	Rb1+ Ka4 Nc3+ Ka5 Ra1+ Kb4 Ra4+ Kxc3 Rxh4
  2	+132.79 	54        	0:00.00	Rb1+ Ka4 Nc3+ Ka5 Ra1+ Kb4
  1	+132.7
```
2018-07-31 11:24:28 +02:00
Stéphane Nicolet ba2a2c34bb Introduce tropism measure in king danger
This patch adds the tropism measure as a new term in the king danger variable.
Since we then trasform this variable as a Score via a quadratic formula, the
main effect of the patch is the positive correlation of the tropism measure
with some checks and pins information already present in the king danger code.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 6805 W: 1597 L: 1431 D: 3777
http://tests.stockfishchess.org/tests/view/5b5df8d10ebc5902bdb91699

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 32872 W: 5782 L: 5523 D: 21567
http://tests.stockfishchess.org/tests/view/5b5e08d80ebc5902bdb917ee

How to continue from there?

• it may be possible to use CloseEnemies=S(7,0)
• we may want to try incorporating other strategic features in the quadratic
  king danger.

Closes https://github.com/official-stockfish/Stockfish/pull/1717

Bench: 5591925
2018-07-30 08:26:48 +02:00
Miguel Lahoz c08e05b494 Increase the mg->eg gradient for the PawnlessFlank malus
Just a change of value to S(19, 84). Also somewhat of a follow up
to the recent tweak in definition of KingFlank.

I tried a lot of other values before this, increasing and decreasing
but with little success, and before giving up I wanted to try tweaking
the middlegame and endgame values in the opposite directions. I guess
this is somewhat lucky.

STC:
LLR: 2.94 (-2.94,2.94) [0.00,4.00]
Total: 67685 W: 15399 L: 14963 D: 37323
http://tests.stockfishchess.org/tests/view/5b5b5ae80ebc5902bdb8e4f8

LTC: (Also thanks to Stephane Nicolet)
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 54635 W: 9505 L: 9172 D: 35958
http://tests.stockfishchess.org/tests/view/5b5b78f20ebc5902bdb8ece5

Closes https://github.com/official-stockfish/Stockfish/pull/1714

Bench: 4883742
2018-07-28 07:34:37 +02:00
VoyagerOne 6184d2b2ac Simplify cmh pruning
Simplify cmh pruning by removing PvNode exception

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 9935 W: 2330 L: 2184 D: 5421
http://tests.stockfishchess.org/tests/view/5b587dc00ebc5902bdb88424

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20635 W: 3585 L: 3464 D: 13586
http://tests.stockfishchess.org/tests/view/5b58910a0ebc5902bdb885b9

Closes https://github.com/official-stockfish/Stockfish/pull/1711

Bench: 4905530
2018-07-27 16:23:45 +02:00
Stéphane Nicolet 9ca014df49 Fix a compilation error for MSVC
The previous commit wouldn't compile on the Microsoft Virtual Studio C++ compiler. So use a more compatible style for the same idea (which we already use in numerous places of evaluate.cpp, for instance in line 563).

Under the Clang compiler, both versions generate exactly the same machine code (same md5 signatures for the two binaries).

No functional change.
2018-07-27 15:46:13 +02:00
Stéphane Nicolet e12fc10b5c Remove a popcount for HinderPassedPawn
Remove a popcount for HinderPassedPawn, and compensate by doubling
 the bonus from S(4,0) to to S(8,0).

Maybe it was pure luck, but we got the idea of this Elo gaining patch by
seing the simplification attempt by Mike Whiteley in pull request #1703.
This suggests that whenever we have a passed evaluation simplification,
we should consider the possibility that the master bonus has become
slightly out of tune with time, and we should try a few Elo gaining [0..4]
tests by hand-tuning the master bonus.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 19136 W: 4388 L: 4147 D: 10601
http://tests.stockfishchess.org/tests/view/5b59be6f0ebc5902bdb8ac06

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 99382 W: 17324 L: 16843 D: 65215
http://tests.stockfishchess.org/tests/view/5b59d2410ebc5902bdb8afa8

Closes https://github.com/official-stockfish/Stockfish/pull/1710

Bench: 4688817
2018-07-27 15:23:57 +02:00
Miguel Lahoz 313f403733 Tweak KingFlank when king is on edge files
This tweak excludes files D and E from the KingFlank bitboard when our
king is on the A or H files respectively. As far as I can tell, this
affects two things: the calculation for CloseEnemies and PawnlessFlank.
Aside from filtering out slightly less relevant attacks in the flank,
I suspect this helps with king prophylaxis, avoiding attacks and moving
towards the center when the pawns start to come off.

STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 56755 W: 12881 L: 12489 D: 31385
http://tests.stockfishchess.org/tests/view/5b58a94c0ebc5902bdb88c72

LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 130205 W: 22536 L: 21957 D: 85712
http://tests.stockfishchess.org/tests/view/5b58b7580ebc5902bdb89029

How to continue: Tweaking the two bonuses mentioned might give some
gain, although as far as I can tell, CloseEnemies is very sensitive to
even small changes.

Closes https://github.com/official-stockfish/Stockfish/pull/1705

Bench: 5026009
2018-07-27 10:38:20 +02:00
Jekaa c9f80660a6 Small reformat in evaluate threats (non functional)
When evaluating threat by safe pawn and pawn push the same expression is used.

STC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19444 W: 4540 L: 4309 D: 10595
http://tests.stockfishchess.org/tests/view/5b5a6e150ebc5902bdb8c5c0

Closes https://github.com/official-stockfish/Stockfish/pull/1709

No functional change.

--------------------

Comments by Stéphane Nicolet:

I don't measure any speed-up on my system, with two parallel benches at depth 22:

Total time (ms) : 74989
Nodes searched : 144830258
Nodes/second : 1931353
master

Total time (ms) : 75341
Nodes searched : 144830258
Nodes/second : 1922329
testedpatch

And anyway, like Stefan Geschwentner, I don't think that a 0.3% speed-up would
be enough to pass a [0..5] LTC test -- as a first approximation, we have this
rule of thumb that 1% speed-up gives about 1 Elo point.

However, considering the facts that the reformatting by itself is interesting,
that this is your first green test and that you played by the rules by running
the SPRT[0..5] test before opening the pull request, I will commit the change.
I will only take the liberty to change the occurrences of safe in lines 590 and
591 to b, to make the code more similar to lines 584 and 585.

So approved, and congrats :-)
2018-07-27 10:30:53 +02:00
ianfab d44701be4b Fix condition for error message of signature script
Use obtained bench instead of reference bench when checking for crash.

No functional change.
2018-07-27 10:16:33 +02:00
protonspring 2660a9145e Remove condition for pawn threats
It appears as though removing squares that are already attacked
by our pawns can be removed.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 51242 W: 11503 L: 11440 D: 28299
http://tests.stockfishchess.org/tests/view/5b58b5a40ebc5902bdb88f52

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35246 W: 6063 L: 5966 D: 23217
http://tests.stockfishchess.org/tests/view/5b58f8e20ebc5902bdb8959b

How to continue after this patch: there is now a slight semantic
overlap between the ThreatByPawnPush and the ThreatBySafePawn bonuses,
so hand-tuning either of these, or both at the same time, is natural.

Closes https://github.com/official-stockfish/Stockfish/pull/1702

Bench 4734881
2018-07-26 09:34:22 +02:00
Stefan Geschwentner a4eda3056e Rank threats on pinned pawns
Add for pinned pawns half of the standard rank based threat bonus.

STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 44010 W: 9987 L: 9635 D: 24388
http://tests.stockfishchess.org/tests/view/5b58aa780ebc5902bdb88c7a

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 29475 W: 5089 L: 4847 D: 19539
http://tests.stockfishchess.org/tests/view/5b58b56c0ebc5902bdb88f37

Closes https://github.com/official-stockfish/Stockfish/pull/1701

Bench: 4503866
2018-07-26 01:29:12 +02:00
Stéphane Nicolet ae98927885 Code clean-up
This patch implements some idea by Alain Savard and Mike Whiteley taken from the perpertual renaming/reformatting thread.

This is a pure code cleaning patch (so no change in functionality), but I use it as a pretext to correct the bogus bench number that I introduced in the previous commit.

Bench: 4413383
2018-07-25 18:31:02 +02:00
Stefan Geschwentner c4c2e08f0d Tweak stat bonus
Increase stat bonus by 1/32 and adjust the divisor of main and capture
history tables to 10692.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 28437 W: 6444 L: 6166 D: 15827
http://tests.stockfishchess.org/tests/view/5b579b4d0ebc5902bdb87139

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 111204 W: 19160 L: 18644 D: 73400
http://tests.stockfishchess.org/tests/view/5b57a7c60ebc5902bdb872d3

Closes https://github.com/official-stockfish/Stockfish/pull/1698

Bench: 4778882
2018-07-25 18:02:07 +02:00
VoyagerOne 6e36860554 CounterMove History Pruning Tweak
STC: (Yellow)
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 40124 W: 8817 L: 8751 D: 22556
http://tests.stockfishchess.org/tests/view/5b5690180ebc5902bdb85c8a

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 21599 W: 3811 L: 3599 D: 14189
http://tests.stockfishchess.org/tests/view/5b5757010ebc5902bdb86b1f

Closes https://github.com/official-stockfish/Stockfish/pull/1697

Bench:  4794161
2018-07-25 17:55:16 +02:00
Stefan Geschwentner bb56779cb6 Revert "Tweak reductions formula: 0.88 * depth + 0.12"
This patch reverts the recent commit called "Tweak reductions formula, etc."
The decisions for the revert decision were as follows:

1) The original commit called "Tweak reductions formula: 0.88 * depth + 0.12"
showed bad scaling at in a Very Long Time Control (VLTC) test:

VLTC (180+1.8):
LLR: -1.59 (-2.94,2.94) [0.00,5.00]
Total: 14968 W: 2247 L: 2257 D: 10464
http://tests.stockfishchess.org/tests/view/5b559ffa0ebc5902bdb84f36

2) So there was a suspicion that the original fast passing LTC test which lead
us to accept the patch may have been a statistical accident, so we organized
a match against the previous master at LTC to get an Elo estimate for the
patch:

LTC match:
ELO: -1.83 +-2.1 (95%) LOS: 4.3%
Total: 36018 W: 6018 L: 6208 D: 23792
http://tests.stockfishchess.org/tests/view/5b55f8110ebc5902bdb8526f

3) Based on these results, we ran a simplification test with [-3..1] bounds
for the revert at LTC:

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41501 W: 7107 L: 7020 D: 27374
http://tests.stockfishchess.org/tests/view/5b5738670ebc5902bdb86932

4) So we revert.

Bench: 4491691
2018-07-25 07:39:06 +02:00
double-beep 38471697b7 Slight decrease of overload value
Set overload value to S(13,6)

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 27606 W: 6371 L: 6094 D: 15141
http://tests.stockfishchess.org/tests/view/5b5455840ebc5902bdb82425

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 112435 W: 19442 L: 18921 D: 74072
http://tests.stockfishchess.org/tests/view/5b546d4a0ebc5902bdb82741

Closes https://github.com/official-stockfish/Stockfish/pull/1694

Bench: 4937000
2018-07-24 08:39:08 +02:00
Stefan Geschwentner 50287a55d3 Tweak reductions formula: 0.88 * depth + 0.12
Replace the depth part in the reduction formula for higher depths
with a slower growing linear function. So for depth > 3 less reductions
are used.

What we can try next:
- move the break point to even higher depths
- tweak the slope for lower and higher depth
- even possibly use a further higher depth threshold for a another
  slower growing function

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 25317 W: 5763 L: 5505 D: 14049
http://tests.stockfishchess.org/tests/view/5b54f9f70ebc5902bdb840ed

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 7451 W: 1320 L: 1167 D: 4964
http://tests.stockfishchess.org/tests/view/5b54feeb0ebc5902bdb84244

Closes https://github.com/official-stockfish/Stockfish/pull/1692

Bench: 4617359
2018-07-23 09:16:29 +02:00
Goodkov Vasiliy Aleksandrovich 0d5fe2f156 Simplify condition for ThreatByRook
Remove stronglyProtected Queen for ThreatByRook. Idea is that in the
current master the  SliderOnQueen bonus and the see_ge() function do
something similar as ThreatByRook for Queen, so this patch removes
some redundancy, in that sense.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21878 W: 4939 L: 4818 D: 12121
http://tests.stockfishchess.org/tests/view/5b53a83b0ebc5902bdb815d1

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35307 W: 5979 L: 5882 D: 23446
http://tests.stockfishchess.org/tests/view/5b53b60b0ebc5902bdb8174c

Close https://github.com/official-stockfish/Stockfish/pull/1690

Bench: 4834554
2018-07-23 00:03:05 +02:00
protonspring af1ddfd83b simplified forward ranks.
This is a non-functional simplification. We change replaces an 'OR'
and a lookup (rank_bb(ksq)) with a bitwise ~.  This is fewer operations
and is probably faster.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25441 W: 5689 L: 5575 D: 14177
http://tests.stockfishchess.org/tests/view/5b52d05a0ebc5902bdb8010e

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26904 W: 4664 L: 4553 D: 17687
http://tests.stockfishchess.org/tests/view/5b543df70ebc5902bdb8212d

No functional change.
2018-07-22 17:59:39 +02:00
Marco Costalba 4bd24da161 Slight tidy up in endgame machinery
No functional change.
2018-07-22 17:55:41 +02:00
Stefan Geschwentner 53c07c34bb Non functional LMR rewrite. 2018-07-22 17:53:31 +02:00
Alain SAVARD 0365b08601 Simplify the "overload" condition
This is a follow-up of the previous pull request (#1686) by Miguel.
We simplify the "Overload" bonus condition by re-using the "weak"
variable, which captures well the essence of the overload condition.
This may also be a small speed optimization because the weak variable
is in a register at this point of the code.

http://tests.stockfishchess.org/tests/view/5b527b440ebc5902bdb7f7db
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 10925 W: 2517 L: 2374 D: 6034

http://tests.stockfishchess.org/tests/view/5b527f930ebc5902bdb7f883
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 15569 W: 2697 L: 2568 D: 10304

Closes https://github.com/official-stockfish/Stockfish/pull/1687

Bench: 5010472
2018-07-21 07:05:50 +02:00
Miguel Lahoz 41bc0d5660 Remove connectivity.
There seems to be some strange interaction between Overload and Connectivity.
Overload encourages us to not have too many defended and attacked pieces,
as this may expose us to various tactics. This feels somewhat like it is in
conflict with Connectivity, where pieces are defended preemptively.

Here I take the "pick one or the other" approach and just remove connectivity,
while strengthening the effect of Overload to compensate. The reasoning is that
if we defend our pieces preemptively, then it does get attacked, we want to do
something about it so we don't get penalized by Overload. On the other
hand, if it doesn't get attacked, then there's no need to defend it.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 27734 W: 6174 L: 6064 D: 15496
http://tests.stockfishchess.org/tests/view/5b5073bd0ebc5902bdb7ba5c

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 51606 W: 8897 L: 8827 D: 33882
http://tests.stockfishchess.org/tests/view/5b50aa900ebc5902bdb7bf29

Bench: 4658006
2018-07-21 06:56:48 +02:00
Stefan Geschwentner 12e79be910 Better check evasion move sorting
Use in addition the counter move history table for sorting quiet
check evasion moves in main and quiecence search. Also rename
"contHistory" to "continuationHistory" while there.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 73284 W: 16433 L: 15938 D: 40913
http://tests.stockfishchess.org/tests/view/5b4f526e0ebc5902bdb7a401

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 12135 W: 2171 L: 1997 D: 7967
http://tests.stockfishchess.org/tests/view/5b4fc0ef0ebc5902bdb7ae0e

Closes https://github.com/official-stockfish/Stockfish/pull/1685

Bench 4817583
2018-07-19 18:27:20 +02:00
Miguel Lahoz 3913726d1c Use single value for KingProtector.
After some recent big tuning session, the values for King Protector were
simplified to only be used on minor pieces. This patch tries to further
simplify by just using a single value, since current S(6,5) and S(5,6)
are close to each other. The value S(6,6) ended up passing, although
S(5,5) was also tried and failed STC.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14261 W: 3288 L: 3151 D: 7822
http://tests.stockfishchess.org/tests/view/5b4ccdf50ebc5902bdb77f65

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19606 W: 3396 L: 3273 D: 12937
http://tests.stockfishchess.org/tests/view/5b4ce4280ebc5902bdb7803b

Bench: 5448998
2018-07-18 08:44:45 +02:00
ElbertoOne 2ac35027d5 Simplify Overload condition
Extend the bonus for Overload to cases where our side
 has more than one attacker to a non pawn piece.
Based on an idea by Bryan in the forum. For instance,
 now black gets the overload bonus in this position:
 8/5R1k/6pb/p6p/P1N4P/1Pp5/2K3P1/2N4r b - - 6 46
 because two black pieces are attacking the knight on c1
 that is defended only by the king.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 57446 W: 12762 L: 12711 D: 31973
http://tests.stockfishchess.org/tests/view/5b4ca9970ebc5902bdb77a88

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 42113 W: 7295 L: 7209 D: 27609
http://tests.stockfishchess.org/tests/view/5b4ccea00ebc5902bdb77f69

Bench: 4667263
2018-07-18 08:37:13 +02:00
DU-jdto a05793517f Minor code style tweaks
No functional change.
2018-07-18 08:26:33 +02:00
ianfab ee0f5cd303 Minor whitespace formatting changes
No functional change.
2018-07-18 08:25:08 +02:00
Gian-Carlo Pascutto e0f317afaa Allow Position::init() to be called multiple times.
For the rationale to allow this, see commit
a66c73deef

This was broken when cuckoo hashing was added, and
subtly broke (for example) lichess' Android application,
thus illustrating the original judgement was sound.

No functional change.
2018-07-18 08:14:57 +02:00
protonspring a6fa6a9e92 Remove rank limitation for MinorBehindPawn
This is a functional simplification.  It seems like the rank restriction
for MinorBehindPawn can be removed.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 61195 W: 13404 L: 13360 D: 34431
http://tests.stockfishchess.org/tests/view/5b47e6f00ebc5978f4be3fc0

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 31701 W: 5367 L: 5264 D: 21070
http://tests.stockfishchess.org/tests/view/5b48a2cb0ebc5978f4be4769

Bench: 4938702
2018-07-14 08:27:33 +02:00
candirufish d2d4e85f25 Tuned Values after 2 million spsa games
Various king and pawn eval values tuned after 2 million games. Rounding
slightly adjusted.

LTC: http://tests.stockfishchess.org/tests/view/5b477a260ebc5978f4be3ed4
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 32783 W: 5852 L: 5588 D: 21343

STC: http://tests.stockfishchess.org/tests/view/5b472d420ebc5978f4be3e4d
LLR: 3.23 (-2.94,2.94) [0.00,4.00]
Total: 44380 W: 10201 L: 9841 D: 24338

I think I reached the limit of the fishtest framework. It frequently
crashed at 2 million games already. The small values also moved a lot
throughout the entire tuning session though with smaller margin. The
passed danger and close enemies values seems the most sensitive (changing
close enemies alone to 6 failed before but now it passes), whether or not
they are close to optimal I don't know, but it seems some parameters are
also correlated to others.

Closes https://github.com/official-stockfish/Stockfish/pull/1670

bench: 5103722
2018-07-14 08:13:15 +02:00
Joost VandeVondele d2752fdc15 Remove offset in thread redistribution scheme.
doesn't have a benefit.

passed STC (8 threads):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 19574 W: 4028 L: 3904 D: 11642
http://tests.stockfishchess.org/tests/view/5b3e48950ebc5902b9fff080

passed LTC (8 threads):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21293 W: 3626 L: 3506 D: 14161
http://tests.stockfishchess.org/tests/view/5b3eefd60ebc5902b9fffa81

Closes https://github.com/official-stockfish/Stockfish/pull/1667

No functional change single threaded.
2018-07-07 10:42:06 +02:00
31m059 0f48095759 Simplify ThreatByKing to be a single Score.
In the current master, ThreatByKing is an array of two Scores, one for
when we have a single attack and one for when we have many. The latter
case is very rarely called during bench and was recently given a strange
negative value during a tuning run, as pointed out by @candirufish on
commit efd4ca2.  Here, we simplify away this second case entirely, and
increase the remaining ThreatByKing to compensate.

Although I derived the parameter tweak independently, with the goal of
preserving the same average bonus, I later noticed that a very similar
Score had already been derived by an ongoing SPSA tuning session.
I therefore recognize @candirufish for first discovering these values.

I would also like to thank @Rocky640 for valuable feedback that pointed
me in the direction of ThreatByKing.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 7677 W: 1772 L: 1623 D: 4282
http://tests.stockfishchess.org/tests/view/5b3db0320ebc5902b9ffe97a

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 108031 W: 18329 L: 18350 D: 71352
http://tests.stockfishchess.org/tests/view/5b3dbf4b0ebc5902b9ffe9db

Closes https://github.com/official-stockfish/Stockfish/pull/1666

Bench: 4678861
2018-07-06 01:34:01 +02:00
VoyagerOne 3279655f12 Capture Stat Tweak
Penalize capture moves that fail to create a cutoff even at quiet move cutoff.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19004 W: 4284 L: 4059 D: 10661
http://tests.stockfishchess.org/tests/view/5b3a7d4d0ebc5902b9ffb6ea

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 23100 W: 3981 L: 3765 D: 15354
http://tests.stockfishchess.org/tests/view/5b3aa4550ebc5902b9ffb8cf

Closes https://github.com/official-stockfish/Stockfish/pull/1664

Bench: 5232010
2018-07-04 01:12:16 +02:00
Joost VandeVondele e7cfa5d020 Simplify saving a TT entry.
Avoid passing TT.generation() to TTEntry::save() at every call,
moving the implementation of TTEntry::save from tt.h to tt.cpp.

tested for no regression:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 53787 W: 11948 L: 11890 D: 29949
http://tests.stockfishchess.org/tests/view/5b2ff37f0ebc5902b2e582fe

Closes https://github.com/official-stockfish/Stockfish/pull/1662

No functional change.
2018-07-04 00:59:15 +02:00
Joost VandeVondele 8c4f0ffa1d Reduce scope of variables
Small cleanup TranspositionTable:clear().

Closes https://github.com/official-stockfish/Stockfish/pull/1659

No functional change.
2018-07-04 00:51:10 +02:00
Ondrej Mosnáček a781535168 Move PSQ score to Position
This patch simplifies Position::do_move() by moving the PSQ score from
StateInfo to Position and updating it inside the put/remove/move_piece
functions.

The downside is that there is now slightly more computation done in
Position::undo_move(), but the fishtest results are Elo neutral.

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 78820 W: 15775 L: 15760 D: 47285
http://tests.stockfishchess.org/tests/view/5b1cd1d00ebc5902ab9c64ab

Passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32966 W: 5716 L: 5615 D: 21635
http://tests.stockfishchess.org/tests/view/5b31e1230ebc5902b2e5a833

Closes https://github.com/official-stockfish/Stockfish/pull/1647

No functional change.
2018-06-27 11:42:25 +02:00
protonspring af6072c8b7 Remove make_bitboard()
In current master, the function make_bitboard() does nothing apart from
helping initialize the SquareBB[] array. This seems like an unnecessary
abstraction layer.

The advantage of make_bitboard() is we can define a bitboard, in a simple
and general way, not only from a single square but also from a list of
squares. It is more elegant, faster and  readable than combining multiple
SquareBB explicitly, but the last complex use case in evaluation was
simplified away a few months ago.

If make_bitboard() becomes useful again to define complicated bitboards,
it will be easy enough to reintroduce it using this pull request as
an implementation reference.

No functional change.
2018-06-26 09:08:15 +02:00
joergoster 1e9397a2df Simplify KingProtector penalty.
Recent tuning results implied this penalty is more important for knights
and bishops, and almost negligible for rooks and queen.

Passed as simplification both
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20873 W: 4592 L: 4469 D: 11812
http://tests.stockfishchess.org/tests/view/5b2fb4d00ebc5902b2e57e84

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 46069 W: 7949 L: 7870 D: 30250
http://tests.stockfishchess.org/tests/view/5b2fcc4b0ebc5902b2e580c5

Closes https://github.com/official-stockfish/Stockfish/pull/1660

Bench: 5487679
2018-06-26 08:55:35 +02:00
Alain SAVARD f0a7bed6fb Simplify HinderPassedPawn bonus
Make sure each piece is not scored more than once as a passed pawn "hinderer",
by scoring only the blockers along the passed pawn path. Inspired by TCEC Game 29.

Passed STC as a simplification
http://tests.stockfishchess.org/tests/view/5b3016d00ebc5902b2e58552
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 75388 W: 16656 L: 16641 D: 42091

Passed LTC as a simplification
http://tests.stockfishchess.org/tests/view/5b302ed90ebc5902b2e587fc
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 49157 W: 8460 L: 8386 D: 32311

Current master was also counting the number of attacks along a passed pawn path,
which might be misleading:

a) a defender might be counted many times for the same pawn path. For example a
   White rook on a1 attacking a black pawn on a7 would score the bonus * 6 but
   would be probably better placed on a8

b) a defender might be counted on different pawn paths and might be overloaded. For
   example a Ke4 or Qe4 against pawns on d6  and f6 would score the bonus * 6.

Counting each blocker or attacker only once is more complicated, and does not help
either: http://tests.stockfishchess.org/tests/view/5b2ff1cb0ebc5902b2e582b2

After this small simplification, there might be ways to increase the HinderPassedPawn
penalty.

Closes https://github.com/official-stockfish/Stockfish/pull/1661

Bench: 4520519
2018-06-26 08:16:37 +02:00
candirufish efd4ca27c4 Another set of tuned values after one million games
Another set of tuned values, obtained by a long session of one million games.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 15810 W: 3687 L: 3458 D: 8665
http://tests.stockfishchess.org/tests/view/5b2d32f60ebc5902b2e55d9e

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 102118 W: 18146 L: 17651 D: 66321
http://tests.stockfishchess.org/tests/view/5b2d372c0ebc5902b2e55e0a

Closes https://github.com/official-stockfish/Stockfish/pull/1658

Bench: 4557946
2018-06-23 09:03:58 +02:00
Stéphane Nicolet 34321fcc2d Increase outflanking weight to 12
Give more incentive to king activity in the endgame by increasing the weight
of the "outflanking" variable from 8 to 12 in the function evaluate_initiative().

Finished yellow after 133102 games at STC:

LLR: -3.07 (-2.94,2.94) [0.00,4.00]
Total: 133102 W: 29535 L: 29179 D: 74388
http://tests.stockfishchess.org/tests/view/5b2b63fe0ebc5902b2e54475

Passed LTC:

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 28027 W: 4918 L: 4672 D: 18437
http://tests.stockfishchess.org/tests/view/5b2ba39e0ebc5902b2e54a64

Closes https://github.com/official-stockfish/Stockfish/pull/1657

Bench: 4721753
2018-06-22 01:29:25 +02:00
candirufish 062eb074c8 Tuned values after 505k games
Various king and pawn tuned eval values after 505k 60 sec 600 nodes time
SPSA games. Adjusted passed rank and file values to be symmetrical.

Passed LTC (after passed rank/file adjustment):
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 37906 W: 6953 L: 6668 D: 24285
http://tests.stockfishchess.org/tests/view/5b2790960ebc5902b8d17ba1

A previous, very similar version with raw tuned values passed STC and LTC:

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 39515 W: 9227 L: 8900 D: 21388
http://tests.stockfishchess.org/tests/view/5b277e3e0ebc5902b8d17ac9

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 14618 W: 2743 L: 2537 D: 9338
http://tests.stockfishchess.org/tests/view/5b2785020ebc5902b8d17b98

Closes https://github.com/official-stockfish/Stockfish/pull/1654

bench: 4777396
2018-06-21 10:23:05 +02:00
Michael An 06a853eb9c Fix GCC 8 cast warnings
Silences the following warnings when compiling with GCC 8.
The fix is to use an intermediate pointer to anonymous function:

```
misc.cpp: In function 'int WinProcGroup::get_group(size_t)':
misc.cpp:241:77: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun1_t' {aka 'bool (*)(_LOGICAL_PROCESSOR_RELATIONSHIP, _SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX*, long unsigned int*)'} [-Wcast-function-type]
   auto fun1 = (fun1_t)GetProcAddress(k32, "GetLogicalProcessorInformationEx");
                                                                             ^
misc.cpp: In function 'void WinProcGroup::bindThisThread(size_t)':
misc.cpp:309:71: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun2_t' {aka 'bool (*)(short unsigned int, _GROUP_AFFINITY*)'} [-Wcast-function-type]
   auto fun2 = (fun2_t)GetProcAddress(k32, "GetNumaNodeProcessorMaskEx");
                                                                       ^
misc.cpp:310:67: warning: cast between incompatible function types from 'FARPROC' {aka 'long long int (*)()'} to 'fun3_t' {aka 'bool (*)(void*, const _GROUP_AFFINITY*, _GROUP_AFFINITY*)'} [-Wcast-function-type]
   auto fun3 = (fun3_t)GetProcAddress(k32, "SetThreadGroupAffinity");
                                                                   ^
```

No functional change.
2018-06-21 09:51:31 +02:00
mstembera 745160572f Fix MSVC error
Compiling the current master with MSVC gives the following error:

```
search.cpp(956): error C2660: 'operator *': function does not take 1 arguments
types.h(303): note: see declaration of 'operator *'
```

This was introduced in commit:
https://github.com/official-stockfish/Stockfish/commit/88de112b84a5285c2afb3e075a05c2ab8ad3fd33

We use a suggestion by @vondele to fix the error, thanks!

No functional change.
2018-06-20 05:43:00 +02:00
Stefano80 9d219c07e4 Slight simplification in scale factor computation
[STC](http://tests.stockfishchess.org/tests/view/5b2614000ebc5902b8d17193)
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 17733 W: 3996 L: 3866 D: 9871

[LTC](http://tests.stockfishchess.org/tests/view/5b264d0f0ebc5902b8d17206)
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 55524 W: 9535 L: 9471 D: 36518

Use pawn count scaling also for opposite bishops endings with additional material, with a slope of 2 instead of 7. This simplifies slightly the code.

This PR is a functionally equivalent refactoring of the version which was submitted.

Four versions tried, 2 passed both STC and LTC. I picked the one which seemed more promising at LTC.

Slope 4 passed STC (-0.54 Elo), LTC not attempted
Slope 3 passed STC (+2.51 Elo), LTC (-0.44 Elo)
Slope 2 passed STC (+2.09 Elo), LTC (+0.04 Elo)
Slope 1 passed STC (+0.90 Elo), failed LTC (-3.40 Elo)

Bench: 4761613
2018-06-20 05:27:24 +02:00
DU-jdto a834bfe833 Remove lmrDepth restriction on quiet see pruning
And tweak the threshold value. With this threshold and the current piece
values, this permits see pruning on quiets to be done up to an lmrDepth
of 9 (beyond that the threshold is below -QueenValueMg and see_ge will
pass unconditionally).

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 110316 W: 24612 L: 24667 D: 61037
http://tests.stockfishchess.org/tests/view/5b20aa760ebc5902ab9c9c1d

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17352 W: 2968 L: 2842 D: 11542
http://tests.stockfishchess.org/tests/view/5b20cf1e0ebc5902ab9c9fb6

Closes https://github.com/official-stockfish/Stockfish/pull/1651

Bench: 5069074
2018-06-13 21:52:53 +02:00
Joost VandeVondele 6c36e65193 Remove depth condition for pruning captures.
The SEE condition alone is sufficient.

passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 109863 W: 24339 L: 24392 D: 61132
http://tests.stockfishchess.org/tests/view/5b1f5b000ebc5902ab9c8668

passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 23390 W: 4020 L: 3903 D: 15467
http://tests.stockfishchess.org/tests/view/5b1f94b40ebc5902ab9c8b5e

Closes https://github.com/official-stockfish/Stockfish/pull/1648

Bench: 4834747
2018-06-13 21:46:12 +02:00
protonspring fc3af7c4fb Optimize an expression in endgame.cpp
I believe using foward_file_bb() here is fewer instructions.

a) Fewer instructions and probably more clear (debatable).
b) Possible that a lookup is slower than a few local operations, but the
   forward_file_bb table is probably used often enough that it is always
   cached.

Passed
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21004 W: 4263 L: 4141 D: 12600
http://tests.stockfishchess.org/tests/view/5b1cad830ebc5902ab9c6239

Closes https://github.com/official-stockfish/Stockfish/pull/1644

No functional change.
2018-06-11 10:22:55 +02:00
candirufish 88de112b84 Simplify capture pruning margin formula
Using just `PawnValueEg * depth` as Capture Prune Margin. There was a bunch
of patches that passed recently regarding captures, maybe this part of the
master code redundant? The patch was tested as a simplification:

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20833 W: 4218 L: 4096 D: 12519
http://tests.stockfishchess.org/tests/view/5b1cf2100ebc5902ab9c6651

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27050 W: 3975 L: 3864 D: 19211
http://tests.stockfishchess.org/tests/view/5b1cfdc80ebc5902ab9c6776

Closes https://github.com/official-stockfish/Stockfish/pull/1643

Bench: 4980482
2018-06-11 10:03:33 +02:00
Stefan Geschwentner 86e467520f Optimize contempt value (21)
After several tests it seems best to increase contempt from 12 to 21. This does
not regress against contempt=0 and gives a gain of around 7-8 elo against SF 7
in comparison to current default contempt.

STC: Test for non-regression contempt=21 against contempt=0
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 71250 W: 13956 L: 13926 D: 43368
http://tests.stockfishchess.org/tests/view/5b19a58d0ebc5902ab9c3bfa

STC: Test contempt 21 against SF 7
ELO: 190.06 +-2.8 (95%) LOS: 100.0%
Total: 40000 W: 22608 L: 2676 D: 14716
http://tests.stockfishchess.org/tests/view/5b19a6520ebc5902ab9c3c0e

STC: Test master against SF 7 for comparison
ELO: 182.95 +-2.7 (95%) LOS: 100.0%
Total: 40000 W: 21905 L: 2595 D: 15500
http://tests.stockfishchess.org/tests/view/5b16f5bc0ebc59214346d5ca

LTC: Test for non-regression contempt=21 against contempt=0
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 47666 W: 6914 L: 6832 D: 33920
http://tests.stockfishchess.org/tests/view/5b1a170b0ebc5902ab9c3fde

LTC: Test contempt 21 against SF 7
ELO: 203.92 +-2.6 (95%) LOS: 100.0%
Total: 40000 W: 22447 L: 1340 D: 16213
http://tests.stockfishchess.org/tests/view/5b1a174b0ebc5902ab9c3fe1

LTC: Test master against SF 7 for comparison
ELO: 196.08 +-2.6 (95%) LOS: 100.0%
Total: 40000 W: 21639 L: 1191 D: 17170
http://tests.stockfishchess.org/tests/view/5b1a17e40ebc5902ab9c3fe4

Closes https://github.com/official-stockfish/Stockfish/pull/1646

Bench: 4786912
2018-06-11 09:54:32 +02:00
Joost VandeVondele 2af2c67650 Remove failedLow from the mainThread struct
failedLow is in fact a local variable in Thread::search().

Also clean some spaces and tabs in code.

No functional change.
2018-06-11 09:46:26 +02:00
candirufish 528507b303 Quiet move soft fail high bonus
Extra bonus for quiet move creating a huge soft fail high (triggered
in 21% of quiet bestmoves on a normal bench run). Pb00067 original idea
using PawnValueMg.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 138207 W: 28060 L: 27295 D: 82852
http://tests.stockfishchess.org/tests/view/5b14471b0ebc5902a81689c1

Passe LTC:
LLR: 2.94 (-2.94,2.94) [0.00,5.00]
Total: 157289 W: 23200 L: 22518 D: 111571
http://tests.stockfishchess.org/tests/view/5b149dde0ebc5902a8b41c5a

bench: 4441320
2018-06-06 10:16:48 +02:00
VoyagerOne 231103d2cf Simplify capture bonus
Simplify capture bonus by simply adding ONE_DEPTH instead of being
dependent on BestValue.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24419 W: 4939 L: 4824 D: 14656
http://tests.stockfishchess.org/tests/view/5b16b2040ebc5963ba37e2a5

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 44560 W: 6524 L: 6438 D: 31598
http://tests.stockfishchess.org/tests/view/5b16ccc00ebc59214346d403

Closes https://github.com/official-stockfish/Stockfish/pull/1640

Bench: 4782637
2018-06-06 09:58:34 +02:00
Stéphane Nicolet e4f8a4fa7f Call cycle detection before qsearch()
This has the property of raising alpha before calling qsearch(), thus
maybe giving some more cuts during qsearch(). The patch is equivalent
to the use of cycle detection inside qsearch() at depth 0, but is in
fact implemented by re-ordering code inside search(), which explains
the [0..4] bounds in the following tests.

STC (interrupted after 124250 games, with LLR=0.87):
http://tests.stockfishchess.org/tests/view/5b1500bd0ebc5902a8b420bf
LLR: 0.87 (-2.94,2.94) [0.00,4.00]
Total: 124250 W: 24973 L: 24470 D: 74807

LTC:
http://tests.stockfishchess.org/tests/view/5b1590eb0ebc5902a84dcd09
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 74234 W: 11098 L: 10733 D: 52403

Closes https://github.com/official-stockfish/Stockfish/pull/1635

Bench: 4326784
2018-06-05 18:15:54 +02:00
Joost VandeVondele 9597ad8cab Stabilize AppVeyor CI
After a helpful suggestion from AppVeyor support staff, moving the Stockfish
execution from ps to cmd seems to work. Alternative to PR #1624 tested in PR #1637.

No functional change.
2018-06-05 18:06:13 +02:00
Guenther Demetz b939788f9d Remove a superfluous subtrahend
The '- 1' subtrahend was introduced for guarding against null move
search at root, which would be nonsense. But this is actually already
guaranteed by the !PvNode condition. This followed from the discussion
in pull request 1609: https://github.com/official-stockfish/Stockfish/pull/1609

No functional change
2018-06-05 10:10:32 +02:00
VoyagerOne 6b5d671cdc Simplify LMR for captures
Simplify LMR for captures by removing capture's stat score logic for reduction.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 37068 W: 7462 L: 7370 D: 22236
http://tests.stockfishchess.org/tests/view/5b115bc30ebc591af58a6fd2

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 80061 W: 11706 L: 11671 D: 56684
http://tests.stockfishchess.org/tests/view/5b117f590ebc59033d2d5315

Closes https://github.com/official-stockfish/Stockfish/pull/1631

Bench: 4470519
2018-06-05 09:58:10 +02:00
joergoster f7bae2de82 Bugfix of Position::has_repeated()
The function Position::has_repeated() is used by Tablebases::root_probe()
to determine whether we can rank all winning moves with the same value, or
if we need to strictly rank by dtz in case the position has already been
repeated once, and we are risking to run into the 50-move rule and thus
losing the win (especially critical in some very complicated endgames).

To check whether the current position or one of the previous positions
after the last zeroing move has already been occured once, we start looking
for a repetition of the current position, and if that is not the case, we
step one position back and repeat the check for that position, and so on.

If you now look at how this was done before the new root ranking patch was
merged two months ago, it seems quite obvious that it is a simple oversight:
https://github.com/official-stockfish/Stockfish/commit/108f0da4d7f993732aa2e854b8f3fa8ca6d3b46c

More specifically, after we stepped one position back with

```
stc = stc->previous;
```

we now have to start checking for a repetition with

```
StateInfo* stp = stc->previous->previous;
```

and not with

```
StateInfo* stp = st->previous->previous;
```

Closes https://github.com/official-stockfish/Stockfish/pull/1625

No functional change
2018-06-04 07:45:12 +02:00
Stéphane Nicolet 8ef9bc5a95 Fix overload ambiguity
Fix an error when compiling current master with MSVC due to the
ambiguity of which operator* overload was intended (reported by
Jarrod Torriero).

No functional change.
2018-06-03 09:19:13 +02:00
pb00068 c5d6ae8c96 Extra bonus for capture creating a huge fail high
STC:
http://tests.stockfishchess.org/tests/view/5b114f3d0ebc596e9e0881f6
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 13007 W: 2730 L: 2541 D: 7736

LTC:
http://tests.stockfishchess.org/tests/view/5b1176740ebc59033d2d52c6
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 15594 W: 2417 L: 2239 D: 10938

See https://github.com/official-stockfish/Stockfish/pull/1627

Bench: 4790240
2018-06-02 18:02:35 +02:00
Stéphane Nicolet a0486ecb40 Fix comments, rename variables
Thanks everybody for the various hints in the perpetual renaming thread:
https://github.com/official-stockfish/Stockfish/issues/1426

No functional change
2018-06-02 17:41:37 +02:00
Joost VandeVondele 6cc5614124 Reallocate TT on threadpool resize.
Makes sure the potential benefit of first touch does not depend on
the order of the UCI commands Threads and Hash, by reallocating the
hash if a Threads is issued. The cost is zeroing the TT once more
than needed. In case the prefered order (first Threads than Hash)
is employed, this amounts to zeroing the default sized TT (16Mb),
which is essentially instantaneous.

Follow up for https://github.com/official-stockfish/Stockfish/pull/1601
where additional data and discussion is available.

Closes https://github.com/official-stockfish/Stockfish/pull/1620

No functional change.
2018-06-02 17:03:01 +02:00
Joost VandeVondele 31b8243272 Simplify depth reduction in IID.
Use a constant reduction instead of a depth dependent reduction.

passed STC:
http://tests.stockfishchess.org/tests/view/5b06eb600ebc5914abc12ba8
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 58086 W: 11710 L: 11657 D: 34719

passed LTC:
http://tests.stockfishchess.org/tests/view/5b07b25e0ebc5914abc12c6d
LLR: 3.09 (-2.94,2.94) [-3.00,1.00]
Total: 110414 W: 16217 L: 16219 D: 77978

Bench: 4521056
2018-05-30 13:38:11 +02:00
Rodrigo Exterckötter Tjäder 626d43a329 Use threads to clear the TT faster.
Stockfish currently takes a while to clear the TT when using larger hash sizes.

On one machine with 128 GB hash it takes about 50 seconds with a single thread,
allowing it to use all allocated cores brought that time down to 4 seconds on
some Linux systems. The patch was further tested on Windows and refined with
NUMA binding of the hash initializing threads (we refer to pull request #1601
for the complete discussion and the speed measurements).

Closes https://github.com/official-stockfish/Stockfish/pull/1601

No functional change
2018-05-30 13:07:18 +02:00
VoyagerOne 3d6995eae8 LMR Capture Tweak
Increase capture reduction by comparing opponent's stat score

STC: (Yellow)
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 92291 W: 18647 L: 18350 D: 55294
http://tests.stockfishchess.org/tests/view/5b04225e0ebc5914abc1291b

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 20176 W: 3061 L: 2868 D: 14247
http://tests.stockfishchess.org/tests/view/5b04bb3f0ebc5914abc129a3

How to continue?
• we could try a different value for the capture history threeshold

Closes https://github.com/official-stockfish/Stockfish/pull/1612

Bench: 5167159
2018-05-24 19:06:24 +02:00
protonspring d5d23bfd2d Simplify BlockedByPawn to one dimension
I was able to get this to pass which reduces BlockedByPawn to one dimension
with NO distance from edge offset.

GOOD)  It's more simple and may provide additional clarity for further
simplifications.  Facilitates migrating unblocked to one dimension as well.

BAD) If there is indeed a distance component to BlockedStorm (may or may
not be the case), this obfuscates this component into ShelterStrength and
UnblockedStorm. This may be more convoluted. Also, it may be more convenient
to have each of the three arrays (ShelterStrength, BlockedStorm, and UnBlocked)
be the same size.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 96173 W: 19326 L: 19343 D: 57504
http://tests.stockfishchess.org/tests/view/5b04544d0ebc5914abc12965

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 49818 W: 7441 L: 7363 D: 35014
http://tests.stockfishchess.org/tests/view/5b0487d50ebc5914abc12990

Closes https://github.com/official-stockfish/Stockfish/pull/1611

Bench: 5133208
2018-05-24 18:46:55 +02:00
Joost VandeVondele beaadd6547 Correctly prevent recursive verification in nmp.
As discussed with @pb00068, the condition to prevent recursive verification
was not completely correct. This patch corrects that condition, and adds an
assert. In the current implementation, recursive verification needs to be
avoided in order not to break the verification of the move closer to the
root (i.e. to not set thisThread->nmp_min_ply to zero prematurely).

This patch is tested as a bug fix, based on and tested against PR #1609 .

passed STC:
http://tests.stockfishchess.org/tests/view/5b050f170ebc5914abc129c5
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 7875 W: 1669 L: 1523 D: 4683

passed LTC:
http://tests.stockfishchess.org/tests/view/5b0513970ebc5914abc129cd
LLR: 3.00 (-2.94,2.94) [-3.00,1.00]
Total: 24048 W: 3525 L: 3407 D: 17116

Bench changes at high depth.

Closes https://github.com/official-stockfish/Stockfish/pull/1610
and    https://github.com/official-stockfish/Stockfish/pull/1609

Bench: 4484288
2018-05-24 18:26:34 +02:00
Joost VandeVondele d702203113 Simplify color usage in search.
define Color us and use this instead of pos.side_to_move() and nmp_odd. The latter allows to clarify the nmp verification criterion.

Tested for no regression:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 76713 W: 15303 L: 15284 D: 46126
http://tests.stockfishchess.org/tests/view/5b046a0d0ebc5914abc12971

No functional change.
2018-05-22 22:43:17 +02:00
VoyagerOne 04407e35d6 Simply reset StatScore to zero at beta cutoff
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 43154 W: 8706 L: 8625 D: 25823
http://tests.stockfishchess.org/tests/view/5b01a7660ebc5914abc1271d

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 48155 W: 7036 L: 6955 D: 34164
http://tests.stockfishchess.org/tests/view/5b01b2e50ebc5914abc1272c

Closes https://github.com/official-stockfish/Stockfish/pull/1608

Bench: 4484288
2018-05-22 07:23:04 +02:00
protonspring 335dc2d021 Simplify shelter: always do strength and danger
This check of pawns before subtracting danger can be removed.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21174 W: 4361 L: 4239 D: 12574
http://tests.stockfishchess.org/tests/view/5b00b9f90ebc5914abc12680

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 56980 W: 8377 L: 8309 D: 40294
http://tests.stockfishchess.org/tests/view/5b00ca750ebc5914abc12683

Closes https://github.com/official-stockfish/Stockfish/pull/1607

Bench: 4746692
2018-05-21 09:51:05 +02:00
Stéphane Nicolet fd4585ef07 Simplifying away the progressKey
Simplifying away all the progressKey stuff gives exactly the same bench,
without any speed impact. Tested for speed against master with two benches
at depth 22 ran in parallel:

**testedpatch**
Total time (ms) : 92350
Nodes searched : 178962949
Nodes/second : 1937877

**master**
Total time (ms) : 92358
Nodes searched : 178962949
Nodes/second : 1937709

We also tested the patch at STC for no-regression with [-3, 1] bounds:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 57299 W: 11529 L: 11474 D: 34296
http://tests.stockfishchess.org/tests/view/5b015a1c0ebc5914abc126e5

Closes https://github.com/official-stockfish/Stockfish/pull/1603

No functional change.
2018-05-21 09:37:44 +02:00
mstembera e8005ebe56 Fix MSVC errors in tbprobe.cpp
Default template parameters values and recursive functions do not play well
together. Fix for below errors that showed up after updating to latest MSVC.

````
tbprobe.cpp(1156): error C2672:
    'search': no matching overloaded function found

tbprobe.cpp(1198): error C2783:
  'Tablebases::WDLScore `anonymous-namespace'::search(Position &,Tablebases::ProbeState *)':
could not deduce template argument for 'CheckZeroingMoves'

````

Closes https://github.com/official-stockfish/Stockfish/pull/1594

No functional change.
2018-05-16 23:40:23 +02:00
Tom Truscott 91a76331ca Use cycle detection to bound search value
A position which has a move which draws by repetition, or which could have
been reached from an earlier position in the game tree, is considered to be
at least a draw for the side to move.

Cycle detection algorithm by Marcel van Kervink:

       https://marcelk.net/2013-04-06/paper/upcoming-rep-v2.pdf

----------------------------

How does the algorithm work in practice? The algorithm is an efficient
method to detect if the side to move has a drawing move, without doing any
move generation, thus possibly giving a cheap cutoffThe most interesting
conditions are both on line 1195:

```
  if (   originalKey == (progressKey ^ stp->key)
      || progressKey == Zobrist::side)
```

This uses the position keys as a sort-of Bloom filter, to avoid the expensive
checks which follow. For "upcoming repetition" consider the opening Nf3 Nf6 Ng1.
The XOR of this position's key with the starting position gives their difference,
which can be used to look up black's repeating move (Ng8). But that look-up is
expensive, so line 1195 checks that the white pieces are on their original squares.

This is the subtlest part of the algorithm, but the basic idea in the above game
is there are 4 positions (starting position and the one after each move). An XOR
of the first pair (startpos and after Nf3) gives a key matching Nf3. An XOR of
the second pair (after Nf6 and after Ng1) gives a key matching the move Ng1. But
since the difference in each pair is the location of the white knight those keys
are "identical" (not quite because while there are 4 keys the the side to move
changed 3 times, so the keys differ by Zobrist::side). The loop containing line
1195 does this pair-wise XOR-ing.

Continuing the example, after line 1195 determines that the white pieces are
back where they started we still need to make sure the changes in the black
pieces represents a legal move. This is done by looking up the "moveKey" to
see if it corresponds to possible move, and that there are no pieces blocking
its way. There is the additional complication that, to match the behavior of
is_draw(), if the repetition is not inside the search tree then there must be
an additional repetition in the game history. Since a position can have more
than one upcoming repetition a simple count does not suffice. So there is a
search loop ending on line 1215.

On the other hand, the "no-progress' is the same thing but offset by 1 ply.
I like the concept but think it currently has minimal or negative benefit,
and I'd be happy to remove it if that would get the patch accepted. This
will not, however, save many lines of code.

-----------------------------

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 36430 W: 7446 L: 7150 D: 21834
http://tests.stockfishchess.org/tests/view/5afc123f0ebc591fdf408dfc

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 12998 W: 2045 L: 1876 D: 9077
http://tests.stockfishchess.org/tests/view/5afc2c630ebc591fdf408e0c

How could we continue after the patch:

• The code in search() that checks for cycles has numerous possible variants.
  Perhaps the check need could be done in qsearch() too.

• The biggest improvement would be to get "no progress" to be of actual benefit,
  and it would be helpful understand why it (probably) isn't. Perhaps there is an
  interaction with the transposition table or the (fantastically complex) tree
  search. Perhaps this would be hard to fix, but there may be a simple oversight.

Closes https://github.com/official-stockfish/Stockfish/pull/1575

Bench: 4550412
2018-05-16 22:51:43 +02:00
VoyagerOne aacee91a5a Update search.cpp
At PvNodes allow bonus for prior counter move that caused a fail low
for depth 1 and 2. Note : I did a speculative LTC on yellow STC patch
since history stats tend to be highly TC sensitive

STC (Yellow):
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 64295 W: 13042 L: 12873 D: 38380
http://tests.stockfishchess.org/tests/view/5af507c80ebc5968e6524153

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22407 W: 3413 L: 3211 D: 15783
http://tests.stockfishchess.org/tests/view/5af85dd40ebc591fdf408b87

Also use local variable excludedMove in NMP (marotear)

Bench:  5294316
2018-05-14 06:52:48 +02:00
Stéphane Nicolet bf01bd0755 Tweak kingAttackersCount and KingAttackWeights
Use the whole kingRing for pawn attackers instead of only the squares directly
around the king. This tends to give quite a lot more kingAttackersCount, so to
compensate and to avoid raising the  king danger too fast we lower the values
in the KingAttackWeights array a little bit.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 51892 W: 10723 L: 10369 D: 30800
http://tests.stockfishchess.org/tests/view/5af6d4dd0ebc5968e652428e

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 24536 W: 3737 L: 3515 D: 17284
http://tests.stockfishchess.org/tests/view/5af709890ebc5968e65242ac

Credits to user @xoroshiro for the idea of using the kingRing for pawn attackers.

How to continue? It seems that the KingAttackWeights[] array stores values
which are quite Elo-sensitive, yet they have not been tuned with SPSA recently.
There might be easy Elo points to get there.

Closes https://github.com/official-stockfish/Stockfish/pull/1597

Bench: 5282815
2018-05-13 07:20:39 +02:00
mstembera c163c2fcb0 Include all blockers in king danger
Simplification: in king danger, include all blockers and not only pinned
pieces, since blockers enemy pieces can result in discovered checks which
are also bad.

STC  http://tests.stockfishchess.org/tests/view/5af35f9f0ebc5968e6523fe9
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 145781 W: 29368 L: 29478 D: 86935

LTC http://tests.stockfishchess.org/tests/view/5af3cb430ebc5968e652401f
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 76398 W: 11272 L: 11232 D: 53894

I also incorrectly scheduled STC with [0,5] which it failed.
http://tests.stockfishchess.org/tests/view/5af283c00ebc5968e6523f33
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 12338 W: 2451 L: 2522 D: 7365

Closes https://github.com/official-stockfish/Stockfish/pull/1593

bench: 4698290

----------------------------------------

Thanks to @vondele and @Rocky640 for a cleaner version of the patch,
and the following comments!

> Most of the pinned, (or for this pull request, blocking) squares were
> already computed in the unsafeChecks, the only missing squares being:
>
> a) squares attacked by a Queen which are occupied by friendly piece
>    or "unsafe". Note that adding such squares never passed SPRT[0,5].
>
> b) squares not in mobilityArea[Us].
>
> There is a strong relationship between the blockers and the unsafeChecks,
> but the bitboard unsafeChecks is still useful when the checker is not
> aligned with the king, and the checking square is occupied by friendly
> piece or is "unsafe". This is always the case for the Knight.
2018-05-10 23:36:16 +02:00
protonspring 255df4ffae Consolidate pawn storm types
Simplification: the Unopposed and Unblocked pawn storm types are mathematically
similar enough to combine with no Elo loss. This reduces the pawn storm types
to BlockedByPawn and UnBlocked.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 116869 W: 23549 L: 23605 D: 69715
http://tests.stockfishchess.org/tests/view/5af2def90ebc5968e6523f82

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39912 W: 6090 L: 5998 D: 27824
http://tests.stockfishchess.org/tests/view/5af3b2e20ebc5968e6524013

Closes https://github.com/official-stockfish/Stockfish/pull/1592

Bench: 5244314
2018-05-10 15:46:32 +02:00
Stefan Geschwentner 76d4a80084 Tune null move pruning
Use a lower stat score threshold of 22500.

Failed yellow at STC after many games (~0.92 Elo):
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 73978 W: 14947 L: 14834 D: 44197
http://tests.stockfishchess.org/tests/view/5af2deec0ebc5968e6523f80

But scales good and passed LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 23147 W: 3453 L: 3237 D: 16457
http://tests.stockfishchess.org/tests/view/5af3cf820ebc5968e6524022

Closes https://github.com/official-stockfish/Stockfish/pull/1591

Bench: 4777674
2018-05-10 12:28:18 +02:00
candirufish fc0a1f37cf Tuned some pawns and evaluation constants
Tuned values in pawns.cpp and evaluate.cpp after a SPSA session:
419k games 60sec 600nodetime.  We have adjusted the PassedRank[]
output of the SPSA session to keep increasing values with rank,
and PassedFile[] output to keep the West <–> East symmetry of
the evaluation.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 58948 W: 12431 L: 12046 D: 34471
http://tests.stockfishchess.org/tests/view/5af2302f0ebc5968e6523f0a

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 22211 W: 3468 L: 3251 D: 15492
http://tests.stockfishchess.org/tests/view/5af264c80ebc5968e6523f1a

See https://github.com/official-stockfish/Stockfish/pull/1587
and https://github.com/official-stockfish/Stockfish/pull/1590

bench: 4437531
2018-05-09 09:22:11 +02:00
Joost VandeVondele 4d647428d8 Remove goto, limit skipping to NMP
This patch simplifies the control flow in search(), removing an if
and a goto. A side effect of the patch is that Stockfish is now a
little bit more selective at low depths, because we allow razoring,
futility pruning and probcut pruning after a null move.

passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32035 W: 6523 L: 6422 D: 19090
http://tests.stockfishchess.org/tests/view/5af142ca0ebc597fb3d39bb6

passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41431 W: 6187 L: 6097 D: 29147
http://tests.stockfishchess.org/tests/view/5af148770ebc597fb3d39bc1

Ideas for further work:

• Use the nodes credit opened by the patch (the increased selectivity)
  to try somewhat higher razoring, futility or probcut margins at [0..4].

Bench: 4855031
2018-05-09 01:42:34 +02:00
Stefan Geschwentner de0889612f Tweak null move pruning conditions
Allow null move pruning only if last stat score < 30000.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 12653 W: 2641 L: 2454 D: 7558
http://tests.stockfishchess.org/tests/view/5af1d8830ebc5968e6523edb

Passed LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 19782 W: 3072 L: 2878 D: 13832
http://tests.stockfishchess.org/tests/view/5af1ee8c0ebc5968e6523ee4

Ideas for further work:

• Tune the stat score threshold.
• Try depth based stat score threshold.
• Try stat score condition for other prunings.

Closes https://github.com/official-stockfish/Stockfish/pull/1589

Bench: 5088156
2018-05-08 23:31:55 +02:00
Alain SAVARD 6634f41ac9 Drop the lever condition for backwards
We can view the patch version as adding some "undermining bonus" for
level pawns, when the defending side can not easily avoid the exchange
by advancing her pawn.

• Case 1) White b2,c3, Black a3,b3:
          Black is breaking through, b2 deserves a penalty

• Case 2) White b2,c3, Black a3,c4:
          if b2xa3 then White ends up with a weak pawn on a3
          and probably a weak pawn on c3 too.

In either case, White can still not safely play b2-b3 and make a
phalanx with c3, which is the essence of a backward pawn definition.

Passed STC in SPRT[0, 4]:
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 131169 W: 26523 L: 26199 D: 78447
http://tests.stockfishchess.org/tests/view/5aefa4d50ebc5902a409a151
ELO 1.19 [-0.38,2.88] (95%)

Passed LTC in SPRT[-3, 1]:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 24824 W: 3732 L: 3617 D: 17475
http://tests.stockfishchess.org/tests/view/5af04d3f0ebc5902a88b2e55
ELO 1.27 [-1.21,3.70] (95%)

Closes https://github.com/official-stockfish/Stockfish/pull/1584

How to continue from there?

There were some promising tests a couple of months ago about adding
a lever condition for king danger in evaluate.cpp, maybe it would
be time to re-try this after all the recent changes in pawns.cpp

Bench: 4773882
2018-05-08 11:14:43 +02:00
Joost VandeVondele 8d1625d6df Remove skipEarlyPruning argument to search()
Remove nine boolean arguments and the corresponding skipEarlyPruning variable.
Instead, skip early pruning only when there is an excluded move, and try null
move pruning only if the previous move was not itself a null move.

passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33623 W: 6853 L: 6755 D: 20015
http://tests.stockfishchess.org/tests/view/5aef462a0ebc5902a409a10e

passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39474 W: 5882 L: 5789 D: 27803
http://tests.stockfishchess.org/tests/view/5aefd4b80ebc5902a409a164

Closes https://github.com/official-stockfish/Stockfish/pull/1585

Bench: 4953556
2018-05-08 10:43:56 +02:00
Joost VandeVondele 155d5417d9 Improve signature script
Catch case of missing bench, indicative of a crash or assert.

No functional change
2018-05-08 10:32:23 +02:00
protonspring cb4bda0f49 Simplify the backward pawns code
The two lines of code in the patch seem to be just as good as master.

1. We now only look at the current square to see if it is currently backward,
whereas master looks there AND further ahead in the current file (master would
declare a pawn "backward" even though it could still safely advance a little).
This simplification allows us to avoid the use of the difficult logic with
`backmost_sq(Us, neighbours | stoppers)`.

2. The condition `relative_rank(Us,s) < RANK_5` is simplified away.

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 68132 W: 14025 L: 13992 D: 40115
http://tests.stockfishchess.org/tests/view/5aedc97a0ebc5902a4099fd6

Passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 23789 W: 3643 L: 3527 D: 16619
http://tests.stockfishchess.org/tests/view/5aee4f970ebc5902a409a03a

Ideas for further work:

• The new code flags some pawns on the 5th rank as backward, which was not the
case in the old master. So maybe we should test a version with that included?

• Further tweaks of the backward condition with [0..5] bounds?

Closes https://github.com/official-stockfish/Stockfish/pull/1583

Bench: 5122789
2018-05-06 09:44:14 +02:00
Stéphane Nicolet d4cb80b210 Tweak the connected[] array value for pawns on rank 5
A recent tuning session by Jerry Donald Watson suggested that the
value for the pawns on the fifth rank in the connected[] array were
a little bit too high in master. We lower here this value from 75 to 65.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 27399 W: 5646 L: 5384 D: 16369
http://tests.stockfishchess.org/tests/view/5aea17c50ebc5902a1bed396

LTC:
LLR: 3.66 (-2.94,2.94) [0.00,4.00]
Total: 95590 W: 14529 L: 14062 D: 66999
http://tests.stockfishchess.org/tests/view/5aea34a40ebc5902a104ebe5

Closes https://github.com/official-stockfish/Stockfish/pull/1580

Bench: 5186783
2018-05-03 22:00:07 +02:00
Stéphane Nicolet 06e0134cbc Correct a bug introduced by Stéphane in the previous patch.
When we are using the "Bitboard + Square" overloaded operators,
the compiler uses the interpediate SquareBB[s] to transform the
square into a Bitboard, and then calculate the result.

For instance, the following code:

```
   b = pos.pieces(Us, PAWN) & s
```

generates in fact the code:

```
   b = pos.pieces(Us, PAWN) & SquareBB[s]`
```

The bug introduced by Stéphane in the previous patch was the
use of `b = pos.pieces(Us, PAWN) & (s + Up)` which can result
in out-of-bounds errors for the SquareBB[] array if s in the
last rank of the board.

We coorect the bug, and also add some asserts in bitboard.h to
make the code more robust for this particular bug in the future.

Bug report by Joost VandeVondele. Thanks!

Bench: 5512000
2018-05-02 13:38:00 +02:00
protonspring 12ef8f71a2 Use special rule for BlockedByKing
Simplification: remove BlockedByKing from storm array and use a special rule.

The BlockedByKing section in the storm array is substantially similar to the
Unopposed section except for two extreme values V(-290), V(-274). Turns out
removing BlockedByKing and using a special rule for these two values shows
no Elo loss. All the other values in the BlockedByKing section are apparently
irrelevant. BlockedByKing now falls under unopposed which (to me) is a bit
more logical since there is no defending pawn on this file. Also, retuning
the Unopposed section may be another improvement.

GOOD) This is a simplification because the entire BlockedByKing section of
the storm array goes away reducing a few lines of code (and less values to
tune). This also brings clarity because the special rule is self documenting.

BAD) It takes execution time to apply the special rule. This should be negli-
gible because it is based on a template parameter and is boiled down to two
bitwise AND's.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 33470 W: 6820 L: 6721 D: 19929
http://tests.stockfishchess.org/tests/view/5ae7b6e60ebc5926dba90e13

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 47627 W: 7045 L: 6963 D: 33619
http://tests.stockfishchess.org/tests/view/5ae859ff0ebc5926dba90e85

Closes https://github.com/official-stockfish/Stockfish/pull/1574

Bench: 5512000

-----------

How to continue after this patch?

This patch may open the possibility to move the special rule to evaluate.cpp
in the evaluate::king() function, where we could refine the rule using king
danger information. For instance, with a king in H2 blocking an opponent pawn
in H3, it may be critical to know that the opponent has no safe check in G2
before giving the bonus :-)
2018-05-01 23:55:30 +02:00
MJZ1977 5a7cdadfb3 Penalty for bad bishop with blocked central files
We increase the penalty for bad bishops by a factor proportional
to the number of our blocked pawns in the center files C, D, E or F.

STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 8868 W: 1870 L: 1700 D: 5298
http://tests.stockfishchess.org/html/live_elo.html?5ae7674f0ebc590e39268b34

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 5813 W: 950 L: 808 D: 4055
http://tests.stockfishchess.org/html/live_elo.html?5ae77bae0ebc5926dba90dd9

Closes https://github.com/official-stockfish/Stockfish/pull/1573

Bench: 5364190
2018-05-01 07:12:27 +02:00
Stefano80 213166ba22 Always scale using pawn contribution
This is a further step in the long quest for a simple way of determining
scale factors for the endgame.

Here we remove the artificial restriction in evaluate_scale_factor()
based on endgame score. Also SCALE_FACTOR_ONEPAWN can be simplified
away. The latter is a small non functional simplification with respect
to the version that was testedin the framework, verified on bench with
depth 22 for good measure.

Passed STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 49438 W: 9999 L: 9930 D: 29509
http://tests.stockfishchess.org/tests/view/5ae20c8b0ebc5963175205c8

Passed LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 101445 W: 15113 L: 15110 D: 71222
http://tests.stockfishchess.org/tests/view/5ae2a0560ebc5902a1998986

How to continue from there?

Maybe the general case could be scaled with pawns from both colors
without losing Elo. If that is the case, then this could be merged
somehow with the scaling in evaluate_initiative(), which also uses
a additive malus down when the number of pawns in the position goes
down.

Closes https://github.com/official-stockfish/Stockfish/pull/1570

Bench: 5254862
2018-04-29 07:26:25 +02:00
Joost VandeVondele d6252ef202 Simplifiy IID condition
do IID for all sufficiently deep searches without TT move.

passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25494 W: 5313 L: 5199 D: 14982
http://tests.stockfishchess.org/tests/view/5ae1e8dd0ebc596317520583

passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 80582 W: 12162 L: 12130 D: 56290
http://tests.stockfishchess.org/tests/view/5ae1f5ab0ebc5963175205a4

Bench: 4966970
2018-04-29 06:59:17 +02:00
Joost VandeVondele 3df8cabb84 Fix 'make strip' for mingw.
Currently the make strip target is broken on mingw as the exe name is wrong (stockfish instead of stockfish.exe).

Needs some testing by mingw users (both profile-build and strip, native and cross).

No functional change.
2018-04-29 06:53:51 +02:00
Stéphane Nicolet 11967e89cd Update various comments
Spotted by Alain Savard, Joost VandeVondele, Ronald de Man, Fabian Fichter, Chris Cain, xoto10

No functional change
2018-04-29 06:48:18 +02:00
Alain SAVARD 45072612d4 No Queen in the MobilityArea
Queen was recently excluded from the mobility area of friendly minor
pieces. Exclude queen also from the mobility area of friendly majors too.

Run as a simplification:

STC
http://tests.stockfishchess.org/tests/view/5ade396f0ebc59602d053742
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 46972 W: 9511 L: 9437 D: 28024

LTC
http://tests.stockfishchess.org/tests/view/5ade64b50ebc5949f20a24d3
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 66855 W: 10157 L: 10105 D: 46593

How to continue from there?

The mobilityArea is used in various places of the evaluation as a
soft proxy for "not attacked by the opponent pawns". Now that the
mobility area is getting smaller and smaller, it may be worth to
hunt for Elo gains by trying the more direct ~attackedBy[Them][PAWN]
instead of mobilityArea[Us] in these places.

Bench: 4650572
2018-04-24 14:41:50 +02:00
Stéphane Nicolet c794c8c801 Simplify the ShelterStrength[] array
Remove the distinction between the king file and the two neighbours
files in the ShelterStrength[] array. Instead we initialize the safety
variable in the evaluate_shelter() function with a -10 penalty if our
king is on a semi-open file (ie. if our king is on a file without a pawn
protection).

Also rename shelter_storm() to evaluate_shelter() while there.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 23153 W: 4795 L: 4677 D: 13681
http://tests.stockfishchess.org/tests/view/5adcb83d0ebc595ec7ff8aa7

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25728 W: 3934 L: 3821 D: 17973
http://tests.stockfishchess.org/tests/view/5adcdcb60ebc595ec7ff8adb

See the commit history in PR#1559 for the proof that the committed
version is equivalent to the version in the tests above:
https://github.com/official-stockfish/Stockfish/pull/1559

Full credit to @protonspring for the renormalized values of the
ShelterStrength[] array used for the simplification. Thanks!

Bench: 4703935
2018-04-23 09:49:05 +02:00
ceebo aef7076c34 Alternative formula for dynamic contempt
Replace the formula involving arctan with something having similar
behaviour that can be implemented using integer-only operations.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 34781 W: 7189 L: 7093 D: 20499
http://tests.stockfishchess.org/tests/view/5ad7c95f0ebc595700526e76

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 39743 W: 5950 L: 5857 D: 27936
http://tests.stockfishchess.org/tests/view/5ad886ee0ebc595700526e9b

Closes https://github.com/official-stockfish/Stockfish/pull/1558

Bench: 5549801
2018-04-23 09:13:27 +02:00
Joost VandeVondele a64d524d02 Remove template W param
Different W template parameters for the different statEntries are
not needed, simplify by consistently using 32.

Passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21683 W: 4545 L: 4424 D: 12714
http://tests.stockfishchess.org/tests/view/5ad797550ebc595700526e59

Passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 110391 W: 16551 L: 16562 D: 77278
http://tests.stockfishchess.org/tests/view/5ad7a88f0ebc595700526e61

Closes https://github.com/official-stockfish/Stockfish/pull/1557

Bench: 5120532
2018-04-23 08:46:32 +02:00
Stéphane Nicolet 82f7d507ea Allow UCI parameters to be double
Change the operators of the Option type in uci.h to accept floating
point numbers in double precision on input as the numerical type for
the "spin" values of the UCI protocol.

The output of Stockfish after the "uci" command is unaffected.

This change is compatible with all the existing GUI (as they will
continue sending integers that we can interpret as doubles in SF),
and allows us to pass double parameters to Stockfish in the console
via the "setoption" command. This will be useful if we implement
another tuner as an alternative for SPSA.

Closes https://github.com/official-stockfish/Stockfish/pull/1556

No functional change.

---------------------

A example of the new functionality in action in the branch `tune_float2'`:
https://github.com/snicolet/Stockfish/commit/876c322d0f20ee232da977b4d3489c4cc929765e

I have added the following lines in ucioptions.cpp:

```C++

void on_pi(const Option& o)
{
      double x = Options["PI"];  // or double x = o;
      std::cerr << "received value is x = " << x << std::endl;
}

...

o["PI"]   << Option(3.1415926, -10000000, 10000000, on_pi);
```

Then I can change the value of Pi in Stockfish via the command line, and
check that Stockfish understands a floating point:

````
> ./stockfish
> setoption name PI value 2.7182818284

received value is x = 2.71828
````

On output, the default value of Pi is truncated to 3 (to remain compatible
with the UCI protocol and GUIs):

````
> uci

[...]
option name SyzygyProbeLimit type spin default 6 min 0 max 6
option name PI type spin default 3 min -10000000 max 10000000
uciok
````
2018-04-23 08:08:27 +02:00
protonspring f7cc0026e3 Simplify BlockedByKing in pawn storms
This patch is non-functional. Current master does four operations to determine
whether an enemy pawn on this file is blocked by the king or not

```
f == file_of(ksq) && rkThem == relative_rank(Us, ksq) + 1 )
```

By adding a direction (based on the template color), this is reduced to two
operations. This works because b is limited to enemy pawns that are ahead of
the king and on the current file.

```
shift<Down>(b) & ksq
```

I've added a line of code, but the number of executing instructions is reduced
(I think). I'm not sure if this counts as a simplification, but it should
theoretically be a little faster (barely). The code line length is also reduced
making it a little easier to read.

Closes https://github.com/official-stockfish/Stockfish/pull/1552

No functional change.
2018-04-18 20:03:52 +02:00
Stéphane Nicolet 73e8daa150 Simplify the shelter mask
The line 246 of pawns.cpp is not necessary, because we restrict the bitboards
to file_of(f) in lines 253 and 256 anyway.

No functional change.
2018-04-18 19:53:39 +02:00
Alain SAVARD 66af80972a Tidy up on Overload
Move new code in the if (defended | weak) {} section.

No functional change
2018-04-18 19:44:41 +02:00
Joost VandeVondele 0143c6f0c2 Document Elo impact of the LMR part of search
Similar to before, document Elo impact of various LMR steps

Tests run by @jerrydonaldwatson

t1 http://tests.stockfishchess.org/tests/view/5abece950ebc591a560aad0b
t2 http://tests.stockfishchess.org/tests/view/5abecf0c0ebc591a560aad0d
t3 http://tests.stockfishchess.org/tests/view/5abecf7b0ebc591a560aad0f
t4 http://tests.stockfishchess.org/tests/view/5abecfe70ebc591a560aad14
t5 http://tests.stockfishchess.org/tests/view/5abed42b0ebc591a560aad33
t6 http://tests.stockfishchess.org/tests/view/5abed0b90ebc591a560aad19
t7 http://tests.stockfishchess.org/tests/view/5abed1240ebc591a560aad1b
t8 http://tests.stockfishchess.org/tests/view/5abed1b90ebc591a560aad1d

No functional change.
2018-04-18 19:17:37 +02:00
syzygy1 108f0da4d7 Tablebases root ranking
This patch corrects both MultiPV behaviour and "go searchmoves" behaviour
for tablebases.

We change the logic of table base probing at root positions from filtering
to ranking. The ranking code is much more straightforward than the current
filtering code (this is a simplification), and also more versatile.

If the root is a TB position, each root move is probed and assigned a TB score
and a TB rank. The TB score is the Value to be displayed to the user for that
move (unless the search finds a mate score), while the TB rank determines which
moves should appear higher in a multi-pv search. In game play, the engine will
always pick a move with the highest rank.

Ranks run from -1000 to +1000:

901 to 1000   : TB win
900           : normally a TB win, in rare cases this could be a draw
1 to 899      : cursed TB wins
0             : draw
-1 to -899    : blessed TB losses
-900          : normally a TB loss, in rare cases this could be a draw
-901 to -1000 : TB loss

Normally all winning moves get rank 1000 (to let the search pick the best
among them). The exception is if there has been a first repetition. In that
case, moves are ranked strictly by DTZ so that the engine will play a move
that lowers DTZ (and therefore cannot repeat the position a second time).

Losing moves get rank -1000 unless they have relatively high DTZ, meaning
they have some drawing chances. Those get ranks towards -901 (when they
cross -900 the draw is certain).

Closes https://github.com/official-stockfish/Stockfish/pull/1467

No functional change (without tablebases).
2018-04-18 18:46:24 +02:00
syzygy1 e9aeaad052 Analysis Contempt combo box
This patch introduces an Analysis Contempt UCI combo box to control
the behaviour of contempt during analysis. The possible values are
Both, Off, White, Black. Technically, the engine is supposed to be in
analysis mode if UCI_AnalyseMode is set by the graphical user interface
or if the user has chosen infinite analysis mode ("go infinite").

Credits: the idea for the combo box is due to Michel Van den Bergh.

No functional change (outside analysis mode).

-----------------------------------------------------

The so-called "contempt" is an optimism value that the engine adds
to one color to avoid simplifications and keep tension in the position
during its search. It was introduced in Stockfish 9 and seemed to give
good results during the TCEC 11 tournament (Stockfish seemed to play a
little bit more actively than in previous seasons).

The patch does not change the play during match or blitz play, but gives
more options for correspondance players to decide for which color(s) they
would like to use contempt in analysis mode (infinite time). Here is a
description of the various options:

* Both  : in analysis mode, use the contempt for both players (alternating)
* Off   : in analysis mode, use the contempt for none of the players
* White : in analysis mode, White will play actively, Black will play passively
* Black : in analysis mode, Black will play actively, White will play passively
2018-04-18 17:49:19 +02:00
syzygy1 ae0bb6dc2b Fix reported DTZ for mate-in-1-ply positions
This corrects a bug in Tablebases::probe_dtz() which sometimes causes
a higher DTZ value to be returned for the position one ply before mate
than for the position two plies before mate.

The problem was reported by Kolja Kühn here:
http://talkchess.com/forum/viewtopic.php?p=757497#757497

It is explained here:
http://talkchess.com/forum/viewtopic.php?p=757506#757506

I have also adjusted some comments to make clear that probe_dtz()
returns -1 for a mate position.

Closes https://github.com/official-stockfish/Stockfish/pull/1546

No functional change
2018-04-14 08:36:41 +02:00
Marco Costalba 6413d9b1f9 Further documentation and coding style on TB code
This patch adds some documentation and code cleanup to tablebase code.

It took me some time to understand the relation among the differrent
structs, although I have rewrote them fully in the past. So I wrote
some detailed documentation to avoid the same efforts for future readers.

Also noteworthy is the use a standard hash table implementation with a
more efficient 1D array instead of a 2D array. This reduces the average
lookup steps of 90% (from 343 to 38 in a bench 128 1 16 run) and reduces
also the table from 5K to 4K
entries.

I have tested on 5-men and no functional and no slowdown reported. It
should be verified on 6-men that the new hash does not overflow. It is
enough to run ./stockfish with 6-men available: if it does not assert at
startup it means everything is ok with 6-men too.

EDIT: verified for 6-men tablebase by Jörg Oster. Thanks!

No functional change.
2018-04-12 09:37:18 +02:00
Jerry Donald Watson 62619fa228 Simplify condition in space definition
We remove an unnecessary condition in the definition of safe squares
in the space evaluation. Only the squares which are occupied by our
pawns or attacked by our opponent's pawns are now excluded.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21096 W: 4321 L: 4199 D: 12576
http://tests.stockfishchess.org/tests/view/5acbf7510ebc59547e537d4e

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 23437 W: 3577 L: 3460 D: 16400
http://tests.stockfishchess.org/tests/view/5acc0f750ebc59547e537d6a

It may be possible to further refine the definition of such safe squares.

Bench: 5351765
2018-04-10 17:47:55 +02:00
VoyagerOne b88374b14a Reset negative statScore on fail high
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 9073 W: 1937 L: 1766 D: 5370

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 53530 W: 8139 L: 7823 D: 37568

Bench: 5170165
2018-04-07 20:26:49 +02:00
Mark Tenzer 759d129e47 Introduce Overload
This patch applies a S(10, 5) bonus for every square that is:

- Occupied by an enemy piece which is not a pawn
- Attacked exactly once by our pieces
- Defended exactly once by enemy pieces

The idea is that these pieces must be defended. Their defenders have
dramatically limited mobility, and they are vulnerable to our future
attack.

As with connectivity, there are probably many more tests to be run in
this area. In particular:

- I believe @snicolet's queen overload tests have demonstrated a potential
  need for a queen overload bonus above and beyond this one; however, the
  conditions for "overload" in this patch are different (excluding pieces
  we attack twice).  My next test after this is (hopefully) merged will be
  to intersect the Bitboard I define here with the enemy's queen attacks and
  attempt to give additional bonus.
- Perhaps we should exclude pieces attacked by pawns--can pawns really be
  overloaded? Should they have the same weight, or less?  This didn't work
  with a previous version, but it could work with this one.
- More generally, different pieces may need more or less bonus. We could
  change bonuses based on what type of enemy piece is being overloaded, what
  type of friendly piece is attacking, and/or what type of piece is being
  defended by the overloaded piece and attacked by us, or any intersection
  of these three.  For example, here attacked/defended pawns are excluded,
  but they're not totally worthless targets, and could be added again with
  a smaller bonus.
- This list is by no means exhaustive.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 17439 W: 3599 L: 3390 D: 10450
http://tests.stockfishchess.org/tests/view/5ac78a2e0ebc59435923735e

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 43304 W: 6533 L: 6256 D: 30515
http://tests.stockfishchess.org/tests/view/5ac7a1d80ebc59435923736f

Closes https://github.com/official-stockfish/Stockfish/pull/1533

Bench: 5248871

----------------

This is my first time opening a PR, so I apologize if there are errors.
There are too many people to thank since I submitted my first test just
over a month ago. Thank you all for the warm welcome and here is to more
green patches!

In particular, I would like to thank:
- @crossbr, whose comment in a FishCooking thread first inspired me to
            consider the overloading of pieces other than queens,
- @snicolet, whose queen overload tests inspired this one and served as
             the base of my first overload attempts,
- @protonspring, whose connectivity tests inspired this one and who provided
                 much of the feedback needed to take this from red to green,
- @vondele, who kindly corrected me when I submitted a bad LTC test,
- @Rocky640, who has helped me over and over again in the past month.

Thank you all!
2018-04-07 01:31:23 +02:00
Alain SAVARD 04a228f9c8 Remove the Queen from the mobility area of minor pieces
In master, we already remove the King from the mobility area of minor pieces
because the King simply stands in the way of other pieces, and since opponent
cannot capture the King, any piece which "protects" the King cannot recapture.

Similarly, this patch introduces the idea that it is rarely a need for a Queen
to be "protected" by a minor (unless it is attacked only by a Queen, in fact).
We used to have a LoosePiece bonus, and in a similar vein the Queen was excluded
from that penalty.

Idea came when reviewing an old game of Kholmov. He was a very good midgame
player, but in the opening his misplace his Queen (and won in the end :-) :
     http://www.chessgames.com/perl/chessgame?gid=1134645

Both white queen moves 10.Qd3 and 13.Qb3 are in the way of some minor piece.
I would prefer to not give a bishop mobility bonus at move 10 for the square d3,
or later a knight mobility bonus at move 13 for the square b3. And the textbook
move is 19.Qe3! which prepares 20.Nb3. This short game sample shows how much a
queen can be "in the way" of minor pieces.

STC
http://tests.stockfishchess.org/tests/view/5ac2c15f0ebc591746423fa3
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22066 W: 4561 L: 4330 D: 13175

LTC
http://tests.stockfishchess.org/tests/view/5ac2d6500ebc591746423faf
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 25871 W: 3953 L: 3738 D: 18180

Closes https://github.com/official-stockfish/Stockfish/pull/1532

Ideas for future work in this area:

• tweak some more mobility areas for other piece type.
• construct a notion of global mobility for the whole piece set.
• bad bishops.

Bench: 4989125
2018-04-04 00:07:40 +02:00
Torsten Franz 06a8fd2154 Simplify ThreatBySafePawn evaluation
Simplify ThreatBySafePawn evaluation by removing the 'if (weak)' speed
optimization check from threats evaluation. This is a non functional
change as it removes just a speed optimization conditional which was
probably useful before but does no longer provide benefits. This section
section had a few more lines not long ago, with ThreatByHangingPawn and
a loop through the threatened pieces, but now there is not much left.

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 47775 W: 9696 L: 9624 D: 28455
http://tests.stockfishchess.org/tests/view/5ac298910ebc591746423f8b

Closes https://github.com/official-stockfish/Stockfish/pull/1531

Non functional change.
2018-04-03 23:34:10 +02:00
Joost VandeVondele 0cfb653eec Simplify Singular Extension condition
Avoid defining a singly used variable, removes one condition.

passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 53489 W: 10814 L: 10752 D: 31923
http://tests.stockfishchess.org/tests/view/5ac08a8d0ebc590e9457cd94

Closes https://github.com/official-stockfish/Stockfish/pull/1530

No functional change.
2018-04-03 00:13:38 +02:00
Joost VandeVondele e408fd7b10 Document Elo impact of various parts of search
In order to understand better the impact of various techniques used in search,
Elo estimates have been run at STC for 60000 games (statistical error ~1.8 Elo),
disabling each feature in turn. This should help future improvements and
simplifications to pick suitable targets.

The list of tests is:
step  7 : http://tests.stockfishchess.org/tests/view/5abcbb4b0ebc5902926cf1ca
step  8 : http://tests.stockfishchess.org/tests/view/5abcbb680ebc5902926cf1cc
step  9 : http://tests.stockfishchess.org/tests/view/5abcbb850ebc5902926cf1ce
step 10 : http://tests.stockfishchess.org/tests/view/5abcbbeb0ebc5902926cf1d2
step 11 : http://tests.stockfishchess.org/tests/view/5abcbbbf0ebc5902926cf1d0
step 13 : http://tests.stockfishchess.org/tests/view/5abd03680ebc5902926cf20b
step 13a: http://tests.stockfishchess.org/tests/view/5abd29660ebc5902926cf22a
step 13b: http://tests.stockfishchess.org/tests/view/5abd29820ebc5902926cf22c
step 14 : http://tests.stockfishchess.org/tests/view/5abd03860ebc5902926cf20f
step 14a: http://tests.stockfishchess.org/tests/view/5abd2b6c0ebc5902926cf230
step 14b: http://tests.stockfishchess.org/tests/view/5abd2b8d0ebc5902926cf232
step 14c: http://tests.stockfishchess.org/tests/view/5abd2bad0ebc5902926cf234
step 14d: http://tests.stockfishchess.org/tests/view/5abd2bcf0ebc5902926cf236
step 14e: http://tests.stockfishchess.org/tests/view/5abd2bf10ebc5902926cf238

This patch documents this in the code.

Note that it will be a waste to recompute these estimates often, even a couple
of [0,5] patches are unlikely to change them by more than the error bars. The
interest of the Elo annotations in the code is not in the details, but in high-
lighting trends such as razoring (2 Elo) and singular extensions (60 Elo). These
estimates should be recomputed at most once a year.

Closes https://github.com/official-stockfish/Stockfish/pull/1522

No functional change.
2018-04-01 03:21:45 +02:00
mstembera 76f9cd4df1 Some small changes
1) Use make_bitboard() in Bitboards::init()

2) Fix MSVC warning: search.h(85): warning C4244: '=': conversion from
   'TimePoint' to 'int', possible loss of data.

Closes https://github.com/official-stockfish/Stockfish/pull/1524

No functional change.
2018-04-01 02:32:55 +02:00
Stéphane Nicolet d9cac9a414 Simplify KBPKB endgame with opposite bishops
When we reach a position with only two opposite colored bishops and
one pawn on the board, current master would give it a scale factor
of 9/64=0.14 in about one position out of 7200, and a scale factor
of 0.0 in the 7199 others. The patch gives a scale factor of 0.0 in
100% of the cases.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 55845 W: 11467 L: 11410 D: 32968
http://tests.stockfishchess.org/tests/view/5abc585f0ebc5902926cf15e

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 11915 W: 1852 L: 1719 D: 8344
http://tests.stockfishchess.org/tests/view/5abc7f750ebc5902926cf18c

We also have exhaustive coverage analysis of this patch effect by
Alain Savard, comparing the perfect evaluation given by the Syzygy
tablebase with the heuristic play after this patch for the set of
all legal positions of the KBPKP endgame with opposite bishops, in
the comments thread for this pull request:
https://github.com/official-stockfish/Stockfish/pull/1520

Alain's conclusion:
> According to this definition and the data, I consider this PR is
> identical to master to "solve for draw" and slightly better than
> master to solve earlier for "wins".

Note: this patch is a side effect of an ongoing effort to improve
the evaluation of positions involving a pair of opposite bishops.
See the GitHub diff of this LTC test which almost passed at sprt[0..5]
for a discussion:
http://tests.stockfishchess.org/tests/view/5ab9030b0ebc5902932cbf93

No functional change (at small bench depths)
2018-04-01 02:19:42 +02:00
Joost VandeVondele f2681232e5 Refine SEE threshold for capture pruning.
eloDoc suggests that this part of search is worth ~18 Elo.
This patch refines the depth dependence of the SEE threshold.

passed STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 21398 W: 4474 L: 4245 D: 12679
http://tests.stockfishchess.org/tests/view/5abfb0630ebc591a560aae07

passed LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 9028 W: 1439 L: 1285 D: 6304
http://tests.stockfishchess.org/tests/view/5abfbff30ebc591a560aae11

Closes https://github.com/official-stockfish/Stockfish/pull/1527

Bench: 6036915
2018-03-31 23:10:35 +02:00
Alain SAVARD d5e3e7d207 Candidate Passed Pawn
Include some not fully supported levers in the (candidate) passed pawns
bitboard, if otherwise unblocked. Maybe levers are usually very short
lived, and some inaccuracy in the lever balance for the definition of
candidate passed pawns just triggers a deeper search.

Here is a example of a case where the patch has an effect on the definition
of candidate passers: White c5/e5 pawns, against Black d6 pawn. Let's say
we want to test if e5 is a candidate passer. The previous master looks
only at files d, e and f (which is already very good) and reject e5 as
a candidate. However, the lever d6 is challenged by 2 pawns, so it should
not fully count. Indirectly, this patch will view such case (and a few more)
to be scored as candidates.

STC
http://tests.stockfishchess.org/tests/view/5abcd55d0ebc5902926cf1e1
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 16492 W: 3419 L: 3198 D: 9875

LTC
http://tests.stockfishchess.org/tests/view/5abce1360ebc5902926cf1e6
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21156 W: 3201 L: 2990 D: 14965

This was inspired by this test of Jerry Donald Watson, except the case of
zero supporting pawns against two levers is excluded, and it seems that
not excluding that case is bad, while excluding is it beneficial. See the
following tests on fishtest:

https://github.com/official-stockfish/Stockfish/pull/1519
http://tests.stockfishchess.org/tests/view/5abccd850ebc5902926cf1dd
http://tests.stockfishchess.org/tests/view/5abcdd490ebc5902926cf1e4

Closes https://github.com/official-stockfish/Stockfish/pull/1521

Bench: 5568461

----

Comments by Jerry Donald Watson:

> My thinking as to why this works:
>
> The evaluation is either called in an interior node or in the qsearch.
> The calls at the end of the qsearch are the more important as they
> ultimately determine the scoring of each move, whereas the internal
> values are mainly used for pruning decisions with a margin. Some strong
> engines don't even call the eval at all nodes. Now the whole point of
> the qsearch is to find quiet positions where captures do not change the
> evaluation of the position with regards to the search bounds - i.e. if
> there were good captures they would be tried.* So when a candidate lever
> appears in the evaluation at the end of the qsearch, the qsearch has
> guaranteed that it cannot just be captured, or if it can, this does not
> take the score past the search bounds. Practically this may mean that
> the side with the candidate lever has the turn, or perhaps the stopping
> lever pawn is pinned, or that side is forced for other reasons to make
> some other move (e.g. d6 can only take one of the pawns in the example
> above).
>
> Hence granting the full score for only one lever defender makes some
> sense, at least, to me.
>
> IMO this is also why huge bonuses for possible captures in the evaluation
> (e.g. threat on queen and our turn), etc. don't tend to work. Such things
> are best left to the search to figure out.
2018-03-31 01:16:51 +02:00
Ondrej Mosnáček c8ef80f466 Use per-thread dynamic contempt
We now use per-thread dynamic contempt. This patch has the following
effects:

 * for Threads=1: **non-functional**
 * for Threads>1:
   * with MultiPV=1: **no regression, little to no ELO gain**
   * with MultiPV>1: **clear improvement over master**

First, I tried testing at standard MultiPV=1 play with [0,5] bounds.
This yielded 2 yellow and 1 red test:

5+0.05, Threads=5:
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 82689 W: 16439 L: 16190 D: 50060
http://tests.stockfishchess.org/tests/view/5aa93a5a0ebc5902952892e6

5+0.05, Threads=8:
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 27164 W: 4974 L: 4983 D: 17207
http://tests.stockfishchess.org/tests/view/5ab2639b0ebc5902a6fbefd5

5+0.5, Threads=16:
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 41396 W: 7127 L: 7082 D: 27187
http://tests.stockfishchess.org/tests/view/5ab124220ebc59029516cb62

Then, I tested with Skill Level=17 (implicitly MutliPV=4), showing
a clear improvement:

5+0.05, Threads=5:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 3498 W: 1316 L: 1135 D: 1047
http://tests.stockfishchess.org/tests/view/5ab4b6580ebc5902932aeca2

Next, I tested the patch with MultiPV=1 again, this time checking for
non-regression ([-3, 1]):

5+0.5, Threads=5:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 65575 W: 12786 L: 12745 D: 40044
http://tests.stockfishchess.org/tests/view/5ab4e8500ebc5902932aecb3

Finally, I ran some tests with fixed number of games, checking if
reverting dynamic contempt gains more elo with Skill Level=17 (i.e.
MultiPV) than applying the "prevScore" fix and this patch. These tests
showed, that this patch gains 15 ELO when playing with Skill Level=17:

5+0.05, Threads=3, "revert dynamic contempt" vs. "WITHOUT this patch":
ELO: -11.43 +-4.1 (95%) LOS: 0.0%
Total: 20000 W: 7085 L: 7743 D: 5172
http://tests.stockfishchess.org/tests/view/5ab636450ebc590295d88536

5+0.05, Threads=3, "revert dynamic contempt" vs. "WITH this patch":
ELO: -26.42 +-4.1 (95%) LOS: 0.0%
Total: 20000 W: 6661 L: 8179 D: 5160
http://tests.stockfishchess.org/tests/view/5ab62e680ebc590295d88524

---
***FAQ***

**Why should this be commited?**
I believe that the gain for multi-thread MultiPV search is a sufficient
justification for this otherwise neutral change. I also believe this
implementation of dynamic contempt is more logical, although this may
be just my opinion.

**Why is per-thread contempt better at MultiPV?**
A likely explanation for the gain in MultiPV mode is that during
search each thread independently switches between rootMoves and via
the shared contempt score skews each other's evaluation.

**Why were the tests done with Skill Level=17?**
This was originally suggested by @Hanamuke and the idea is that with
Skill Level Stockfish sometimes plays also moves it thinks are slightly
sub-optimal and thus the quality of all moves offered by the MultiPV
search is checked by the test.

**Why are the ELO differences so huge?**
This is most likely because of the nature of Skill Level mode --
since it slower and weaker than normal mode, bugs in evaluation have
much greater effect.

---

Closes https://github.com/official-stockfish/Stockfish/pull/1515.

No functional change -- in single thread mode.
2018-03-30 10:48:57 +02:00
Joost VandeVondele c959871a59 Integrate syzygy in automated testing (v2).
Extends valgrind/sanitizer testing to cover syzygy code.

The script downloads 4 man syzygy as needed. The time needed for the
additional testing is small (in fact hard to see a difference compared
to the large fluctuations in testing time in travis).

Possible follow-ups:

 * include more TB sensitive positions in bench.
 * include the test script of recent commit "Refactor tbprobe.cpp".
 * verify unchanged bench with TB (with a long run).
 * make the TB part of the continuation integration tests optional.

Closes https://github.com/official-stockfish/Stockfish/pull/1518
and    https://github.com/official-stockfish/Stockfish/pull/1490

No functional change.
2018-03-30 10:23:48 +02:00
Jerry Donald Watson 9953bfff62 Change reduction/improving interaction
Adjust criterion for applying extra reduction if not improving.
We now add an extra ply of reduction if r > 1.0, instead of the
previous condition Reductions[NonPV][imp][d][mc] >= 2.

Why does this work? Previously, reductions when not improving had
a discontinuity as the depth and/or move count increases due to the
Reductions[NonPV][imp][d][mc] >= 2 condition. Hence, values of r
such that 0.5 < r < 1.5 would be mapped to a reduction of 1, while
1.5 < r < 2.5 would be mapped to a reduction of 3. This patch allows
values of r satisfying 1.0 < r < 1.5 to be mapped to a reduction of 2,
making the reduction formula more continuous.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 35908 W: 7382 L: 7087 D: 21439
http://tests.stockfishchess.org/tests/view/5aba723a0ebc5902a4743e8f

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 23087 W: 3584 L: 3378 D: 16125
http://tests.stockfishchess.org/tests/view/5aba89070ebc5902a4743ea9

Ideas for future work:
- We could look at retuning the LMR formula.
- We could look at adjusting the reductions in PV nodes if not improving.

Bench: 5326261
2018-03-28 00:31:47 +02:00
Stéphane Nicolet 5ce630584c Tweak queen values for midgame and endgame
Queen midgame value: -1%
Queen endgame value: +1%

This patch failed STC, but was exceptionally tested at LTC,
where it passed two independant SPRT tests with bounds [0..4]:

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 20651 W: 3191 L: 2980 D: 14480
http://tests.stockfishchess.org/tests/view/5ab73faa0ebc5902932cbdca

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 89111 W: 13647 L: 13233 D: 62231
http://tests.stockfishchess.org/tests/view/5ab8c0e50ebc5902932cbeae

See https://github.com/official-stockfish/Stockfish/pull/1511
for the discussion of this patch.

How to continue from there?
• there is probably some more Elo gains to get from the tuning
of the queen value.

Bench: 5460229
2018-03-27 17:51:11 +02:00
protonspring 29070bd01a Simplification: remove pawn shelter/storm masks
Encode the pawn shelter/storm masks into the danger score

This highly specialized rule directly contradicts the VERY high
danger score for blocked pawns. Reducing the danger score for
blocked pawns and removing this rule is apparently an effective
compromise.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 36597 W: 7522 L: 7429 D: 21646
http://tests.stockfishchess.org/tests/view/5ab935f30ebc5902932cc016

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 100690 W: 15373 L: 15372 D: 69945
http://tests.stockfishchess.org/tests/view/5ab9501f0ebc5902932cc042

Closes https://github.com/official-stockfish/Stockfish/pull/1514

Bench: 5980822
2018-03-27 17:30:52 +02:00
Ondrej Mosnáček 367304e930 Fix dynamic contempt for MultiPV
Use rootMoves[PVIdx].previousScore instead of bestValue for
dynamic contempt. This is equivalent for MultiPV=1 (bench remained the
same, even for higher depths), but more correct for MultiPV.

STC (MultiPV=3):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2657 W: 1079 L: 898 D: 680
http://tests.stockfishchess.org/tests/view/5aaa47cb0ebc590297330403

LTC (MultiPV=3):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2390 W: 874 L: 706 D: 810
http://tests.stockfishchess.org/tests/view/5aaa593a0ebc59029733040b

VLTC 240+2.4 (MultiPV=3):
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 2399 W: 861 L: 694 D: 844
http://tests.stockfishchess.org/tests/view/5aaf983e0ebc5902a182131f

LTC (MultiPV=4, Skill Level=17):
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 747 W: 333 L: 175 D: 239
http://tests.stockfishchess.org/tests/view/5aabccee0ebc5902997ff006

Note: although the ELO differences seem huge, they are inflated by the
nature of Skill Level / MultiPV search, so I don't think they can be
reasonably compared with classic ELO strength.

See https://github.com/official-stockfish/Stockfish/pull/1491 for some
verifications searches with MultiPV = 10 at depths 12 and 24 from the
starting position and the position after 1.e4, comparing the outputs
of the full PV by the old master and by this patch.

No functional change for MultiPV=1
2018-03-27 17:14:55 +02:00
Ondrej Mosnáček 8ff2fcf299 Refactor tbprobe.cpp
This involves:
 * replacing the union hacks with simply reusing the EntryPiece arrays
   for the no-pawns case
 * merging the PairsData structure with the EntryPiece/-Pawn structs
   (with credit to Marco: @mcostalba)
 * simplifying some HashTable functions
 * thanks to previous changes, removing the ugly memsets
 * simplifying the template logic for WDL/DTZ distinction
   (now we distinguish based on an enum type, not the entry classes)
 * removing the unneeded Atomic wrapper

-----------------------------

For reference, here is a manual way to check that patches concerning
table bases code are non-functional changes:

0) Download the Syzygy table bases (up to 6 men).
1) Make sure you have branches master and the pull request pointing to
   the right commits.
2) Download the bench calculation scripts from the following URL:

        https://gist.github.com/WOnder93/b5fcf9c989b4a1715684d5c82367cdbe

   and copy into src inside your Stockfish repo.
3) Make the scripts executable (chmod +x *.sh).
4) Run the following command to use TBs located at <path>:

       export SYZYGY_PATH='<path>'

5) After that, run this (it will take a long time, this is a deep bench):

       BENCH_ARGS='128 1 22' ./check_benches.sh master tbprobe_cleanup 2>/dev/null`

==> You should see two equal numbers printed.
    (Of course, now we have to trust that the script itself is correct :)

-----------------------------

Closes https://github.com/official-stockfish/Stockfish/pull/1477

No functional change.
2018-03-27 17:08:08 +02:00
Marco Costalba 280022baf7 Fix indentation in movepicker.cpp
Fix indentation in movepicker.cpp and evaluate.cpp. Also move
the PickType enum out of global space.

No functional change
2018-03-27 16:45:12 +02:00
Stéphane Nicolet a03e98dcd3 Switch time management to 64 bits
This is a patch to fix issue #1498, switching the time management variables
to 64 bits to avoid overflow of time variables after 25 days.

There was a bug in Stockfish 9 causing the output to be wrong after
2^31 milliseconds search. Here is a long run from the starting position:

info depth 64 seldepth 87 multipv 1 score cp 23 nodes 13928920239402
nps 0 tbhits 0 time -504995523 pv g1f3 d7d5 d2d4 g8f6 c2c4 d5c4 e2e3 e7e6 f1c4
c7c5 e1g1 b8c6 d4c5 d8d1 f1d1 f8c5 c4e2 e8g8 a2a3 c5e7 b2b4 f8d8 b1d2 b7b6 c1b2
c8b7 a1c1 a8c8 c1c2 c6e5 d1c1 c8c2 c1c2 e5f3 d2f3 a7a5 b4b5 e7c5 f3d4 d8c8 d4b3
c5d6 c2c8 b7c8 b3d2 c8b7 d2c4 d6c5 e2f3 b7d5 f3d5 e6d5 c4e5 a5a4 e5d3 f6e4 d3c5
e4c5 b2d4 c5e4 d4b6 e4d6 g2g4 d6b5 b6c5 b5c7 g1g2 c7e6 c5d6 g7g6

We check at compile time that the TimePoint type is exactly 64 bits long for
the compiler (TimePoint is our alias in Stockfish for std::chrono::milliseconds
-- it is a signed integer type of at least 45 bits according to the C++ standard,
but will most probably be implemented as a 64 bits signed integer on modern
compilers), and we use this TimePoint type consistently across the code.

Bug report by user "fischerandom" on the TCEC chat (thanks), and the
patch includes code and suggestions by user "WOnder93" and Ronald de Man.

Fixes issue:          https://github.com/official-stockfish/Stockfish/issues/1498
Closes pull request:  https://github.com/official-stockfish/Stockfish/pull/1510

No functional change.
2018-03-27 16:25:41 +02:00
Jerry Donald Watson f0f6da2d30 Make kingRing always 8 squares
Make kingRing always eight squares, extending the bitboard to the
F file if the king is on the H file, and to the C file if the king
is on the A file. This may deal with cases where Stockfish (like
many other engines) would shift the king around on the back rank
like g1h1, not because there is some imminent threat, but because
it makes king safety look a little better just because the king ring
had a smaller area.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 34000 W: 7167 L: 6877 D: 19956
http://tests.stockfishchess.org/tests/view/5ab8216d0ebc5902932cbe64

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22574 W: 3576 L: 3370 D: 15628
http://tests.stockfishchess.org/tests/view/5ab84e6a0ebc5902932cbe72

How to continue from there?

This patch probably makes it easier to tune the king safety evaluation,
because the new regularity of the king ring size will make the king
safety function more continuous.

Closes https://github.com/official-stockfish/Stockfish/pull/1512

Bench: 5934103
2018-03-26 09:28:37 +02:00
Chess13234 62937d1007 IID margin reduced to 128
Change the internal iterative deepening margin from +256 to +128.

Passed STC:
LLR: 3.26 (-2.94,2.94) [0.00,5.00]
Total: 22006 W: 4646 L: 4400 D: 12960
http://tests.stockfishchess.org/tests/view/5ab54a240ebc590295d8847d

Passed LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 46665 W: 7187 L: 6898 D: 32580
http://tests.stockfishchess.org/tests/view/5ab5ba690ebc590295d884b4

Possible work in this area after this change:

• Reduce even more the margin?
• Try to do IID for smaller depths too?

Bench: 5851102
2018-03-24 08:59:48 +01:00
Joost VandeVondele d40e7ee209 Join refutation stages in the movepicker
Unifies a bit further the three refuation stages in the MovePicker
class. Also treat the skipping of TT move now always via select_move(),
as discussed in pull request #1454.

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 16608 W: 3461 L: 3331 D: 9816
http://tests.stockfishchess.org/tests/view/5ab0aaf00ebc59029fb6f6c3

Closes https://github.com/official-stockfish/Stockfish/pull/1502

No functional change.
2018-03-21 01:48:47 +01:00
DU-jdto ed26d71354 Fix a few minor code style inconsistencies
No functional change.
2018-03-21 01:35:05 +01:00
Joost VandeVondele bd59560480 Join common code in the stages of next_move()
Rewrite the MovePicker class using lambda expressions for move filtering.
Includes code style changes by @mcostalba.

Verified for speed, passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 43191 W: 9391 L: 9312 D: 24488
http://tests.stockfishchess.org/tests/view/5a99b9df0ebc590297cc8f04

This rewrite of MovePicker.cpp seems to trigger less random crashes on Ryzen
machines than the version in previous master (reported by Bojun Guo).

Closes https://github.com/official-stockfish/Stockfish/pull/1454

No functional change.
2018-03-19 00:59:30 +01:00
Joost VandeVondele 1940485030 Simplify razoring logic
passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 36574 W: 7523 L: 7430 D: 21621
http://tests.stockfishchess.org/tests/view/5aae11450ebc590299abf52f

No functional change.
2018-03-19 00:17:08 +01:00
Ronald de Man 759b3c79cf Mark all compile-time constants as constexpr.
To more clearly distinguish them from "const" local variables, this patch
defines compile-time local constants as constexpr. This is consistent with
the definition of PvNode as constexpr in search() and qsearch(). It also
makes the code more robust, since the compiler will now check that those
constants are indeed compile-time constants.

We can go even one step further and define all the evaluation and search
compile-time constants as constexpr.

In generate_castling() I replaced "K" with "step", since K was incorrectly
capitalised (in the Chess960 case).

In timeman.cpp I had to make the non-local constants MaxRatio and StealRatio
constepxr, since otherwise gcc would complain when calculating TMaxRatio and
TStealRatio. (Strangely, I did not have to make Is64Bit constexpr even though
it is used in ucioption.cpp in the calculation of constexpr MaxHashMB.)

I have renamed PieceCount to pieceCount in material.h, since the values of
the array are not compile-time constants.

Some compile-time constants in tbprobe.cpp were overlooked. Sides and MaxFile
are not compile-time constants, so were renamed to sides and maxFile.

Non-functional change.
2018-03-18 23:48:16 +01:00
mstembera 350dff4464 Fix an MSVC 2017 error and warnings.
Improved version by @mcostalba.

No functional change.
2018-03-16 22:37:03 +01:00
Stéphane Nicolet 96362fe3df Renaming some variables in code
Implements renaming suggestions by Marco Costalba, Günther Demetz,
Gontran Lemaire, Ronald de Man, Stéphane Nicolet, Alain Savard,
Joost VandeVondele, Jerry Donald Watson, Mike Whiteley, xoto10,
and I hope that I haven't forgotten anybody.

Perpetual renaming thread for suggestions:
https://github.com/official-stockfish/Stockfish/issues/1426

No functional change.
2018-03-15 10:44:26 +01:00
Joost VandeVondele 8ab12c9012 Make using quiescence search implicit
If search depth is less than ONE_PLY call qsearch(), no need to check the
depth condition at various call sites of search().

Passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14568 W: 3011 L: 2877 D: 8680
http://tests.stockfishchess.org/tests/view/5aa846190ebc59029781015b

Also helps gcc to find some optimizations (smaller binary, some speedup).
Thanks to Aram and Stefan for identifying an oversight in an early version.

Closes https://github.com/official-stockfish/Stockfish/pull/1487

No functional change.
2018-03-15 09:25:59 +01:00
Stéphane Nicolet 8db75dd9ec Check the clock every 1024 nodes
This patch checks the clock every 1024, instead of 4096 in current master.

This is a step towards a solution to alleviate the problem reported by
Gian-Carlo Pascutto in pull request #1471, about a situation where SF
would sometimes lose on time in endgames at time control 1min+1sec when
using Syzygy EGTB on a spinning drive.

We made four tests called "Estimate the Elo cost of calling now() every
N nodes", with N in {256,1024} and time control in {STC,LTC}. Each test
was 40000 games with auto-purge off, against master branch.

http://tests.stockfishchess.org/tests/view/5aa4f37c0ebc59029780ff2d
http://tests.stockfishchess.org/tests/view/5aa4f3080ebc59029780ff2b
http://tests.stockfishchess.org/tests/view/5aa5cbb50ebc59029780ffce
http://tests.stockfishchess.org/tests/view/5aa5cb730ebc59029780ffcc

Here is a summary of the results (speed-up and Elo gain against master):

|   Time   |       N=256       |      N=1024      |
|:--------:|:-----------------:|:----------------:|
| Speed-up |    0.1% slower    |   0.2% faster    |
|   STC    |  -0.33 +-2.2 Elo  |  1.33 +-2.2 Elo  |
|   LTC    |   1.08 +-2.3 Elo  |  2.34 +-2.2 Elo  |

Based on these results, we tested N=1024 as a non-regression against
master at LTC:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 80746 W: 12430 L: 12399 D: 55917
http://tests.stockfishchess.org/tests/view/5aa7ba660ebc590297810116

Closes https://github.com/official-stockfish/Stockfish/pull/1482

No functional change.
2018-03-14 22:01:58 +01:00
lucasart edf4c07d25 Use intrinsics only for LSB/MSB
The NO_BSF does not cover any real life use-case today. The only compilers that
can compile SF today, with the current Makefile and no source code changes, are
either GCC compatible (define __GNUC__) or MSVC compatible (define _MSC_VER). So
they all support LSB/MSB intrinsics.

This patch simplifies away the software fall-backs of LSB/MSB that were still
in Stockfish code, but unused in any of the officially supported compilers.

Note the (legacy) MSVC/WIN32 case, where we use a 32-bit BSF/BSR solution, as
64-bit intrinsics aren't available there.

Discussed in: https://github.com/official-stockfish/Stockfish/pull/1447
and:          https://github.com/official-stockfish/Stockfish/pull/1479

No functional change.
2018-03-14 01:31:21 +01:00
Jerry Donald Watson b605103a34 Simplify condition for ProbCut move count pruning
We retire depth-dependence in ProbCut move count pruning,
the move count limit condition is now : "probCutCount < 3"

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 33895 W: 6995 L: 6897 D: 20003
http://tests.stockfishchess.org/tests/view/5aa6eaba0ebc59029781009d

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 28959 W: 4441 L: 4333 D: 20185
http://tests.stockfishchess.org/tests/view/5aa73dfa0ebc5902978100be

Ideas for future work:

• Is a flat move count limit in ProbCut ideal? Depth dependence, or
  dependence on some other variable, could possibly be reintroduced.

• The move count limit 3 is untuned and a better value may exist.

Closes https://github.com/official-stockfish/Stockfish/pull/1486

Bench: 5741807
2018-03-13 23:32:45 +01:00
Jerry Donald Watson c5f6bd517c Tweak probcut margin with 'improving' flag
Adjust ProbCut rBeta by whether the score is improving, and also
set improving to false when in check. More precisely, this patch
has two parts:

1) the increased beta threshold for ProbCut is now adjusted based
   on whether the score is improving
2) when in check, improving is always set to false.

Co-authored by Joost VandeVondele (@vondele) and Bill Henry (@VoyagerOne).

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 13480 W: 2840 L: 2648 D: 7992
http://tests.stockfishchess.org/tests/view/5aa693fe0ebc59029781004c

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 25895 W: 4099 L: 3880 D: 17916
http://tests.stockfishchess.org/tests/view/5aa6ac940ebc59029781006e

In terms of opportunities for future work opened up by this patch,
the ProbCut rBeta formula could probably be tuned to gain more Elo.

Closes https://github.com/official-stockfish/Stockfish/pull/1485

Bench: 5328254
2018-03-13 23:21:46 +01:00
VoyagerOne b2961aded6 Use quiescence search for Probcut
Perform qsearch for the preliminary search in Probcut

Passed STC with sprt[-3..1] bounds:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 31090 W: 6386 L: 6283 D: 18421
http://tests.stockfishchess.org/tests/view/5aa598ed0ebc59029780ff9f

Passed LTC with sprt[0..4] bounds:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 104056 W: 15990 L: 15531 D: 72535
http://tests.stockfishchess.org/tests/view/5aa5b0f30ebc59029780ffa9

Closes https://github.com/official-stockfish/Stockfish/pull/1483

Bench: 5404567
2018-03-13 08:21:11 +01:00
Joost VandeVondele efe702e9f5 qsearch(): remove inCheck as a template parameter
Simplifies a bit, and avoids bugs as in #1478

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 104862 W: 21302 L: 21337 D: 62223
http://tests.stockfishchess.org/tests/view/5aa6de1b0ebc590297810097

Closes https://github.com/official-stockfish/Stockfish/pull/1484

No functional change
2018-03-13 08:13:26 +01:00
Jerry Donald Watson 840605c14e Adjust initiative in pure pawn endgames
King and pawn endgames are typically decisive, and a small
advantage is often sufficient to win. Therefore we now take
this into account when computing the initiative adjustment.

This idea came from a series of patches by Gian-Carlo Pascutto.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 48770 W: 10203 L: 9845 D: 28722
http://tests.stockfishchess.org/tests/view/5aa58cce0ebc59029780ff8d

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22252 W: 3572 L: 3366 D: 15314
http://tests.stockfishchess.org/tests/view/5aa5b27c0ebc59029780ffad

Ideas for future developement:

- There have been a number of changes to the initiative
  calculation lately. Perhaps the coefficients could be
  tuned again.

- It may be possible to add special knowledge for other
  endgames in the initiative calculation.

Closes https://github.com/official-stockfish/Stockfish/pull/1481

Bench: 5750110
2018-03-12 22:57:21 +01:00
protonspring 6e9337b107 MovePicker: combine countermove with killers.
Handle the countermove in the same way we use stages to progress
through the killer moves, using a common array called "refutations".
Removes some lines of code and simplifies a bit the jump table.

STC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 71707 W: 14622 L: 14595 D: 42490
http://tests.stockfishchess.org/tests/view/5aa003cf0ebc590297cb6276

LTC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 22320 W: 3470 L: 3352 D: 15498
http://tests.stockfishchess.org/tests/view/5aa051020ebc590297cb62ba

Closes https://github.com/official-stockfish/Stockfish/pull/1468

No functional change.
2018-03-12 02:49:14 +01:00
protonspring c3af52c43b Connectivity Score S(3,1) is stronger than S(2,2)
I believe my tests were conclusive enough to demonstrate
that a connectivity score of S(3,1) is stronger than S(2,2).

STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 16175 W: 3386 L: 3165 D: 9624
http://tests.stockfishchess.org/tests/view/5aa48b150ebc59029780fef6

LTC
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 98685 W: 15209 L: 14765 D: 68711
http://tests.stockfishchess.org/tests/view/5aa496f50ebc59029780fefa

Bench 5601228
2018-03-12 01:52:12 +01:00
protonspring 5dc381a566 Implement an old Russian proverb
"Loose pieces drop, in blitz keep everything protected"

Adding a small S(2,2) bonus for knights, bishops, rooks, and
queens that are "connected" to each other (in the sense that
they are under attack by our own pieces) apparently is a good
thing. It probably helps the pieces work together a bit better.

STC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 12317 W: 2655 L: 2467 D: 7195
http://tests.stockfishchess.org/tests/view/5aa2d86b0ebc590297cb6474

LTC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 35725 W: 5516 L: 5263 D: 24946
http://tests.stockfishchess.org/tests/view/5aa2fc6f0ebc590297cb64a8

How to continue from there (by Stefan Geschwentner)?

• First we should identify all other eval terms which have an overlap
  with new connectivity bonus (like the outpost bonus). A simple way
  would be subtract the connectivity bonus from them and look if this
  better, or use a SPSA session for these terms.

• Tuning Connectivity himself with SPSA seems not so promising because
  of the small range which is useful. Here manual testing changes of
  Connectivity like +-1 seems better.

• The eg value is more important because in endgame the position gets
  more open and so attacks on pieces are easier. Another important point
  is that when defending/fortress-like positions each defending piece
  needs a protection, otherwise attacks on them can break defense.

Closes https://github.com/official-stockfish/Stockfish/pull/1474

Bench: 5318575
2018-03-10 12:04:03 +01:00
Joost VandeVondele 2c5dfb3122 Assign improving only once
Avoid duplicated code after recent commit "Use evaluation trend
to adjust futility margin". We initialize the improving variable
to true in the check case, which allows to avoid redundant code
in the general case.

Tested for speed by snicolet, patch seems about 0.4% faster.

No functional change.

Note: initializing the improving variable to false in the check
case was tested as a functional change, ending yellow in both STC
and LTC. This change is not included in the commit, but it is an
interesting result that could become part of a future patch about
improving or LMR. Reference of the LTC yellow test:
http://tests.stockfishchess.org/tests/view/5aa131560ebc590297cb636e
2018-03-10 11:06:53 +01:00
Alain SAVARD 1093047e7d Two steps slider threats on queen
Allow a potential slider threat from a square currently occupied
by a harmless attacker, just as the recent "knight on queen" patch.
Also from not completely safe squares, use the mobilityArea instead
of excluding all pawns for both SlidersOnQueen and KnightOnQueen

We now compute the potential sliders threat on queen only if opponent
has one queen.

Run as SPRT [0,4] since it is some kind of simplification but maybe
not clearly one.

STC:
http://tests.stockfishchess.org/tests/view/5aa1ddf10ebc590297cb63d8
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 22997 W: 4817 L: 4570 D: 13610

LTC:
http://tests.stockfishchess.org/tests/view/5aa1fe6b0ebc590297cb63e5
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 11926 W: 1891 L: 1705 D: 8330

After this patch is committed, we may try to:

• re-introduce some "threat by queen" bonus to make Stockfish's queen
  more aggressive (attacking aspect)

• introduce a concept of "queen overload" to force the opponent queen
  into passivity and protecting duties (defensive aspect)

• more generally, re-tune the queen mobility array since patches in the
  last three months have affected a lot the location/activity of queens.

Closes https://github.com/official-stockfish/Stockfish/pull/1473

bench: 5788691
2018-03-09 21:47:44 +01:00
IIvec 82697f1193 Simplification: use Arctan for the optimism S-curve
This was an idea of Gontran Lemaire (gonlem), but the graphs
he published did not seem accurate to me. I did my own graphs,
got my own constants and here is the result:

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 42179 W: 8704 L: 8622 D: 24853
http://tests.stockfishchess.org/tests/view/5a9db6270ebc590297cb611b

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 17737 W: 2702 L: 2577 D: 12458
http://tests.stockfishchess.org/tests/view/5a9eb5b20ebc590297cb61b2

Note: we are now back to 70% draw rate in selfplay mode!

Bench: 5544908
2018-03-07 22:51:31 +01:00
Jerry Donald Watson 65c3bb8586 Use evaluation trend to adjust futility margin
Adjust futility margin in the child node based on whether the
static evaluation is improving.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 15271 W: 3157 L: 2958 D: 9156
http://tests.stockfishchess.org/tests/view/5a9f2f8c0ebc590297cb6216

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 6617 W: 1053 L: 908 D: 4656
http://tests.stockfishchess.org/tests/view/5a9f98390ebc590297cb6241

Ideas for future work:

- Tune the new margins.
- Try to get this idea to work for futility pruning in parent
  nodes as well.

Bench: 5779242
2018-03-07 22:34:49 +01:00
Stéphane Nicolet 6d8f583af2 Knight threats on Queen
We give a S(21,11) bonus for knight threats on the next moves
against enemy queen. The threats are from squares which are
"not strongly protected" and which may be empty, contain enemy
pieces or even one of our piece at the moment (N,B,Q,R) -- hence
be two-steps threats in the later case because we will have to
move our piece and *then* attack the enemy queen with the knight.

STC: http://tests.stockfishchess.org/tests/view/5a9e442e0ebc590297cb6162
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 35129 W: 7346 L: 7052 D: 20731

LTC: http://tests.stockfishchess.org/tests/view/5a9e6e620ebc590297cb617f
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 42442 W: 6695 L: 6414 D: 29333

How to continue from there?

• Trying to refine the threat condition ("not strongly protected")
• Trying the two-steps idea for bishops or rooks threats against queen

Bench: 6051247
2018-03-07 22:12:29 +01:00
Torsten Franz 56a104e2e0 Remove popcount trick from space evaluation
Similar removal of superposition code trick as in the
"Simplify tropism computation" patch. This simplification
of the space() function will allow us to specify space
masks which can reach into enemy territory.

passed STC:
LLR: 3.38 (-2.94,2.94) [-3.00,1.00]
Total: 184630 W: 40581 L: 40758 D: 103291
http://tests.stockfishchess.org/tests/view/5a8433360ebc590297cc80c5

passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 231799 W: 37647 L: 37858 D: 156294
http://tests.stockfishchess.org/tests/view/5a96a34a0ebc590297cc8cfd

No functional change.
2018-03-07 21:58:16 +01:00
Stéphane Nicolet d42e6338ea Fix a warning in GCC for Windows
No functional change
2018-03-06 22:11:45 +01:00
joergoster 43682d08f7 Code style in Razoring and ProbCut
No functional change.
2018-03-06 01:54:36 +01:00
Chris Cain 3192b09fe0 Introduce variadic make_bitboard()
Adds a helper function to make a bitboard from a list of squares.

No functional change
2018-03-06 01:33:00 +01:00
Stefano Cardanobile 450f04969c Using a S-curve for the optimism measure
Add a logarithmic term in the optimism computation, increase
the maximal optimism and lower the contempt offset.

This increases the dynamics of the optimism aspects, giving
a boost for balanced positions without skewing too much on
unbalanced positions (but this version will enter panic mode
faster than previous master when behind, trying to draw faster
when slightly behind). This helps, since optimism is in general
a good thing, for instance at LTC, but too high optimism
rapidly contaminates play.

passed STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 159343 W: 34489 L: 33588 D: 91266
http://tests.stockfishchess.org/tests/view/5a8db9340ebc590297cc85b6

passed LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 47491 W: 7825 L: 7517 D: 32149
http://tests.stockfishchess.org/tests/view/5a9456a80ebc590297cc8a89

It must be mentioned that a version of the PR with contempt 0
did not pass STC [0,5]. The version in the patch, which uses
default contempt 12, was found to be as strong as current master
on different matches against SF7 and SF8, both at STC and LTC.

One drawback maybe is that it raises the draw rate in self-play
from 56% to 59%, giving a little bit less sensitivity for SF
developpers to find evaluation improvements by selfplay tests
in fishtest.

Possible further work:

• tune the values accurately, while keeping in mind the drawrate issue
• check whether it is possible to remove linear and offset term
• try to simplify the S-shape curve

Bench: 5934644
2018-03-04 16:55:58 +01:00
Joost VandeVondele cad300cfab No extra stage for QS recaptures.
remove the QRECAPTURES and QSEARCH_RECAPTURES stages as they can be
incoorporated in QCAPTURES stage with a simple condition on depth.

passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 214964 W: 46188 L: 46443 D: 122333
http://tests.stockfishchess.org/tests/view/5a98544b0ebc590297cc8e35

passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14552 W: 2404 L: 2274 D: 9874
http://tests.stockfishchess.org/tests/view/5a9947b20ebc590297cc8e93

Bench: 6361582
2018-03-03 12:41:35 +01:00
VoyagerOne b87308692a Do move-count pruning in probcut
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 4928 W: 1163 L: 1007 D: 2758

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 20368 W: 3441 L: 3238 D: 13689
http://tests.stockfishchess.org/tests/view/5a98bea40ebc590297cc8e5d

Bench: 6361568
2018-03-03 12:04:25 +01:00
Marco Costalba f35e52f030 Merge Stats tables
Use a recursive std::array with variadic template
parameters to get rid of the last redundacy.

The first template T parameter is the base type of
the array, the W parameter is the weight applied to
the bonuses when we update values with the << operator,
the D parameter limits the range of updates (range is
[-W * D, W * D]), and the last parameters (Size and
Sizes) encode the dimensions of the array.

This allows greater flexibility because we can now tweak
the range [-W * D, W * D] for each table.

Patch removes more lines than what adds and streamlines
the Stats soup in movepick.h

Closes PR#1422 and PR#1421

No functional change.
2018-03-03 11:35:33 +01:00
Stéphane Nicolet 94b3cdd908 Better indentation in Makefile
No functional change
2018-03-03 11:07:23 +01:00
Tom Vijlbrief 94abc2a0cf Reintroduce depth 2 razoring (with additional margin)
The first depth 2 margin triggers the verification quiescence search.
This qsearch() result has to be better then the second lower margin,
so we only skip the razoring when the qsearch gives a significant
improvement.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 32133 W: 7395 L: 7101 D: 17637
http://tests.stockfishchess.org/tests/view/5a93198b0ebc590297cc8942

Passed LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 17382 W: 3002 L: 2809 D: 11571
http://tests.stockfishchess.org/tests/view/5a93b18c0ebc590297cc89c2

This Elo-gaining version was further simplified following a suggestion
of Marco Costalba:

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 15553 W: 3505 L: 3371 D: 8677
http://tests.stockfishchess.org/tests/view/5a964be90ebc590297cc8cc4

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 13253 W: 2270 L: 2137 D: 8846
http://tests.stockfishchess.org/tests/view/5a9658880ebc590297cc8cca

How to continue after this patch?

Reformating the razoring code (step 7 in search()) to unify the
depth 1 and depth 2 treatements seems quite possible, this could
possibly lead to more simplifications.

Bench: 5765806
2018-02-28 13:42:32 +01:00
Stefan Geschwentner 59d10374ca Blocked pawn storm
In pawn structures like white pawns f6,h6 against black pawns f7,g6,h7
the attack on the king is blocked by the own pawns. So decrease the
penalty for king safety.

See diagram and discussion in
https://github.com/official-stockfish/Stockfish/pull/1434

A sample position that this patch wants to avoid is the following
1rr2bk1/3q1p1p/2n1bPpP/pp1pP3/2pP4/P1P1B3/1PBQN1P1/1K3R1R w - - 0 1

White pawn storm on the king side was a disaster, it locked the king
side completely. Therefore, all the king tropism bonus that white have
on the king side are useless, and kingadjacent attacks too. Master
gives White a static +4.5 advantage, but White cannot win that game.
The patch is lowering this evaluation artefact.

STC:
LLR: 2.94 (-2.94,2.94) [0.00,5.00]
Total: 16467 W: 3750 L: 3537 D: 9180
http://tests.stockfishchess.org/tests/view/5a92102d0ebc590297cc87d0

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 64242 W: 11130 L: 10745 D: 42367
http://tests.stockfishchess.org/tests/view/5a923dc80ebc590297cc8806

This version includes reformatting and speed optimization by Alain Savard.

Bench: 5643527
2018-02-28 13:03:35 +01:00
Leonid Pechenik ad5d86c771 Tweak time management
Using a SPSA tuning session to optimize the time management
parameters.

With SPSA tuning it is not always possible to say where improvements
came from. Maybe some variables changed randomly or because result
was not sensitive enough to them. So my explanation of changes will
not be necessarily correct, but here it is.

• When decrease of thinking time was added by Joost a few months ago
if best move has not changed for several plies, one more competing
indicator  was introduced for the same purpose along with increase
in score and absence of fail low at root. It seems that tuning put
relatively more importance on that new indicator what allowed to save
time.
• Some of this saved time is distributed proportionally between all
moves and some more time were given to moves when score dropped a lot
or best move changed.
• It looks also that SPSA redistributed more time from the beginning to
later stages of game via other changes in variables - maybe because
contempt made game to last longer or for whatever reason.

All of this is just small tweaks here and there (a few percentages changes).

STC (10+0.1):
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 18970 W: 4268 L: 4029 D: 10673
http://tests.stockfishchess.org/tests/view/5a9291a40ebc590297cc8881

LTC (60+0.6):
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 72027 W: 12263 L: 11878 D: 47886
http://tests.stockfishchess.org/tests/view/5a92d7510ebc590297cc88ef

Additional non-regression tests at other time controls

Sudden death 60s:
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 14444 W: 2715 L: 2608 D: 9121
http://tests.stockfishchess.org/tests/view/5a9445850ebc590297cc8a65

40 moves repeating at LTC:
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 10309 W: 1880 L: 1759 D: 6670
http://tests.stockfishchess.org/tests/view/5a9566ec0ebc590297cc8be1

This is a functional patch only for time management, but the bench
does not reflect this because it uses fixed depth search, so the number
of nodes does not change during bench.

No functional change.
2018-02-28 12:37:20 +01:00
Stéphane Nicolet de642f16db Simplify tropism computation
Simplification. Tests show that the "shift-and-superpose" trick is no longer
necessary. The speed benefit of avoiding a popcount is no longer relevant
on modern machines.

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41675 W: 9168 L: 9086 D: 23421
http://tests.stockfishchess.org/tests/view/5a840bcc0ebc590297cc80b5

Passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 117728 W: 19875 L: 19911 D: 77942
http://tests.stockfishchess.org/tests/view/5a8444800ebc590297cc80ca

No functional change.
2018-02-27 19:10:40 +01:00
Stefan Geschwentner cccbecb6f8 Stat score initialization: grandchildren
This is the sequel of the previous patch, we now let the parent node initialize
stat score to zero once for all grandchildren.

Initialize statScore to zero for the grandchildren of the current position.
So statScore is shared between all grandchildren and only the first grandchild
starts with statScore = 0. Later grandchildren start with the last calculated
statScore of the previous grandchild. This influences the reduction rules in
LMR which are based on the statScore of parent position.

Tests results against the previous patch:

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 23676 W: 5417 L: 5157 D: 13102
http://tests.stockfishchess.org/tests/view/5a9423a90ebc590297cc8a46

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 35485 W: 6168 L: 5898 D: 23419
http://tests.stockfishchess.org/tests/view/5a9435550ebc590297cc8a54

Bench: 5643520
2018-02-27 18:42:59 +01:00
Stefan Geschwentner 1463881153 Stat score initialization: children
Let the parent node initialize stat score to zero once for all siblings.

Initialize statScore to zero for the children of the current position.
So statScore is shared between sibling positions and only the first sibling
starts with statScore = 0. Later siblings start with the last calculated
statScore of the previous sibling. This influences the reduction rules in
in LMR which are based on the statScore of parent position.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 22683 W: 5202 L: 4946 D: 12535
http://tests.stockfishchess.org/tests/view/5a93315f0ebc590297cc894f

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 48548 W: 8346 L: 8035 D: 32167
http://tests.stockfishchess.org/tests/view/5a933ba90ebc590297cc8962

Bench: 5833683
2018-02-27 18:18:56 +01:00
Stéphane Nicolet 29bc128384 No Tempo for draw scores given by heuristic functions
The current master applies Eval::Tempo even to leaves evaluated
as draw by some of the static evaluation functions of endgame.cpp
(for instance KNN vs K or stalemates in KP vs K). This results in
some lines being reported as +0.07 or -0.07 when the terminal
position has reached such endgames (0.07 being about the value
of a tempo for Stockfish).

This patch does not apply Eval::tempo to these positions. This leads
to more nodes being evaluated as VALUE_DRAW during search, giving more
opportunities for cut-offs in alpha-beta.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 52602 W: 11776 L: 11403 D: 29423
http://tests.stockfishchess.org/tests/view/5a8cb8f60ebc590297cc8546

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,4.00]
Total: 156613 W: 26820 L: 26158 D: 103635
http://tests.stockfishchess.org/tests/view/5a8f452d0ebc590297cc865a

Bench: 4924749
2018-02-27 08:06:46 +01:00
Marco Costalba ad2a0e356e Speedup and simplify pinners and blockers
To compute dicovered check or pinned pieces we use some bitwise
operators that are not really needed because already accounted for
at the caller site.

For instance in evaluation we compute:

     pos.pinned_pieces(Us) & s

Where pinned_pieces() is:

     st->blockersForKing[c] & pieces(c)

So in this case the & operator with pieces(c) is useless,
given the outer '& s'.

There are many places where we can use the naked blockersForKing[]
instead of the full pinned_pieces() or discovered_check_candidates().

This path is simpler than original and gives around 1% speed up for me.
Also tested for speed by mstembera and snicolet (neutral in both cases).

No functional change.
2018-02-27 01:19:06 +01:00
erbsenzaehler d438720a1c Unify use of -mdynamic-no-pic
Apply -mdynamic-no-pic in a single place in the Makefile instead of 5 places.

Verified on three different Macs:
- a MacBook from 2013
- a MacBook running MacOS 10.9.5
- an iMac running MacOS 10.13.3

No functional change.
2018-02-27 00:30:47 +01:00
AndyGrant 71cc01c2ef Shallow search to verify probcut
Perform a preliminary shallow search to verify a probcut before doing
the normal "depth - 4 plies" search.

STC:
LLR: 4.73 (-2.94,2.94) [0.00,5.00]
Total: 36281 W: 8221 L: 7830 D: 20230
http://tests.stockfishchess.org/tests/view/5a921cb90ebc590297cc87f6

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 22907 W: 3954 L: 3738 D: 15215
http://tests.stockfishchess.org/tests/view/5a92672b0ebc590297cc8814

Happy to see something from Ethereal work for Stockfish :)

Bench: 5882274
2018-02-26 03:01:46 +01:00
Stéphane Nicolet 7a03450bd9 Count passed pawns in asymmetry measure
The previous asymmetry measure of the pawn structure only used to
consider the number of pawns on semi-opened files in the position.
With this patch we also increase the measure by the number of passed
pawns for both players.

Many thanks to the community for the nice feedback on the previous
version, with special mentions to Alain Savard and Marco Costalba
for clarity and speed suggestions.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 13146 W: 3038 L: 2840 D: 7268
http://tests.stockfishchess.org/tests/view/5a91dd0c0ebc590297cc877e

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 27776 W: 4771 L: 4536 D: 18469
http://tests.stockfishchess.org/tests/view/5a91fdd50ebc590297cc879b

How to continue after this patch?

Stockfish will now evaluate more positions with passed pawns, so
tuning the passed pawns values may bring Elo. The patch has also
consequences on the initiative term, where we might want to give
different weights to passed pawns and semi-openfiles (idea by
Stefano Cardanobile).

Bench: 5302866
2018-02-26 01:06:45 +01:00
Joost VandeVondele ebb3e7df65 Combine killer moves
Move the first killer move out of the capture stage, combining treatment
of first and second killer move.

passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 55777 W: 12367 L: 12313 D: 31097
http://tests.stockfishchess.org/tests/view/5a88617e0ebc590297cc8351

Similar to an earlier proposition of Günther Demetz, see pull request #1075.
I think it is more robust and readable than master, why hand-unroll the loop
over the killer array, and duplicate code ?

This version includes review comments from Marco Costalba.

Bench: 5227124
2018-02-25 23:11:56 +01:00
Stéphane Nicolet 2ec36f8ae8 Revert "Count passed pawns in asymmetry measure" 2018-02-25 15:12:19 +01:00
Stéphane Nicolet 2e21aba8d9 Count passed pawns in asymmetry measure
The previous asymmetry measure of the pawn structure only used to
consider the number of pawns on semi-opened files in the postions.
With this patch we also increase the measure by the number of passed
pawns for both players.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 13146 W: 3038 L: 2840 D: 7268
http://tests.stockfishchess.org/tests/view/5a91dd0c0ebc590297cc877e

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 27776 W: 4771 L: 4536 D: 18469
http://tests.stockfishchess.org/tests/view/5a91fdd50ebc590297cc879b

How to continue from there: Stockfish will now evaluate more positions
with passed pawns, so tuning the passed pawns values may bring Elo.
The patch also has consequences on the initiative term.

Bench: 5302866
2018-02-25 13:12:23 +01:00
Joost VandeVondele 8dd6875240 Join all capture init stages in MovePicker
Passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16789 W: 3685 L: 3554 D: 9550
http://tests.stockfishchess.org/tests/view/5a91a8bb0ebc590297cc875b

Passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21293 W: 3527 L: 3407 D: 14359
http://tests.stockfishchess.org/tests/view/5a920a730ebc590297cc87ba

No functional change
2018-02-25 12:49:51 +01:00
DU-jdto 16b31bb249 More robust interaction of singular search and iid
When iid (Internal iterative deepening) is invoked, the prior value of ttValue is
not guaranteed to be VALUE_NONE. As such, it is currently possible to enter a state
in which ttValue has a specific value which is inconsistent with tte->bound() and
tte->depth(). Currently, ttValue is only used within the search in a context that
prevents this situation from making a difference (and so this change is non-functional,
but this is not guaranteed to remain the case in the future.

For instance, just changing the tt depth condition in singular extension node to be

    tte->depth() >= depth - 4 * ONE_PLY

instead of

    tte->depth() >= depth - 3 * ONE_PLY

interacts badly with the absence of ttMove in iid. For the ttMove to become a singular
extension candidate, singularExtensionNode needs to be true. With the current master,
this requires that tte->depth() >= depth - 3 * ONE_PLY. This is not currently possible
if tte comes from IID, since the depth 'd' used for the IID search is always less than
depth - 4 * ONE_PLY for depth >= 8 * ONE_PLY (below depth 8 singularExtensionNode can
never be true anyway). However, with DU-jdto/Stockfish@251281a , this condition can be
met, and it is possible for singularExtensionNode to become true after IID. There are
then two mechanisms by which this patch can affect the search:

• If ttValue was VALUE_NONE prior to IID, the fact that this patch sets ttValue allows
  the 'ttValue != VALUE_NONE' condition of singularExtensionNode to be met.

• If ttValue wasn't VALUE_NONE prior to IID, the fact that this patch modifies ttValue's
  value causes a different 'rBeta' to be calculated if the singular extension search is
  performed.

Tested at STC for non-regression:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 76981 W: 17060 L: 17048 D: 42873
http://tests.stockfishchess.org/tests/view/5a7738b70ebc5902971a9868

No functional change
2018-02-25 01:15:38 +01:00
DU-jdto 5d57bb467a Simplification: do razoring only for depth 1
The razoring heuristic is quite a drastic pruning technique,
using a depth 0 search at internal nodes of the search tree
to estimate the true value of depth n nodes. This patch limits
this razoring to the case of internal nodes of depth 1.
Author: Jarrod Torriero (DU-jdto)

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8043 W: 1865 L: 1716 D: 4462
http://tests.stockfishchess.org/tests/view/5a90a9290ebc590297cc86c1

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32890 W: 5577 L: 5476 D: 21837
http://tests.stockfishchess.org/tests/view/5a90c8510ebc590297cc86d5

Opportunities opened by this patch: it would be interesting to
know if it brings Elo to re-introduce razoring or soft razoring
at depth >= 2, maybe using a larger margin to compensate for the
increased pruning effect.

Bench: 5227124
2018-02-24 13:12:04 +01:00
Tom Vijlbrief 9246e4a6f9 Lower razor depth to < 3 and adjust margin
Various margins were tested: 600, 560, 585, 580, 590 and 595.

Only 590 (this patch) passed both STC and LTC.
Higher margins appear to be better for longer time controls.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 24496 W: 5470 L: 5210 D: 13816
http://tests.stockfishchess.org/tests/view/5a8c6d040ebc590297cc8508

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 74540 W: 12888 L: 12491 D: 49161
http://tests.stockfishchess.org/tests/view/5a8d14c70ebc590297cc8566

Bench: 5475941
2018-02-23 22:13:11 +01:00
Marco Costalba a09eee5798 Reformat SEE to better document the function
This is one of the most difficult to understand but also
most important and speed critical functions of SF.

This patch rewrites some part of it to hopefully
make it clearer and drop some redundant variables
in the process.

Same speed than master (or even a bit more).

Thanks to Chris Cain for useful feedback.

No functional change.
2018-02-23 22:02:44 +01:00
Stéphane Nicolet 52f92d05a9 Move pawn_attacks_bb() helper to bitboard.h
No functional change.
2018-02-21 22:31:38 +01:00
Mike Whiteley 820c5c25b6 rename shift variables.
Where variable names are explicitly incorrect, I feel morally obligated to at least
suggest an alternative. There are many, but these two are especially egregious.

No functional change.
2018-02-21 21:47:04 +01:00
Alain SAVARD 6c898a10be Avoid a compilation warning
Avoid a warning while compiling with gcc version 4.9.2

No functional change.
2018-02-21 00:52:59 +01:00
Marco Costalba 67f5f54a29 Code style in evaluate.cpp
Passed STC
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 75666 W: 16482 L: 16616 D: 42568
http://tests.stockfishchess.org/tests/view/5a8953af0ebc590297cc83ab

No functional change.
2018-02-20 17:11:18 +01:00
Stéphane Nicolet b2b0013d11 Update list of authors after Stockfish 9
No functional change.
2018-02-18 01:51:35 +01:00
Mike Whiteley 80ea80e451 Simplify trapped rook
As far as can tell, semiopenFiles are set if there is a pawn anywhere on
the file. The removed condition would be true even if the pawns were very
advanced, which doesn't make sense if we're looking for a trapped rook.
Seems the engine fairs better with this removed. My guess s that the
condition that mobility is 3 or less does this well enough.

Begs the question whether this is a mobility issue alone... not sure.
Should I do LTC test?

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 13377 W: 3009 L: 2871 D: 7497
http://tests.stockfishchess.org/tests/view/5a855be40ebc590297cc8166

Passed LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 16288 W: 2813 L: 2685 D: 10790
http://tests.stockfishchess.org/tests/view/5a8575a80ebc590297cc817e

Bench: 5006365
2018-02-15 19:38:09 +01:00
Ronald de Man 860223c5e6 Fix gcc PGO build on Windows
This fixes the issue #1375 of the PGO builds failing under Windows:
https://github.com/official-stockfish/Stockfish/issues/1375

Solution found during this discussion in the fishcooking forum:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/RjIPgeFFLPQ

Closes #1408.

No functional change.
2018-02-12 23:14:33 +01:00
Marco Costalba 4c57cf0ead Code style fixes in search.cpp
Some code style triviality.

No functional change.
2018-02-12 22:58:25 +01:00
Marco Costalba baab8be324 Update travis CI to g++7
Use newer g++ 7 instead of 6 in travis CI tests.

No functional change.
2018-02-12 22:42:39 +01:00
Stefano Cardanobile cb1324312d Introduce dynamic contempt
Make contempt dependent on the current score of the root position.

The idea is that we now use a linear formula like the following to decide
on the contempt to use during a search :

    contempt = x + y * eval

where x is the base contempt set by the user in the "Contempt" UCI option,
and y * eval is the dynamic part which adapts itself to the estimation of
the evaluation of the root position returned by the search. In this patch,
we use x = 18 centipawns by default, and the y * eval correction can go
from -20 centipawns if the root eval is less than -2.0 pawns, up to +20
centipawns when the root eval is more than 2.0 pawns.

To summarize, the new contempt goes from -0.02 to 0.38 pawns, depending if
Stockfish is losing or winning, with an average value of 0.18 pawns by default.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 110052 W: 24614 L: 23938 D: 61500
http://tests.stockfishchess.org/tests/view/5a72e6020ebc590f2c86ea20

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 16470 W: 2896 L: 2705 D: 10869
http://tests.stockfishchess.org/tests/view/5a76c5b90ebc5902971a9830

A second match at LTC was organised against the current master:

ELO: 1.45 +-2.9 (95%) LOS: 84.0%
Total: 19369 W: 3350 L: 3269 D: 12750
http://tests.stockfishchess.org/tests/view/5a7acf980ebc5902971a9a2e

Finally, we checked that there is no apparent problem with multithreading,
despite the fact that some threads might have a slightly different contempt
level that the main thread.

Match of this version against master, both using 5 threads, time control 30+0.3:
ELO: 2.18 +-3.2 (95%) LOS: 90.8%
Total: 14840 W: 2502 L: 2409 D: 9929
http://tests.stockfishchess.org/tests/view/5a7bf3e80ebc5902971a9aa2

Include suggestions from Marco Costalba, Aram Tumanian, Ronald de Man, etc.

Bench: 5207156
2018-02-09 19:07:19 +01:00
Leonid Pechenik d71adc5bd9 Retire "Extra thinking before accepting draw PVs"
This patch simplifies the time management code, removing the extra
thinking time for moves with draw PV and increasing thinking time
for all moves proportionally by around 4%.

Last time when the time management was carefully tuned was 1.5-2 years
ago. As new patches were getting added, time management was drifting out
of optimum. This happens because when search becomes more precise pv and
score are becoming more stable, there are less fail lows, best move is
picked earlier and there are less best move changes. All this factors are
entering in time management, and average time per move is decreasing with
more and more good patches. For individual patches such effect is small
(except some) and may be up or down, but when there are many of them,
effect is more substantial. The same way benchmark with more and more
patches is slowly drifting down on average.

So my understanding that back in October adding more think time for draw
PV showed positive Elo because time management was not well tuned, there
was more time available, and think_hard patch applied this additional time
to moves with draw PV, while just retuning back to optimum would recover Elo
anyway. It is possible that absence of contempt also helped, as SF9 is showing
less 0.0 scores than the October version.

Anyway, to me it seems that proper place to deal with draw PV is search, and
contempt sounds as much better solution. In time management there is little
additional elo, and if some code is not helping like removed here, it is better
to discard it. It is simpler to find genuine improvement if code is clean.

• Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20487 W: 4558 L: 4434 D: 11495
http://tests.stockfishchess.org/tests/view/5a7706ec0ebc5902971a9854

• Passed LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41960 W: 7145 L: 7058 D: 27757
http://tests.stockfishchess.org/tests/view/5a778c830ebc5902971a9895

• Passed an additional non-regression [-5..0] test at the time control
of 60sec for the game (sudden death) with disabled draw adjudication:
LLR: 2.95 (-2.94,2.94) [-5.00,0.00]
Total: 8438 W: 1675 L: 1586 D: 5177
http://tests.stockfishchess.org/tests/view/5a7c3d8d0ebc5902971a9ac0

• Passed an additional non-regression [-5..0] test at the time control
of 1sec+1sec per move with disabled draw adjudication:
LLR: 2.97 (-2.94,2.94) [-5.00,0.00]
Total: 27664 W: 5575 L: 5574 D: 16515
http://tests.stockfishchess.org/tests/view/5a7c3e820ebc5902971a9ac3

This is a functional change for the time management code.

Bench: 4983414
2018-02-09 10:41:32 +01:00
Stéphane Nicolet 211ebc5c7a Fix bug for 'eval' command in terminal
The 'eval' debugging command in Terminal did not initialize the Eval::Contempt
variable, leading to random output during debugging sessions (normal search
was unaffected by the bug).

Example of session where the two 'eval' commands should give the same output,
but did not:

./stockfish
position startpos
d
eval
go depth 20
d
eval

The bug is fixed by initializing Eval::Contempt to SCORE_ZERO in Eval::trace

No functional change.
2018-02-09 01:12:08 +01:00
FauziAkram 917fe69f84 A combo of 3 successful tuning patches
Shelter Weakness by Fauzi Akram Dabat
Threats by Alain Savard
Passed Pawns by Alain Savard

STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 51378 W: 11592 L: 11223 D: 28563
http://tests.stockfishchess.org/tests/view/5a79e2fe0ebc5902971a99d1

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21631 W: 3888 L: 3661 D: 14082
http://tests.stockfishchess.org/tests/view/5a7aefe80ebc5902971a9a39

Bench: 4983414
2018-02-08 15:14:02 +01:00
protonspring 69067e1988 Obey skipQuiets strictly in MovePicker
The current logic in master is to continue return quiet moves if their
history score is above 0. It appears as though this check can be
removed, which is also more logically consistent with the “skipQuiets”
semantics used in search.cpp.

This patch may open new opportunitiesto get Elo by changing or
tuning the definition of 'moveCountPruning' in line 830 of search.cpp,
because obeying skipQuiets without checking the history scores makes
the search more sensitive to 'moveCountPruning'.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 34780 W: 7680 L: 7584 D: 19516
http://tests.stockfishchess.org/tests/view/5a79f8d80ebc5902971a99db

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 38757 W: 6732 L: 6641 D: 25384
http://tests.stockfishchess.org/tests/view/5a7afebe0ebc5902971a9a46

Bench 4954595
2018-02-08 10:46:27 +01:00
Joost VandeVondele 312a248fa9 More robust bench extraction
Allow travis.yml to recognize a variety of bench formats in commit messages, for instance:

Bench: 5023593. (really).
bench: 5023593 (it was 1234567)
bench : 5023593 (blah blah)
Bench:5023593
Bench: 5023593. 567 something (1234567) 563

No functional change.
2018-02-07 01:29:53 +01:00
syzygy1 ef61886332 Enable LTO for clang
Enable link-time optimization in the Makefile when compiling with clang.
Also update travis.yml to use clang++-5.0 and llvm-5.0-dev.

No functional change.
2018-02-06 00:46:50 +01:00
protonspring 414a3e6ee3 Don’t score and sort all captures in RECAPTURES stage.
For these recaptures, we’re are only considering those captures
    that recapture the recapture square (small portion of all the
    captures). Therefore, scoring all of the captures and pick_besting
    out of the whole group is not necessary.

    STC
    LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
    Total: 85583 W: 18978 L: 18983 D: 47622
    http://tests.stockfishchess.org/tests/view/5a717faa0ebc590f2c86e9a7

    LTC
    LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
    Total: 20231 W: 3533 L: 3411 D: 13287
    http://tests.stockfishchess.org/tests/view/5a73ad330ebc5902971a96ba

    Bench: 5023593
2018-02-05 17:27:59 +01:00
Stéphane Nicolet e316e432d0 Revert "Implement old 'multipv' search"
This revert the following commit:
https://github.com/official-stockfish/Stockfish/commit/44a7db0f9ac02d2461aff39e25f1ac9107ffbfac

Bug report by Ronald de Man in issue:
https://github.com/official-stockfish/Stockfish/issues/1392

Bench: 5023629
2018-02-04 21:42:56 +01:00
protonspring d93baae220 Simplify qsearch stages in MovePicker
The difference between QCAPTURES_1 and QCAPTURES_2 quiescence search stages
boils down to a simple check of depth. The way it's being done now is
unnecessarily complex.

This patch is simpler, clearer, and easier to understand.

Passed SPRT[-3..1] test at STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 99755 W: 22158 L: 22192 D: 55405
http://tests.stockfishchess.org/tests/view/5a71f41c0ebc590f2c86e9cb

No functional change.
2018-02-04 14:57:57 +01:00
joergoster 44a7db0f9a Implement 'old' multipv search.
It seems to be a waste of time to loop through all remaining root moves
after finishing each PV line. This patch skips this until we have reached
the last PV line (this is the way it was done in Glaurung and very early
versions of Stockfish).

No functional change in Single PV mode.

MultiPV=3 STC and LTC tests
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 3113 W: 1248 L: 1064 D: 801

LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2260 W: 848 L: 679 D: 733

Bench: 5023629
2018-02-04 13:35:44 +01:00
Stéphane Nicolet 83c828f31e Restore development version
No functional change.
2018-02-04 02:08:09 +01:00
Stéphane Nicolet 0f6f42cd83 Improved spelling, grammar and comment
Author: Ben Koshy

No functional change
2018-02-04 00:13:56 +01:00
mstembera 378c8bdbb8 Stockfish 9
Official release version of Stockfish 9

Bench 5023629
2018-01-31 11:41:09 +01:00
Torsten Franz 2e11388581 Retire lever (#1378)
Retire the lever bonus from pawns evaluation

STC: http://tests.stockfishchess.org/tests/view/5a6aef5c0ebc590d945d59c8
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 88290 W: 19549 L: 19560 D: 49181

LTC: http://tests.stockfishchess.org/tests/view/5a6b70140ebc590d945d59f7
LLR: 3.22 (-2.94,2.94) [-3.00,1.00]
Total: 104603 W: 18105 L: 18108 D: 68390

Bench 5023629
2018-01-30 09:40:56 +01:00
Rocky640 0a5b03af3f Limit the king distance factor when evaluating passed pawns (#1373)
Limit the king distance factor when evaluating passed pawns
Passed STC
http://tests.stockfishchess.org/tests/view/5a6bf7290ebc590d945d5a3a
LLR: 3.31 (-2.94,2.94) [0.00,5.00]
Total: 23987 W: 5550 L: 5281 D: 13156

and LTC
http://tests.stockfishchess.org/tests/view/5a6c57710ebc590297c36af2
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 16926 W: 3014 L: 2820 D: 11092

Bench: 5059457
2018-01-28 14:56:45 +01:00
syzygy1 fd4d800c98 This modifies the in-search TB probing to continue searching for a mate "behind" a TB win (or loss). (#1285)
It does the following:

- If a TB win or loss value allows an alpha or beta cutoff, the cutoff is taken.
- Otherwise, the search of the current subtree continues. In PV nodes, the final value returned is adjusted to reflect that the position is a TB win (or loss).

The patch also fixes a potential problem caused by root_probe() and root_probe_wdl() dirtying the root-move scores.

This patch removes the limitation of current master that a mate is never found if the root position is not yet in the TBs, but the path to mate at some point enters the TBs. The patch is intended to preserve the efficiency and effectiveness of the current TB probing approach.

No functional change (withouth TB)
2018-01-28 14:40:07 +01:00
mibere a5a98d822e Top CPU Contributors as of January 2018 (#1367)
No functional change.
2018-01-24 16:54:04 +01:00
Stéphane Nicolet 254d995e18 Contempt 20
Set the default contempt value of Stockfish to 20 centipawns.

The contempt feature of Stockfish tries to prevent the engine from
simplifying the position too quickly when it feels that it is very
slightly behind, instead keeping the tension a little bit longer.

Various tests in November 2017 have proved that our current imple-
mentation works well against SF7 (which is about 130 Elo weaker than
current master) and than the Elo gain is an increasing function of
contempt, going (against SF7) from +0 Elo when contempt is set at
zero centipawns, to +30 Elo when contempt is 40 centipawns.

See pull request 1325 for details:

https://github.com/official-stockfish/Stockfish/pull/1325

This november discussion left open the decision of which "default"
value for contempt we should use for Stockfish, taking into account
the various uses ofStockfish (opening preparation for humans, computer
online tournaments,analysis tool for web pages, human/computer play,
etc).

This pull request proposes to set the default contempt value of SF
to twenty centipawns, which turns out to be the highest value which
is not a regression against current master, as this seemed to be a
good compromise between risk and safety. A couple of SPRT[-3..1]
tests were done to bisect this value:

Contempt 10: http://tests.stockfishchess.org/tests/view/5a5d42d20ebc5902977e2901 (PASSED)
Contempt 15: http://tests.stockfishchess.org/tests/view/5a5d41740ebc5902977e28fa (PASSED)
Contempt 20: http://tests.stockfishchess.org/tests/view/5a5d42060ebc5902977e28fc (PASSED)
Contempt 25: http://tests.stockfishchess.org/tests/view/5a5d433f0ebc5902977e2904 (FAILED)

Surprisingly, a test at "very long time control" hinted that using
contempt 20 is not only be non-regressive against contempt 0, but
may actually exhibit some small Elo gain, giving a likehood of superio-
rity of 88.7% after 8500 games:

VLTC:
ELO: 2.28 +-3.7 (95%) LOS: 88.7%
Total: 8521 W: 1096 L: 1040 D: 6385
http://tests.stockfishchess.org/tests/view/5a60b2820ebc590297b9b7e0

Finally, there was some concerns that a contempt value of 20 would
be worse than a value of 7, but a test with 20000 games at STC was
neutral:

STC:
ELO: 0.45 +-3.1 (95%) LOS: 61.2%
Total: 20000 W: 4222 L: 4196 D: 11582
http://tests.stockfishchess.org/tests/view/5a64d2fd0ebc590297903868

See the comments in pull request 1361 for the long, nice discussion
(180 entries :-)) leading to the decision to propose contempt 20 as
the default value:

https://github.com/official-stockfish/Stockfish/pull/1361

Whether Stockfish should strictly adhere to the Komodo and Houdini
semantics and add the UCI commands to force the contempt to be White
in the so-called "analysis mode" is still under discussion, and may
be or may not be the object of a future commit.

Bench: 5783344
2018-01-23 14:26:45 +01:00
Tom Vijlbrief 5451687efb Make razor margin depth independent
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 37171 W: 6680 L: 6587 D: 23904

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 11632 W: 1574 L: 1442 D: 8616

bench: 5098576
2018-01-23 14:15:05 +01:00
Fabian Fichter b61759e907 Simplify away redundant SEE pruning condition (#1363)
SEE immediately returns true for promotions,
so excluding them before checking SEE is redundant.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 56758 W: 10166 L: 10106 D: 36486
http://tests.stockfishchess.org/tests/view/5a645eaf0ebc590297903833

No functional change.
2018-01-23 14:05:48 +01:00
Marco Costalba b508f9561c Fix bench number
Incorrect bench number in master. Fix it.

bench: 4971497
2018-01-13 09:21:23 +01:00
Günther Demetz 1b6459195c Simplify verification search (#1362)
1. avoid recursive call of verification.
   For the interested side to move recursion makes no sense.
   For the other side it could make sense in case of mutual zugzwang,
   but I was not able to figure out any concrete problematic position.
   Allows the removal of 2 local variables.
   
2. avoid further reduction by removing R += ONE_PLY;

Benchmark with zugzwang-suite (see #1338), max 45 secs per position:
Patch  solves 33 out of 37
Master solves 31 out of 37

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 76188 W: 13866 L: 13840 D: 48482
http://tests.stockfishchess.org/tests/view/5a5612ed0ebc590297da516c

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40479 W: 5247 L: 5152 D: 30080
http://tests.stockfishchess.org/tests/view/5a56f7d30ebc590299e4550e

bench: 5340015
2018-01-13 09:01:23 +01:00
IIvec aa88261a8f Revert to old time management (#1351)
As many users reported some problems with new time management,
and recent tests on longer time controls

http://tests.stockfishchess.org/tests/view/5a460e160ebc590ccbb8c35d
http://tests.stockfishchess.org/tests/view/5a462f4d0ebc590ccbb8c37a

are even little in favor of old time management, this revert seems as a logical step.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14060 W: 2562 L: 2430 D: 9068

LTC:
LLR: 3.44 (-2.94,2.94) [-3.00,1.00]
Total: 31611 W: 3958 L: 3827 D: 23826

bench: 5365777 (same as master)
2018-01-13 08:59:20 +01:00
mbootsector 33682bfb98 Use mobility in kingsafety (#1360)
Use mobility in kingsafety

STC: http://tests.stockfishchess.org/tests/view/5a55134d0ebc590296938a28
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 32651 W: 6057 L: 5792 D: 20802

LTC: http://tests.stockfishchess.org/tests/view/5a5618f40ebc590297da516f
LLR: 2.94 (-2.94,2.94) [0.00,5.00]
Total: 17875 W: 2421 L: 2245 D: 13209

bench: 5365777
2018-01-13 08:56:42 +01:00
hxim 66ce8ad5fd Remove ThreatByHangingPawn bonus (#1356)
* no ThreatByHangingPawn

* bench: 4919682
2018-01-04 22:31:30 +01:00
Joost VandeVondele 9afa1d7330 New Year 2018
Adjust copyright headers.

No functional change.
2018-01-01 13:18:10 +01:00
Joost VandeVondele 2ba47416cb Explicitly zero TT upon resize.
as discussed in issue #1349, the way pages are allocated with calloc might imply some overhead on first write.
This overhead can be large and slow down the first search after a TT resize significantly, especially for large TT.
Using an explicit clear of the TT on resize fixes this problem.

Not implemented, but possibly useful for large TT, is to do this zero-ing using all search threads. Not only would this be faster, it could also lead to a more favorable memory allocation on numa systems with a first touch policy.

No functional change.
2018-01-01 13:17:51 +01:00
Guenther Demetz 7d4d3a2c3a Include x-ray attacks through all queens independently of the color.
When calculating attacks from rooks/bishops current master includes
x-rays through own queen. This patch includes also x-rays through
opponent queen.

Credits go to Brian who inspired for this idea
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/Z3APRYpQeMU

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 41549 W: 7544 L: 7244 D: 26761
Elo	2.05 [-0.29,4.19] (95%)
http://tests.stockfishchess.org/tests/view/5a3b5fe50ebc590ccbb8bf9a

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 110010 W: 14208 L: 13739 D: 82063
Elo	1.20 [-0.27,2.55] (95%)
http://tests.stockfishchess.org/tests/view/5a3b8e7c0ebc590ccbb8bfad

bench: 5544445
2017-12-26 10:42:24 +01:00
Joost VandeVondele 1c50d8cbf5 Upon changing the number of threads, make sure all threads are bound
The heuristic to avoid thread binding if less than 8 threads are requested resulted in the first 7 threads not being bound.
The branch was verified to yield a roughly 13% speedup by @CoffeeOne on the appropriate hardware and OS, and an earlier version of this patch tested well on his machine:

http://tests.stockfishchess.org/tests/view/5a3693480ebc590ccbb8be5a
ELO: 9.24 +-4.6 (95%) LOS: 100.0%
Total: 5000 W: 634 L: 501 D: 3865

To make sure all threads (including mainThread) are bound as soon as the total number exceeds 7, recreate all threads on a change of thread number.
To do this, unify Threads::init, Threads::exit and Threads::set are unified in a single Threads::set function that goes through the needed steps.
The code includes several suggestions from @joergoster.

Fixes issue #1312 

No functional change
2017-12-26 10:40:42 +01:00
Joost VandeVondele 2198cd0524 Allow for general transposition table sizes. (#1341)
For efficiency reasons current master only allows for transposition table sizes that are N = 2^k in size, the index computation can be done efficiently as (hash % N) can be written instead as (hash & 2^k - 1). On a typical computer (with 4, 8... etc Gb of RAM), this implies roughly half the RAM is left unused in analysis.

This issue was mentioned on fishcooking by Mindbreaker:
http://tests.stockfishchess.org/tests/view/5a3587de0ebc590ccbb8be04

Recently a neat trick was proposed to map a hash into the range [0,N[ more efficiently than (hash % N) for general N, nearly as efficiently as (hash % 2^k):

https://lemire.me/blog/2016/06/27/a-fast-alternative-to-the-modulo-reduction/

namely computing (hash * N / 2^32) for 32 bit hashes. This patch implements this trick and now allows for general hash sizes. Note that for N = 2^k this just amounts to using a different subset of bits from the hash. Master will use the lower k bits, this trick will use the upper k bits (of the 32 bit hash).

There is no slowdown as measured with [-3, 1] test:

http://tests.stockfishchess.org/tests/view/5a3587de0ebc590ccbb8be04
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 128498 W: 23332 L: 23395 D: 81771

There are two (smaller) caveats:

1) the patch is implemented for a 32 bit hash (so that a 64 bit multiply can be used), this effectively limits the number of clusters that can be used to 2^32 or to 128Gb of transpostion table. That's a change in the maximum allowed TT size, which could bother those using 256Gb or more regularly.

2) Already in master, an excluded move is hashed into the position key in rather simple way, essentially only affecting the lower 16 bits of the key. This is OK in master, since bits 0-15 end up in the index, but not in the new scheme, which picks the higher bits. This is 'fixed' by shifting the excluded move a few bits up. Eventually a better hashing scheme seems wise.

Despite these two caveats, I think this is a nice improvement in usability.

Bench: 5346341
2017-12-18 16:32:21 +01:00
Günther Demetz b53239d641 Enhanced verify search (#1338)
by disabling null-move-pruning for the side to move for first part of
the remaining search tree. This helps to better recognize zugzwang.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 18220 W: 3379 L: 3253 D: 11588
http://tests.stockfishchess.org/tests/view/5a2fa6460ebc590ccbb8bc2f

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41899 W: 5359 L: 5265 D: 31275
http://tests.stockfishchess.org/tests/view/5a2fcf440ebc590ccbb8bc47

For further detail see commit notes and discussion at 
https://github.com/pb00068/Stockfish/commit/6401a80ab91df5c54390ac357409fef2e51ff5bb

bench: 5776193
2017-12-18 16:30:27 +01:00
Gontran Lemaire 83e829c9dc Remove QueenMinorsImbalance array #1340
Remove QMI array and adjust bishop, knight and queen coefficients
in QuadraticOurs and QuadraticTheirs arrays in compensation of
this removal.

STC : http://tests.stockfishchess.org/tests/view/5a21d8350ebc590ccbb8b5fe
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 49659 W: 9029 L: 8957 D: 31673

LTC : http://tests.stockfishchess.org/tests/view/5a33c0dd0ebc590ccbb8bd7e
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 45905 W: 5834 L: 5745 D: 34326

Bench: 5176807
2017-12-17 09:20:17 +01:00
Joost VandeVondele f6981b1321 Make staticEval independent of the search path
Current master can yield different staticEvals depending on the path
used to reach the position. The reason for this is that the evaluation after a
null move is always computed subtracting 2 * Eval::Tempo, while this is not
the case for lazy or specialized evals. This patch always adds tempo to evals,
which doesn't affect playing strength:

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 59911 W: 7616 L: 7545 D: 44750

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 104947 W: 18897 L: 18919 D: 67131

Fixes issue #1335

Bench: 5208264
2017-12-17 09:11:55 +01:00
Rocky640 be6fafd079 Simplify other checks (#1337)
Replace an intricate definition with a more natural one.

Master was excluding squares occupied by a pawn which was blocked by a pawn.
This version excludes any squares occupied by a pawn which is blocked by "something"

Passed STC
http://tests.stockfishchess.org/tests/view/5a2f557b0ebc590ccbb8bc0d
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 44211 W: 8009 L: 7928 D: 28274

and LTC
http://tests.stockfishchess.org/tests/view/5a301d440ebc590ccbb8bc80
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 31958 W: 4108 L: 4002 D: 23848

Bench 5000136
2017-12-17 08:50:45 +01:00
Alain SAVARD 020dd69a35 Simplify other checks #1334
Simplify the other check penalty computation. Compared to current master,

a) it uses a 143 kingDanger penalty instead of S(10, 10) for the "otherCheck"
(credits to ElbertoOne for finding a suitable kingDanger range to replace the score
and to Guardian for showing this could also be a neutral change at LTC).
This makes our king safety model more consistent and simpler.

b) it might also score more than one "otherCheck" penalty for a given piece type instead of just one

c) it might score many pinned penalties instead of just one.

d) It also remove 3 conditionals and uses simpler expressions.
So it was tested as a SPRT[-3, 1]

Passed STC
http://tests.stockfishchess.org/tests/view/5a2b560b0ebc590ccbb8ba6b
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 11705 W: 2217 L: 2080 D: 7408

And LTC
http://tests.stockfishchess.org/tests/view/5a2bfd0d0ebc590ccbb8bab0
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 26812 W: 3575 L: 3463 D: 19774

Trying to improve on b) another attempt was made to score also the
"otherchecks" for piece types which had some safe checks, but this
failed STC http://tests.stockfishchess.org/tests/view/5a2c79e60ebc590ccbb8badd

bench: 5149133
2017-12-11 15:27:44 +01:00
Ben Koshy 70262f2027 Add Resources to understand code base (#1332)
No functional change.
2017-12-10 13:46:43 +01:00
Günther Demetz 196ae7c38b Don't consider defending queen as check blocker (#1328)
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 22636 W: 4212 L: 3990 D: 14434
http://tests.stockfishchess.org/tests/view/5a2506140ebc590ccbb8b75a

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 63448 W: 8287 L: 7965 D: 47196
http://tests.stockfishchess.org/tests/view/5a253a610ebc590ccbb8b776

bench: 5767699
2017-12-06 14:23:41 +01:00
Stéphane Nicolet be382bb0cf A better contempt implementation for Stockfish (#1325)
* A better contempt implementation for Stockfish

The round 2 of TCEC season 10 demonstrated the benefit of having a nice contempt implementation: it gives the strongest programs in the tournament the ability to slow down the game when they feel the position is slightly worse, prefering to stay in a complicated (even if slightly risky) middle game rather than simplifying by force into a drawn endgame.

The current contempt implementation of Stockfish is inadequate, and this patch is an attempt to provide a better one.

Passed STC non-regression test against master:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 83360 W: 15089 L: 15075 D: 53196
http://tests.stockfishchess.org/tests/view/5a1bf2de0ebc590ccbb8b370

This contempt implementation is showing promising results in certains situations. For instance, it obtained a nice +30 Elo gain when playing with contempt=40 against Stockfish 7, compared to current master:

• master against SF 7 (20000 games at LTC): +121.2 Elo
• this patch with contempt=40 (20000 games at LTC): +154.11 Elo

This was the result of real cooperative work from the Stockfish team, with key ideas coming from Stefan Geschwentner (locutus2) and Chris Cain (ceebo) while most of the community helped with feedback and computer time.

In this commit the bench is unchanged by default, but you can test at home with the new contempt in the UCI options. The style of play will change a lot when using contempt different of zero (I repeat: not done in this version by default, however)!

The Stockfish team is still deliberating over the best default contempt value in self-play and the best contempt modeling strategy, to help users choosing a contempt value when playing against much weaker programs. These informations will be given in future commits when available :-)

Bench: 5051254

* Remove the prefetch

No functional change.
2017-12-05 07:25:42 +01:00
Joost VandeVondele d193482213 Pawn endgames directly skip early pruning.
Instead of checking individual steps. Idea by @Stefano80.

passed STC
http://tests.stockfishchess.org/tests/view/5a23e5d20ebc590ccbb8b6d5
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 37445 W: 6866 L: 6773 D: 23806

passed LTC
http://tests.stockfishchess.org/tests/view/5a24260c0ebc590ccbb8b716
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 38780 W: 4946 L: 4848 D: 28986

Bench: 5466219
2017-12-04 17:57:36 +01:00
Stéphane Nicolet 54f6ce70fd Avoid warnings by the Clang compiler
Clang gave a couple of warnings for unused parameters after the recnet commit "Use constexpr when makes sense".

No functional change.
2017-12-04 17:53:42 +01:00
syzygy1 822695d4d3 Use a Direction enum for Square deltas
Currently the NORTH/WEST/SOUTH/EAST values are of type Square, but conceptually they are not squares but directions. This patch separates these values into a Direction enum and overloads addition and subtraction to allow adding a Square to a Direction (to get a new Square).

I have also slightly trimmed the possible overloadings to improve type safety. For example, it would normally not make sense to add a Color to a Color or a Piece to a Piece, or to multiply or divide them by an integer. It would also normally not make sense to add a Square to a Square.

This is a non-functional change.
2017-12-04 17:52:31 +01:00
Joost VandeVondele 2acda1fde3 Use bool(Bitboard b) instead of !!b (#1321)
The idiom !!b is confusing newcomers (e.g. Stefan needs explaining here https://groups.google.com/d/msg/fishcooking/vYqnsRI4brY/Gaf60QuACwAJ).

No functional change.
2017-12-03 18:29:55 +01:00
Guy Vreuls 28b6a457c2 Use constexpr when makes sense
No functional change.
2017-12-03 12:44:24 +01:00
Stéphane Nicolet ccd6bad512 Compile without exceptions
Add the -fno-exceptions flag to the Makefile to avoid the unecessary exceptions support in the executable (we do not use any exception in Stockfish at the moment).

This change gives a 9.2% reduction in size for the executable binary.

Before : executable size = 376956 bytes
After: executable size = 347652 bytes

No functional change.
2017-12-03 12:30:09 +01:00
syzygy 8a5a64eac5 Minor cleanup of search.cpp
Four very minor edits. Note that tte->save() uses posKey and
not pos.key() in other places.

Originally I also added a futility_move_counts() function to
make things more consistent with the futility_margin() and
reduction() functions. But then razor_margin[] should probably
also be turned into a function, etc. Maybe a good idea, maybe not.
So I did not include it.

Non functional change.
2017-12-03 12:24:46 +01:00
Stefan Geschwentner a87a1005ad Attack threats
Give bonus for safe attack threats from bishops and rooks on opponent queen

STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 8629 W: 1599 L: 1438 D: 5592
http://tests.stockfishchess.org/tests/view/5a1ad4490ebc590ccbb8b30d

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 7093 W: 987 L: 846 D: 5260
http://tests.stockfishchess.org/tests/view/5a1aec5d0ebc590ccbb8b317

Bench: 5051254
2017-12-03 10:41:48 +01:00
basepr1me 7dd1f4a7c0 OpenBSD friendly start. 2017-11-18 16:45:33 +01:00
Stefan Geschwentner d64ffd9621 Simplify good/bad capture detection. bench 5336313 2017-11-18 16:27:44 +01:00
Stefano Cardanobile c769d4df84 Fix comments. Bench: 5109559. 2017-11-18 14:08:06 +01:00
Stéphane Nicolet 03a9b3bd8d Simplify away the PawnSet[] imbalance array (#1308)
Simplify away the PawnSet[] imbalance array

STC:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 52977 W: 9550 L: 9484 D: 33943
http://tests.stockfishchess.org/tests/view/5a06b4780ebc590ccbb8a833

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 83717 W: 10599 L: 10559 D: 62559
http://tests.stockfishchess.org/tests/view/5a0aa36a0ebc590ccbb8aa99

Bench: 5340212
2017-11-18 11:24:23 +01:00
Rocky640 53239d7d3d Simplify some kingring penalties expressions
The new "weak" expression helps simplify the safe check calculations for rooks or minors, (but the end result for all the safe checks is the exactly the same as in current master)

The only functional change is for the "outer king ring" (for example, squares f3 g3 h3 when white king is on g1). In current master, there was a 191 penalty if any of these was not defended at all.
With this pr, there is this 191 penalty if any of these is not defended at all or is only defended by a white queen.

Tested as a simplification
STC
http://tests.stockfishchess.org/tests/view/59fb03d80ebc590ccbb89fee
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 66167 W: 12015 L: 11971 D: 42181
(against master (Update Copyright year inMakefile))

LTC
http://tests.stockfishchess.org/tests/view/5a0106ae0ebc590ccbb8a55f
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 15790 W: 2095 L: 1968 D: 11727
(against master (Handle BxN trade as good capture when history scor))

same as #1296 but rebased on latest master
bench: 5109559
2017-11-11 13:37:29 +01:00
ceebo 3f44f5303b Add comments to pos.see_ge()
In terms of technical changes this patch eliminates the return
statements from the main loop of pos.see_ge() and replaces two conditional
computations with a single bitwise negation.

No functional change
2017-11-10 12:14:53 +01:00
VoyagerOne 87452f3a8c Capture Stat Simplification- Bench: 5363761 2017-11-10 12:12:58 +01:00
Gian-Carlo Pascutto 8cfcca12d1 Always do MaxCardinality checks.
Stockfish currently relies on the "filter_root_moves" function also
having the side effect of clamping Cardinality against MaxCardinality
(the actual piece count in the tablebases). So if we skip this function,
we will end up probing in the search even without tablebases installed.

We cannot bail out of this function before this check is done, so move
the MultiPV hack a few lines below.
2017-11-08 13:45:14 +01:00
Joost VandeVondele 0a74c16ffe Simplify Null Move Search condition
Removes depth condition, adjust parameters.

passed STC:
http://tests.stockfishchess.org/tests/view/5a008cbc0ebc590ccbb8a512
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 29282 W: 5317 L: 5210 D: 18755

passed LTC:
http://tests.stockfishchess.org/tests/view/5a00d8530ebc590ccbb8a541
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26893 W: 3458 L: 3345 D: 20090

Bench: 5015773
2017-11-08 13:44:24 +01:00
Günther Demetz 652199d840 Handle BxN trade as good capture when history score is good
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19374 W: 3499 L: 3294 D: 12581
http://tests.stockfishchess.org/tests/view/59fc23f50ebc590ccbb8a0bf

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 91030 W: 11680 L: 11274 D: 68076
http://tests.stockfishchess.org/tests/view/59fc43ad0ebc590ccbb8a0d0

Bench: 5482249
2017-11-05 22:05:51 +01:00
Stefan Geschwentner 4bc11984fc Introduce capture history table for capture move sorting
Introduce capture move history table indexed by moved piece,
target square and captured piece type for sorting capture moves.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 11374 W: 2096 L: 1924 D: 7354
http://tests.stockfishchess.org/tests/view/59fac8dc0ebc590ccbb89fc5

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 24791 W: 3196 L: 3001 D: 18594
http://tests.stockfishchess.org/tests/view/59fae4d20ebc590ccbb89fd9

Bench: 5536775
2017-11-03 13:57:18 +01:00
Joost VandeVondele 486c8175c4 Replace easyMove with simple scheme
Reduces time for a stable bestMove, giving some of the won time for the next move.

the version before the pvDraw passed both STC and LTC

passed STC:
http://tests.stockfishchess.org/tests/view/59e98d5a0ebc590ccbb896ec
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 78561 W: 13945 L: 13921 D: 50695
elo =    0.106 +-    1.445 LOS:   55.716%

passed LTC:
http://tests.stockfishchess.org/tests/view/59eb9df90ebc590ccbb897ae
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 29056 W: 3640 L: 3530 D: 21886
elo =    1.315 +-    1.982 LOS:   90.314%

This version, rebased on pvDrawPR with the obvious change, was verified again on STC:

http://tests.stockfishchess.org/tests/view/59ee104e0ebc590ccbb89899
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19890 W: 3648 L: 3525 D: 12717
elo =    2.149 +-    2.895 LOS:   92.692%

and LTC:
http://tests.stockfishchess.org/tests/view/59f9673a0ebc590ccbb89ea0
Total             :    17966
Win               :     2273 (  12.652%)
Loss              :     2149 (  11.961%)
Draw              :    13544 (  75.387%)
Score             :   50.345%
Sensitivity       :    0.014%
2*(W-L)/(W+L)     :    5.608%

LLR  [-3.0,  1.0] :     2.95

BayesElo range    : [  -1.161,   4.876,  10.830] (DrawElo:  341.132)
LogisticElo range : [  -0.501,   2.105,   4.677]
LOS               :   94.369 %

LTC again:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17966 W: 2273 L: 2149 D: 13544
LogisticElo range : [ -0.501, 2.105, 4.677]
LOS : 94.369 %

unchanged bench: 5234652
2017-11-03 13:51:53 +01:00
Cooffe e0d2fdc843 Update Copyright year inMakefile
No functional change.
2017-10-28 12:35:44 +02:00
Joost VandeVondele e50af36a00 Extra thinking before accepting draw PVs.
If the PV leads to a draw (3-fold / 50-moves) position
and we're ahead of time, think a little longer, possibly
finding a better way.

As this is most likely effective at higher draw rates,
tried speculative LTC after a yellow STC:

STC:
http://tests.stockfishchess.org/tests/view/59eb173a0ebc590ccbb8975d
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 56095 W: 10013 L: 9902 D: 36180
elo =    0.688 +-    1.711 LOS:   78.425%

LTC:
http://tests.stockfishchess.org/tests/view/59eba1670ebc590ccbb897b4
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 59579 W: 7577 L: 7273 D: 44729
elo =    1.773 +-    1.391 LOS:   99.381%

bench: 5234652
2017-10-28 12:33:48 +02:00
IIvec 287e2e2f74 Fix premature using of all available time in x/y TC
In x/y time controls there was a theoretical possibility
to use all available time few moves before the clock will
be updated with new time. This patch fixes that issue.

Tested at 60/15 time control:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 113963 W: 20008 L: 20042 D: 73913

The test was done without adjudication rules!

Bench 5234652
2017-10-22 07:43:37 +02:00
Gian-Carlo Pascutto 86ac50403d Don't filter root moves if MultiPV mode is enabled
A band-aid patch to workaround current TB code
limitations with multi PV.

Hopefully this will be removed after committing the
big update of TB impementation, now under discussion.

No functional change.
2017-10-22 07:18:48 +02:00
ceebo 0dc3b0978d Add initiative to trace
No functional change
2017-10-22 07:00:46 +02:00
Joost VandeVondele 9d79138682 Fix issue #1268
If the search is quit before skill.pick_best is called,
skill.best_move might be MOVE_NONE.

Ensure skill.best is always assigned anyhow.

Also retire the tricky best_move() and let the underlying
semantic to be clear and explicit.

No functional change.
2017-10-11 11:47:50 +02:00
Alain SAVARD 43c186c645 Simplify bonus for bishop on long diagonal
Removing 2 conditions, and increase the ThreatbyPawn to compensate.

Passed STC
http://tests.stockfishchess.org/tests/view/59dbde900ebc5916ff64be6d
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14236 W: 2615 L: 2483 D: 9138

Passed LTC
http://tests.stockfishchess.org/tests/view/59dc26470ebc5916ff64be92
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16552 W: 2136 L: 2010 D: 12406

Bench: 5234652
2017-10-11 11:13:44 +02:00
Niklas Fiekas b36489742b WLDEntryPiece -> WDLEntryPiece for consistency
No functional change.
2017-10-05 11:00:01 +02:00
Stéphane Nicolet 452e5154cf Good bishops on the main diagonals
Bonus in midgame for bishops on long diagonals when the central squares are not occupied by pawns.

Author: ElbertoOne

Passed STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 10801 W: 1955 L: 1786 D: 7060
http://tests.stockfishchess.org/tests/view/59cf5c1d0ebc5916ff64b9da

and LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 83978 W: 10685 L: 10303 D: 62990
http://tests.stockfishchess.org/tests/view/59cf6f6c0ebc5916ff64b9e4

Bench: 5620312
2017-10-02 07:53:28 +02:00
VoyagerOne 07b5a28a68 Decrease reduction for exact PV nodes
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 59004 W: 10621 L: 10249 D: 38134

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 25801 W: 3306 L: 3108 D: 19387

Bench: 5742466
2017-09-30 20:56:27 +02:00
Joost VandeVondele c33af32dad Measure nodes after search finished.
Only affects nmpsec in the multithreaded case.

No functional change.
2017-09-29 16:39:12 +02:00
GuardianRM 0e949ac2c9 Tweak statScore condition
The first change (ss->statScore >= 0) does nothing.

The second change ((ss-1)->statScore >= 0 ) has a massive change.
(ss-1)->statScore is not set until (ss-1) begins to apply LMR to moves.
So we now increase the reduction for bad quiets when our opponent is
running through the first captures and the hash move.

STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 57762 W: 10533 L: 10181 D: 37048

LTC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19973 W: 2662 L: 2480 D: 14831

Bench: 5037819
2017-09-22 16:48:08 +02:00
syzygy ba4e215493 Let ss->ply denote the number of plies from the root to the current node
This patch lets ss->ply be equal to 0 at the root of the search.

Currently, the root has ss->ply == 1, which is less intuitive:

- Setting the rootNode bool has to check (ss-1)->ply == 0.

- All mate values are off by one: the code seems to assume that mated-in-0
  is -VALUE_MATE, mate-1-in-ply is VALUE_MATE-1, mated-in-2-ply is VALUE_MATE+2, etc.
  But the mate_in() and mated_in() functions are called with ss->ply, which is 1 in
  at the root.

- The is_draw() function currently needs to explain why it has "ply - 1 > i" instead
  of simply "ply > i".

- The ss->ply >= MAX_PLY tests in search() and qsearch() already assume that
  ss->ply == 0 at the root. If we start at ss->ply == 1, it would make more sense to
  go up to and including ss->ply == MAX_PLY, so stop at ss->ply > MAX_PLY. See also
  the asserts testing for 0 <= ss->ply && ss->ply < MAX_PLY.

The reason for ss->ply == 1 at the root is the line "ss->ply = (ss-1)->ply + 1" at
the start for search() and qsearch(). By replacing this with "(ss+1)->ply = ss->ply + 1"
we keep ss->ply == 0 at the root. Note that search() already clears killers in (ss+2),
so there is no danger in accessing ss+1.

I have NOT changed pv[MAX_PLY + 1] to pv[MAX_PLY + 2] in search() and qsearch().
It seems to me that MAX_PLY + 1 is exactly right:

- MAX_PLY entries for ss->ply running from 0 to MAX_PLY-1, and 1 entry for the
  final MOVE_NONE.

I have verified that mate scores are reported correctly. (They were already reported
correctly due to the extra ply being rounded down when converting to moves.)

The value of seldepth output to the user should probably not change, so I add 1 to it.
(Humans count from 1, computers from 0.)

A small optimisation I did not include: instead of setting ss->ply in every invocation
of search() and qsearch(), it could be set once for all plies at the start of
Thread::search(). This saves a couple of instructions per node.

No functional change (unless the search searches a branch MAX_PLY deep), so bench
does not change.
2017-09-17 10:44:10 +02:00
Stéphane Nicolet 043a469f83 Score unopposed weak pawns only if majors
Do not use the opposed flag for scoring backward and isolated pawns
in pawns.cpp, instead give a S(5,25) bonus for each opponent unopposed
weak pawns when we have a rook or a queen on the board.

STC run stopped after 113188 games:
LLR: 1.63 (-2.94,2.94) [0.00,5.00]
Total: 113188 W: 20804 L: 20251 D: 72133
http://tests.stockfishchess.org/tests/view/59b58e4d0ebc5916ff64b12e

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 66673 W: 8672 L: 8341 D: 49660
http://tests.stockfishchess.org/tests/view/59b902580ebc5916ff64b231

This is Alain Savard's idea, just with a different bonus.
Original patch there:

green STC, http://tests.stockfishchess.org/tests/view/597dcd2b0ebc5916ff64a09b
yellow LTC, http://tests.stockfishchess.org/tests/view/597ea69e0ebc5916ff64a0e6

Bench: 6259498
2017-09-17 09:52:27 +02:00
IIvec 21926ce2d8 Higher Move Overhead
This shoudl reduce time losses experienced by
users after new time management code.

Verified for no regression in very short TC (4sec + 0.1)
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35262 W: 7426 L: 7331 D: 20505

Bench 5322108
2017-09-12 12:31:53 +02:00
ianfab ed8286eb1b Extend ShelterWeakness array by dimension isKingFile
Use different penalties for weaknesses in the pawn shelter
depending on whether it is on the king's file or not.

STC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 71617 W: 13471 L: 13034 D: 45112

LTC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 48708 W: 6463 L: 6187 D: 36058

Bench: 5322108
2017-09-09 12:23:34 +02:00
VoyagerOne 3ac47c84d3 Streamlline reduction based on movecount
Use MoveCount History only at quiet moves and simply reduce
reduction by one depth instead of increasing moveCount in formula.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 27511 W: 5171 L: 4919 D: 17421

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 92337 W: 12135 L: 11748 D: 68454

Bench: 6351176
2017-09-05 11:03:50 +02:00
syzygy 741523eda8 Small simplication of see_ge()
Two simplifications:

- Remove the initialisation to 0 of occupied, which is now unnecessary.
- Remove the initial check for nextVictim == KING

If nextVictim == KING, then PieceValue[MG][nextVictim] will be 0, so that
balance >= threshold is true. So see_ge() returns true anyway.

No functional change.
2017-09-05 10:57:10 +02:00
Marco Costalba 04eb87fd08 Travis CI: Make all warnings into errors
Compile with -Werror flag. To make debugging easier
also show compile ourput.

This flag is enabled only in Travis CI, not in the shipped
Makefile becuase we can't test on every possible platform.
2017-09-05 10:40:34 +02:00
Joost VandeVondele 323925b91c Remove unneeded compile options.
In light of issue #1232, a test was performed about the value of '-fno-exceptions' and a second one of the combination '-fno-exceptions -fno-rtti'. It turns out these options are can be removed without introducing slowdown.

STC for removing '-fno-exceptions'
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 13678 W: 2572 L: 2439 D: 8667

STC for removing '-fno-exceptions -fno-rtti' (current patch)
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32557 W: 6074 L: 5973 D: 20510

No functional change.
2017-09-02 16:58:23 +02:00
syzygy 5ba4373522 Prevent Stockfish from exiting if DTZ table is not present
During TB initialisation, Stockfish checks for the presence of WDL
tables but not for the presence of DTZ tables. When attempting to probe
a DTZ table, it is therefore possible that the table is not present.
In that case, Stockfish should neither exit nor report an error.

To verify the bug:
$ ./stockfish
setoption name SyzygyTable value <path_to_WDL_dir>
position fen 8/8/4r3/4k3/8/1K2P3/3P4/6R1 w - -
go infinite
Could not mmap() /opt/tb/regular/KRPPvKR.rtbz
$

(On my system, the WDL tables are in one directory and the DTZ tables
in another. If they are in the same directory, it will be difficult
to trigger the bug.)

The fix is trivial: check the file descriptor/handle after opening
the file.

No functional change.
2017-09-02 09:59:04 +02:00
Marco Costalba 3fd7e37868 Fix a warning with MSVC
warning C4244: '*=': conversion from 'double' to 'int', possible loss of data

No functional change.
2017-09-01 20:23:31 +02:00
Joost VandeVondele 9d95d43c57 Multi-threaded search testing with valgrind
Also check with valgrind the multi-threaded search.

On top of the fix for issue #1227 (PR #1235).

No functional change.
2017-09-01 20:19:43 +02:00
Joost VandeVondele e385f194e9 Fix uninitialized memory usage
After increasing the number of threads, the histories were not cleared,
resulting in uninitialized memory usage.

This patch fixes this by clearing threads histories in Thread c'tor as
is the idomatic way.

This fixes issue 1227

No functional change.
2017-09-01 20:16:56 +02:00
VoyagerOne 7b4c9852e1 Adjust moveCount history only at LMR
STC:
LLR: 3.32 (-2.94,2.94) [-3.00,1.00]
Total: 17584 W: 3277 L: 3131 D: 11176

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 26412 W: 3447 L: 3334 D: 19631

Bench: 5417521
2017-08-31 08:53:37 +02:00
Joost VandeVondele bf485f4aff Simplify away non-normal moves in SEE
credit goes to @mstembera for suggesting this approach.
SEE now deals with castling, promotion and en passant in a similar way.

passed STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32902 W: 6079 L: 5979 D: 20844

passed LTC
LLR: 3.92 (-2.94,2.94) [-3.00,1.00]
Total: 110698 W: 14198 L: 14145 D: 82355

Bench: 5713905
2017-08-30 15:02:40 +02:00
Marco Costalba a2b8f91cfa Appveyor: do a Debug and Release build
And set x86 and x64 platforms for real.

Currently this is broken and the same binary is compiled for all platforms.

This is becuase we use a custom build step. OTH the default
build step seems not compatible with cmake generated *sln file.

No functional change.
2017-08-26 11:50:27 +02:00
Joost VandeVondele d5f883ab29 Improve multi-threaded mate finding
If any thread found a 'mate in x' stop the search. Previously only
mainThread would do so. Requires the bestThread selection to be
adjusted to always prefer mate scores, even if the search depth is less.

I've tried to collect some data for this patch. On 30 cores, mate finding
seems 5-30% faster on average. It is not so easy to get numbers for this,
as the time to find a mate fluctuates significantly with multi-threaded runs,
so it is an average over 100 searches for the same position. Furthermore,
hash size and position make a difference as well.

Bench: 5965302
2017-08-26 09:53:34 +02:00
Stéphane Nicolet 92c39522b1 Count all weak squares in the king ring with a single popcount
Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 26966 W: 4993 L: 4745 D: 17228
http://tests.stockfishchess.org/tests/view/599e798a0ebc5916ff64aa8c

and LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 39570 W: 5104 L: 4857 D: 29609
http://tests.stockfishchess.org/tests/view/599ee5230ebc5916ff64aabe

Bench: 5965302
2017-08-26 09:34:27 +02:00
Joost VandeVondele 5ef94eb970 Use moveCount history for reduction
Use less reduction for moves with larger moveCount if your
opponent did an unexpected (== high moveCount) move in the
previous ply... unexpected moves might need unexpected answers.

passed STC:
http://tests.stockfishchess.org/tests/view/599f08cc0ebc5916ff64aace
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 9638 W: 1889 L: 1720 D: 6029

passed LTC:
http://tests.stockfishchess.org/tests/view/599f1e5c0ebc5916ff64aadc
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 28308 W: 3742 L: 3533 D: 21033

Bench: 5747429
2017-08-26 09:30:42 +02:00
Stéphane Nicolet 002bf4d8db Avoid constructing an empty tuple in qsearch
Avoid constructing, passing as a parameter and binding a useless empty tuple of pointers in the qsearch move picker constructor.

Also reformat the scoring function in movepicker.cpp and do some cleaning in evaluate.cpp while there.

No functional change.
2017-08-22 10:16:19 +02:00
Marco Costalba 5ea327d924 Improve appeyor build
Check bench number and do not
hard-code *.cpp file names.

No functional change.
2017-08-20 19:22:11 +02:00
lucasart fe60caba94 Restore safety margin of 60ms
What this patch does is:
* increase safety margin from 40ms to 60ms. It's worth noting that the previous
  code not only used 60ms incompressible safety margin, but also an additional
  buffer of 30ms for each "move to go".
* remove a whart, integrating the extra 10ms in Move Overhead value instead.
  Additionally, this ensures that optimumtime doesn't become bigger than maximum
  time after maximum time has been artificially discounted by 10ms. So it keeps
  the code more logical.

Tested at 3 different time controls:

Standard 10+0.1
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 58008 W: 10674 L: 10617 D: 36717

Sudden death 16+0
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 59664 W: 10945 L: 10891 D: 37828

Tournament 40/10
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 16371 W: 3092 L: 2963 D: 10316

bench: 5479946
2017-08-20 20:00:31 +08:00
Marco Costalba 7aa7dfd4df Fix some Clang warnings
Found by Clang in extra verbose mode :-)

No functional change.
2017-08-19 14:32:31 +02:00
Marco Costalba fa5b0936ee Wide bench coverage
Add tests for:

- Positions with move list
- Chess960 positions

Now bench covers almost all cases, only few endgames
are still out of reach (verified with lcov)

It is a non functionality patch, but bench
changed because we added new test positions.

bench: 5479946
2017-08-19 13:58:12 +02:00
Marco Costalba 45e254a0a0 Restore perft
Rewrite perft to be placed naturally inside new
bench code. In particular we don't have special
custom code to run perft anymore but perft is
just a new parameter of 'go' command.

So user API is now changed, old style command:

$perft 5

becomes

$go perft 4

No functional change.
2017-08-18 09:04:38 -07:00
Marco Costalba 444d99b6d2 Rewrite benchmark
First step in improving bench to handle
arbitrary UCI commands so to test many
more code paths.

This first patch just set the new code
structure.

No functional change.
2017-08-18 09:04:38 -07:00
Marco Costalba e10255339f Reformat time manager code
In particular clarify that 'sd'
parameter is used only in !movesToGo
case.

Verified with Ivan's check tool it is
equivalent to original code.

No functional change.
2017-08-18 08:54:38 -07:00
Joost VandeVondele daf0fe1f57 Collect more corrections to optimum/maximum
The only call site of Time.maximum() corrected by 10.
Do this directly in remaining().

Ponder increased Time.optimum by 25% in init(). Idem.
Delete unused includes.

No functional change.
2017-08-18 08:38:40 -07:00
Marco Costalba 4d511512d2 Speed up Trevis CI
Avoid a couple of redundant rebuilds and compile
with 2 threads since travis gives 2vCPUs.

Also enable -O1 optimization for valgrind and
sanitizers, it should be safe withouth false
positives and it gives a very sensible speed
up, especially with valgrind.

The spee dup allow us to increase testing to
depth 10, useful for thread sanitizer.

No functional change.
2017-08-18 03:07:41 -07:00
Marco Costalba 9c35b9365d Clarify stats range
Current update formula ensures that the
possible value range is [-32 * D, 32 * D].

So we never overflow if abs(32 * D) < INT16_MAX

Thanks to Joost and mstembera to clarify this.

No functional change.
2017-08-18 02:02:35 -07:00
IIvec 01d97521fd Time management simplification
STC (http://tests.stockfishchess.org/tests/view/598188a40ebc5916ff64a21b):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25363 W: 4658 L: 4545 D: 16160

LTC (http://tests.stockfishchess.org/tests/view/5981d59a0ebc5916ff64a229):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 75356 W: 9690 L: 9640 D: 56026

40/10 TC (http://tests.stockfishchess.org/tests/view/5980c5780ebc5916ff64a1ed):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19377 W: 3650 L: 3526 D: 12201

15+0 TC (http://tests.stockfishchess.org/tests/view/5982cb730ebc5916ff64a25d):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 5913 W: 1217 L: 1069 D: 3627

This time management handles base time and movestogo cases separatelly. One can test one case without affecting the other. Also, increment usage can be tested separately without (necessarily) affecting sudden death or x moves in y seconds performance.

On stable machines there are no time losses on 0.1+0.001 time control (tested on i7 + Windows 10 platform).

Bench 5608839
2017-08-17 14:42:22 -07:00
Marco Costalba d482e3a890 Fix involuntary conversions of ExtMove to Move
The trick is to create an ambiguity for the
compiler in case an unwanted conversion to
Move is attempted like in:

    ExtMove m1{Move(17),4}, m2{Move(4),17};

    std::cout << (m1 < m2) << std::endl; // 1
    std::cout << (m1 > m2) << std::endl; // 1(!)

This fixes issue #1204

No functional change.
2017-08-17 02:04:00 -07:00
Marco Costalba 9001f55147 Unify stats update()
Now that is a bit bigger makes sense to
unify some duplicated code.

No functional change.
2017-08-17 01:52:26 -07:00
lucasart ae6a4ebf1f Use int16_t in History values
Reduces memory footprint by ~1.2MB (per thread).

Strong pressure: small but mesurable gain
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 258430 W: 46977 L: 45943 D: 165510

Low pressure: no regression
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 73542 W: 13058 L: 13026 D: 47458

Strong pressure + LTC: elo gain confirmed
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 31489 W: 4532 L: 4295 D: 22662

Tested for crashing on overflow and after 70K
games at STC we have only 4 time losses,
possible candidate for an overflow.

No functional change.
2017-08-17 00:32:44 -07:00
Marco Costalba 232c50fed0 Fix incorrect StateInfo
We use Position::set() to set root position across
threads. But there are some StateInfo fields (previous,
pliesFromNull, capturedPiece) that cannot be deduced
from a fen string, so set() clears them and to not lose
the info we need to backup and later restore setupStates->back().
Note that setupStates is shared by threads but is accessed
in read-only mode.

This fixes regression introduced by df6cb446ea

Tested with 3 threads at STC:
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 14436 W: 2304 L: 2196 D: 9936

Bench: 5608839
2017-08-14 23:01:58 -07:00
Marco Costalba c3e964f35e Run clang-tidy 'modernize'
Some warnings after a run of:

$ clang-tidy-3.8 -checks='modernize-*' *.cpp syzygy/*.cpp -header-filter=.* -- -std=c++11

I have not fixed all suggestions, for instance I still prefer
to declare the type instead of a spread use of 'auto'. I also
perfer good old 'typedef' to the new 'using' form.

I have not fixed some warnings in the last functions of
syzygy code because those are still the original functions
and need to be completely rewritten anyhow.

Thanks to erbsenzaehler for the original idea.

No functional change.
2017-08-13 05:46:21 -07:00
Marco Costalba df6cb446ea Thread code reformat
Simplify out low level sync stuff (mutex
and friends) and avoid to use them directly
in many functions.

Also some renaming and better comment while
there.

No functional change.
2017-08-13 04:41:59 -07:00
Marco Costalba bdeda52efd Retire States global variable
And other small touches in uci.cpp

No functional change.
2017-08-12 23:54:48 -07:00
tthsqe12 5837228aa0 Fix the handling of opposite bishops in KXK endgame evaluation
The case of three or more bishops against a long king must look at all of the
bishops and not just the first two in the piece lists. This patch makes sure
that the position is treated as a win when there are bishops on opposite
colors. This functional change is very small because bench remains the same.

LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 24249 W: 4349 L: 4275 D: 15625
http://tests.stockfishchess.org/tests/view/598186530ebc5916ff64a218

Bench: 5608839
2017-08-12 02:47:11 -07:00
Joost VandeVondele 2783203428 Simplify finished search in ponder/infinite mode.
In this rare case (e.g. go infinite on a stalemate),
just spin till ponderhit/stop comes.

The Thread::wait() is a renmant of the old YBWC
code, today with lazy SMP, threads don't need to
wait when outside of their idle loop.

No functional change.
2017-08-10 22:42:52 -07:00
Marco Costalba 66c5eaebd8 Re-apply the fix for Limits::ponder race
But this time correctly set Threads.ponder

We avoid using 'limits' for passing pondering
flag because we don't want to have 2 ponder
variables in search scope: Search::Limits.ponder
and Threads.ponder. This would be confusing also
because limits.ponder is set at the beginning of
the search and never changes, instead Threads.ponder
can change value asynchronously during search.

No functional change.
2017-08-10 12:47:31 -07:00
Marco Costalba 44236f4ed9 Revert "Fix a race on Limits::ponder"
This reverts commit 5410424e3d.

After the commit pondering is broken, so revert for now. I will
resubmit with a proper fix.

The issue is mine, Joost original code is correct.

No functional change.
2017-08-10 10:59:38 -07:00
Joost VandeVondele 5410424e3d Fix a race on Limits::ponder
Limits::ponder was used as a signal between uci and search threads,
but is not an atomic variable, leading to the following race as
flagged by a sanitized binary.

Expect input:
```
 spawn  ./stockfish
 send "uci\n"
 expect "uciok"
 send "setoption name Ponder value true\n"
 send "go wtime 4000 btime 4000\n"
 expect "bestmove"
 send "position startpos e2e4 d7d5\n"
 send "go wtime 4000 btime 4000 ponder\n"
 sleep 0.01
 send "ponderhit\n"
 expect "bestmove"
 send "quit\n"
 expect eof
```

Race:
```
WARNING: ThreadSanitizer: data race (pid=7191)
  Read of size 4 at 0x0000005c2260 by thread T1:

  Previous write of size 4 at 0x0000005c2260 by main thread:

  Location is global 'Search::Limits' of size 88 at 0x0000005c2220 (stockfish+0x0000005c2260)
```

The reason of teh race is that ponder is not just set in UCI go()
assignment but also is signaled by an async ponderhit in uci.cpp:

      else if (token == "ponderhit")
          Search::Limits.ponder = 0; // Switch to normal search

The fix is to add an atomic bool to the threads structure to
signal the ponder status, letting Search::Limits to reflect just
what was passed to 'go'.

No functional change.
2017-08-10 10:46:46 -07:00
Marco Costalba 750dfa0521 Fix some races and clarify the code
Better split code that should be run at
startup from code run at ucinewgame. Also
fix several races when 'bench', 'perft' and
'ucinewgame' are sent just after 'bestomve'
from the engine threads are still running.

Also use a specific UI thread instead of
main thread when setting up the Position
object used by UI uci loop. This fixes a
race when sending 'eval' command while searching.

We accept a race on 'setoption' to allow the
GUI to change an option while engine is searching
withouth stalling the pipe. Note that changing an
option while searchingg is anyhow not mandated by
UCI protocol.

No functional change.
2017-08-10 10:19:56 -07:00
AndyGrant dbc984d9f8 Make variable naming consistent
moved_piece is the only variable in search not using camel case
2017-08-10 10:04:30 -07:00
Joost VandeVondele f731bcadb7 Unify scoring functions in MovePicker
No functional change.
2017-08-10 02:06:21 -07:00
Joost VandeVondele b40e45c1cc Remove Stack/thread dependence in movepick
as a lower level routine, movepicker should not depend on the
search stack or the thread class, removing a circular dependency.
Instead of copying the search stack into the movepicker object,
as well as accessing the thread class for one of the histories,
pass the required fields explicitly to the constructor (removing
the need for thread.h and implicitly search.h in movepick.cpp).
The signature is thus longer, but more explicit:

Also some renaming of histories structures while there.

passed STC [-3,1], suggesting a small elo impact:

LLR: 3.13 (-2.94,2.94) [-3.00,1.00]
Total: 381053 W: 68071 L: 68551 D: 244431
elo =   -0.438 +-    0.660 LOS:    9.7%

No functional change.
2017-08-06 01:45:54 -07:00
snicolet 53c2d9df5e Tweak connected pawns seed[] array values
Raise a little bit the values in the connected pawns seed[] array.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 99033 W: 17939 L: 17448 D: 63646
http://tests.stockfishchess.org/tests/view/597355630ebc5916ff649e3e

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 48044 W: 6371 L: 6099 D: 35574
http://tests.stockfishchess.org/tests/view/597596610ebc5916ff649eba

Bench: 5608839

Closes #1182
2017-08-01 18:41:29 -07:00
Rocky640 b24bd762b2 Rework the "unsupported" penalty into a "supported" bonus
A pawn (according to all the searched positions of a bench run) is not supported 85% of the time,
(in current master it is either isolated, backward or "unsupported").

So it made sense to try moving the S(17, 8) "unsupported" penalty value into the base pawn value hoping for a more representative pawn value, and accordingly
a) adjust backward and isolated so that they stay more or less the same as master
b) increase the mg connected bonus in the supported case by S(17, 0) and let the Connected formula find a suitable eg value according to rank.

Tested as a simplification SPRT(-3, 1)

Passed STC
http://tests.stockfishchess.org/tests/view/5970dbd30ebc5916ff649dd6
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 19613 W: 3663 L: 3540 D: 12410

Passed LTC
http://tests.stockfishchess.org/tests/view/597137780ebc5916ff649de3
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24721 W: 3306 L: 3191 D: 18224

Bench: 5581946

Closes #1179
2017-08-01 18:37:54 -07:00
VoyagerOne 722e1e0da6 Remove redundant if-statements
No functional change

Closes #1173
2017-07-27 02:14:18 -07:00
mstembera 973ede008a Tuned PSQT using a custom tuner.
bench: 5878420

Closes #1177
2017-07-23 17:35:44 -07:00
VoyagerOne a6ae2d3a31 Simplify aspiration window
Don't modify alpha window on fail-high

Bench: 5875983

Closes #1172
2017-07-23 17:25:23 -07:00
Joost VandeVondele 272e4d1ac7 Faster travis checks
in the last month a couple of timeouts have been seen in travis valgrind testing, leading to undesired false positives. The precise cause of this is unclear: a normal valgrind instrumented run is about 6min, the timeout is 10min. Either there are rare hangs (not reproduced locally), or maybe the actual runtime fluctuates on the travis infrastructure (which uses VMs on AWS as far as I know). This patch leads to roughly a 2x speedup of the instrumented testing by reducing the depth from 10 to 9. If timeouts persist, it needs further analysis.

No functional change.

Closes #1171
2017-07-23 17:23:14 -07:00
Marco Costalba e551afbab7 Move game_phase() to material.cpp
For some reason, although game phase is used
only in material, it is computed in Position.

Move computation to material, where it belongs,
and remove the useless call chain.

No functional change.
2017-07-15 07:28:38 +02:00
Joona Kiiski d31f068312 Revert "Remove questionable gcc flags from profile-build"
This reverts commit 0371a8f8c4.
2017-07-13 16:36:27 -07:00
joergoster 377d77dbe9 Provide selective search depth info for each pv move
No functional change

Closes #1166
2017-07-13 16:30:03 -07:00
Joost VandeVondele 36a93d90f7 Move stop signal to Threads
Instead of having Signals in the search namespace,
make the stop variables part of the Threads structure.
This moves more of the shared (atomic) variables towards
the thread-related structures, making their role more clear.

No functional change

Closes #1149
2017-07-13 16:08:37 -07:00
Joona Kiiski 0371a8f8c4 Remove questionable gcc flags from profile-build
Optimization options for official stockfish should be
consistent, easy, future proof and simple.

We don't want to optimize for any specific version of gcc

No functional change

Closes #1165
2017-07-08 14:20:46 -07:00
GuardianRM c8e5384c3a Queen vs. Minors imbalance
Addition of correction values in case of Imbalance of queens,
depending on the number of light pieces on the side without a queen.

Passed patch:

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 29036 W: 5379 L: 5130 D: 18527

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 13680 W: 1836 L: 1674 D: 10170

Bench: 6258930

Closes #1155
2017-07-08 14:14:14 -07:00
Marco Costalba 802fca6fdd Don't uselessy share rootDepth
It is not needed becuase the only case is a real special
one (bench on depth with many threads) and can be easily
rewritten to avoid sharing rootDepth.

Verified with ThreadSanitizer.

No functional change.

Closes #1159
2017-07-02 22:06:47 -07:00
Marco Costalba 01b6cdb76b Fix some warnings with clang static analyzer
Only one remains (also in tbprobe.cpp), but is bougus.

As a side note, tbprobe.cpp is almost clean, only the last 3
functions probe_wdl(), root_probe() and root_probe_wdl()
are still the original ones and are quite hacky.

Somewhere in the future we will reformat also the last 3
ones. The reason why has not been done before it is because
these functions are really wrong by design and should be
rewritten entirely, not only reformatted.

No functional change.

Closes #1160
2017-07-02 22:02:11 -07:00
Marco Costalba c0cb713a00 Indentation fix in index()
No functional change.

Closes #1158
2017-07-02 22:00:29 -07:00
Alain SAVARD 6d24ef8585 Tidy up
No functional change

Closes #1148
2017-07-02 21:53:45 -07:00
mstembera 69eb391cd7 Magic::index()
Make magic_index() a member of Magic since it uses all it's members
and keep us from having to pass the function pointer around to
init_magics().

No functional change

Closes #1146
2017-06-28 17:11:17 -07:00
Joost VandeVondele 7e897a64f2 Remove race suppression.
Pull #1134 fixed another race, so that can be removed from the thread sanitizer suppressions.

No functional change.

Closes #1150
2017-06-28 17:06:52 -07:00
Marco Costalba 05513a6641 Only main thread checks time
The main change of the patch is that now time check
is done only by main thread. In the past, before lazy
SMP, we needed all the threds to check for available
time because main thread could have been blocked on
a split point, now this is no more the case and main
thread can do the job alone, greatly simplifying the logic.

Verified for regression testing on STC with 7 threads:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 11895 W: 1741 L: 1608 D: 8546

No functional change.

Closes #1152
2017-06-28 17:03:35 -07:00
Marco Costalba fa1e3427bd Simplify pos_is_ok()
Now we don't need anymore the tricky pointer to
show the failed test. Added some few tests too.

Also small rename in see_ge() while there.

No functional change

Closes #1151
2017-06-28 16:54:59 -07:00
VoyagerOne 77342126d8 Increase reduction if tt-move is a capture
The idea is that chances are the tt-move is best and will be difficult to raise alpha when playing a quiet move.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 7582 W: 1415 L: 1259 D: 4908

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 59553 W: 7885 L: 7573 D: 44095

Bench: 5725676

Closes #1147
2017-06-21 14:06:05 -07:00
snicolet 612d93234b Improve readability of evaluation functions
This patch puts the evaluation helper functions inside EvalInfo struct, which simplifies a bit their signature and (most importantly, IMHO) makes their C++ code much cleaner and simpler to read (by removing the "ei." qualifiers all around in evaluate.cpp).

Also rename the EvalInfo struct into Evaluation class to get a natural invocation v = Evaluation(p).value() to evaluation position p.

The downside is an increase of 20 lines in evaluate.cpp (for the prototypes of the helper functions). The upsides are better readability and a speed-up of 0.6% (by generating all the helpers for the NO_TRACE case together, which helps the instruction cache).

No functional change

Closes #1135
2017-06-21 14:01:59 -07:00
VoyagerOne 0149a4c3d6 Update Top CPU - Bench: 6599721
Closes #1145
2017-06-21 13:47:10 -07:00
Joona Kiiski 336901fdb0 Revert "Prefetch earlier in qsearch()"
This reverts commit b73016bb41.

No functional change

Closes #1144
2017-06-21 13:45:03 -07:00
Joost VandeVondele 3cb0200459 Fix four data races.
the nodes, tbHits, rootDepth and lastInfoTime variables are read by multiple threads, but not declared atomic, leading to data races as found by -fsanitize=thread. This patch fixes this issue. It is based on top of the CI-threading branch (PR #1129), and should fix the corresponding CI error messages.

The patch passed an STC check for no regression:

http://tests.stockfishchess.org/tests/view/5925d5590ebc59035df34b9f
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 169597 W: 29938 L: 30066 D: 109593

Whereas rootDepth and lastInfoTime are not performance critical, nodes and tbHits are. Indeed, an earlier version using relaxed atomic updates on the latter two variables failed STC testing (http://tests.stockfishchess.org/tests/view/592001700ebc59035df34924), which can be shown to be due to x86-32 (http://tests.stockfishchess.org/tests/view/592330ac0ebc59035df34a89). Indeed, the latter have no instruction to atomically update a 64bit variable. The proposed solution thus uses a variable in Position that is accessed only by one thread, which is copied every few thousand nodes to the shared variable in Thread.

No functional change.

Closes #1130
Closes #1129
2017-06-21 13:37:58 -07:00
Alain SAVARD 2c237da546 Misc coding style fixes
a few comment and blank fixes.

No functional change

Closes #1141
2017-06-16 19:55:30 -07:00
snicolet b73016bb41 Prefetch earlier in qsearch()
Closes #1139
2017-06-16 19:52:38 -07:00
Marco Costalba 27ba611a3d Better naming in endgame code
And small clean-up of magic bitboards code.

No functional change.

Closes #1138
2017-06-16 19:33:44 -07:00
Brian Sheppard f907d5b7d9 Move depth calculation in probCut
The change passed an STC regression:

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 59350 W: 10793 L: 10738 D: 37819

I verified that there was no change in performance on my machine, but of course YMMV:

Results for 40 tests for each version:

                Base      Test      Diff
        Mean    2014338   2016121   -1783
        StDev   62655     63441     3860

p-value: 0.678
speedup: 0.001

No functional change.

Closes #1137
2017-06-16 19:30:19 -07:00
Joost VandeVondele ebc563059c Call TT.new_search() earlier.
TT.new_search() was being called by mainThread in Thread::search(). However, mainThread is the last to start searching, and helper threads could reach a measured rootDepth 10 (on 64 cores) before mainThread increments the TT generation. Fixed by moving the call to MaintThread::search() before helper threads start searching.

No functional change.

Closes #1134
2017-06-16 19:20:01 -07:00
mstembera 659990b43f Reordering magic data
Gather all magic relevant data into a struct.

This changes memory layout putting everything necessary for processing a single square
in the same memory location thus speeding up access.

Original patch by @snicolet

No functional change.

Closes #1127
Closes #1128
2017-06-06 10:22:12 -07:00
atumanian 6d89d0b64a Don't score as an immediate draw 2-fold repetitions of the root position
In the current version a search stops when the current position is the same as
any position earlier in the search stack,
including the root position but excluding positions before the root.
The new version makes an exception for repeating the root position.

This gives correct scores for moves in the MultiPV > 1 mode.

Fixes #948 (see it for the detailed description of the bug).

LTC: http://tests.stockfishchess.org/tests/view/587910bc0ebc5915193f754b
ELO: 0.38 +-1.7 (95%) LOS: 66.8%
Total: 40000 W: 5166 L: 5122 D: 29712

STC: http://tests.stockfishchess.org/tests/view/5922e6230ebc59035df34a50
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 94622 W: 17059 L: 17064 D: 60499

 LTC: http://tests.stockfishchess.org/tests/view/59273a000ebc59035df34c03
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 61259 W: 7965 L: 7897 D: 45397

Bench: 6599721

Closes #1126
2017-06-06 10:15:13 -07:00
Joost VandeVondele 1781439fc2 use auto& for histories
No functional change.

Closes #1113
2017-06-06 09:56:13 -07:00
Marco Costalba ecd3218b6b History code rewrite (#1122)
Rearrange and rename all history heuristic code. Naming
is now based on chessprogramming.wikispaces.com conventions
and the relations among the various heuristics are now more
clear and consistent.

No functional change.
2017-05-26 08:42:50 +02:00
Nathan Rugg 24df0f72c0 Changed spelling back to "Bishops" in eval output
No functional change.

Closes #1124
2017-05-23 20:27:30 -07:00
VoyagerOne 1d31065e1d Evasion Pruning Tweak
Use moveCount to decide when to prune for evasion pruning

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 24476 W: 4518 L: 4289 D: 15669

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 18362 W: 2476 L: 2298 D: 13588

Bench: 6469989

Closes #1120
2017-05-21 18:27:57 -07:00
snicolet c216dcbe7b Do check analysis later in the game
The previous patch has added a fraction of the king danger score to the
endgame score of the tapered eval, so it seems natural to perform the
king danger computation later in the endgame.

With this patch we extend the limit of such check analysis down to the
material of Rook+Knight, when we have at least two pieces attacking the
opponent king zone.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 7446 W: 1409 L: 1253 D: 4784
http://tests.stockfishchess.org/tests/view/591c097c0ebc59035df3477c

and LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 14234 W: 1946 L: 1781 D: 10507
http://tests.stockfishchess.org/tests/view/591c24f10ebc59035df3478c

Bench: 5975183

Closes #1121
2017-05-17 18:24:43 -07:00
snicolet cf893bcded Use a fraction of king danger in endgame score
When SF has an attack on the opponent king in one flank, the huge
midgame -> endgame gradient of the tapered eval prevents us to properly
evaluate neutral exchanges on the other flank as the current king
danger score is a pure midgame term. This may affect SF's ability to
switch to defense in some positions. We add a small contribution
of the king danger to the endgame score to limit this
effect.

Again suggested in the following forum thread:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/xrUCQ7b0ObE

Passed STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 12719 W: 2371 L: 2192 D: 8156
http://tests.stockfishchess.org/tests/view/5919761a0ebc59035df3468f

And LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 31293 W: 4194 L: 3974 D: 23125
http://tests.stockfishchess.org/tests/view/591980450ebc59035df34695

Bench: 5961548

Closes #1118
2017-05-17 18:19:47 -07:00
Joost VandeVondele 732aa34e3d Fix memory access in Search::clear()
Fixes a bug in Search::clear, where the filling of CounterMoveStats&, overwrote (currently presumably unused) memory because sizeof(cm) returns the size in bytes, whereas elements was needed.

No functional change

Closes #1119
2017-05-17 18:15:01 -07:00
snicolet 862934d7ae Limit king ring to eight squares
In current master the size of the king ring varies abruptly from eight
squares when the king is in g8, to 12 squares when it is in g7. Because
the king ring is used for estimating attack strength, this may lead to
an overestimation of king danger in some positions. This patch limits
the king ring to eight squares in all cases.

 Inspired by the following forum thread:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/xrUCQ7b0ObE

Passed STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 9244 W: 1777 L: 1611 D: 5856

and LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 87121 W: 11765 L: 11358 D: 63998

Bench: 6121121

Closes #1115
2017-05-15 19:28:37 -07:00
Joost VandeVondele 7edd1f7ccd Execute an implied ucinewgame at startup
execute an implied ucinewgame upon entering the UCI::loop,
to make sure that searches starting with and without an (optional) ucinewgame
command yield the same search.

This is needed now that seach::clear() initializes tables to non-zero default values.

No functional change

Closes #1101
Closes #1104
2017-05-15 18:54:13 -07:00
Marco Costalba 0c1f119069 Default argument for see_ge()
No functional change.

Closes #1111
2017-05-10 18:20:45 +02:00
Joost VandeVondele 99d914985f Remove int to int conversion, unused include.
No functional change.

Closes #1112
2017-05-09 18:36:32 -07:00
FauziAkram 6b4959e3e0 Linear Protector bonus by distance
Replacing the old Protector table with a simple linear formula which takes into account a different slope for each different piece type.

The idea of this simplification of Protector is originated by Alain (Rocky)

STC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 70382 W: 12859 L: 12823 D: 44700

LTC: LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 61554 W: 8098 L: 8031 D: 45425

Bench: 6107863

Closes #1099
2017-05-07 21:11:51 -07:00
IIvec ae97941628 King safety and rook mobility parameters tweak
STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 19280 W: 3595 L: 3373 D: 12312

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 221405 W: 28940 L: 28220 D: 164245

Bench: 6506664

Closes #1105
2017-05-07 20:57:29 -07:00
Stefan Geschwentner 69ec09bd4b Bonus for pawn scrifice which create passed pawn
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16752 W: 3141 L: 2944 D: 10667

LTC:
LLR: 3.34 (-2.94,2.94) [0.00,5.00]
Total: 33928 W: 4544 L: 4300 D: 25084

Bench: 5639223

Closes #1092
2017-05-07 20:51:52 -07:00
Marco Costalba 25296547d0 Move Pieces[] out of global visibility
It is an helper array used only in position.cpp

Also small code tidy up while there.

No functional change.

Closes #1106
2017-05-07 20:20:02 -07:00
mstembera 321a27fbe3 Avoid *begin always being included in the sorted list regardless of its value.
This was a minor criticism by @zamar in the original pull request
https://github.com/official-stockfish/Stockfish/pull/1065
necessitating a comment explanation.

No functional change.

Closes #1091
2017-05-07 20:15:56 -07:00
joergoster 8b15961349 Fix multiPV issue #502
In general, this patch handles the cases where we don't have a valid score for each PV line in a multiPV search. This can happen if the search has been stopped in an unfortunate moment while still in the aspiration loop. The patch consists of two parts.

Part 1: The new PVIdx was already part of the k-best pv's in the last iteration, and we therefore have a valid pv and score to output from the last iteration. This is taken care of with:

      bool updated = (i <= PVIdx && rootMoves[i].score != -VALUE_INFINITE);

Case 2: The new PVIdx was NOT part of the k-best pv's in the last iteration, and we have no valid pv and score to output. Not from the current nor from the previous iteration. To avoid this, we are now also considering the previous score when sorting, so that the PV lines with no actual but with a valid previous score are pushed up again, and the previous score can be displayed.

  bool operator<(const RootMove& m) const {
    return m.score != score ? m.score < score : m.previousScore < previousScore; } // Descending sort

I also added an assertion in UCI::value() to possibly catch similar issues earlier.

No functional change.

Closes #502
Closes #1074
2017-05-03 19:46:40 -07:00
Joost VandeVondele e9f26cccdd gcc 7 port
Testing the release candidate revealed only one minor issue, namely a new warning -Wimplicit-fallthrough (part of -Wextra) triggers in the movepicker. This can be silenced by adding a comment, and once we move to c++17 by adding a standard annotation [[fallthrough]];.

No functional change.

Closes #1090
2017-04-30 08:43:43 -07:00
VoyagerOne a18c2c2c3f Don't do InCheck Pruning at the root of QS
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 34603 W: 6441 L: 6167 D: 21995

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 24474 W: 3274 L: 3076 D: 18124

Bench: 5934421

Closes #1089
2017-04-28 20:40:45 -07:00
Rocky640 b948b037a5 Remove cap in kingDanger initialization
Passed STC
http://tests.stockfishchess.org/tests/view/58fd53be0ebc59035df33eb5
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 52048 W: 9397 L: 9329 D: 33322

Passed LTC
http://tests.stockfishchess.org/tests/view/58ff9e0a0ebc59035df33f5c
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 35382 W: 4650 L: 4549 D: 26183

Bench: 5872717

Closes #1087
2017-04-28 20:38:03 -07:00
Marco Costalba e06a117d5e Retire the misdesigned StepAttacks[] array.
StepAttacks[] is misdesigned, the color dependance is specific
to pawns, and trying to generalise to king and knights, proves
neither useful nor convinient in practice.

So this patch reformats the code with the following changes:

- Use PieceType instead of Piece in attacks_() functions

- Use PseudoAttacks for KING and KNIGHT

- Rename StepAttacks[] into PawnAttacks[]

Original patch and idea from Alain Savard.

No functional change.

Closes #1086
2017-04-28 20:33:30 -07:00
Joost VandeVondele b1b19343cd Copy killers in the movepicker
ss->killers can change while the movepicker is active.
The reason ss->killers changes is related to the singular
extension search in the moves loop that calls search<>
recursively with ss instead of ss+1,
effectively using the same stack entry for caller and callee.
By making a copy of the killers,
the movepicker does the right thing nevertheless.

Tested as a bug fix

STC:
http://tests.stockfishchess.org/tests/view/58ff130f0ebc59035df33f37
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 70845 W: 12752 L: 12716 D: 45377

LTC:
http://tests.stockfishchess.org/tests/view/58ff48000ebc59035df33f3d
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 28368 W: 3730 L: 3619 D: 21019

Bench: 6465887

Closes #1085
2017-04-28 20:29:04 -07:00
snicolet 49a9d4cf99 Avoid misuse of StepAttacksBB for pawns
Make it explicit that first index of StepAttacksBB is a piece, not a piece type.

No functional change

Closes #1083
2017-04-25 17:57:49 -07:00
Joost VandeVondele d06a8d0c18 Zero unused constant
No functional change

Closes #1081
2017-04-25 17:21:36 -07:00
Marco Costalba b48439e906 Assorted code style issues
I have removed the check for

 pieceCount[PAWN] > FILE_NB

because totally useless.

No functional change.
2017-04-24 09:49:44 +02:00
Joost VandeVondele 6b9a22b40d Sort moves partially: linear depth dependence
STC: http://tests.stockfishchess.org/tests/view/58f98d260ebc59035df33d5e
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 58958 W: 10862 L: 10485 D: 37611

LTC: http://tests.stockfishchess.org/tests/view/58fa45d40ebc59035df33d86
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 18607 W: 2427 L: 2251 D: 13929

Bench: 6065528

Closes #1079
2017-04-23 08:37:55 -07:00
IIvec 0868de705d King safety parameters improved
STC:
LLR: 2.97 (-2.94,2.94) [0.00,4.00]
Total: 58648 W: 10883 L: 10524 D: 37241

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 52546 W: 7131 L: 6844 D: 38571

Bench 6121479

Closes #1078
2017-04-23 08:04:03 -07:00
Joost VandeVondele 9da3b44ddc Use int instead of Value for history related stats.
history related scores are not related to evaluation based scores.
For example, can easily exceed the range -VALUE_INFINITE,VALUE_INFINITE.
As such the current type is confusing, and a plain int is a better match.

tested for no regression:

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 43693 W: 7909 L: 7827 D: 27957

No functional change.

Closes #1070
2017-04-23 07:59:28 -07:00
Joost VandeVondele ced29248c9 simplify logic for history based pruning
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 34255 W: 6292 L: 6194 D: 21769

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 89914 W: 11769 L: 11739 D: 66406

Bench: 6581936

Closes #1066
2017-04-20 11:28:11 -07:00
Joost VandeVondele 2d96e8fbe9 Partial insertion sort
the order of elements returned by std::partition is implementation defined (since not stable) and could depend on the version of libstdc++ linked.
As std::stable_partition was tested to be too slow (http://tests.stockfishchess.org/tests/view/585cdfd00ebc5903140c6082).
Instead combine partition with our custom implementation of insert_sort, which fixes this issue.
Implementation based on a patch by mstembera (http://tests.stockfishchess.org/tests/view/58d4d3460ebc59035df3315c), which suggests some benefit by itself.
Higher depth moves are all sorted (INT_MIN version), as in current master.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33116 W: 6161 L: 6061 D: 20894

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 88703 W: 11572 L: 11540 D: 65591

Bench: 6256522

Closes #1058
Closes #1065
2017-04-20 11:19:01 -07:00
Stefano Cardanobile d4b9ee0f1d Update Readme.md
Update number of threads.

Closes #1072
2017-04-17 10:38:37 -07:00
Joost VandeVondele 57a3334322 Prefer std::find over a hand-coded loop
tested for no regression.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 37378 W: 6649 L: 6556 D: 24173

No functional change.

Closes #1071
2017-04-17 09:25:27 -07:00
VoyagerOne 9d3ed9ed2e Move-Count Formula Tweak
STC:
LLR: 3.18 (-2.94,2.94) [0.00,4.00]
Total: 55004 W: 10289 L: 9930 D: 34785

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 48184 W: 6401 L: 6128 D: 35655

Bench: 5960754
2017-04-17 09:22:10 -07:00
Stefano80 06175c6055 Remove cap from space score contribution and increase bonus
STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 58462 W: 10615 L: 10558 D: 37289

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 65061 W: 8539 L: 8477 D: 48045

It is worth noting that an attempt to only increase the bonus passed STC but failed LTC, and
an attempt to remove the cap without increasing the bonus is still running at STC, but will probably fail after more than 100k.

Bench: 6188591

Closes #1063
2017-04-17 09:19:36 -07:00
Stéphane Nicolet 1454831220 Doubled and supported pawns
Do not give the doubled pawn penalty when the frontmost pawn is
supported, for instance f2-g2-g3

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 55927 W: 10418 L: 10052 D: 35457
http://tests.stockfishchess.org/tests/view/58eb9fc20ebc59035df33858

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 32078 W: 4257 L: 4035 D: 23786
http://tests.stockfishchess.org/tests/view/58ec48420ebc59035df3388b

Bench: 5995472

Closes #1062
2017-04-16 06:22:48 -07:00
Stefano80 b258b4fee7 Remove minimum to contribution from king danger to score.
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24858 W: 4559 L: 4445 D: 15854

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40789 W: 5338 L: 5244 D: 30207

Bench: 7027489

Closes #1059
2017-04-09 07:48:58 -07:00
joergoster 72a501c6fe Fix zugzwang pruning issues
By adding pos.non_pawn_material(pos.side_to_move()) as a precondition in step 13,
which is already in use in Futility Pruning (child node) and Null Move Pruning for similar reasons.

Pawn endgames, especially those with only 1 or 2 pawns, are simply heavily influenced by zugzwang situations.

Since we are using a bitbase for KPK endgames, I see no reason to accept buggy evals as shown in #760

Patch looks neutral at STC
LLR: 2.32 (-2.94,2.94) [-3.00,1.00]
Total: 79580 W: 10789 L: 10780 D: 58011

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27071 W: 3502 L: 3390 D: 20179

Bench: 6259071

Closes #1051
Closes #760
2017-04-07 17:15:00 -07:00
VoyagerOne 35b77b120e Standardize stat penalty
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 90631 W: 16325 L: 16323 D: 57983

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 97679 W: 12779 L: 12759 D: 72141

Bench: 6340591

Closes #1053
2017-04-07 17:02:31 -07:00
VoyagerOne ebe021f6a5 Don't update TT at excluded move ply
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 38906 W: 7125 L: 6835 D: 24946

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 28600 W: 3752 L: 3543 D: 21305

Bench:  6861050

Closes #1048
2017-04-02 20:32:54 -07:00
Daniel Dugovic 06eba14dc9 Add assertion for the maximum number of pawns
No functionl change

Closes #1039
2017-03-27 15:55:48 -07:00
Joost VandeVondele c5de4080db Introduce assert for stats update
Make sure updates to the stats are done in a stable way.

No functional change

Closes #1038
Closes #1037
2017-03-25 17:57:07 -07:00
Joost VandeVondele 3b7c1a17e4 Increase maximum number of threads
a single Xeon Phi can present itself as a single numa node with up to 288 threads (4 threads per hardware core).
Tested to work as expected with a Xeon Phi CPU 7230 up to 256 threads.

No functional change

Closes #1045
2017-03-25 10:35:17 -07:00
joergoster afe75571d8 Simplify ThreatBySafePawn scoring
Bench: 6197938

Closes #1047
2017-03-25 10:22:20 -07:00
VoyagerOne 30c583204f Singular extension and check extension tweak
If singular extension fails to trigger extension then don't consider check extension.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 69428 W: 12663 L: 12271 D: 44494

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 44023 W: 5875 L: 5612 D: 32536

Bench: 6170444

Closes #1043
2017-03-25 10:13:25 -07:00
VoyagerOne 352bd6f5aa Skip quiet moves based on moveCount pruning threshold and history stats
If we can moveCountPrune and next quiet move has negative stats,
then go directly to the next move stage (Bad_Captures).

Reduction formula is tweaked to compensate for the decrease in move count that is used in LMR.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 6847 W: 1276 L: 1123 D: 4448

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 48687 W: 6503 L: 6226 D: 35958

Bench: 5919519

Closes #1036
2017-03-18 15:44:49 -07:00
Joost VandeVondele c80d52c845 History stat bonus: Move condition to bonus calculation
about 0.5% speedup.

No functional change

Closes #1034
2017-03-17 14:46:47 -07:00
joergoster c076216a32 Pawns count imbalance table
Instead of having a continuous increasing bonus for our number of pawns when calculating imbalance, use a separate lookup array with tuned values.
Idea by GuardianRM.

STC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16155 W: 2980 L: 2787 D: 10388

LTC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 100478 W: 13055 L: 12615 D: 74808

Bench: 6128779

Closes #1030
2017-03-17 14:41:08 -07:00
Marco Costalba a6d6a2c2fa Assorted code style fixes
No functional change

Closes #1029
2017-03-14 21:02:21 -07:00
mstembera d01b66ae8f Fix pawn entry prefetch
No functional change

Closes #1026
2017-03-14 20:56:26 -07:00
snicolet c3d2e6aba9 Helper functions to count material for both sides
Syntactic sugar: helper functions to count material or pieces for both sides.

No functional change

Closes #1025
2017-03-08 18:45:38 -08:00
Joost VandeVondele d490bb9973 Always have counterMoves associated
Simplifies away all associated checks, leading to a ~0.5% speedup.
The code now explicitly checks if moves are OK, rather than using nullptr checks.

Verified for no regression:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32218 W: 5762 L: 5660 D: 20796

No functional change

Closes #1021
2017-03-08 18:35:23 -08:00
pb00068 cc76524c2e Further simplify skipping of plies with threads
No functional change

Closes #1020
2017-03-08 18:03:01 -08:00
VoyagerOne 3627348e2b Allow pruning advance pawn pushes if not near endgame
STC:
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 101088 W: 18016 L: 17717 D: 65355

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 61194 W: 8108 L: 7791 D: 45295

Bench: 5803228

Closes #1023
2017-03-05 18:58:06 -08:00
snicolet 728ce2195c Speed-up some arrays reading
This patch removes the empty rows at the beginning and at the end of
MobilityBonus[] and Protector[] arrays:

• reducing the size of MobilityBonus from 768 bytes to 512 bytes
• reducing the size of Protector from 1024 to 512 bytes

Also adds some comments and cleaner code for the arrays in pawns.cpp

No speed penalty (measured speed-up of 0.4%).

No functional change.

Closes #1018
2017-03-05 18:20:27 -08:00
Joost VandeVondele 1810c4d758 Simplify skipping of plies with helper threads
Replaces the HalfDensity array with an equivalent, compact implementation.
Includes suggestions by mcostalba & snicolet.

No functional change

Closes #1004
2017-02-26 16:41:58 -08:00
snicolet 8f7e032b8c Change definition of "weak" in threats calculation
By defining "strongly protected" as "protected by a pawn, or protected
by two pieces and not attacked by two enemy pieces".

Passed STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 17050 W: 3128 L: 2931 D: 10991

Passed LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 120995 W: 15852 L: 15343 D: 89800

Bench : 6269229

Closes #1016
2017-02-25 17:43:54 -08:00
mstembera f1e3dfea74 Reorder members of Material::Entry
This eliminates alignment padding and reduces size from 48 to 40 bytes.
This makes the material HashTable smaller and more cache friendly.

No functional change

Closes #1013
2017-02-23 21:33:03 -08:00
GuardianRM 9f48e1ec15 Pieces protecting king
Initial protective idea by Snicolet for knight, for other pieces too
Patch add penalties and bonuses for pieces, depending on the distance from the own king

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 21192 W: 3919 L: 3704 D: 13569

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 26177 W: 3642 L: 3435 D: 19100

Bench : 6687377

Closes #1012
2017-02-23 21:26:59 -08:00
snicolet eefbe967c2 Keep pawns on both flanks
Positions with pawns on only one flank tend to be more drawish. We add
a term to the initiative bonus to help the attacking player keep pawns
on both flanks.

STC: yellowish run stopped after 257137 games
LLR: -0.92 (-2.94,2.94) [0.00,5.00]
Total: 257137 W: 46560 L: 45511 D: 165066

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 15602 W: 2125 L: 1956 D: 11521

Bench : 6976310

Closes #1009
2017-02-19 14:27:03 -08:00
FauziAkram c243cd5f4a Variable tuning
A tuning patch which cover the following changes:

increase the importance of queen and rook mobility in endgame and
decrease it in mg, since if we use the heavy pieces too early in the game
we will just make opponent develop their pieces by threatening ours.

King Psqt:
1)King will be encouraged more to stay in the first ranks in the MG
2)and will be encouraged more to go to the middle of the board/last ranks in the EG

Bishop scale better in EG
Logical changes on various psqt tables
1/6 of the changes of the last tuning session on mobility tables

STC: LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 227879 W: 41240 L: 40313 D: 146326
LTC : LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 167047 W: 21871 L: 21291 D: 123885

Bench: 5695960

Closes #1008
2017-02-19 14:00:44 -08:00
VoyagerOne 05cf45f2d1 Razor Simplification
Remove code that restrict using tt-moves for razoring.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 67442 W: 12039 L: 11997 D: 43406

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 38541 W: 5044 L: 4947 D: 28550

Bench: 5667216

Closes #1002
2017-02-18 22:50:37 -08:00
torfranz faedcf08a8 Retire loose enemies bonus
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 44727 W: 7943 L: 7862 D: 28922

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 148343 W: 19044 L: 19088 D: 110211

Bench: 5669076

Closes #1005
2017-02-14 21:26:08 -08:00
VoyagerOne 83fb4547f6 search(): Move nullValue variable into local scope
No functional change

Closes #1003
2017-02-14 21:22:58 -08:00
Joost VandeVondele 1e814e0ca0 Fix makefile: 32 bit builds without optimization.
Fixes failing build for

make ARCH=x86-32 clean && make ARCH=x86-32 optimize=no build

by passing -m32 also to the link step.

Extend travis testing accordingly.

No functional change.

Closes #999
2017-02-14 21:11:44 -08:00
torfranz e0d91f4c44 Retire small bonus in passed pawn evaluation
STC: http://tests.stockfishchess.org/tests/view/5899824d0ebc59099759f3ee
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 33370 W: 6061 L: 5961 D: 21348

LTC: http://tests.stockfishchess.org/tests/view/5899e3820ebc59099759f415
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 36206 W: 4664 L: 4563 D: 26979

Bench: 6072262

Closes #998
2017-02-10 16:52:34 -08:00
Joona Kiiski a753e20bd4 A small tweak in doEasyMove()
Time.elapsed() > Time.optimum() * 5 / 44
instaed of:
Time.elapsed() > Time.optimum() * 5 / 42

This was yellow on STC:
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 156856 W: 28317 L: 27942 D: 100597

Passed on LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 36909 W: 4926 L: 4682 D: 27301

Note: Patch was originally submitted by user GuardianRM.
However his repo was deleted before merge.

No functional change

Closes #995
2017-02-10 16:44:13 -08:00
Stefano80 5205d44f87 Simplify scale factor computation
Minor non-functional simplifications in computing the scale factor.

In my opinion, the code is now slightly more readable:

- remove one condition which can never be satisfied.
- immediately return instead of assigning the sf variable.

Tested for non-regression:

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 62162 W: 11166 L: 11115 D: 39881

No functional change

Closes #992
2017-02-05 16:06:37 -08:00
VoyagerOne 0553b46829 Simplify Queen Mobility
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 85997 W: 15550 L: 15540 D: 54907

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 25333 W: 3370 L: 3256 D: 18707

Bench: 6459194

Closes #991
2017-02-05 15:40:30 -08:00
Stéphane Nicolet ddecdc97d7 Simplify away QueenContactChecks
Changing the definition of safe checks to include
squares protected only by the king, but twice
attacked by the opponent.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8691 W: 1683 L: 1541 D: 5467
http://tests.stockfishchess.org/tests/view/588f53b50ebc5915193f7dc7

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 31266 W: 4150 L: 4043 D: 23073
http://tests.stockfishchess.org/tests/view/588f68ab0ebc5915193f7dda

Bench : 5885815
2017-01-31 09:54:38 +01:00
VoyagerOne fa24cc25a4 Simplify TT penalty stat (#980)
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20251 W: 3692 L: 3570 D: 12989

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 16432 W: 2155 L: 2029 D: 12248

Bench: 5941174
2017-01-29 08:54:58 +01:00
VoyagerOne 5254a6040c Penalty for a quiet ttMove that fails low
Also the penalty/bonus function is misleading, we
should simply change it to stat_bonus(depth) for
bonus and -stat_bonus(depth+ ONE_PLY) for extra
penalty.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 11656 W: 2183 L: 2008 D: 7465

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 11152 W: 1531 L: 1377 D: 8244

Bench: 6101931
2017-01-28 09:32:07 +01:00
Stefan Geschwentner 471f7a1b5c Candidate passed pawns
Detect safe candidate passers.

STC: http://tests.stockfishchess.org/tests/view/5882395c0ebc5915193f78b3
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 53569 W: 9925 L: 9570 D: 34074

LTC: http://tests.stockfishchess.org/tests/view/5882b4fb0ebc5915193f78e2
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 77576 W: 10387 L: 10014 D: 57175

Bench: 5325829
2017-01-28 09:04:24 +01:00
pb00068 58c181de9a Simplify away pinnedPieces bitboard in EvalInfo (#975)
Results for 20 tests for each version (pgo-builds):

            Base      Test      Diff      
    Mean    2110519   2118116   -7597     
    StDev   8727      4906      10112     

p-value: 0,774
speedup: 0,004

Further verified for no regression:
http://tests.stockfishchess.org/tests/view/5885abd10ebc5915193f79e6
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21786 W: 3959 L: 3840 D: 13987

No functional change
2017-01-28 08:43:54 +01:00
Alain SAVARD cf4a38e0cb Simplification of lazy threshold
Passed STC
http://tests.stockfishchess.org/tests/view/587846c10ebc5915193f74ec
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 217236 W: 39041 L: 39254 D: 138941

Passed LTC
http://tests.stockfishchess.org/tests/view/587e157a0ebc5915193f76e7
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 52396 W: 6883 L: 6804 D: 38709

This submitted version (using if (abs(mg + eg) > 1500) )
seems more logical than the following other green simplification (using if (abs(mg)>1500))
since it can happen than mg_value is > eg_value (about 20% of the time)
and the submitted version seems stronger at LTC

STC
http://tests.stockfishchess.org/tests/view/5879702d0ebc5915193f7585
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39958 W: 7315 L: 7227 D: 25416

LTC
http://tests.stockfishchess.org/tests/view/5879af3e0ebc5915193f7592
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 204322 W: 26529 L: 26648 D: 151145

bench: 6406285
2017-01-21 10:47:54 +01:00
Alain SAVARD 9eed183489 Outpost array simplification
The ReachableOutpost values were almost exactly half the Outpost values.

Passed STC
http://tests.stockfishchess.org/tests/view/588020510ebc5915193f781e
LLR: 3.86 (-2.94,2.94) [-3.00,1.00]
Total: 119238 W: 21462 L: 21460 D: 76316

Passed LTC
http://tests.stockfishchess.org/tests/view/5880ae090ebc5915193f7843
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 23540 W: 3097 L: 2980 D: 17463

Curiously, using a division by 2, with slightly different values, did not passed
http://tests.stockfishchess.org/tests/view/587fece00ebc5915193f780a

bench: 5828283
2017-01-21 10:36:46 +01:00
Alain SAVARD 243a9f5484 Reformat eval_init()
Move more code into eval_init, removing some
clutter in the main routine.

Write eval_init only from "our" point of view
(do not init the attackedBy[Them] bitboards).

Add mobilityArea to the evalinfo

A few edits while being there

tested for non-regression at STC
http://tests.stockfishchess.org/tests/view/587fab230ebc5915193f77d9
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39585 W: 7183 L: 7094 D: 25308

Non functional change.
2017-01-21 10:13:49 +01:00
Stéphane Nicolet 9f8f093fd6 Update some comments (#973)
Use somewhat more precise comments in a couple of places.

No functional change.
2017-01-17 14:50:03 +01:00
Stefano Cardanobile de02768af7 Introduce lazy evaluation
After we have taken into account all cheap evaluation
terms, we check whether the score exceeds a given threshold.
If this is the case, we return a scaled down evaluation.

STC:
LLR: 3.35 (-2.94,2.94) [0.00,5.00]
Total: 12575 W: 2316 L: 2122 D: 8137

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 67480 W: 9016 L: 8677 D: 49787

Current version is the one rewritten by ceebo
further edited by me.

Bench: 5367704
2017-01-13 09:17:48 +01:00
loco-loco 99cd513264 Removing CM parameter from the Stats struct. (#970)
After the history simplifications, we are only using Value Stats for CounterMoveHistory table. Therefore the parameter CM is not necessary.

No functional change.
2017-01-12 08:46:46 +01:00
Stéphane Nicolet d2971f3fca Check for overflow in Score * int multiplication (#969)
Add asserts to check for overflow in Score * int multiplication.

There is no overflow in current master, but it would be easy to
create one as the scale of the current eval does not leave many
spare bits. For instance, adding the following unused variables
in master at the end of evaluate() (line 882 of evaluate.cpp)
overflows:

Score s1 = score * 4;  // no overflow
Score s2 = score * 5;  // overflow

Assertion failed: (eg_value(result) == (i * eg_value(s))),
function operator*, file ./types.h, line 336.

Same md5 checksum as current master for non debug compiles.

No functional change.
2017-01-11 18:11:17 +01:00
Rocky640 d40351243b StormDanger CleanUp (#964)
Order the enum and the array the same way they appear around line 250.
Makes it much easier to follow.

Add comments in the array definition and critical rows.
Use same terminology as elsewhere in pawns.cpp

No functional change.
2017-01-11 08:56:38 +01:00
Joost VandeVondele d8f683760c Adjust copyright headers to 2017 (#965)
No functional change.
2017-01-11 08:46:29 +01:00
Marco Costalba 332b5013b5 Travis: fix bench fetch in case of PR (#968)
When Travis tests a PR, a commit merge is created
but master branch is not updated, although HEAD is.

No functional change.
2017-01-11 08:44:06 +01:00
lucasart 34e47ca87d Rename FromTo -> History (#963)
Previously, we had duplicated History:

- one with (piece,to) called History
- one with (from,to) called FromTo

Now that we have only one, rename it to History, which is the generally accepted
name in the chess programming litterature for this technique.

Also correct some comments that had not been updated since the introduction of CMH.

No functional change.
2017-01-10 08:47:56 +01:00
lucasart e0504ab876 Remove HistoryStats
STC:
LLR: 3.44 (-2.94,2.94) [-3.00,1.00]
Total: 120831 W: 21572 L: 21594 D: 77665

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26565 W: 3519 L: 3406 D: 19640

bench 5920493
2017-01-09 15:50:12 +01:00
Marco Costalba d9dd520896 Fix previous patch for OS X (#961)
Use posix version of sed that is available on all
platforms.

No functional change.
2017-01-09 15:37:09 +01:00
Marco Costalba 394e9cd892 Trevis CI: use commit bench number as a reference
No functional change.
2017-01-09 12:51:28 +01:00
Joost VandeVondele ba15781be8 New shell scripts for testing, used for travis CI (#957)
Perform more complex verification and validation.

- signature.sh : extract and optionally compare Bench/Signature/Node count.
- perft.sh : verify perft counts for a number of positions.
- instrumented.sh : run a few commands or uci sequences through valgrind/sanitizer instrumented binaries.
- reprosearch.sh : verify reproducibility of search.

These script can be used from directly from the command line in the src directory.

Update travis script to use these shell scripts.

No functional change.
2017-01-09 10:30:57 +01:00
ElbertoOne d39ffbeea6 Simplified select best thread (#958)
Only select best thread if score is better and depth equal or larger.

STC (7 threads): http://tests.stockfishchess.org/tests/view/586a4d090ebc5903140c64b2
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 120297 W: 18652 L: 18682 D: 82963

LTC (7 threads): http://tests.stockfishchess.org/tests/view/586e31b30ebc5903140c663d
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 57187 W: 7035 L: 6959 D: 43193

bench: 4940355
2017-01-09 10:26:50 +01:00
Marco Costalba 3ab3e55bb5 Drop Stats c'tors
Now taht we correctly value-initialize Thread objects,
we don't need c'tors anymore because tables will be
zero-initialized by the compier when Thread object
is instanced.

Verified that we have no errors with Valgrind.

No functional change.
2017-01-07 10:14:39 +01:00
Joost VandeVondele 1c316c41bb Correctly zero-initialize MainThread
It can be used uninitialized in time management.
Fixes all valgrind errors on './stockfish go wtime 8000 btime 8000 winc 500 binc 500'

This is one (of the many) quirks of C++. There is a subtle difference between:

new Foo
new Foo()

The first statement calls the default constructor (that in case of a POD leaves data members
uninitialized), the second one performs a value-initialization (that in case of POD is
equivalent to a zero-initialization)

See:
http://stackoverflow.com/questions/620137/do-the-parentheses-after-the-type-name-make-a-difference-with-new
http://stackoverflow.com/questions/5116541/difference-between-creating-object-with-or-without

No functional change.
2017-01-07 10:02:43 +01:00
Joost VandeVondele 90b052462c Zero init fromToStats in constructor. (#953)
Extend commit fe99de to fromToStats, which fixes the last valgrind errors on 
a simple 'go depth 12' at startup.

No functional change.
2017-01-06 10:43:18 +01:00
pb00068 8b2c81d3ea Rejoin lines that belong to HalfDensity map (#952)
No functional change.
2017-01-05 09:00:41 +01:00
Marco Costalba fe99de20ff Correct zero-init of Thread data members
If not explicitly initialized in a class constructor,
then all data members are default-initialized when
the corresponing struct/class is instanced.

For array and built-in types (int, char, etc..)
default-initialization is a no-op and we need to
explicitly zero them.

No functional change.
2017-01-05 08:50:17 +01:00
Joost VandeVondele 6b16ebc825 Use consistent variable names for counterMoveStats (#949)
Unify naming in movepick and search, by adopting the latter convention (cmh,fmh,fmh2).

No functional change.
2017-01-02 09:36:43 +01:00
Stéphane Nicolet e3a8b8bcff Simplify unstoppable again (#950)
Assign a small bonus for our passed pawns when the opponent has no
pieces left.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 51124 W: 9036 L: 8966 D: 33122

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 114560 W: 14604 L: 14604 D: 85352

Bench: 4940355
2017-01-02 09:33:40 +01:00
Jonathan Calovski 1052ce74f6 Tweak best thread selection logic
STC 7 threads:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 26881 W: 4161 L: 3941 D: 18779
http://tests.stockfishchess.org/tests/view/58667a830ebc5903140c632f

LTC 7 threads:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 22988 W: 2767 L: 2583 D: 17638
http://tests.stockfishchess.org/tests/view/586722690ebc5903140c636d

bench: 5468995
2017-01-01 11:28:50 +01:00
lucasart e258c5a779 WDL: rename WDLCursedLoss into WDLBlessedLoss
Tested using syzygy bench method:

- 2016 random positions ranging between 3 and 10 pieces
- each searched using bench at depth=10

Same node count (and no speed regression).

No functional change.
2017-01-01 11:11:52 +01:00
Aram Tumanian b7b9d7c9c7 Don't clear EasyMove in search()
EasyMove is cleared after every iteration of the
search if the 3rd move in the PV of the main thread
changes from the previous iteration. Therefore,
clearing EasyMove during a search iteration may be
excessive. The tests show that this is indeed unnecessary.
In the new version the EasyMove variable is used only in
the Thread::search function.

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 47719 W: 8438 L: 8362 D: 30919

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 122841 W: 15448 L: 15457 D: 91936

bench: 5468995
2017-01-01 11:04:24 +01:00
Sergei Antonov 881a9dfb0a Threefold repetition detection
Implement a threefold repetition detection. Below are the examples of
problems fixed by this change.

    Loosing move in a drawn position.
    position fen 8/k7/3p4/p2P1p2/P2P1P2/8/8/K7 w - - 0 1 moves a1a2 a7a8 a2a1
    The old code suggested a loosing move "bestmove a8a7", the new code suggests "bestmove a8b7" leading to a draw.

    Incorrect evaluation (happened in a real game in TCEC Season 9).
    position fen 4rbkr/1q3pp1/b3pn2/7p/1pN5/1P1BBP1P/P1R2QP1/3R2K1 w - - 5 31 moves e3d4 h8h6 d4e3
    The old code evaluated it as "cp 0", the new code evaluation is around "cp -50" which is adequate.

Brings 0.5-1 ELO gain. Passes [-3.00,1.00].

STC: http://tests.stockfishchess.org/tests/view/584ece040ebc5903140c5aea
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 47744 W: 8537 L: 8461 D: 30746

LTC: http://tests.stockfishchess.org/tests/view/584f134d0ebc5903140c5b37
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 36775 W: 4739 L: 4639 D: 27397

Patch has been rewritten into current form for simplification and
logic slightly changed so that return a draw score if the position
repeats once earlier but after or at the root, or repeats twice
strictly before the root. In its original form, repetition at root
was not returned as an immediate draw.

After retestimng testing both version with SPRT[-3, 1], both passed
succesfully, but this version was chosen becuase more natural. There is
an argument about MultiPV in which an extended draw at root may be sensible.
See discussion here:

   https://github.com/official-stockfish/Stockfish/pull/925

For documentation, current version passed both at STC and LTC:

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 51562 W: 9314 L: 9245 D: 33003

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 115663 W: 14904 L: 14906 D: 85853

bench: 5468995
2017-01-01 10:56:46 +01:00
Alain SAVARD 43f6b33e50 Small eval cleanup and renaming
Non-functional changes

a) splitting the threat array to avoid using an enum
b) reorder the scores according to functions where they are used.
c) declarations in evaluate_pieces after the const(s) like elsewhere
d) more compact definitions of KingFlank,
now that we need it also for the PanwLessFlank penalty.
e) reuse CenterFiles in evaluate_space
f) move one line inside next popcount

No functional change.
2016-12-31 14:15:57 +01:00
lucasart ab4f498bbc Remove SafeCheck (#946)
It was a bit of a hack, without intrinsic value, but rather compensating for the
fact that checks were mistuned.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 88308 W: 15553 L: 15545 D: 57210

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 53115 W: 6741 L: 6662 D: 39712

bench 5468995
2016-12-31 13:52:42 +01:00
Jörg Oster 8765f9ce16 Further simplify unstoppable (#938)
By finally moving it into passed pawns eval.

Tested for no regression:
STC
LLR: 3.25 (-2.94,2.94) [-3.00,1.00]
Total: 57109 W: 10023 L: 9947 D: 37139

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21511 W: 2800 L: 2680 D: 16031

Bench: 5255881
2016-12-25 10:44:56 +01:00
hxim 3728e833aa Fix psqt format and use smaller numbers in king psqt (#940)
Fix minus sign in pawn psqt and use smaller numbers in king psqt.

No functional change.
2016-12-25 10:40:17 +01:00
Joost VandeVondele af7412e58a Explicitly use alpha+1 for beta in NonPV search (#939)
Fixes the only exception, in razoring.

The code already does assert(PvNode || (alpha == beta - 1)), and it can be verified by studying the program flow that this is indeed the case, also for the modified line.

No functional change.
2016-12-25 10:34:48 +01:00
Joost VandeVondele 1ceaea701b Simplify threshold handling for probcut. (#936)
Just use greater equal as this is what see_ge does now.

passed STC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 226506 W: 39755 L: 39978 D: 146773

passed LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 138483 W: 17450 L: 17479 D: 103554

Bench: 5212921
2016-12-22 16:02:32 +01:00
Stefano Cardanobile f72b7dc99a piecesCount (#932)
All counts in search.cpp are of the form xxxCount. Conform piecesCnt to this unwritten rule.

No functional change.
2016-12-20 11:18:19 +01:00
Joost VandeVondele ee22b61f5e Use DEPTH_ZERO initializer for depth in qsearch (#931)
Simplifies the main search function.

No functional change.
2016-12-20 11:17:38 +01:00
VoyagerOne 8c61bbda54 Another simplification for SEE pruning
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 29310 W: 5225 L: 5118 D: 18967

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 128454 W: 16424 L: 16442 D: 95588

Bench: 4556848
2016-12-17 08:37:58 +01:00
Stefan Geschwentner 847bc0e80f Simplify pruning
STC: http://tests.stockfishchess.org/tests/view/5842be140ebc5903140c5619
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16014 W: 2839 L: 2710 D: 10465

LTC: http://tests.stockfishchess.org/tests/view/584316a50ebc5903140c5638
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 182360 W: 22830 L: 22914 D: 136616

Retested at LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 45502 W: 5821 L: 5732 D: 33949

Bench: 4684146
2016-12-12 12:01:16 +01:00
Joost VandeVondele 1b62d413c1 Clean-up skipEarlyPruning (#921)
make skipEarlyPruning a search argument instead of managing this by hand.

Verified for no regression at STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 96754 W: 17089 L: 17095 D: 62570

No functional change.
2016-12-11 20:05:25 +01:00
Jonathan Calovski 589049a0e5 Simplify unstoppable condition
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 86389 W: 15165 L: 15153 D: 56071

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 98143 W: 12311 L: 12288 D: 73544

Bench: 5437987
2016-12-11 19:57:02 +01:00
Joost VandeVondele a47bbca0ea Refactor bonus and penalty calculation (#917)
* Refactor bonus and penalty calculation

Compute common terms in a helper function.

No functional change.

* Further refactoring

Remove some parenthesis that are now useless.
Define prevSq once, use repeatedly.

No functional change.

bench: 5884767 (bench of previous patch is wrong)
2016-12-05 18:58:12 +01:00
Stéphane Nicolet 0d33466bcd Pawn flank attacks
This patch tweaks some pawn values to favor flank attacks.

The first part of the patch increases the midgame psqt values of external pawns to launch more attacks (credits to user GuardianRM for this idea), while the second part increases the endgame connection values for pawns on upper ranks.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 34997 W: 6328 L: 6055 D: 22614

and LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 13844 W: 1832 L: 1650 D: 10362

Bench: 5884767
2016-12-05 18:49:07 +01:00
ElbertoOne 46d066b041 Remove piece condition in decrease lmr reduction check
STC: http://tests.stockfishchess.org/tests/view/584154780ebc5903140c55cf
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40866 W: 7251 L: 7164 D: 26451

LTC: http://tests.stockfishchess.org/tests/view/5841e6e50ebc5903140c5605
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 62413 W: 7948 L: 7881 D: 46584

Bench: 5807913
2016-12-04 09:28:55 +01:00
Andrey Neporada bf8b45fe63 Help GCC to optimize msb() to single instruction
GCC compiles builtin_clzll to “63 ^ BSR”. BSR is processor instruction "Bit Scan Reverse".
So old msb() function is basically 63 - 63 ^ BSR.
Unfortunately, GCC fails to simplify this expression.

Old function compiles to

    bsrq    %rdi, %rdi
    movl    $63, %eax
    xorq    $63, %rdi
    subl    %edi, %eax
    ret

New function compiles to

    bsrq    %rdi, %rax
    ret

BTW, Clang compiles both function to the same (optimal) code.

No functional change.
2016-12-03 09:37:07 +01:00
goodkov e70da0d2eb Simplify pruning rule
STC: http://tests.stockfishchess.org/tests/view/583df86d0ebc5903140c5481
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 43283 W: 7761 L: 7678 D: 27844

LTC: http://tests.stockfishchess.org/tests/view/583f42670ebc5903140c5525
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 56587 W: 7232 L: 7157 D: 42198

bench: 5084980
2016-12-02 09:04:45 +01:00
Joost VandeVondele 1e76ba7cec WeakQueen Parameter tweak
New tuned values.

passed STC
http://tests.stockfishchess.org/tests/view/5834573c0ebc5903140c507b
LLR: 3.16 (-2.94,2.94) [0.00,4.00]
Total: 157415 W: 27917 L: 27227 D: 102271

passed LTC
http://tests.stockfishchess.org/tests/view/58388d2b0ebc5903140c523b
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 95405 W: 12350 L: 11959 D: 71096

Bench: 4912054
2016-12-01 14:55:00 +01:00
ElbertoOne 535435b7fc TrappedRook simplification
Just remove rank checks for rook and king for TrappedRook evaluation.

STC: http://tests.stockfishchess.org/tests/view/5833fdfc0ebc5903140c5050
LLR: 3.03 (-2.94,2.94) [-3.00,1.00]
Total: 34474 W: 6088 L: 5986 D: 22400

LTC: http://tests.stockfishchess.org/tests/view/58392f3f0ebc5903140c5276
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 48358 W: 6221 L: 6136 D: 36001

bench: 5536128
2016-11-27 17:42:20 +01:00
theo77186 7a3844e6ef Fix PGO build with GCC (#904) 2016-11-27 14:43:52 +01:00
Michael Byrne fbb2ffacfd Fix PGO Build for clang
This fixes https://github.com/official-stockfish/Stockfish/issues/167.

Additional improvments by Joost VandeVondele.
2016-11-27 10:03:52 +01:00
joergoster 8f30d233f8 Fix trace in case of space evaluation
We only compute space eval during the opening/early midgame.
Apply the same logic for DoTrace.

No functional change.
2016-11-27 09:28:32 +01:00
Marco Costalba ec83e8a72c Fix regression: print const position
Fix a regression introduced with new TB code.

No functional change.
2016-11-27 09:11:56 +01:00
mbootsector e7289465b9 Rank based threats
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19404 W: 3581 L: 3374 D: 12449

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 16204 W: 2194 L: 2023 D: 11987

Bench: 5757843
2016-11-27 09:05:57 +01:00
Joost VandeVondele 8ceb1ff53b Fix undefined behavior
This fixes #892. Undefined behavior as seen with
clang -fsanitize=undefined.

No functional change.
2016-11-26 16:49:59 +01:00
Marco Costalba 2ec626ddae Fix compile under Windows XP
The needed Windows API for processor groups could be missed from old Windows
versions, so instead of calling them directly (forcing the linker to resolve
the calls at compile time), try to load them at runtime. To do this we need
first to define the corresponding function pointers.

Also don't interfere with running fishtest on numa hardware with Windows.
Avoid all stockfish one-threaded processes will run on the same node

No functional change.
2016-11-26 07:04:17 +01:00
Aram Tumanian 9eccba7761 Fix the pawn hash failure when the pawn key is 0
This patch fixed bugs #859 and #882.
At initialization we generate a new random key (Zobrist::noPawns).
It's added to the pawn key of all positions, so that the pawn key
of a pawnless position is no longer 0.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21307 W: 3738 L: 3618 D: 13951

LTC:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 45270 W: 5737 L: 5648 D: 33885

No functional change.
2016-11-25 08:48:35 +01:00
erbsenzaehler ca464fc89e Cleanup Makfile for MacOs
1) Explicitly setting the default lib to the system-default is not
   needed on a Mac. See:
   http://libcxx.llvm.org/docs/UsingLibcxx.html

2) We do no longer need to exclude bmi2-builds from LTO. See:
   https://llvm.org/bugs/show_bug.cgi?id=19416

Changes tested and discussed on FishCooking:
   https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/acUQtKtEzMM

No functional change.
2016-11-25 08:46:20 +01:00
Marco Costalba 0d9a9f5e98 Handle Windows Processors Groups
Under Windows it is not possible for a process to run on more than one
logical processor group. This usually means to be limited to use max 64
cores. To overcome this, some special platform specific API should be
called to set group affinity for each thread. Original code from Texel by
Peter sterlund.

Tested by Jean-Paul Vael on a Xeon E7-8890 v4 with 88 threads and confimed
speed up between 44 and 88 threads is about 30%, as expected.

No functional change.
2016-11-22 07:56:04 +01:00
Joost VandeVondele 6036303bb6 Avoid touching source files in profile-build
This refines the profile-build target to avoid 'touch'ing the sources,
keeping meaningful modification dates and avoiding editor warnings like vi's:

WARNING: The file has been changed since reading it!!!
Do you really want to write to it (y/n)?

Instead of touching sources, the (instrumented) object files are removed,
which has the same effect of rebuilding them in the next step.

As a side effect, this simplifies the Makefile a bit.

No functional change.
2016-11-20 10:51:42 +01:00
Fabian Beuke b5d10d17c2 Reduce variable scope in swap_byte
Added a specialization to remove the 'if' condition

No functional change.
2016-11-19 21:31:24 +01:00
Aram Tumanian 797602938d Start searching for a repetition from the 4th ply behind
A position can never repeat the one on the previous move.
Thus we can start searching for a repetition from the 4th
ply behind. In the case:

 std::min(st->rule50, st->pliesFromNull) < 4

We don't need to do any more calculations. This case happens
very often - in more than a half of all calls of the function.

No functional change.
2016-11-19 10:20:28 +01:00
Alain SAVARD 76d113f5f0 Pawn shelter and pawn storm tuned
Based on SPSA tuned values

Passed STC
http://tests.stockfishchess.org/tests/view/582363b30ebc5910626b9ca8
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 40628 W: 7380 L: 7087 D: 26161

and passed LTC
http://tests.stockfishchess.org/tests/view/5823b73b0ebc5910626b9cb5
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 273312 W: 35991 L: 35131 D: 202190

bench: 5773672
2016-11-19 09:42:11 +01:00
Marco Costalba 18df1698f4 Fix compile error from previous patch
Due to different types on some platforms.

No functional change.
2016-11-19 09:35:57 +01:00
Stéphane Nicolet 7f4de0196b Do not use GCC extension for anonymous unions
Anonymous struct inside anonymous unions are a GCC extension.
This patch uses named structs to stick to the C+11 standard.

Avoids a string of warnings on the Clang compiler.

Non functional change (same bench and same MD5 signature,
so compiled code is exactly the same as in current master)
2016-11-19 09:22:49 +01:00
Joost VandeVondele cddc8d4546 More accurate 'go nodes' searches at low count
Makes the actual number of nodes searched match closely
the number of nodes requested, by increasing the frequency
of checking the number of nodes searched at low node count.
All other searches retain the default checking frequency of
once per 4096 nodes, and are thus unaffected.

Passed STC as non-regression
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 26643 W: 4766 L: 4655 D: 17222

No functional change.
2016-11-19 08:47:41 +01:00
Stefan Geschwentner a90fc4c877 Non-quiet pruning tweak
Count in the difference of static evaluation
and alpha for pruning threshold.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 16885 W: 3061 L: 2866 D: 10958

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 33566 W: 4428 L: 4202 D: 24936

Bench: 5513149
2016-11-19 08:37:52 +01:00
Aram Tumanian e6c2899020 Make a version of Position::do_move() without the givesCheck parameter
In 10 of 12 calls total to Position::do_move()the givesCheck argument is
simply gives_check(m). So it's reasonable to make an overload without
this parameter, which wraps the existing version.

No functional change.
2016-11-12 09:55:12 +01:00
joergoster de269ee18e FEN parsing: add a second check for correctly setting e.p. square
Currently, we only check if there is a pawn in place
to make the en-passant capture. Now also check that
there is a pawn that could just have advanced two
squares. Also update the corresponding comment.

This makes the parsing of FENs a bit more robust, and
now correctly handles positions like the one reported by Dann Corbit.

position fen rnbqkb1r/ppp3pp/3p1n2/3P4/8/2P5/PP3PPP/RNBQKB1R w KQkq e6
d

 +---+---+---+---+---+---+---+---+
 | r | n | b | q | k | b |   | r |
 +---+---+---+---+---+---+---+---+
 | p | p | p |   |   |   | p | p |
 +---+---+---+---+---+---+---+---+
 |   |   |   | p |   | n |   |   |
 +---+---+---+---+---+---+---+---+
 |   |   |   | P |   |   |   |   |
 +---+---+---+---+---+---+---+---+
 |   |   |   |   |   |   |   |   |
 +---+---+---+---+---+---+---+---+
 |   |   | P |   |   |   |   |   |
 +---+---+---+---+---+---+---+---+
 | P | P |   |   |   | P | P | P |
 +---+---+---+---+---+---+---+---+
 | R | N | B | Q | K | B |   | R |
 +---+---+---+---+---+---+---+---+

Fen: rnbqkb1r/ppp3pp/3p1n2/3P4/8/2P5/PP3PPP/RNBQKB1R w KQkq - 0 1

No functional change.
2016-11-10 11:45:51 +01:00
Mira dc4655e1f9 Stack offset changed from -5 to -4
Non functional change, tests under sanitizers OK.

Rationales for change

- Offset in code is in range -4 ... 2
- There was an error by (pathological) corner case MAX_PLY=0

No functional change.
2016-11-10 11:43:21 +01:00
atumanian 0fa80c9ba3 Update comments related after new see_ge()
Update comments according to changes from my patch: #822

No functional change.
2016-11-10 11:40:31 +01:00
Joost VandeVondele 61c727fdcb Allow benches with more than 2G nodes.
./stockfish bench 128 1 4000000000 default nodes

    crashes before, works after.

    No functional change.
2016-11-07 13:35:28 +01:00
Joost VandeVondele 876f07cbee Fix undefined behaviour with unaligned loads in syzygy code
Casting a pointer to a different type with stricter alignment
requirements yields to implementation dependent behaviour.
Practicaly everything is fine for common platforms because the
CPU/OS/compiler will generate correct code, but anyhow it is
better to be safe than sorry.

Testing with dbg_hit_on() shows that the unalignment accesses are
very rare (below 0.1%) so it makes sense to split the code in a
fast path for the common case and a slower path as a fallback.

No functional change (verified with TB enabled).
2016-11-06 11:48:07 +01:00
ppigazzini f5d3f0ded6 Update AUTHORS for SF8
And format top contriutor (space instead of tabs, proper Unix
line endings).
2016-11-06 10:28:17 +01:00
Marco Costalba e4659693de Fix a warning with debug=no
Warning in TB code due to unused variable.

Verified same bench with TB code enabled.

No functional change.
2016-11-06 09:45:25 +01:00
Joost VandeVondele 52cb348023 Reproducible searches after ucinewgame
Fixes issue #859.

thisThread->callsCnt in search<>() was different (by 1) for the first and second game played.

No functional change.
2016-11-05 22:25:55 +01:00
Marco Costalba 0d669be76c Restore development version
No functional change.
2016-11-05 09:32:39 +01:00
Marco Costalba 23f384cac3 Add explicit braces and fix a warning
Warning under both gcc and clang.

No functional change.
2016-11-05 09:29:22 +01:00
VoyagerOne b915fdc889 Reduction Simplification
Simplify reduction formula by removing a parameter.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 35798 W: 6368 L: 6272 D: 23158

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 125375 W: 15827 L: 15839 D: 93709

Bench: 4735038
2016-11-05 08:17:42 +01:00
Joost VandeVondele 67d19447f4 Makefile fix for sanitize
Small fixes for compilation with sanitize=yes optimize=no,
by always adding -fsanitize=undefined to the LDFLAGS as required.
Updates config-sanity to check&report the status of the flag.

No functional change.
2016-11-05 08:15:56 +01:00
VoyagerOne d4abf54247 Top CPU Contributors
Give some well needed credit...

No functional change.
2016-11-05 08:12:22 +01:00
Miroslav Fontán 445aade39c Simplify code, delete else after return 2016-11-05 08:08:25 +01:00
Miroslav Fontán f3cd7002aa Sync variable names in decl vs def 2016-11-05 08:05:22 +01:00
Joost VandeVondele c43146edb7 Travis-ci: integrate perft checking
makes verifying perft numbers for a few positions
part of travis-ci. Adds <5s testing time.

No functional change.
2016-11-05 08:03:34 +01:00
Marco Costalba c0bb041539 Rewrite syzygy in C++
Rewrite the code in SF style, simplify and
document it.

Code is now much clear and bug free (no mem-leaks and
other small issues) and is also smaller (more than
600 lines of code removed).

All the code has been rewritten but root_probe() and
root_probe_wdl() that are completely misplaced and should
be retired altogheter. For now just leave them in the
original version.

Code is fully and deeply tested for equivalency both in
functionality and in speed with hundreds of games and
test positions and is guaranteed to be 100% equivalent
to the original.

Tested with tb_dbg branch for functional equivalency on
more than 12M positions.

stockfish.exe bench 128 1 16 syzygy.epd

Position: 2016/2016
Total 12121156 Hits 0 hit rate (%) 0
Total time (ms) : 4417851
Nodes searched : 1100151204
Nodes/second : 249024

Tested with 5,000 games match against master, 1 Thread,
128 MB Hash each, tc 40+0.4, which is almost equivalent
to LTC in Fishtest on this machine. 3-, 4- and 5-men syzygy
bases on SSD, 12-moves opening book to emphasize mid- and endgame.

Score of SF-SyzygyC++ vs SF-Master: 633 - 617 - 3750  [0.502] 5000
ELO difference: 1

No functional change.
2016-11-05 07:55:08 +01:00
Marco Costalba 369eff437c Stockfish 8
Bench: 5926706

No functional change
2016-11-01 10:19:17 +01:00
Stefan Geschwentner 344616e917 Bonus for attacked passed pawn promotion path
Bonus for each attacked square on the promotion path
of an enemy passed pawn.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 141511 W: 25295 L: 24579 D: 91637

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 23161 W: 3022 L: 2831 D: 17308

Bench: 5926706
2016-10-30 13:48:03 +01:00
Joost Vandevondele 40b1b27178 Fix a series of undefined behaviours
Avoid shifting negative signed integers and use typed
enum to avoids decrementing a variable beyond its defined
range, like:
       for (Rank r = RANK_8; r >= RANK_1; --r)

Changes were tested individually and passed SPRT[-3, 1].

With this patch gcc --sanitize builds cleanly.

No functional change.
2016-10-27 06:44:41 +02:00
Joost VandeVondele bf51b4796a travis-ci: Enable undefined behavior checking 2016-10-27 06:29:24 +02:00
syzygy e18e557e77 Output PV if last iteration does not complete
Instead of outputting "info nodes ... time ..." when the last
iteration is interrupted, simply call UCI::pv() to output the PV.

I thought about calling UCI:pv() with bounds -VALUE_INFINITE, VALUE_INFINITE
to avoid "lowerbound" or "upperbound" appearing in it, but I'm not sure that
would be any better.

This patch fixes rare inconsistencies between the first move of
the last PV output and the bestmove played. It also makes sure
that all the latest statistics are sent to the GUI (not only nodes
and time but also nps, tbhits, hashfull).

No functional change.
2016-10-27 06:26:22 +02:00
Stéphane Nicolet 818b4a126d Endgame malus for having a king in a pawnless flank
Original idea by "ElbertoOne", while "FauziAkram" suggested to put a
small midgame penalty too.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 71808 W: 13038 L: 12610 D: 46160

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 150874 W: 19828 L: 19221 D: 111825

Bench: 6077005
2016-10-25 06:57:29 +02:00
VoyagerOne e77f38c431 History Stat Comparison
Adjust LMR by comparing history stats
with opponent (prior ply).

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 27754 W: 5066 L: 4824 D: 17864

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 216596 W: 28157 L: 27343 D: 161096

Bench: 5437729
2016-10-25 06:44:19 +02:00
Marco Costalba e18321f55a Correcty resey TB hit counter
Restore original behaviour to reset
the counter before a new move search.

Also fixed some warnings and added const
qualifier to a couple of functions, as
suggested by m_stembera.

Thanks to Werner Bergmans for reporting
the regression.

No functional change.
2016-10-22 08:22:13 +02:00
syzygy ca67752645 Per-thread TB hit counters
Use a per-thread counter to reduce contention
with many cores and endgame positions.

Measured around 1% speed-up on a 12 core and 8%
on 28 cores with 6-men, searching on:
 7R/1p3k2/2p2P2/3nR1P1/8/3b1P2/7K/r7 b - - 3 38

Also retire the unused set_nodes_searched() and fix
a couple of return types and naming conventions.

No functional change.
2016-10-21 06:15:45 +02:00
Joost Vandevondele 3686e719a1 Simplify next_move by always scoring evasions
For a default bench, this fixes the last valgrind
error (jump on uninitialised value).

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 187869 W: 33303 L: 33463 D: 121103

No functional change.
2016-10-20 17:17:14 +02:00
Joost Vandevondele 9893e7fd53 Make valgrind testing part of travis ci. 2016-10-19 21:29:00 +02:00
ajithcj 99f3ad6858 Remove useless assignments to currentMove
We reference (ss-1)->currentMove, i.e. we peek
current move of the parent node, so currentMove
should be valid in the main move loop, when we
search() the subtree, but outside of main loop
it is useless.

No functional change.
2016-10-18 09:00:52 +02:00
VoyagerOne e27d3bb884 Use explicit logic for pruning
Also a speedup since we don't need to recalculate SEE
for extensions...as it already determined to be positive.

Results for 12 tests for each version:

        Base      Test      Diff
Mean    2132395   2191002   -58607
StDev   128058    85917     134239
p-value: 0.669
speedup: 0.027

Non functional change.
2016-10-18 08:53:51 +02:00
Jacques 16e1881126 Fixes for ARM compilation: take 2
The target:

Odroid U3 (http://www.hardkernel.com/main/products/prdt_info.php?g_code=g138745696275)
Debian Jessie
As listed in #550 and #638 three modifications are needed for compilation to work:

float-abi flag for GCC If an FPU is present and supported by the installed os then passed value need to be hard.
I didn't find any better solution than using readelf to check for the availibilty of Tag_ABI_VFP_args which sould indicate support for the FPU. The check is only done if the arch is arm and if readelf is not present
on the system, there will be an error (/bin/sh: 1: readelf: not found) but it will not break and will continue with the default softfp value. Outputing the error is not really acceptable but I wanted some feedback on the
check itself.

-lpthread is needed on armv7 outside of Android
I replaced UNAME with KERNEL and OS to allow to differentiate Android.

m32 flag
My understanding is that outside of Android the flag is generating errors on armv7.

These modifications should introduce change only for non Android armv7 build.

No functional change.
2016-10-14 08:58:07 +02:00
Marco Costalba e1f600f186 Revert "Fixes for ARM compilation"
This reverts commit a3fe80c36a.

Break compilation on mingw for me.
2016-10-13 08:36:30 +02:00
Jacques a3fe80c36a Fixes for ARM compilation
The target:

Odroid U3 (http://www.hardkernel.com/main/products/prdt_info.php?g_code=g138745696275)
Debian Jessie
As listed in #550 and #638 three modifications are needed for compilation to work:

float-abi flag for GCC If an FPU is present and supported by the installed os then passed value need to be hard.
I didn't find any better solution than using readelf to check for the availibilty of Tag_ABI_VFP_args which sould indicate support for the FPU. The check is only done if the arch is arm and if readelf is not present
on the system, there will be an error (/bin/sh: 1: readelf: not found) but it will not break and will continue with the default softfp value. Outputing the error is not really acceptable but I wanted some feedback on the
check itself.

-lpthread is needed on armv7 outside of Android
I replaced UNAME with KERNEL and OS to allow to differentiate Android.

m32 flag
My understanding is that outside of Android the flag is generating errors on armv7.

These modifications should introduce change only for non Android armv7 build.

No functional change.
2016-10-13 08:34:04 +02:00
Marco Costalba fdf3a51c68 AppVeyor: run bench after build
And show resulting bench signature.

The run is very slow becuase optimizations
are all disabled by default /Od /RTC1

No functional change.
2016-10-10 21:00:59 +02:00
Marco Costalba e61f7b1e6d Add AppVeyor integration
It is like Trevis CI but for Windows platform.

Currently just compile builds, wthouth benching
the resulting executable.

No functional change.
2016-10-10 16:29:29 +02:00
ajithcj f799610d4b Simplify futility pruning return value
Return eval as it is while doing futility pruning.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 167687 W: 29778 L: 29904 D: 108005

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26905 W: 3503 L: 3390 D: 20012

Bench: 5936728
2016-10-09 09:54:43 +02:00
atumanian 073eed590e Optimisation of Position::see and Position::see_sign
Stephane's patch removes the only usage of Position::see, where the
returned value isn't immediately compared with a value. So I replaced
this function by its optimised and more specific version see_ge. This
function also supersedes the function Position::see_sign.

bool Position::see_ge(Move m, Value v) const;

This function tests if the SEE of a move is greater or equal than a
given value. We use forward iteration on captures instread of backward
one, therefore we don't need the swapList array. Also we stop as soon
as we have enough information to obtain the result, avoiding unnecessary
calls to the min_attacker function.

Speed tests (Windows 7), 20 runs for each engine:
Test engine: mean 866648, st. dev. 5964
Base engine: mean 846751, st. dev. 22846
Speedup: 1.023

Speed test by Stephane Nicolet

Fishtest STC test:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 26040 W: 4675 L: 4442 D: 16923
http://tests.stockfishchess.org/tests/view/57f648990ebc59038170fa03

No functional change.
2016-10-08 06:38:36 +02:00
Stéphane Nicolet 1e586288ca Do not use SEE in evasion scoring
Idea by Aram Tumanian (atumanian)

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 43889 W: 7849 L: 7767 D: 28273

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 29333 W: 3809 L: 3700 D: 21824

Bench: 6421663
2016-10-06 00:00:27 +02:00
Stefano Cardanobile 0162fb83c2 Retire implicit malus for stonewalls
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 75864 W: 13466 L: 13437 D: 48961

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 99050 W: 12472 L: 12451 D: 74127

bench: 6098474
2016-10-05 09:32:08 +02:00
VoyagerOne ab26c61971 Allow inCheck pruning
This is a bit tricky because we don't want
to prune the only legal evasions, even if
with negative SEE. So add an assert to avoid
this subtle bug to slip in later.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 14140 W: 2625 L: 2421 D: 9094

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 11558 W: 1555 L: 1379 D: 8624

bench: 5256717
2016-10-03 16:18:53 +02:00
Marco Costalba eccccba0ce Remove useless razoring condition
Condition is always true! For any value of the
array index! Even an out of bound array, like
razor_margin[120]!!!!

No functional change.
2016-09-29 15:24:36 +02:00
HiraokaTakuya b77bae0529 Make razor_margin[4] ONE_PLY value independent
No functional change.
2016-09-29 15:20:07 +02:00
Stéphane Nicolet 7ae3c05795 Rename shift_bb() to shift()
Rename shift_bb() to shift(), and DELTA_S to SOUTH, etc.
to improve code readability, especially in evaluate.cpp
when they are used together:

    old b = shift_bb<DELTA_S>(pos.pieces(PAWN))
    new b = shift<SOUTH>(pos.pieces(PAWN))

While there fix some small code style issues.

No functional change.
2016-09-25 10:45:10 +02:00
joergoster 351844061e Allowing singular extension in mate positions
Drop useless condition

abs(ttValue) < VALUE_KNOWN_WIN

And extend singular extension search to cases when ttValue
stores a mate score. This improves mate finding and does
not introduce any regression.

Yery tested this patch against current master on the 6500+
Chest mate suite with 200K fixed nodes:

    shortest mates found: master: 1206 patch:1205
    any mate found: master: 1903 patch: 2003

with 1 sec time:

    shortest mates found: master: 2667 patch: 2628
    any mate found: master: 3585 patch: 3646

Verified for no regression:

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25655 W: 4578 L: 4465 D: 16612

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 66247 W: 8618 L: 8557 D: 49072

bench: 6335042
2016-09-24 19:56:02 +02:00
Marco Costalba 8662bdfa12 Fix crash when passing a mate/stalemate position
Both Tablebases::filter_root_moves() and
extract_ponder_from_tt(9 were unable to handle
a mate/stalemate position.

Spotted and reported by Dann Corbit.

Added some mate/stalemate positions to bench so
to early catch this regression in the future.

No functional change.
2016-09-24 07:37:52 +02:00
Stéphane Nicolet 28240d375c Simplify pinners conditions in SEE()
Use the following transformations:

- to check that A is included in B, testing "(A & ~B) == 0" is faster
than "(A & B) == A"

- to remove the intersection of A and B from A, doing "A &= ~B;" is as
fast as "if (A & B) A &= ~B;" but is simpler.

Overall, the simpler patch version is 0.3% than current master.

No functional change.
2016-09-22 08:31:23 +02:00
Guenther Demetz 943ae89be1 Fix pin-aware SEE
Correct pinners calculation and fix bug with pinned
pieces giving check. With this patch 'pinners' only
returns sliders with exactly one defensive piece between
the slider and the attacked square (in other words, pinners
returns exact pinners).

This was a co-operation between Marco Costalba,
Stphane Nicolet and me.

Special thanks to Ronald de Man for reporting the bug with
pinned pieces giving check, discussed here:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/S_4E_Xs5HaE

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 132118 W: 23578 L: 23645 D: 84895

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 36424 W: 4770 L: 4670 D: 26984

bench: 6272231
2016-09-21 08:42:25 +02:00
Joost Vandevondele 4b0043ae7c Use fixed depth bench to make PGO builds more reproducible
Discussed on fishcooking

proposal and objdump verification:
https://groups.google.com/d/msg/fishcooking/4_ausUwMXP0/EGPsMYqOFAAJ

verified no significant speed difference between depth and time:
https://groups.google.com/d/msg/fishcooking/4_ausUwMXP0/KazW5QZmFgAJ

stockfish_time - stats:
mean = 2207232.56        std = 7079.51        std/mean = 0.003207

stockfish_depth - stats:
mean = 2201783.57        std = 6356.69        std/mean = 0.002887

No functional change
2016-09-18 08:13:34 +02:00
Marco Costalba 92f01aa2bd Fix a warning with MSVC
warning C4706: assignment within conditional expression

No functional change.
2016-09-17 10:14:28 +02:00
Stéphane Nicolet ea41f18e6e Swap mg and eg in internal representation of Score
Instrumentation shows that in make_score(mg, eg) calls, the mg value is
zero in 25,9% of the calls while the eg value is zero in 36,8% of the
calls.

Swapping the internal fields of mg and eg in the internal
representation of Score allows the compiler to optimize away the shift
in (eg << 16) + mg in more cases, thus resulting in a 0.3% speed-up
overall.

No functional change
2016-09-17 09:56:36 +02:00
Marco Costalba 057d710fc2 Fix indentation in struct FromToStats
And other little trivial stuff.

No functional change.
2016-09-17 09:51:20 +02:00
Stéphane Nicolet 01f2466f6e Retire KingDanger array
Rescales the king danger variables in evaluate_king() to
suppress the KingDanger[] array. This avoids the cost of
the memory accesses to the array and simplifies the non-linear
transformation used.

Full credits to "hxim" for the seminal idea and implementation,
see pull request #786.
https://github.com/official-stockfish/Stockfish/pull/786

Passed STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 9649 W: 1829 L: 1689 D: 6131

Passed LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 53494 W: 7254 L: 7178 D: 39062

Bench: 6116200
2016-09-16 08:30:06 +02:00
Marco Costalba 5c58d1f5cb Use per-thread counterMoveHistory
Drops a scalability bottleneck due to memory contention
of a single shared table across threads. The effect starts
to be sensible with a high number of threads. Specifically
we have a small regression with 7 threads both at 60 and
180 seconds TC:

10000 @ 60+0.6 th 7
ELO: -2.46 +-3.2 (95%) LOS: 6.5%
Total: 9896 W: 1037 L: 1107 D: 7752

5000 @ 180+0.6 th 7
ELO: -1.95 +-4.1 (95%) LOS: 17.7%
Total: 5000 W: 444 L: 472 D: 4084

We have a regression because counterMoveHistory table is
quite big and it takes time for a single thread to fill it.
Sharing the table yields to a higher fill rate and better
quality of moves and up to 7 threads the benefits of sharing
more then compensate the loss in speed due to contention.
Interestingly even with a 3X longer TC, so with more time
for the single thread to catch up, the improvment is quite
limited and below noise level. It seems we really need much
longer TC to saturate the table.

When we move to high threads number it's another story:

5000 @ 60+0.6 th 22
ELO: 3.49 +-4.3 (95%) LOS: 94.6%
Total: 4880 W: 490 L: 441 D: 3949

2000 @ 60+0.6 th 32
ELO: 8.34 +-6.9 (95%) LOS: 99.1%
Total: 2000 W: 229 L: 181 D: 1590

As expected the speed-up more than compensates the filling
rate, and we expect that with tournament TC, where single
thread is able to saturate the table, the difference will
be even stronger. For instance for TCEC 9 super-final time
control will be 180 minutes + 15 seconds and this scalability
improvement seems definitely the way to go.

So, summarizing:

GOOD:

Measured big improvement in high core scenario

Suitable for TCEC 9 superfinal (big hardware, very long TC)

Consistent and natural patch that extends to counterMoveHistory
what we already do for remaining history tables, that are all per-thread

Non functional change for the common case of a single core

Very simple (just 6 lines modified, no added ones)

BAD:

Small regression (within 2-3 ELO) with few threads and short TC

bench: 5341477
2016-09-16 08:15:07 +02:00
Marco Costalba b96dd754ed Renaming in MovePicker
Rename stages and simplify a bit the code.

No functional change.
2016-09-15 09:07:49 +02:00
Marco Costalba 01ee509a5c Retire MovePicker::see_sign()
No more used after last patch.

No functional change.
2016-09-14 15:43:56 +02:00
VoyagerOne 95ad2b51b7 Tweak SEE margin in pruning conditions
Use 35 * depth^2 to calculate see_margin.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 22636 W: 4212 L: 3990 D: 14434

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 47241 W: 6314 L: 6041 D: 34886

The Movepick SEE is now dead code, retire it.

Bench: 5341477
2016-09-14 15:38:38 +02:00
syzygy 438805aee8 Integrate next_stage() logic into next_move()
Measured bench speed up goes from 0,7% to 2%,
given the unreliable measure a reverse simmplification
test was done on fishtest:

master vs patch
LLR: -2.94 (-2.94,2.94) [-3.00,1.00]
Total: 15499 W: 2685 L: 2867 D: 9947

Test result is positive, master is weaker.

No functional change.
2016-09-13 07:14:09 +02:00
Guenther Demetz ace8e951d7 Simplify code for pinaware SEE
This is the most compact and neatest version
is was able to produce.

On normal builds I have a small slowdown:
normal builds base vs. simplification (gcc 4.8.1 Win7-64 i7-3770 @ 3.4GHz x86-64-modern)
Results for 20 tests for each version:

        Base      Test      Diff
Mean    1974744   1969333   5411
StDev   11825     10281     5874
p-value: 0,178
speedup: -0,003

On pgo-builds however I measure a nice 1.1% speedup

pgo-builds base vs. simplification
Results for 20 tests for each version:

        Base      Test      Diff
Mean    1974119   1995444   -21325
StDev   8703      5717      4623
p-value: 1
speedup: 0,011

No functional change.
2016-09-12 15:45:00 +02:00
Guenther Demetz 90ce24b11e Pinned aware SEE
Don't allow pinned pieces to attack the exchange-square as long all
pinners (this includes also potential ones) are on their original
square.
As soon a pinner moves to the exchange-square or get captured on it, we
fall back to standard SEE behaviour.

This correctly handles the majority of cases with absolute pins.

bench: 6883133
2016-09-12 09:31:09 +02:00
Stefano Cardanobile 4c95edddbf Reorder evaluation start
In evaluate, we start by initializing the pos.psq_score
and adding the material imbalance. After that, we check
whether a specialized eval exists and if yes we return
that value and discard whatever we have computed until now.

It sounds more logical to first probe material entry and
return if we have a specialized eval, and only if it is
not the case initialize eval with some values. There is
no measurable speed-difference on my computer.

Non functional change.
2016-09-11 07:42:12 +02:00
Marco Costalba 602d7fbb07 Use Movepick SEE value in search
This halves the calls to the costly pos.see_sign(),
speed up is about 1-1.3%

Non functional change.
2016-09-09 17:11:54 +02:00
Marco Costalba d909d10f33 Refactor previous patch
No functional change.
2016-09-08 06:02:42 +02:00
ajithcj 38428ada54 Prune dangerous moves at low depth
At very low depths prune captures,
promotions and checks if see is negative.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 6772 W: 1328 L: 1173 D: 4271

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8917 W: 1270 L: 1122 D: 6525

bench: 6024713
2016-09-08 05:55:10 +02:00
Marco Costalba e340ce221c Syntactic sugar to loop across pieces
Also add some comments to the new operator~(Piece).

No functional change.
2016-09-04 15:33:17 +02:00
syzygy ca6c9f85a5 Change from [Color][PieceType] to [Piece]
Speed up of almost 1% in both normal and
pgo builds.

No functional change.
2016-09-04 09:22:09 +02:00
Marco Costalba c5828c4eba Fix syzygy with partial TB
In case we have installed a not complete set of 6-men tables and
there is 6 piece position on board, but no corresponding
tablebase engine is not using any syzygy at all.

Reported by Jouni Uski, fix by Peter Österlund,
confirmed as a bug by Ronald de Man.

bench: 7591630
2016-09-03 08:21:05 +02:00
Stéphane Nicolet d37dfe9ae4 Space bonus in presence of open files
If the opponent has a cramped position, opening a file often
helps him/her to exchange pieces, so it makes sense to reduce
the space bonus if there are open files.

Credits: Leonardo Ljubičić for the strategic idea, Alain Savard for the
implementation of the open files calculation, "CrunchyNYC" for the
compensation of the numerator.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 49112 W: 9239 L: 8900 D: 30973

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 89415 W: 12014 L: 11601 D: 65800

Bench: 7591630
2016-09-03 00:04:20 +02:00
lucasart 13b4444d9e Change exclusion key setup
Should depend on which move is excluded. This
allow us to remove the dedicated Position::exclusion_key().

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 59814 W: 11136 L: 11083 D: 37595

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 31023 W: 4187 L: 4080 D: 22756

bench 7553379
2016-09-02 08:37:01 +02:00
Stefano80 7f2eb10e93 Retire linear imbalance
Retire linear imbalance and compensate
in piece values enumeration.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 43596 W: 8105 L: 8023 D: 27468

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24482 W: 3352 L: 3237 D: 17893

Bench: 7777707
2016-09-02 08:25:17 +02:00
ajithcj 5cffc032da Optimize order of a few conditions in search
Also fix size of KingDanger array to reduce memory footprint.

Small speed up of around 0.5%

No functional change.
2016-08-31 13:47:45 +02:00
VoyagerOne 2731bbaf6b Remove condition on killers in history pruning
Now allows main killer to be history prune.

STC:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 15852 W: 2910 L: 2781 D: 10161

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 56428 W: 7610 L: 7537 D: 41281

Bench: 8032058
2016-08-30 09:09:55 +02:00
Stefan Geschwentner 6aa9308f08 Tweak probcut threshold
Use better threshold for capture move generation.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 23265 W: 4415 L: 4188 D: 14662

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 36618 W: 5083 L: 4836 D: 26699

bench: 7030088
2016-08-29 22:15:03 +02:00
Andrew Grant b4f6728e61 Removed an extra space
No functional change.
2016-08-28 09:47:30 +02:00
Alain SAVARD 2b57b61cb1 Move king tropism to evaluate_king
No functional change.
2016-08-28 08:49:40 +02:00
Marco Costalba 1ee2838214 Retire CheckInfo
Move its content directly under StateInfo.

Verified for no speed regression.

No functional change.
2016-08-28 08:08:13 +02:00
Marco Costalba 0b944c7186 Silence some warnings with MSVC 2013
No functional change.
2016-08-27 12:16:13 +02:00
Stéphane Nicolet 805afcbf3d Move CheckInfo under StateInfo
This greately simplifies usage because hides to the
search the implementation specific CheckInfo.

This is based on the work done by Marco in pull request #716,
implementing on top of it the ideas in the discussion: caching
the calls to slider_blockers() in the CheckInfo structure,
and simplifying the slider_blockers() function by removing its
first parameter.

Compared to master, bench is identical but the number of calls
to slider_blockers() during bench goes down from 22461515 to 18853422,
hopefully being a little bit faster overall.

archlinux, gcc-6
make profile-build ARCH=x86-64-bmi2
50 runs each

bench:
base = 2356320 +/- 981
test = 2403811 +/- 981
diff = 47490 +/- 1828

speedup = 0.0202
P(speedup > 0) = 1.0000

perft 6:
base = 175498484 +/- 429925
test = 183997959 +/- 429925
diff = 8499474 +/- 469401

speedup = 0.0484
P(speedup > 0) = 1.0000

perft 7 (but only 10 runs):
base = 185403228 +/- 468705
test = 188777591 +/- 468705
diff = 3374363 +/- 476687

speedup = 0.0182
P(speedup > 0) = 1.0000

$ ./pyshbench ../Stockfish/master ../Stockfish/test 20
run base     test     diff
...

base = 2501728 +/- 182034
test = 2532997 +/- 182034
diff = 31268 +/- 5116

speedup = 0.0125
P(speedup > 0) = 1.0000

No functional change.
2016-08-27 09:53:26 +02:00
Marco Costalba 4c5cbb1b14 Make engine ONE_PLY value independent
This non-functional change patch is a deep work to allow SF to be independent
from the actual value of ONE_PLY (currently set to 1). I have verified SF is
now independent for ONE_PLY values 1, 2, 4, 8, 16, 32 and 256.

This patch gives consistency to search code and enables future work, opening
the door to safely tweaking the ONE_PLY value for any reason.

Verified for no speed regression at STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 95643 W: 17728 L: 17737 D: 60178

No functional change.
2016-08-27 09:12:25 +02:00
gamander 133808851d Fixed wrong definition of WhiteCamp and BlackCamp
No functional change.
2016-08-27 08:48:07 +02:00
hxim 9bd856e28e Simplify stats update
Simplify code by moving countermove and follow-up move
history update into procedure.

No functional change.
2016-08-27 08:39:32 +02:00
Marco Costalba 3e739f883f Reformat stats update
Rewritten in a way to have explicit in the search
the bonus/penalty we apply: hopefully this will lead
to further simplification/fix of current rather messy
stats update code.

No functional change.
2016-08-25 09:51:50 +02:00
VoyagerOne 0b4f6e562b Refutation penalty on captures
Apply refutation penalty for prior PV quiet move on captures

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8208 W: 1153 L: 1008 D: 6047
http://tests.stockfishchess.org/tests/view/57bc5a9f0ebc59030fbe47b5

Only LTC because a very similar patch already passed STC + LTC

bench: 7038730
2016-08-24 08:26:32 +02:00
VoyagerOne 5596492f6e Simplify IID
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 30468 W: 5687 L: 5582 D: 19199
http://tests.stockfishchess.org/tests/view/57b1ddd80ebc591c761f63e2

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 87406 W: 11756 L: 11725 D: 63925
http://tests.stockfishchess.org/tests/view/57b212590ebc591c761f63f9

bench: 6554900
2016-08-19 08:59:18 +02:00
VoyagerOne 8493c8c6b8 Do LMR on captures
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 5361 W: 1086 L: 936 D: 3339
http://tests.stockfishchess.org/tests/view/57b31b0f0ebc591c761f643d

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 54694 W: 7591 L: 7287 D: 39816
http://tests.stockfishchess.org/tests/view/57b3442b0ebc591c761f6450

bench: 6881120
2016-08-19 08:50:14 +02:00
Marco Costalba e3af492142 Remove a stale assignment
No more used after previous patch.

Spotted by Jekaa .

No functional change.
2016-08-18 10:40:56 +02:00
Stefano80 9585f8ef58 Retire pawn span
Retire pawn span and replace with pawn count in evaluate_scale_factor.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 26482 W: 4929 L: 4818 D: 16735

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 61938 W: 8400 L: 8335 D: 45203

Bench: 7662861
2016-08-18 10:16:56 +02:00
VoyagerOne 7396c08b79 Use predicted depth for history pruning
STC: (Yellow)
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 69115 W: 12880 L: 12797 D: 43438

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 124163 W: 16923 L: 16442 D: 90798

Note: Note based off past experiments / patches... history pruning
is quite TC sensitive. I believe the reason for this TC dependency
is that the CMH/FMH is a very large table that takes time to fill
up with. In addition having more time for will increase the accuracy
of the stats' value.

Bench: 7351698
2016-08-15 14:45:40 +02:00
Luca Brivio 3d10cfcdd2 Cap space evaluation bonus
When computing space evaluation, limit the bonus square count to 16.

STC @ 10+0.1 th 1:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 30793 W: 5910 L: 5648 D: 19235

LTC @ 60+0.6 th 1:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 31361 W: 4410 L: 4184 D: 22767

Bench: 7165385
2016-08-13 10:12:09 +02:00
Alain SAVARD 8abb98455f Simplify space formula
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 16868 W: 3260 L: 3132 D: 10476

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 16910 W: 2381 L: 2255 D: 12274

bench: 6663531
2016-08-10 12:49:24 +02:00
VoyagerOne b3525fa9ea Use Color-From-To history stats to help sort moves
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 33502 W: 6498 L: 6223 D: 20781
http://tests.stockfishchess.org/tests/view/578abb940ebc5972faa169e2

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 50782 W: 7124 L: 6832 D: 36826
http://tests.stockfishchess.org/tests/view/578b8e5d0ebc5972faa169fd

LTC: (Sanity test against latest master)
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 32759 W: 4600 L: 4370 D: 23789
http://tests.stockfishchess.org/tests/view/5798b7d30ebc591c761f5b72

bench: 6985912

P.S. Thanks @mstembera for rewriting my code to make it smp compatible. A BIG thank you!
2016-08-02 09:17:14 +02:00
VoyagerOne 85924db496 Futility tweak
Use a different margin for pruning child nodes.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16692 W: 3251 L: 3051 D: 10390
http://tests.stockfishchess.org/tests/view/579b95d10ebc591c761f5c03

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 24140 W: 3501 L: 3297 D: 17342
http://tests.stockfishchess.org/tests/view/579bb15d0ebc591c761f5c0b

Bench: 7927017
2016-08-01 21:47:42 +02:00
ajithcj f2f3a06a1a Allow null pruning at depth 1
This removes a check that prevents null pruning at depth 1 PLY.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 23445 W: 4638 L: 4521 D: 14286

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 61416 W: 8627 L: 8563 D: 44226

bench: 8145304
2016-07-24 09:54:07 +02:00
Stefano80 714329dbdc See prune at higher depth
Allow SEE pruning at higher depths in shallow depth
pruning using a threshold increasing with depth.

STC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 35366 W: 7011 L: 6724 D: 21631

LTC
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 15578 W: 2243 L: 2070 D: 11265

Bench: 8417887
2016-07-24 08:55:58 +02:00
ajithcj 76971d8acb Gradually relax the NMP staticEval check
Gradually relax the NMP staticEval check as we go to
higher depths.

Use tuned values.

STC
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16745 W: 3371 L: 3168 D: 10206

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 5906 W: 875 L: 736 D: 4295

bench: 8548212
2016-07-16 07:35:31 +02:00
Lyudmil Antonov f619f1d38c Workaround gcc stack alignment bug
GCC uses SSE instructions to move data but in 32-bit gcc version used
by abrok the stack is not 16-byte aligned due to a bug.

This patch workaround teh bug by not using the stack
to store KingFlank[]

Fixes issue #721.

No functional change.
2016-07-16 07:06:13 +02:00
joergoster 82d02a3133 Fix extract_ponder_from_tt()
Checking for legality of a possible ponder move
must be done before we undo the first pv move,
of course. (spotted by mohammed li.)

This obviously only has any effect when playing in ponder mode.

No functional change.
2016-07-12 08:47:21 +02:00
Stéphane Nicolet 1d09ee70f7 King tropism
Bonus for each square that we attack in the flank where the opponent
king is. Squares that we attack twice and are not protected by an enemy
pawn count double.

Passed STC:
http://tests.stockfishchess.org/tests/view/577dfca60ebc5972faa166b8
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 48373 W: 9832 L: 9481 D: 29060

And LTC:
http://tests.stockfishchess.org/tests/view/577e77870ebc5972faa166df
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 8881 W: 1408 L: 1255 D: 6218

Bench: 7577046
2016-07-08 09:35:22 +02:00
Alain SAVARD eb20a87c67 More safe checks
Consider a check given by a rook or a minor to be a "safe check"
also in the case where supported by another piece,
and given on a square only defended by a queen

Was yellow STC
http://tests.stockfishchess.org/tests/view/576fcbc80ebc5972faa163e8
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 55453 W: 10431 L: 10315 D: 34707

Passed LTC
http://tests.stockfishchess.org/tests/view/57733a0b0ebc5972faa164b7
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 54550 W: 7671 L: 7365 D: 39514

bench: 7398346
2016-07-07 08:39:58 +02:00
ajithcj ade3bb9a4e Use staticEval in null prune condition
Don't null prune at depth < 12 if staticEval < beta

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 62858 W: 12035 L: 11632 D: 39191

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 49784 W: 7009 L: 6720 D: 36055

bench: 8054611
2016-07-02 16:43:49 +02:00
loco-loco 716a145a6c Removing inCheck condition for counter move bonus
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20206 W: 3946 L: 3823 D:

LTC:
LLR: 3.10 (-2.94,2.94) [-3.00,1.00]
Total: 25004 W: 3512 L: 3390 D: 18102

Bench: 8172428
2016-06-26 10:25:05 +02:00
Marco Costalba 190d2ea4bc Restore standard passed pawn definition
Use the usual and accepted passed pawn semantic
instead of a non-standard one and remove a FIXME.

STC (http://tests.stockfishchess.org/tests/view/576401350ebc5972faa1608d):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 29646 W: 5663 L: 5557 D: 18426

LTC (http://tests.stockfishchess.org/tests/view/5764e4e90ebc5972faa160c3):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 40224 W: 5578 L: 5484 D: 29162

bench: 7543902
2016-06-25 11:46:17 +02:00
ElbertoOne c2c0e6b07d Remove redundant PvNode condition
After commit 6d58bf777c we always call PvNodes
with cutNode set to false.

No functional change.
2016-06-24 08:46:36 +02:00
VoyagerOne c94145b65c Comment out a redundant condition
Take advantage that VALUE_NONE = 32002 to remove
the condition.

Commented out and not removed becuase it is tricky
to rely on the hidden value of VALUE_NONE and code
can break in case we change VALUE_NONE in the future.

No functional change.
2016-06-24 08:26:39 +02:00
Jonathan Calovski 09efbf915e Remove scalefactor dependency
STC
http://tests.stockfishchess.org/tests/view/5764539e0ebc5972faa160a4
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 43878 W: 8289 L: 8208 D: 27381

LTC
http://tests.stockfishchess.org/tests/view/5764f0130ebc5972faa160c9
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39338 W: 5408 L: 5313 D: 28617

bench: 7977279
2016-06-21 09:01:39 +02:00
ElbertoOne 6d58bf777c On IID do not always search with cutNode = true
On IID now search with cutNode value instead of fixed value true.

STC (http://tests.stockfishchess.org/tests/view/575fa3860ebc5972faa15f67):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 163974 W: 30744 L: 30874 D: 102356

LTC (http://tests.stockfishchess.org/tests/view/5763b0640ebc5972faa16075):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 48363 W: 6611 L: 6528 D: 35224

Bench: 7806393
2016-06-19 18:04:22 +02:00
VoyagerOne 1c0c4db677 Simplify Check Extension
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32704 W: 6146 L: 6045 D: 20513

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 146622 W: 19967 L: 20017 D: 106638

Bench: 8245662
2016-06-18 08:32:25 +02:00
ajithcj 8e45e70e55 Don't insert pv back into tt
This code was added before the accurate pv patch, when
we retrieved PV directly from TT.

It's not required for correct (and long) PVs any more and
should be safe to remove it.

Also, allowing helper threads to repeatedly over-write
TT doesn't seem to make sense(that was probably an un-intended
side-effect of lazy smp). Before Lazy SMP only Main Thread used
to run ID loop and insert PV into TT.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 74346 W: 13946 L: 13918 D: 46482

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 47265 W: 6531 L: 6447 D: 34287

bench: 8819179
2016-06-15 08:55:40 +02:00
lucasart 126036abb0 Do not hardcode Debug Log File
Allow to specifiy the log file name, this comes
handy in case of self-matches so that each SF
instance writes into a different log file.

No functional change.
2016-06-15 08:47:08 +02:00
Marco Costalba ca14345ba2 Filter root moves filter before copy to threads
Currently root moves are copied to all teh threads
but are DTZ filtered only in main thread at the
beginning of teh search.

This patch moves the TB filtering before the
copy of root moves fixing issue #679

https://github.com/official-stockfish/Stockfish/issues/679

No bench change.
2016-06-11 09:24:40 +02:00
VoyagerOne 7d2a79f037 Stat Formula Tweak
bonus = d * d + 2 * d - 2

STC:
LLR: 2.94 (-2.94,2.94) [0.00,4.00]
Total: 99444 W: 18274 L: 17778 D: 63392

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 89757 W: 12285 L: 11896 D: 65576

bench: 8276130
2016-06-10 07:27:47 +02:00
VoyagerOne 7c5d724724 Tweak check extension condition
There are two concepts with this patch:

Limit check extensions by using move count.
The idea is to limit search explosion.

Always extend check if the first move gives check.
The idea is to save expensive SEE calls, since the vast
majority of first move will have SEE value >= 0, also
first move may still be strong even if the SEE is negative.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16503 W: 3068 L: 2873 D: 10562

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 37202 W: 5261 L: 5014 D: 26927

bench: 8543366
2016-06-10 07:15:56 +02:00
Alain SAVARD 6fed8ff22a Small Queen simplification
Moving a few lines from evaluate_threats to evaluate_pieces allows to
a) Remove a condition pos.count<QUEEN>(Them) == 1
b) use precalculated s instead of pos.square(Them)
c) do not check the condition at all in queenless endings

Passed STC
http://tests.stockfishchess.org/tests/view/5752e0410ebc59029919b1f4
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 67175 W: 12194 L: 12152 D: 42829

and LTC
http://tests.stockfishchess.org/tests/view/57587b140ebc59029919b2f4
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 20276 W: 2774 L: 2653 D: 14849

bench: 7907962
2016-06-10 06:53:07 +02:00
mstembera 0c076f1136 Avoid some redundant scaling function calls
Posted by Mohammed Li here:
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/N-PHfN0O79o

No functional change.
2016-06-10 06:43:37 +02:00
ElbertoOne e48c7547c5 LMR reduction parameter tweak
More reduction for cut nodes, less for moves that escape a capture:

STC (http://tests.stockfishchess.org/tests/view/57548c1e0ebc59029919b247):
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 60165 W: 11519 L: 11149 D: 37497

LTC (http://tests.stockfishchess.org/tests/view/57555b570ebc59029919b260):
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 10353 W: 1493 L: 1317 D: 7543

Bench: 8902859
2016-06-07 15:15:49 +02:00
Marco Costalba 6e2ca97d93 Fix syzygy DTZ bug
In this position: 3K4/8/3k4/8/4p3/4B3/5P2/8 w - - 0 5

Current DTZ probe returns 1 instead of 15

What happens is that the double push f4 is erroneously detected as a win move.

After the push we have:

[D]3K4/8/3k4/8/4pP2/4B3/8/8 b - f3 0 5

And here the code misses the possible ep capture exf3.

The bug is in probe_dtz_no_ep() where is used probe_ab() that is
blind to ep captures so it returns v == 2 (win) for position

3K4/8/3k4/8/4pP2/4B3/8/8 b - f3 0 5

Note that at the caller site the original position did not have any
possible ep capture, so probe_dtz() returns immediately after calling
probe_dtz_no_ep().

The fix is to call the ep-aware probe_wdl() instead of probe_ab()

I have verified that DTZ is correct now and also there are no more
mistmatches compared to the new 'syzygy' branch. Tested on a set of
more than 600 endgame positions, included some tricky ones.

For people interested to redo the test or doing additional tests
please pull branch tb_dbg from https://github.com/mcostalba/Stockfish repo.

bench: 8450534 (bench unaffected because syzygy is not exercized during bench)
2016-06-07 15:06:26 +02:00
VoyagerOne 5f096e9bef Simplify Futility Pruning
Don't update bestValue when futility pruning.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21933 W: 4031 L: 3912 D: 13990

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 46225 W: 6115 L: 6028 D: 34082

Bench: 8450534
2016-06-03 19:58:42 +02:00
ElbertoOne 20023ac9b8 LMR Simplification
LMR simplification that also gives a slight ELO gain, especially at LTC:

STC (http://tests.stockfishchess.org/tests/view/574ec8e20ebc59029919b147):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32402 W: 5967 L: 5866 D: 20569

LTC (http://tests.stockfishchess.org/tests/view/574fbebf0ebc59029919b16d):
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 15103 W: 2103 L: 1975 D: 11025

Bench: 8248133
2016-06-03 19:53:04 +02:00
joergoster 3549d98d07 Tuned values for piece check and attack unit factors
A middle ground patch of two successful tuning patches,
one at STC, the other at LTC, which now passed both.

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 67893 W: 12777 L: 12384 D: 42732

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 30165 W: 4189 L: 3960 D: 22016

bench: 9209507
2016-06-03 07:39:27 +02:00
Stéphane Nicolet 07b247f943 Pins or discovered attacks on the opponent's queen
Bonus for pins or discovered attacks on the opponent's queen

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 32020 W: 5914 L: 5652 D: 20454

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 10946 W: 1530 L: 1375 D: 8041

Bench: 7031649
2016-05-28 14:57:58 +02:00
Stéphane Nicolet abac509ccb Teach check_blockers to check also non-king pieces
This is a prerequisite for next patch

No functional change.
2016-05-28 14:52:21 +02:00
Stéphane Nicolet ab0f4c0353 Simplify doubled pawn
Only use doubled pawn malus when the doubled pawns are on consecutive squares.

Passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7678 W: 1469 L: 1325 D: 4884

And LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 26739 W: 3562 L: 3449 D: 19728

Bench: 8211685
2016-05-26 12:33:44 +02:00
Leonid Pechenik 71bfbb22fc More detailed dependence of time allocation on the magnitude of score change
10+0.1:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 5657 W: 1130 L: 979 D: 3548

60+0.6:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 36884 W: 5002 L: 4762 D: 27120

bench: 8428997
2016-05-20 19:44:50 +02:00
loco-loco 7cb8cbb403 Assorted pruning tweaks
LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 38257 W: 5206 L: 4961 D: 28090

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 16550 W: 3110 L: 2914 D: 10526

Bench: 8428997
2016-05-20 19:34:49 +02:00
mstembera 0784bd542b Fix a multiPV bug in lazy SMP
Where the helper threads were not doing multiPV at all.

Regression tested sprt @ 5+0.05 th 7

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 73918 W: 11891 L: 11853 D: 50174

bench: 8716243
2016-05-14 21:34:55 +02:00
Stéphane Nicolet 16c603ce9b Double pawn simplification
Try doubled pawn simplification, with psq
table compensation.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 36094 W: 6558 L: 6463 D: 23073

LTC:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 102352 W: 13417 L: 13404 D: 75531

Bench: 8716243
2016-05-13 13:01:59 +02:00
loco-loco 969982406c Merge good and bad quiets
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 58613 W: 10779 L: 10723 D: 37111

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33608 W: 4539 L: 4436 D: 24633

Bench: 9441294
2016-05-10 18:15:16 +02:00
Alain SAVARD 4b9ed6566a Unsafe checks
Introducing a new multi-purpose penalty related to King safety, which
includes all kind of potential checks (from unsafe or unavailable
squares currently occupied by some other piece)

This will indirectly detect and reward some pins, discovered checks, and
motifs such as square vacation, or rook behind its pawn and aligned with
King (example Black Rg8, g7 against Kg1),
and penalize some pawn blockers (if they move, it allows a discovered
check by the pawn).

And since it looks also at protected squares, it detects some potential
defense overloading.

Finally, the rook contact checks had been removed some time ago. This
test will give a small bonus for them, as well as for bishop contact
checks.

Passed STC
http://tests.stockfishchess.org/tests/view/5729ec740ebc59301a354b36
LLR: 2.94 (-2.94,2.94) [0.00,5.00]
Total: 13306 W: 2477 L: 2296 D: 8533

and LTC
http://tests.stockfishchess.org/tests/view/572a5be00ebc59301a354b65
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 20369 W: 2750 L: 2565 D: 15054

bench: 9298175
2016-05-06 20:04:57 +02:00
Marco Costalba 5e4cd3fc0d Retire __popcnt64 intrinsic
Just use _mm_popcnt_u64() that is available
both for MSVC abd Intel compiler.

Verified on MSVC that the produced assembly
has the hardware 'popcnt' instruction.

No functional change.
2016-05-05 09:09:07 +02:00
VoyagerOne 5486911e01 Simplify History LMR Formula
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41713 W: 7589 L: 7504 D: 26620

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41353 W: 5484 L: 5391 D: 30478

Bench: 8946983
2016-05-05 08:53:50 +02:00
Marco Costalba 3487eb9f9e Fix a warning with MSVC
Introduced by 2dd24dc4e6 ("Use popcount intrinsic with Intel")

No functional change.
2016-05-01 15:10:33 +02:00
joergoster dc0030de4f Fix LazySMP when searching to a fixed depth.
Currently, helper threads will only search up to the
specified depth limit. Now let them search until the
main thread has finished the specified depth.

On the other hand, we don't want to pick a thread with
a higher search depth.

This may be considered cheating. ;-)

No functional change.
2016-05-01 14:30:50 +02:00
erbsenzaehler 2dd24dc4e6 Use popcount intrinsic with Interl compiler
It seems that icc used our fallback version of popcount.
Now use intrinsics.

icc version 16.0.2 (gcc version 5.3.0 compatibility)
bmi2 compile
uname -r 4.5.1-1-ARCH

20xbench gives a nice speedup
./stockfish-icc-master 2161515 +- 34462
./stockfish-icc-sse42 2260857 +- 50349
2016-05-01 14:18:16 +02:00
Krgp 8f934dff9a Remove useless -mbmi flag in Makefile
I could not find anything documented that is necessary that prepending -mbmi to -mbmi2 gives some benefit.
Instead at
https://gcc.gnu.org/onlinedocs/gcc/x86-Built-in-Functions.html#x86-Built-in-Functions

The following built-in functions are available when -mbmi is used. All of them generate the machine instruction that is part of the name.
unsigned int __builtin_ia32_bextr_u32(unsigned int, unsigned int);
unsigned long long __builtin_ia32_bextr_u64 (unsigned long long, unsigned long long);

The following built-in functions are available when -mbmi2 is used. All of them generate the machine instruction that is part of the name.
unsigned int _bzhi_u32 (unsigned int, unsigned int)
unsigned int _pdep_u32 (unsigned int, unsigned int)
unsigned int _pext_u32 (unsigned int, unsigned int)
unsigned long long _bzhi_u64 (unsigned long long, unsigned long long)
unsigned long long _pdep_u64 (unsigned long long, unsigned long long)
unsigned long long _pext_u64 (unsigned long long, unsigned long long)

and at
https://gcc.gnu.org/ml/gcc/2014-02/msg00204.html

( "... The real optimization comes from being able to use pext
(parallel bit extract), which can implement several bextr expressions in
parallel.")

Apart from that we don't use all -msse -msse2 -msse3 -msse4.2 etc. but just -msse3 (or -msse4.2) only.

As regards to the speedup within noise level - this pull request is actually reversal of mcostalba#198 wherein prepending -mbmi to -mbmi2 was claimed to be 0.3% faster and here (removing -mbmi) gives 0.4% speed gain.
2016-05-01 14:11:28 +02:00
Stéphane Nicolet 2694ef23c3 Isolated pawn simplification
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 117822 W: 21697 L: 21744 D: 74381

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 92307 W: 12330 L: 12305 D: 67672

Bench: 8813983

Resolves #659
2016-04-30 22:23:22 +01:00
VoyagerOne e082112cfe Use FMHs to assist with LMR formula.
STC:
LLR: 2.99 (-2.94,2.94) [0.00,5.00]
Total: 52232 W: 9654 L: 9304 D: 33274

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 115988 W: 15550 L: 15049 D: 85389

Bench: 7890808

Resolves #651
2016-04-24 01:00:58 +01:00
erbsenzaehler 4048bae47b Use -O3 for all compilers (including ICC)
There seems to be no benefit from using -fast over -O3 with icc.
So use -O3 everywhere.

No functional change

Resolves #652
2016-04-24 00:55:56 +01:00
DU-jdto c737062436 Remove some pointless micro-optimizations
Seems to give around 1% speed-up for CPUs with popcnt support.
Seems to give a very minor speed-up for CPUs without popcnt.

No functional change

Resolves #646
2016-04-23 02:04:28 +01:00
Marco Costalba 94e41274bb Fix incorrect draw detection
In this position we should have draw for repetition:

position fen rnbqkbnr/2pppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 moves g1f3 g8f6 f3g1
go infinite

But latest patch broke it.

Actually we had two(!) very subtle bugs, the first is that Position::set()
clears the passed state and in particular 'previous' member, so
that on passing setupStates, 'previous' pointer was reset.

Second bug is even more subtle: SetupStates was based on std::vector
as container, but when vector grows, std::vector copies all its contents
to a new location invalidating all references to its entries. Because
all StateInfo records are linked by 'previous' pointer, this made pointers
go stale upon adding more element to setupStates. So revert to use a
std::deque that ensures references are preserved when pushing back new
elements.

No functional change.
2016-04-18 00:13:16 +02:00
loco-loco ec6aab0136 Add a second level of follow-up moves
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 6438 W: 1229 L: 1077 D: 4132

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 4000 W: 605 L: 473 D: 2922

bench: 7378965

Resolves #636
2016-04-17 15:19:20 +01:00
Marco Costalba 7eaea3848c StateInfo is usually allocated on the stack by search()
And passed in do_move(), this ensures maximum efficiency and
speed and at the same time unlimited move numbers.

The draw back is that to handle Position init we need to
reserve a StateInfo inside Position itself and use at
init time and when copying from another Position.

After lazy SMP we don't need anymore this gimmick and we can
get rid of this special case and always pass an external
StateInfo to Position object.

Also rewritten and simplified Position constructors.

Verified it does not regress with a 3 threads SMP test:
ELO: -0.00 +-12.7 (95%) LOS: 50.0%
Total: 1000 W: 173 L: 173 D: 654

No functional change.
2016-04-17 08:29:33 +02:00
Niklas Fiekas ee7a68ea5f Fix last search info carried over to mate position
When starting search in a mate or stalemate position, Stockfish does not
even care to reinitialize and start worker threads. However after search
all threads are checked for the best move.

This can lead to bestmove and info beeing carried over from the last
search.

Example session:

    setoption name threads value 7
    go movetime 4000
    position startpos moves f2f3 e7e5 g2g4 d8h4
    go movetime 4000

Actual output is like (almost always):

    [...]
    bestmove e2e4
    info depth 0 score mate 0
    info depth 20 seldepth 29 multipv 1 score cp 28 [...] pv e2e4
    bestmove e2e4

Expected output / output after fix:

    [...]
    bestmove e2e4 ponder e7e6
    info depth 0 score mate 0
    bestmove (none)

Resolves #623
2016-04-16 10:22:36 +01:00
Marco Costalba d30994ecd5 Hide global visibility when not needed
Also move PieceValue definition in psqt.cpp,
where it is initialized.

Fix a warning in popcount16() with Intel compiler

No functional change.
2016-04-09 10:42:04 +02:00
Marco Costalba bd04f9a0f1 Fix Travis Cl
Broken after "32-bit/64-bit Makefile fix" commit.

Ubuntu "Precise" 12.04.5 supports multilib only until
g++ 4.6 that is not enough to compile Stockfish.

So move to Ubuntu 14.04.4 LTS (Trusty Tahr)

No functional change.
2016-04-09 09:35:19 +02:00
DU-jdto 1cbba8d6fa Small passed pawn simplification
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 21993 W: 4197 L: 4078 D: 13718

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 67213 W: 9135 L: 9077 D: 49001

Bench: 7482426

Resolves #622
2016-04-08 19:48:03 +01:00
Alain SAVARD e9e5f72c74 Undefended King Ring
There was already a penalty for squares only defended by King (undefended)

This test records a penalty for completely undefended squares in the so called extended king-ring
(so if we exclude squares defended by a Kg8 for example, we only look at h6 g6 and f6)

We also exclude squares occupied by opponent pieces in this computation,
based on the following results

Was yellow at STC
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 112499 W: 20649 L: 20293 D: 71557

and passed LTC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 36805 W: 5100 L: 4857 D: 26848

Bench: 8430233

Resolves: #619
2016-04-08 19:22:33 +01:00
Alain SAVARD 29b5842da8 Backward simplication
On top of the usual conditions
a) some opponent in front (but no lever)
b) some neighbours (in front) (but no neighbour behind or same rank)
c) < rank_5

to find out if a pawn is backward we look at the squares in front of this pawn to reach the same rank as the next neighbour.

In current master, a pawn is backward if any of those squares is controlled by an enemy pawn on an adjacent file

In this version, a pawn is ALSO backward if any of those squares is occupied by an enemy pawn.

STC:
http://tests.stockfishchess.org/tests/view/56fe7efd0ebc59301a3541f1
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 19051 W: 3557 L: 3433 D: 12061

LTC:
http://tests.stockfishchess.org/tests/view/56febc2d0ebc59301a354209
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40810 W: 5619 L: 5526 D: 29665

Bench: 7525245

Resolves #614
2016-04-08 19:12:55 +01:00
mstembera 8fb45caade Simplify popcnt
Also a speedup(about 1%) on 64-bit w/o hardware popcnt

Retire Max15 and Full template parameters
(Contributed by Marco Costalba)

Now that we have just SW and HW versions, use
template default parameter to get rid of explicit
template parameters.

Retire bitcount.h and move the only defined
function to bitboard.h

No functional change

Resolves #620
2016-04-08 18:52:15 +01:00
lucasart 900279a06f 32-bit/64-bit Makefile fix
Counter intuitively, make build ARCH=x86-32 does NOT produce a 32-bit compile
when running a 64-bit OS. Nor would ARCH=x86-64 produce a 64-bit compile when
running a 32-bit OS (assuming it compiled w/o errors).

No functional change

Resolves #621
2016-04-08 18:47:31 +01:00
NicklasPersson d5b24ad77b A combo patch of two tuning patches
STC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 14223 W: 2700 L: 2494 D: 9029

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 66294 W: 9065 L: 8739 D: 48490

Bench: 7607385

Resolves #612
2016-03-31 13:26:00 +01:00
lucasart f256388e08 Guard against UB in lsb/msb
lsb(b) and msb(b) are undefined when b == 0. This can lead to subtle bugs, where
the resulting code behaves differently on different configurations:
- It can be the home grown software LSB/MSB
- It can be the compiler generated software LSB/MSB (when using compiler
  intrinsics without the right compiler flags to allow compiler to use hardware
  LSB/MSB). Which of course depends on the compiler.
- It can be hardware LSB/MSB generated by the compiler.
- Not to mention that hardware LSB/MSB can return different value on different
  hardware when b == 0.

No functional change

Resolves #610
2016-03-31 13:22:37 +01:00
Marco Costalba db4b0d8b7d Rewrite bsfq management
Use compiler intrinsics when possible to
avoid writing platform specific asm code.

Tested on Windows 7 with MSVC 2013 and mingw 4.8.3 (32 and 64 bit)
and on Linux Mint with g++ 4.8.4 and clang 3.4 (32 and 64 bit).

No functional change

Resolves #609
2016-03-28 15:46:55 +01:00
snicolet 24dac5ccd3 Bonus for loose enemies
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 30504 W: 5743 L: 5485 D: 19276

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 11936 W: 1651 L: 1493 D: 8792

Bench: 8880041

Resolves #606
2016-03-27 20:56:54 +01:00
mbootsector 8788612828 Raise endgame passed pawn and material values
STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 136149 W: 25213 L: 24588 D: 86348

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 54637 W: 7533 L: 7238 D: 39866

Bench: 8546808

Resolves #608
2016-03-27 20:22:48 +01:00
fanon 5d1644ba69 Simplify pawns King Safety calculation
STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 130209 W: 23516 L: 23581 D: 83112

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33541 W: 4563 L: 4460 D: 24518

Bench: 8644370

Resolves #604
2016-03-27 20:08:56 +01:00
VoyagerOne 60590577f2 A small simplification in movepick.h
No functional change

Resolves #597
2016-03-14 20:54:49 -07:00
mstembera 04be84e0e2 Simplify Safe Checks
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 11796 W: 2211 L: 2074 D: 7511

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14324 W: 1935 L: 1806 D: 10583

Bench: 8075202

Resolves #600
2016-03-14 20:51:16 -07:00
Marco Costalba 647402ff79 Assorted cleanup of latest commits
No functional change.

Resolves #601
2016-03-14 20:42:44 -07:00
Stefan Geschwentner a273b6ef8c Add followup moves history for move ordering
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 7955 W: 1538 L: 1378 D: 5039

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 5323 W: 778 L: 642 D: 3903

Bench: 8261839

Resolves #599
2016-03-10 14:26:06 -08:00
mbootsector e1a7d135b2 Passed pawn bonus simplification
STC: (yellow)

LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 86114 W: 16063 L: 15921 D: 54130

LTC:

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14347 W: 2025 L: 1896 D: 10426

Bench: 8576437

Resolves #595
2016-03-02 19:10:56 +00:00
snicolet c1be0c68c7 Tweak initiative formula
Give more weight to the pawns number and
the vertical king distance in evaluate_initiative()

Passed STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 26729 W: 5067 L: 4825 D: 16837

and LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 60480 W: 8338 L: 8016 D: 44126

Bench: 8295162

Resolves #594
2016-03-02 00:02:03 +00:00
ElbertoOne c13052f344 Clean up depth reduction calculation
Might also be a slight speed up

No functional change

Resolves #593
2016-02-28 13:40:47 +00:00
joergoster 8de29390f2 Pass endgame value to evaluate_scale_factor()
No functional change

Resolves #592
2016-02-28 13:35:34 +00:00
VoyagerOne 45a309d92e Simplify Reduction Formula
Formula now only contains one coefficient. Making it much easier to tune.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 187443 W: 34858 L: 35028 D: 117557

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 88329 W: 11982 L: 11953 D: 64394

Bench: 7521394

Resolves #591
2016-02-28 13:31:25 +00:00
Leonid Pechenik d5ba8e827d Revert "Remove slowMover"
This reverts commit 77fa960f89.

Resolves #590
2016-02-28 13:25:05 +00:00
IIvec 77fa960f89 Remove slowMover
Removes a slowMover and one paramater from move_importance function.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 77023 W: 14456 L: 14433 D: 48134

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 37175 W: 5190 L: 5092 D: 26893

Resolves #589
2016-02-21 20:16:28 +00:00
VoyagerOne bfe9044ad9 History Stat Formula Simplification
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 67476 W: 12561 L: 12521 D: 42394

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 111923 W: 15147 L: 15149 D: 81627

Bench: 8430465

Resolves #588
2016-02-21 20:12:33 +00:00
VoyagerOne 744ed85a4d Fix futility pruning bug
PredictedDepth can be negative, causing the futility_margin to be negative.
It will be very difficult to tweak moveCount pruning and reduction formula, as they are tuned to prevent this behavior.

No functional change

Resolves #587
2016-02-14 19:48:46 +00:00
hxim 56dd58e6f9 Remove Weights
Removed remaining redundant weights for pawn structure,
passed pawns, space and king safety by redistributing them
into individual evaluation terms.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 15173 W: 2790 L: 2659 D: 9724

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 43433 W: 5936 L: 5846 D: 31651

Bench: 7156237

Resolves #586
2016-02-07 21:00:24 +00:00
Marco Costalba 4f6aa15228 Document HalfDensityMap
No functional change.

Resolves #584
2016-02-07 20:54:25 +00:00
Leonid Pechenik aedebe35cf Time management simplification
10+0.1:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 41963 W: 7967 L: 7883 D: 26113

60+0.6:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 132314 W: 17939 L: 17969 D: 96406

Resolves #580
2016-01-29 00:47:07 +00:00
Guenther Demetz 9a10313a9d rotating symmetric patterns with increasing skipsize
STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00] sprt @ 5+0.1 th 21
Total: 7068 W: 1121 L: 975 D: 4972

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00] sprt @ 12+0.12 th 21
Total: 26691 W: 3594 L: 3481 D: 19616

No functional change with a single thread

Resolves #574
2016-01-24 14:08:01 +00:00
Joona Kiiski 8c3a5bbc52 Do not probe syzygy bases when castling is possible
Almost no functional change. Bench is unchanged.

Resolves #230
Resolves #573
2016-01-20 15:24:21 +00:00
lucasart 28933a580e Retire RootNode template
There is no reason to compile 3 different copies of search(). PV nodes are on
the cold path, and PvNode is a template parameter, so there is no cost in
computing:

const bool RootNode = PvNode && (ss-1)->ply == 0;

And this simplifies code a tiny bit as well.

Speed impact is negligible on my machine (i7-3770k, linux 4.2, gcc 5.2):

            nps   +/-
test    2378605  3118
master  2383128  2793
diff      -4523  2746

Bench: 7751425

No functional change.

Resolves #568
2016-01-18 22:21:42 +00:00
Guenther Demetz 12eb345ebd Depth margin parameter-tweak in TT-save
Verified that is improvement with multiple threads:

LLR: 2.95 (-2.94,2.94) [0.00,4.00] sprt @ 30+0.3 th 3
Total: 14817 W: 2103 L: 1915 D: 10799

LLR: 2.96 (-2.94,2.94) [0.00,4.00] sprt @ 15+0.15 th 7
Total: 10264 W: 1498 L: 1321 D: 7445

Verified that is not a significant regression with a single thread:

LLR: 2.96 (-2.94,2.94) [-4.00,0.00] sprt @ 60+0.6 th 1
Total: 23975 W: 3294 L: 3210 D: 17471

Resolves #575
2016-01-18 22:04:38 +00:00
Joona Kiiski ebec2fa48e Remove redundant -std=c++0x flag
This flag is functionally identical to '-std=c++11' flag which
is part of standard flags.

No functional change

Resolves #571
2016-01-18 21:54:40 +00:00
Joona Kiiski 552b3ccb66 Makefile: Allow specifying compiler executable
No functional change

Resolves #570
2016-01-18 21:47:52 +00:00
Marco Costalba 356147d99a Rewrite time formula
Time management is really too complex, our aim is
to simplify it, but for time being at least rewrite
in an understandable way.

No functional change.
2016-01-18 17:12:18 +01:00
Lyudmil Antonov 89723339d9 Assorted English grammar changes
No functional change

Resolves #567
2016-01-16 21:34:29 +00:00
Stefano80 74e2fa97b7 Adjust reductions based on history and cmh tables
STC:
LLR: 4.06 (-2.94,2.94) [0.00,5.00]
Total: 149395 W: 28029 L: 27208 D: 94158

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 9628 W: 1368 L: 1217 D: 7043

bench: 8076724

Resolves #565
2016-01-13 16:18:35 +00:00
Stefano80 dcd8ce7094 Update comments in LMR step
No functional change

Resolves #564
2016-01-13 16:03:53 +00:00
Leonid Pechenik 4d1220d672 Tune time management for LTC
60+0.6:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 102533 W: 14270 L: 13842 D: 74421

Resolves #558
2016-01-09 09:21:52 +00:00
lucasart aa31f7f096 Retire CenterBind
And compensate in the PSQT.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27714 W: 5161 L: 5052 D: 17501

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 36354 W: 5008 L: 4909 D: 26437

Bench: 8603285

Resolves #556
2016-01-03 17:33:36 +00:00
Alain SAVARD 1de97337c2 Fine tuning of unsupported pawn penalty
Adjust the unsupported pawn penalty when the pawn is supporting 2 pawns
(for example g7 in f6-g7-h6)

Passed STC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 23833 W: 4384 L: 4158 D: 15291

Passed LTC
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 42711 W: 5918 L: 5655 D: 31138

Bench: 8390233

Resolves #549
2016-01-03 14:15:29 +00:00
Leonid Pechenik 9eceb894ac Adjust time used for move based on previous score
Use less time if evaluation is not worse than for previous move and even less time if in addition no fail low encountered for current iteration.

STC: 10+0.1
ELO: 5.37 +-2.9 (95%) LOS: 100.0%
Total: 20000 W: 3832 L: 3523 D: 12645

STC: 10+0.1
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 17527 W: 3334 L: 3132 D: 11061

LTC: 60+0.6
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 28233 W: 3939 L: 3725 D: 20569

LTC: 60+0.6
ELO: 2.43 +-1.4 (95%) LOS: 100.0%
Total: 60000 W: 8266 L: 7847 D: 43887

LTC: 60+0.06
LLR: 2.95 (-2.94,2.94) [-1.00,3.00]
Total: 38932 W: 5408 L: 5207 D: 28317

Resolves #547
2016-01-03 14:01:15 +00:00
Joona Kiiski 5972c4a678 Restore development version 2016-01-03 13:52:01 +00:00
Joona Kiiski dd9cf30581 Stockfish 7
Bench: 8355485

No functional change
2016-01-02 20:26:02 +00:00
ppigazzini d4af15f682 Update AUTHORS and copyright notice
No functional change

Resolves #555
2016-01-02 09:43:51 +00:00
Marco Costalba 9742fb10fd Update Copyright year
No functional change.

Resolves #554
2016-01-01 10:17:36 +00:00
Joona Kiiski 7cafbab529 Stockfish 7 Beta 2
Bench: 8355485

No functional change
2015-12-30 11:32:45 +00:00
Alain SAVARD 6a79d5d10e Correct Pawn Trace Score + Code Clean up
No functional change

Resolves #542
2015-12-30 11:30:28 +00:00
Marco Costalba 3e4fed3a91 Fix assert with very high score position
In case of a very high material score, we can
overflow VALUE_INFINITE.

This patch fixes an assert with:

position fen 7k/QQQQR3/2B5/4KN1Q/3QQ3/8/8/4R3 b - - 0 1
go depth 1

No functional change.

Resolves #546
2015-12-30 11:16:48 +00:00
Joona Kiiski a5c76d69c3 Stockfish 7 Beta 1
Bench: 8355485

No functional change
2015-12-27 19:36:56 +00:00
Marco Costalba 1b5b900a29 Move some globals into main thread scope
Make it explicit that those variables are not globals, but
are used only by main thread. I think it is a sensible
clarification because easy move is already tricky enough
and current patch makes the involved actors explicit.

No functional change.

Resolves #537
2015-12-27 19:29:16 +00:00
Marco Costalba 90f5937373 Revert "Fix compiling of 32 bit binary on 64-bit Windows"
This reverts commit 1e8836d921

Broken compile on mingw under Windows:

Config:
debug: 'yes'
optimize: 'yes'
arch: 'i386'
bits: '32'
prefetch: 'yes'
bsfq: 'no'
popcnt: 'no'
sse: 'yes'
pext: 'no'

Flags:
CXX: i686-w64-mingw32-c++
CXXFLAGS: -Wall -Wcast-qual -fno-exceptions -fno-rtti -std=c++11  -Wextra -Wshadow -g -O3 -msse
LDFLAGS:  -static

Testing config sanity. If this fails, try 'make help' ...

mingw32-make[1]: Leaving directory 'C:/stockfish/src'
c:/MinGw/bin/mingw32-make ARCH=x86-32 COMP=mingw all
mingw32-make[1]: Entering directory 'C:/stockfish/src'
sh: C:\Program: No such file or directory
i686-w64-mingw32-c++ -Wall -Wcast-qual -fno-exceptions -fno-rtti -std=c++11  -Wextra -Wshadow -g -O3 -msse   -c -o benchmark.o benchmark.cpp
<builtin>: recipe for target 'benchmark.o' failed
process_begin: CreateProcess(NULL, i686-w64-mingw32-c++ -Wall -Wcast-qual -fno-exceptions -fno-rtti -std=c++11 -Wextra -Wshadow -g -O3 -msse -c -o benchmark.o benchmark.cpp, ...) failed.
make (e=2): Impossibile trovare il file specificato.

mingw32-make[1]: *** [benchmark.o] Error 2
mingw32-make[1]: Leaving directory 'C:/stockfish/src'
makefile:401: recipe for target 'build' failed
mingw32-make: *** [build] Error 2

No functional change.
2015-12-23 09:41:56 +01:00
braich 1e8836d921 Fix compiling of 32 bit binary on 64-bit Windows
Two versions of mingw-w64 (targeting Win64 and Win32)
can be installed on Windows too.

No functional change

Resolves #532
2015-12-22 19:07:10 +00:00
DU-jdto e3c85c314d Remove another unnecessary Search::Stack field
No functional change

Resolves #535
2015-12-22 17:16:15 +00:00
Alain SAVARD da34023cfc New mobility bonus
Tuned the global mobility factor for each piece, as well as some +- delta,

The master mobility factor was {266,334} and tuning gave
{267, 362} +S(-2,-2) for the Knight
{249, 328} +S( 0,-2) for the Bishop
{298, 353} +S(1,1) for the Rook
{265, 358} +S(2,-1) for the Queen

Passed STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 49402 W: 9367 L: 9037 D: 30998

and LTC
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 26831 W: 3871 L: 3658 D: 19302

Bench: 8355485

Resolves #536
2015-12-22 10:02:37 +00:00
VoyagerOne ed72a1e9ba Remove killer move conditions from LMR
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8459 W: 1619 L: 1477 D: 5363

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32239 W: 4404 L: 4299 D: 23536

Bench: 7597031

Resolves #534
2015-12-20 20:41:17 +00:00
loco-loco 5bbd944099 Remove unused field SearchStack::ttMove
No functional change

Resolves #533
2015-12-20 20:37:18 +00:00
Guenther Demetz 38adb487ca Distinct iteration paths for Lazy SMP threads
STC 5+0.1, threads 7
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 6026 W: 1047 L: 901 D: 4078

LTC: 20+0.2, threads 7
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 19739 W: 2910 L: 2721 D: 14108

STC 5+0.1, threads 20
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 2493 W: 462 L: 331 D: 1700

LTC 30+0.3, threads 20
ELO: 8.86 +-3.7 (95%) LOS: 100.0%
Total: 8000 W: 1076 L: 872 D: 6052

Bench: 8012530

Resolves #525
2015-12-18 21:59:09 +00:00
mstembera 2c1797ab81 Fix easy move bug in SMP mode
Fix a bug where we could stop the search after only 10% of time used due to a matching easy move but later switch to a different move that was never pre-screened as easy due to SMP thread select.

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27227 W: 4910 L: 4800 D: 17517

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 40368 W: 5826 L: 5733 D: 28809

Resolves #521
2015-12-14 18:47:12 +00:00
Alain SAVARD 9fac5774f4 Threats retuned
STC:

LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 45239 W: 8913 L: 8591 D: 27735

LTC:

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21046 W: 3200 L: 2989 D: 14857

Bench: 8012530

Resolves #526
2015-12-14 18:11:13 +00:00
Leonid Pechenik 69240a982d Simplify time management and fix 'ponder on' bug
Simplify time management code by removing hard stops for unchanging first root moves.
Search is now stopped earlier at the end iteration if it did not have fail-lows at root.

This simplification also fixes pondering bug. Ponder flag was true by default
and cutechess-cli doesn't change it to false even though no pondering is possible.
Fix the issue by setting the default value of 'Ponder' flag to false.

10+0.1:
ELO: 3.51 +-3.0 (95%) LOS: 99.0%
Total: 20000 W: 3898 L: 3696 D: 12406

40+0.4:
ELO: 1.39 +-2.7 (95%) LOS: 84.7%
Total: 20000 W: 3104 L: 3024 D: 13872

60+0.06:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 37231 W: 5333 L: 5236 D: 26662

Stopped run at 100+1:
LLR: 1.09 (-2.94,2.94) [-3.00,1.00]
Total: 37253 W: 4862 L: 4856 D: 27535

Resolves #523
Fixes #510
2015-12-14 18:00:52 +00:00
mstembera 7904a7d930 Fix MultiPv and Skill in SMP.
7 threads, 5+0.1:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 55460 W: 9665 L: 9601 D: 36194

No functional change in normal playing mode
2015-12-12 10:32:28 +00:00
absimaldata 8315620522 New Tuned Weights
More accurate evaluation weights

Performed better at STC

LLR: 1.32 (-2.94,2.94) [0.00,4.00]
Total: 190043 W: 37433 L: 36675 D: 115935

Passed LTC

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 30157 W: 4540 L: 4303 D: 21314

Bench: 9264977

Resolves #515
2015-12-05 10:31:00 +00:00
Marco Costalba 52eac1e535 Simplify outpost code
Also inline defintions of SpaceMask and CenterBindMask.

Verified from assembly that compiler computes the values
at compile time, so it is also theoretical faster.

While there factor out scale factor evaluation.

No functional change.
2015-12-05 11:16:19 +01:00
mstembera 5d077bb482 Proper Makefile for cross compiling 64 or 32 bit PGO + LTO + static Windows binaries under Linux.
No functional change

Resolves #511
2015-12-03 14:01:38 +00:00
Stéphane Nicolet 5a42427cb8 Introduce new Threats weights = {350, 256}
Raise the midgame threats weight by 37%.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 8165 W: 1675 L: 1487 D: 5003

and LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 28181 W: 4141 L: 3912 D: 20128

Bench: 7824961

Resolves #512
2015-12-02 11:18:26 +00:00
Joona Kiiski 5339ab57d8 Revert "Allow cross compilation of Windows binaries on a Linux system"
This reverts commit 388630ae28.

Confuses fishtest build system
2015-11-24 17:24:45 -08:00
mstembera 388630ae28 Allow cross compilation of Windows binaries on a Linux system
that are PGO, LTO, and statically linked.
Credit: pasquale....@gmail.com

No functional change

Resolves #505
2015-11-23 17:06:14 -08:00
mstembera 79f393084a Clean up RootMove less operator
This is used by std::stable_sort() to sort moves from highest score to lowest score.

1) The comment is incorrect since highest to lowest means descending.
2) It's more natural to implement a less operator using another less operator rather than a greater operator.

No functional change.

Resolves #504
2015-11-20 23:30:07 -08:00
lucasart 328098d027 Fix TT comment and static_assert()
Comment is based on a misunderstanding of what unaligned memory access is. Here
is an article that explains it very clearly:
https://www.kernel.org/doc/Documentation/unaligned-memory-access.txt

No matter how we define TTEntry or TTCluster, there will never be any unaligned
memory access. This is because the complier knows the alignment rules, and does
the necessary adjustments to make sure unaligned memory access does not occur.

The issue being adressed here has nothing to do with unaligned memory access. It
is about cache performance. In order to achieve best cache performance:
- we prefetch the cacheline as soon as possible.
- we ensure that TT clusters do not spread across two cachelines. If they did,
  we would need to prefetch 2 cachelines, which could hurt cache performance.

Therefore the true conditions to achieve this are:
1/ start adress of TT is cache line aligned. void TranspositionTable::resize()
enforces this.
2/ TT cluster size should *divide* the cache line size. Currently, we pack 2
clusters per cache lines. It used to be 1 before "TT sardines". Does not matter
what the ratio is, all we want is to fit an integer number of clusters per cache
line.

No functional change.

Resolves #506
2015-11-20 23:23:53 -08:00
Marco Costalba 93195555ed Rewrite how threads are spawned
Instead of creating a running std::thread and
returning, wait in Thread c'tor that the native
thread of execution goes to sleep in idle_loop().

In this way we can simplify how search is started,
because when main thread is idle we are sure also
all other threads will be idle, in any case, even
at thread creation and startup.

After lazy smp went in, we can simpify and rewrite
a lot of logic that is now no more needed. This is
hopefully the final big cleanup.

Tested for no regression at 5+0.1 with 3 threads:
LLR: 2.95 (-2.94,2.94) [-5.00,0.00]
Total: 17411 W: 3248 L: 3198 D: 10965

No functional change.
2015-11-21 07:48:50 +01:00
VoyagerOne 07e0741dfb History Pruning: Don't prune the main killer move.
Also increased pruned depth to 4.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 23380 W: 4581 L: 4350 D: 14449

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 28934 W: 4329 L: 4105 D: 20500

Bench: 8369743

Resolves #498
2015-11-15 20:40:09 -08:00
Ken T Takusagawa 27ca89d479 Do not conceal the invocation of the benchmark program
It is better to be able to see what arguments it is being called with.

No functional change

Resolves #497
2015-11-15 20:34:37 -08:00
Stefan Geschwentner cccf3c815c Bonus for reachable outpost
Give a bonus for outpost squares which in reach of a bishop or knight.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 22725 W: 4570 L: 4339 D: 13816

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 15019 W: 2333 L: 2157 D: 10529

Bench: 8503181

Resolves #495
2015-11-14 06:31:45 -08:00
Marco Costalba 76ed0ab501 Retire ThreadBase
Now that we don't have anymore TimerThread, there is
no need of this long class hierarchy.

Also assorted reformatting while there.

To verify no regression, passed at STC with 7 threads:
LLR: 2.97 (-2.94,2.94) [-5.00,0.00]
Total: 30990 W: 4945 L: 4942 D: 21103

No functional change.
2015-11-13 08:22:44 +01:00
Marco Costalba ce84ab6e9d Fix broken UCI 'wait for stop'
When we reach the maximum depth, we can finish the
search without a raise of Signals.stop. However, if
we are pondering or in an infinite search, the UCI
protocol states that we shouldn't print the best move
before the GUI sends a "stop" or "ponderhit" command.

It was broken by lazy smp. Fix it by moving the stopping
of the threads after waiting for GUI.

No functional change.
2015-11-13 08:13:59 +01:00
lucasart fca8dbc029 Avoid friend
operator<<(os, pos) does not need to access any private members of pos.

No functional change.

Resolves #492
2015-11-10 21:46:02 +00:00
lucasart e6eeb17aa6 Ensure that rootDepth < DEPTH_MAX
Indeed, if we use a depth >= DEPTH_MAX, we start having negative depth in the
TT (due to int8_t cast).

No functional change in single thread mode

Resolves #490
2015-11-10 21:41:42 +00:00
Marco Costalba 9c9205860c Get rid of timer thread
Unfortunately std::condition_variable::wait_for()
is not accurate in general case and the timer thread
can wake up also after tens or even hundreds of
millisecs after time has elapsded. CPU load, process
priorities, number of concurrent threads, even from
other processes, will have effect upon it.

Even official documentation says: "This function may
block for longer than timeout_duration due to scheduling
or resource contention delays."

So retire timer and use a polling scheme based on a
local thread counter that counts search() calls and
a small trick to keep polling frequency constant,
independently from the number of threads.

Tested for no regression at very fast TC 2+0.05 th 7:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32969 W: 6720 L: 6620 D: 19629

TC 2+0.05 th 1:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 7765 W: 1917 L: 1765 D: 4083

And at STC TC, both single thread
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 15587 W: 3036 L: 2905 D: 9646

And with 7 threads
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 8149 W: 1367 L: 1227 D: 5555

bench: 8639247
2015-11-03 11:27:00 +01:00
mbootsector 27c5cb5912 Pick bestmove from the deepest thread.
STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 26930 W: 4441 L: 4214 D: 18275

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 7783 W: 1017 L: 876 D: 5890

No functional change in single thread mode

Resolves #485
2015-11-02 10:05:43 +00:00
Marco Costalba 86f04dbcc0 Assorted trivia in search.cpp
The only interesting change is the moving of
stack[MAX_PLY+4] back to its original position
in id_loop (now renamed Thread::search).

No functional change.
2015-10-31 19:26:35 +01:00
VoyagerOne aa242d2f84 New History Bonus Formula
bonus = d^2 + d - 1

Bench: 8639247

Resolves #484
2015-10-31 18:17:00 +00:00
Ajith 8d858783e1 Reduce variation in rootDepth between different threads
Reduce the variation in Root Depth between different threads. This
prevents threads from searching at a depth much higher than Main Thread.

Performed well at STC 24 Threads:
ELO: 3.44 +-3.8 (95%) LOS: 96.1%
Total: 10000 W: 1627 L: 1528 D: 6845

And LTC 24 Threads
LLR: 1.43 (-2.94,2.94) [0.00,4.00]
Total: 3804 W: 500 L: 420 D: 2884
ELO : +7.31
p-value: 73.16%

Passed no regression at STC 3 Threads:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40457 W: 7148 L: 7060 D: 26249

And LTC 3 Threads:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 17704 W: 2489 L: 2364 D: 12851

Raising a pull request early as 24 Thread tests are very expensive and
this is clearly a positive gain at high thread counts and high time
controls. The change is a small parameter tweak with no additional
logic.

No functional change for single thread mode.

Resolves #481
2015-10-29 17:06:13 -07:00
Stéphane Nicolet 80d7556af7 Some code and comment cleanup
- Remove all references to split points
- Some grammar and spelling fixes

No Functional change

Resolves #478
2015-10-29 15:28:59 +00:00
lucasart 00d9e9fd28 Use atomics instead of volatile
Rely on well defined behaviour for message passing, instead of volatile. Three
versions have been tested, to make sure this wouldn't cause a slowdown on any
platform.

v1: Sequentially consistent atomics

No mesurable regression, despite the extra memory barriers on x86. Even with 15
threads and extreme time pressure, both acting as a magnifying glass:

threads=15, tc=2+0.02
ELO: 2.59 +-3.4 (95%) LOS: 93.3%
Total: 18132 W: 4113 L: 3978 D: 10041

threads=7, tc=2+0.02
ELO: -1.64 +-3.6 (95%) LOS: 18.8%
Total: 16914 W: 4053 L: 4133 D: 8728

v2: Acquire/Release semantics

This version generates no extra barriers for x86 (on the hot path). As expected,
no regression either, under the same conditions:

threads=15, tc=2+0.02
ELO: 2.85 +-3.3 (95%) LOS: 95.4%
Total: 19661 W: 4640 L: 4479 D: 10542

threads=7, tc=2+0.02
ELO: 0.23 +-3.5 (95%) LOS: 55.1%
Total: 18108 W: 4326 L: 4314 D: 9468

As suggested by Joona, another test at LTC:

threads=15, tc=20+0.05
ELO: 0.64 +-2.6 (95%) LOS: 68.3%
Total: 20000 W: 3053 L: 3016 D: 13931

v3: Final version: SeqCst/Relaxed

threads=15, tc=10+0.1
ELO: 0.87 +-3.9 (95%) LOS: 67.1%
Total: 9541 W: 1478 L: 1454 D: 6609

Resolves #474
2015-10-25 09:15:45 +00:00
Stefano80 3428a28751 KRPPKRP endgame: Simplify ugly switch statement
No functional change

Resolves #470
2015-10-24 23:17:08 +01:00
Marco Costalba 307a5a4f63 Cleanup history stats
And other assorted trivia.

No functional change.
2015-10-24 17:29:12 +02:00
Alain SAVARD 55758344d3 Simplify threats
Using less parameters and code to compute Threats
Includes also a few spacing edits.

Run as a simplification.

Passed STC 10+0.1
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 18879 W: 3725 L: 3600 D: 11554

Passed LTC 60+0.4
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 74116 W: 11001 L: 10958 D: 52157

bench: 8004751
2015-10-24 10:08:26 +02:00
VoyagerOne 4eca981e55 History pruning
Prune moves with negative History
and CMH scores at low depth.

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 24182 W: 4672 L: 4439 D: 15071

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 12579 W: 1959 L: 1792 D: 8828

bench: 8907701
2015-10-24 09:30:31 +02:00
Marco Costalba 5066e17eba Update authors
Fishtest is a key factor of SF success.

Thanks to Fishtest we have not only greately
improved ELO but, even more important, we
have enabled a kind of joint development and
testing that it is the herat of on open
source project like SF.

Open sourcing is not just about open code, it is
about commuity development. In case of a chess engine
this has never been possible before due to missing
a strong and strict testing environment that allows
many people to contribute in a safe and coordinate way.

Fishtest is a new way of developing chess engines,
something that has never exsisted before.

No functional change.
2015-10-22 07:03:03 +02:00
Stefano80 328d314f2f Almost passed tuning attempts
Collect and give a second try to some almost passed tuning attempts and
one-line tweaks from the last month.

Passed STC

LLR: 3.07 (-2.94,2.94) [0.00,4.00]
Total: 15124 W: 2974 L: 2756 D: 9394

And LTC

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 21577 W: 3507 L: 3289 D: 14781

Bench: 8855226

Resolves #464
2015-10-20 19:49:01 -07:00
mbootsector ecc5ff6693 Lazy SMP
Start all threads searching on root position and
use only the shared TT table as synching scheme.

It seems this scheme scales better than YBWC for
high number of threads.

Verified for nor regression at STC 3 threads
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40232 W: 6908 L: 7130 D: 26194

Verified for nor regression at LTC 3 threads
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 28186 W: 3908 L: 3798 D: 20480

Verified for nor regression at STC 7 threads
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 3607 W: 674 L: 526 D: 2407

Verified for nor regression at LTC 7 threads
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 4235 W: 671 L: 528 D: 3036

Tested with fixed games at LTC with 20 threads
ELO: 44.75 +-7.6 (95%) LOS: 100.0%
Total: 2069 W: 407 L: 142 D: 1520

Tested with fixed games at XLTC (120secs) with 20 threads
ELO: 28.01 +-6.7 (95%) LOS: 100.0%
Total: 2275 W: 349 L: 166 D: 1760

Original patch of mbootsector, with additional work
from Ivan Ivec (log formula), Joerg Oster (id loop
simplification) and Marco Costalba (assorted formatting
and rework).

Bench: 8116244
2015-10-20 06:58:08 +02:00
Stéphane Nicolet 7ea5659c5f Asymmetry bonus for the attacking side
Use asymmetry in the position (king separation, pawn structure) to
compute an "initiative bonus" for the attacking side.

Passed STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 14563 W: 2826 L: 2636 D: 9101

And LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 14363 W: 2317 L: 2141 D: 9905

Bench: 8116244

Resolves #462
2015-10-15 21:27:52 -07:00
VoyagerOne 8fd34d7763 Combination of two ideas:
Apply bonus for the prior CMH that caused a fail low.

Balance Stats: CMH and History bonuses are updated differently.
This eliminates the "fudge" factor weight when scoring moves. Also
eliminated discontinuity in the gravity history stat formula. (i.e. stat
scores will no longer inverse when depth exceeds 22)

STC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 21802 W: 4107 L: 3887 D: 13808

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 46036 W: 7046 L: 6756 D: 32234

Bench: 7677367
2015-10-12 14:00:54 -07:00
Jonathan Calovski 55b46ffa90 Retire rook contact checks
STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 34114 W: 6363 L: 6265 D: 21486

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 61776 W: 9349 L: 9289 D: 43138

LTC (after rebasing):
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 15261 W: 2343 L: 2214 D: 10704

Bench: 7523382

Resolves #442
2015-10-07 19:43:20 +01:00
Marco Costalba 08d6465d90 Travis CI: add gcc 4.8 for osx
This setup was still missing.

Suggested by Stéphane Nicolet.

No functional change.
2015-10-07 20:33:37 +02:00
Marco Costalba 5f4d9309c8 Travis CI: add clang and osx
Extend builds to clang and osx platforms.

And check bench numbers.

No functional change.
2015-10-06 12:35:42 +02:00
Stefano80 436c303731 Tuning of assorted values
Passed STC

LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 45401 W: 8590 L: 8274 D: 28537

Passed LTC

LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 36089 W: 5589 L: 5331 D: 25169

Bench: 8397672

Resolves #445
2015-10-05 19:49:52 -07:00
Alain SAVARD 540b49a152 Remove queen threat evaluation
Threats by queen seem to be worthless.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 13627 W: 2607 L: 2473 D: 8547

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19146 W: 2950 L: 2827 D: 13369

Bench: 8222484

Resolves #439
2015-10-06 03:35:17 +01:00
Marco Costalba 2e45447957 Add Trevis CI support
Add Travis CI support to GitHub repo.

After every push to master, Travis will build
the sources directly from GitHub repo according
to .travis.yml and verify everything is ok.

No functional change.
2015-10-05 12:39:34 +02:00
Marco Costalba dc3508d157 Fix a comment in TTEntry::save
Comment was slightly incorrect.

No functional change.
2015-10-05 09:16:16 +02:00
Marco Costalba ca38358574 Run PVS-STUDIO analyzer
Fix issues after a run of PVS-STUDIO analyzer.
Mainly false positives but warnings are anyhow
useful to point out not very readable code.

Noteworthy is the memset() one, where PVS prefers ss-2
instead of stack. This is because memeset() could
be optimized away by the compiler when using 'stack',
due to stack being a local variable no more used after
memset. This should normally not happen, but when
it happens it leads to very sublte and difficult
to find bug, so better to be safe than sorry.

No functional change.
2015-10-05 09:13:33 +02:00
Stefan Geschwentner 83e19fbed5 File based passed pawn bonus
Add file based bonus for passed pawns. Values tuned by SPSA.

STC:
LLR: 3.33 (-2.94,2.94) [0.00,5.00]
Total: 36889 W: 6805 L: 6507 D: 23577

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 32301 W: 5101 L: 4858 D: 22342

Bench: 8073614

Resolves #436
2015-10-03 03:50:06 -07:00
Jonathan Calovski 9f5b31c21d Bonus for checking moves
STC:
LLR: 2.97 (-2.94,2.94) [0.00,5.00]
Total: 14531 W: 2765 L: 2576 D: 9190

LTC:
LLR: 3.20 (-2.94,2.94) [0.00,5.00]
Total: 52518 W: 8107 L: 7782 D: 36629

Bench: 7556477

Resolves #435
2015-10-03 03:39:21 -07:00
Marco Costalba 3c0fe1d9b2 Rework lock protecting
When changing 'search' and 'splitPointsSize' we have to
use thread locks, not split point ones, because can_join()
is called under the formers.

Verified succesfully with 24 hours toruture tests with 20
cores machine by Louis Zulli: it does not hangs.

Verifyed for no regressions with STC, 7 threads:
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 52804 W: 8159 L: 8087 D: 36558

No functional change.
2015-09-30 10:47:20 +02:00
Jonathan Calovski 77b4f4c2e7 Refine ranks and increase resulting bonus.
STC:
LLR: 2.94 (-2.94,2.94) [0.00,4.00]
Total: 272379 W: 51773 L: 50658 D: 169948

LTC:
LLR: 3.06 (-2.94,2.94) [0.00,4.00]
Total: 41504 W: 6555 L: 6273 D: 28676

bench: 7658406

Resolves #430
2015-09-19 09:14:35 -07:00
mstembera 68fbb1e052 Reduce writes in TT::probe().
Only refresh TT entry when it's really necessary.
This should give a small speed boost for some machines.
And it's a risk-free change.

No functional change.

Resolves #429
2015-09-17 17:33:40 -07:00
mstembera 01fab4d432 Remove unnecessary generation check in TT save
Checking for generation is unnecessary because if the key matches then the entry was probed and refreshed earlier.

STC 2MB
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 57391 W: 10671 L: 10613 D: 36107
http://tests.stockfishchess.org/tests/view/55ef59fa0ebc5976a2d6da5d

LTC 8MB
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 60732 W: 9260 L: 9199 D: 42273
http://tests.stockfishchess.org/tests/view/55ef8fe60ebc5976a2d6da6b

STC 16MB
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 23443 W: 4369 L: 4293 D: 14781
http://tests.stockfishchess.org/tests/view/55ef8fe60ebc5976a2d6da6b

No functional change

Resolves #427
2015-09-17 17:13:45 -07:00
Stefan Geschwentner 660c38f781 Scales the endgame score by the number of pawns.
Credits goes also to Stephane Nicolet for his great idea of scaling by pawns.

STC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 9994 W: 1929 L: 1760 D: 6305

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 11240 W: 1789 L: 1626 D: 7825

bench 7298564

Resolves #423
2015-09-15 05:55:09 -07:00
Joona Kiiski 613dc66c12 Careful SMP locking - Fix very occasional hangs
Louis Zulli reported that Stockfish suffers from very occasional hangs with his 20 cores machine.

Careful SMP debugging revealed that this was caused by "a ghost split point slave", where thread
was marked as a split point slave, but wasn't actually working on it.

The only logical explanation for this was double booking, where due to SMP race, the same thread
is booked for two different split points simultaneously.

Due to very intermittent nature of the problem, we can't say exactly how this happens.

The current handling of Thread specific variables is risky though. Volatile variables are in some
cases changed without spinlock being hold. In this case standard doesn't give us any kind of
guarantees about how the updated values are propagated to other threads.

We resolve the situation by enforcing very strict locking rules:
- Values for key thread variables (splitPointsSize, activeSplitPoint, searching)
can only be changed when the thread specific spinlock is held.
- Structural changes for splitPoints[] are only allowed when the thread specific spinlock is held.
- Thread booking decisions (per split point) can only be done when the thread specific spinlock is held.

With these changes hangs didn't occur anymore during 2 days torture testing on Zulli's machine.

We probably have a slight performance penalty in SMP mode due to more locking.

STC (7 threads):
ELO: -1.00 +-2.2 (95%) LOS: 18.4%
Total: 30000 W: 4538 L: 4624 D: 20838

However stability is worth more than 1-2 ELO points in this case.

No functional change

Resolves #422
2015-09-10 19:15:43 +01:00
mstembera 3e2591d83c Minor clean up of some function parameters
No function change

Resolves #416
2015-09-07 20:17:39 +01:00
mstembera 46b5a5f0db Fix syzygy en passant issue
v = value without ep capture being considered
v1 = value of the ep capture

The correct logic is:
if without e.p. capture we are losing, and the value of e.p is either draw, or win or "loss, but 50 move rule saves us", then we should use the value of ep capture.

Credit and thanks to syzygy1 and lantonov !

No functional change (except with syzygy bases)

Resolves #415
Resolves #394
2015-09-06 22:19:33 +01:00
gguliash 84a641b8bb A small code simplification
No functional change

Resolves #411
2015-08-30 19:58:32 +01:00
Joona Kiiski 7f300a7698 History gravity
Instead of using hard coded Min and Max values for history,
always adjust the old value slightly downwards before adding a new value.

The adjustment acts like gravity that prevents the value escaping too
far from zero.

Bench: 8020484

Resolves #407
2015-08-29 15:09:00 +01:00
Marco Costalba 087b638f6c Reformat trace code
Apart from usual renaiming, take advantage of
C++11 function template default parmeter to
get rid of Eval trampoline functions.

Some triviality fixes while there.

No functional change.
2015-08-29 08:28:01 +02:00
lucasart 7ad85fca6d Prune castling moves
Align the behaviour with reductions. Initially castling moves had to be
treated differently, because the SEE did not handle them correctly. But now it
does.

STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 83750 W: 15722 L: 15711 D: 52317

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 97183 W: 15120 L: 15115 D: 66948

bench 7759837

Resolves #403
2015-08-28 20:33:12 +01:00
mstembera bf6bc1b37b Better document entry age calculation used in TT replace.
No functional change.

Resolves #401
Resolves #400
2015-08-20 20:27:06 +01:00
lucasart c052e03426 Retire dangerous flag
Replace by its value where it is used. Code is more clear that
way.

No functional change.

Resolves #402
2015-08-20 20:18:23 +01:00
Alain SAVARD 69a1a808c8 Retire PawnSafePush bonus
PawnSafePush, with the value S(5,5) proved not "necessary"
possibly due to recent changes to MobilityArea and other changes to Connected bonus.

STC:
LLR: 3.22 (-2.94,2.94) [-3.00,1.00]
Total: 98528 W: 18757 L: 18759 D: 61012

LTC:
LLR: 5.30 (-2.94,2.94) [-3.00,1.00]
Total: 204194 W: 31698 L: 31734 D: 140762

Bench: 7620871

Resolves #396
2015-08-17 19:32:43 +01:00
Marco Costalba 600234f2e2 Reformat PassedPawnsBonus
Align to SF coding standards.

No functional change.
2015-08-15 16:04:16 +02:00
mstembera 9dbb3ae8b2 TT entry value based on depth and relative age
Calculate TT replace value as depth minus eight times relative age.

STC 2MB
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 45258 W: 8595 L: 8279 D: 28384

LTC 8MB
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 56401 W: 8809 L: 8489 D: 39103

STC 16MB
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 34764 W: 6565 L: 6529 D: 21670

Bench: 9069474

Resolves #395
2015-08-15 12:24:48 +01:00
DiscanX f4ace94f91 Tuned values for mid and end game passed pawns.
STC :
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 22691 W: 4468 L: 4228 D: 13995

LTC :
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 13620 W: 2216 L: 2023 D: 9381

Bench: 8384669

Resolves #391
2015-08-10 01:26:59 +08:00
mstembera 1aae35b8b7 Revert TT replacement strategy changes (#380)
It could cause problems with high depths and long time controls

Bench: 8626315

Resolves #390
2015-08-08 17:49:53 +01:00
Marco Costalba e6310b3469 Rename Position::list
Use Position::square and Position::squares instead.

This allow us to remove king_square(), simplify
endgames and to have more naming uniformity.

Moreover, this is a prerequisite step in case
in the future we decide to retire piece lists
altoghter and use pop_lsb() to loop across
pieces and serialize the moves. In this way
we just need to change definition of Position::square
to something like:

template<PieceType Pt> inline
Square Position::square(Color c) const {
  return lsb(byColorBB[c]);
}

No functional change.
2015-08-04 09:51:06 +02:00
Marco Costalba 68d61b80c6 Simplify IID depth formula
Restore original formula messed up during
half-ply removal.

STC
LLR: 4.11 (-2.94,2.94) [-3.00,1.00]
Total: 21349 W: 4091 L: 3909 D: 13349

LTC
LLR: 5.42 (-2.94,2.94) [-3.00,1.00]
Total: 52819 W: 8321 L: 8122 D: 36376

bench: 8040572
2015-07-30 09:54:25 +02:00
VoyagerOne a2410227cc PV refutation penalty
Extra penalty for PV move in previous ply when it gets refuted.

STC:
LLR: 4.49 (-2.94,2.94) [-1.50,4.50]
Total: 41094 W: 7889 L: 7620 D: 25585

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 12304 W: 1967 L: 1811 D: 8526

Bench: 8373608

Resolves #386
2015-07-30 00:51:49 +08:00
Alain SAVARD 4a6d59c6c7 MobilityArea (simplified)
Based off of Pull request #383:

Include squares occupied by some pawns in the MobilityArea
a) not blocked
b) on rank 4 and above
c) or captures

Passed STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 8157 W: 1644 L: 1516 D: 4997

And LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 26086 W: 4274 L: 4051 D: 17761

-----------

Then, a simplification test failed, trying to remove b and c)
LLR: -2.95 (-2.94,2.94) [-3.00,1.00]
Total: 6048 W: 1117 L: 1288 D: 3643

Another simplification test, was run to remove just (c)
Passed STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 28073 W: 5364 L: 5255 D: 17454

And LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 34652 W: 5448 L: 5348 D: 23856

A parameter tweak test showed that changing b) for "on rank 3 and above"
does not work
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 5233 W: 937 L: 1077 D: 3219

Finally, a small rewrite, and we have this version

Include squares occupied by some pawns in the MobilityArea which are
a) not blocked
b) on rank 4 and above

Bench: 8977899

Resolves #385
2015-07-30 00:47:51 +08:00
mstembera 51330414c4 Tuned version of TT replacement policy
If the used multiplier of 8 was any number larger than DEPTH_MAX
this would be a non functional patch.

STC 2MB
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 16353 W: 3216 L: 3066 D: 10071
http://tests.stockfishchess.org/tests/view/55a6d0630ebc590abbe1babd

LTC 8MB
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 134618 W: 21276 L: 20716 D: 92626
http://tests.stockfishchess.org/tests/view/55a74d760ebc590abbe1bad6

STC 16MB
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 22549 W: 4257 L: 4178 D: 14114
http://tests.stockfishchess.org/tests/view/55a9a2f90ebc590abbe1bb16

Bench: 7372460

Resolves #380
2015-07-25 00:33:53 +08:00
Marco Costalba 42b77ea7ee Tidy up in movepick.cpp
Some formattng after recent changes.

No functional change.
2015-07-19 12:38:33 +02:00
Marco Costalba 4095ff0ee5 Fix formatting of previous patch
No functional change.
2015-07-16 08:17:40 +02:00
VoyagerOne 1d977aa79e LMR Simplification: Remove countermove condition
Removed countermove condition for decreasing reduction.

LTC:
LLR: 3.01 (-2.94,2.94) [-3.00,1.00]
Total: 32410 W: 5092 L: 4986 D: 22332

STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 24450 W: 4632 L: 4517 D: 15301

Bench: 6943812

Resolves #378
2015-07-15 20:21:14 +01:00
mstembera d8aeaab79c Consistent TT replace policy
This fixes an inconsistency bug where TT entries were valued differently
depending on which pointer they were accessed through.

STC 2MB
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 11301 W: 2176 L: 2038 D: 7087

LTC 8MB
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17732 W: 2870 L: 2745 D: 12117

STC 16MB
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 17401 W: 3324 L: 3227 D: 10850

Bench: 8248164

Resolves #377
2015-07-15 20:17:16 +01:00
joergoster 3054db989d Use distance<file>() function in endgame.cpp
This one occurance of distance function was most likely overlooked.

No functional change.

Resolves #376
2015-07-12 20:48:50 +01:00
Marco Costalba b2edac7075 Remove useless razoring condition
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20626 W: 3977 L: 3855 D: 12794

LTC:
LLR: 3.10 (-2.94,2.94) [-3.00,1.00]
Total: 87334 W: 13675 L: 13648 D: 60011

Retire also the now unused pawn_on_7th() helper.

bench: 8248166
2015-07-09 08:13:30 +02:00
VoyagerOne aa0dc16b75 CMH Fix: Exclude captures for TT move refutation penalty
This will make sure we store only quiet moves for TT Penalty.

STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 34748 W: 6617 L: 6420 D: 21711

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 19975 W: 3259 L: 3137 D: 13579

Bench: 8063826

Resolves #373
2015-07-04 14:22:41 +01:00
Marco Costalba 112607bf49 Correctly check for no-makefile compile
Under Windows with MSVC we use the IDE to compile,
in this case we infer some compiler flags usually
set by Makefile.

The condition to check this was wrong, namely when compiling
with mingw under Windows 64 bit we always set IS_64BIT and
USE_BSFQ even if compiled with ARCH=x86-32 (this is how I
found it).

Small code style touches while there.

No functional change.
2015-06-28 10:56:42 +02:00
VoyagerOne 7e9f776fca LMR Tweak: Decrease reduction if cmh>0 && history>0.
STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 9627 W: 1879 L: 1748 D: 6000

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 21543 W: 3433 L: 3234 D: 14876

Bench: 8646407

Resolves #370 #371
2015-06-27 09:45:59 +01:00
Marco Costalba 3f1bc9e013 Fix compile on icc
Error is:

  a value of type "int" cannot be assigned to an entity of type "Value"

Also retire the now unused squares_of_color() function.

No functional change.
2015-06-25 05:29:56 +02:00
joergoster e8cc5c9968 Small coding style fix for Outpost array
No functional change

Resolves #367
2015-06-17 15:02:57 +01:00
lucasart 00563be5cc Retire -Wno-long-long
long long is part of the C++11 standard.

No functional change.

Resolves #364
2015-06-13 16:59:50 +01:00
lucasart ad87d707ff Simplify outpost evaluation
STC

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 114149 W: 21986 L: 22032 D: 70131

LTC

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 52232 W: 8468 L: 8396 D: 35368

Bench: 6716940

No functional change

Resolves #363
2015-06-07 09:49:01 +01:00
Vince Negri 7b8ffe0f5d Remove intermediate re-search in LMR
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20149 W: 3830 L: 3707 D: 12612

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 45384 W: 7089 L: 7006 D: 31289

Bench: 8110365

Resolves #361
2015-06-05 21:03:10 +01:00
Marco Costalba c1d3bd2dba Rename stages
Hopefully more clear.

No functional change.
2015-06-04 15:51:45 +02:00
lucasart 5ebdf2f8c8 Tune pawn shelter/storm
STC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 31679 W: 6183 L: 5912 D: 19584

LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 43550 W: 6885 L: 6602 D: 30063

Bench: 9219343

Resolves #360
2015-06-02 12:09:19 -07:00
Stéphane Nicolet 4502a1934f Tweak backward pawns definition
Advanced pawns cannot be backward. Also lower the backward penalty in
endgame.

Passed STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 18534 W: 3588 L: 3433 D: 11513

and LTC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 21319 W: 3415 L: 3217 D: 14687

Bench: 7271152

Resolves #359
2015-06-02 12:07:02 -07:00
Marco Costalba 2795aedbc3 Checking for rook color when setting castling
In Chess960 we can have legal positions with
opponent rook in A or H file and with castling
available, for instance:

4k3/pppppppp/8/8/8/8/PPPPPPPP/rR2K3 w Q - 0 1

In those cases we pick up the wrong rook when
setting castling.

Fix it by checking the color of the rook.

Bug reported by Matthew Lai.

No functional change.
2015-05-29 05:38:40 +02:00
lucasart 8e95c39bfa Simplify backward pawn scoring
STC

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 52322 W: 10011 L: 9945 D: 32366

LTC

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14143 W: 2334 L: 2203 D: 9606

Bench: 7976423

Resolves #354
2015-05-27 20:05:20 +01:00
Joona Kiiski 80a759d544 Resolve build failure for Mac
Remove '-Wl' switch from gcc arguments when compiling for Mac

No functional change

Resolves #353
2015-05-24 13:07:15 +01:00
lucasart 2e0ec4ad6f Fix merge error for Tuned PSQT
Fall-out from 411e704f

Bench: 7907776

Resolves #352
2015-05-21 19:56:01 +01:00
Stefan Geschwentner e14046517e Remove Gain Stats
Additionally in futility pruning the margin is raised for compensation.

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 48481 W: 9229 L: 9156 D: 30096

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32058 W: 5134 L: 5031 D: 21893

Bench: 8098149

Resolves #350
2015-05-18 13:59:30 -07:00
lucasart 411e704fdf Tuned PSQT
STC:
LLR: 3.11 (-2.94,2.94) [-0.50,4.50]
Total: 58764 W: 11530 L: 11185 D: 36049

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.50]
Total: 282710 W: 46339 L: 45209 D: 191162

Bench: 8512947

Resolves #349
2015-05-18 13:54:45 -07:00
Marco Costalba ee0371f86e Cleanup work in misc.cpp
Also some code style tidy up of latest patches.

Also renamed checkSq -> checkSquares because it
is a bitboard and not a square.

No functional change.
2015-05-10 09:42:26 +02:00
mstembera eaeb63f1d0 Smart TT save
Don't overwrite more valuable data with less valuable data

STC 2MB
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 21132 W: 4108 L: 3946 D: 13078
http://tests.stockfishchess.org/tests/view/5547d59f0ebc5940ca5d6883

LTC 8MB
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 13381 W: 2149 L: 1987 D: 9245
http://tests.stockfishchess.org/tests/view/5549b5a80ebc5940ca5d68b9

STC 16MB regression w/ zero effective hash pressure
LLR: 2.96 (-2.94,2.94) [-5.00,0.00]
Total: 18944 W: 3607 L: 3564 D: 11773
http://tests.stockfishchess.org/tests/view/554b0fda0ebc5940ca5d68ea

Bench: 8787152

Resolves #347
2015-05-09 17:43:57 +01:00
lucasart 92dbee7875 Edge distance
Instead of crafting a clever formula to calculate the array offset, simply use a
3 dimensional array. Remove the comment while at it, because now the code is
self-documenting.

No functional change.

Resolves #344
2015-05-09 14:06:19 +01:00
lucasart 2e86d1febc Restore deterministic search state
Introduce helper function Search::reset() which clears all kind of search
memory, in order to restore a deterministic search state.

Generalize TT.clear() into Search::reset() for the following use cases:
- bench: needed to guarantee deterministic bench (ie. if you call bench from
interactive command line twice in a row you get the same value).
- Clear Hash: restore clean search state, which is the purpose of this button.
- ucinewgame: ditto.

No functional change.

Resolves #346
2015-05-07 14:20:32 -07:00
lucasart 946fa47625 Never clear stats
Based on an idea and patch by VoyagerOne.

Small simplification, but was tedted for an ELO gain anyway.

STC:
LLR: 2.95 (-2.94,2.94) [-1.00,4.00]
Total: 5375 W: 1119 L: 977 D: 3279

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,5.00]
Total: 17893 W: 2984 L: 2792 D: 12117

bench 8322847
2015-05-06 19:17:33 +08:00
Marco Costalba 7231b18af1 Halve PSQT row data
Use symmetry along vertical middle axis of the board
to reduce the number of parameters.

For instance psqt value of SQ_A5 == SQ_A4 and value of
SQ_F8 == SQ_F1.

This is always true, at least until now nobody came in
with an asymmetric psqt table that worked.

Original patch by Lucas.

No functional change.
2015-05-03 20:07:52 +02:00
Marco Costalba 578b21bbee Split PSQT init from Position init
Easier for tuning psq tables:

TUNE(myParameters, PSQT::init);

Also move PSQT code in a new *.cpp file, and retire the
old and hacky psqtab.h that required to be included only
once to work correctly, this is not idiomatic for a header
file.

Give wide visibility to psq tables (previously visible only
in position.cpp), this will easy the use of psq tables outside
Position, for instance in move ordering.

Finally trivial code style fixes of the latest patches.

Original patch of Lucas Braesch.

No functional change.
2015-05-03 20:07:15 +02:00
Stefano80 59f64fda4f Improve ordering of good captures using rank term
Rank based term improved approximation of pos.see() for scoring good
captures.

STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 4632 W: 945 L: 827 D: 2860

LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 25770 W: 4184 L: 3964 D: 17622

Bench: 7593704

Resolves #342
2015-05-03 10:19:34 -07:00
Stefano80 cc54a91ca5 Replace MVV/LVA by MVV for good captures
Passed STC

LLR: 3.71 (-2.94,2.94) [-3.00,1.00]
Total: 64363 W: 12299 L: 12214 D: 39850

and LTC

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 69976 W: 11056 L: 11011 D: 47909

Bench: 8012532

Resolves #340
2015-04-28 22:07:48 +01:00
VoyagerOne b7063ef65b Change extra ply LMR condition to: cmh <= 0 && hist < 0
Extra ply LMR condition is now cmh <= 0 && h < 0
Instead of cmh + h < 0

STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 55210 W: 10812 L: 10557 D: 33841

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 13212 W: 2239 L: 2045 D: 8928

Bench: 8420865

Resolves #339
2015-04-26 20:12:25 +01:00
Alain SAVARD 078625e333 Exclude queen from Rook Contact Check computation
In ei.attackedBy, Queen does not x-ray through Rook, but the Rook does
X-ray through the Queen.

So most of the rook contact checks supported by queen are, in fact,
Queen Contact Checks and they are already scored separately.

Bench: 7762189

Resolves #338
2015-04-18 22:17:46 +01:00
lucasart 6c040c821a Retire FORCE_INLINE
No speed regression on my machine (i7-3770k, gcc 4.9.1, linux 3.16):

        stat        test     master   diff
        mean   2,482,415  2,474,987  7,906
        stdev      4,603      5,644  2,497

        speedup        0.32%
        P(speedup>0)  100.0%

Fishtest 9+0.03:

ELO: 0.26 +-1.8 (95%) LOS: 61.2%
Total: 60000 W: 12437 L: 12392 D: 35171

No functional change.

Resolves #334
2015-04-15 21:21:45 +01:00
VoyagerOne 20e92895af Removed extra condition (history < 0) in LMR to help sync up with move ordering.
LMR condition is now cmh+history<0
Instead of history<0 OR cmh+history<0

STC:
LLR: 2.96 (-2.94,2.94) [-3.00, 1.00]
Total: 26446 W: 5092 L: 4980 D: 16374

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00, 1.00]
Total: 14129 W: 2340 L: 2209 D: 9580

Bench: 7815183

Resolves #331
2015-04-12 20:05:59 +01:00
Marco Costalba fb03188fc7 Assorted cleanup of last patches
No functional change.
2015-04-11 23:24:43 +02:00
Marco Costalba a66c73deef Allow Position::init() to be called more than once
Currently Zobrist::castling[] are not properly zeroed
and rely on the compiler to do this at startup, but this
makes Position::init() to set different values every time
it is called!

This is a bit odd, and although not impacting normal usage,
can yield to subtle misbehaviour, very difficult to track
down, in case we happen to call it more than once for some
reason. I found this while developing tuning support and
it took me a while to track it down.

So properly init Zobrist::castling[]

No functional change.

Resolves #329
2015-04-10 20:39:15 +01:00
Stefan Geschwentner 27efc5ac99 Update stats at pv nodes
If a quiet best move is found at a pv node then always update stats.

STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 41485 W: 8047 L: 7830 D: 25608

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 14351 W: 2420 L: 2250 D: 9681

Bench: 6985247

Resolves #330
2015-04-10 20:34:55 +01:00
mstembera 36f2133df3 New formula for quiet move scoring: 3 * cmh + 1 * hist
STC

LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 45363 W: 8759 L: 8532 D: 28072

LTC

LLR: 3.51 (-2.94,2.94) [0.00,4.00]
Total: 125092 W: 20032 L: 19468 D: 85592

Bench: 7058819

Resolves #328
2015-04-10 20:14:11 +01:00
Stefan Geschwentner ef4d89c9bd update stats also in check
Update stats also if in check (drop condition).

STC:
LLR: 3.22 (-2.94,2.94) [-3.00,1.00]
Total: 87472 W: 16929 L: 16913 D: 53630

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 39971 W: 6436 L: 6345 D: 27190

Bench: 7086031

Resolves #327
2015-04-09 20:41:08 +01:00
lucasart aaf17326e2 Prune evasions when we can castle
A minor simplification.

STC:

LLR: 2.95 (-2.94,2.94) [-3.50,0.50]
Total: 67877 W: 12882 L: 12904 D: 42091

STC:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20677 W: 4023 L: 3901 D: 12753

LTC:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 12221 W: 2022 L: 1888 D: 8311

Bench: 7911336

Resolves #326
2015-04-09 20:34:06 +01:00
Stéphane Nicolet 2ca142a5b4 Use minimumSplitDepth = 5
Using minimumSplitDepth = 5 seems to be the best compromise in the
current SMP implementation

STC, 11 threads:

ELO: 14.87 +-4.1 (95%) LOS: 100.0%
Total: 8509 W: 1497 L: 1133 D: 5879

STC, 4 threads:

ELO: 0.30 +-2.8 (95%) LOS: 58.2%
Total: 20000 W: 3365 L: 3348 D: 13287

STC, 2 threads:

ELO: -1.02 +-2.0 (95%) LOS: 16.4%
Total: 40000 W: 7087 L: 7204 D: 25709

Resolves #324
2015-04-09 20:32:36 +01:00
Marco Costalba ffbaa1c5a8 Fix MSVC warning from previous patch
No functional change.
2015-04-03 19:29:29 +02:00
Marco Costalba 57dfe41ddb Fix elapsed()
Messed up during merge.

No functional change.
2015-04-03 04:53:53 +02:00
Marco Costalba 926f215061 Add support for playing in 'nodes as time' mode
When running more games in parallel, or simply when running a game
with a background process, due to how OS scheduling works, there is no
guarantee that the CPU resources allocated evenly between the two
players. This introduces noise in the result that leads to unreliable
result and in the worst cases can even invalidate the result. For
instance in SF test framework we avoid running from clouds virtual
machines because are a known source of very unstable CPU speed.

To overcome this issue, without requiring changes to the GUI, the idea
is to use searched nodes instead of time, and to convert time to
available nodes upfront, at the beginning of the game.

When nodestime UCI option is set at a given nodes per milliseconds
(npmsec), at the beginning of the game (and only once), the engine
reads the available time to think, sent by the GUI with 'go wtime x'
UCI command. Then it translates time in available nodes (nodes =
npmsec * x), then feeds available nodes instead of time to the time
management logic and starts the search. During the search the engine
checks the searched nodes against the available ones in such a way
that all the time management logic still fully applies, and the game
mimics a real one played on real time. When the search finishes,
before returning best move, the total available nodes are updated,
subtracting the real searched nodes. After the first move, the time
information sent by the GUI is ignored, and the engine fully relies on
the updated total available nodes to feed time management.

To avoid time losses, the speed of the engine (npms) must be set to a
value lower than real speed so that if the real TC is for instance 30
secs, and npms is half of the real speed, the game will last on
average 15 secs, so much less than the TC limit, providing for a
safety 'time buffer'.

There are 2 main limitations with this mode.

1. Engine speed should be the same for both players, and this limits
the approach to mainly parameter tuning patches.

2. Because npms is fixed while, in real engines, the speed increases
toward endgame, this introduces an artifact that is equivalent to an
altered time management. Namely it is like the time management gives
less available time than what should be in standard case.

May be the second limitation could be mitigated in a future with a
smarter 'dynamic npms' approach.

Tests shows that the standard deviation of the results with 'nodestime'
is lower than in standard TC, as is expected because now all the introduced
noise due the random speed variability of the engines during the game is
fully removed.

Original NIT idea by Michael Hoffman that shows how to play in NIT mode
without requiring changes to the GUI. This implementation goes a bit
further, the key difference is that we read TC from GUI only once upfront
instead of re-reading after every move as in Michael's implementation.

No functional change.
2015-04-03 04:40:55 +02:00
Marco Costalba df722521ba Rename of TimeMgr and friends
More natural naming IMO.

No functional change.
2015-04-03 04:19:29 +02:00
Marco Costalba 5d1b92e8f9 Introduce elapsed_time()
And reformat a bit time manager code.

Note that now we set starting search time in think() and
no more in ThreadPool::start_thinking(), the added delay
is less than 1 msec, so below timer resolution (5msec) and
should not affect time lossses ratio.

No functional change.
2015-04-03 04:19:26 +02:00
mstembera 6661a31541 Simplification to use only one counter move.
STC http://tests.stockfishchess.org/tests/view/5518dca30ebc5902160ec5d0
LLR: 2.95 (-2.94,2.94) [-3.50,0.50]
Total: 18868 W: 3638 L: 3530 D: 11700

LTC http://tests.stockfishchess.org/tests/view/5518f7ed0ebc5902160ec5d4
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 69767 W: 11019 L: 10973 D: 47775

Extracted from http://tests.stockfishchess.org/tests/view/5511028a0ebc5902160ec40b

Original patch by hxim.  All credit goes to him.

Bench: 7664249

Resolves #320
2015-04-03 01:16:15 +08:00
Marco Costalba 6c42575208 Assorted code style of latest commits
No functional chnage.
2015-03-29 10:16:10 +02:00
Joona Kiiski 61a6e46664 Fix indentations for hanging pawns code
No functional change
2015-03-29 08:54:25 +01:00
lucasart 60beb18efc Remove some difficult to understand C++11 constructs
Code like this is more a case of showing off one's C++ knowledge, rather than
using it adequately, IMHO.

**First loop (std::generate)**

Iterators are inadequate here, because they lose the key information which is
idx. As a result, we need to carry a redundant idx variable, and increment it
along the way. Very clumsy.
Usage of std::generate and a lambda function only obfuscate the code, which is
merely a simple and stupid loop over the elements of a vector.

**Second loop (std::accumulate)**

This code is thoroughlly incomprehensible. Restore the original, which was much
simpler to understand.

**Third loop (range based loop)**

Again, a range based loop is inadequate, because we lose idx! To resolve this
artificially created problem, the data model was made redundant (idx is a data
member of db[] elements!?), which is ugly and unjustified. A simple and stupid
for loop with idx does the job much better.

No functional change.

Resolves #313
2015-03-29 08:40:24 +01:00
Alain SAVARD 8463fa479e PSV3_1
Small speed-up in pawn.cpp
Results for 10 tests for each version:

Base      Test      Diff
Mean    1435636   1445238   -9602
StDev   22576     23189     1848

p-value: 1
speedup: 0.007

No functional change

Resolves #295
2015-03-29 07:30:46 +08:00
Ajith 63500274db Give a reduced bonus for threats by hanging pawns
Passed STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 105539 W: 20389 L: 20001 D: 65149

and LTC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 9629 W: 1577 L: 1432 D: 6620

Bench: 7658627

Resolves #317
2015-03-29 07:28:47 +08:00
mbootsector 1d5eaba573 Retire follow-up move heuristic
STC: http://tests.stockfishchess.org/tests/view/5501d0f30ebc5902160ec0fd
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 34891 W: 6904 L: 6808 D: 21179

LTC: http://tests.stockfishchess.org/tests/view/550328540ebc5902160ec133
LLR: 3.10 (-2.94,2.94) [-3.00,1.00]
Total: 182653 W: 29866 L: 29993 D: 122794

Bench: 8396161

Resolves #310
2015-03-28 22:12:06 +00:00
VoyagerOne ac8e6ff000 Use CounterMoveHistory when calculating LMR for cut nodes
If the sum of CounterMoveHistory heuristic and History heuristic is below zero,
then reduce an extra ply in cut nodes

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6479 W: 1099 L: 967 D: 4413

Bench: 7773299

Resolves #315
2015-03-28 21:15:49 +00:00
Marco Costalba 3a6753328c Clean up previous patch
No functional change.
2015-03-25 07:05:27 +01:00
joergoster 50182f1100 Tuned values for the pawn piece square table
Quick measure at very fast tc:
ELO: 4.77 +-2.2 (95%) LOS: 100.0%
Total: 40124 W: 8711 L: 8160 D: 23253

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 52284 W: 8880 L: 8559 D: 34845

Bench: 8865736

Resolves #311
2015-03-25 07:39:44 +08:00
VoyagerOne e51965aa57 Introduce a new counter move history penalty
Extra penalty for TT move in previous ply when it gets refuted

STC:

LLR: 2.94 (-2.94,2.94) [-1.50,4.50]
Total: 31303 W: 6216 L: 6025 D: 19062

LTC:

LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 6950 W: 1189 L: 1054 D: 4707

Bench: 8191926

Resolves #309
2015-03-24 23:04:08 +00:00
Joona Kiiski a7381d5e81 Fully yielding locks, no spinning
7 threads:

ELO: 2.00 +-2.7 (95%) LOS: 92.4%
Total: 20000 W: 3276 L: 3161 D: 13563

There is no functional change in single thread mode

Resolves #304
2015-03-24 21:34:19 +00:00
Marco Costalba dc3a5f791e Allow Bitbases::init() to be called more than once
Currently if we call it more than once, we crash.

This is not a real problem, because this function is
indeed called just once. Nevertheless with this small fix,
that gets rid of a hidden 'static' variable, we cleanly
resolve the issue.

While there, fix also ThreadPool::exit to return in a
consistent state. Now all the init() functions but
UCI::init() are reentrant and can be called multiple
times.

No functional change.
2015-03-23 17:14:31 +01:00
Marco Costalba 35a082064f Double magics generation speed
Profiling shows that resetting attacks table after
a failed candidate magic attempt is the biggest
time consumer, so rewrite the logic avoiding the
memset()

Magics init for rook+bishop goes from 200msecs to
under 100msec.

No functional change.
2015-03-23 17:10:38 +01:00
Marco Costalba be77406a55 Get rid of nativeThread
No functional change.
2015-03-23 09:02:52 +01:00
Marco Costalba 26dabb1e6b Use only one ConditionVariable to sync UI
To sync UI with main thread it is enough a single
condition variable because here we have a single
producer / single consumer design pattern.

Two condition variables are strictly needed just for
many producers / many consumers case.

Note that this is possible because now we don't send to
sleep idle threads anymore while searching, so that now
only UI can wake up the main thread and we can use the
same ConditionVariable for both threads.

The natural consequence is to retire wait_for_think_finished()
and move all the logic under MainThread class, yielding the
rename of teh function to join()

No functional change.
2015-03-21 07:55:33 +01:00
lucasart 35b6079852 Fix comment
We always probe, but we do not prune at PV nodes.

No functional change.

Resolves #300
2015-03-20 22:40:03 +00:00
Marco Costalba ebf3735754 Retire ConditionVariable
Now that we use spinlocks everywhere and don't put
threads to sleep while idle, we can use the slower
(but no more in hot path) std::condition_variable_any
instead of our homwgrown ConditionVariable struct.

Verified fo rno regression at STC with 7 threads:
ELO: -0.66 +-2.7 (95%) LOS: 31.8%
Total: 20000 W: 3210 L: 3248 D: 13542

No functional change
2015-03-20 03:05:25 +01:00
joergoster 966bc477af Tuned mobility with another SPSA run
Further improved mobility values after another SPSA session, 50k
iterations.

Elo measure at very fast 9+0.05":
ELO: 3.40 +-2.2 (95%) LOS: 99.9%
Total: 40000 W: 8434 L: 8042 D: 23524

and LTC SPRT[0, 4]:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 11052 W: 1874 L: 1687 D: 7491

Bench: 8226843

Resolves #301
2015-03-20 08:08:21 +08:00
joergoster c6f987d1ad Fix the comment for Position::is_draw()
We no longer check for insufficient material.

No functional change

Resolves #299
2015-03-18 20:30:50 +00:00
lucasart d477a0d611 connected should be bool, not Bitboard
There's no reason to define it as a Bitboard, so for consistency, use bool.

This is even a speedup on my machine: i7-3770k, using gcc 4.9.1 (linux):

    stat        test     master    diff
    mean   2,341,338  2,327,998  13,134
    stdev     15,765     14,717   5,405

    speedup       0.56%
    P(speedup>0) 100.0%

No functional change.

Resolves #298
2015-03-18 20:21:41 +00:00
Marco Costalba 54889618c2 Reformat FastMove
Align to SF coding style.

Verified no regression:

LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 55938 W: 10893 L: 10835 D: 34210

No functional change.
2015-03-18 08:12:59 +01:00
Marco Costalba 9a6cfee73b Simplify nosleep logic
Avoid redundant 'while' conditions. It is enough to
check them in the outer loop.

Quick tested for no regression 10K games at 4 threads
ELO: -1.32 +-3.9 (95%) LOS: 25.6%
Total: 10000 W: 1653 L: 1691 D: 6656

No functional change.
2015-03-18 08:01:50 +01:00
Marco Costalba 2e8552db76 Fix a bogus use of mutex
Spinlock must be used instead.

Tested for no regression at 15+0.05 th 4:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25928 W: 4303 L: 4190 D: 17435

No functional change.

Resolves #297
2015-03-17 08:19:29 +00:00
Marco Costalba a4b2eeea75 Re-arrange history update code
Unify the quites moves loop for both cases,
the compiler optimizes away the

 if (is_ok((ss-1)->currentMove))

inside loop, so that the result is same
speed as original.

No functional change.
2015-03-16 15:14:09 +01:00
Marco Costalba 13d4df95cd Use acquire() and release() for spinlocks
It is more idiomatick than lock() and unlock()

No functional change.
2015-03-16 08:14:08 +01:00
Vince Negri bae4679de2 Fix dependency generation for MacOSX
No functional change

Resolves #290
2015-03-15 20:31:43 +00:00
Joona Kiiski a3943b2aa7 Fix dependency generation for C++11
No functional change

Resolves #291
2015-03-15 20:26:08 +00:00
Joona Kiiski f04f50b368 Do not sleep, but yield
During the search, do not block on condition variable, but instead use std::this_thread::yield().

Clear gain with 16 threads. Again results vary highly depending on hardware, but on average it's a clear gain.

ELO: 12.17 +-4.3 (95%) LOS: 100.0%
Total: 7998 W: 1407 L: 1127 D: 5464

There is no functional change in single thread mode

Resolves #294
2015-03-15 19:45:30 +00:00
joergoster a4b98a052e New values for Mobility and Outposts.
Both are the result of a SPSA tuning session with a custom book, 50k iterations each.

After an additional tuning session of the mobility values, tuning the delta values, with following result.

40k games at 9+0.05:
ELO: 4.13 +-2.2 (95%) LOS: 100.0%
Total: 40000 W: 8581 L: 8106 D: 23313

and LTC
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 36518 W: 6049 L: 5782 D: 24687

Bench: 8567402

Resolves #284
2015-03-14 21:35:52 +00:00
Marco Costalba e5da0e4b79 Link with -static in mingw
Fixes reported startup error about missing libwinpthread-1.dll
when the dll is not in the path.

The current -static-xxxx flags, introduced with:

https://github.com/official-stockfish/Stockfish/commit/373503f4a9a990054b5

Only take in account standard libraries, but not thread
library.

No functional change.

Resolves #289
2015-03-14 19:23:54 +00:00
Joona Kiiski d71f707040 Introduce yielding spin locks
Idea and original implementation by Stephane Nicolet

7 threads 15+0.05
ELO: 3.54 +-2.9 (95%) LOS: 99.2%
Total: 17971 W: 2976 L: 2793 D: 12202

There is no functional change in single thread mode
2015-03-14 19:14:52 +00:00
Joona Kiiski 558b0c848c Remove check for gcc version from Makefile.
This check is obsolete.
very old gcc versions can't compile c++11 code.

No functional change

Resolves #285
2015-03-13 20:29:28 +00:00
mstembera 062ca91db5 New easy move implementation
Spend much less time in positions where one move is much better than all other alternatives.
We carry forward pv stability information from the previous search to identify such positions.
It's based on my old InstaMove idea but with two significant improvements.

1) Much better instability detection inside the search itself.
2) When it's time to make a FastMove we no longer make it instantly but still spend at least 10% of normal time verifying it.

Credit to Gull for the inspiration.
BIG thanks to Gary because this would not work without accurate PV!

20K
ELO: 8.22 +-3.0 (95%) LOS: 100.0%
Total: 20000 W: 4203 L: 3730 D: 12067

STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 23266 W: 4662 L: 4492 D: 14112

LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 12470 W: 2091 L: 1931 D: 8448

Resolves #283
2015-03-12 19:49:30 +00:00
Stefan Geschwentner 13c11f4048 Introduce Counter Move History tables
Introduce a counter move history table which additionally is indexed by the last move's piece and target square.
For quiet move ordering use now the sum of standard and counter move history table.

STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 4747 W: 1005 L: 885 D: 2857

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 5726 W: 1001 L: 872 D: 3853

Because of reported low NPS on multi core test
STC (7 threads):
ELO: 7.26 +-3.3 (95%) LOS: 100.0%
Total: 14937 W: 2710 L: 2398 D: 9829

Bench: 7725341

Resolves #282
2015-03-12 07:29:57 +00:00
Joona Kiiski 81c7975dcd Use thread specific mutexes instead of a global one.
This is necessary to improve the scalability with high number of cores.

There is no functional change in a single thread mode.

Resolves #281
2015-03-11 21:59:34 +00:00
Marco Costalba 4b59347194 Retire spinlocks
Use Mutex instead.

This is in preparaation for merging with master branch,
where we stilll don't have spinlocks.

Eventually spinlocks will be readded in some future
patch, once c++11 has been merged.

No functional change.
2015-03-11 21:20:47 +01:00
Marco Costalba 6027652773 Cleanup thread_win.h
No functional change.
2015-03-10 17:13:52 +01:00
Marco Costalba 04372316b3 Disable spinlocks
To allow testing on fishtest.

No functional change.
2015-03-10 12:47:49 +01:00
Marco Costalba 8725494966 Add thread_win32.h header
Workaround slow std::thread implementation in mingw
and gcc for Windows with our own old low level thread
functions.

No functional change.
2015-03-10 12:42:40 +01:00
Marco Costalba be50952533 Sync with master
bench: 8285241
2015-03-07 10:56:01 +01:00
Marco Costalba e53774bc49 Sync with master
No functional change.
2015-03-07 10:45:58 +01:00
hxim 3017e8c604 Transform minKingPawnDistance into a local variable
minKingPawnDistance is used only as local variable in one place so we don't need it to be part of "Pawns::Entry" structure.

No functional change.

Resolves #277
2015-03-07 09:03:19 +00:00
Stéphane Nicolet 8fcaa78a04 Update Makefile for Mac OS X compilation
This change in the Makefile restores the possibility to compile
Stockfish on Mac OS X 10.9 and 10.10 after the C++11 has been merged.

To use the default (fastest) settings, compile with:

make build ARCH=x86-64-modern

To test the clang settings, compile with

make build ARCH=x86-64-modern COMP=clang

Beware that the clang settings may provide a slightly slower (6%)
executable.

Backported from master.

No functional change

Resolves #275
2015-03-07 08:39:21 +01:00
Marco Costalba a590d1d52d Re-enable spinlocks
For branch C++11, that doe snot run on fishtest,
there is no need of this kludge, let only master
have it.

No functional change.
2015-03-07 08:38:26 +01:00
Joona Kiiski 856a5f3aaa Revert C++11 merge
Restore the state of repo back to commit 'Simplify pawn code a bit' (1e6d21dbb6)

No functional change
2015-03-07 07:38:22 +00:00
Stéphane Nicolet 6fa6da3ee1 Update Makefile for Mac OS X compilation
This change in the Makefile restores the possibility to compile
Stockfish on Mac OS X 10.9 and 10.10 after the C++11 has been merged.

To use the default (fastest) settings, compile with:

make build ARCH=x86-64-modern

To test the clang settings, compile with

make build ARCH=x86-64-modern COMP=clang

Beware that the clang settings may provide a slightly slower (6%)
executable.

No functional change

Resolves #275
2015-03-05 04:18:59 +00:00
Marco Costalba cb2111f0b6 Disable spinlocks
Now that c++11 branch has been merged in master,
disable unconditionally the spinlocks and use mutex
instead. This will allow to run fishtest even on HT
machines withouth changes.

In the future we will reintorduce spinlocks, once
we will have took care of fishtest.

No functional change.
2015-03-02 08:11:39 +01:00
Marco Costalba 6645115377 Allow to disable spinlocks
And use mutex instead. You may never want to do this.
It is a workaround to run c++11 on fishtest where many
machiens have HTenabled and this can be a problem when
number of cores set is higher than number of physical cores.

To disable spinlocks, just compile with -DNO_SPINLOCK flag

No functional change.
2015-03-01 17:16:05 +01:00
Marco Costalba 63a5fc2366 Rename available_to()
Change this API to be more natural and simple.

Inspired by a patch by Joona.

No functional change.
2015-03-01 12:33:05 +01:00
Marco Costalba 0da7295795 Sync with master
bench: 8285241
2015-02-28 20:22:28 +01:00
Stéphane Nicolet 1e6d21dbb6 Simplify pawn code a bit
Simplify a bit the number of bitwise operators used to calculate the
pawn evaluation in pawns.cpp

No functional change.

Resolves #269
2015-03-01 01:01:37 +08:00
Stéphane Nicolet 9369f4963d Raise penalty for knight attacked by pawn
Raise a bit the penalty for knight attacked by pawn.

STC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 27744 W: 5563 L: 5380 D: 16801

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 87571 W: 14747 L: 14314 D: 58510

Bench: 8285241

Resolves #270
2015-03-01 00:58:32 +08:00
Marco Costalba a459709fcc Retire apply_weight()
Use the more natural operator*() instead.

No functional change.
2015-02-27 10:38:12 +01:00
Marco Costalba 2dbb1adf2a Sync with master
bench: 8069601
2015-02-26 20:50:39 +01:00
Marco Costalba 8a2c4134a6 Normalize twice supported pawns
Align codying style to current conventions and move
formula for twice supported pawns to Pawns::init()
where it should be.

No functional change.
2015-02-27 01:56:16 +08:00
Alain SAVARD fce799294c Apex Pawns
Pawns which are supported already have a bonus. Apex are pawns which are
supported twice.
This patch gives an additional 50% bonus for them.

STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 6549 W: 1333 L: 1209 D: 4007

LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 18002 W: 3037 L: 2850 D: 12115

Bench: 8069601

Resolves #267
2015-02-27 01:52:23 +08:00
Marco Costalba 0b36ba74fc Don't assume the type of Time::point
But instead use the proper definition. Also
rewrite chrono functions while there.

No functional change.
2015-02-24 14:08:14 +01:00
Marco Costalba 153fb216a1 Small tweaks in pawns.cpp
No functional change.
2015-02-24 11:45:29 +01:00
Marco Costalba d3d26a94b3 Improve spinlock implementation
Calling lock.test_and_set() in a tight loop creates expensive
memory synchronizations among processors and penalize other
running threads. So syncronize only only once at the beginning
with fetch_sub() and then loop on a simple load() that puts much
less pressure on the system.

Reported about 2-3% speed up on various systems.

Patch by Ronald de Man.

No functional change.
2015-02-23 19:48:46 +01:00
Marco Costalba 38112060dc Use spinlock instead of mutex for Threads and SplitPoint
It is reported to be defenitly faster with increasing
number of threads, we go from a +3.5% with 4 threads
to a +15% with 16 threads.

The only drawback is that now when testing with more
threads than physical available cores, the speed slows
down to a crawl. This is expected and was similar at what
we had setting the old sleepingThreads to false.

No functional change.
2015-02-23 13:47:07 +01:00
Marco Costalba 775f8239d3 Introduce Spinlock class
Initialization is more complex than what I'd like due
to MSVC compatibility that for some reason does not like:

std::atomic_flag lock = ATOMIC_FLAG_INIT;

No functional change.
2015-02-23 13:37:46 +01:00
Marco Costalba 098f645d26 Sync with master
bench: 8253813
2015-02-23 13:36:15 +01:00
Marco Costalba 29ff925fd8 Fix build under OS X
Reported by Vince Negri

No functional change.
2015-02-22 16:33:24 +01:00
Marco Costalba 8d16111ffd Sync with master
bench: 8253813
2015-02-22 14:20:23 +01:00
Marco Costalba e2226cbb20 Use only 'level' as late join metric
It seems other metric are useless, this allow us
to simplify the code and to prune useless stuff.

STC 20K games 4 threads
ELO: -0.76 +-2.8 (95%) LOS: 29.9%
Total: 20000 W: 3477 L: 3521 D: 13002

STC 10K games 16 threads
ELO: 1.36 +-3.9 (95%) LOS: 75.0%
Total: 10000 W: 1690 L: 1651 D: 6659

bench: 8253813
2015-02-22 12:59:34 +01:00
Marco Costalba 5fd5453e59 Further refine SMP code
Backported from C++11 branch:

https://github.com/official-stockfish/Stockfish/commit/7ff965eebfbc17d2b
https://github.com/official-stockfish/Stockfish/commit/e74c2df907d5336d3d2b

Fully verified it is equivalent to master (see log msg
of individual commits for details).

No functional change.
2015-02-21 11:33:03 +01:00
Marco Costalba e74c2df907 Use sp->master instead of bestThread
Verified with:

dbg_hit_on(th != sp->master);

It is 100% equivalent on more than 200K hits.

No functional change.
2015-02-21 10:40:59 +01:00
Stéphane Nicolet 41ccc885ec Fix comment for kingAdjacentZoneAttacksCount
The comment for kingAdjacentZoneAttacksCount[] was bogus, using
reversed semantics for color.

No functional change

Resolves #262
2015-02-20 19:59:26 +00:00
snicolet 2f46592736 Mobile phalanxes
Try to create mobile phalanxes

STC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 52393 W: 10912 L: 10656 D: 30825

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 30398 W: 5315 L: 5063 D: 20020

Bench: 8253813

Resolves #261
2015-02-20 19:56:57 +00:00
Marco Costalba 7ff965eebf Improve comments in SMP code
No functional change.
2015-02-20 12:38:54 +01:00
Marco Costalba a6f873cd8d Use range-based-for in late join
No functional change.
2015-02-20 10:50:47 +01:00
Marco Costalba 40548c9153 Sync with master
bench: 7911944
2015-02-20 10:37:29 +01:00
Marco Costalba 667f350737 Clarify we don't late join with only 2 threads
Thanks to Gary for pointing this out.

No functional change.
2015-02-19 23:12:59 +01:00
Marco Costalba 950c8436ed Use size_t consistently across thread code
No functional change.
2015-02-19 10:43:28 +01:00
Marco Costalba 8d47caa16e Retire redundant sp->slavesCount field
It should be used slavesMask.count() instead.

Verified 100% equivalent when sp->allSlavesSearching:

dbg_hit_on(sp->allSlavesSearching, sp->slavesCount != sp->slavesMask.count());

No functional change.
2015-02-19 10:36:15 +01:00
Marco Costalba b9d4e6f7fd Fix a warning under MSVC
Assignment of size_t to int.

No functional change.
2015-02-19 10:18:24 +01:00
Marco Costalba 193a7ae35b Add a couple of asserts to late join
Document and clarify that we cannot rejoin on ourselves
and that we never late join if we are master and all
slaves have finished, inded in this case we exit idle_loop.

No functional change.
2015-02-19 10:08:29 +01:00
Marco Costalba 4f906a2589 Remove useless condition in late join
In case of Threads.size() == 2 we have that sp->allSlavesSearching
is always false (because we have finished our search), bestSp is
always NULL and we never late join, so there is no need to special
case here.

Tested with dbg_hit_on(sp && sp->allSlavesSearching) and
verified it never fires.

No functional change.
2015-02-19 09:53:39 +01:00
Marco Costalba dccaa145d2 Compute SplitPoint::spLevel on the fly
And retire a redundant field. This is important also
from a concept point of view becuase we want to keep
SMP structures as simple as possible with the only
strictly necessary data.

Verified with

dbg_hit_on(sp->spLevel != level)

that the values are 100% the same out of more 50K samples.

No functional change.
2015-02-18 21:50:35 +01:00
Marco Costalba 6656ed8904 Simplify attackUnits formula
Use '/ 8' instead of '* 31 / 256'

Passed STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 55077 W: 10999 L: 10940 D: 33138

And LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14751 W: 2530 L: 2400 D: 9821

bench: 7911944
2015-02-17 08:23:35 +01:00
Joona Kiiski d65f75c153 Improve smp performance for high number of threads
Balance threads between split points.

There are huge differences between different machines and autopurging makes it very difficult to measure the improvement in fishtest, but the following was recorded for 16 threads at 15+0.05:

    For Bravone (1000 games): 0 ELO
    For Glinscott (1000 games): +20 ELO
    For bKingUs (1000 games): +50 ELO
    For fastGM (1500 games): +50 ELO

The change was regression for no one, and a big improvement for some, so it should be fine to commit it.
Also for 8 threads at 15+0.05 we measured a statistically significant improvement:
ELO: 6.19 +-3.9 (95%) LOS: 99.9%
Total: 10325 W: 1824 L: 1640 D: 6861

Finally it was verified that there was no (significant) regression for

4 threads:
ELO: 0.09 +-2.8 (95%) LOS: 52.4%
Total: 19908 W: 3422 L: 3417 D: 13069

2 threads:
ELO: 0.38 +-3.0 (95%) LOS: 60.0%
Total: 19044 W: 3480 L: 3459 D: 12105

1 thread:
ELO: -1.27 +-2.1 (95%) LOS: 12.3%
Total: 40000 W: 7829 L: 7975 D: 24196

Resolves #258
2015-02-16 20:36:13 +00:00
lucasart f8f5dcbb68 Compute checkers from scratch
This micro-optimization only complicates the code and provides no benefit.
Removing it is even a speedup on my machine (i7-3770k, linux, gcc 4.9.1):

stat        test     master    diff
mean   2,403,118  2,390,904  12,214
stdev     12,043     10,620   3,677

speedup       0.51%
P(speedup>0) 100.0%

No functional change.
2015-02-16 09:34:26 +08:00
Marco Costalba 686b45e121 Retire one do_move() overload
After Lucas patch it is almost useless.

No functional change.
2015-02-15 12:23:03 +01:00
lucasart dc13004283 Compute checkers from scratch
This micro-optimization only complicates the code and provides no benefit.
Removing it is even a speedup on my machine (i7-3770k, linux, gcc 4.9.1):

stat        test     master    diff
mean   2,403,118  2,390,904  12,214
stdev     12,043     10,620   3,677

speedup       0.51%
P(speedup>0) 100.0%

No functional change.
2015-02-15 12:11:05 +01:00
Marco Costalba 901bfb1f55 Revert "Delayed killers checking"
It seems a slowdown when run with fishbench.

No functional change.
2015-02-15 11:32:53 +01:00
Marco Costalba 20a5c07472 Further simplify KingDanger init
And remove a tale whitespace while there.

No functional change.
2015-02-14 15:56:54 +01:00
Marco Costalba e38eb2302d Further simplify KingDanger init
And remove a tale whitespace while there.

No functional change.
2015-02-14 15:55:11 +01:00
Marco Costalba 0af24a1445 Sync with master
Bench: 7369224
2015-02-14 15:30:16 +01:00
snicolet a8f9c7a790 Small bonus for all safe pawn pushes
Pawn flexibility: add a small bonus for all safe pawn pushes

STC:
LLR: 2.70 (-2.94,2.94) [-1.50,4.50]
Total: 18233 W: 3705 L: 3557 D: 10971

LTC:
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 17684 W: 3042 L: 2854 D: 11788

Bench: 7369224

Resolves #253
2015-02-13 21:33:00 +00:00
Marco Costalba c0a80afe89 Reformat tracing functions
No functional change.
2015-02-13 11:54:46 +01:00
hxim 44a571c1c7 Fix KingDanger[] array initialization
Use integer arithmetic instead of floating point arithmetic.
Floating point arithmetic was causing different results for some 32-bit compiles

No functional change

Resolves #249
Resolves #250
2015-02-09 22:02:35 +00:00
Marco Costalba ce8ac7997c Sync with master
bench: 7699138
2015-02-08 21:32:14 +01:00
Joona Kiiski e118570038 Pawn Center Bind Bonus
Bonus for two pawns controlling the same central square

STC:

LLR: 3.14 (-2.94,2.94) [-1.50,4.50]
Total: 15974 W: 3291 L: 3133 D: 9550

LTC:

LLR: 3.24 (-2.94,2.94) [0.00,6.00]
Total: 10449 W: 1837 L: 1674 D: 6938

Idea from Lyudmil Tsvetkov.

Bench: 7699138

Resolves #248
2015-02-08 19:28:01 +00:00
Marco Costalba 8f10f6c9cd Shuffle put_piece() and friends signatures
It is more consistent with the others member functions.

No functional change.
2015-02-08 18:17:08 +01:00
Marco Costalba 3184852bdc Small tweaks in do_move and friends
Also remove useless StateCopySize64 optimization:
compiler uses SSE movups instruction anyhow and
does not need this trick (verified with fishbench).

No functional change.
2015-02-08 13:09:29 +01:00
Marco Costalba 99c9cae586 Avoid casting to char* in prefetch()
Funny enough, gcc __builtin_prefetch() expects
already a void*, instead Windows's _mm_prefetch()
requires a char*.

The patch allows to remove ugly casts from caller
sites.

No functional change.
2015-02-07 19:13:41 +01:00
Marco Costalba 152a4dc5cd Rewrite pos_is_ok()
No functional change.
2015-02-07 15:02:28 +01:00
Marco Costalba 47a0768102 Micro-optimize SEE
Results for 10 tests for each version (gcc 4.8.3 on mingw):

            Base      Test      Diff
    Mean    1502447   1507917   -5470
    StDev   3119      1364      4153

p-value: 0,906
speedup: 0,004

Results for 10 tests for each version (MSVC 2013):

            Base      Test      Diff
    Mean    1400899   1403713   -2814
    StDev   1273      2804      2700

p-value: 0,851
speedup: 0,002

No functional change.
2015-02-07 12:21:39 +01:00
Marco Costalba 170bdf40cd Rename dbg_hit_on_c() to dbg_hit_on()
Use an overload instead of a new named function.

I have found this handier and easier when adding
some quick debug code.

No functional change.
2015-02-07 11:18:06 +01:00
Marco Costalba 8b0fee9998 Rename dbg_hit_on_c() to dbg_hit_on()
Use an overload instead of a new named function.

I have found this handier and easier when adding
some quick debug code.

No functional change.
2015-02-07 11:15:38 +01:00
Marco Costalba 1277a42823 Sync with master
bench: 7696257
2015-02-07 10:32:28 +01:00
lucasart 35aa21c1fe Removes useless templates, some of which lead to code duplication: is_K*() functions.
No functional change

Resolves #245
2015-02-07 09:12:04 +00:00
Stefan Geschwentner 18b0809639 Add bonus for pawn attack threats
Latent pawn attacks: Add a bonus to safe pawn pushes which attacks an
enemy piece.  Based on an idea of Lyudmil Tsvetkov.

STC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7925 W: 1666 L: 1537 D: 4722

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 40109 W: 6841 L: 6546 D: 26722

Bench: 7696257

Resolves #240
2015-02-03 11:19:33 +08:00
mstembera f4136c5434 Profile build options
I went through all the individual compile options that differ between
-fprofile-generate/-fprofile-use  and  -fprofile-arcs/-fbranch-probabilities
and distilled the speed difference down to only turning off
-fno-peel-loops and -fno-tracer.  Using this we still get the full speedup
(maybe a bit more because other optimizations stay on) and it's also much cleaner
because we can get rid of the "@rm -f ucioption.gc*" hack for all versions of gcc.

No functional change.

Resolves #237
2015-02-03 11:09:37 +08:00
NicklasPersson ddccb5355c Improved King Safety values
From an SPSA-session on king safety.

STC:
ELO: 3.21 +-2.1 (95%) LOS: 99.8%
Total: 40000 W: 8181 L: 7812 D: 24007

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 47765 W: 8091 L: 7785 D: 31889

Bench: 8589262

Resolves #241
2015-02-03 04:00:52 +08:00
Marco Costalba 1c7a727795 Use move assignment in movegen.h
No functional change and same speed (tested with perft)
2015-02-01 16:42:31 +01:00
Marco Costalba e56e5045dd Allow to assign a Move to an ExtMove
After defining ExtMove::operator Move(), this is a
natural extension.

No fnctional change.
2015-02-01 13:31:49 +01:00
Marco Costalba 20d6a8e57f Delay checking for duplicated killer moves
Follow the usual approach to delay computation
as far as possible, in case an earlier killer
cut-offs we avoid to do useless work.

This also greatly simplifies the code.

No functional change.
2015-02-01 13:17:42 +01:00
Marco Costalba 0dc6f16992 Small tweaks in movepick.cpp
No functional change.
2015-02-01 12:01:35 +01:00
Marco Costalba 21120288a3 Silence a warning under MSVC
warning C4100: 'ci' : unreferenced formal parameter

It is a silly and wrong one, but just silent it.

No functional change.
2015-02-01 10:40:22 +01:00
Marco Costalba 519b2fe849 More readable score<CAPTURES>()
No functional change.
2015-01-31 20:13:38 +01:00
Marco Costalba f3189bdc9a Use C++11 loops in MovePicker
No functional change.
2015-01-31 20:02:08 +01:00
Marco Costalba 65f46794af Implicit conversion from ExtMove to Move
Verified with perft there is no speed regression,
and code is simpler. It is also conceptually correct
becuase an extended move is just a move that happens
to have also a score.

No functional change.
2015-01-31 19:22:07 +01:00
Marco Costalba 81d6c4a0d6 Another small tweak to skills
No functional change.
2015-01-31 18:24:39 +01:00
Marco Costalba 60c121f3b1 Sync with master
bench: 7374604
2015-01-31 13:05:51 +01:00
Marco Costalba 45eac9507c Use C++ loops in insert_pv_in_tt
Also small tweak to extract_ponder_from_tt

No functional change.
2015-01-31 12:57:52 +01:00
Marco Costalba bfd0f95f06 Move uci_pv under UCI namespace
That's the correct place.

No functional change.
2015-01-31 12:18:24 +01:00
Marco Costalba 1b947aafbf Convert Reductions[] from int8_t to Depth
This is the type anyhow. Assorted cleanup while there.

No functional change.
2015-01-31 11:44:35 +01:00
Marco Costalba a7592e69d7 Fix a MSVC warning
warning C4805: '|' : unsafe mix of type 'Bitboard' and type 'bool' in operation

No functional change.
2015-01-31 10:19:00 +01:00
Jean-Francois Romang a3b4e9e23c Ressurrect hashfull patch
This is an old patch from Jean-Francois Romang to send
UCI hashfull info to the GUI:
https://github.com/mcostalba/Stockfish/pull/60/files

It was wrongly judged as a slowdown, but it takes much
less than 1 ms to run, indeed on my core i5 2.6Ghz it
takes about 2 microsecs to run!

Regression test is good:

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7352 W: 1548 L: 1401 D: 4403

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 61432 W: 10307 L: 10251 D: 40874

I have set the name of the author to the original
one.

No functional change.
2015-01-30 18:07:20 +01:00
Marco Costalba f6f1d24223 Sync with master
bench: 7374604
2015-01-30 17:58:18 +01:00
Marco Costalba ce0a95c2c0 Simplify skill level and reduce ELO
This patch has two positive effects:

- Retire a hackish formula and leave
  just a natural, simple and plain one.

- Reduce strenght at very low level, but
  don't impact medium/high levels.

Actually even at level 0, SF is still too
strong for many beginners (this was reported
many times for instance on Droidfish user
comments on Google Play).

Test on fishtest shows that ELO drop is around
170 ELO at level 0 (good!), 130 ELO at level 1
and smoothly reduces (as expected) until level
10 where the drop is just of 8 ELO.

No functional change.
2015-01-29 22:35:24 +01:00
Alain SAVARD 8aa8608c2a Simplify backward pawn definition
Make use of 'lever' attribute

No functional change

Resolves #234
2015-01-28 21:29:04 +00:00
NicklasPersson 7837fb2aca King safety tuning with values obtained by SPSA.
Part I:

LTC:

LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 11529 W: 2075 L: 1882 D: 7572

Part II:

LTC:

ELO: 2.07 +-2.1 (95%) LOS: 97.3%
Total: 34859 W: 5967 L: 5759 D: 23133

Bench: 7374604

Resolves #228
2015-01-28 21:05:21 +00:00
Joona Kiiski 9f0d5241bf Restore development version
No functional change
2015-01-28 20:53:50 +00:00
Joona Kiiski 5b555525d2 Stockfish 6
Stockfish bench signature is: 8918745
2015-01-27 20:27:38 +00:00
Joona Kiiski c4518e395e Stockfish 6 Release Candidate 3
- Fix a skill level problem: Don't allow move pruning at root node
- Revert "Fix profile build for gcc on Mac OSX". Results for a faster binary in x86-64.
- Fix a MSVC warning

Bench: 8918745
2015-01-25 22:03:57 +00:00
Stefan Geschwentner d8b3ad2208 Fix a skill level problem: Don't allow move pruning at root node
Bench: 8918745

Resolves #231
2015-01-25 21:57:13 +00:00
Joona Kiiski ec36b8dea9 Revert "Fix profile build for gcc on Mac OSX"
Seems to be a performance regression for standard build.

For SF6 people compiling on Mac OSX using profile-build option
just need to make necessary adjustments manually...

No functional change

Resolves #223
2015-01-25 21:51:09 +00:00
Marco Costalba 14f7d9a629 Re-arrange Skill struct
Instead of swapping sub-optimal move in Skill
d'tor, make it explicitly at the end of the search.

Also streamline and clarify relation with multiPV
and pass it directly instead of relying on the hacky
'candidates' member.

No functional change.
2015-01-25 11:16:54 +01:00
Marco Costalba 8ebf30d44a Fix a MSVC warning at W4
Warning is C4512 (assignment operator could not be generated)

Now, apart the foreign syzygy code, everything compiles
without warnings at warning level 4.

Backported from C++11 branch.

No functional change.
2015-01-25 09:36:05 +01:00
Joona Kiiski 5154ac9cff Stockfish 6 Release Candidate 2
- Fix a compilation issue related to BMI2 PEXT instruction
- Retrieve a ponder move from TT if PV is only one move long

Bench: 8080602

No functional change
2015-01-24 19:42:49 +00:00
Marco Costalba 54b5b528d9 Don't use _pext_u64() directly
This intrinsic to call BMI2 PEXT instruction is
defined in immintrin.h. This header should be
included only when USE_PEXT is defined, otherwise
we define _pext_u64 as 0 forcing a nop.

But under some mingw platforms, even if we don't
include the header, immintrin.h gets included
anyhow through an include chain that starts with
STL <algorithm> header. So we end up both defining
_pext_u64 function and at the same time defining
_pext_u64 as 0 leading to a compile error.

The correct solution is of not using _pext_u64 directly.

This patch fixes a compile error with some mingw64
package when compiling with x86-64.

No functional change.

Resolves #222
2015-01-24 19:38:06 +00:00
Marco Costalba 44643c2770 Try hard to retrieve a ponder move
In case we stop the search during a fail-high
it is possible we return to GUI without a ponder
move. This patch try harder to find a ponder move
retrieving it from TT. This is important in games
played with 'ponder on'.

bench: 8080602

Resolves #221
2015-01-24 19:35:31 +00:00
Marco Costalba 69407ae9b8 Additional work in bitbases
Verified the generated bitbases are unchanged.

No functional change.
2015-01-24 09:29:35 +01:00
Marco Costalba 18375e3bfb Rearrange bitbases C++11 way
No functional change.
2015-01-22 11:05:31 +01:00
Marco Costalba 6390a3da94 Document how to enable PEXT with MSVC
When not using Makefile, e.g. with MSVC, if hardware
supports BMI2 instructions, then USE_PEXT should be
added in project configuration to enable pext support.

No functional change.
2015-01-21 19:54:15 +01:00
Marco Costalba 36f8814aa6 Rearrange Endgames
Remove references to EndgameBase and use instead
Value and ScaleFactor as template parameters of
the endgames maps.

No functional change.
2015-01-21 17:35:53 +01:00
Marco Costalba 96e36a7897 Explicitly defaulted and deleted members
Better than a bit obscure implicit ones.

No functional change.
2015-01-21 13:18:19 +01:00
Marco Costalba 2ca2c3f35b Fun with lambdas
Use lambda functions instead of has_positive_value()
and toggle_case()

No functional change.
2015-01-21 11:33:53 +01:00
Marco Costalba f54c44e6be Don't use _pext_u64() directly
This intrinsic to call BMI2 PEXT instruction is
defined in immintrin.h. This header should be
included only when USE_PEXT is defined, otherwise
we define _pext_u64 as 0 forcing a nop.

But under some mingw platforms, even if we don't
include the header, immintrin.h gets included
anyhow through an include chain that starts with
STL <algorithm> header. So we end up both defining
_pext_u64 function and at the same time defining
_pext_u64 as 0 leading to a compile error.

The correct solution is of not using _pext_u64 directly.

This patch fixes a compile error with some mingw64
package when compiling with x86-64.

No functional change.
2015-01-20 22:17:22 +01:00
Marco Costalba ca3622f8e8 Fix a MSVC warning at W4
Warning is C4512 (assignment operator could not be generated)

Now, apart the foreign syzygy code, everything compiles
without warnings at warning level 4.

No functional change.
2015-01-19 08:11:43 +01:00
Joona Kiiski 97a034ad3e Stockfish 6 Release Candidate 1
Bench: 8080602

No functional change

Resolves #218
2015-01-18 14:53:11 +00:00
Joona Kiiski 1b62b47b62 Fix profile build for gcc on Mac OSX
Switch back to using -fprofile-generate and
-fprofile-use flags

No functional change

Resolves #219
Resolves #210
2015-01-18 14:42:30 +00:00
Joona Kiiski 7f51610103 Don't print fail-high or fail-lows in MultiPV mode
Supposed to give a better user experience when using MultiPV mode

No functional change

Resolves #217
2015-01-18 14:38:46 +00:00
Marco Costalba 712bbb052e Fix compile error with MSVC 2013
Without this std::min chokes with windows 'min' macro.

No functional change.
2015-01-18 15:14:05 +01:00
Marco Costalba fe4fda474f Fun with initializer lists in UCI::square
No functional change.
2015-01-18 11:04:51 +01:00
Marco Costalba f7d8ea3866 Fix a coverity scan warning
Coverity scan warns about uninitialized 'sf' argument when
calling probe(). Actually it is a false positive because
argument is passed by reference and assigned inside
probe(). Nevertheless it is a hint that fucntion signature
is a bit tricky, so rewrite it in a more conventional way,
assigning 'sf' from probe() return value.

No functional change.
2015-01-18 10:41:56 +01:00
Marco Costalba f53aea45e3 Add syzygy support
bench: 8080602
2015-01-18 08:27:46 +01:00
Marco Costalba 3c07603dac Import C++11 branch
Import C++11 branch from:

https://github.com/mcostalba/Stockfish/tree/c++11

The version imported is teh last one as of today:
https://github.com/mcostalba/Stockfish/commit/6670e93e50a7a4e739ac2ac8b6026ffa3a12150a

Branch is fully equivalent with master but syzygy
tablebases that are missing (but will be added with
next commit).

bench: 8080602
2015-01-18 08:00:50 +01:00
Marco Costalba c73f33f37e Fix syzygy warnings with Intel compiler
Quick hack, a better job would require to
first move all syzygy code up to SF standard.

No functional change.
2015-01-18 07:29:51 +01:00
Marco Costalba 05cb58f4fc Fix some missing rename from previous patch
No functional change.
2015-01-17 22:15:15 +01:00
Marco Costalba 595fc342cf Fix a possible overflow in TT resize
On platforms where size_t is 32 bit, we
can have an overflow in this expression:

(mbSize * 1024 * 1024)

Fix it setting max hash size of 2GB on platforms
where size_t is 32 bit.

A small rename while there: now struct Cluster
is definied inside class TranspositionTable so
we should drop the redundant TT prefix.

No functional change.
2015-01-17 21:42:47 +01:00
mstembera 58fdb84b0d Simplify and optimize value extractors
Speed up results by Joona:

gcc-4.7 (1.5%)
gcc-4.8 (0.5%)
gcc-4.9 (1.0%)

Speed up results by mstembera:

gcc 474
p-value: 0.719

gcc 482
p-value: 1

gcc 492
p-value: 0.859

No functional change

Resolves #211
2015-01-16 19:03:49 +00:00
lucasart ab276357d7 Enable futility pruning for PV nodes in qsearch
STC:

LLR: 4.20 (-2.94,2.94) [-3.00,1.00]
Total: 85573 W: 17195 L: 17125 D: 51253

LTC:

LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 43385 W: 7298 L: 7214 D: 28873

Bench: 8080602

Resolves #206
2015-01-14 20:21:54 +00:00
Stefan Geschwentner 4abe333e1f Enable Futility pruning in PV nodes
STC:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21553 W: 4342 L: 4221 D: 12990

LTC:

LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7675 W: 1351 L: 1209 D: 5115

Bench: 8668014

Resolves #205
2015-01-14 20:13:48 +00:00
hxim 7b20bb6e1a Use VALUE_ZERO instead of 0 when comparing with Value
This is the correct practice.

No functional change

Resolves #208
2015-01-13 21:07:18 +00:00
Marco Costalba 4eb2d8ce09 Assorted headers cleanup
Mostly comments fixing and other small things.

No functional change.
2015-01-11 22:56:35 +01:00
Marco Costalba b97df4c236 Fix a crash of syzygy on Android
On Android-ARM current TB code crashes at
random times even in single thread mode.

Reported, debugged, fixed and verified
by Peter Osterlund.

No functional change.

Resolves #201
2015-01-11 20:10:31 +00:00
Stefano80 17bcdb9212 Higher movetime resolution for benchmarks
Use a millisecond movetime precision for benchmarks

No functional change

Resolves #198
2015-01-11 20:07:11 +00:00
Marco Costalba 202a78e8e4 Fix compile for Android 5
Android 5 can only run position independent executables.

Note that this breaks Android 4.0 and earlier.

See here for more info:
http://stackoverflow.com/questions/24818902/running-a-native-library-on-android-l-error-only-position-independent-executab

Thanks to Peter Osterlund for the support.

No functional change
2015-01-11 10:52:42 +01:00
Marco Costalba 42b48b08e8 Update copyright year
No functional change.
2015-01-10 11:46:28 +01:00
Marco Costalba aea2fde611 Assorted formatting and comment tweaks in position.h
No functional change.
2015-01-07 09:09:41 +01:00
lucasart b73ae56ee1 Tune null search reduction
STC

LLR: 2.96 (-2.94,2.94) [-0.50,4.50]
Total: 107289 W: 21851 L: 21325 D: 64113

LTC

LLR: 2.96 (-2.94,2.94) [0.00,5.00]
Total: 83837 W: 14378 L: 13916 D: 55543

Bench: 7604776

Resolves #194
2015-01-05 22:09:12 +00:00
Marco Costalba 3fda064a66 Retire one implementation of pop_lsb()
We have two implementations that are equivalent,
so retire one.

Plus usual tidy up of comments and code reshuffle.

No functional change.
2015-01-03 22:16:30 +01:00
lucasart a6e292034a Remove a useless optimization
This optimization is aimed at old hardware only (withouth popcount), and even on
non popcount compile (ARCH=x86-64), it provides no mesurable speedup:

stat        test     master     diff
mean   2,341,779  2,354,699  -12,920
stdev     12,910     14,770   18,150

speedup      -0.55%
P(speedup>0)  23.8%

No functional change.

Resolves #187
2015-01-03 20:40:57 +00:00
Joona Kiiski c7332d5610 Smoother king safety
STC:

LLR: 4.03 (-2.94,2.94) [-1.50,4.50]
Total: 35707 W: 7352 L: 7106 D: 21249

LTC:

LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 24563 W: 4330 L: 4095 D: 16138

Bench: 8411409

Resolves #190
2015-01-03 20:34:20 +00:00
Marco Costalba 9d1e2c0e76 Assorted work in uci.cpp
- Change UCI::value() signature

This function should only return the value,
lowerbound and upperbound info is up to the
caller because it requires external knowledge,
out of the scope of this little helper.

- Retire 'key' command

It is not an UCI command and is absolutely
useless: never used.

- Comments fixing and other trivia

No functional change.
2015-01-03 18:46:59 +01:00
Marco Costalba c150f07291 Fix a silly warning with Intel compiler
Intel compiler cries for an explicit cast

warning #2259: non-pointer conversion from "double" to "int"
may lose significant bits

No functional change.
2015-01-03 10:17:04 +01:00
Marco Costalba 62f531254e Fix comments in thread.cpp
And reshuffle a bit the functions to place
them in a consistent order.

To be on the safe side, patch has been
validated for no regression/crashes with
a small 8K games test with 3 threads:

ELO: 3.98 +-4.4 (95%) LOS: 96.3%
Total: 8388 W: 1500 L: 1404 D: 5484

No functional change.
2015-01-03 09:34:58 +01:00
Marco Costalba 91cc82aa25 Let material probing to access per-thread table
It is up to material (and pawn) table look up
code to know where the per-thread tables are,
so change API to reflect this.

Also some comment fixing while there

No functional change.
2015-01-02 21:31:02 +01:00
Matthew Lai 19b8249ff4 Disable the default copy constructor for Position class
No functional change

Resolved #183
2015-01-02 20:19:11 +00:00
Stefan Geschwentner 25b492ab58 Big King Safety tuning
All king safety related terms (shelterweakness, stormdanger,
attackunits, ..) was tuned together. Additionally for attack units a
finer granularity (factor 4) is used.

STC
ELO: 9.22 +-3.1 (95%) LOS: 100.0%
Total: 19514 W: 4340 L: 3822 D: 11352

LTC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6399 W: 1192 L: 1056 D: 4151

Bench: 8224782

Resolves #184
2015-01-03 03:33:02 +08:00
Marco Costalba 4c9b423161 Retire Material::space_weight()
Move all in evaluation.

Simplify the code and concentrate in a single place
all the logic behind space evaluation, making it much
more clear.

Verified also at STC it does not regress due to a possible
slow down:

LLR: 3.91 (-2.94,2.94) [-3.00,1.00]
Total: 65744 W: 13285 L: 13194 D: 39265

No functional change.
2015-01-02 10:02:28 +01:00
Marco Costalba 2416242c96 Additional tidy up in timeman.cpp
Fixed some comments and moved/renamed some
variables.

No functional change.
2014-12-30 09:49:54 +01:00
Marco Costalba 6933f05f4b Use score and value consistently
And other assorted small fixing, code style
tweaks and reshuffles in evaluate.cpp

No functional change.
2014-12-28 19:06:56 +01:00
Arjun Temurnikar f9571e8d57 Remove some comments in timeman.cpp
Commenst are obsolete now, an updated description
would be quite obscure, so better let the code
to talk and remove them all together.

No functional change.
2014-12-28 11:58:29 +01:00
Marco Costalba 2bfacf136c Reformat promotion generation
Use the same template of other pawns moves generation,
make the code more uniform, simplify generate_promotions
that has now been renamed.

No functional change (verified also with perft).
2014-12-25 11:09:07 +01:00
Marco Costalba 296534f234 Prefer names to numbers in storm code
Also replaces some tabs with spaces and
change StormDanger order to reflect
ShelterWeakness one.

No functional change.
2014-12-22 08:33:07 +01:00
Joona Kiiski e5c7b44f7a Use "rm -f" instead of "rm" for gcc profiling hack in Makefile
In some UNIX systems "rm" prompts user for confirmation.
However "rm -f" is always a guaranteed forced deletion.

Also move gcc profiling hack under the correct target

No Functional change

Resolves #168
2014-12-20 17:55:18 +00:00
Stefan Geschwentner 3231038262 Big King Safety tuning
ShelterWeakness and Stormdanger array are now indexed additionally by
file pair (a/h,b/g,c/f,d/e). The special case of king blocking a pawn
is incorporated in the StormDanger array.  Finally the 93 parameters
are tuned by SPSA on LTC.

STC
ELO: 3.46 +-2.2 (95%) LOS: 99.9%
Total: 40000 W: 8275 L: 7877 D: 23848

LTC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 10311 W: 1876 L: 1721 D: 6714

Bench: 9498821

Resolves #163
2014-12-21 01:53:44 +08:00
Marco Costalba 9cae6e66ce Don't account for Tempo in specialized endgames
The evaluation is already done by the specialized
function, don't need to add something elese later.

With this patch following positions are evaluated
correctly as draws:

8/6p1/1Pkp1p1p/2nNn2P/2P1K1P1/8/8/3B4 w - - 7
8/1k4p1/1P1p1p1p/3NnK1P/2P3P1/1n6/4B3/8 w - -

Verified it not regress with an STC test:
LLR: 3.15 (-2.94,2.94) [-3.00,1.00]
Total: 49812 W: 10095 L: 10016 D: 29701

Reported by Arjun Temurnikar.

bench: 8289983
2014-12-19 11:06:40 +01:00
mstembera 46d5fff01f Change profile-build options to produce 1% to 2% faster executables.
The "@rm ucioption.gc*" line is necessary to avoid a gcc 4.7.x bug.
Confirmed for gcc 4.7.4, 4.8.1, and 4.9.1
Suggested by Kiran Panditrao on fishcooking forum.
https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/AY8gN53nG18

No functional change.

Resolves #160
2014-12-19 03:57:04 +08:00
Marco Costalba b8fd1a78dc Improve comments in UCI
And simplify naming while there.

No functional change.

Resolves #159
2014-12-14 23:50:33 +00:00
Marco Costalba 413b243809 Coding style in TT code
In particular seems more natural to return
bool and TTEntry on the same line, actually
we should pass and return them as a pair,
but due to limitations of C++ and not wanting
to use std::pair this can be an acceptable
compromise.

No functional change.

Resolves #157
2014-12-14 23:49:00 +00:00
Gary Linscott 0edb6348d2 Fix compile for some versions of mingw
The bswap intrinsics are specific to the compiler, not the
host platform.

No functional change.

Resolves #155
2014-12-14 14:45:43 -05:00
mstembera 14cf27e6f6 Avoid searching TT twice for the same key/position during probe() and store().
Just keep the pointer and remove code from tt.cpp

STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 13620 W: 2810 L: 2665 D: 8145

LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 13021 W: 2238 L: 2073 D: 8710STC http://tests.stockfishchess.org/tests/view/548436860ebc59331739b90c

STC 4MB
ELO: 2.41 +-2.2 (95%) LOS: 98.6%
Total: 40000 W: 8175 L: 7897 D: 23928

LTC 16MB
ELO: 1.78 +-2.0 (95%) LOS: 96.1%
Total: 39683 W: 6763 L: 6560 D: 26360

Resolves #151

Bench: 8116521
2014-12-13 07:22:37 +00:00
Gary Linscott 7b4828b68c Only use _ReadWriteBarrier on MSVC
It was causing compile errors when cross-compiling using mingw.

No functional change.
2014-12-11 14:56:24 -05:00
joergoster f6d220ab14 Halve StormDanger bonus for blocked pawn on A/H file
STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 3410 W: 758 L: 641 D: 2011

LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 33225 W: 5708 L: 5445 D: 22072

Bench: 8465811

Resolves #153
2014-12-11 13:08:29 -05:00
Gary Linscott afafdf7b73 MSVC compiling fixes
No functional change.

Resolves #150
2014-12-11 13:03:44 -05:00
Joona Kiiski b15dcd9774 Fix profile build for syzygy
Touch source files under syzygy directory to force recompilation
after collecting profile data.

No functional change

Resolves #149
2014-12-10 17:59:41 +00:00
Joona Kiiski 94dd204c3b Retire 'os' flag from Makefile
Appears to be unused

No functional change

Resolves #147
2014-12-10 17:57:55 +00:00
Marco Costalba 5943600a89 Assorted nitpicking code-style
No functional change.
2014-12-10 12:38:13 +01:00
Marco Costalba 589c711449 Clarify when forcing the moves loop
In some cases we want to go direcly to the moves loop
without checking for early return. The patch make this
logic more clear and consistent.

Tested for no regression, passed STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25282 W: 5136 L: 5022 D: 15124

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 72007 W: 12133 L: 12095 D: 47779

bench: 9316798
2014-12-10 12:35:21 +01:00
Ernesto Gatti 158864270a Simpler PRNG and faster magics search
This patch replaces RKISS by a simpler and faster PRNG, xorshift64* proposed
by S. Vigna (2014). It is extremely simple, has a large enough period for
Stockfish's needs (2^64), requires no warming-up (allowing such code to be
removed), and offers slightly better randomness than MT19937.

Paper: http://xorshift.di.unimi.it/
Reference source code (public domain):
http://xorshift.di.unimi.it/xorshift64star.c

The patch also simplifies how init_magics() searches for magics:

- Old logic: seed the PRNG always with the same seed,
  then use optimized bit rotations to tailor the RNG sequence per rank.

- New logic: seed the PRNG with an optimized seed per rank.

This has two advantages:
1. Less code and less computation to perform during magics search (not ROTL).
2. More choices for random sequence tuning. The old logic only let us choose
from 4096 bit rotation pairs. With the new one, we can look for the best seeds
among 2^64 values. Indeed, the set of seeds[][] provided in the patch reduces
the effort needed to find the magics:

64-bit SF:
Old logic -> 5,783,789 rand64() calls needed to find the magics
New logic -> 4,420,086 calls

32-bit SF:
Old logic -> 2,175,518 calls
New logic -> 1,895,955 calls

In the 64-bit case, init_magics() take 25 ms less to complete (Intel Core i5).

Finally, when playing with strength handicap, non-determinism is achieved
by setting the seed of the static RNG only once. Afterwards, there is no need
to skip output values.

The bench only changes because the Zobrist keys are now different (since they
are random numbers straight out of the PRNG).

The RNG seed has been carefully chosen so that the
resulting Zobrist keys are particularly well-behaved:

1. All triplets of XORed keys are unique, implying that it
   would take at least 7 keys to find a 64-bit collision
   (test suggested by ceebo)

2. All pairs of XORed keys are unique modulo 2^32

3. The cardinality of { (key1 ^ key2) >> 48 } is as close
   as possible to the maximum (65536)

Point 2 aims at ensuring a good distribution among the bits
that determine an TT entry's cluster, likewise point 3
among the bits that form the TT entry's key16 inside a
cluster.

Details:

     Bitset   card(key1^key2)
     ------   ---------------
RKISS
     key16     64894   = 99.020% of theoretical maximum
     low18    180117   = 99.293%
     low32    305362   = 99.997%

Xorshift64*, old seed
     key16     64918   = 99.057%
     low18    179994   = 99.225%
     low32    305350   = 99.993%

Xorshift64*, new seed
     key16     65027   = 99.223%
     low18    181118   = 99.845%
     low32    305371   = 100.000%

Bench: 9324905

Resolves #148
2014-12-08 08:18:26 +08:00
Gary Linscott a87da2c4b3 Add some tablebase positions to bench
This makes it easier to check for regressions in the tablebase code.

Bench: 9489202
5-man bench: 8943906

Resolves #145
2014-12-08 07:58:05 +08:00
hxim fbb53524ef Rename some variables for more clarity.
No functional change.

Resolves #131
2014-12-08 07:53:33 +08:00
Marco Costalba ba1464751d Explicitly pass RootMoves to TB probes
Currently Search::RootMoves is accessed and even
modified by TB probing functions in a hidden
and sneaky way.

This is bad practice and makes the code tricky.
Instead explicily pass the vector as function
argument so to clarify that the vector is modified
inside the functions.

No functional change.
2014-12-06 15:08:21 +00:00
Marco Costalba eeb6d923fa Move TB stuff under Tablebases namespace
Simplified also some logic while there.

TBLargest needs renaming too, but itis for
a future patch because touches also syzygy
directory stuff.

No functional change.
2014-12-06 14:58:00 +00:00
Marco Costalba c30eb4c9c9 Refactor syzygy code in search
Move to SF coding style.

Also skip calculating piece count in search()
when TB are not found (!TBCardinality)

No functional change.
2014-12-06 14:35:50 +00:00
Joona Kiiski 35c1ccef39 Retire support for Haiku installation directory from Makefile
- It is out of the scope of the project.
- It is the responsibility of Haiku package maintainer to
  configure this.

No functional change

Resolves #143
2014-12-06 14:23:08 +00:00
Joona Kiiski 0935dca9a6 Retire hackish support for aCC and HP-UX from Makefile
- It is out of scope of the project.
- We have no way to verify that it even works anymore

No functional change

Resolves #142
2014-12-06 14:19:39 +00:00
Marco Costalba 314d446518 Retire total_piece_count()
We really don't need to uglify in this way
our nice count() API with this ad-hoc hack.

So remove the hack and use the already
existing infrastructure.

No functional change.

Resolves #134
2014-11-30 20:37:24 +00:00
hxim c014444f09 Remove CONNECTED_KINGS from Syzygy code
No functional change

Resolves #140
2014-11-30 20:24:32 +00:00
Rodrigo Exterckötter Tjäder 9b4e123fbe Cleaning Syzygy profiling data
Updating the makefile so that the clean and gcc-profile-clean targets also
remove the profiling data files in the syzygy directory.

No functional change.

Resolves #136
2014-11-30 19:53:04 +00:00
Marco Costalba a43f633c19 Rewrite TBScore in uci_pv()
Streamline the code and make
it understandable.

No functional change.

Resolves #135
2014-11-30 19:35:35 +00:00
Marco Costalba 66f5cd3f9d Retire #ifdef SYZYGY macro
It just clutters the code for no
real reason.

No functional change.

Resolves #139
2014-11-30 19:23:17 +00:00
mstembera fe07ae4cb4 Bitbase index() from ADD to OR.
No functional change.

Resolves #132
2014-11-26 07:56:48 +08:00
lucasart 2c52147dbf Introduce ratio operation
Just like in Physics, the ratio of 2 things in the same unit, should be
without unit.

Example use case:
- Ratio of a Depth by a Depth (eg. ONE_PLY) should be an int.
- Ratio of a Value by a Value (eg. PawnValueEg) should be an int.

Remove a bunch of useless const while there.

No functional change.

Resolves #128
2014-11-26 07:55:57 +08:00
Ronald de Man 7caa6cd338 Syzygy tablebases
Adds support for Syzygy tablebases to Stockfish.  See
the Readme for information on using the tablebases.

Tablebase support can be enabled/disabled at the Makefile
level as well, by setting syzygy=yes or syzygy=no.

Big/little endian are both supported.

No functional change (if Tablebases are not used).

Resolves #6
2014-11-26 07:49:58 +08:00
Gary Linscott 4509eb1342 Fix out-of-bound array access printing ponder move
It is possible that we won't have a ponder move if our PV
is too short.  In that case, just don't print a ponder move.

No functional change

Resolves #130
2014-11-24 08:53:00 +08:00
Gary Linscott 7ad59d9ac9 Fix pondering
The UCI specification states that an engine can never exit the search
while pondering.

No functional change.

Resolves #118
2014-11-24 08:50:36 +08:00
Jonathan Calovski 48127fe5d3 Amend defended
Amend defended to remove now redundant condition.

No functional change.

Resolves #125
2014-11-22 05:46:59 +08:00
Marco Costalba 84408e5cd6 Fix doubled pawns asymmetry
When evaluating double pawns we use always
lsb() to extract the frontmost square.

This breaks evaluation color symmetry as is
possible to verify with an instrumented evaluate()

  Value evaluate(const Position& pos) {

    Value v = do_evaluate<false>(pos);
    Position p = pos;
    p.flip();
    assert(v == do_evaluate<false>(p));
    return v;
  }

Passed no regression test:

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21035 W: 4244 L: 4122 D: 12669

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 39839 W: 6662 L: 6572 D: 26605

bench: 8255966
2014-11-21 20:40:25 +01:00
Marco Costalba 79232be02a Further tweak accurate pv
It is a non functional change, but because
we now skip copying of pv[] in SpNode, patch
has been tested for regression with 3 threads:

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 54668 W: 9873 L: 9809 D: 34986

No functional change.
2014-11-21 20:37:45 +01:00
hxim 0a1f54975f Fix some comments
No functional change.

Resolves #123
2014-11-19 06:39:17 +08:00
Gary Linscott bffe32f4fe Fix fen output for castling rights
This is a regression from 428962a

We have to cast to char here, otherwise the compiler
interprets it as an integer, and writes a number.

No functional change

Resolves #122
2014-11-19 06:37:59 +08:00
Marco Costalba 4aca11ae2a Codying style in accurate PV
This is the first of a patch series to
rearrange and simplify accurate PV.

In this patch there is simple coding
style and reformatting stuff.

Verified with fishtest it does not crash
with MAX_PLY = 8

No functional change.
2014-11-18 11:57:57 +01:00
Marco Costalba 1a939cd8c8 Fix a warning on Intel C++
warning #2259: non-pointer conversion from "int" to
"uint8_t={unsigned char}" may lose significant bits

No functional change.
2014-11-17 12:56:48 +01:00
sf-x d65c9a3262 Use PHASE_MIDGAME in game_phase()
No functional change

Resolves #117
2014-11-17 07:50:33 +08:00
Rodrigo Exterckötter Tjäder 99f2c1a2a6 Clear token before reading from input
Previously token would keep its value from the previous line when an empty
line was input, leading to unexpected behaviour.

No functional change

Resolves #119
2014-11-17 07:48:30 +08:00
lucasart 3b1f552b08 Half History Max
STC
LLR: 3.35 (-2.94,2.94) [-0.50,3.50]
Total: 17993 W: 3740 L: 3508 D: 10745

LTC
LLR: 3.25 (-2.94,2.94) [0.00,4.00]
Total: 21346 W: 3691 L: 3453 D: 14202

Bench: 7694316

Resolves #120
2014-11-17 07:04:58 +08:00
Marco Costalba 4840643fed Use DEPTH_MAX instead of MAX_PLY
When comparing to a Depth it is more
consistent to use DEPTH_MAX instead
of a int.

This is a subtle difference because we use
ply and depth almost interchangably in SF,
but they are different. FOr counting plies
makes ense to continue using ints, while
for Depth we have our specific enum.

This cleanly fixes a new Clang 3.5 warning:

No functional change.
2014-11-15 05:36:49 +01:00
Gary Linscott 4739037f96 100% accurate PV display
This gives SF accurate PVs, such that the evaluation of the leaf node in
the PV matches the score backed up to the root (99% of the time.
q-search will use the value stored in the hash table instead of the eval
value sometimes).

One drawback is that fail-high/low only get a minimal 2 move PV.

It doesn't add any additional overhead to the non-PV codepath except an
extra eight bytes to the SearchStack structure in multi-threaded
searches.

A core part of this is not pruning based on TT score in PV nodes. This
was measured as not being a regression at multiple TCs, except for one
exception, fast TC with huge hash, which is not realistic for longer
searches.

STC - 1 thread, 128 mb hash
ELO: 1.42 +-3.1 (95%) LOS: 81.9%
Total: 20000 W: 4078 L: 3996 D: 11926

STC - 3 thread, 128 mb hash
ELO: -3.60 +-2.9 (95%) LOS: 0.8%
Total: 20000 W: 3575 L: 3782 D: 12643

STC - 3 thread, 8 mb hash
ELO: 0.12 +-2.9 (95%) LOS: 53.3%
Total: 20000 W: 3654 L: 3647 D: 12699

LTC - 3 thread, 32mb hash
ELO: 2.29 +-2.0 (95%) LOS: 98.8%
Total: 35740 W: 5618 L: 5382 D: 24740

Bench: 6984058

Resolves #102
2014-11-12 16:16:33 -05:00
lucasart 234344500f Use quiet ttMove in qsearch() (7962287)
Daniel Jose reported that it was an elo gain in his engine:
http://www.talkchess.com/forum/viewtopic.php?t=54290

STC: Hash=16
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33067 W: 6670 L: 6571 D: 19826

LTC: Hash=64
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 41181 W: 7008 L: 6920 D: 27253

And another one to verify no regression with hash pressure:

STC: Hash=4
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 25085 W: 5059 L: 4991 D: 15035

Verified that qsearch does not explode after this patch (recapture threshold).

Bench 7962287

Resolves #112
2014-11-12 21:06:14 +00:00
Marco Costalba b777b17f6f Use if/else instead of goto
Real men jump/branch by hand...but
we prefer the humble way.

Moved also some uci info code where it
belongs, while there.

No functional change.

Resolves #110
2014-11-12 21:02:20 +00:00
Marco Costalba db4b8ee000 Use Depth instead of int in search
And make it more ONE_PLY value independent,
although we are not there yet.

No functional change.

Resolves #111
2014-11-12 21:00:16 +00:00
lucasart c6d45c60b5 Profile Build with Hash=16
16MB for 1" searches is more comensurate with the average use case.

And 16 is the default used by stockfish bench, so it makes sense to be
consistent, if only to have the same minimum memory requirement for using
SF and compiling it with PGO.

No functional change.

Resolves #109
2014-11-10 23:06:12 +00:00
lucasart d709a5f1c5 Fix bounds of FutilityMoveCounts
This is a left-over from ONE_PLY == 2.

No functional change.

Resolves #107
2014-11-09 20:13:56 +00:00
Marco Costalba 1b0df1ae14 Retire pvMove in search()
Now we can directly replace it with
the definition resulting in simpler
and possibly faster code because
PvNode is evaluated at compile time.

No functional change.
2014-11-09 20:36:28 +01:00
Marco Costalba 57fdfdedcf Assorted code-style triviality
No functional change.
2014-11-09 20:17:29 +01:00
Marco Costalba 6fb0a1bc40 Introduce distance() and unify some API
Original work by Lucas.

No functional change.
2014-11-09 10:27:04 +01:00
lucasart 8631b08d97 Codestyle massage Search::init()
* remove some erroneous comments, that were based on the ONE_PLY == 2.
* rename hd to d, because there's no more half-depth in SF.
* rescope variables into the for loops.
* reindent the for loops correctly.
* add a comment to explain the eval improving part (not so obvious to read
this code as array has 4 dimensions).

No functional change.
2014-11-08 10:56:51 -05:00
lucasart 3d2aab11d8 Be more optimistic in aspiration window
Be more optimistic wrt search instability, and set the unviolated bound
half window.

STC
LLR: 2.96 (-2.94,2.94) [-1.00,4.00]
Total: 16362 W: 3371 L: 3197 D: 9794

LTC
LLR: 2.94 (-2.94,2.94) [0.00,5.00]
Total: 21666 W: 3780 L: 3569 D: 14317

Bench: 6807896

Resolves #105
2014-11-08 10:47:56 -05:00
lucasart 7ebb872409 Prune ttMove like any other
This should reduce search inconsistencies, and doesn't seem to have a measurable ELO Impact:

STC with Hash=16
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 49264 W: 10076 L: 10007 D: 29181

LTC with Hash=64
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 82149 W: 14044 L: 14023 D: 54082

Plus an extra test, to make sure it doesn't regress with strong hash pressure:

STC with Hash=4
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 21498 W: 4327 L: 4246 D: 12925

Bench: 7302735

Resolves #100
2014-11-07 21:40:24 +00:00
lucasart 375797d51c Retire CACHE_LINE_ALIGNMENT
Speed tests showed no benefit.

No functional change.

Resolves #97
2014-11-07 14:27:04 -05:00
lucasart 8e98bd616e Apply King Safety later in the endgame
Idea is to apply king safety later in the endgame. Previously, we didn't
apply KS in a RR vs. Q ending for example, which causes poor play.
Now we calculate king attacks when the attacking side has a queen or more.

STC with 8moves_v3
LLR: 3.06 (-2.94,2.94) [0.00,4.00]
Total: 38481 W: 6228 L: 5952 D: 26301

LTC with 2moves_v1
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 51053 W: 8670 L: 8353 D: 34030

Bench: 7514010

Resolves #98
2014-11-06 13:01:47 -05:00
mstembera bc83515c9e Removing some superfluous extern declarations
No functional change.

Resolves #93
2014-11-05 21:17:19 +00:00
Marco Costalba bcbab19376 Assume UCI 'nodes' is int64_t instead of int
UCI specification is not clear on the size of
integers that are exchanged in the protocol, so
instead of a simple int, assume 'nodes' is a
int64_t because we need a bigger size to store
this value in many real cases, especialy with
very long searches.

No functional change.

Resolves #75
2014-11-05 21:11:05 +00:00
Marco Costalba d29a68f585 Rearrange check_time()
Remove an ugly workaround for a gcc
warning while there.

No functional change.
2014-11-05 21:09:21 +00:00
Ajith 0608d6aaec Add bonuses for each threat instead of max threat value.
Use SPSA tuned values for all threat bonuses

STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 6074 W: 1284 L: 1160 D: 3630

LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 13563 W: 2402 L: 2232 D: 8929

Bench: 6981908

Resolves #94
2014-11-04 23:50:54 +08:00
mstembera 2fd075d1ea Optimize TranspositionTable::store() and TranspositionTable::probe() for speed.
No functional change.

Resolves #85
2014-11-03 18:40:49 +00:00
lucasart d12378497c Do not assume that enum are signed
Clang 3.5 issues warning on constructs like: abs(f1 - f2). The thing is that
f1 and f2 are enum types, and the range given (all positive) allows the
compiler to choose an unsigned type (efficiency being one reason to prefer
unsigned arithmetic). If f1 < f2 are unsigned, then f1 - f2 wraps around zero
and the abs() becomes a no-op. It's the reinterpretation of the unsigned
result (large value) as a signed int that happens to give the correct result,
thanks to 2's complement. This is all tricky and dangerous!

In the spirit of the standard, we assume nothing on the signedness of enums,
and simply calculate the rank and file distances as:
- rank_dist(r1, r2) = r1 < r2 ? r2 - r1 : r1 - r2
- file_dist(f1, f2) = f1 < f2 ? f2 - f1 : f1 - f2
this logic can in fact be applied to any enum we may use, so for better
generality and to avoid code duplication, we use a template function diff()
here.

No functional change.

Resolves #95
2014-11-04 00:35:02 +08:00
lucasart 8ab9c2511a Cleanup MAX_PLY
This area has become obscure and tricky over the course of incremental
changes that did not respect the original's consistency and clarity. Now,
it's not clear why we use MAX_PLY = 120, or why we use MAX_PLY+6, among
other things.

This patch does the following:

* ID loop: depth ranges from 1 to MAX_PLY-1, and due to TT constraint (depth
must fit into an int8_t), MAX_PLY should be 128.

* stack[]: plies now range from 0 to MAX_PLY-1, hence stack[MAX_PLY+4],
because of the extra 2+2 padding elements (for ss-2 and ss+2). Document this
better, while we're at it.

* Enforce 0 <= ply < MAX_PLY:
  - stop condition is now ss->ply >= MAX_PLY and not ss->ply > MAX_PLY.
  - assert(ss->ply < MAX_PLY), before using ss+1 and ss+2.
  - as a result, we don't need the artificial MAX_PLY+6 range. Instead we
  can use MAX_PLY+4 and it's clear why (for ss-2 and ss+2).

* fix: extract_pv_from_tt() and insert_pv_in_tt() had no reason to use
MAX_PLY_PLUS_6, because the array is indexed by plies, so the range of
available plies applies (0..MAX_PLY before, and now 0..MAX_PLY-1).

Tested with debug compile, using MAX_PLY=16, and running bench at depth 17,
using 1 and 7 threads. No assert() fired. Feel free to submit to more severe
crash-tests, if you can think of any.

No functional change.
2014-11-03 23:36:24 +08:00
Marco Costalba fc0733087a Restore std::dec after std::hex
Code is leaking a std::hex, and causes subsequent
sync_cout output to be in hexadecimal.

Spotted by Lucas

No functional change.
2014-11-02 08:03:52 +01:00
Marco Costalba 42a20920e5 Retire ScalePawnSpan[]
Obscure, undocmented and misnamed array. Replace with
the direct formula: it is much more clear what the
code does.

No functional change.

Resolves #90
2014-11-01 22:10:25 +00:00
Marco Costalba 79fa72f392 Merge pull request #89 from official-stockfish/pull_no_pretty
Prefer operator<<() to pretty()

No functional change.
2014-11-01 22:24:33 +01:00
Marco Costalba d3091971b7 Retire PawnsFileSpan
It is useless. Tested as no regression:

STC
LLR: 4.06 (-2.94,2.94) [-3.00,1.00]
Total: 140718 W: 28527 L: 28568 D: 83623

LTC
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 60034 W: 10359 L: 10303 D: 39372

bench: 6564212

Resolves #88
2014-11-01 20:50:52 +00:00
lucasart d9caede324 Correctly describe POPCNT compile
SSE4.2 has nothing to do with POPCNT. We must dispell this myth, because
Stockfish is a reference and many will copy this mistake if they see it in Stockfish:
* SSE is an SIMD instruction set, relative to vectorization (using special 128-bit registers).
* POPCNT/LZCNT work on normal registers (eg. AL, AX, EAX, RAX).

The confusion comes from the fact that, in the Intel product line, it just
so happens that SSE4.2 and POPCNT/LZCNT came out at the same time. But this
is not true for AMD. For example, all AMD Pheniom II have SSE3 but no
POPCNT/LZCNT, and that is why the modern compile uses  -msse3 -popcnt and not -msse4.2.

No functional change.

Resolves #86
2014-11-01 20:43:57 +00:00
lucasart 8a7876d48d Consistent use of anonymous namespace
Objects that are only accessible at file-scope should be put in the anonymous namespace.
This is what the  C++ standard recommends, rather than using static, which is really C-style and results in static linkage.

Stockfish already does this throughout the code. So let's weed out the few exceptions,
because... they have no reason to be exceptional.

No functional change.

Resolves #84
2014-11-01 20:35:10 +00:00
lucasart 2ee1250294 Remove dead code
No functional change.

Closes #87
2014-11-01 20:16:29 +00:00
Marco Costalba 5644e14d0e Prefer operator<<() to pretty()
It is more idiomatic, we didn't used it
in the past because Position::pretty(Move)
had a calling argument, but now we can.

As an added benefit, we avoid a lot of string
copies in the process because now we avoid
std::ostringstream ss.

No functional change.
2014-11-01 19:02:35 +01:00
Marco Costalba d07a875398 Merge pull request #82 from official-stockfish/clean_up_bishop_code
Code style clean-up

No functional change.
2014-11-01 18:05:03 +01:00
Marco Costalba 460892382a Code style clean-up
This piece of code needs some love.

No functional change.
2014-10-30 12:32:43 +01:00
mstembera 5605cc7684 max_piece_type cleanup, and slight speed increase.
No functional change.

Resolves #81
2014-10-28 22:23:01 +08:00
Joona Kiiski 7d42752158 A small clean up of previous patch suggested by Marco
No functional change
2014-10-27 20:26:12 +00:00
Joona Kiiski fefb27bab4 Speed up max_piece_type()
Write code in the way that allows compiler to perform loop unrolling.

My measurement (32 cycles each):

Orig:

Time (Mean: 2466.59375, Trimmed mean: 2464.25, Std: 12.6869487803348)
Nodes (Mean: 4294458, Trimmed mean: 4294458, Std: 0)
Speed (Mean: 1741.09247987678, Trimmed mean: 1742.72879715475, Std: 8.93612608292678)

Time (Mean: 2470.15625, Trimmed mean: 2468.75, Std: 12.7484581610433)
Nodes (Mean: 4294458, Trimmed mean: 4294458, Std: 0)
Speed (Mean: 1738.58176151341, Trimmed mean: 1739.54618465403, Std: 8.95585822316946)

Mod:

Time (Mean: 2449.90625, Trimmed mean: 2445.9375, Std: 12.1000116635508)
Nodes (Mean: 4294458, Trimmed mean: 4294458, Std: 0)
Speed (Mean: 1752.94829372932, Trimmed mean: 1755.75934908231, Std: 8.61478453124504)

Time (Mean: 2442.78125, Trimmed mean: 2441.1875, Std: 8.17839157228837)
Nodes (Mean: 4294458, Trimmed mean: 4294458, Std: 0)
Speed (Mean: 1758.03872783803, Trimmed mean: 1759.16825356261, Std: 5.81131316346191)

No functional change
2014-10-27 20:25:57 +00:00
snicolet 7de40076ac Tune PawnsFileSpan
Passed the following SPRT tests:

STC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 24428 W: 5056 L: 4880 D: 14492

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 26590 W: 4715 L: 4472 D: 17403

Bench: 6615949

Resolves #78
2014-10-27 11:48:44 +00:00
Pascal Romaret f1359845de Improve compatibility with Shredder Classic GUI
This commit fixes two issues:

1) Don't print PVs after the search has been interrupted

    This solves the "mate 0 upperbound" scores that sometimes
    creep up when a multi-PV analysis gets interrupted with
    the `stop` command.

2) Print multipv before score

    Shredder Classic fails to identify the main PV
    (the one with multipv 1) if `score` comes first.
    This leads to an eval graph that doesn't reflect
    the scores actually reported by Stockfish when
    doing a multiPV analysis.

No functional change

Closes #76
2014-10-27 11:07:35 +00:00
Marco Costalba 5ab55827b8 Fix an obscure gcc warning
warning: narrowing conversion from ‘int’ to ‘char’ inside { }
is ill-formed in C++11 [-Wnarrowing]

When pedantic meets esoteric!

No functional change.
2014-10-26 19:40:20 +00:00
Marco Costalba 9ba391c5cb Retire notation.cpp
Now we can finally retire notation.cpp
and move UCI helpers under uci.cpp

No functional change.
2014-10-26 19:40:13 +00:00
Marco Costalba 2469daebb1 Final UCI helpers renaming
To reflect new changes, specifically that
now are all under UCI namespace.

No functional change.
2014-10-26 19:40:04 +00:00
Marco Costalba aa60c80ade Retire notation.h
And move the few remaining content
under UCI namespace where they belong.

No functional change.
2014-10-26 19:39:56 +00:00
Marco Costalba 5cbcff55cc Rename ucioption.h to uci.h
We are going to add all UCI related
functions here, so first rename it
to a more proper name.

No functional change.
2014-10-26 19:39:46 +00:00
Marco Costalba 428962a2e8 Retire to_char() helpers
Remove some useless wrappers and make
the conversion explicit and starightforward.

No functional change.
2014-10-26 19:39:37 +00:00
Marco Costalba 4f6b1bf3be Reformat max_threat()
Helper function should just know how to find the
biggest piece type in a bitboard. All the threat
logic and data shoud be in evaluate_threats().

This nicely separates the scope of the two functions
in a more consistent way and simplifies the code.

No functional change.
2014-10-26 09:12:56 +00:00
snicolet 8d62ece945 Calculate maximum threat for hanging pieces
Use the max_threat() helper function to estimate more precisely the
best hanging piece threat.  Also retunes the Threat array using SPSA.

STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7598 W: 1596 L: 1468 D: 4534

LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 7896 W: 1495 L: 1350 D: 5051

Bench: 6816504

Resolves #73
2014-10-24 01:10:11 +08:00
Marco Costalba 480682b670 Document why initing eval tables
Instead of hard-code the weights in a big table,
we prefer to calculate them out of few parameters
at startup. This allows to keep low the number of
independent parameters and hence is good for tuning
and for a better insight in the meaning of the numbers.

No functional change.
2014-10-15 19:36:22 +01:00
Marco Costalba 89d9db2979 Rename Tracing methods
Easier to understand and more in line with
standard Trace classes naming like:

http://msdn.microsoft.com/en-us/library/system.diagnostics.trace.aspx

No functional change.
2014-10-12 20:35:19 +01:00
Marco Costalba 907f912463 Account for Tempo in do_evaluate()
This is more correct because we let evaluate()
to be a pure dispatcher and also now evaluate
and tracing outputs are consistent.

No functional change.
2014-10-12 20:35:07 +01:00
Marco Costalba c97b702f4d Fix some warnings with Intel C++ compiler
No functional change.
2014-10-12 20:34:54 +01:00
lucasart 91de6b0f37 Further streamline connected pawn evaluation
Make even more clear what are the terms that
contribute to evaluate connected pawns, and
completely separate them from the weights
that are now fully looked up in a table.

For future tuning makes sense to init the table with
a formula instead of hard-code it. This allows to
reduce problem space cardinality and makes tuning
easier.

And fix a MSVC warning while there:
warning C4804: '>>' : unsafe use of type 'bool' in operation

No functional change.
2014-10-12 20:03:49 +01:00
lucasart 069073ea68 Merge Connected and Candidate
These two notions are very correlated. Since connected has the most
generality, it makes sense to generalize it to encompass what is
covered by candidate.

STC:
LLR: 4.03 (-2.94,2.94) [-3.00,1.00]
Total: 11970 W: 2577 L: 2379 D: 7014

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 13194 W: 2389 L: 2255 D: 8550

bench 7328585
2014-10-07 07:42:17 +01:00
joergoster 15e2191111 Remove the now redundant TT prefetch call from Position::do_move.
Tested for no regression and passed both
STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 48250 W: 9757 L: 9686 D: 28807

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 51412 W: 8887 L: 8816 D: 33709

No functional change.

Resolves #66
2014-10-06 23:59:56 +08:00
Luca Brivio 8a9b9ec96a Simplify futility move count array formula
No functional change
2014-10-05 12:49:18 +01:00
Marco Costalba 74829342ef Reformat and rename hash_after_move()
Align to standard coding style and properly use
enum types. Rename while there.

No functional change.
2014-10-04 10:36:29 +01:00
Ajith 4b926f227d Extend King Threats to all pieces (other than pawns).
STC
LLR: 2.99 (-2.94,2.94) [-1.50,4.50]
Total: 20559 W: 4261 L: 4095 D: 12203

LTC
LLR: 2.96 (-2.94,2.94) [0.00,4.00]
Total: 75232 W: 13097 L: 12696 D: 49439

Bench: 7543790

Resolves #63
2014-10-04 03:54:12 +08:00
joergoster 82d065b011 Speculative prefetch
Idea by Peter Oesterlund.
Implemented and tested by Joerg Oester

STC 3 threads
ELO: 3.19 +-2.1 (95%) LOS: 99.9%
Total: 40000 W: 7576 L: 7209 D: 25215

LTC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 22026 W: 3829 L: 3619 D: 14578

STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 7291 W: 1531 L: 1404 D: 4356

No functional change

Resolves #61
2014-10-02 22:19:14 +01:00
lucasart e60cdca9b0 Convert TT depth to int8_t
Now that half plies have been removed from the engine, we can encode
TT depth into an int8_t.

Range is -128 to +127, so it goes still further than the previous
limit of 121 plies (with ONE_PLY == 2 where depth - DEPTH_NONE was
encoded as an uint8_t).

No functional change.

Resolved #60
2014-10-01 20:51:32 +01:00
Marco Costalba a1b62d68ec Trivial code style fixes
Mainly to sync mine and official repo.

No functional change.
2014-09-30 09:05:20 +02:00
Marco Costalba 222f59b9c1 Move ONE_PLY to be 1 instead of 2: search()
Now that half-plies are no more used we can simplify
the code assuming that ONE_PLY is 1 and no more 2.

Verified with a SMP test:
LLR: 2.95 (-2.94,2.94) [-4.50,0.00]
Total: 8926 W: 1712 L: 1607 D: 5607

No functional change.
2014-09-29 15:17:12 +02:00
lucasart 27a1877299 Clean up VALUE_KNOWN_WIN conditions
A patch (+ some extra changes) passed with:

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 14575 W: 3101 L: 2967 D: 8507

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 42579 W: 7580 L: 7496 D: 27503

Bench: 6545733

Resolves #52
2014-09-28 17:45:49 +01:00
Uri Blass ea9c424bba Remove use of half-ply reductions from LMR, Null-move, IID and
Singular extensions.

STC:
ELO: 3.80 +-3.1 (95%) LOS: 99.2%
Total: 19727 W: 4190 L: 3974 D: 11563

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 7647 W: 1356 L: 1214 D: 5077

Bench: 6545733

Resolves #55
2014-09-28 04:33:28 +08:00
Joona Kiiski 7ed15af371 Cap evaluation based null move extra reduction to three plies
It's a zero elo patch, and a reasonable safeguard against uncontrolled extreme reductions.

STC:
ELO: -0.08 +-2.0 (95%) LOS: 46.9%
Total: 40000 W: 6728 L: 6737 D: 26535

LTC:
ELO: -0.14 +-1.8 (95%) LOS: 44.0%
Total: 40000 W: 5557 L: 5573 D: 28870

Bench: 7201830
2014-09-25 20:42:25 +01:00
Uri Blass d6613b7589 Change history reduction in LMR to be a full ply.
STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 9829 W: 2142 L: 1998 D: 5689

LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 27162 W: 4802 L: 4692 D: 17668

Bench: 7284120

Resolves #53
2014-09-26 02:22:39 +08:00
uriblass fed3e752ae Evaluate king safety when no queen is present.
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 16657 W: 3547 L: 3391 D: 9719

LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 31258 W: 5664 L: 5403 D: 20191

Bench: 8331165

Resolves #51
2014-09-22 01:32:47 +08:00
Gary Linscott 766fb9c67d Fix spacing.
No functional change.
2014-09-22 01:27:34 +08:00
Marco Costalba b652720903 Rearrange evaluation constants definitions
Make them more uniform and consistent.

No functional change.

Conflicts:
	src/evaluate.cpp
2014-09-21 08:26:25 +08:00
Marco Costalba 4d0a6c5a6f Rename time variable to reflect UCI parameters
On top of previous patch, rename time variables to
reflect the simplification of UCI parameters.

It is more correct to use as varibales directly the
corresponding UCI option, without intorducing redundant
intermediate variables.

This allows also to simplify the code.

No functional change.
2014-09-21 08:23:12 +08:00
mbootsector 2eec710318 King-pawn threat bonus for endgames.
STC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10224 W: 1765 L: 1638 D: 6821

LTC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6923 W: 1027 L: 899 D: 4997

bench 7818100

Resolves #49
2014-09-19 22:36:05 +08:00
Marco Costalba aa10d0aaa6 Rename "Contempt Factor" to "Contempt"
Suggested by Ronald and Lucas on talkchess.

No functional change.

Conflicts:
	src/ucioption.cpp
2014-09-16 21:14:00 +01:00
Marco Costalba ff480bdb83 Retire struct Log
No more used now that we have removed
"Write Search Log" UCI option.

No functional change.
2014-09-16 21:13:50 +01:00
lucasart b66552fc27 Simplify Time Management UCI options
No functional change
2014-09-15 20:24:51 +01:00
Marco Costalba cd065dd584 Small tweak to idle_loop()
In case of a succesful late join we set again
'searching' flag, so we can restart search
immediately without an useless lock/unlock
cycle.

No functional change.
2014-09-04 20:19:03 +01:00
Joona Kiiski 95b24083fb Simplify idle_loop()
No functional change
2014-08-30 21:03:41 +01:00
lucasart 877313a413 Retire Search Log
No functional change

Bench: 7461881
2014-08-24 16:28:51 +01:00
Marco Costalba 8b8885ab07 Fix perft 1
Compute correct number of moves for this corner case.

A smal bug crept in after recent perft rework.

No functional change.
2014-08-10 07:36:22 +08:00
lucasart a903ed07e0 Retire move_to_san()
Now "Write Search Log" will pring moves in UCI format, consistent with all the rest. This functionality is
not aimed at end-users anyway. It's hardly useful at all, in fact. Also, pretty-printing SAN moves is
something that better belongs in the GUI than in the engine.

No functional change.
2014-08-10 07:28:00 +08:00
lucasart 880e3cd7c8 Move to_char() and to_string() to notation
Where they better belong.

Also, this removes '#include <string>' from types.h, which reduces the amount of code to compile (every
translation unit includes types.h).

No functional change.
2014-08-09 13:25:05 +08:00
Marco Costalba a67c22611a Rework perft implementation
Unify various perft functions and move all the code
to search.cpp.

Avoid perft implementation to be splitted between
benchmark.cpp (where it has no reason to be) and
search.cpp

No functional and no speed change (tested).
2014-08-09 13:00:59 +08:00
lucasart 2efeded6e3 Write perft(N-1) into cout
So that one can redirect cout to /dev/null and only print print cerr in the terminal (for more accurate speed
tests).

Suggested by Marco.

No functional change.
2014-08-07 21:15:05 +08:00
lucasart 6044f25d71 Fix Hash in bench
The compiler tries to cast Options["Hash"] into a string, using:

Option::operator std::string() const {
  assert(type == "string");
  return currentValue;
}

And, as expected, the assert() fails.

std::to_string() would be the right solution, but it's C++11. And using a stringstream is too much code to
achieve so little. Let's keep it the way it was: hardcoded (ie. default hash defined in two places).

No functional change.
2014-08-07 18:51:07 +08:00
joergoster 9da015517c Remove insufficient material rule
The eval already returns zero in KK, KBK, KNK (see material.cpp). The difference is:
- we lose the "TB pruning" benefit of the draw rule (ie. search goes on even if eval is zero)
- we gain some speed by removing a useless test from the hot path

STC:
LLR: 0.05 (-2.94,2.94) [-3.00,1.00]
Total: 128000 W: 21357 L: 21560 D: 85083

LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 33023 W: 4613 L: 4509 D: 23901

bench 7461881
2014-08-06 18:42:10 +08:00
lucasart 888a1d3445 Remove useless code in Position::pretty()
First, remove some dead code (function never called with a Move argument).

Then, remove printing of legal moves, which does not belong here. Let's keep commands orthogonal and minimal:
- the "d" command should display the board, nothing more, or less.
- "perft 1" will display the list of legal moves.

No functional change.
2014-08-06 07:04:27 +08:00
lucasart 8b88ca9017 Reduce minimum memory requirement by 16MB
Stockfish allocates the default hash (32MB) in main(), before entering UCI::loop(). If there is not enough
memory, the program will crash even before UCI::loop() is entered and the GUI is given a change to specify a
lower Hash value.

This defective design could be resolved by doing a lazy allocation upon "isready" command, as the UCI protocol
guarantees that "isready" will be sent at least once before any search. But it's a bit cumbersome when using
Stockfish "manually" to have to remember to type "isready" everytime.

So leave the current design, but reduce the default hash to 16MB instread of 32MB. In order to perform such
quick searches (depth=13), there is no reason to use so much Hash anyway. Another benefit is to introduce a
bit of hash pressure in bench, which increases chances to detect rare bugs related to TT replacement, for
example.

This is not a functional change, although it obviously changes the bench.

bench 7461879
2014-08-05 11:42:48 +08:00
lucasart 94fe366779 Default Hash defined in a single place
Instead of defining it both in ucioption.cpp and benchmark.cpp. Obviously changing the default Hash will
change the bench as a result.

No functional change.
2014-08-05 11:40:25 +08:00
lucasart 535f70088e Retire divide command
The main purpose of perft is to help debugging. But without the breakdown in sum of perft(N-1), it is a
completely useless debugging tool.

So perft now displays the breakdown, and divide is therefore removed.

No functional change.
2014-08-04 13:54:09 +08:00
Marco Costalba 1b69910865 Ensure printing UCI info in multi-pv case
After commit 94b1bbb68b, in case available root moves are less than multiPV, we
could never reach condition:

PVIdx + 1 == multiPV

and as a consequence UCI output is not printed.

Fixed suggested by Joerg Oster.

No functional change.
2014-08-02 22:16:45 +08:00
Oskar Werkelin Ahlin 29451de874 Correct bench timing
No functional change.
2014-08-01 19:32:53 +08:00
lucasart 5adc678628 Document Threat[] indices
From Marco's repo.

No functional change.
2014-07-30 07:06:48 +08:00
Marco Costalba f2053ba19f Fix a warning with MSVC 2010
Warning C4267: 'argument' : conversion from 'size_t' to 'int', possible loss of data

No functional change.
2014-07-29 06:53:09 +08:00
David Zar 68171ecaca Small code style reformatting
No functional change.
2014-07-29 06:52:32 +08:00
lucasart 6b82a234d0 Aspiration: widen slower
STC:
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 182323 W: 30664 L: 30234 D: 121425

LTC:
LLR: 3.51 (-2.94,2.94) [0.00,4.00]
Total: 59841 W: 8345 L: 8006 D: 43490

bench 7962536
2014-07-27 09:18:04 +08:00
David Zar 4758fd31b1 Outpost tuning
double mg bonus and half eg bonus.

STC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 18142 W: 3094 L: 2948 D: 12100

LTC:
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 83561 W: 11706 L: 11329 D: 60526

bench 7831429
2014-07-23 07:05:10 +08:00
lucasart 7f68fc611c Revert Contempt = 20
Despite being neutral at STC, it turned out to be regressive at LTC:

40k games at LTC with Hash=8
ELO: -2.06 +-1.9 (95%) LOS: 1.4%
Total: 39720 W: 5740 L: 5976 D: 28004

40k games at LTC with Hash=128
ELO: -2.69 +-1.9 (95%) LOS: 0.2%
Total: 39149 W: 5702 L: 6005 D: 27442

bench 7477963
2014-07-20 08:50:54 +08:00
Marco Costalba 896bd917f8 Simplify evaluate_passed_pawns
From a suggestion by David Zar.

No functional change.
2014-07-15 20:59:32 +08:00
lucasart 67a5e1ecf9 Contempt = 20
Also raise the admissible bounds to (-100,100), as there is no reason to prevent users from using high
values if they want to.

Does not regress in self play:
ELO: 0.10 +-2.0 (95%) LOS: 53.7%
Total: 40000 W: 7084 L: 7073 D: 25843

master vs SF 3
ELO: 182.86 +-2.7 (95%) LOS: 100.0%
Total: 40000 W: 21843 L: 2541 D: 15616

Contempt = 20 vs SF 3
ELO: 189.25 +-2.8 (95%) LOS: 100.0%
Total: 40000 W: 22721 L: 2859 D: 14420

Diff is therefore 6.4 +/- 3.9 elo against a 180-190 elo weaker engine, which is significantly positive,
as expected. This elo difference is likely understated, because of FishTest aggressive draw adjudication
though.

We could push Contempt further, but after 20cp, it would get in the way of FishTest draw adjudication
rule, and is likely to reduce the testing throughput as a result.

bench 8198667
2014-07-15 07:14:58 +08:00
David Zar a9b8e8b931 Bonus passed pawn blocked by our pieces
passed STC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 63965 W: 10950 L: 10692 D: 42323

and LTC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 19237 W: 2740 L: 2562 D: 13935

bench: 7477963
2014-07-14 21:19:07 +08:00
Marco Costalba 94b1bbb68b Small reformat to Skill class
Steamline a bit the implementation of
skill levels. As a side effect we can
retire MultiPV global and use a local
variable instead.

No functional change.
2014-07-11 18:08:27 +08:00
lucasart f133f61e3f Remove useless condition
Small simplification, suggested by Uri Blass.

passed STC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 25839 W: 4464 L: 4351 D: 17024

and LTC:
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 220039 W: 29981 L: 30131 D: 159927

No functional change.
2014-07-10 18:25:20 +08:00
Joona Kiiski eb50793cff Retire FakeSplit
- Currently broken
    - Never been really useful
    - Does not work well with new splitting model

Verified for no regression at STC with 3 threads:
LLR: 2.96 (-2.94,2.94) [-6.00,0.00]
Total: 81905 W: 12122 L: 12381 D: 57402

No functional change
2014-07-09 07:19:06 +08:00
Marco Costalba 9b30913996 Avoid 'double assigments' tricks
Bitboard init code is already noteasy to follow,
so don't make it even harder using 'smart' code.

Also reindent a while loop in standard way.

No functional change.
2014-07-07 21:53:25 +08:00
joergoster ac7780bd35 Tune trapped rook penalty
Passed STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 15687 W: 3352 L: 3199 D: 9136

and LTC (parameter tweaks)
LLR: 2.95 (-2.94,2.94) [0.00,4.00]
Total: 27983 W: 5046 L: 4797 D: 18140

bench: 8330705
2014-07-03 19:38:54 +08:00
lucasart c192b692cf size_t cast in TranspositionTable::first_entry()
32-bit truncation would make this function bogus when clusterCount >= 2^33 (ie. Hash >= 256 GB).

No function change.
2014-07-03 18:23:56 +08:00
lucasart 85b08ce3ad Use compiler intrinsic instead of assembly for popcnt
This time, do not break compatibility with some AMD machines that have SSE3 and popcnt, but do
not have SSE4.2.

No functional change.
2014-07-03 18:22:53 +08:00
Gary Linscott b9a88da4ab Revert "Use compiler intrinsic instead of assembly for popcnt"
This reverts commit a69f1d7c20.
2014-07-01 17:01:54 -04:00
lucasart a69f1d7c20 Use compiler intrinsic instead of assembly for popcnt
No functional change.
2014-07-01 20:50:33 +08:00
lucasart 24ba204931 Raise max Hash to 1TB
And use size_t where appropriate, as suggested on FishCooking.

No functional change.
2014-07-01 18:37:18 +08:00
Ajith 6b354305e1 Add bonuses for Minors attacking enemy pieces(except pawns) even when they are protected by enemy pawns.
Patch passed STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 8206 W: 1426 L: 1304 D: 5476

and LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 19534 W: 2821 L: 2640 D: 14073

Bench: 9942172
2014-06-30 10:55:10 -04:00
joergoster ffedfa3354 Fix Singular extension condition to handle mate scores
With Eelco's patch "Don't special case for abs(beta) >= VALUE_MATE_IN_MAX_PLY" condition "abs(ttValue) < VALUE_KNOWN_WIN" has been removed from singular extension search, and condition "abs(beta) < VALUE_KNOWN_WIN" was added to the SingularExtensionNode definition.
This might lead to problems, especially in positions, where a mate is due.
For example, this position 5rk1/4K1pp/8/5PPP/8/8/8/1R6 w - - 12 1 triggers an assert.
stockfish: search.cpp:434: Value {anonymous}::search(Position&, Search::Stack*, Value, Value, Depth, bool) [with {anonymous}::NodeType NT = (<unnamed>::NodeType)2u; bool SpNode = false]: Assertion `-VALUE_INFINITE <= alpha && alpha < beta && beta <= VALUE_INFINITE' failed.

So let's re-insert the removed condition.
First spotted by Uri Blass, fix by me.

Bench: 8759675
2014-06-29 20:17:40 +01:00
Joseph R. Prostko e4e423bb05 Change the install prefix for Haiku
* /boot/common was removed from Haiku
* The equivalent path now that package management has been implemented is /boot/system/non-packaged

No functional change

Bench: 8759681
2014-06-29 15:15:10 +01:00
Ron Britvich ccd823a4ff Pack 3 TT entries in 32 bytes cluster
Idea from Ron Britvich

Code reworked by Marco Costalba and Joona Kiiski

Bench: 8095369

Resolves #3
Resolves #10
2014-06-28 14:06:32 -04:00
Gary Linscott 7ff865b924 Merge pull request #9 from glinscott/pawnspan
Scale down endgames with pawns on one or two adjacent files

Passed STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 16081 W: 2745 L: 2604 D: 10732

Passed LTC
LLR: 2.95 (-2.94,2.94) [1.00,6.00]
Total: 123832 W: 17292 L: 16584 D: 89956

128k games to measure ELO at 15+0.05:
ELO: 2.07 +-1.1 (95%) LOS: 100.0%
Total: 128000 W: 21632 L: 20869 D: 85499

New bench: 8028792
2014-06-26 12:21:06 -07:00
Gary Linscott ec627911f0 A bit more cleanup 2014-06-26 15:20:30 -04:00
Gary Linscott 22ffb588e5 Merge pull request #8 from glinscott/revert_ce1c260
Revert "Check for an available slave early on"
2014-06-26 08:20:45 -07:00
Gary Linscott e6a8d03dd8 Revert "Check for an available slave early on"
This reverts commit ce1c260ea9.
2014-06-26 11:16:36 -04:00
Gary Linscott 747e035c7d Original version of shane's patch 2014-06-25 16:45:14 -04:00
shane31 6c9f4cf36f Scale down endgames with pawns on one or two adjacent files 2014-06-25 16:01:00 -04:00
Gary Linscott ab580106fd Merge pull request #5 from glinscott/authors
Add AUTHORS
2014-06-23 09:14:03 -07:00
Marco Costalba aa23eb4440 Some reformatting in pawns.cpp
No functional change.
2014-06-21 15:08:31 +02:00
Marco Costalba e8baf2b772 Ensure ttValue != VALUE_NONE in singular extension search
The assert:

  assert(ttValue != VALUE_NONE);

Could fire for multiple reasons (although is very rare),
for instance after an IID we can have ttMove != MOVE_NONE
while ttValue is still set at VALUE_NONE.

But not only this, actually SMP is a source of corrupted
ttValue and anyhow we can detect the condition:

 ttMove != MOVE_NONE && ttValue == VALUE_NONE

even north of IID.

Reported by Ronald de Man.

It is so rare that bench didn't change.

bench: 7710548
2014-06-21 13:07:29 +02:00
Gary Linscott 58bb23d0c3 Remove some duplicates 2014-06-21 00:49:49 -04:00
Gary Linscott bbacff5e01 Add AUTHORS 2014-06-20 21:35:09 -04:00
Marco Costalba 43efd7fad7 Better value clipping in game_phase()
No functional change.
2014-06-21 00:05:14 +02:00
Marco Costalba 3b315c9ada Move game_phase() to Position
It seems a more natural to place this
function there.

No functional change.
2014-06-20 23:44:46 +02:00
Marco Costalba f7926ea41e Small renaming in material weights
Also dropped some temporary variable: compiler
is more than able to push on stack temp values
by itself (verified).

No functional change.
2014-06-20 23:15:31 +02:00
Marco Costalba 264c8637a3 Simplify a condition in is_KXK()
No functional change.
2014-06-19 15:44:42 +02:00
Eelco de Groot 55a3e0af8d Don't special case for abs(beta) >= VALUE_MATE_IN_MAX_PLY
Remove from the search this special case and apply
null search and razoring also in mate positions.

Tested in no-regression mode and passed both

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 65431 W: 10860 L: 10810 D: 43761

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 34928 W: 4814 L: 4713 D: 25401

This patch kicks in only in mate positions and in
these cases it seems beneficial in finding mates
faster as Yery Spark measured on the Chest mate suite:

Total number of positions 6425
Fixed nodes 200K per position

master: 1049
new:    1154

And also the 5446 'hard' positions again with 2000K nodes
(those not found by both engines in 200K nodes):

master: 1069
new:    1395

bench: 7710548
2014-06-16 21:50:14 +02:00
Leonid Pechenik 66c93245e0 Simplify unstoppable pawns
Tested in no-regression mode and passed both

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17919 W: 3103 L: 2978 D: 11838

and LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 31460 W: 4414 L: 4308 D: 22738

bench: 7709279
2014-06-16 12:39:50 +02:00
Marco Costalba 73ca93f3c0 Remove -ansi flag for Clang
It seems this flag is only for gcc and
yields a warning under OSX Mavericks:

clang: warning: argument unused during compilation: '-ansi'

No functional change.
2014-06-15 10:31:57 +02:00
Marco Costalba c6fc51c5ea Fix a warning with MSVC in 'analyze' mode
Here MSVC is worried that

StepAttacksBB[PAWN][psq]

could overflow, so change psq initialization
to clarify psq is always less than 64.

No functional change.
2014-06-14 12:46:58 +02:00
Marco Costalba 2cb4c7052e Triviality in UCI::loop
Code style paranoid in action here :-)

No functional change.
2014-06-14 12:26:08 +02:00
Marco Costalba 3c1201c20c Factor out pawn attacks in 'lever'
Improves readibility and possibly speed.

No functional change.
2014-06-14 12:20:54 +02:00
Ajith 84dabe5982 Simplify pawn threats and merge into ThreatenedByPawn[]
Tested in no-regression mode,

passed STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 14477 W: 2493 L: 2362 D: 9622

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 51964 W: 7091 L: 7013 D: 37860

bench: 7875814
2014-06-12 08:32:16 +09:00
Reuven Peleg e10276f45d Tweak outpost name
This name is more accurate, since that function evaluates only one
outpost in every call.
No functional change.
2014-06-11 23:42:47 +09:00
Reuven Peleg 2312c26722 Add bonus for advanced lever
Passed both STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 15413 W: 2670 L: 2530 D: 10213

And LTC:
LLR: 2.95 (-2.94,2.94) [1.00,6.00]
Total: 66908 W: 9398 L: 8960 D: 48550

Bench : 7859385
2014-06-11 23:38:25 +09:00
Marco Costalba b6cd89aeaf Small renaming in Tracing
No functional change.
2014-06-09 05:30:18 +09:00
Marco Costalba ce1c260ea9 Check for an available slave early on
Don't take the split lock if we don't have
available slaves (about 30-40% of times).

This new condition allows to retire the now
redundant one on number of threads.

No functional change.
2014-06-09 04:56:31 +09:00
Marco Costalba 4d30126e4b Use unsigned char as argument of std::isspace
Although signature allows an int:

int isspace( int ch );

The behavior is undefined if the value of ch is
not representable as unsigned char and is not
equal to EOF.

See

http://en.cppreference.com/w/cpp/string/byte/isspace
http://www.greenend.org.uk/rjk/tech/cfu.html

This is really a tricky corner case of C standard!

Spotted and reported by Ron Britvich.

No functional change.
2014-06-07 01:21:47 +02:00
Marco Costalba 7f56d2949d Avoid to use nullChild
Use instead:

(ss-1)->currentMove == MOVE_NULL

No functional change.
2014-06-06 11:12:05 +02:00
Marco Costalba d10ae90dea Remove 'update gains' hack
Use (move != MOVE_NONE) condition to
filtering out updating gains at root.

bench: 8454456
2014-06-06 11:10:40 +02:00
Marco Costalba 2f75639485 Temporary revert previous patch
Split previous patch in 2 steps: first remove
the MOVE_NULL hack, then retire nullChild.

The first step is a prerequisite
for second one and affects bench.

The second step (next patch) just removes nullChild
without affecting bench.

bench: 8205159
2014-06-06 11:08:35 +02:00
Marco Costalba ad1167c482 Avoid to use nullChild
Should be a non functional change, but
for some reason bench is changed.

bench: 8454456
2014-06-06 10:07:48 +02:00
Marco Costalba 08753771fc Move Tempo to evaluation
No functional change.
2014-06-06 09:40:01 +02:00
Marco Costalba 69ac45d903 Revert "Score extractors"
Are broken for big-endian case and
I have verified with MSVC 2013 Premium
bench is correct and there is no
miscompilation, so the main reason
to change the original code drops.

No functional change.
2014-06-05 23:55:18 +02:00
Marco Costalba 323a006666 Fix a warning with MSVC Premium 2013
Reported by Ron Britvich.

No functional change.
2014-06-05 23:46:58 +02:00
pellanda 16e170d105 Update Readme.md
Retire line about polyglot.ini and
book as they are not used anymore.

No functional change.
2014-06-04 22:31:59 +02:00
Lucas Braesch adeded29fb Symmetric King Safety: take 2
Another attempt at retiring current asymmetric
king evaluation and use a much simpler symmetric
one. As a good side effect we can avoid recalculating
eval after a null move.

Tested in no-regression mode and passed

STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 21580 W: 3752 L: 3632 D: 14196

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 18253 W: 2593 L: 2469 D: 13191

And a LTC regression test against SF DD to
verify we don't have regression against
weaker engines due to some kind of 'contempt'
effect:

ELO: 54.69 +-2.1 (95%) LOS: 100.0%
Total: 40000 W: 11072 L: 4827 D: 24101

bench: 8205159
2014-06-03 20:01:36 +02:00
Marco Costalba 187a9fe5e7 Use see() instead of see_sign() in previous patch
Before it was working by accident in case of
see_sign() and failing with see() due to how
castle moves are coded (king captures the rook).

Better to explicitly filter out castling moves
and use see() without any surprise/trick.

No functional case.
2014-06-03 12:38:58 +02:00
kinderchocolate 6f48367094 Add some const qualifier
No functional change.
2014-06-03 11:43:52 +02:00
Reuven Peleg 83a574ff27 Decrease reduction for moves that escape a capture
Passed both STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 10094 W: 1833 L: 1704 D: 6557

and LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 27738 W: 4147 L: 3928 D: 19663

bench : 8599236
2014-06-03 11:33:19 +02:00
Marco Costalba f4dcec0b94 Retire Polyglot Book management
Book handling belongs to GUI, we kept this code
for historical reasons, but nowdays there is
really no need of this old, (mostly) unused
and especially incorrect designed functionality.

It is up to the GUI to choose the book (far easier for
the user) and to select the book parameters. In no
place, including fishtest, TCEC, rating lists, etc.
the "own book" is used, moreover currently SF is
released without any book and even if in the future we
bundle a book in the release package, it will be the GUI
that will take care of it.

This corrects a wrong design decision that Galurung
and later Stockfish inherited from what was common
practice many yeas ago.

No functional change.
2014-06-01 15:29:34 +02:00
Marco Costalba e4fc9d84d7 Retire eval weights UCI options
There is really little that user can achieve (apart
from a weakened engine) tweaking these parameters
that are already tuned and have no immediate or visible
effect.

So better do not expose them to the user and avoid the
typical "What is the best setup for my machine?" kind of
question (by far the most common, by far the most useless).

No functional change.
2014-06-01 13:37:21 +02:00
Marco Costalba 07a525cdfd Retire polyglot.ini
All actively maintained GUI are able
to understand UCI protocol, so there
is no more need for this obsolete file.

No functional change.
2014-05-31 23:48:23 +02:00
Marco Costalba f5622cd5ec Restore development version
bench: 8732553
2014-05-31 23:34:36 +02:00
Marco Costalba 54f8a9cb13 Stockfish 5
Stockfish bench signature is: 8732553
2014-05-31 09:16:54 +02:00
Marco Costalba ad937d0b2d Revert "Symmetric king safety"
Regression test of 40K games at 60 secs shows
this commit to be a 2-3 ELO regression.

So revert to original king safety.

bench: 8732553
2014-05-26 21:39:48 +02:00
Marco Costalba 495a0fa699 Fix a warning with Intel compiler
warning #2259: non-pointer conversion from
"int" to "int16_t={short}" may lose significant
bits.

No functional change.
2014-05-25 00:21:46 +02:00
Marco Costalba 585655b16e Tidy up tt.h
Backport some non-functional changes
found working on 'dense TT' patch.

No functional change.
2014-05-25 00:02:09 +02:00
Marco Costalba e49eb67119 Add perft 'divide' command
To show perft numbers for each move. Just
use 'divide' instead of 'perft', for instance:

position startpos moves e2e4 e7e5
divide 4

Inspired by Ronald de Man.

No functional change.
2014-05-24 09:56:32 +02:00
Marco Costalba 88b5100e29 Update polyglot.ini after last patch
No functional change.
2014-05-19 21:19:47 +02:00
Lucas Braesch 40f5abba10 Symmetric king safety
Retire current asymmetric king evaluation
and use a much simpler symmetric one.

As a side effect retire the infamous
'Aggressiveness' and 'Cowardice' UCI
options.

Tested in no-regression mode,

Passed both STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 33855 W: 5863 L: 5764 D: 22228

And LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 40571 W: 5852 L: 5760 D: 28959

bench: 8321835
2014-05-19 14:24:39 +02:00
Marco Costalba 5e03734eac Fix an off-by-one bug in extract_pv_from_tt
At root we start counting plies from 1,
instead pv[] array starts from 0. So
the variable 'ply' we use in extract_pv_from_tt
to index pv[] is misnamed, indeed it is
not the real ply, but ply-1.

The fix is to leave ply name in extract_pv_from_tt
but assign it the correct start value and
consequentely change all the references to pv[].
Instead in insert_pv_in_tt it's simpler to rename
the misnamed 'ply' in 'idx'.

The off-by-one bug was unhidden when trying to use
'ply' for what it should have been, for instance in
this position:

position fen 8/6R1/8/3k4/8/8/8/2K5 w - - 0 1

at depth 24 mate line is erroneusly truncated due
to value_from_tt() using the wrong ply.

Spotted by Ronald de Man.

bench: 8732553
2014-05-17 22:59:07 +02:00
Marco Costalba e46a72dd1d Extract a reliable PV line
Truncate the extracted PV from the point where
the score stored in hash starts to deviate from
the root score.

Idea from Ronald de Man.

bench: 8732553
2014-05-17 12:49:52 +02:00
Michel Van den Bergh 5ec63eb6b6 Drop to qsearch at low depth in razoring
If razoring conditions are satisfied and
depth is low, then directly drop in qsearch.

Passed both STC
LLR: 2.98 (-2.94,2.94) [-1.50,4.50]
Total: 12914 W: 2345 L: 2208 D: 8361

And LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 50600 W: 7548 L: 7230 D: 35822

bench: 8739659
2014-05-13 22:37:28 +02:00
Arjun Temurnikar a3c8c4b70d Remove undefended minors
Tested in "no regression" mode.

Passed both STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 68026 W: 12277 L: 12236 D: 43513

And LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 85682 W: 12861 L: 12836 D: 59985

bench: 7311935
2014-05-13 21:37:42 +02:00
Marco Costalba 696d6cedb9 Save stalemates in TT
When there aren't legal moves after
a search, instead of returning imediately,
save bestValue in TT as in the usual case.

There is really no reason to special case
this one.

With this patch is fully fixed (again) follwing
position:

    7k/6p1/6B1/5K1P/8/8/8/8 w - - 0 1

Also in SMP case.

bench: 8802105
2014-05-11 10:56:25 +02:00
Marco Costalba 9f843adf89 Retire "Idle Threads Sleep" UCI option
After last Joona's patch there is no measurable
difference between the option set or unset.

Tested by Andreas Strangmüller with 16 threads
on his Dual Opteron 6376.

After 5000 games at 15+0.05 the result is:

1 Stockfish_14050822_T16_on   : 3003  5000 (+849,=3396,-755), 50.9 %
2 Stockfish_14050822_T16_off  : 2997  5000 (+755,=3396,-849), 49.1 %

bench: 880215
2014-05-11 10:29:56 +02:00
Arjun Temurnikar bfd8704a7d Make imbalance table more clear
No functional change.
2014-05-10 08:54:31 +02:00
Reuven Peleg f89a8f0769 Pass Position as const ref in update_stats()
No functional change.
2014-05-08 22:36:30 +02:00
Marco Costalba 6ba1d3ead6 Clarify some comments in SMP code
Spotted by Joona.

No functional change.
2014-05-08 09:09:35 +02:00
Marco Costalba 7e3dba4f4c Reformat and simplify previous patch
No functional change.
2014-05-07 08:56:16 +02:00
Joona Kiiski f6e98a924a Allow a slave to 'late join' another splitpoint
Instead of waiting to be allocated, actively search
for another split point to join when finishes its
search. Also modify split conditions.

This patch has been tested with 7 threads SMP and
passed both STC:

LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 2885 W: 519 L: 410 D: 1956

And a reduced-LTC at  25+0.05
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 4401 W: 684 L: 566 D: 3151

Was then retested against regression in 3 thread case
at standard LTC of  60+0.05:

LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 40809 W: 5446 L: 5406 D: 29957

bench: 8802105
2014-05-07 08:38:56 +02:00
Ron Britvich 8f6a494ad7 Rewrite Score extractors
Less tricky and even a bit faster. With this
version Visual Studio Ultimate 2013 Update 2 RC
runs fine even in O2 optimization.

No functional change.
2014-05-05 09:05:29 +02:00
Marco Costalba b8e6f83cfb Change search() signature
Pass SpNode as template parameter.

No functional change.
2014-05-04 13:35:30 +02:00
Marco Costalba 5413fda739 Revert dynamic contempt
On a final fixed game number test it failed
to prove better than standard version.

STC 15+0.05

ELO: -0.86 +-1.7 (95%) LOS: 15.8%
Total: 57578 W: 10070 L: 10213 D: 37295

bench: 8802105
2014-05-04 09:58:49 +02:00
Marco Costalba 145d293142 Revert stalemate detection in evaluation
Unfortunatly we have a slow down that causes
a regression in STC with no-regression mode:

LLR: -2.96 (-2.94,2.94) [-3.00,1.00]
Total: 22454 W: 3836 L: 4029 D: 14589

bench: 8678654
2014-05-04 09:42:32 +02:00
Marco Costalba f1240483fb Revert back KBPsK to latest Gary's version
The bug was found to be elsewhere. This version
is correct and also is able to detect as draw
positions like:

8/8/5b2/8/8/4k1p1/6P1/5K2 b - - 6 133

bench: 8678654
2014-05-04 09:34:22 +02:00
Ronald de Man c0d3010438 Fix KXK endgame
Position is win also if strong side has a bishop
and a knight (plus other material, otherwise
KBNK would be triggered instead of KXK).

This fixes a subtle bug where a search on position

k7/8/8/8/8/P7/PB6/K7 b - - 6 1

Instead of returning a draw score, suddendly returns
a big score. This happens because at one point in
search we reach this position:

8/Pk6/8/8/8/4B3/P7/K7 w - - 3 8

Where white can promote. In case of rook promotion (and also in case of
queen promotion) the resutling position gets a huge static eval that is
above VALUE_KNOWN_WIN (from the point of view of white). So for rook
promotion it is

          &&  futilityBase > -VALUE_KNOWN_WIN

that prevents futility pruning in qsearch. (Removing this condition indeed
lets the problem occur). Raising the static eval for K+B+N+X v K to a value
higher than VALUE_KNOWN_WIN fixes this particular problem without having to
introduce an extra futility pruning condition in qsearch.

I just checked and it seems K+R v K, K+2B v K and even K+B+N v K already get
a huge static eval. Why not K+B+N+P v K?

I think this fix corrects an oversight. There is special code for KBNK, but
KBNXK is handled by KXK, so the test for sufficient material should also test
for B+N.

bench: 8678654
2014-05-03 21:40:09 +02:00
Marco Costalba a9e93fa6a5 Fully correct stealmate detection
In the (rare) cases when the two conditions
are true, then fully check again with a slow
but correct MoveList<LEGAL>(pos).size().

This is able to detect false positives like
this one:

8/8/8/Q7/5k1p/5P2/4KP2/8 b - - 0 17

When we have a possible simple pawn push that
is not stored in attacks[] array. Because the
third condition triggers very rarely, even if
it is slow, it does not alters in a measurable
way the average speed of the engine.

bench: 8678654
2014-05-03 12:12:22 +02:00
Marco Costalba 9e8bf82350 Add stealmate detection to evaluation
Currently a stealmate position is misevaluated
in a negative/positive score, this leads qsearch(),
that does not detects stealmates too, to return the
wrong score and this yields to some kind of endgames
to be completely misevaluated.

With this patch is fully fixed follwing position

7k/6p1/6B1/5K1P/8/8/8/8 w - - 0 1

Also in SMP case.

Correct root cause analysys by Ronald de Man.

bench: 8678654
2014-05-03 11:47:49 +02:00
Marco Costalba dc87ec7258 Revert to Galurung's KBPsK endgame
After reverting to the original Tord's
endgame, a search on position

7k/6p1/6B1/5K1P/8/8/8/8 w - - 0 1

Reports, correctly, a draw score instead of
an advantage for white.

Issue reported by Uri Blass.

bench: 8678654
2014-05-02 11:04:18 +02:00
Marco Costalba 43973f43c6 Use only standard conforming eg_value()
Remove the optimization for Intel, is not
standard and can break at any time, moreover
our release build is not done with Intel C++
anymore so we don't need to sqeeze the extra
speed out from this compiler.

No functional change.
2014-05-01 23:08:07 +02:00
Marco Costalba bee4f1cf09 Don't save stale value in TT after split
If we return from split with a stale value
due to a stop or a cutoff upstream occurred,
then we exit moves loop and save a stale value
in TT before returning search().

This patch, from Joona, fixes this.

bench: 8678654
2014-05-01 16:26:18 +02:00
Marco Costalba da91a34c09 Better document search stop condition
Handling a stop or a cutoff in the search is
a tricky business, so better document this
difficult part of the code.

No functional change.
2014-05-01 08:48:59 +02:00
Marco Costalba 626dc8a03b Remove dead code in search
We can never have bestValue == -VALUE_INFINITE at
the end of move loop because if no legal move exists
we detect it with previous condition on !moveCount,
if a legal move exists we never prune it due to
futility pruning condition:

bestValue > VALUE_MATED_IN_MAX_PLY

So this code never executes, as I have also verified
directly.

Issue reported by Joona.

No functional change.
2014-05-01 07:46:44 +02:00
Marco Costalba cf50e265fa Fix a compile error with Intel C++
Intel compiler is very picky:
"error: this operation on an enumerated type requires an
applicable user-defined operator function"

Reported by Tony Gaor.

No functional change.
2014-04-30 08:55:45 +02:00
snicolet d3ffd0ffca Penalize hanging pieces
Passed both STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 8519 W: 1565 L: 1440 D: 5514

And LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 60618 W: 9141 L: 8777 D: 42700

bench: 8678654
2014-04-30 08:36:44 +02:00
Marco Costalba db229504e2 Rearrange interpolation formula
Put the division at the end to reduce
rounding errors. This alters the bench
due to different rounding errors, but
should not alter ELO in any way.

bench: 7615217
2014-04-28 17:27:49 +02:00
mstembera 918c29f83a Minor stuff scattered around
Just random minor stuff I found while browsing the code.

No functional change.
2014-04-28 17:07:43 +02:00
Marco Costalba a1f39c1ef9 Remove other useless floor()
No functional change.
2014-04-27 19:17:40 +02:00
Marco Costalba 9f2a64abd2 Don't need floor() in timeman.cpp
For positive numbers result is equivalent:
http://stackoverflow.com/questions/3300290/cast-to-int-vs-floor

Spotted by Joseph Ellis.

No functional change.
2014-04-27 19:10:00 +02:00
Marco Costalba 93e3b06fe2 Fix Intel compiler warnings
Fallout from previous patch: Intel compiler
is very noisy.

No functional change.
2014-04-27 12:02:36 +02:00
Marco Costalba 86c20416c8 Remove some useless casts
No functional change.
2014-04-27 11:44:16 +02:00
Marco Costalba c9e396b542 We can add an integer to a Value
We have defined corresponding operators,
so rely on them to streamline the code
and increase readibility.

No functional change.
2014-04-27 11:25:42 +02:00
Arjun Temurnikar fe23f27086 Remove rook passers eval completely
Tested in no-regression mode.

Passed STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17727 W: 3248 L: 3122 D: 11357

And (a very long!) LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 106327 W: 16258 L: 16265 D: 73804

bench: 7396783
2014-04-27 09:51:48 +02:00
Marco Costalba de6fe2cb4e Reformatting in material.h
No functional change.
2014-04-27 09:31:22 +02:00
Marco Costalba 057c3d60cd Move game phase constants to enum Value
No functional change.
2014-04-27 09:23:45 +02:00
Marco Costalba cb4ac4221e Speed up by almost 3%
This apparentely silly tweak allows
to speed up the bench by almost 3%.

Not clear why, repeating with perft,
the speed up vanishes.

Suggested by Jonathan Calovski.

No functional change.
2014-04-27 00:25:47 +02:00
Marco Costalba 55604f156b Fix issues detected by Coverity Scan
Most of Coverity Scan reports are false
positives, but in rare cases we have
confirmed (very small) issues.

No functional change.
2014-04-26 09:33:50 +02:00
Marco Costalba 7ddbcf7e87 Speed up picking of killers
Changing the order of the conditions gives
about 1% speed up!

No functional change.
2014-04-25 12:53:51 +02:00
Arjun Temurnikar 3705559fdb Remove RookOn7th and merge values into psqt
Tested in no-regression mode:

STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 55678 W: 9954 L: 9892 D: 35832

LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 9238 W: 1448 L: 1311 D: 6479

bench: 7905850
2014-04-24 08:53:46 +02:00
Arjun Temurnikar 6579a65bbb Remove penalty for knight when few enemy pawns
Tested in standard mode at STC and no-regression
mode at LTC:

STC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 19503 W: 3502 L: 3349 D: 12652

LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 67474 W: 9974 L: 9921 D: 47579

bench: 8331217
2014-04-24 08:47:13 +02:00
Jonathan Calovski f70cef3b79 Shuffle movepicker score
Believed to be a speed optimization as benched
on Windows with bench realtime affinity 0x1 deleting
highest and lowest runs:

Base	Test
1549259	1608202
1538115	1583934
1543168	1556938
1536365	1554179
1533026	1582010

Signature remains unchanged and gives anywhere from 1-2% nps
boost in analysis depending on number of cores used.

No functional change.
2014-04-24 08:38:11 +02:00
Marco Costalba a89b26bedd Correctly apply previous patch
Apply the correct values from previous patch.

bench: 8082049
2014-04-24 08:35:13 +02:00
joergoster 8bfb53efe2 Move queen vs. 3 minors rule to imbalance tables
Tuned with CLOP after 57k games.

Simplification: tested in no-regression mode.

Passed both STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17254 W: 3159 L: 3032 D: 11063

And LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 32413 W: 4967 L: 4864 D: 22582

bench: 8082049
2014-04-23 08:51:34 +02:00
Marco Costalba a66e6e5ad9 Revert "Generalize shift_bb() to handle double pushes"
Seems to intorduce some compiler warning as
reported by Gary. Warning seems bogus, but
revert anyhow.

No functional change.
2014-04-21 20:56:12 +02:00
Marco Costalba 56273fca1e Reset DrawValue[] before new search
Spotted by Ronald de Man

bench: 7384368
2014-04-21 14:30:27 +02:00
Leonid Pechenik ef43e6b05d Rise contempt when in advantage
This is a very discussed patch with many
argumentations pro and against. The fact is
it passed both STC:

LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 16305 W: 3001 L: 2855 D: 10449

And LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 34273 W: 5180 L: 4931 D: 24162

Although it is true that a correct test should
include foreign engines, we commit it anyhow so
people can test it out in the wild, under broader
conditions.

bench: 7384368
2014-04-21 12:23:03 +02:00
Marco Costalba 223ebe7b40 Streamline implementation of Position::pretty()
Simpler and more in line with Bitboards::pretty()

No functional change.
2014-04-21 11:47:01 +02:00
Stefan Geschwentner 0e8ad40ef0 Raise penalty blocked enemy pawn on 6th rank
Idea from Lyudmil Tsvetkov.

The value seems to be raised a bit abruptly, but as
Gary said, a blocked pawn on the sixth rank has been
instrumental in limiting king mobility in multiple
losses that I've seen from SF. A blocked pawn on fifth
rank is much less serious on the king safety impact.

Passed both STC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 14551 W: 2750 L: 2607 D: 9194

and LTC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 43595 W: 6917 L: 6618 D: 30060

And even a retest at 60" fixed games 40K
ELO: 1.79 +-1.9 (95%) LOS: 97.0%
Total: 39889 W: 6018 L: 5813 D: 28058

bench: 7154916
2014-04-21 07:47:50 +02:00
Marco Costalba eced15fe36 Generalize shift_bb() to handle double pushes
And use it in evaluate_space.

No functional change.
2014-04-20 15:52:37 +02:00
Arjun Temurnikar a4d058bca2 Small simplification to passed pawns
Tested in no-regression mode.

Passed both STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 28521 W: 5066 L: 4958 D: 18497

And LTC
LLR: 3.04 (-2.94,2.94) [-3.00,1.00]
Total: 21939 W: 3261 L: 3138 D: 15540

bench: 8165681
2014-04-20 10:06:51 +02:00
Joseph Hellis 619d66b7ab Remove supporting pawns
Tested in no-regression mode

Passed both STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 44957 W: 7984 L: 7903 D: 29070

and LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 12396 W: 1916 L: 1783 D: 8697

Bench: 7907885
2014-04-17 08:45:31 +02:00
Jean-Francois Romang 9073866491 Enable BMI1 too when using BMI2 ARCH
Adding BMI1 allows the compiler to use _blsr_u64
automatically (the advertised 0.3% speed gain).
I verified that the compiler does not use this
instruction with the -mbmi2 flag only. Also, all
processors supporting BMI2 is also supporting BMI1.

No functional change
2014-04-14 07:54:24 +02:00
Marco Costalba 7bce8831d3 More readable trapped rook condition
Prefer

file_of(s) < file_of(ksq)

to the inidrect

file_of(ksq) < FILE_E

To evaluate if semiopen side to check is the left side.

Also other small touches while there.

No functional change.
2014-04-13 14:29:42 +02:00
Daylen Yang 81a8c1118b Simplify Makefile by removing OS X targets
Right now the Makefile is cluttered with OS X equivalents
of all the x86 targets. We can get rid of all of them and
just check UNAME against "Darwin" for the few OS X-specific
things we need to do.

We also disable Clang LTO when using BMI2 instructions. For
some reason, LLVM cannot find the PEXT instruction when using
LTO. I don't know why, but disabling LTO for BMI2 fixes it.

No functional change.
2014-04-13 09:05:21 +02:00
Marco Costalba b2c0634d48 Move args parsing to UCI::loop
This leaves a very clean main.cpp

No functional change.
2014-04-12 13:51:52 +02:00
Marco Costalba ada55c5d0a Reshuffle in uci.cpp
Move function definitions before call site.

No functional change.
2014-04-12 12:05:25 +02:00
Marco Costalba 17cb7e7fa3 Reshuffle in timeman.cpp
Move template definitions before call site.

No functional change.
2014-04-12 12:00:37 +02:00
Marco Costalba 800ba28e83 Fix a typo in evaluate
Spotted by Lyudmil Antonov.

No functional change.
2014-04-12 10:17:41 +02:00
Marco Costalba 0b2794ae12 Retire signature-build
Does not seem used.

No functional change.
2014-04-12 09:18:55 +02:00
Jean-Francois Romang 226bbc1e63 Add ARCH x86-64-bmi2 support
Intel Haswell and newer CPUs can calculate sliders
attacks using special PEXT asm instructions instead
of magic bitboards. This gives a +3% speed up.

To enable it just compile with ARCH=x86-64-bmi2

No functional change.
2014-04-12 09:15:14 +02:00
Marco Costalba da2f8880b9 Switch to hardware PEXT
Retire software pext and introduce hardware
call when USE_PEXT is defined during compilation.

This is a full complete implementation of sliding
attacks using PEXT.

No functional change.
2014-04-12 08:55:30 +02:00
Marco Costalba c556fe1d71 Implement PEXT based attacks
According to:

https://chessprogramming.wikispaces.com/BMI2#PEXTBitboards

No functional change.
2014-04-12 08:55:30 +02:00
Marco Costalba 2bfe61c33b Add PEXT software implementation
For development/debug purposes.

No functional change.
2014-04-12 08:55:30 +02:00
Marco Costalba 2f92e3b525 Big reshuffle in evaluate.cpp
Reshuffle functions to define them in reverse
calling order (C style).

This allow us to define templates before they are
used. Currently it is not like this, for instance
evaluate_pieces is defined after do_evaluate that
calls it. This happens to work for some strange
reason (two phase lookup?) but we want to avoid
code that works 'by magic'.

As a nice side-effect we can now remove the function
prototypes.

No functional change.
2014-04-12 08:39:18 +02:00
Gary Linscott 0510112f91 Move LowMobPenalty into psq/mobility tables
Tested in no-regression mode.

Passed both STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 36705 W: 6537 L: 6443 D: 23725

and LTC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 9358 W: 1495 L: 1358 D: 6505

bench: 6921356
2014-04-12 07:15:51 +02:00
Gary Linscott 4597324572 Reduce penalty for doubled pawns further away from each other
Passed both STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 38339 W: 6849 L: 6649 D: 24841

and LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 22693 W: 3455 L: 3256 D: 15982

bench: 7508468
2014-04-12 07:05:02 +02:00
Marco Costalba ce6b7a1b85 Further simplification in evaluate
No functional change
2014-04-11 08:38:09 +02:00
Marco Costalba e3b54235ad Get rid of an evaluate_pieces() overload
Rewrite and greatly simplify that part of code.

No functional change.
2014-04-10 19:45:18 +02:00
Gary Linscott 8863afeb84 Add a penalty for low mobility pieces
Passed both STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 81857 W: 14652 L: 14342 D: 52863

and LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 45400 W: 6999 L: 6697 D: 31704

bench: 7716978
2014-04-10 08:35:10 +02:00
Gary Linscott 5c75455c8e Restrict queen mobility to safe squares
Passed both STC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 16188 W: 3119 L: 2971 D: 10098

and LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 6336 W: 1010 L: 882 D: 4444

bench: 7533692
2014-04-08 22:25:54 +02:00
Lucas Braesch 0d8a4c7565 Rescale UCI scores to PawnValueEg
This is more consistent with what other engines are doing.
Often people thinks that SF's scores are overblown. In the
end, it just boils down to the arbitrary way of rescaling them.

No functional change.
2014-04-06 11:53:28 +02:00
Marco Costalba 64d29a6330 Sync some common names
No functional change.
2014-04-06 11:26:12 +02:00
Marco Costalba fcf2a34080 Some more work in pretty_pv
No functional change.
2014-04-06 10:15:11 +02:00
Marco Costalba 698b645e10 Small tidy up in move_to_san
No functional change.
2014-04-05 19:05:37 +02:00
mstembera d28ea7b518 Smaller and faster profile-build binaries
I have noticed that increasing the bench depth produces
progressively smaller and slightly faster executables at
the cost of longer compile times.  Also using bench "time"
instead of "depth" seems to produce slightly smaller/faster
executables  given comparable compile times.

I have made a new Makefile that generates smaller and
about 1% to 2% faster profile executables at only a
little extra compile time.  On  my mobile 2GHz i7 a
full profile build time goes from 3'48" to 4'13" and
the exe goes down by 5% from 416,310 bytes to 395,567
bytes.

No functional change.
2014-04-05 12:48:20 +02:00
Lucas Braesch be641e881f Remove QueenOn7th and QueenOnPawn
Small simplification.

Passed SPRT(-3,1) both at STC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17051 W: 3132 L: 3005 D: 10914

and LTC:
LLR: 4.55 (-2.94,2.94) [-3.00,1.00]
Total: 24890 W: 3842 L: 3646 D: 17402

The rationale behind this is that I've never managed to add a
Queen on 7th rank bonus in DiscoCheck, because it never showed
to be positive (evne slightly) in testing. The only thing that
worked is Rook on 7th rank.

In terms of SF code, it seemed natural to group it with QueenOnPawn
as well as those are done together. I know you're against groupping
in general, but when it comes to non regression test, you are being
more conservative by groupping. If the group passes SPRT(-3,1) it's
safer to commit, than test every component in SPRT(-3,1) and end up
with the risk of commiting several -1 elo regression instead of just
one -1 elo regression.

In chess terms, perhaps it's just easier to manouver a Queen (which
can more also diagonaly) than a Rook. Therefore you can let the search
do its job without needing eval ad-hoc terms to guide it. For the Rook
which takes more moves to manouver such eval terms can be (marginally)
useful.

bench: 7473314
2014-04-05 11:26:44 +02:00
mstembera 3b19ea6ae5 Speed up apply_weight
Speed up by about 2% this hot path function pre-calculating
midgame and endgame values of the weight.

No functional change.
2014-04-05 11:12:18 +02:00
Marco Costalba ccf59b0228 Update Readme to 128 max threads
No functional change.
2014-04-03 11:33:42 +02:00
Joerg Oster 299afcd886 Queen vs. 3 pieces imbalance
Passed both STC
LLR: 7.32 (-2.94,2.94) [-1.50,4.50]
Total: 98108 W: 18087 L: 17576 D: 62445

And LTC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 15082 W: 2417 L: 2248 D: 10417

bench: 7717336
2014-04-03 10:42:06 +02:00
Marco Costalba c15b132f03 Make operator<< to return void
This should help preventing misuse.

No functional change.
2014-04-03 10:34:25 +02:00
Stefan Geschwentner 0ba814b3ca Drop not defended by pawn condition
Passed no-regression test both at STC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 96554 W: 17563 L: 17572 D: 61419

and at LTC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 31891 W: 4905 L: 4801 D: 22185

bench: 7720896
2014-04-01 11:59:25 +02:00
Marco Costalba d2caba1f66 Do not workaround function argument evaluation
Rewrite options initialization to do not trying to
hack upon the undefined evaluation order of function
arguments.

No functional change.
2014-04-01 09:39:41 +02:00
Marco Costalba 1efc19ade0 Fix a compile error with Intel C++
Error: a value of type "int" cannot be assigned
to an entity of type "Value"

No functional change.
2014-03-30 14:20:12 +02:00
Marco Costalba 678425f274 Fix a warning with MSVC 2010
Warning C4804: '<' : unsafe use of type 'bool' in operation

No functional change.
2014-03-30 14:25:57 +02:00
Marco Costalba 422c9c2acd Show evaluation from white POV in trace
We chose this instead of negamax sign convention
(ie. from the point of view of the side to move)
because it is more in line to how the eval
table is presented.

Also some tweak to formatting while there.

No functional change.
2014-03-30 10:45:46 +02:00
Marco Costalba 9350d0dce5 Raise VALUE_INFINITE
In some legal positions like this one:
R6R/3Q4/1Q4Q1/4Q3/2Q4Q/Q4Q2/Np1Q4/kB1N1KB1  b -- 0 1

We can have a very high score, in this case 30177 and 29267
for midgame and endgame respectively, and because
VALUE_INFINITE = 30001 we have an assert in interpolate()

Midgame and endgame scores are stored in 16 bit signed integers
so we can rise VALUE_INFINITE a little bit. This does not fix
the possibility of overflow in general case, just makes the
condition more difficult to trigger and anyhow better uses all
the score width.

Raising VALUE_INFINITE to 32000 seems to fix the problem for this
particular case.

No functional change.
2014-03-29 11:13:42 +01:00
Jean-Francois Romang 4833887842 Further simplification of TT replace strategy
No functional change
2014-03-29 10:05:02 +01:00
Marco Costalba f811a5693e Restore old aspiration window to 16
Tested directly at LTC because previous long
test series on this topic shows it is TC dependant.

Tested with no-regression mode because gets rid of
an ugly and ad-hoc rule.

Test at LTC:
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 67918 W: 10590 L: 10541 D: 46787

bench: 7926803
2014-03-29 08:45:40 +01:00
Stefan Geschwentner af0c13ba6a Pinned pieces affect king safety
Here the new idea is to link pinned pieces
with king safety.

Passed both STC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 10047 W: 1867 L: 1737 D: 6443

And LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 10419 W: 1692 L: 1543 D: 7184

bench: 8325087
2014-03-29 08:37:55 +01:00
Marco Costalba c7cf45241c Fix a bug in pawns eval tracing
Instead of totals we were showing white and
black values.

Spotted by Sven Schüle

No functional change.
2014-03-26 07:06:29 +01:00
Marco Costalba 865b71309c Simplify TT replace strategy
Tested for no-regression with SPRT[-3, 1] at STC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 32046 W: 6020 L: 5918 D: 20108

No functional change.
2014-03-24 08:32:08 +01:00
snicolet e7362dae78 Introduce penalty for weak (=unsupported) pawns.
We add a penalty for each pawn which is not protected by another pawn
of the same color.

Passed both short TC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 12107 W: 2411 L: 2272 D: 7424

And long TC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 9204 W: 1605 L: 1458 D: 6141

bench: 7682173
2014-03-24 08:27:40 +01:00
Marco Costalba 9e72e35942 Fix an incorrect 'friend' declaration
Spotted by Lee David.

No functional change.
2014-03-23 11:17:38 +01:00
Marco Costalba bc183b0c04 Retire last usage of operator|(File f, Rank r)
This for some reason was missed.

No functional change.
2014-03-23 10:42:37 +01:00
mstembera ffdf63ff7c Refresh TT entries generation automatically on probe
And other assorted simplifications, tested with SPRT[-3, 1]

Passed both short TC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 18814 W: 3600 L: 3475 D: 11739

And long TC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 20731 W: 3217 L: 3096 D: 14418

No functional change.
2014-03-23 09:46:15 +01:00
Marco Costalba f12449d492 Rename score to value in ExtMove
We use 'score' for the value mid-endgame pair.

No functional change.
2014-03-23 08:54:10 +01:00
Marco Costalba 441a34a1c6 Trivial formatting in ucioption.cpp
No functional change.
2014-03-23 00:06:26 +01:00
Marco Costalba 5f12069cbf Retire operator|(File f, Rank r)
Use make_square() instead. Less fancy but
more clear.

No functional change.
2014-03-22 23:54:18 +01:00
Marco Costalba 4eee603433 Fix a subtle bug in UCI options printing
We want all the UCI options are printed in the order in which are
assigned, so we use an index that, depending on Options.size(),
increases after each option is added to the map. The problem is
that, for instance, in the first assignment:

o["Write Debug Log"] = Option(false, on_logger);

Options.size() can value 0 or 1 according if the l-value (that
increments the size) has been evaluated after or before the
r-value (that uses the size value).

The culprit is that assignment operator in C++ is not a
sequence point:

http://en.wikipedia.org/wiki/Sequence_point

(Note: to be nitpick here we actually use std::map::operator=()
 that being a function can evaluate its arguments in any order)

So there is no guarantee on what term is evaluated first and
behavior is undefined by standard in this case. The net result
is that in case r-value is evaluated after l-value the last
idx is not size() - 1, but size() and in the printing loop
we miss the last option!

Bug was there since ages but only recently has been exposed by
the removal of UCI_Analyze option so that the last one becomes
UCI_Chess960 and when it is missing engine cannot play anymore
Chess960.

The fix is trivial (although a bit hacky): just increase the
last loop index.

Reported by Eric Mullins that found it on an ARM and MIPS
platforms with gcc 4.7

No functional change.
2014-03-22 11:30:06 +01:00
Marco Costalba c714b90594 Fix a typo
Spotted by Isaac Haïk Dunn.

No functional change.
2014-03-18 18:36:39 +01:00
Marco Costalba aab5863dd4 Increase max threads to 128
Thanks to std::bitset we can easily increase
the limit of active threads above 64.

Thanks to Lucas Braesch for pointing at the
correct solution of using std::bitset.

No functional change.
2014-03-18 12:07:26 +01:00
Marco Costalba fa3f6dcbea Fix a crash under MSVC
Using memset on a std::vector is undefined behavior,
so manually init all the data memebers of LimitsType.

Bug intorduced in 41641e3b1e

No functional change.
2014-03-16 10:55:58 +01:00
Marco Costalba a091ae4cc8 Split also if no slaves are found
Because we test for available slaves before
entering split(), we almost always allocate a
slave, only in the rare case of a race (less
then 2% of cases) this is not true, but to
special case this occurrence is not worth
the added complexity.

bench: 7451319
2014-03-15 23:43:35 +01:00
Marco Costalba a1a7bc84da Remove "Max Threads per Split Point" UCI option
Experimental patch to verify if drop of nps
in endgames at very long TC is due to this.

Suggested by Ronald de Man.

bench: 7451319
2014-03-15 21:26:04 +01:00
Marco Costalba 6e4b4c42ed Merge default tests in pos_is_ok
No functional change.
2014-03-15 15:34:31 +01:00
Marco Costalba 142874b058 Microptimize castling in undo_move()
We don't need to set 'captured' and 'pt' after we
castle back.

No functional change.
2014-03-15 11:47:30 +01:00
Mysseno 36c381154b Depth dependant aspiration window delta
Split delta value in aspiration window so that when
search depth is less than 24 a smaller delta value
is used. The idea is that the search is likely to
be more accurate at lower depths and so we can exclude
more possibilities, 25% to be exact.

Passed STC
LLR: 2.96 (-2.94, 2.94) [-1.50, 4.50]
Total: 20430 W: 3775 L: 3618 D: 13037

And LTC
LLR: 2.96 (-2.94, 2.94) [0.00, 6.00]
Total: 5032 W: 839 L: 715 D: 3478

Bench: 7451319
2014-03-14 19:47:41 +01:00
Marco Costalba d0587f2c7f Use legal positions for endgame initialization
During endgame initialization we get the material
hash key of each endgame forging and ad-hoc position
that in same cases is illegal (leaves teh king under
capture). This is not a problem for the material key,
but rises an assert when SF is run in debug mode with
'testKingCapture' set in pos_is_ok().

So rewrite the code to always produce legal positions.

No functional change.
2014-03-14 09:57:34 +01:00
Marco Costalba 1f4798a173 Further work in pos_is_ok()
No functional change.
2014-03-14 09:43:19 +01:00
Daylen Yang 61e4443342 Re-add link time optimization on OS X
In the new version of clang, -O4 does not imply -flto, so we set the
flag
2014-03-14 08:49:08 +01:00
Marco Costalba 6571acffaa Reformat do_castling()
No functional change.
2014-03-13 12:53:03 +01:00
Marco Costalba 6f3d787692 Further merge StateInfo setup functions
No functional change.
2014-03-12 22:46:17 +01:00
Marco Costalba 1e032ece92 Merge hash key computation functions
No functional change.
2014-03-12 09:14:38 +01:00
Marco Costalba cc3002ff04 Use std::count in pos_is_ok()
No functional change.
2014-03-11 23:19:47 +01:00
Marco Costalba c40d4e0133 Small simplification in gives_check
Use switch statement also for normal case.

No speed regression.

No functional change.
2014-03-11 22:58:08 +01:00
Marco Costalba ca0804dfe4 Print dbg counters after bench
Print last debug counters update just
before to exit benchmark.

Suggested by Stephane Nicolet.

No functional change.
2014-03-11 22:19:14 +01:00
Marco Costalba d2a8ba3299 Simplify pseudo_legal()
Big simplification of pawn move check.

Code has been tested with a brute force approach: for
every position reached during a bench search, the function
has been called for each combinations of Move(from, to)
and verified the result is the same of old code.

Actually this function is very critical becuase is the
one that ensures corrupted TT moves are discarded, so
to properly test it a simple bench is not enough.

Verified also speed is not changed.

No functional chnage.
2014-03-10 08:38:23 +01:00
Marco Costalba 20ff12e1be Simplify generate<EVASIONS>
No speed regression, tested with both perft and
bench.

No functional change.
2014-03-09 12:16:27 +01:00
Marco Costalba 5cf9e0b254 Retire SERIALIZE macros
Explicitly write the 'while' loops. This adds some
code but makes clear what's the code behind the
macro.

No functional change.
2014-03-09 11:10:33 +01:00
Marco Costalba 1d1b7df7c6 Rename castling flag to castling right
This is a more conventional naming as
reported also in:

http://chessprogramming.wikispaces.com/Castling+rights

No functional change.
2014-03-08 15:08:55 +01:00
Stephane Nicolet 13b9e1e098 Fix dbg_mean_of() for negative numbers
Type should be int64_t instead of uint64_t

No functional change.
2014-03-05 08:55:12 +01:00
mstembera 553ead429d Some minor cleanup stuff
I came across while browsing the code.

No functional change.
2014-03-03 08:57:20 +01:00
Marco Costalba bbd69c0260 Revert dynamic draw value
When tested with weaker engines did not
performed as expected, actually it was even
a regression from standard version.

bench: 8430785
2014-03-03 08:39:34 +01:00
Marco Costalba 708cb311a0 Pass CastlingFlag argument only
Instead of Color and CastlingSide. Change functions API
accordingly.

No functional change.
2014-03-02 13:21:19 +01:00
Marco Costalba 3e5470d88f Remove limit of minimumSplitDepth
There is no reason why an user cannot set
it at a value less than 4.

No functional change.
2014-03-01 23:22:14 +01:00
Marco Costalba 3d8c0f16c2 Rename xxx_to_char() -> to_char()
No functional change.
2014-03-01 22:07:41 +01:00
Marco Costalba de2ba70830 Simplify Bitboards::pretty
No functional change.
2014-03-01 13:10:05 +01:00
Marco Costalba 9f0485e343 Retire UCI_AnalyseMode option
It has been obsoleted out already some time ago
and currently there is no point in changing eval
score according to if we are in game or analyzing.

So retire the option.

No functional change.
2014-03-01 12:10:42 +01:00
Joerg Oster b917cd275e Dynamic draw value
Try to avoid repetition draws at early midgame,
this should give an edge against weaker opponents
and reduce draw rate.

Tested for regressions with SPRT[-3, 1] and
passed both short TC
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 68498 W: 12928 L: 12891 D: 42679

And long TC
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 40212 W: 6386 L: 6295 D: 27531

bench: 7990513
2014-02-26 19:33:52 +01:00
Marco Costalba 0949f06a60 Fix a warning with Intel compiler
warning #2259: non-pointer conversion from "int" to
"uint8_t={unsigned char}" may lose significant bits

No functional change
2014-02-22 12:00:14 +01:00
Marco Costalba 012f20d66e Fix an assert in Probcut
When running the following position:

8/kPp5/2P3p1/p1P1p1P1/2PpPp2/3p1p2/3P1P2/5K2 w - - 0 1

An assert is raised at depth 92:

assert(-VALUE_INFINITE <= alpha && alpha < beta && beta <= VALUE_INFINITE);

This is because it happens that beta = 29832,
so rbeta = 30032 that is > VALUE_INFINITE

Bug spotted and analyzed by Uri, fix suggested by Joerg.

Other fixes where possible but this one is pointed
exactly at the source of the bug, so it is the best
from a code documentation point of view.

bench: 8430785
2014-02-22 10:38:21 +01:00
Leonid Pechenik b8cfc255d4 Distribute part of first move time to other moves
Passed both short TC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 18907 W: 3475 L: 3322 D: 12110

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 19044 W: 2997 L: 2811 D: 13236

bench: 8430785
2014-02-20 08:39:00 +01:00
Marco Costalba 9fcefb2760 Avoid recalculating CheckInfo in generate_castling()
No functional change.
2014-02-20 08:27:13 +01:00
Marco Costalba 64a71c3c02 Don't update pieceCount for ALL_PIECES
It is currently unused and only adds
overhead for nothing.

Also set proper type of chess960.

No functional change.
2014-02-18 16:41:24 +01:00
Marco Costalba 7b0a2f2a90 Update SEE to return a Value
It seems more natural because the actual returned
value is from PieceValue[] array.

No functional change.
2014-02-16 13:06:31 +01:00
Marco Costalba 17ffc22279 Sync code style in material.cpp
Update to use common code style.

No functional change.
2014-02-16 12:56:54 +01:00
Marco Costalba 52ebf87238 Increase MAX_PLY from 100 to 120
Under some very rare case 100 plies of search
could be not enough. Increasing more could lead
to crashes due to reached stack size limit on
some platforms.

Strongly requested by Uri.

bench: 8430785
2014-02-16 12:20:37 +01:00
Marco Costalba d91079d4b0 Fix material key for King
Currently king has no material key associated because
it can never happen to find a legal position without
both kings, so there is no need to keep track of it.

The consequence is that a position with only the two
kings has material key set at zero and if the material
hash table is empty any entry will match and this is
wrong.

Normally bug is hidden becuase the checking for a draw
with pos.is_draw() is done earlier than evaluate() call,
so that we never check in gameplay the material key of a
position with two kings.

Nevertheless the bug is there and can be reproduced setting
at startup a position with only two kings and typing
'eval' from prompt.

The fix is very simple: add a random key also for the king.

Also fixed the condition in material.cpp to avoid asserting
when a 'just 2 kings' postion is evaluated.

No functional change.
2014-02-16 11:51:30 +01:00
Marco Costalba 62c0dc5dea Restore PorbCut name
Actually MultiCut is too different from current scheme.
Note that neither ProbCut is exactly what we do because
we try just a handful of captures instead of all moves,
nevertheless it seems more in line with what we do.

Suggested by Joona.

No functional change.
2014-02-15 22:21:39 +01:00
Marco Costalba 7bdb8c9c5c Reformat of eval tracing code
Also assorted rename while there.

No functional change.
2014-02-15 14:11:26 +01:00
Marco Costalba 3336963185 Rename ProbCut to Multicut
The teqnique used is actually MultiCut:

https://chessprogramming.wikispaces.com/Multi-Cut

And not ProbCut:

https://chessprogramming.wikispaces.com/ProbCut

No functional change.
2014-02-15 09:37:55 +01:00
Marco Costalba 4808d15a85 Assorted renaming in search
Inspired by DON.

No functional change.
2014-02-15 09:20:27 +01:00
Joerg Oster 67f88f5e3e Return static eval when reaching MAX_PLY
Makes more sense than returning a draw score. Tested
with reduced MAX_PLY = 30 and passed both short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 17434 W: 3345 L: 3194 D: 10895

And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 2610 W: 488 L: 373 D: 1749

With current limit of MAX_PLY = 100 the patch should not
introduce any measurable change, nevertheless is the correct
approach.

Idea of returning eval is from  Michel Van den Bergh.

bench: 8430785
2014-02-15 08:20:33 +01:00
Marco Costalba b3470d7ab1 Fix magic boosters conversion
Fix small overflow error while converting
magic boosters from right rotate to left rotate,
in particular booster 38 was converted to 4122
instead of the corrcet value 26.

Formula used was:

s1 = original & 63, s2 = (original >> 6) & 63;
new = (64 - s1) | ((64 - s2) << 6);

Instead of:

s1 = original & 63, s2 = (original >> 6) & 63;
new = ((64 - s1) & 63) | (((64 - s2) & 63) << 6);

This has no impact in number of cycles needed, but
just in the resultig number that yields to a rotate
amount bigger than 63.

Spotted by Ehsan Rashid.

No functional change.
2014-02-14 10:43:37 +01:00
Marco Costalba e4695f15bc Additional renaming from DON
Assorted renaming and triviality.

No functional change.
2014-02-14 09:42:50 +01:00
joergoster ffd6685f79 Fix a compiler warning
Latest master triggers a compiler warning due
to comparing int64_t to uint64_t.

notation.cpp: In Funktion »std::string pretty_pv(Position&, int, Value, int64_t, Move*)«:
notation.cpp:230:30: Warnung: Vergleich zwischen vorzeichenbehafteten und vorzeichenlosen Ganzzahlausdrücken [-Wsign-compare]

This patch should fix it.

No functional change.
2014-02-12 21:52:21 +01:00
Leonid Pechenik 72e8640f4d Simplify time management
Tested with simplification mode SPRT[-4, 0]

Passed both short TC
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 34102 W: 6184 L: 6144 D: 21774

And long TC
LLR: 2.96 (-2.94,2.94) [-4.00,0.00]
Total: 16518 W: 2647 L: 2545 D: 11326

And also 40/10 TC
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 22406 W: 4390 L: 4312 D: 13704

bench: 8430785
2014-02-12 20:01:11 +01:00
Marco Costalba e6523e56b8 Move magic random to RKISS
When initializing the magic numbers used to
compute sliding attacks, we endless generate a
random and test it as a possible magic.

In the general case this takes a lot of iterations,
but here, insteaad of picking a casual random, we
rotate it a couple of times and generate a number that
we know has a good probability to be a magic candidate.

This is becuase the quantities by which we rotate the
number are known in advance to produce quickly a good
canidate.

The patch, inspired by DON, just moves the shuffle to RKISS
changing the boosters to take in account a left rotation
instead of a right rotation as in the original.

No functional change.
2014-02-12 14:47:36 +01:00
Marco Costalba b534176d4a Revert "Retire null search verification"
Although does not change ELO level, it seems
verification is useful in many zugzwang positions
as reported by many sources.

So revert this simplification.

bench: 8430785
2014-02-12 14:16:21 +01:00
Henri Wiechers ddeb01612b Faster handling of king captures in Position::see
Another SEE speed up that passed the SPRT short TC test!

LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 81337 W: 15060 L: 14745 D: 51532

No functional change.
2014-02-09 19:42:24 +01:00
Marco Costalba 41641e3b1e Assorted tweaks from DON
Mainly renames and some little code style improvment,
inspired by looking at DON sources:

https://github.com/erashid/DON

No functional change.
2014-02-09 17:31:45 +01:00
Marco Costalba 2f5aaf7de6 Rewrite previous patch removing the macro
No functional change.
2014-02-08 13:27:44 +01:00
Thanar2 d48a304262 Inline common path of pos.gives_check
Test for common case which, when running default
stockfish bench, avoids 96% of 19 million calls to
pos.gives_check().

Yields to a 2-4% speed up according to platform.

Passed fishtest at STC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 12441 W: 2333 L: 2196 D: 7912
http://tests.stockfishchess.org/tests/view/52f02d790ebc5901df50f887

Passed fishtest at LTC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 42175 W: 6702 L: 6409 D: 29064
http://tests.stockfishchess.org/tests/view/52f0dbe00ebc5901df50f8a0

No functional change.
2014-02-08 13:17:29 +01:00
Marco Costalba 0a1092f64e Don't fear races when are harmless
Actually race conditions do exist in an engine, just
think for a moment to TT concurrent access. Racy code
is not a problem per se, if the consequences are well
known and correctly handled.

In case of TT access we ensure that the TT move is validated
before to be tried, here we just retry the same move in less
that 1 case out of a million: this is totally harmless considering
that very probably the second time the move is tried we get
immediately a TT hit and search quickly returns.

So we simplify the code for no harm.

No fuctional change (in single thread case)
2014-02-08 13:07:57 +01:00
Lucas Braesch 399968f1d0 Retire null search verification
Tested with SPRT in simplification mode [-4.00,0.00],
this ensures that the patch is (very probably) not
a regression.

Passed both short TC
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 27543 W: 4278 L: 4209 D: 19056

And long TC
LLR: 2.95 (-2.94,2.94) [-4.00,0.00]
Total: 39483 W: 7325 L: 7305 D: 24853

bench: 8347121
2014-02-08 08:25:08 +01:00
Lucas Braesch e5c3effdb1 Better document null search window
Hopefully this patch makes the code more:

* Self-documenting: Null search is always a zero window search,
  because it is testing for a fail high. It should never be done
  on a full window! The current code only works because we don't
  do it at PV nodes, and therefore (alpha, beta) = (beta-1, beta):
  that's the kind of "clever" trick we should avoid.

* Idiot-proof: If we want to enable null search at PV nodes, all we
  need to do now is comment out the !PvNode condition. It's that simple!

In theory, null search should not be done at PV nodes, because PV nodes
should never fail high. But in practice, they DO fail high, because of
aspiration windows, and search inconsistencies, for example. So it makes
sense to keep that flexibility in the code.

No functional change.
2014-02-04 08:24:46 +01:00
Lucas Braesch e88ef801af Better document razoring
Use ralpha instead of rbeta

* rbeta is confusing people. It took THREE attempts to code razoring
at PV nodes correctly in a recent test, because of the rbeta trick.
Unnecessary tricks should be avoided.

* The more correct and self-documenting way of doing this, is to say
that we use a zero window around alpha-margin, not beta-margin.
The fact that, because we only do it at PV nodes, alpha happens to be
beta-1 and that the current stuff with rbeta works, may be correct,
but is confusing.

Remove the misleading and partially erroneous comment about returning
v + margin:

* comments should explain what the code does, not what it could have done.

* this comment is partially wrong in saying that v+margin is "logical",
  and that it is "surprising" that is doesn't work.

From a theoretical perspective, at least 3 ways of doing this are equally
defendable:

1/ fail hard: return alpha: The most conservative. We bet that the search
will fail low, but we don't know by how much and don't want to take risks.

2/ aggressive fail soft: return v (what the current code does). This
corresponds to normal fail soft, with the added assumption that we don't
care about the reduction effect (see below point 3/)

3/ conservative fail soft: return v + margin. If the reduced search (qsearch)
gives us a score <= v, we bet that the non reduced search will give us a
score <= v + margin.

* Saying that 2/ is "logical" implies that 1/ and 3/ are not, which is
arguably wrong. Besides, experimental results tell us that 2/ beats 3/,
and that's not something we can argue against: experimental results are
the only trusted metric.

* Also, with the benefit of hindsight, I don't think the fact that 2/ is
better than 3/ is surprising at all. The point is that it is YOUR turn to
move, and you are assuming that by NOT playing (and letting the opponent
capture your hanging pieces in QS) you cannot generally GAIN razor_margin(depth).

No functional change.
2014-02-03 21:37:14 +01:00
Joona Kiiski d8796fe7e9 H-file penalty and center bonus
After almost 50K at long TC it seems
slightly positive:

ELO: 1.73 +-1.8 (95%) LOS: 97.3%
Total: 47122 W: 7507 L: 7273 D: 32342

bench: 8430785
2014-02-03 21:23:49 +01:00
Joona Kiiski 77341f67f3 Tweak bishop PSQT tables
Tuned after 49K iterations of SPSA.

Passed both short TC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 14231 W: 2684 L: 2542 D: 9005

And long TC:
LLR: 2.94 (-2.94,2.94) [0.00,4.00]
Total: 87556 W: 13757 L: 13342 D: 60457

bench: 6875743
2014-01-29 18:02:13 +01:00
Uri Blass 93f95cc936 Reduce VALUE_KNOWN_WIN to 10000
With some positions like

8/8/8/2p2K2/1pp5/br1p1b2/2p2r2/qqkqq3 w - -

The eval score is higher than VALUE_INFINITE because
is the sum of VALUE_KNOWN_WIN plus a big material
advantage. This leads to an assert. Here are the
steps to reproduce:

Compile SF with debug=yes then do

./stockfish
position fen 8/8/8/2p2K2/1pp5/br1p1b2/2p2r2/qqkqq3 w - -
go depth 1

This patch fixes the issue in this case, but do exsist
other positions for which the patch is not enough and
we will need to limit the eval score to be sure not
overflow the limit.

Note that is not possible to increase the value of
VALUE_INFINITE because should remain within int16_t
type to be stored in a TT entry.

bench: 7356053
2014-01-27 20:09:51 +01:00
Marco Costalba f434cea287 Fix null reduction formula
Depth is already dependent on the actual value
of ONE_PLY, in particular can be expressed like:

Depth = n * ONE_PLY

And because formula is used to calculate R that is
also dependent on the value of ONE_PLY and can be
expressed like:

R = x * ONE_PLY

We don't want to divide depth by a 'ply' value but
directly by an integer number.

Spotted by sf-x

No functional change.
2014-01-27 08:18:48 +01:00
Marco Costalba 216972186e Grammar fix in MovePicker::next_move
Thanks to Lyudmil Antonov and Michel Van den Bergh
for spotting this and suggesting the fix.

No functional change.
2014-01-26 23:09:22 +01:00
Stefan Geschwentner 074c7a3c30 Variable null-move value based reduction
Instead of a fixed reduction of ONE_PLY, now
Null move dynamic reduction based on value can
grow larger in case we are above beta of a value
much higher then PawnValueMg.

Note that now an eval returning VALUE_KNOWN_WIN
makes null search to drop in qsearch.

Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 26141 W: 4871 L: 4699 D: 16571

And long TC:
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 33695 W: 5309 L: 5056 D: 23330

bench: 7356053
2014-01-26 10:23:31 +01:00
Joona Kiiski cf95a55d76 Do not set default value for architeture in Makefile
Fixes a regression that ARCH parameter was not properly validated.
Invalid value would default to generic 32-bit build.

No functional change.
2014-01-25 11:29:32 +01:00
Henri Wiechers a08a21d5a0 Small simplification to Position::see
Verified there are no hidden bugs and is
actually a speed optimization:

Fixed games at 15+0.05 TC
ELO: 1.72 +-2.9 (95%) LOS: 87.5%
Total: 20000 W: 3741 L: 3642 D: 12617

No functional change
2014-01-19 11:16:34 +01:00
Joona Kiiski cabd512916 Fix +M0 score when low on time
When time remaining is less than Emergency Move Time,
we won't even complete one iteration and engine reports
a stale +M0 score.

To reproduce run "go wtime 10"

info depth 1 seldepth 1 score mate 0 upperbound nodes 2 nps 500 time 4 multipv 1 pv a2a3
info nodes 2 time 4
bestmove a2a3 ponder (none)

This patch fixes the issue.

Tested by Binky at very short TC: 0.05+0.05
ELO: 5.96 +-12.9 (95%) LOS: 81.7%
Total: 1458 W: 394 L: 369 D: 695

And at a bit longer TC:
ELO: 1.56 +-3.7 (95%) LOS: 79.8%
Total: 16511 W: 3983 L: 3909 D: 8619

bench: 7804908
2014-01-19 11:09:44 +01:00
Marco Costalba 40c863d41a Increase max hash size to 16GB
TCEC season 3, which is due to start in a few weeks, just
had its server upgraded to 64GB RAM and will therefore allow
16GB hash to be used per engine.

This is almost the upper limit without changing the
type of size and hashMask. After this we need to
move to uint64_t instead of uint32_t.

No functional change.
2014-01-18 18:22:32 +01:00
Joona Kiiski be61edbd19 Tweak knight PSQT tables
Passed both short TC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 31765 W: 6103 L: 5913 D: 19749

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 38867 W: 6268 L: 5988 D: 26611

bench: 7804908
2014-01-18 17:52:37 +01:00
Chris Cain df201175c6 Simplify pawnless endgame evaluation
Retire KmmKm evaluation function. Instead give a very drawish
scale factor when the material advantage is small and not much
material remains.

Retire NoPawnsSF array. Pawnless endgames without a bishop will
now be scored higher. Pawnless endgames with a bishop pair will
be scored lower. The effect of this is hopefully small.

Consistent results both at short TC (fixed games):
ELO: -0.00 +-2.1 (95%) LOS: 50.0%
Total: 40000 W: 7405 L: 7405 D: 25190

And long TC (fixed games):
ELO: 0.77 +-1.9 (95%) LOS: 78.7%
Total: 39690 W: 6179 L: 6091 D: 27420

bench: 7213723
2014-01-18 17:22:54 +01:00
Stefan Geschwentner 53ab32ef0b Introduce 'follow up' moves
When we have a fail-high of a quiet move, store it in
a Followupmoves table indexed by the previous move of
the same color (instead of immediate previous move as
is in countermoves case).

Then use this table for quiet moves ordering in the same
way we are already doing with countermoves.

These followup moves will be tried just after countermoves
and before remaining quiet moves.

Passed both short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10350 W: 1998 L: 1866 D: 6486

And long TC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 14066 W: 2303 L: 2137 D: 9626

bench: 7205153
2014-01-14 09:24:35 +01:00
Ralph Stößer 402a7ae151 Ad-hoc shelter rule
This hacky rule allows to get an about right eval out of this position:
r2qk2r/ppp2p2/2npbn2/2b1p3/2P1P1P1/2NB1PPp/PPNP3K/R1BQ1R2 b kq - 0 13

And, more importantly, passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 6239 W: 1249 L: 1127 D: 3863

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 38371 W: 6165 L: 5888 D: 26318

bench: 8183238
2014-01-12 22:48:08 +01:00
Marco Costalba 3d58092ec1 Retire KBBKN endgame
As pointed out by Joona, Lucas and otehr people in
the forum, this endgame is not a known, there are many
positions where it takes more than 50 moves to claim the
win and becasue exact rules is not possible better to
retire and allow the search to workout the endgame for us.

bench: 8502826
2014-01-12 22:32:41 +01:00
Henri Wiechers db05b1f9f5 Rename Position::hidden_checkers to check_blockers
No functional change.
2014-01-12 04:10:46 +09:00
Henri Wiechers 262c380c4b Position::gives_check - use ci.ksq
Also remove a couple of local variables while
there.

No functional change.
2014-01-09 07:38:04 +09:00
Marco Costalba 6a6fd0b5f5 Fix early stop condition
While editing original Uri's messy patch
I have incorrectly simplified the logic
condition. Here is the correct original
version, as it was tested.

bench: 8502826
2014-01-09 06:58:25 +09:00
Ralph Stoesser f14cd1bb89 Retire easy move
Remove the easy move code and add the condition to
play instantly if only one legal move is available.

Verified there is no regression at 60+0.05
ELO: 0.17 +-1.9 (95%) LOS: 57.0%
Total: 40000 W: 6397 L: 6377 D: 27226

bench: 8502826
2014-01-08 23:57:06 +09:00
Uri Blass a5869d8d25 Stop earlier if iteration is taking too long
If we are still at first move, without a fail-low and
current iteration is taking too long to complete then
stop the search.

Passed short TC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 26030 W: 4959 L: 4785 D: 16286

Long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 18019 W: 2936 L: 2752 D: 12331

And performed well at 40/30
ELO: 4.33 +-2.8 (95%) LOS: 99.9%
Total: 20000 W: 3480 L: 3231 D: 13289

bench: 8502826
2014-01-08 23:45:55 +09:00
renouve 45dbd9cd03 Retire grain size code
Seems useless at long TC.

Tested at 60+0.05
ELO: 2.98 +-2.2 (95%) LOS: 99.6%
Total: 30440 W: 4934 L: 4673 D: 20833

And at 120+0.05
ELO: 2.50 +-2.6 (95%) LOS: 97.1%
Total: 19633 W: 2848 L: 2707 D: 14078

bench: 8502826
2014-01-08 23:22:17 +09:00
Marco Costalba a646f74e6a Assorted grammar fixes
Mainly from Lyudmil Antonov and
one from Henri Wiechers and Louis Zulli.

No functional change.
2014-01-07 14:30:41 +09:00
Joona Kiiski 1fee23a598 Tweak King PST tables
First tested with 50K games at very short TC of 5+0.05
ELO: 3.11 +-2.0 (95%) LOS: 99.9%
Total: 49665 W: 10941 L: 10497 D: 28227

Then retested with usual SPRT at short TC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 16875 W: 3198 L: 3049 D: 10628

And at long TC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 5890 W: 985 L: 857 D: 4048

bench: 7800379
2014-01-07 13:41:16 +09:00
Marco Costalba 1574428f64 Fix a typo
Spotted by Isaac H. Dunn.

No functional change.
2014-01-06 01:22:44 +01:00
Pablo Vazquez 0118623495 Remove duplicated code
Introduce update_stats() and remove correspondng
duplicated code.

No functional change.
2014-01-05 14:10:29 +01:00
H. Felix Wittmann 8454d871ec Ensure move_importance() is non-zero
In case ply is very high, function will round
to zero (although mathematically it is always
bigger than zero). On my system this happens at
movenumber 6661.

Although 6661 moves in a game is, of course,
probably impossible, for safety and to be locally
consistent makes sense to ensure returned value
is positive.

Non functional change.
2014-01-02 13:01:24 +01:00
shane31 153309e287 Scale eval when down to only one pawn
Passed both short TC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 11921 W: 2346 L: 2208 D: 7367

And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 21002 W: 3395 L: 3197 D: 14410

bench: 7602383
2014-01-02 12:44:46 +01:00
Marco Costalba c9dcda6ac4 Update copyright year
No functional change.
2014-01-02 01:49:18 +01:00
Marco Costalba 29e0d8caa7 Simplify move_importance(): take 3
Use pow() of a negative number instead of 1/x

No functional change.
2014-01-01 13:43:58 +01:00
Marco Costalba a795187c50 Further simplify move_importance()
Function move_importance() is already always
positive, so we don't need to add a constant
term to ensure it.

Becuase move_importance() is used to calculate
ratios of a linear combination (as explained in
previous patch), result is not affected. I have
also verified it directly.

No functional change.
2014-01-01 13:35:11 +01:00
H. Felix Wittmann 92faa74dfa Simplify move_importance()
Drop a useless parameter. This works because ratio1 and ratio2
are ratios of linear combinations of thisMoveImportance and
otherMovesImportance and so the yscale cancels out.

Therefore the values of ratio1 and ratio2 are independent
of yscale and yscale can be retired.

The same applies to yshift, but here we want to ensure
move_importance() > 0, so directly hard-code this safety
guard in function definition.

Actually there are some small differences due to rounding errors
and usually are at most few millisecond, that's means below 1% of
returned time, apart from very short intervals in which a difference
of just 1 msec can raise to 2-3% of total available time.

No functional change.
2014-01-01 12:58:10 +01:00
Marco Costalba c5d478b923 Rename pawn chain to connected
The flag raises also in case of a pawn duo, i.e.
when we have two adjacent pawns on the same rank,
and not only in case of a chain, i.e. when the two
pawns are on a diagonal line.

See this for a reference:
http://en.wikipedia.org/wiki/Connected_pawns

Renaming suggested by Ralph.

No functional change.
2014-01-01 10:56:57 +01:00
Gary Linscott 9b1d594456 Remove bishop pin bonus
Shows no regression at LTC after 20K games:

ELO: 0.03 +-2.7 (95%) LOS: 51.0%
Total: 20608 W: 3252 L: 3250 D: 14106

bench: 7516178
2013-12-31 15:27:52 +01:00
Arjun Temurnikar 71440cf77b Retire KingExposed[] array
And merge its values into KPSQT table.

Passed blazingly fast both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 5348 W: 1091 L: 971 D: 3286

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 3029 W: 530 L: 415 D: 2084

bench: 8702197
2013-12-31 12:05:22 +01:00
Henri Wiechers 082fbf4264 Remove asymmThreshold stale comment
No functional change.
2013-12-30 10:34:20 +01:00
Marco Costalba f7742669cb Retire asymmThreshold
Verified with 40K games at long TC does not regress:
ELO: 1.74 +-1.9 (95%) LOS: 96.2%
Total: 39624 W: 6402 L: 6203 D: 27019

bench: 7762310
2013-12-30 01:27:57 +01:00
Matt Sullivan 59a9bc9351 Retire MoveImportance[]
Use a skew-logistic function to replace the
MoveImportance[] array.

Verified it does not regress at fixed number
of games both at short TC:
LLR: -2.91 (-2.94,2.94) [-1.50,4.50]
Total: 39457 W: 7539 L: 7538 D: 24380

And long TC:
ELO: -0.49 +-1.9 (95%) LOS: 31.0%
Total: 39358 W: 6135 L: 6190 D: 27033

bench: 7335588
2013-12-29 10:33:39 +01:00
Stefan Geschwentner 8ca80cb0f1 Fine tune previous patch
Passed short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 18331 W: 3608 L: 3453 D: 11270

And scored above 50% on a very long test in long TC
LLR: -2.97 (-2.94,2.94) [0.00,6.00]
Total: 51533 W: 8181 L: 8047 D: 35305

bench: 7335588
2013-12-27 18:46:49 +01:00
Marco Costalba 14aebe2b7c Further simplify previous patch
Use a single XOR instead of NEGATE + AND

No functional change.
2013-12-26 12:08:23 +01:00
Stefan Geschwentner cf0a2a26a9 Bonus for file distance of the outermost pawns
In endgame it's better to have pawns on both wings.
So give a bonus according to file distance between left
and right outermost pawns.

Passed both short TC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 39073 W: 7749 L: 7536 D: 23788

And long TC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6149 W: 1040 L: 910 D: 4199

bench: 7665034
2013-12-25 21:31:41 +01:00
Ralph Stößer 899a2c033e Loosened trigger condition for king safety
Reduce eval discontinuity becuase now we kick in
king safety evaluation in many more cases.

Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 8708 W: 1742 L: 1613 D: 5353

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 6743 W: 1122 L: 990 D: 4631

bench: 6835416
2013-12-23 20:55:30 +01:00
Chris Caino 8aa4f3fde4 Increase pawn king attack weight
Tighter lower bound for pawn attacks so to
activate king safety in some cases like here:

6k1/2B3p1/2Pp1p2/2nPp3/2Q1P2K/P2n1qP1/R6P/1R6 w

Original patch by Chris, further simplified by
Jörg Oster.

Passed both short TC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 30171 W: 5887 L: 5700 D: 18584

And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 20706 W: 3402 L: 3204 D: 14100

bench: 7607562
2013-12-23 20:52:12 +01:00
Gary Linscott 26689d8c2a Faster and simplified threat eval
Add a bonus according if the attacking
pieces are minor or major.

Passed both short TC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 13142 W: 2625 L: 2483 D: 8034

And long TC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 18059 W: 3031 L: 2844 D: 12184

bench: 7425809
2013-12-19 18:52:34 +01:00
Marco Costalba f196c1bad8 Further simplify Makefile
No functional change.
2013-12-17 10:16:00 +01:00
Marco Costalba e460ab74ad Reformat Makefile
No functional change.
2013-12-17 10:02:14 +01:00
Lucas Braesch f5727deee3 Remove threat move stuff
A great simplification that shows no regression
and it seems even a bit scalable.

Tested with fixed number of games:

Short TC
ELO: 0.60 +-2.1 (95%) LOS: 71.1%
Total: 39554 W: 7477 L: 7409 D: 24668

Long TC
ELO: 2.97 +-2.0 (95%) LOS: 99.8%
Total: 36424 W: 5894 L: 5583 D: 24947

bench: 8184352
2013-12-15 09:43:29 +01:00
Marco Costalba b96079f86b Sync history and counter moves updating
Change updating rule after a TT hit to match
the same one at the end of the search.

Small change in functionality, but we want to
have uniform rules in the code.

bench: 7767864
2013-12-10 07:05:06 +01:00
Lucas Braesch 86347100a5 Update History and Counter move on TT hit
We already update killers so it is natural to extend to
history and counter move too.

Passed both short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 52690 W: 9955 L: 9712 D: 33023

And long TC
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 5555 W: 935 L: 808 D: 3812

bench: 7876473
2013-12-10 06:57:06 +01:00
Ralph Stößer 8e9d4081ee Research at intermediate depth if LMR is very high
After a fail high in LMR, if reduction is very high do
a research at lower depth before teh full depth one.
Chances are that the re-search will fail low and the
full depth one is skipped.

Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 11363 W: 2204 L: 2069 D: 7090

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 7292 W: 1195 L: 1061 D: 5036

bench: 7869223
2013-12-09 08:03:47 +01:00
Marco Costalba 4630ab5743 More work on Bitboards::init()
No functional change.
2013-12-07 12:32:18 +01:00
Marco Costalba 1ff5ce8863 More readable init of MS1BTable[]
Because now it uses lsb(), the BSFTable[] must be
already initialized.

No functional change.
2013-12-07 11:35:26 +01:00
Marco Costalba b71418defa Further simplify Bitboards init()
No functional change.
2013-12-07 10:57:05 +01:00
Marco Costalba 47b89f2e37 Clarify definition of capture_or_promotion()
No functional change.
2013-12-06 10:43:17 +01:00
Arjun Temurnikar 431c3ac485 Even more spelling fixes
No functional change.
2013-12-06 09:03:24 +01:00
Arjun Temurnikar 190aea4cdc Assorted spelling/grammar/captitalization
No functional change.
2013-12-04 23:30:37 +01:00
Chris Caino f026517e5e Micro-optimise dangerous condition
Since all ENPASSANT moves are now considered dangerous, this
change of order should give a slight speedup.

Also simplify futilityValue formula.

No functional change.
2013-12-04 17:51:25 +01:00
Marco Costalba b71cedb2b0 Retire TheirHalf[]
We avoid to use an ad-hoc table at the cost of a
relative_rank() call in advanced_pawn_push().

On my 32 bit system it is even slightly faster (on 64bit
may be different). This is the speed in nps alternating
old and new bench runs:

new

368890
368825
369972

old

367798
367635
368026

No functional change.
2013-12-04 17:45:09 +01:00
Chris Caino 69a14554ee Broader condition for dangerous pawn moves
Instead of a passed pawn now we just require the pawn to
be in the opponent camp to be considered a dangerous
move. Added some renaming to reflect the change.

Passed both short TC test
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10358 W: 2033 L: 1900 D: 6425

And long TC
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 21459 W: 3486 L: 3286 D: 14687

bench: 8322172
2013-12-04 17:19:45 +01:00
Marco Costalba 2408243cf4 Shrink Position::is_draw()
No functional change.
2013-12-03 11:40:31 +01:00
Marco Costalba 342fd6385b Remove redundant argument from hidden_checkers()
No functional change.
2013-12-03 11:15:39 +01:00
Marco Costalba e6c9ce6358 Small improvment to Position::fen()
No functional change.
2013-12-03 10:53:21 +01:00
Jerry Donald c7e7d9217b Re-fix a comment
No functional change.
2013-12-03 08:50:12 +01:00
Jerry Donald a8af78c833 Another round of spelling fixes
And also renamed a loop variable while there.

No functional change.
2013-12-02 23:51:29 +01:00
Richard Lloyd 13a73f67c0 Big assorted spelling fixes
No functional change.
2013-12-02 20:29:35 +01:00
Jerry Donald 500b9b0eb3 Assorted spelling fixes
No functional change.
2013-12-02 18:41:30 +01:00
Marco Costalba f99cb3dc27 Rename CASTLE to CASTLING
It is call 'castling move', not 'castle move'

Noticed while reading DiscoCheck sources.

No functional change.
2013-12-01 11:16:47 +01:00
Marco Costalba dd4e5db2be Simplify a condition in gives_check()
Now that aligned() is quite fast we can skip
some logic.

No functional change.
2013-12-01 10:09:30 +01:00
Marco Costalba 5f2bf91ad1 Rename Bitboards print to pretty
To align to same named Position function and
avoid using std::cout directly.

Also remove some stale <iostream> include while
there.

No functional change.
2013-11-30 11:32:49 +01:00
Marco Costalba 034a2b04f2 Rewrite some bitboard init code
And move the static function Position::attacks_from() to
bitboard code renaming it attacks_bb()

No functional change.
2013-11-30 11:02:56 +01:00
Daylen Yang 6ea5dc294c Makefile improvements for compiling on OS X
Add a Mac SSE4.2 target. Also change the Mac OS X minimum version to
10.6. Rationale: 97% of Macs run at least 10.6, version 10.9 is now
free, and using 10.6 as the minimum version gives a small 5% boost in
benchmark speed over versions using 10.0 as the minimum version.

Finally, enable Clang’s Link Time Optimization when compiling for the
Mac.

No functional change.
2013-11-30 00:09:17 +01:00
Marco Costalba 8f5deaea61 Restore development version
bench: 8596156
2013-11-29 10:50:43 +01:00
Marco Costalba c5bb9b9da9 Stockfish DD
Stockfish bench signature is: 8596156
2013-11-29 10:23:14 +01:00
Kelly Wilson 67e5581e37 Add support for PPC 64bit on Linux
In particular Debian Linux-3.9.8-1- PPC64

No functional change.
2013-11-29 10:06:54 +01:00
Joona Kiiski 3ede7daab3 Generate Qsearch checks only at depth 0
An old idea retested at SPRT(0, 3) with 60+0.05 TC:
LLR: 2.95 (-2.94,2.94) [0.00,3.00]
Total: 98872 W: 15549 L: 15123 D: 68200

This is a very small elo increase patch so it really
stresses the limits of fishtest.

bench: 8596156
2013-11-24 10:26:49 +01:00
Marco Costalba dbd6156fce Revert previous fix
It seems to intorduce a regression when tested
with 3 threads at 15+0.05:

ELO: -2.26 +-2.2 (95%) LOS: 2.4%
Total: 30000 W: 4813 L: 5008 D: 20179

bench: 8331357
2013-11-19 07:20:50 +01:00
Hongzhi Cheng 691ed425ba Get correct excluded moves for split nodes
Tested setting FakeSplit to true and running

./stockfish bench 128 2

There is a different signature with and without
the patch so it affects functionality but
only in SMP case.

bench: 8331357
2013-11-18 16:41:49 +01:00
Marco Costalba c376ffce0f Revert previous patch
It seems a regression at 15+0.05:
ELO: -4.82 +-2.1 (95%) LOS: 0.0%
Total: 40000 W: 7181 L: 7736 D: 25083

bench: 8331357
2013-11-17 23:47:18 +01:00
Marco Costalba 917944e9c5 Fix an assert in SMP case
SMP case is very tricky and raises an assert in stage_moves():

assert(stage == KILLERS_S1 || stage == QUIETS_1_S1 || stage == QUIETS_2_S1)

So rewrite the code to just return moves[] when we are sure
we are in quiet moves stages.

Also rename stage_moves to quiet_moves to reflect that.

No functional change (but needs testing in SMP case)
2013-11-17 10:24:25 +01:00
Marco Costalba d9c7cad630 Retire quietsSearched[]
Use MovePicker moves[] to access already tried
quiet moves. A bit of care shall be taken
to avoid calling stage_moves() when we are still
at ttMove stage, because moves are yet to be
generated. Actually our staging move generation
makes this code a bit more tricky than what I'd
like, but removing an ausiliary redundant
array like quietsSearched[] is a good thing.

Idea by DiscoCheck

bench: 9355734
2013-11-17 09:51:04 +01:00
Marco Costalba 9763c69fa5 Simplify generate<EVASIONS>
Use the newly introduced LineBB[] to simplify this
super hot-path function.

Verified with perft we don't have any speed regression, although
the number of squares removed is less than before in case of
contact check.

Insipred by DiscoCheck implementation.

Perft numbers are the same, but we have an harmless functional
change due to reorder of moves, because now some illegal moves
are no more detected at generation time, but in the search.

bench: 8331357
2013-11-11 19:53:19 +01:00
Gregor Cramer 555d9a8711 Faster castling in Chess960 case
Only rook attackers has to be considered, all other attackers are
already handled in the lines above.

No functional change.
2013-11-11 15:55:08 +01:00
Joona Kiiski b9768b8bc5 Reintroduce gains
This seems a die hard idea :-)

Passed both short TC
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 17485 W: 3307 L: 3156 D: 11022

And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 38181 W: 6002 L: 5729 D: 26450

bench: 8659830
2013-11-11 08:46:11 +01:00
Jörg Oster 4ef6b2c32a Remove opposed flag for doubled pawns
Actually, it is not used, as both arrays have the
same values. Some local tests in either direction
showed no improvement.

Also some minor corrections in the comments.

No functional change.
2013-11-10 17:21:59 +01:00
Marco Costalba 1d18647e73 Rename squares_aligned()
Rename to the shorter but still
clear aligned()

No functional change.
2013-11-10 17:14:46 +01:00
Marco Costalba a518d5d3ad Simplify squares_aligned()
Use newly introduced LineBB[]

No functional change.
2013-11-10 12:05:19 +01:00
Chris Caino 091aff0445 Evaluate mobility of pinned pieces exactly
Previously some squares could be "incorrectly" awarded
to a pinned piece.

e.g. in 3k4/1q6/3b4/3Q4/8/5K2/B7/8 b - - 0 1 the black
bishop get 4 squares too many and the white queen gets 6.

Passed both short TC.
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 4871 W: 934 L: 817 D: 3120

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 38968 W: 6113 L: 5837 D: 27018

bench: 9282549
2013-11-10 11:52:38 +01:00
Chris Caino 3ed86ed3f9 Remove RedundantMajor
But compensate by reducing rook and queen
value by 53 = (160 / 3)

Material imbalances are affected as follows:

       Red. Major   Rook  Queen  Total
QRR      +160      -2*53    -53     +1
QR       +160        -53    -53    +54
RR       +160      -2*53      0    +54
R           0        -53      0    -53
Q           0          0    -53    -53

so that the imbalance changes by at most 54 + 53 = 107 units.
This corresponds to appromximately 3.5cp in the final evaluation.

Verified with fixed number 40000 games at both short
and long TC it does not regress.

Short TC 15+0.05
ELO: 1.93 +-2.1 (95%) LOS: 96.6%
Total: 40000 W: 7520 L: 7298 D: 25182

Long TC 60+0.05
ELO: -0.33 +-1.9 (95%) LOS: 36.5%
Total: 39663 W: 6067 L: 6105 D: 27491

bench: 6703846
2013-11-10 09:48:49 +01:00
Marco Costalba e4d34e1815 Fix printing of incorrect PV in some cases
As, Gary (that analyzed the bug) says:

SF does not print a PV when the original best move fails low,
we hit our time allowance, and stop the search.  The output from
the SF search is below.  It was failing low on Ne1 at depth 34.
Then, we get bestmove Qd3, but no PV change.

info depth 34 seldepth 45 score cp 38 upperbound nodes 483484489 nps 15464575 time 31264 multipv 1 pv f3e1 h5h4 e1d3 h4g3 f2g3 a6f6 f1f6 e7f6 d1a4 f6e7 a1f1 d8f8 a4b3 b7b6 b3c2 f7f6 c2a4 h3g5 b2b3 g5f7 a4c6 f7d6 h1g2 f6f5 e4f5 d6f5
info depth 34 seldepth 45 score cp 38 upperbound nodes 483484489 nps 15464575 time 31264 multipv 1 pv f3e1 h5h4 e1d3 h4g3 f2g3 a6f6 f1f6 e7f6 d1a4 f6e7 a1f1 d8f8 a4b3 b7b6 b3c2 f7f6 c2a4 h3g5 b2b3 g5f7 a4c6 f7d6 h1g2 f6f5 e4f5 d6f5
info depth 34 seldepth 47 score cp 30 upperbound nodes 2112334132 nps 17255517 time 122415 multipv 1 pv f3e1 h5h4 d1a4 a6f6 e1d3 d8f8 a4c2 h4g3 f2g3 f6f1 a1f1 h7g8 b2b3 f7f6 a3a4 b7b6
info depth 34 seldepth 47 score cp 30 upperbound nodes 2112334132 nps 17255517 time 122415 multipv 1 pv f3e1 h5h4 d1a4 a6f6 e1d3 d8f8 a4c2 h4g3 f2g3 f6f1 a1f1 h7g8 b2b3 f7f6 a3a4 b7b6
info nodes 18235667001 time 969824
bestmove e2d3 ponder c8d7

Looking at the code, if we hit Signals.stop, we return from id_loop
before printing any PV.  It is possible for us to have resorted the
RootMove list though, which will change the move that is actually
played.

No functional change.
2013-11-09 19:05:43 +01:00
Marco Costalba 42caebfaa5 Fix compile in debug mode
No functional change.
2013-11-09 18:41:51 +01:00
Lucas Braesch eed508b444 Futility pruning simplification
1/ eval margin and gains removed:
16bit are now free on TT entries, due to the removal of eval margin. may be useful
in the future :) gains removed: use instead by Value(128). search() and qsearch()
are now consistent in this regard.

2/ futility_margin()
linear formula instead of complex (log(depth), movecount) formula.

3/ unify pre & post futility pruning
pre futility pruning used depth < 7 plies, while post futility pruning used
depth < 4 plies. Now it's always depth < 7.

Tested with fixed number of games both at short TC:
ELO: 0.82 +-2.1 (95%) LOS: 77.3%
Total: 40000 W: 7939 L: 7845 D: 24216

And long TC
ELO: 0.59 +-2.0 (95%) LOS: 71.9%
Total: 40000 W: 6876 L: 6808 D: 26316

bench 7243575
2013-11-09 10:17:27 +01:00
Marco Costalba 343544f3f7 Revert "Retire eval margin and gains"
This reverts commit ecd07e51d0.

Patch was incorrect and partial. It will be reapplied in
the correct form.

bench: 9189063
2013-11-07 22:32:13 +01:00
Gary Linscott 13d1f0ae43 Restrict mobility of pinned pieces
Passed both short TC:
LLR: 3.00 (-2.94,2.94) [-1.50,4.50]
Total: 54342 W: 10950 L: 10692 D: 32700

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 61976 W: 10654 L: 10251 D: 41071

This patch introduces a slowdown of 3.5 % !!!!!

bench: 7911558
2013-11-07 22:26:03 +01:00
Lucas Braesch ecd07e51d0 Retire eval margin and gains
1/ eval margin and gains removed:
 - gains removed by Value(128): search() and qsearch() now behave consistently!

2/ futility_margin()
 - testing showed that there is no added value in this weird (log(depth), movecount)
   formula, and a much simpler linear formula is just as good. In fact, it is most
   likely better, as it is not yet optimally tuned.
 - the new simplified formula also means we get rid of FutilityMargins[], its
   initialization code, and more importantly ss->futilityMoveCount, and the hacky
   code that updates it throughout the search().
 - the current formula gives negative futility margins, and there is a hidden interaction
   between the move coutn pruning formula and the futility margin one: what happens is
   that MCP is supposed to be triggered before we use the non-sensical negative futility
   margins.

3/ unify pre & post futility pruning
 - pre futility pruning (what SF calls value based pruning) used depth < 7 plies,
   while post futility pruning (what SF calls static null move pruning) used depth < 4 plies.
 - also the condition depth < 7 in pre futility pruning was not obvious, and it seemd
   to be depth < 16 (futility_margin() returns an infinite value when depth >= 7).

Tested with fixed number of games both at short TC:
ELO: 0.82 +-2.1 (95%) LOS: 77.3%
Total: 40000 W: 7939 L: 7845 D: 24216

And long TC
ELO: 0.59 +-2.0 (95%) LOS: 71.9%
Total: 40000 W: 6876 L: 6808 D: 26316

bench: 10206576
2013-11-07 19:46:51 +01:00
Chris Caino 52ae0efccf Two more parameters eliminated
RedundantRook and RedundantQueen replaced by simple
variable RedundantMajor. Also the SameColor coefficient
for Queen<->Queen has been set by definition to 0.

The remaining 5 parameters:

LinearCoefficients[ROOK]
LinearCoefficients[QUEEN]
QuadraticCoefficientsSameColor[ROOK][ROOK]
QuadraticCoefficientsSameColor[QUEEN][ROOK]
RedundantMajor

are sufficient to equate the material imbalances for the
5 common material configurations of R, RR, Q, QR and QRR
to any desired values simultaneously.

With the chosen parameters there should be no functional
change unless one side has more than 2 rooks or more
than 1 queen. For example bench from the start position
using the commands:

./stockfish
go depth 16

produces identical output except for one extra node
in the last iteration.

bench: 8198094
2013-11-07 19:20:24 +01:00
Chris Caino 53c04c0429 Zero more redundant coefficients
Coefficients for Bishop<->BishopPair and Bishop<->Bishop
are also pretty much redundant. By altering the values
in LinearCoefficients[] these coefficients can be zeroed
without changing the imbalance calculations in any position
with less than 3 bishops for one side.

bench: 7995098
2013-11-05 20:08:39 +01:00
Chris Caino 1064288b38 Zero redundant material imbalance terms
First coefficient in the SameColor array does an
equivalent job when folded into the LinearCoefficients
array.

All of the diagonal terms in the OppositeColor array
are redundant due to cancellation.

No functional change.
2013-11-05 20:08:39 +01:00
Joona Kiiski d34bb889b1 Test Easy Move if no BestMoveChanges
In case we find a very good move after a
troubled start, we don't return immediately
anymore.

Tested directly at long TC where it passed:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 13910 W: 2397 L: 2228 D: 9285

bench: 7995098
2013-11-02 11:34:42 +01:00
Marco Costalba a3a0df92a3 Set timer to a fixed interval
And remove a complex (and broken) formula.

Indeed previous code was broken in case of TC with big
time increments where available_time() was too similar
to total time yielding to many time losses, so for instance:

go wtime 2600 winc 2600
info nodes 4432770 time 2601 <-- time forfeit!

maximum search time = 2530 ms
available_time = 2300 ms

For a reference and further details see:

https://groups.google.com/forum/?fromgroups=#!topic/fishcooking/dCPAvQDcm2E

Speed tested with bench disabling timer alltogheter vs timer set at
max resolution, showed we have no speed regressions both in single
core and when using all physical cores.

No functional change.
2013-11-01 08:56:15 +01:00
Ralph Ster e8f9447b11 Use a formula for chain membership bonus
Passed both short TC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5087 W: 1072 L: 951 D: 3064

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 28620 W: 5042 L: 4798 D: 18780

bench: 7995098
2013-10-31 06:13:30 +01:00
Marco Costalba 3cc47edf62 Tweak bishop pair and knight weight
A combo of two patches that failed SPRT with score
higher than 50% but togheter they succeed:

SPRT at 60+0.05
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 7312 W: 1276 L: 1139 D: 4897

bench: 8029334
2013-10-28 19:31:25 +01:00
Matthew Sullivan d454cd4216 Fix divide by zero bug in late game
If the game got late enough that move_importance(currentPly) * slowMover / 100
rounds to 0, then we ended up dividing 0 by 0 when only looking 1 move ahead.

This apparently caused the search to almost immediately abort and Stockfish
would blunder in long games. So convert thisMoveImportance to a double.

No functional change.
2013-10-27 08:03:58 +01:00
Marco Costalba 48f38f3092 Retire mirror()
Inline the only caller site.

No functional change.
2013-10-24 20:40:26 +02:00
Marco Costalba 281472e50e Prefer file_bb() to FileBB[]
No functional change.
2013-10-24 20:34:23 +02:00
Jörg Oster f011a5af11 Penalty for Knight when enemy pawns are few
This seems more a material imbalance topic,
anyhow test is good and so patch is applied
as is.

Passed both short TC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 17391 W: 3548 L: 3393 D: 10450

And long TC:
LLR: 3.00 (-2.94,2.94) [0.00,6.00]
Total: 34660 W: 5972 L: 5700 D: 22988

bench: 8291883
2013-10-24 20:11:33 +02:00
Marco Costalba 67b0da83da Further smplify pawn endgames
Dumb down a bit the code and trade some possible
speed (but this is far from hot path anyhow) for
some added readability for the layman.

No functional change.
2013-10-23 19:56:35 +02:00
Chris Caino 3674f18b97 Use flip_sq idea in endgame.cpp
The normalising transformation is computed all at
once by the helper function get_flip_sq and then
applied immediately to the relevant squares as soon
as they are loaded from the position class.

bench: 8350690
2013-10-23 16:00:49 +02:00
Chris Caino 72f7282ad4 Simplify futility move count formula
Simpler formula but introduces some slight changes if d >= 10

Original code grows like  0.225 * d^1.8
New code grows like       0.222 * d^1.8

Full list of values:

d old new diff
--------------
0 2 2 0
1 2 2 0
2 3 3 0
3 4 4 0
4 5 5 0
5 6 6 0
6 7 7 0
7 9 9 0
8 11 11 0
9 13 13 0
10 15 16 1
11 18 19 1
12 21 21 0
13 24 24 0
14 27 28 1
15 31 31 0
16 35 35 0
17 39 38 -1
18 42 42 0
19 47 46 -1
20 51 51 0
21 55 55 0
22 60 60 0
23 65 65 0
24 70 70 0
25 75 75 0
26 81 80 -1
27 87 86 -1
28 92 91 -1
29 98 97 -1
30 104 103 -1
31 111 109 -2

Test code:

int main() {

  for(int d=0; d<32; d++)
  {
     int a = int(3 + 0.3 * pow(double(d), 1.8)) * 3/4 + (2 < d && d < 5);
     int b = int(2.4 + 0.222 * pow(d + 0.0, 1.8));

     std::cout << d << " " << a << " " << b << " " << b-a << std::endl;
  }

  return 0;
}

bench: 8350690
2013-10-22 23:09:40 +02:00
Chris Caino fbfce2132a Simplify futility margins formula
New formula mathces the old formula until d = 45

Test code:

int main() {

  for(int d=1; d<=45; d++)
  {
     int a = int(log(double(d * d) / 2) / log(2.0) + 1.001);
     int b = int(2.9 * log(double(d)));

     if (a != b) std::cout << d << std::endl;
  }

  return 0;
}

bench: 8455956
2013-10-22 23:06:06 +02:00
Marco Costalba 2c825294ec Tweak again chain pawn bonus
This is the first chain bonus version
from Ralph that also passed both

Short TC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 23460 W: 4727 L: 4556 D: 14177

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 31858 W: 5497 L: 5240 D: 21121

And performed better against current
committed version, always at 60secs:

LLR: -2.94 (-2.94,2.94) [-3.00,3.00]
Total: 26301 W: 4477 L: 4580 D: 17244

This test was done by Leonid.

bench: 8455956
2013-10-22 17:47:16 +02:00
Marco Costalba f86d2aee29 Re-add "Further increase safe checks bonus"
After 40K games at 60 secs, result is still
not clear, but not a regression against SF 4

After
ELO: 50.11 +-2.1 (95%) LOS: 100.0%
Total: 40000 W: 10547 L: 4817 D: 24636

Before
ELO: 49.51 +-2.1 (95%) LOS: 100.0%
Total: 40000 W: 10483 L: 4821 D: 24696

So re-apply the patch to avoid to
special-case this one.

bench: 7403882
2013-10-22 17:33:11 +02:00
Marco Costalba 35ea39bed2 Restore behaviour after count<ALL_PIECES> fix
Because pos.count<ALL_PIECES>(Us) was always zero,
rewrite the formula as if this would still be
the case.

bench: 8510004
2013-10-22 17:27:58 +02:00
Ralph Stößer 97015afce8 Further improve chain pawn evaluation
Passed both short TC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 28299 W: 5854 L: 5667 D: 16778

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 9738 W: 1797 L: 1644 D: 6297

bench: 9294116
2013-10-20 23:41:40 +02:00
Marco Costalba f22a63ce67 Fix pos.count<ALL_PIECES>()
It was never updated !

Currently it only affects evaluate_passed_pawns()
and in particularly the rule to increase the bonus
if we have more non-pawn pieces. We could simply use
popcount() instead and avoid the little slowdown
in put_piece() and remove_piece(), but this would
leave a very subtle and tricky hole where people
are forced to remember that pos.count<ALL_PIECES>()
does not work. This is not obvious and so dangerous.

Thanks to Ronald de Man for spotting this.

bench: 7931424
2013-10-20 23:36:46 +02:00
Marco Costalba c08e7419a0 Fix build on Intel compiler
Due to a strange issue (bug?) the ternary
operator does not return a BitCountType for
icc, so revert to the expression.

The same patch was already applied in
9749f1f14c

Thanks to NssY Wanyonyi for pointing out
this.

No functional change.
2013-10-20 23:19:08 +02:00
Marco Costalba 67f91bc5ea Revert "Further increase safe checks bonus"
This reverts commit 4bc2374450 for
two reasons.

First regression testing shows almost equal
score:

Before the patch:
ELO: 49.75 +-2.5 (95%) LOS: 100.0%
Total: 27205 W: 7113 L: 3244 D: 16848

After the patch:
ELO: 48.87 +-2.9 (95%) LOS: 100.0%
Total: 20860 W: 5478 L: 2563 D: 12819

Second, and more sensible to me, this patch
increases safe check bonuses to 4 times their
original value (!) and considering:

- Values were already well tuned

- Values are highly critical

- King safety is highly critical, very TC
  dependent and very difficult to test

- Our testing coverage is partial (self-testing,
  blitz times)

I think is better to be safe than sorry and so
I revert the patch.

bench: 8440524
2013-10-20 10:04:43 +02:00
Ralph Stößer 4bc2374450 Further increase safe checks bonus
Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 10466 W: 2087 L: 1953 D: 6426

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 26334 W: 4540 L: 4310 D: 17484

And also proved stronger than a slightly
different patch, also succesful against master:

https://github.com/mcostalba/Stockfish/commit/dc6830a3b4ed12

But losing against current one in a match
at 60secs with SPRT [-3, 3]:

LLR: -2.96 (-2.94,2.94) [-3.00,3.00]
Total: 44484 W: 7360 L: 7463 D: 29661

bench: 9160831
2013-10-19 12:19:36 +02:00
Marco Costalba f5e872a0e3 Some evaluation code reshuffle
No functional change.
2013-10-18 09:49:38 -07:00
Jörg Oster 25cb851f8a Score chain pawn also by rank
Use the (rescaled) CandidatePassed[] table
that is already rank based.

Passed both short TC
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 11048 W: 2272 L: 2135 D: 6641

And long TC
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 4116 W: 769 L: 645 D: 2702

bench: 8440524
2013-10-18 10:27:55 +02:00
Chris Caino 3cddb0c076 Simplification of KPsK function
Also the drawing criteria has been slightly loosened.
It now detects a draw if the king is ahead of all the
pawns and on the same file or the adjacent file.

bench: 7700683
2013-10-15 07:36:01 +02:00
Chris Caino 2bf18bfc63 Bug fix for KQKRPs endgame
This lost position 8/8/3q4/8/5k2/2P1R3/2K2P2/8 w - - 0 1
was previously evaluated as a draw.

The king and rook need to be correctly placed with
respect to the _same_ pawn.

(Note also that the check for the pawn being on RANK_2
in the old version is redundant: it must be on RANK_2 if
it hopes to protect a rook on RANK_3)

Signed-off-by: Marco Costalba <mcostalba@gmail.com>
2013-10-15 07:34:29 +02:00
Ralph Stößer 5aeb907fa1 Double king safety weights
Good both at short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 5448 W: 1133 L: 1012 D: 3303

And at long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 40509 W: 6836 L: 6541 D: 27132

bench: 7700683
2013-10-14 23:24:29 +02:00
Chris Caino 0c68971c13 Remove a drawing rule from KBPsK function
The rule can be incorrect if the attacking king is
well placed e.g. 8/6K1/8/8/7k/1B6/7P/8 w - - 0 1

bench: 8279065
2013-10-14 19:55:07 +02:00
Marco Costalba d9be00342c Massive stronger/weaker renming
No functional change.
2013-10-14 19:38:08 +02:00
Chris Caino 027d85e82a Add helper function verify_material
Allows to remove a lot of assert code in endgames.

No functional change.
2013-10-14 19:23:57 +02:00
ceebo 3bc3c069f1 Add some knowledge for KRPKB endgame
bench: 8279065
2013-10-14 07:44:02 +02:00
ceebo e9366fa155 Improve KBPsK endgame
Better endgame with bishop and blocked g-pawn

bench: 8279065
2013-10-14 07:44:01 +02:00
Marco Costalba 549b5c478f Remove unuseful optimization in RKISS
Don't need a struct here. Speed test shows
result is teh same. Moreover RKISS is used
mainly at startup to compute magics, so
prefer to keep it simple...RKISS ;-)

Also some assorted triviality while there.

No functional change.
2013-10-13 03:35:17 -07:00
Joona Kiiski b15e148b5e Smoother transition for LMR
Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 12376 W: 2596 L: 2454 D: 7326

And long TC:
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 14798 W: 2584 L: 2409 D: 9805

bench: 8279065
2013-10-09 19:13:41 +02:00
Uri Blass bb83a417cb Increase slowmover and reduce instability
These two changes go in opposite directions and it
seems that the combination is stronger than original.

Here are the positive tests at various TC:

15+0.05
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 24561 W: 4946 L: 4772 D: 14843

60+0.05
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 15259 W: 2598 L: 2423 D: 10238

40/30
LLR: 2.96 (-2.94,2.94) [-3.00,3.00]
Total: 2570 W: 527 L: 422 D: 1621

Unfortunately there is also a bad result
with one sec time increment that needs
to be further investigated:

12+1
LLR: -2.97 (-2.94,2.94) [-3.00,3.00]
Total: 2694 W: 438 L: 543 D: 1713

bench: 8340585
2013-10-08 21:24:21 +02:00
Lucas Braesch 984ee9d05b Use TT refined value to stand pat
Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 17811 W: 3520 L: 3366 D: 10925

And long TC:
LLR: 2.95 (-2.94,2.94) [0.00,6.00]
Total: 30255 W: 5070 L: 4825 D: 20360

bench: 8340585
2013-10-08 19:51:08 +02:00
Lucas Braesch a0cc15ccbc Use double everywhere
Rationale:

- Speed of double and float is about the same (not on the hot path anyway)

- Double makes code prettier (no need to write 1.0f, just 1.0)

- Only practical advantage of float is to use less memory, but since we never
  store large arrays of double, we don't care.

No functional change.
2013-10-05 18:12:52 +02:00
Lucas Braesch 7f142d6817 Use prefix operators wherever possible
No functional change.
2013-10-05 18:10:43 +02:00
Marco Costalba bd1c3ed7e3 Add more depth/positions to bench
Increase bench default depth from 12 to 13 and
add 15 new endgame positions to have broader
coverage and also more reliable nps calulcation
used for fishtest framework.

Due to the new endgame positions, where nps is higher,
the total nps is increased of about 15%.

Thanks to Lucas and Jörg for the suggestions.

No functional change, but bench number is now:

bench: 8336338
2013-09-29 09:43:10 +02:00
Marco Costalba cca34e234c Drop 'is' prefix from query functions
Most but not all.

No functional change.
2013-09-28 06:47:59 -07:00
Marco Costalba ed95ad1c0e Fix build on Mac OS X
For some users -stack_size,0x4000 does not work,
so revert for now.

osX 10.6.8
gcc version 4.7.3 (MacPorts gcc47 4.7.3_2)

g++: error: unrecognized command line option '-stack_size,0x4000'
make[2]: *** [stockfish] Error 1
make[1]: *** [gcc-profile-make] Error 2
make: *** [profile-build] Error 2

No functional change.
2013-09-28 04:16:16 -07:00
Marco Costalba c65d67feb5 Revert "Use a per-thread array"
This reverts commit 800410eef1 and instead increases
stack size.

I went through the old emails with Daylen that reported the
crash issue on Mac OS X and was fixed by 0049d3f337.

It was reported default stack size for a thread in Mac OS X is 8
megabytes while the patch that we are reverting allows to reduce
stack size at max of about 217KB, so the reason for the crash was
only marginal in MAX_MOVES value. On those emails Daylen also
hinted how to increase stack size for Mac OS X to 16MB.

So prefer to increase stack size to 16MB instad of re-inventing
the wheel and do our home grown stack as we did with the patch
that we are now reverting (it will remain anyhow in git history
for documentation purposes).

No functional change.
2013-09-28 10:10:51 +02:00
Lucas Braesch bc6faf633e Simplify extensions
Unify extensions between PV and not PV nodes
and remove all but check extensions.

This is a simplification so tested at fixed number
of games where proved to not regress.

About 45k games at 15+0.05
ELO: 1.23 +-2.0 (95%) LOS: 88.5%
Total: 45643 W: 9107 L: 8946 D: 27590

About 45k games at 60+0.05
ELO: 1.07 +-1.8 (95%) LOS: 87.8%
Total: 46786 W: 7728 L: 7584 D: 31474

bench: 3172206
2013-09-28 09:54:22 +02:00
Reuven Peleg 8d1c1074d5 Simplify tte use condition
No functional change.
2013-09-27 09:40:48 +02:00
Raminder Singh e654209211 Fix best move lookup bug
If the uci option 'Best Book Move' is set to true the lookup still
returns a move at random instead of the move with the highest
weight.

No functional change.
2013-09-27 09:04:24 +02:00
Marco Costalba b742a3f29a Increase MAX_MOVES to 256
This should be enough for any legal position, even
the handcrafted ones, like the one presented by Reuven:

1Q5R/4Q1K1/B1Q5/B4Q2/N2Q4/pQ4Q1/pn2Q3/krQ4R w - -

Where currently we crash. This reverts the patch
0049d3f337 of 8/4/2012 where stack
was shrinked due to crashes while in deep analysys.

No functional change.
2013-09-27 08:59:03 +02:00
Marco Costalba 800410eef1 Use a per-thread array for generated moves
This greately reduces stack usage and is a
prerequisite for next patch.

Verified with 40K games both in single and SMP
case that there are no regressions.

No functional change.
2013-09-27 08:44:36 +02:00
Jean-Francois Romang 7b2cda95d9 Update disabled warnings for Intel compiler
No functional change.
2013-09-23 08:11:43 +02:00
Jean-Francois Romang c2cefa6de0 Replace -O3 with -fast for intel compiler
No functional change.
2013-09-23 08:11:14 +02:00
Ralph Stößer d7f5f15d69 Reduce negative quiets by ONE_PLY / 2
Passed both short TC:
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 3402 W: 708 L: 593 D: 2101

And long TC:
LLR: 2.93 (-2.94,2.94) [0.00,6.00]
Total: 23379 W: 3972 L: 3759 D: 15648

bench: 3529630
2013-09-23 08:08:52 +02:00
Marco Costalba 84906b83ec Final time management setup
This is an even safer setup proposed and tested
by Alexandre Meirelles.

Regression testing of 40K games at 10+0.05 show
result is stable both against current master:

ELO: -0.29 +-2.2 (95%) LOS: 39.7%
Total: 40000 W: 8010 L: 8043 D: 23947

and again original master (the one with smallest
time parameters):

ELO: 1.71 +-2.2 (95%) LOS: 93.8%
Total: 40000 W: 8325 L: 8128 D: 23547

Alexandre verified with LittleBlitzer time losses are
greately reduced with this setup:

Games Completed = 2100 of 3000 (Avg game length = 35.745 sec)

Settings = RR/128MB/15000ms+50ms/M 1000cp for 12 moves, D 150 moves/
Time = 39200 sec elapsed, 16800 sec remaining
 1.  Stockfish 190913             1091.5/2100    803-720-577      (L: m=313 t=1 i=0 a=406)    (D: r=278 i=91 f=136 s=8 a=64)    (tpm=212.5 d=14.75 nps=925427)
 2.  Houdini 2.0 w32              1008.5/2100    720-803-577      (L: m=250 t=299 i=0 a=254)    (D: r=278 i=91 f=136 s=8 a=64)    (tpm=204.1 d=12.04 nps=1326351)

No functional change.
2013-09-23 07:59:51 +02:00
Marco Costalba 274079990a Increase Emergency Move Time to 20
Goes in the direction of avoiding time losses and seems
equivalent after almost 40K games at super fast TC of 10+0.05

ELO: 2.61 +-2.2 (95%) LOS: 99.1%
Total: 39869 W: 8258 L: 7959 D: 23652

No functional change.
2013-09-19 07:26:36 +02:00
Marco Costalba 10cb19d534 Increase Emergency Move Time to 10
Goes in the direction of avoiding time losses and seems
equivalent after almost 40K games at super fast TC of 10+0.05

ELO: 2.41 +-2.3 (95%) LOS: 98.1%
Total: 37222 W: 7843 L: 7585 D: 21794

No functional change.
2013-09-17 16:32:39 +02:00
Marco Costalba d50b33cacd Fix a silly unstoppable eval bug
The logic is broken for black side because we get more bonus
for pawn in 7th rank than for pawn in 2nd rank!

Spotted by Reuven Peleg

bench:3884409
2013-09-16 23:59:37 +02:00
Joona Kiiski 77b5ee0117 Fix time parameters for blitz games
The ideal setting for super-blitz might be something like:

    "Emergency Base Time" = 50
    "Emergency Move Time" = 5

This would give a total emergency time buffer of:

    50 + 40 * 5 = 250 ms

This setup replaces the previous half cooked hack
"Don't blunder under extreme time pressure".

Test results are very good at super blitz, but keep good even
at 60 secs.

At 5+0.05
ELO: 24.30 +-2.4 (95%) LOS: 100.0%
Total: 37802 W: 10060 L: 7420 D: 20322

At 15+0.05
ELO: 13.41 +-2.9 (95%) LOS: 100.0%
Total: 22271 W: 4853 L: 3994 D: 13424

At 60+0.05
ELO: 5.30 +-3.2 (95%) LOS: 99.9%
Total: 16000 W: 2897 L: 2653 D: 10450

No functional change.
2013-09-16 09:07:47 +02:00
Marco Costalba af750bd2ef Rewrite unstoppable pawns evaluation
Instead of current code, give a bonus according to the frontmost
square among candidate + passed pawns.

This is a big simplification that removes a lot of accurate code
substituting it with a statistically based one using the common
'bonus' scheme, leaving to the search to sort out the details.

Results are equivalent but code is much less and, as an added bonus,
we now store candidates bitboard in pawns hash and allow this
info to be used in evaluation. This paves the way to possible
candidate pawns evaluations together with all the other pieces,
as we do for passed.

Patch passed short TC
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 16927 W: 3462 L: 3308 D: 10157

Then failed (quite quickly) at long TC
LLR: -2.95 (-2.94,2.94) [0.00,6.00]
Total: 8451 W: 1386 L: 1448 D: 5617

But when ran with a conclusive 40K fixed games at 60 secs it proved
almost equivalent to original one.

ELO: 1.08 +-2.0 (95%) LOS: 85.8%
Total: 40000 W: 6739 L: 6615 D: 26646

bench: 3884003
2013-09-16 08:57:37 +02:00
Reuven Peleg 21cbfafc03 Code style at passed pawn eval
No functional change.
2013-09-15 21:49:06 +02:00
Reuven Peleg d3947b2f3e Nicer operator declerations
No functional change.
2013-09-15 21:45:18 +02:00
Kojirion a71209868b Use pre-increment also for native types
Now that we use pre-increment on enums, it
make sense, for code style uniformity, to
swith to pre-increment also for native types,
although there is no speed difference.

No functional change.
2013-09-15 09:17:21 +02:00
Marco Costalba 7a1ff6d8ff Fix operator++ definition
ENABLE_OPERATORS_ON has incorrect definitions of
post-increment and post-decrement operators.

In particularly the returned value is the variable
already incremented/decremented, while instead they
should return the variable _before_ inc/dec.

This has no real effect because are only used in loops
and where the returned value is never used, neverthless
it is wrong. The fix would be to copy the variable to a
dummy, then inc/dec the variable, then return the dummy.

So instead, rename to pre-increment that can be implemented
without the dummy, actually the current implementation
it is already the correct pre-increment, with the only change
to return a reference (an l-value) and not a copy, so
to properly mimic the pre-increment on native integers.

Spotted by Kojirion.

No functional change.
2013-09-15 09:09:06 +02:00
Marco Costalba 82f6779c2e Don't blunder under extreme time pressure
We always attempt to keep at least this emergencyBaseTime
at clock. But if available time is very low it means that
we will force ourself to play immediately to satisfy the
emergencyBaseTime constrain and so leading to blunders.

Patch is good at short and very short TC (15secs and 5secs respectively)
LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 26590 W: 5426 L: 5245 D: 15919

LLR: 2.96 (-2.94,2.94) [-1.50,4.50]
Total: 5767 W: 1397 L: 1268 D: 3102

Instead seems has no influence at longer TC (60 secs)
LLR: -2.96 (-2.94,2.94) [0.00,6.00]
Total: 79862 W: 13623 L: 13339 D: 52900

So it is committed to have a broader testing but is
to be consider still EXPERIMENTAL and can be reverted
easily.

No functional change.
2013-09-15 07:59:09 +02:00
Marco Costalba 3abccdc82d Move classify_leaf() to c'tor in bitbases
No functional change.
2013-09-14 13:08:37 +02:00
Marco Costalba 8d6d0223bf Small touches to bitbase.cpp
Inspired by Lucas's code:

https://github.com/lucasart/chess/blob/master/src/kpk.cc

No functional change.
2013-09-14 11:19:12 +02:00
Reuven Peleg fc17d0de77 Increase passed bonus for having more pieces
Passed both short TC:
LLR: 2.95 (-2.94,2.94) [-1.50,4.50]
Total: 36463 W: 7575 L: 7365 D: 21523

And long TC:
LLR: 2.97 (-2.94,2.94) [0.00,6.00]
Total: 2953 W: 564 L: 446 D: 1943

bench: 3846852
2013-09-13 18:08:22 +02:00
Marco Costalba 27f2ce8f6e Revert "Move draw by material check"
Possible regression

bench: 4554579
2013-09-12 08:44:11 +02:00
Marco Costalba 45b0aea875 Revert "Fix random moves when time < 10ms"
Possible regression.

No functional change.
2013-09-12 08:38:19 +02:00
Marco Costalba 4803d5772c Extend checks more when below alpha
Passed both short TC:
LLR: 2.97 (-2.94,2.94) [-1.50,4.50]
Total: 8739 W: 1830 L: 1698 D: 5211

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 6716 W: 1238 L: 1101 D: 4377

bench: 4554576
2013-09-11 19:15:28 +02:00
Uri Blass 738c5595ad Extend checks more in losing positions
Passed both short TC:
LLR: 2.98 (-2.94,2.94) [-1.50,4.50]
Total: 3974 W: 860 L: 741 D: 2373

And long TC:
LLR: 2.96 (-2.94,2.94) [0.00,6.00]
Total: 16807 W: 2917 L: 2733 D: 11157

bench: 3767999
2013-09-11 09:15:47 +02:00
Reuven Peleg bebd6e16f6 Simplify unstoppable pawns evaluation
No functional change
2013-09-10 23:02:05 +02:00
Marco Costalba 49e110c52b Fix random moves when time < 10ms
In case we have less then 10ms to think as soon as
we wake up the timer, it immediately fires and calls
check_time() where due to condition:

elapsed > TimeMgr.maximum_time() - 2 * TimerResolution

the stop flag is set and search returns immediately, without
actually search anything.

Here the somewhat hacky fix is to start the timer after
at least one iteration as been completed.

No functional change.
2013-09-10 21:23:20 +02:00
Reuven Peleg 4d90aeb0ab More readable space mask
No functional change.
2013-09-10 19:18:10 +02:00
homoSapiensSapiens 03cd049c68 Change condition to use relative rank
No functional change
2013-09-10 00:47:31 +02:00
Marco Costalba 6ab8b9b6c6 Fix some comments in position.cpp
No functional change.
2013-09-08 06:28:53 -07:00
Marco Costalba 490f67a3f8 Move draw by material check
It is more natural to test this case among
others material distributions.

No functional change.
2013-09-08 06:11:35 -07:00
Marco Costalba 0515ad0fb0 Remove unreachable values in mobility table
The possible maximum mobility cardinality (plus one in case of
zero squares available) is:

- Knights: max. 8  squares -> max. 9  entries
- Bishops: max. 13 squares -> max. 14 entries
- Rooks:   max. 14 squares -> max. 15 entries
- Queen:   max. 27 squares -> max. 28 entries

So remove the extra entries in the table.

Spotted by Dariusz Orzechowski.

No functional change.
2013-09-07 18:25:24 +02:00
Marco Costalba bf51db2526 Fix warning: double to float truncation
MSVC 2013 says:
warning C4305: '*=' : truncation from 'double' to 'float'

No functional change.
2013-09-07 12:30:44 +02:00
Lucas Braesch 59702aca0d Singular extension at 8 plies also for PV nodes
Passed both short TC:
LLR: 2.96 (-2.94,2.94)
Total: 11451 W: 2455 L: 2282 D: 6714

And long TC
LLR: 2.96 (-2.94,2.94)
Total: 15813 W: 2907 L: 2723 D: 10183

bench: 3864419
2013-09-07 09:34:22 +02:00
Lucas Braesch c86eee3918 Union of 2 changes
Union of

- LMR >= 3 plies from Gary tests.stockfishchess.org/tests/view/522522960ebc595d328fcafd

- allows() tweak from Reuven tests.stockfishchess.org/tests/view/5225fa1c0ebc595d328fcb53

Both passed Step I and failed Step II.

Instead this union passed both short TC:
LLR: 2.95 (-2.94,2.94)
Total: 14525 W: 3063 L: 2874 D: 8588

And long TC
LLR: 2.94 (-2.94,2.94)
Total: 31075 W: 5566 L: 5308 D: 20201

bench: 4238160
2013-09-07 09:25:45 +02:00
Lucas Braesch 10b53e1c5e Do not prune useless checks in QS
Passed both SPRT tests in "simplification mode", so with
elo0: -3.00 alpha: 0.05 elo1: 3.00 beta: 0.05

Short TC:
LLR: 2.96 (-2.94,2.94)
Total: 6243 W: 1302 L: 1195 D: 3746

Long TC
LLR: 2.96 (-2.94,2.94)
Total: 22972 W: 4124 L: 4020 D: 14828

bench: 4633330
2013-09-05 18:50:16 +02:00
Marco Costalba a30d3571ca Revert "Fix check for bishop pair in material imbalance"
Idea is sound but implementation is partial. Ryan and Joona noticed that
    we leave an hole in material table. Also we got another report by an user
    of an odd behaviour. Namely, if you start stockfish and from the prompt
    give 'bench' you get 3453941, then if you run again bench you get 3453940.

    The reason is that two different positions with the same number of pieces,
    but one with a bishop pair and another without have the same material key.
    But after Eelco patch also different material imbalance and this yields
    to this issue.

    Restesting at long TC shows the patch does not really contribute at
    ELO improvement. Actually patch failed at long TC.

    LLR: -2.97 (-2.94,2.94)
    Total: 23109 W: 4104 L: 4092 D: 14913

    So revert.

    bench: 3453945
2013-09-05 06:34:48 +02:00
Reuven Peleg 457ac26de5 Rewrite backward pawn detection
Use the new backmost_sq() instead of a loop.

No functional change.
2013-09-03 20:11:00 +02:00
Eelco de Groot 679c2ea227 Fix check for bishop pair in material imbalance
Prefer pos.bishop_pair() to pos.count<BISHOP>(WHITE) > 1
because the first checks that the two bishops are on
different color squares.

Although the change seems to kick in only in very rare cases,
quite surprisingly it was able to pass SPRT test at short TC.

LLR: 2.95 (-2.94,2.94)
Total: 39818 W: 8174 L: 7956 D: 23688

bench: 3453941
2013-09-03 19:40:34 +02:00
Marco Costalba 9ff594c3a9 Rewrite KBBKN endgame
This was thought to be a draw but the bishops generally win. However,
it takes up to 66 moves. The position in the diagram was thought to be
a draw for over one hundred years, but tablebases show that White wins
in 45 moves. All of the long wins go through this type of semi-fortress
position. It takes several moves to force Black out of the temporary
fortress in the corner; then precise play with the bishops prevents Black
from forming the temporary fortress in another corner (Nunn 1995:265ff).

Before computer analysis, Speelman listed this position as unresolved,
but "probably a draw" (Speelman 1981:109).

bench: 3453945
2013-09-02 11:03:01 +02:00
Marco Costalba 849b089a63 Don't use lpthread for Android
Thanks to Peter Osterlund for the feedback.

No functional change.
2013-09-01 13:48:09 -07:00
Marco Costalba aee404f532 Improve ARM compatibility
STANDALONE-TOOLCHAIN.html in Android NDK says:

It is recommended to use the -mthumb compiler flag to force the generation
of 16-bit Thumb-1 instructions (the default being 32-bit ARM ones).

If you want to target the 'armeabi-v7a' ABI, you will need ensure that the
following two flags are being used:

  CFLAGS='-march=armv7-a -mfloat-abi=softfp'

Note: The first flag enables Thumb-2 instructions, and the second one
      enables H/W FPU instructions while ensuring that floating-point
      parameters are passed in core registers, which is critical for
      ABI compatibility. Do *not* use these flags separately!

Thanks to Peter Osterlund for pointout this doc and for showing me
an example Makefile to follow.

No functional change.
2013-09-01 09:18:37 -07:00
Uri Blass 0915f85895 Union of 2 changes that failed with good score
This is a union of 2 changes:

A tweak of recaptures limit from Joona Kiiski
http://tests.stockfishchess.org/tests/view/52166d7c0ebc59319a242400

and a tweak of move count pruning from Leonid Pechenik
http://tests.stockfishchess.org/tests/view/5217c7e60ebc59319a242456

The set passed both short TC at 30+0.05
LLR: 2.96 (-2.94,2.94)
Total: 18936 W: 3723 L: 3566 D: 11647

And the usual long TC at 60+0.05
LLR: 2.95 (-2.94,2.94)
Total: 48962 W: 8837 L: 8487 D: 31638

bench: 3453945
2013-09-01 08:07:21 -07:00
Marco Costalba 5e8bc6ac2a Assorted clean up in endgames
No functional change.
2013-09-01 07:39:04 -07:00
Marco Costalba 3e4dcaa06e Fix a bogus assert in allows()
Becuase castle is coded as "king captures the rook"
the to_sq(move), A1/8 or H1/8 is empty after the move,
leading to assert assert(p != NO_PIECE) in color_of().

Teach allows() asserts about castle and fix the crash.

Bug reported by Ryan Takker and tracked down by Tom Vijlbrief.

No functional change.
2013-08-30 16:42:18 +02:00
Marco Costalba 14f47c8ac6 Use frontmost_sq() and backmost_sq helpers
Should easier to read than the lsb() / msb() low
level functions.

No functional change.
2013-08-30 16:22:22 +02:00
Gary Linscott 5d90c149b5 Enable LMR for dangerous moves
Passed both short TC
LLR: 2.96 (-2.94,2.94)
Total: 5598 W: 1250 L: 1125 D: 3223

And long TC
LLR: 2.97 (-2.94,2.94)
Total: 16441 W: 3102 L: 2912 D: 10427

bench: 4620975
2013-08-29 23:02:18 +02:00
homoSapiensSapiens 4b9e338541 Bonus for rook behind a passed
If our rook is behind a passed pawn, all
squares are defended.

One of the longest tests to pass !

Passed both short TC
LLR: 2.97 (-2.94,2.94)
Total: 44560 W: 9518 L: 9281 D: 25761

And long TC
LLR: 2.96 (-2.94,2.94)
Total: 61348 W: 11618 L: 11192 D: 38538

bench: 3787694
2013-08-29 22:53:21 +02:00
homoSapiensSapiens a0cf424cfc Replace hardcoded 128 by constant
No functional change.
2013-08-29 13:59:49 +02:00
Gary Linscott aecdbfc4a0 Add lsb() overload
Helper to find least significant bit relative to
the given color.

No functional change.

Signed-off-by: Marco Costalba <mcostalba@gmail.com>
2013-08-28 19:50:11 +02:00
Uri Blass e6482b7d97 Time management: move faster if PV is stable
Move faster but compensate by allocating more
time when the best move changes.

Passed short TC at 15+0.05
LLR: 2.93 (-2.94,2.94)
Total: 13895 W: 3030 L: 2882 D: 798

Long TC at 60+0.05
LLR: 2.96 (-2.94,2.94)
Total: 9266 W: 1777 L: 1624 D: 5865

At time increment 30+0.5
LLR: 2.96 (-2.94,2.94)
Total: 6703 W: 1238 L: 1134 D: 4331

And at fixed game number, longer TC 120+0.05
ELO: 5.17 +-2.8 (95%) LOS: 100.0%
Total: 19306 W: 3378 L: 3091 D: 12837

bench: 4728533
2013-08-26 10:29:58 -07:00
homoSapiensSapiens 6e6c5b6103 Simplify chain detection
No functional change.
2013-08-25 18:36:49 +02:00
homoSapiensSapiens 04fd7efdfa Simplify kf definition in shelter_storm()
No functional change.
2013-08-25 18:31:47 +02:00
Chris Caino 5e331f9618 Fix KBPK bug
With

position fen 7k/8/8/8/8/7P/6K1/7B w - - 0 1
go depth 25

The evaluation at depth 22 is not draw as it should be. The reason is that
when search reaches the position 8/6kP/8/8/8/3B4/6K1/8 w - - 0 1 if white plays
h8R or h8N then we get a position that is a "KNOWN_WIN" and is _not_ a check, so
futility pruning in qsearch kicks in and black may think that it is "futile"
to reply Kxh8 since, according to the logic of the code, it cannot raise the score
back towards a draw.

bench: 4728533
2013-08-24 10:08:03 +02:00
homoSapiensSapiens b9f5d1c6ff Simplify condition for backward pawn
No functional changes.
2013-08-22 14:39:08 +02:00
Marco Costalba c4533e0d94 Retire redundant endgames
The case of two lone kings on the board is already considered
by the "No pawns" scaling factor rules in material.cpp as is
KBK and KNK.

Moreover we had a small leak in endgames map because for
KK endgame it happens white and black material keys are the
same (both equal to zero), so when adding the black endgame in
Endgames::add() we were overwriting the already exsisting
white one, leading to a memory leak found by Valgrind.

So remove the endgames althogheter and rely on scaling
to correctly set the endgames value to a draw.

No functional change.
2013-08-22 13:13:06 +02:00
Joona Kiiski f39cf1b008 Use null move when depth >= 2 plies
Passed both short TC:
LLR: 2.96 (-2.94,2.94)
Total: 23725 W: 5031 L: 4855 D: 13839

And long TC:
LLR: 2.96 (-2.94,2.94)
Total: 15730 W: 2939 L: 2754 D: 10037

bench: 4729333
2013-08-22 09:06:48 +02:00
Marco Costalba c6baefb79d Restore development version
No functional change.
2013-08-21 08:41:47 +02:00
56 changed files with 8811 additions and 8264 deletions
+74
View File
@@ -0,0 +1,74 @@
language: cpp
sudo: required
dist: trusty
matrix:
include:
- os: linux
compiler: gcc
addons:
apt:
sources: ['ubuntu-toolchain-r-test']
packages: ['g++-7', 'g++-7-multilib', 'g++-multilib', 'valgrind', 'expect', 'curl']
env:
- COMPILER=g++-7
- COMP=gcc
- os: linux
compiler: clang
addons:
apt:
sources: ['ubuntu-toolchain-r-test', 'llvm-toolchain-trusty-5.0']
packages: ['clang-5.0', 'llvm-5.0-dev', 'g++-multilib', 'valgrind', 'expect', 'curl']
env:
- COMPILER=clang++-5.0
- COMP=clang
- LDFLAGS=-fuse-ld=gold
- os: osx
compiler: gcc
env:
- COMPILER=g++
- COMP=gcc
- os: osx
compiler: clang
env:
- COMPILER=clang++ V='Apple LLVM 6.0' # Apple LLVM version 6.0 (clang-600.0.54) (based on LLVM 3.5svn)
- COMP=clang
branches:
only:
- master
before_script:
- cd src
script:
# Obtain bench reference from git log
- git log HEAD | grep "\b[Bb]ench[ :]\+[0-9]\{7\}" | head -n 1 | sed "s/[^0-9]*\([0-9]*\).*/\1/g" > git_sig
- export benchref=$(cat git_sig)
- echo "Reference bench:" $benchref
#
# Verify bench number against various builds
- export CXXFLAGS=-Werror
- make clean && make -j2 ARCH=x86-64 optimize=no debug=yes build && ../tests/signature.sh $benchref
- make clean && make -j2 ARCH=x86-32 optimize=no debug=yes build && ../tests/signature.sh $benchref
- make clean && make -j2 ARCH=x86-32 build && ../tests/signature.sh $benchref
- make clean && make -j2 ARCH=x86-64 build && ../tests/signature.sh $benchref
#
# Check perft and reproducible search
- ../tests/perft.sh
- ../tests/reprosearch.sh
#
# Valgrind
#
- export CXXFLAGS=-O1
- if [ -x "$(command -v valgrind )" ]; then make clean && make -j2 ARCH=x86-64 debug=yes optimize=no build > /dev/null && ../tests/instrumented.sh --valgrind; fi
- if [ -x "$(command -v valgrind )" ]; then ../tests/instrumented.sh --valgrind-thread; fi
#
# Sanitizer
#
# Use g++-7 as a proxy for having sanitizers, might need revision as they become available for more recent versions of clang/gcc
- if [[ "$COMPILER" == "g++-7" ]]; then make clean && make -j2 ARCH=x86-64 sanitize=undefined optimize=no debug=yes build > /dev/null && ../tests/instrumented.sh --sanitizer-undefined; fi
- if [[ "$COMPILER" == "g++-7" ]]; then make clean && make -j2 ARCH=x86-64 sanitize=thread optimize=no debug=yes build > /dev/null && ../tests/instrumented.sh --sanitizer-thread; fi
+129
View File
@@ -0,0 +1,129 @@
# List of authors for Stockfish, updated for version 10
Tord Romstad (romstad)
Marco Costalba (mcostalba)
Joona Kiiski (zamar)
Gary Linscott (glinscott)
Aditya (absimaldata)
Ajith Chandy Jose (ajithcj)
Alain Savard (Rocky640)
Alexander Kure
Ali AlZhrani (Cooffe)
Andrew Grant (AndyGrant)
Andrey Neporada (nepal)
Andy Duplain
Aram Tumanian (atumanian)
Arjun Temurnikar
Auguste Pop
Balint Pfliegel
Ben Koshy (BKSpurgeon)
Bill Henry (VoyagerOne)
braich
Brian Sheppard (SapphireBrand)
Bryan Cross (crossbr)
Bujun Guo (noobpwnftw)
Chris Cain (ceebo)
Dan Schmidt
Daniel Dugovic (ddugovic)
Dariusz Orzechowski
David Zar
Daylen Yang (daylen)
DiscanX
Eelco de Groot
ElbertoOne
erbsenzaehler
Ernesto Gatti
Fabian Beuke (madnight)
Fabian Fichter (ianfab)
fanon
Fauzi Akram Dabat (FauziAkram)
Felix Wittmann
gamander
gguliash
Gian-Carlo Pascutto (gcp)
Gontran Lemaire (gonlem)
Goodkov Vasiliy Aleksandrovich (goodkov)
Gregor Cramer
GuardianRM
Günther Demetz (pb00067, pb00068)
Guy Vreuls (gvreuls)
Henri Wiechers
Hiraoka Takuya (HiraokaTakuya)
homoSapiensSapiens
Hongzhi Cheng
Ivan Ivec (IIvec)
Jacques B. (Timshel)
Jan Ondruš (hxim)
Jared Kish (Kurtbusch)
Jarrod Torriero (DU-jdto)
Jean-Francois Romang (jromang)
Jerry Donald Watson (jerrydonaldwatson)
Jonathan Calovski (Mysseno)
Jonathan D. (SFisGOD)
Joost VandeVondele (vondele)
Jörg Oster (joergoster)
Joseph Ellis (jhellis3)
Joseph R. Prostko
jundery
Justin Blanchard
Kelly Wilson
Ken Takusagawa
kinderchocolate
Kiran Panditrao (Krgp)
Kojirion
Leonardo Ljubičić (ICCF World Champion)
Leonid Pechenik (lp--)
Linus Arver
loco-loco
Luca Brivio (lucabrivio)
Lucas Braesch (lucasart)
Lyudmil Antonov (lantonov)
Matthew Lai (matthewlai)
Matthew Sullivan
Mark Tenzer (31m059)
Michael Byrne (MichaelB7)
Michael Stembera (mstembera)
Michael Chaly (Vizvezdenec)
Michel Van den Bergh (vdbergh)
Miguel Lahoz (miguel-l)
Mikael Bäckman (mbootsector)
Mike Whiteley (protonspring)
Miroslav Fontán (Hexik)
Moez Jellouli (MJZ1977)
Mohammed Li (tthsqe12)
Nathan Rugg (nmrugg)
Nicklas Persson (NicklasPersson)
Niklas Fiekas (niklasf)
Ondrej Mosnáček (WOnder93)
Oskar Werkelin Ahlin
Pablo Vazquez
Pascal Romaret
Pasquale Pigazzini (ppigazzini)
Patrick Jansen (mibere)
pellanda
Peter Zsifkovits (CoffeeOne)
Ralph Stößer (Ralph Stoesser)
Raminder Singh
renouve
Reuven Peleg
Richard Lloyd
Rodrigo Exterckötter Tjäder
Ron Britvich (Britvich)
Ronald de Man (syzygy1)
Ryan Schmitt
Ryan Takker
Sebastian Buchwald (UniQP)
Sergei Antonov (saproj)
sf-x
shane31
Steinar Gunderson (sesse)
Stefan Geschwentner (locutus2)
Stefano Cardanobile (Stefano80)
Stéphane Nicolet (snicolet)
Thanar2
thaspel
Tom Vijlbrief (tomtor)
Torsten Franz (torfranz)
Uri Blass (uriblass)
Vince Negri
+71 -14
View File
@@ -1,5 +1,8 @@
### Overview ### Overview
[![Build Status](https://travis-ci.org/official-stockfish/Stockfish.svg?branch=master)](https://travis-ci.org/official-stockfish/Stockfish)
[![Build Status](https://ci.appveyor.com/api/projects/status/github/official-stockfish/Stockfish?svg=true)](https://ci.appveyor.com/project/mcostalba/stockfish)
Stockfish is a free UCI chess engine derived from Glaurung 2.1. It is Stockfish is a free UCI chess engine derived from Glaurung 2.1. It is
not a complete chess program and requires some UCI-compatible GUI not a complete chess program and requires some UCI-compatible GUI
(e.g. XBoard with PolyGlot, eboard, Arena, Sigma Chess, Shredder, Chess (e.g. XBoard with PolyGlot, eboard, Arena, Sigma Chess, Shredder, Chess
@@ -7,11 +10,13 @@ Partner or Fritz) in order to be used comfortably. Read the
documentation for your GUI of choice for information about how to use documentation for your GUI of choice for information about how to use
Stockfish with it. Stockfish with it.
This version of Stockfish supports up to 64 CPUs. The engine defaults This version of Stockfish supports up to 512 cores. The engine defaults
to one search thread it is therefore recommended to inspect the value of to one search thread, so it is therefore recommended to inspect the value of
the *Threads* UCI parameter, and to make sure it equals the number of CPU the *Threads* UCI parameter, and to make sure it equals the number of CPU
cores on your computer. cores on your computer.
This version of Stockfish has support for Syzygybases.
### Files ### Files
@@ -22,19 +27,59 @@ This distribution of Stockfish consists of the following files:
* Copying.txt, a text file containing the GNU General Public License. * Copying.txt, a text file containing the GNU General Public License.
* src, a subdirectory containing the full source code, including a Makefile * src, a subdirectory containing the full source code, including a Makefile
that can be used to compile Stockfish on Unix-like systems. For further that can be used to compile Stockfish on Unix-like systems.
information about how to compile Stockfish yourself read section below.
* polyglot.ini, for using Stockfish with Fabien Letouzey's PolyGlot
adapter.
### Opening books ### Syzygybases
This version of Stockfish has support for PolyGlot opening books. For **Configuration**
information about how to create such books, consult the PolyGlot
documentation. The book file can be selected by setting the *Book File* Syzygybases are configured using the UCI options "SyzygyPath",
UCI parameter. "SyzygyProbeDepth", "Syzygy50MoveRule" and "SyzygyProbeLimit".
The option "SyzygyPath" should be set to the directory or directories that
contain the .rtbw and .rtbz files. Multiple directories should be
separated by ";" on Windows and by ":" on Unix-based operating systems.
**Do not use spaces around the ";" or ":".**
Example: `C:\tablebases\wdl345;C:\tablebases\wdl6;D:\tablebases\dtz345;D:\tablebases\dtz6`
It is recommended to store .rtbw files on an SSD. There is no loss in
storing the .rtbz files on a regular HD.
Increasing the "SyzygyProbeDepth" option lets the engine probe less
aggressively. Set this option to a higher value if you experience too much
slowdown (in terms of nps) due to TB probing.
Set the "Syzygy50MoveRule" option to false if you want tablebase positions
that are drawn by the 50-move rule to count as win or loss. This may be useful
for correspondence games (because of tablebase adjudication).
The "SyzygyProbeLimit" option should normally be left at its default value.
**What to expect**
If the engine is searching a position that is not in the tablebases (e.g.
a position with 8 pieces), it will access the tablebases during the search.
If the engine reports a very large score (typically 123.xx), this means
that it has found a winning line into a tablebase position.
If the engine is given a position to search that is in the tablebases, it
will use the tablebases at the beginning of the search to preselect all
good moves, i.e. all moves that preserve the win or preserve the draw while
taking into account the 50-move rule.
It will then perform a search only on those moves. **The engine will not move
immediately**, unless there is only a single good move. **The engine likely
will not report a mate score even if the position is known to be won.**
It is therefore clear that behaviour is not identical to what one might
be used to with Nalimov tablebases. There are technical reasons for this
difference, the main technical reason being that Nalimov tablebases use the
DTM metric (distance-to-mate), while Syzygybases use a variation of the
DTZ metric (distance-to-zero, zero meaning any move that resets the 50-move
counter). This special metric is one of the reasons that Syzygybases are
more compact than Nalimov tablebases, while still storing all information
needed for optimal play and in addition being able to take into account
the 50-move rule.
### Compiling it yourself ### Compiling it yourself
@@ -46,11 +91,23 @@ Stockfish has support for 32 or 64-bit CPUs, the hardware POPCNT
instruction, big-endian machines such as Power PC, and other platforms. instruction, big-endian machines such as Power PC, and other platforms.
In general it is recommended to run `make help` to see a list of make In general it is recommended to run `make help` to see a list of make
targets with corresponding descriptions. When not using Makefile to targets with corresponding descriptions. When not using the Makefile to
compile (for instance with Microsoft MSVC) you need to manually compile (for instance with Microsoft MSVC) you need to manually
set/unset some switches in the compiler command line; see file *types.h* set/unset some switches in the compiler command line; see file *types.h*
for a quick reference. for a quick reference.
### Resource For Understanding the Code Base
* [Chess Programming Wiki](https://www.chessprogramming.org/Main_Page)
has good overall chess engines explanations
(techniques used here are well explained like hash maps etc), it was
also recommended by the [support team at stockfish.](http://support.stockfishchess.org/discussions/questions/1132-how-to-understand-stockfish-sources)
* [Here](https://www.chessprogramming.org/Stockfish) you can find a set
of features and techniques used by Stockfish and each of them is explained
at the wiki, however, it's a generic way rather than focusing on Stockfish's
own implementation, but it will still help you.
### Terms of use ### Terms of use
@@ -67,4 +124,4 @@ to where the source code can be found. If you make any changes to the
source code, these changes must also be made available under the GPL. source code, these changes must also be made available under the GPL.
For full details, read the copy of the GPL found in the file named For full details, read the copy of the GPL found in the file named
*Copying.txt* *Copying.txt*.
+146
View File
@@ -0,0 +1,146 @@
Contributors with >10,000 CPU hours as of November 4, 2018
Thank you!
Username CPU Hours Games played
noobpwnftw 3730975 292309380
mibere 535242 43333774
crunchy 375564 29121434
cw 371664 28748719
fastgm 318178 22283584
JojoM 295354 20958931
dew 215476 17079219
ctoks 214031 17312035
glinscott 204517 13932027
bking_US 187568 12233168
velislav 168404 13336219
CSU_Dynasty 168069 14417712
Thanar 162373 13842179
spams 149531 10940322
Fisherman 141137 12099359
drabel 134441 11180178
leszek 133658 9812120
marrco 133566 10115202
sqrt2 128420 10022279
vdbergh 123230 9200516
tvijlbrief 123007 9498831
vdv 120381 8555423
malala 117291 8126488
dsmith 114010 7622414
BrunoBanani 104938 7448565
CoffeeOne 100042 4593596
Data 94621 8433010
mgrabiak 92248 7787406
bcross 89440 8506568
brabos 81868 6647613
BRAVONE 80811 5341681
psk 77195 6156031
nordlandia 74833 6231930
robal 72818 5969856
TueRens 72523 6383294
sterni1971 71049 5647590
sunu 65855 5360884
mhoram 65034 5192880
davar 64794 5457564
nssy 64607 5371952
Pking_cda 64499 5704075
biffhero 63557 5480444
teddybaer 62147 5585620
solarlight 61278 5402642
ElbertoOne 60156 5504304
jromang 58854 4704502
dv8silencer 57421 3961325
tinker 56039 4204914
Freja 50331 3808121
renouve 50318 3544864
robnjr 47504 4131742
grandphish2 47377 4110003
eva42 46857 4075716
ttruscott 46802 3811534
finfish 46244 3481661
rap 46201 3219490
ronaldjerum 45641 3964331
xoto 44998 4170431
gvreuls 44359 3902234
bigpen0r 41780 3448224
Bobo1239 40767 3657490
Antihistamine 39218 2792761
mhunt 38991 2697512
racerschmacer 38929 3756111
VoyagerOne 35896 3378887
homyur 35561 3012398
rkl 33217 2978536
pb00067 33034 2803485
speedycpu 32043 2531964
SC 31954 2848432
EthanOConnor 31638 2143255
oryx 30962 2899534
gri 30108 2429137
csnodgrass 29396 2808611
Garf 28887 2873564
Pyafue 28885 1986098
jkiiski 28014 1923255
slakovv 27017 2031279
Prcuvu 26300 2307154
hyperbolic.tom 26248 2200777
jbwiebe 25663 2129063
anst 25525 2279159
Patrick_G 24222 1835674
nabildanial 23524 1586321
achambord 23495 1942546
Sharaf_DG 22975 1790697
chriswk 22876 1947731
ncfish1 22689 1830009
cuistot 22201 1383031
Zirie 21171 1493227
Isidor 20634 1736219
JanErik 20596 1791991
xor12 20535 1819280
team-oh 20364 1653708
nesoneg 20264 1493435
dex 20110 1682756
rstoesser 19802 1335177
Vizvezdenec 19750 1695579
eastorwest 19531 1841839
sg4032 18913 1720157
horst.prack 18425 1708197
cisco2015 18408 1793774
ianh2105 18133 1668562
MazeOfGalious 18022 1644593
ville 17900 1539130
j3corre 17607 975954
eudhan 17502 1424648
jmdana 17351 1287546
iisiraider 17175 1118788
jundery 17172 1115855
wei 16852 1822582
SFTUser 16635 1363975
purplefishies 16621 1106850
DragonLord 16599 1252348
chris 15274 1575333
IgorLeMasson 15201 1364148
dju 15074 914278
Flopzee 14700 1331632
OssumOpossum 14149 1029265
enedene 13762 935618
ako027ako 13442 1250249
AdrianSA 13324 924980
bpfliegel 13318 886523
Nikolay.IT 13260 1155612
jpulman 12776 854815
joster 12438 988413
fatmurphy 12015 901134
Nesa92 11711 1132245
Adrian.Schmidt123 11542 898699
modolief 11228 926456
Dark_wizzie 11214 1017910
mschmidt 10973 818594
Andrew Grant 10780 947859
infinity 10762 746397
SapphireBrand 10692 1024604
Thomas A. Anderson 10553 736094
basepi 10434 935168
lantonov 10325 972610
pgontarz 10294 878746
Spprtr 10189 823246
crocogoat 10115 1017325
stocky 10083 718114
+71
View File
@@ -0,0 +1,71 @@
version: 1.0.{build}
clone_depth: 50
branches:
only:
- master
- appveyor
# Operating system (build VM template)
os: Visual Studio 2015
# Build platform, i.e. x86, x64, AnyCPU. This setting is optional.
platform:
- x86
- x64
# build Configuration, i.e. Debug, Release, etc.
configuration:
- Debug
- Release
matrix:
# The build fail immediately once one of the job fails
fast_finish: true
# Scripts that are called at very beginning, before repo cloning
init:
- cmake --version
- msbuild /version
before_build:
- ps: |
# Get sources
$src = get-childitem -Path *.cpp -Recurse | select -ExpandProperty FullName
$src = $src -join ' '
$src = $src.Replace("\", "/")
# Build CMakeLists.txt
$t = 'cmake_minimum_required(VERSION 3.8)',
'project(Stockfish)',
'set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_SOURCE_DIR}/src)',
'set(source_files', $src, ')',
'add_executable(stockfish ${source_files})'
# Write CMakeLists.txt withouth BOM
$MyPath = (Get-Item -Path "." -Verbose).FullName + '\CMakeLists.txt'
$Utf8NoBomEncoding = New-Object System.Text.UTF8Encoding $False
[System.IO.File]::WriteAllLines($MyPath, $t, $Utf8NoBomEncoding)
# Obtain bench reference from git log
$b = git log HEAD | sls "\b[Bb]ench[ :]+[0-9]{7}" | select -first 1
$bench = $b -match '\D+(\d+)' | % { $matches[1] }
Write-Host "Reference bench:" $bench
$g = "Visual Studio 14 2015"
If (${env:PLATFORM} -eq 'x64') { $g = $g + ' Win64' }
cmake -G "${g}" .
Write-Host "Generated files for: " $g
build_script:
- cmake --build . --config %CONFIGURATION% -- /verbosity:minimal
before_test:
- cd src/%CONFIGURATION%
- stockfish bench 2> out.txt >NUL
- ps: |
# Verify bench number
$s = (gc "./out.txt" | out-string)
$r = ($s -match 'Nodes searched \D+(\d+)' | % { $matches[1] })
Write-Host "Engine bench:" $r
Write-Host "Reference bench:" $bench
If ($r -ne $bench) { exit 1 }
-47
View File
@@ -1,47 +0,0 @@
[PolyGlot]
EngineDir = .
EngineCommand = ./stockfish
Book = false
BookFile = book.bin
Log = false
LogFile = stockfish.log
Resign = true
ResignScore = 600
[Engine]
Write Debug Log = false
Write Search Log = false
Search Log Filename = SearchLog.txt
Book File = book.bin
Best Book Move = false
Contempt Factor = 0
Mobility (Midgame) = 100
Mobility (Endgame) = 100
Pawn Structure (Midgame) = 100
Pawn Structure (Endgame) = 100
Passed Pawns (Midgame) = 100
Passed Pawns (Endgame) = 100
Space = 100
Aggressiveness = 100
Cowardice = 100
Min Split Depth = 0
Max Threads per Split Point = 5
Threads = 1
Idle Threads Sleep = false
Hash = 128
Ponder = true
OwnBook = false
MultiPV = 1
Skill Level = 20
Emergency Move Horizon = 40
Emergency Base Time = 200
Emergency Move Time = 70
Minimum Thinking Time = 20
Slow Mover = 100
UCI_Chess960 = false
UCI_AnalyseMode = false
+244 -271
View File
@@ -1,6 +1,7 @@
# Stockfish, a UCI chess playing engine derived from Glaurung 2.1 # Stockfish, a UCI chess playing engine derived from Glaurung 2.1
# Copyright (C) 2004-2008 Tord Romstad (Glaurung author) # Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
# Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad # Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
# Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
# #
# Stockfish is free software: you can redistribute it and/or modify # Stockfish is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by # it under the terms of the GNU General Public License as published by
@@ -20,28 +21,30 @@
### Section 1. General Configuration ### Section 1. General Configuration
### ========================================================================== ### ==========================================================================
### Establish the operating system name
UNAME = $(shell uname)
### Executable name ### Executable name
ifeq ($(COMP),mingw)
EXE = stockfish.exe
else
EXE = stockfish EXE = stockfish
endif
### Installation dir definitions ### Installation dir definitions
PREFIX = /usr/local PREFIX = /usr/local
# Haiku has a non-standard filesystem layout
ifeq ($(UNAME),Haiku)
PREFIX=/boot/common
endif
BINDIR = $(PREFIX)/bin BINDIR = $(PREFIX)/bin
### Built-in benchmark for pgo-builds and signature ### Built-in benchmark for pgo-builds
PGOBENCH = ./$(EXE) bench 32 1 10 default depth PGOBENCH = ./$(EXE) bench
SIGNBENCH = ./$(EXE) bench
### Object files ### Object files
OBJS = benchmark.o bitbase.o bitboard.o book.o endgame.o evaluate.o main.o \ OBJS = benchmark.o bitbase.o bitboard.o endgame.o evaluate.o main.o \
material.o misc.o movegen.o movepick.o notation.o pawns.o position.o \ material.o misc.o movegen.o movepick.o pawns.o position.o psqt.o \
search.o thread.o timeman.o tt.o uci.o ucioption.o search.o thread.o timeman.o tt.o uci.o ucioption.o syzygy/tbprobe.o
### Establish the operating system name
KERNEL = $(shell uname -s)
ifeq ($(KERNEL),Linux)
OS = $(shell uname -o)
endif
### ========================================================================== ### ==========================================================================
### Section 2. High-level Configuration ### Section 2. High-level Configuration
@@ -51,138 +54,88 @@ OBJS = benchmark.o bitbase.o bitboard.o book.o endgame.o evaluate.o main.o \
# ---------------------------------------------------------------------------- # ----------------------------------------------------------------------------
# #
# debug = yes/no --- -DNDEBUG --- Enable/Disable debug mode # debug = yes/no --- -DNDEBUG --- Enable/Disable debug mode
# sanitize = undefined/thread/no (-fsanitize )
# --- ( undefined ) --- enable undefined behavior checks
# --- ( thread ) --- enable threading error checks
# optimize = yes/no --- (-O3/-fast etc.) --- Enable/Disable optimizations # optimize = yes/no --- (-O3/-fast etc.) --- Enable/Disable optimizations
# arch = (name) --- (-arch) --- Target architecture # arch = (name) --- (-arch) --- Target architecture
# os = (name) --- --- Target operating system
# bits = 64/32 --- -DIS_64BIT --- 64-/32-bit operating system # bits = 64/32 --- -DIS_64BIT --- 64-/32-bit operating system
# prefetch = yes/no --- -DUSE_PREFETCH --- Use prefetch x86 asm-instruction # prefetch = yes/no --- -DUSE_PREFETCH --- Use prefetch asm-instruction
# bsfq = yes/no --- -DUSE_BSFQ --- Use bsfq x86_64 asm-instruction (only # popcnt = yes/no --- -DUSE_POPCNT --- Use popcnt asm-instruction
# with GCC and ICC 64-bit)
# popcnt = yes/no --- -DUSE_POPCNT --- Use popcnt x86_64 asm-instruction
# sse = yes/no --- -msse --- Use Intel Streaming SIMD Extensions # sse = yes/no --- -msse --- Use Intel Streaming SIMD Extensions
# pext = yes/no --- -DUSE_PEXT --- Use pext x86_64 asm-instruction
# #
# Note that Makefile is space sensitive, so when adding new architectures # Note that Makefile is space sensitive, so when adding new architectures
# or modifying existing flags, you have to make sure there are no extra spaces # or modifying existing flags, you have to make sure there are no extra spaces
# at the end of the line for flag values. # at the end of the line for flag values.
### 2.1. General ### 2.1. General and architecture defaults
debug = no
optimize = yes optimize = yes
debug = no
sanitize = no
bits = 32
prefetch = no
popcnt = no
sse = no
pext = no
### 2.2 Architecture specific ### 2.2 Architecture specific
# General-section
ifeq ($(ARCH),general-64)
arch = any
os = any
bits = 64
prefetch = no
bsfq = no
popcnt = no
sse = no
endif
ifeq ($(ARCH),general-32) ifeq ($(ARCH),general-32)
arch = any arch = any
os = any
bits = 32
prefetch = no
bsfq = no
popcnt = no
sse = no
endif endif
# x86-section ifeq ($(ARCH),x86-32-old)
arch = i386
endif
ifeq ($(ARCH),x86-32)
arch = i386
prefetch = yes
sse = yes
endif
ifeq ($(ARCH),general-64)
arch = any
bits = 64
endif
ifeq ($(ARCH),x86-64) ifeq ($(ARCH),x86-64)
arch = x86_64 arch = x86_64
os = any
bits = 64 bits = 64
prefetch = yes prefetch = yes
bsfq = yes
popcnt = no
sse = yes sse = yes
endif endif
ifeq ($(ARCH),x86-64-modern) ifeq ($(ARCH),x86-64-modern)
arch = x86_64 arch = x86_64
os = any
bits = 64 bits = 64
prefetch = yes prefetch = yes
bsfq = yes
popcnt = yes popcnt = yes
sse = yes sse = yes
endif endif
ifeq ($(ARCH),x86-32) ifeq ($(ARCH),x86-64-bmi2)
arch = i386 arch = x86_64
os = any bits = 64
bits = 32
prefetch = yes prefetch = yes
bsfq = no popcnt = yes
popcnt = no
sse = yes sse = yes
pext = yes
endif endif
ifeq ($(ARCH),x86-32-old)
arch = i386
os = any
bits = 32
prefetch = no
bsfq = no
popcnt = no
sse = no
endif
#arm section
ifeq ($(ARCH),armv7) ifeq ($(ARCH),armv7)
arch = armv7 arch = armv7
os = any
bits = 32
prefetch = yes prefetch = yes
bsfq = yes
popcnt = no
sse = no
endif endif
# osx-section ifeq ($(ARCH),ppc-32)
ifeq ($(ARCH),osx-ppc-64)
arch = ppc64
os = osx
bits = 64
prefetch = no
bsfq = no
popcnt = no
sse = no
endif
ifeq ($(ARCH),osx-ppc-32)
arch = ppc arch = ppc
os = osx
bits = 32
prefetch = no
bsfq = no
popcnt = no
sse = no
endif endif
ifeq ($(ARCH),osx-x86-64) ifeq ($(ARCH),ppc-64)
arch = x86_64 arch = ppc64
os = osx
bits = 64 bits = 64
prefetch = yes
bsfq = yes
popcnt = no
sse = yes
endif
ifeq ($(ARCH),osx-x86-32)
arch = i386
os = osx
bits = 32
prefetch = yes
bsfq = no
popcnt = no
sse = yes
endif endif
@@ -191,148 +144,164 @@ endif
### ========================================================================== ### ==========================================================================
### 3.1 Selecting compiler (default = gcc) ### 3.1 Selecting compiler (default = gcc)
CXXFLAGS += -Wall -Wcast-qual -fno-exceptions -std=c++11 $(EXTRACXXFLAGS)
DEPENDFLAGS += -std=c++11
LDFLAGS += $(EXTRALDFLAGS)
ifeq ($(COMP),) ifeq ($(COMP),)
COMP=gcc COMP=gcc
endif endif
ifeq ($(COMP),mingw)
comp=mingw
CXX=g++
profile_prepare = gcc-profile-prepare
profile_make = gcc-profile-make
profile_use = gcc-profile-use
profile_clean = gcc-profile-clean
endif
ifeq ($(COMP),gcc) ifeq ($(COMP),gcc)
comp=gcc comp=gcc
CXX=g++ CXX=g++
profile_prepare = gcc-profile-prepare CXXFLAGS += -pedantic -Wextra -Wshadow
profile_make = gcc-profile-make
profile_use = gcc-profile-use ifeq ($(ARCH),armv7)
profile_clean = gcc-profile-clean ifeq ($(OS),Android)
CXXFLAGS += -m$(bits)
LDFLAGS += -m$(bits)
endif
else
CXXFLAGS += -m$(bits)
LDFLAGS += -m$(bits)
endif
ifneq ($(KERNEL),Darwin)
LDFLAGS += -Wl,--no-as-needed
endif
endif
ifeq ($(COMP),mingw)
comp=mingw
ifeq ($(KERNEL),Linux)
ifeq ($(bits),64)
ifeq ($(shell which x86_64-w64-mingw32-c++-posix),)
CXX=x86_64-w64-mingw32-c++
else
CXX=x86_64-w64-mingw32-c++-posix
endif
else
ifeq ($(shell which i686-w64-mingw32-c++-posix),)
CXX=i686-w64-mingw32-c++
else
CXX=i686-w64-mingw32-c++-posix
endif
endif
else
CXX=g++
endif
CXXFLAGS += -Wextra -Wshadow
LDFLAGS += -static
endif endif
ifeq ($(COMP),icc) ifeq ($(COMP),icc)
comp=icc comp=icc
CXX=icpc CXX=icpc
profile_prepare = icc-profile-prepare CXXFLAGS += -diag-disable 1476,10120 -Wcheck -Wabi -Wdeprecated -strict-ansi
profile_make = icc-profile-make
profile_use = icc-profile-use
profile_clean = icc-profile-clean
endif endif
ifeq ($(COMP),clang) ifeq ($(COMP),clang)
comp=clang comp=clang
CXX=clang++ CXX=clang++
profile_prepare = gcc-profile-prepare CXXFLAGS += -pedantic -Wextra -Wshadow
profile_make = gcc-profile-make
profile_use = gcc-profile-use
profile_clean = gcc-profile-clean
endif
### 3.2 General compiler settings ifneq ($(KERNEL),Darwin)
CXXFLAGS = -Wall -Wcast-qual -fno-exceptions -fno-rtti $(EXTRACXXFLAGS) ifneq ($(KERNEL),OpenBSD)
LDFLAGS += -latomic
endif
endif
ifeq ($(comp),gcc) ifeq ($(ARCH),armv7)
CXXFLAGS += -ansi -pedantic -Wno-long-long -Wextra -Wshadow ifeq ($(OS),Android)
endif CXXFLAGS += -m$(bits)
LDFLAGS += -m$(bits)
ifeq ($(comp),mingw) endif
CXXFLAGS += -Wextra -Wshadow else
CXXFLAGS += -m$(bits)
LDFLAGS += -m$(bits)
endif
endif endif
ifeq ($(comp),icc) ifeq ($(comp),icc)
CXXFLAGS += -wd383,981,1418,1419,1476,10187,10188,11505,11503 -Wcheck -Wabi -Wdeprecated -strict-ansi profile_make = icc-profile-make
endif profile_use = icc-profile-use
else
ifeq ($(comp),clang) ifeq ($(comp),clang)
CXXFLAGS += -ansi -pedantic -Wno-long-long -Wextra -Wshadow profile_make = clang-profile-make
profile_use = clang-profile-use
else
profile_make = gcc-profile-make
profile_use = gcc-profile-use
endif
endif endif
ifeq ($(os),osx) ifeq ($(KERNEL),Darwin)
CXXFLAGS += -arch $(arch) -mmacosx-version-min=10.0 CXXFLAGS += -arch $(arch) -mmacosx-version-min=10.9
LDFLAGS += -arch $(arch) -mmacosx-version-min=10.9
endif endif
### 3.3 General linker settings ### Travis CI script uses COMPILER to overwrite CXX
LDFLAGS = $(EXTRALDFLAGS) ifdef COMPILER
COMPCXX=$(COMPILER)
endif
ifeq ($(comp),mingw) ### Allow overwriting CXX from command line
LDFLAGS += -static-libstdc++ -static-libgcc ifdef COMPCXX
CXX=$(COMPCXX)
endif endif
### On mingw use Windows threads, otherwise POSIX ### On mingw use Windows threads, otherwise POSIX
ifneq ($(comp),mingw) ifneq ($(comp),mingw)
# On Android Bionic's C library comes with its own pthread implementation bundled in
ifneq ($(OS),Android)
# Haiku has pthreads in its libroot, so only link it in on other platforms # Haiku has pthreads in its libroot, so only link it in on other platforms
ifneq ($(UNAME),Haiku) ifneq ($(KERNEL),Haiku)
LDFLAGS += -lpthread LDFLAGS += -lpthread
endif endif
endif
endif endif
ifeq ($(os),osx) ### 3.2.1 Debugging
LDFLAGS += -arch $(arch) -mmacosx-version-min=10.0
endif
### 3.4 Debugging
ifeq ($(debug),no) ifeq ($(debug),no)
CXXFLAGS += -DNDEBUG CXXFLAGS += -DNDEBUG
else else
CXXFLAGS += -g CXXFLAGS += -g
endif endif
### 3.5 Optimization ### 3.2.2 Debugging with undefined behavior sanitizers
ifneq ($(sanitize),no)
CXXFLAGS += -g3 -fsanitize=$(sanitize) -fuse-ld=gold
LDFLAGS += -fsanitize=$(sanitize) -fuse-ld=gold
endif
### 3.3 Optimization
ifeq ($(optimize),yes) ifeq ($(optimize),yes)
CXXFLAGS += -O3
ifeq ($(comp),gcc) ifeq ($(comp),gcc)
CXXFLAGS += -O3 ifeq ($(OS), Android)
CXXFLAGS += -fno-gcse -mthumb -march=armv7-a -mfloat-abi=softfp
endif
endif
ifeq ($(os),osx) ifeq ($(comp),$(filter $(comp),gcc clang icc))
ifeq ($(arch),i386) ifeq ($(KERNEL),Darwin)
CXXFLAGS += -mdynamic-no-pic CXXFLAGS += -mdynamic-no-pic
endif endif
ifeq ($(arch),x86_64)
CXXFLAGS += -mdynamic-no-pic
endif
endif
ifeq ($(arch),armv7)
CXXFLAGS += -fno-gcse
endif
endif
ifeq ($(comp),mingw)
CXXFLAGS += -O3
endif
ifeq ($(comp),icc)
ifeq ($(os),osx)
CXXFLAGS += -fast -mdynamic-no-pic
else
CXXFLAGS += -O3
endif
endif
ifeq ($(comp),clang)
### -O4 requires a linker that supports LLVM's LTO
CXXFLAGS += -O3
ifeq ($(os),osx)
ifeq ($(arch),i386)
CXXFLAGS += -mdynamic-no-pic
endif
ifeq ($(arch),x86_64)
CXXFLAGS += -mdynamic-no-pic
endif
endif
endif endif
endif endif
### 3.6. Bits ### 3.4 Bits
ifeq ($(bits),64) ifeq ($(bits),64)
CXXFLAGS += -DIS_64BIT CXXFLAGS += -DIS_64BIT
endif endif
### 3.7 prefetch ### 3.5 prefetch
ifeq ($(prefetch),yes) ifeq ($(prefetch),yes)
ifeq ($(sse),yes) ifeq ($(sse),yes)
CXXFLAGS += -msse CXXFLAGS += -msse
@@ -342,31 +311,49 @@ else
CXXFLAGS += -DNO_PREFETCH CXXFLAGS += -DNO_PREFETCH
endif endif
### 3.8 bsfq ### 3.6 popcnt
ifeq ($(bsfq),yes)
CXXFLAGS += -DUSE_BSFQ
endif
### 3.9 popcnt
ifeq ($(popcnt),yes) ifeq ($(popcnt),yes)
ifeq ($(comp),icc)
CXXFLAGS += -msse3 -DUSE_POPCNT CXXFLAGS += -msse3 -DUSE_POPCNT
else
CXXFLAGS += -msse3 -mpopcnt -DUSE_POPCNT
endif
endif endif
### 3.10 Link Time Optimization, it works since gcc 4.5 but not on mingw. ### 3.7 pext
ifeq ($(pext),yes)
CXXFLAGS += -DUSE_PEXT
ifeq ($(comp),$(filter $(comp),gcc clang mingw))
CXXFLAGS += -mbmi2
endif
endif
### 3.8 Link Time Optimization, it works since gcc 4.5 but not on mingw under Windows.
### This is a mix of compile and link time options because the lto link phase ### This is a mix of compile and link time options because the lto link phase
### needs access to the optimization flags. ### needs access to the optimization flags.
ifeq ($(comp),gcc) ifeq ($(optimize),yes)
ifeq ($(optimize),yes) ifeq ($(debug), no)
ifeq ($(debug),no) ifeq ($(comp),$(filter $(comp),gcc clang))
GCC_MAJOR := `$(CXX) -dumpversion | cut -f1 -d.` CXXFLAGS += -flto
GCC_MINOR := `$(CXX) -dumpversion | cut -f2 -d.` LDFLAGS += $(CXXFLAGS)
ifeq (1,$(shell expr \( $(GCC_MAJOR) \> 4 \) \| \( $(GCC_MAJOR) \= 4 \& $(GCC_MINOR) \>= 5 \))) endif
ifeq ($(comp),mingw)
ifeq ($(KERNEL),Linux)
CXXFLAGS += -flto CXXFLAGS += -flto
LDFLAGS += $(CXXFLAGS) LDFLAGS += $(CXXFLAGS)
endif endif
endif endif
endif
endif endif
endif
### 3.9 Android 5 can only run position independent executables. Note that this
### breaks Android 4.0 and earlier.
ifeq ($(OS), Android)
CXXFLAGS += -fPIE
LDFLAGS += -fPIE -pie
endif
### ========================================================================== ### ==========================================================================
### Section 4. Public targets ### Section 4. Public targets
@@ -376,14 +363,12 @@ help:
@echo "" @echo ""
@echo "To compile stockfish, type: " @echo "To compile stockfish, type: "
@echo "" @echo ""
@echo "make target ARCH=arch [COMP=comp]" @echo "make target ARCH=arch [COMP=compiler] [COMPCXX=cxx]"
@echo "" @echo ""
@echo "Supported targets:" @echo "Supported targets:"
@echo "" @echo ""
@echo "build > Standard build" @echo "build > Standard build"
@echo "signature-build > Standard build with embedded signature"
@echo "profile-build > PGO build" @echo "profile-build > PGO build"
@echo "signature-profile-build > PGO build with embedded signature"
@echo "strip > Strip executable" @echo "strip > Strip executable"
@echo "install > Install executable" @echo "install > Install executable"
@echo "clean > Clean up" @echo "clean > Clean up"
@@ -392,13 +377,12 @@ help:
@echo "" @echo ""
@echo "x86-64 > x86 64-bit" @echo "x86-64 > x86 64-bit"
@echo "x86-64-modern > x86 64-bit with popcnt support" @echo "x86-64-modern > x86 64-bit with popcnt support"
@echo "x86-64-bmi2 > x86 64-bit with pext support"
@echo "x86-32 > x86 32-bit with SSE support" @echo "x86-32 > x86 32-bit with SSE support"
@echo "x86-32-old > x86 32-bit fall back for old hardware" @echo "x86-32-old > x86 32-bit fall back for old hardware"
@echo "osx-ppc-64 > PPC-Mac OS X 64 bit" @echo "ppc-64 > PPC 64-bit"
@echo "osx-ppc-32 > PPC-Mac OS X 32 bit" @echo "ppc-32 > PPC 32-bit"
@echo "osx-x86-64 > x86-Mac OS X 64 bit" @echo "armv7 > ARMv7 32-bit"
@echo "osx-x86-32 > x86-Mac OS X 32 bit"
@echo "armv7 > ARMv7 32 bit"
@echo "general-64 > unspecified 64-bit" @echo "general-64 > unspecified 64-bit"
@echo "general-32 > unspecified 32-bit" @echo "general-32 > unspecified 32-bit"
@echo "" @echo ""
@@ -409,52 +393,39 @@ help:
@echo "clang > LLVM Clang compiler" @echo "clang > LLVM Clang compiler"
@echo "icc > Intel compiler" @echo "icc > Intel compiler"
@echo "" @echo ""
@echo "Non-standard targets:" @echo "Simple examples. If you don't know what to do, you likely want to run: "
@echo ""
@echo "make hpux > Compile for HP-UX. Compiler = aCC"
@echo ""
@echo "Examples. If you don't know what to do, you likely want to run: "
@echo "" @echo ""
@echo "make build ARCH=x86-64 (This is for 64-bit systems)" @echo "make build ARCH=x86-64 (This is for 64-bit systems)"
@echo "make build ARCH=x86-32 (This is for 32-bit systems)" @echo "make build ARCH=x86-32 (This is for 32-bit systems)"
@echo "" @echo ""
@echo "Advanced examples, for experienced users: "
@echo ""
@echo "make build ARCH=x86-64 COMP=clang"
@echo "make profile-build ARCH=x86-64-modern COMP=gcc COMPCXX=g++-4.8"
@echo ""
.PHONY: build profile-build embed-signature
build: .PHONY: help build profile-build strip install clean objclean profileclean help \
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) config-sanity config-sanity icc-profile-use icc-profile-make gcc-profile-use gcc-profile-make \
clang-profile-use clang-profile-make
build: config-sanity
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) all $(MAKE) ARCH=$(ARCH) COMP=$(COMP) all
profile-build: profile-build: config-sanity objclean profileclean
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) config-sanity
@echo "" @echo ""
@echo "Step 0/4. Preparing for profile build." @echo "Step 1/4. Building instrumented executable ..."
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_prepare)
@echo ""
@echo "Step 1/4. Building executable for benchmark ..."
@touch *.cpp *.h
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_make) $(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_make)
@echo "" @echo ""
@echo "Step 2/4. Running benchmark for pgo-build ..." @echo "Step 2/4. Running benchmark for pgo-build ..."
@$(PGOBENCH) > /dev/null $(PGOBENCH) > /dev/null
@echo "" @echo ""
@echo "Step 3/4. Building final executable ..." @echo "Step 3/4. Building optimized executable ..."
@touch *.cpp $(MAKE) ARCH=$(ARCH) COMP=$(COMP) objclean
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_use) $(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_use)
@echo "" @echo ""
@echo "Step 4/4. Deleting profile data ..." @echo "Step 4/4. Deleting profile data ..."
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) $(profile_clean) $(MAKE) ARCH=$(ARCH) COMP=$(COMP) profileclean
embed-signature:
@echo "Running benchmark for getting the signature ..."
@$(SIGNBENCH) 2>&1 | sed -n 's/Nodes searched : \(.*\)/\/string Version\/s\/"\\(.*\\)"\/"sig-\1"\//p' > sign.txt
@sed -f sign.txt misc.cpp > misc2.cpp
@mv misc2.cpp misc.cpp
@rm sign.txt
signature-build: build embed-signature
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) all
signature-profile-build: build embed-signature profile-build
strip: strip:
strip $(EXE) strip $(EXE)
@@ -464,8 +435,19 @@ install:
-cp $(EXE) $(BINDIR) -cp $(EXE) $(BINDIR)
-strip $(BINDIR)/$(EXE) -strip $(BINDIR)/$(EXE)
clean: #clean all
$(RM) $(EXE) $(EXE).exe *.o .depend *~ core bench.txt *.gcda clean: objclean profileclean
@rm -f .depend *~ core
# clean binaries and objects
objclean:
@rm -f $(EXE) *.o ./syzygy/*.o
# clean auxiliary profiling files
profileclean:
@rm -rf profdir
@rm -f bench.txt *.gcda ./syzygy/*.gcda *.gcno ./syzygy/*.gcno
@rm -f stockfish.profdata *.profraw
default: default:
help help
@@ -480,14 +462,16 @@ config-sanity:
@echo "" @echo ""
@echo "Config:" @echo "Config:"
@echo "debug: '$(debug)'" @echo "debug: '$(debug)'"
@echo "sanitize: '$(sanitize)'"
@echo "optimize: '$(optimize)'" @echo "optimize: '$(optimize)'"
@echo "arch: '$(arch)'" @echo "arch: '$(arch)'"
@echo "os: '$(os)'"
@echo "bits: '$(bits)'" @echo "bits: '$(bits)'"
@echo "kernel: '$(KERNEL)'"
@echo "os: '$(OS)'"
@echo "prefetch: '$(prefetch)'" @echo "prefetch: '$(prefetch)'"
@echo "bsfq: '$(bsfq)'"
@echo "popcnt: '$(popcnt)'" @echo "popcnt: '$(popcnt)'"
@echo "sse: '$(sse)'" @echo "sse: '$(sse)'"
@echo "pext: '$(pext)'"
@echo "" @echo ""
@echo "Flags:" @echo "Flags:"
@echo "CXX: $(CXX)" @echo "CXX: $(CXX)"
@@ -497,22 +481,32 @@ config-sanity:
@echo "Testing config sanity. If this fails, try 'make help' ..." @echo "Testing config sanity. If this fails, try 'make help' ..."
@echo "" @echo ""
@test "$(debug)" = "yes" || test "$(debug)" = "no" @test "$(debug)" = "yes" || test "$(debug)" = "no"
@test "$(sanitize)" = "undefined" || test "$(sanitize)" = "thread" || test "$(sanitize)" = "no"
@test "$(optimize)" = "yes" || test "$(optimize)" = "no" @test "$(optimize)" = "yes" || test "$(optimize)" = "no"
@test "$(arch)" = "any" || test "$(arch)" = "x86_64" || test "$(arch)" = "i386" || \ @test "$(arch)" = "any" || test "$(arch)" = "x86_64" || test "$(arch)" = "i386" || \
test "$(arch)" = "ppc64" || test "$(arch)" = "ppc" || test "$(arch)" = "armv7" test "$(arch)" = "ppc64" || test "$(arch)" = "ppc" || test "$(arch)" = "armv7"
@test "$(os)" = "any" || test "$(os)" = "osx"
@test "$(bits)" = "32" || test "$(bits)" = "64" @test "$(bits)" = "32" || test "$(bits)" = "64"
@test "$(prefetch)" = "yes" || test "$(prefetch)" = "no" @test "$(prefetch)" = "yes" || test "$(prefetch)" = "no"
@test "$(bsfq)" = "yes" || test "$(bsfq)" = "no"
@test "$(popcnt)" = "yes" || test "$(popcnt)" = "no" @test "$(popcnt)" = "yes" || test "$(popcnt)" = "no"
@test "$(sse)" = "yes" || test "$(sse)" = "no" @test "$(sse)" = "yes" || test "$(sse)" = "no"
@test "$(pext)" = "yes" || test "$(pext)" = "no"
@test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang" @test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang"
$(EXE): $(OBJS) $(EXE): $(OBJS)
$(CXX) -o $@ $(OBJS) $(LDFLAGS) $(CXX) -o $@ $(OBJS) $(LDFLAGS)
gcc-profile-prepare: clang-profile-make:
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) gcc-profile-clean $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
EXTRACXXFLAGS='-fprofile-instr-generate ' \
EXTRALDFLAGS=' -fprofile-instr-generate' \
all
clang-profile-use:
llvm-profdata merge -output=stockfish.profdata *.profraw
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
EXTRACXXFLAGS='-fprofile-instr-use=stockfish.profdata' \
EXTRALDFLAGS='-fprofile-use ' \
all
gcc-profile-make: gcc-profile-make:
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
@@ -522,18 +516,12 @@ gcc-profile-make:
gcc-profile-use: gcc-profile-use:
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
EXTRACXXFLAGS='-fprofile-use' \ EXTRACXXFLAGS='-fprofile-use -fno-peel-loops -fno-tracer' \
EXTRALDFLAGS='-lgcov' \ EXTRALDFLAGS='-lgcov' \
all all
gcc-profile-clean:
@rm -rf *.gcda *.gcno bench.txt
icc-profile-prepare:
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) icc-profile-clean
@mkdir profdir
icc-profile-make: icc-profile-make:
@mkdir -p profdir
$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \ $(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
EXTRACXXFLAGS='-prof-gen=srcpos -prof_dir ./profdir' \ EXTRACXXFLAGS='-prof-gen=srcpos -prof_dir ./profdir' \
all all
@@ -543,23 +531,8 @@ icc-profile-use:
EXTRACXXFLAGS='-prof_use -prof_dir ./profdir' \ EXTRACXXFLAGS='-prof_use -prof_dir ./profdir' \
all all
icc-profile-clean:
@rm -rf profdir bench.txt
.depend: .depend:
-@$(CXX) $(DEPENDFLAGS) -MM $(OBJS:.o=.cpp) > $@ 2> /dev/null -@$(CXX) $(DEPENDFLAGS) -MM $(OBJS:.o=.cpp) > $@ 2> /dev/null
-include .depend -include .depend
### ==========================================================================
### Section 6. Non-standard targets
### ==========================================================================
hpux:
$(MAKE) \
CXX='/opt/aCC/bin/aCC -AA +hpxstd98 -mt +O3 -DNDEBUG -DNO_PREFETCH' \
CXXFLAGS="" \
LDFLAGS="" \
all
+77 -66
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -22,22 +23,20 @@
#include <istream> #include <istream>
#include <vector> #include <vector>
#include "misc.h"
#include "position.h" #include "position.h"
#include "search.h"
#include "thread.h"
#include "tt.h"
#include "ucioption.h"
using namespace std; using namespace std;
static const char* Defaults[] = { namespace {
const vector<string> Defaults = {
"setoption name UCI_Chess960 value false",
"rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1", "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1",
"r3k2r/p1ppqpb1/bn2pnp1/3PN3/1p2P3/2N2Q1p/PPPBBPPP/R3K2R w KQkq - 0 10", "r3k2r/p1ppqpb1/bn2pnp1/3PN3/1p2P3/2N2Q1p/PPPBBPPP/R3K2R w KQkq - 0 10",
"8/2p5/3p4/KP5r/1R3p1k/8/4P1P1/8 w - - 0 11", "8/2p5/3p4/KP5r/1R3p1k/8/4P1P1/8 w - - 0 11",
"4rrk1/pp1n3p/3q2pQ/2p1pb2/2PP4/2P3N1/P2B2PP/4RRK1 b - - 7 19", "4rrk1/pp1n3p/3q2pQ/2p1pb2/2PP4/2P3N1/P2B2PP/4RRK1 b - - 7 19",
"rq3rk1/ppp2ppp/1bnpb3/3N2B1/3NP3/7P/PPPQ1PP1/2KR3R w - - 7 14", "rq3rk1/ppp2ppp/1bnpb3/3N2B1/3NP3/7P/PPPQ1PP1/2KR3R w - - 7 14 moves d4e6",
"r1bq1r1k/1pp1n1pp/1p1p4/4p2Q/4Pp2/1BNP4/PPP2PPP/3R1RK1 w - - 2 14", "r1bq1r1k/1pp1n1pp/1p1p4/4p2Q/4Pp2/1BNP4/PPP2PPP/3R1RK1 w - - 2 14 moves g2g4",
"r3r1k1/2p2ppp/p1p1bn2/8/1q2P3/2NPQN2/PPP3PP/R4RK1 b - - 2 15", "r3r1k1/2p2ppp/p1p1bn2/8/1q2P3/2NPQN2/PPP3PP/R4RK1 b - - 2 15",
"r1bbk1nr/pp3p1p/2n5/1N4p1/2Np1B2/8/PPP2PPP/2KR1B1R w kq - 0 13", "r1bbk1nr/pp3p1p/2n5/1N4p1/2Np1B2/8/PPP2PPP/2KR1B1R w kq - 0 13",
"r1bq1rk1/ppp1nppp/4n3/3p3Q/3P4/1BP1B3/PP1N2PP/R4RK1 w - - 1 16", "r1bq1rk1/ppp1nppp/4n3/3p3Q/3P4/1BP1B3/PP1N2PP/R4RK1 w - - 1 16",
@@ -47,49 +46,77 @@ static const char* Defaults[] = {
"3r1rk1/p5pp/bpp1pp2/8/q1PP1P2/b3P3/P2NQRPP/1R2B1K1 b - - 6 22", "3r1rk1/p5pp/bpp1pp2/8/q1PP1P2/b3P3/P2NQRPP/1R2B1K1 b - - 6 22",
"r1q2rk1/2p1bppp/2Pp4/p6b/Q1PNp3/4B3/PP1R1PPP/2K4R w - - 2 18", "r1q2rk1/2p1bppp/2Pp4/p6b/Q1PNp3/4B3/PP1R1PPP/2K4R w - - 2 18",
"4k2r/1pb2ppp/1p2p3/1R1p4/3P4/2r1PN2/P4PPP/1R4K1 b - - 3 22", "4k2r/1pb2ppp/1p2p3/1R1p4/3P4/2r1PN2/P4PPP/1R4K1 b - - 3 22",
"3q2k1/pb3p1p/4pbp1/2r5/PpN2N2/1P2P2P/5PP1/Q2R2K1 b - - 4 26" "3q2k1/pb3p1p/4pbp1/2r5/PpN2N2/1P2P2P/5PP1/Q2R2K1 b - - 4 26",
"6k1/6p1/6Pp/ppp5/3pn2P/1P3K2/1PP2P2/3N4 b - - 0 1",
"3b4/5kp1/1p1p1p1p/pP1PpP1P/P1P1P3/3KN3/8/8 w - - 0 1",
"2K5/p7/7P/5pR1/8/5k2/r7/8 w - - 0 1 moves g5g6 f3e3 g6g5 e3f3",
"8/6pk/1p6/8/PP3p1p/5P2/4KP1q/3Q4 w - - 0 1",
"7k/3p2pp/4q3/8/4Q3/5Kp1/P6b/8 w - - 0 1",
"8/2p5/8/2kPKp1p/2p4P/2P5/3P4/8 w - - 0 1",
"8/1p3pp1/7p/5P1P/2k3P1/8/2K2P2/8 w - - 0 1",
"8/pp2r1k1/2p1p3/3pP2p/1P1P1P1P/P5KR/8/8 w - - 0 1",
"8/3p4/p1bk3p/Pp6/1Kp1PpPp/2P2P1P/2P5/5B2 b - - 0 1",
"5k2/7R/4P2p/5K2/p1r2P1p/8/8/8 b - - 0 1",
"6k1/6p1/P6p/r1N5/5p2/7P/1b3PP1/4R1K1 w - - 0 1",
"1r3k2/4q3/2Pp3b/3Bp3/2Q2p2/1p1P2P1/1P2KP2/3N4 w - - 0 1",
"6k1/4pp1p/3p2p1/P1pPb3/R7/1r2P1PP/3B1P2/6K1 w - - 0 1",
"8/3p3B/5p2/5P2/p7/PP5b/k7/6K1 w - - 0 1",
// 5-man positions
"8/8/8/8/5kp1/P7/8/1K1N4 w - - 0 1", // Kc2 - mate
"8/8/8/5N2/8/p7/8/2NK3k w - - 0 1", // Na2 - mate
"8/3k4/8/8/8/4B3/4KB2/2B5 w - - 0 1", // draw
// 6-man positions
"8/8/1P6/5pr1/8/4R3/7k/2K5 w - - 0 1", // Re5 - mate
"8/2p4P/8/kr6/6R1/8/8/1K6 w - - 0 1", // Ka2 - mate
"8/8/3P3k/8/1p6/8/1P6/1K3n2 b - - 0 1", // Nd2 - draw
// 7-man positions
"8/R7/2q5/8/6k1/8/1P5p/K6R w - - 0 124", // Draw
// Mate and stalemate positions
"6k1/3b3r/1p1p4/p1n2p2/1PPNpP1q/P3Q1p1/1R1RB1P1/5K2 b - - 0 1",
"r2r1n2/pp2bk2/2p1p2p/3q4/3PN1QP/2P3R1/P4PP1/5RK1 w - - 0 1",
"8/8/8/8/8/6k1/6p1/6K1 w - -",
"7k/7P/6K1/8/3B4/8/8/8 b - -",
// Chess 960
"setoption name UCI_Chess960 value true",
"bbqnnrkr/pppppppp/8/8/8/8/PPPPPPPP/BBQNNRKR w KQkq - 0 1 moves g2g3 d7d5 d2d4 c8h3 c1g5 e8d6 g5e7 f7f6",
"setoption name UCI_Chess960 value false"
}; };
} // namespace
/// benchmark() runs a simple benchmark by letting Stockfish analyze a set /// setup_bench() builds a list of UCI commands to be run by bench. There
/// of positions for a given limit each. There are five parameters; the /// are five parameters: TT size in MB, number of search threads that
/// transposition table size, the number of search threads that should /// should be used, the limit value spent for each position, a file name
/// be used, the limit value spent for each position (optional, default is /// where to look for positions in FEN format and the type of the limit:
/// depth 12), an optional file name where to look for positions in fen /// depth, perft, nodes and movetime (in millisecs).
/// format (defaults are the positions defined above) and the type of the ///
/// limit value: depth (default), time in secs or number of nodes. /// bench -> search default positions up to depth 13
/// bench 64 1 15 -> search default positions up to depth 15 (TT = 64MB)
/// bench 64 4 5000 current movetime -> search current position with 4 threads for 5 sec
/// bench 64 1 100000 default nodes -> search default positions for 100K nodes each
/// bench 16 1 5 default perft -> run a perft 5 on default positions
void benchmark(const Position& current, istream& is) { vector<string> setup_bench(const Position& current, istream& is) {
string token; vector<string> fens, list;
Search::LimitsType limits; string go, token;
vector<string> fens;
// Assign default values to missing arguments // Assign default values to missing arguments
string ttSize = (is >> token) ? token : "32"; string ttSize = (is >> token) ? token : "16";
string threads = (is >> token) ? token : "1"; string threads = (is >> token) ? token : "1";
string limit = (is >> token) ? token : "12"; string limit = (is >> token) ? token : "13";
string fenFile = (is >> token) ? token : "default"; string fenFile = (is >> token) ? token : "default";
string limitType = (is >> token) ? token : "depth"; string limitType = (is >> token) ? token : "depth";
Options["Hash"] = ttSize; go = "go " + limitType + " " + limit;
Options["Threads"] = threads;
TT.clear();
if (limitType == "time")
limits.movetime = 1000 * atoi(limit.c_str()); // movetime is in ms
else if (limitType == "nodes")
limits.nodes = atoi(limit.c_str());
else if (limitType == "mate")
limits.mate = atoi(limit.c_str());
else
limits.depth = atoi(limit.c_str());
if (fenFile == "default") if (fenFile == "default")
fens.assign(Defaults, Defaults + 16); fens = Defaults;
else if (fenFile == "current") else if (fenFile == "current")
fens.push_back(current.fen()); fens.push_back(current.fen());
@@ -97,12 +124,12 @@ void benchmark(const Position& current, istream& is) {
else else
{ {
string fen; string fen;
ifstream file(fenFile.c_str()); ifstream file(fenFile);
if (!file.is_open()) if (!file.is_open())
{ {
cerr << "Unable to open file " << fenFile << endl; cerr << "Unable to open file " << fenFile << endl;
return; exit(EXIT_FAILURE);
} }
while (getline(file, fen)) while (getline(file, fen))
@@ -112,34 +139,18 @@ void benchmark(const Position& current, istream& is) {
file.close(); file.close();
} }
int64_t nodes = 0; list.emplace_back("ucinewgame");
Search::StateStackPtr st; list.emplace_back("setoption name Threads value " + threads);
Time::point elapsed = Time::now(); list.emplace_back("setoption name Hash value " + ttSize);
for (size_t i = 0; i < fens.size(); i++) for (const string& fen : fens)
{ if (fen.find("setoption") != string::npos)
Position pos(fens[i], Options["UCI_Chess960"], Threads.main()); list.emplace_back(fen);
cerr << "\nPosition: " << i + 1 << '/' << fens.size() << endl;
if (limitType == "perft")
{
size_t cnt = Search::perft(pos, limits.depth * ONE_PLY);
cerr << "\nPerft " << limits.depth << " leaf nodes: " << cnt << endl;
nodes += cnt;
}
else else
{ {
Threads.start_thinking(pos, limits, vector<Move>(), st); list.emplace_back("position fen " + fen);
Threads.wait_for_think_finished(); list.emplace_back(go);
nodes += Search::RootPos.nodes_searched();
}
} }
elapsed = Time::now() - elapsed + 1; // Assure positive to avoid a 'divide by zero' return list;
cerr << "\n==========================="
<< "\nTotal time (ms) : " << elapsed
<< "\nNodes searched : " << nodes
<< "\nNodes/second : " << 1000 * nodes / elapsed << endl;
} }
+71 -66
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,7 +18,9 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#include <algorithm>
#include <cassert> #include <cassert>
#include <numeric>
#include <vector> #include <vector>
#include "bitboard.h" #include "bitboard.h"
@@ -25,15 +28,15 @@
namespace { namespace {
// The possible pawns squares are 24, the first 4 files and ranks from 2 to 7 // There are 24 possible pawn squares: the first 4 files and ranks from 2 to 7
const unsigned IndexMax = 2*24*64*64; // stm * psq * wksq * bksq = 196608 constexpr unsigned MAX_INDEX = 2*24*64*64; // stm * psq * wksq * bksq = 196608
// Each uint32_t stores results of 32 positions, one per bit // Each uint32_t stores results of 32 positions, one per bit
uint32_t KPKBitbase[IndexMax / 32]; uint32_t KPKBitbase[MAX_INDEX / 32];
// A KPK bitbase index is an integer in [0, IndexMax] range // A KPK bitbase index is an integer in [0, IndexMax] range
// //
// Information is mapped in a way that minimizes number of iterations: // Information is mapped in a way that minimizes the number of iterations:
// //
// bit 0- 5: white king square (from SQ_A1 to SQ_H8) // bit 0- 5: white king square (from SQ_A1 to SQ_H8)
// bit 6-11: black king square (from SQ_A1 to SQ_H8) // bit 6-11: black king square (from SQ_A1 to SQ_H8)
@@ -41,7 +44,7 @@ namespace {
// bit 13-14: white pawn file (from FILE_A to FILE_D) // bit 13-14: white pawn file (from FILE_A to FILE_D)
// bit 15-17: white pawn RANK_7 - rank (from RANK_7 - RANK_7 to RANK_7 - RANK_2) // bit 15-17: white pawn RANK_7 - rank (from RANK_7 - RANK_7 to RANK_7 - RANK_2)
unsigned index(Color us, Square bksq, Square wksq, Square psq) { unsigned index(Color us, Square bksq, Square wksq, Square psq) {
return wksq + (bksq << 6) + (us << 12) + (file_of(psq) << 13) + ((RANK_7 - rank_of(psq)) << 15); return wksq | (bksq << 6) | (us << 12) | (file_of(psq) << 13) | ((RANK_7 - rank_of(psq)) << 15);
} }
enum Result { enum Result {
@@ -51,27 +54,26 @@ namespace {
WIN = 4 WIN = 4
}; };
inline Result& operator|=(Result& r, Result v) { return r = Result(r | v); } Result& operator|=(Result& r, Result v) { return r = Result(r | v); }
struct KPKPosition { struct KPKPosition {
KPKPosition() = default;
operator Result() const { return res; } explicit KPKPosition(unsigned idx);
Result classify_leaf(unsigned idx); operator Result() const { return result; }
Result classify(const std::vector<KPKPosition>& db) Result classify(const std::vector<KPKPosition>& db)
{ return us == WHITE ? classify<WHITE>(db) : classify<BLACK>(db); } { return us == WHITE ? classify<WHITE>(db) : classify<BLACK>(db); }
private:
template<Color Us> Result classify(const std::vector<KPKPosition>& db); template<Color Us> Result classify(const std::vector<KPKPosition>& db);
Color us; Color us;
Square bksq, wksq, psq; Square ksq[COLOR_NB], psq;
Result res; Result result;
}; };
} // namespace } // namespace
bool Bitbases::probe_kpk(Square wksq, Square wpsq, Square bksq, Color us) { bool Bitbases::probe(Square wksq, Square wpsq, Square bksq, Color us) {
assert(file_of(wpsq) <= FILE_D); assert(file_of(wpsq) <= FILE_D);
@@ -80,24 +82,23 @@ bool Bitbases::probe_kpk(Square wksq, Square wpsq, Square bksq, Color us) {
} }
void Bitbases::init_kpk() { void Bitbases::init() {
std::vector<KPKPosition> db(MAX_INDEX);
unsigned idx, repeat = 1; unsigned idx, repeat = 1;
std::vector<KPKPosition> db(IndexMax);
// Initialize db with known win / draw positions // Initialize db with known win / draw positions
for (idx = 0; idx < IndexMax; idx++) for (idx = 0; idx < MAX_INDEX; ++idx)
db[idx].classify_leaf(idx); db[idx] = KPKPosition(idx);
// Iterate through the positions until no more of the unknown positions can be // Iterate through the positions until none of the unknown positions can be
// changed to either wins or draws (15 cycles needed). // changed to either wins or draws (15 cycles needed).
while (repeat) while (repeat)
for (repeat = idx = 0; idx < IndexMax; idx++) for (repeat = idx = 0; idx < MAX_INDEX; ++idx)
if (db[idx] == UNKNOWN && db[idx].classify(db) != UNKNOWN) repeat |= (db[idx] == UNKNOWN && db[idx].classify(db) != UNKNOWN);
repeat = 1;
// Map 32 results into one KPKBitbase[] entry // Map 32 results into one KPKBitbase[] entry
for (idx = 0; idx < IndexMax; idx++) for (idx = 0; idx < MAX_INDEX; ++idx)
if (db[idx] == WIN) if (db[idx] == WIN)
KPKBitbase[idx / 32] |= 1 << (idx & 0x1F); KPKBitbase[idx / 32] |= 1 << (idx & 0x1F);
} }
@@ -105,71 +106,75 @@ void Bitbases::init_kpk() {
namespace { namespace {
Result KPKPosition::classify_leaf(unsigned idx) { KPKPosition::KPKPosition(unsigned idx) {
wksq = Square((idx >> 0) & 0x3F); ksq[WHITE] = Square((idx >> 0) & 0x3F);
bksq = Square((idx >> 6) & 0x3F); ksq[BLACK] = Square((idx >> 6) & 0x3F);
us = Color ((idx >> 12) & 0x01); us = Color ((idx >> 12) & 0x01);
psq = File ((idx >> 13) & 0x03) | Rank(RANK_7 - (idx >> 15)); psq = make_square(File((idx >> 13) & 0x3), Rank(RANK_7 - ((idx >> 15) & 0x7)));
// Check if two pieces are on the same square or if a king can be captured // Check if two pieces are on the same square or if a king can be captured
if ( wksq == psq || wksq == bksq || bksq == psq if ( distance(ksq[WHITE], ksq[BLACK]) <= 1
|| (StepAttacksBB[KING][wksq] & bksq) || ksq[WHITE] == psq
|| (us == WHITE && (StepAttacksBB[PAWN][psq] & bksq))) || ksq[BLACK] == psq
return res = INVALID; || (us == WHITE && (PawnAttacks[WHITE][psq] & ksq[BLACK])))
result = INVALID;
if (us == WHITE) // Immediate win if a pawn can be promoted without getting captured
{ else if ( us == WHITE
// Immediate win if pawn can be promoted without getting captured && rank_of(psq) == RANK_7
if ( rank_of(psq) == RANK_7 && ksq[us] != psq + NORTH
&& wksq != psq + DELTA_N && ( distance(ksq[~us], psq + NORTH) > 1
&& ( square_distance(bksq, psq + DELTA_N) > 1 || (PseudoAttacks[KING][ksq[us]] & (psq + NORTH))))
||(StepAttacksBB[KING][wksq] & (psq + DELTA_N)))) result = WIN;
return res = WIN;
}
// Immediate draw if is stalemate or king captures undefended pawn
else if ( !(StepAttacksBB[KING][bksq] & ~(StepAttacksBB[KING][wksq] | StepAttacksBB[PAWN][psq]))
|| (StepAttacksBB[KING][bksq] & psq & ~StepAttacksBB[KING][wksq]))
return res = DRAW;
return res = UNKNOWN; // Immediate draw if it is a stalemate or a king captures undefended pawn
else if ( us == BLACK
&& ( !(PseudoAttacks[KING][ksq[us]] & ~(PseudoAttacks[KING][ksq[~us]] | PawnAttacks[~us][psq]))
|| (PseudoAttacks[KING][ksq[us]] & psq & ~PseudoAttacks[KING][ksq[~us]])))
result = DRAW;
// Position will be classified later
else
result = UNKNOWN;
} }
template<Color Us> template<Color Us>
Result KPKPosition::classify(const std::vector<KPKPosition>& db) { Result KPKPosition::classify(const std::vector<KPKPosition>& db) {
// White to Move: If one move leads to a position classified as WIN, the result // White to move: If one move leads to a position classified as WIN, the result
// of the current position is WIN. If all moves lead to positions classified // of the current position is WIN. If all moves lead to positions classified
// as DRAW, the current position is classified DRAW otherwise the current // as DRAW, the current position is classified as DRAW, otherwise the current
// position is classified as UNKNOWN. // position is classified as UNKNOWN.
// //
// Black to Move: If one move leads to a position classified as DRAW, the result // Black to move: If one move leads to a position classified as DRAW, the result
// of the current position is DRAW. If all moves lead to positions classified // of the current position is DRAW. If all moves lead to positions classified
// as WIN, the position is classified WIN otherwise the current position is // as WIN, the position is classified as WIN, otherwise the current position is
// classified UNKNOWN. // classified as UNKNOWN.
const Color Them = (Us == WHITE ? BLACK : WHITE); constexpr Color Them = (Us == WHITE ? BLACK : WHITE);
constexpr Result Good = (Us == WHITE ? WIN : DRAW);
constexpr Result Bad = (Us == WHITE ? DRAW : WIN);
Result r = INVALID; Result r = INVALID;
Bitboard b = StepAttacksBB[KING][Us == WHITE ? wksq : bksq]; Bitboard b = PseudoAttacks[KING][ksq[Us]];
while (b) while (b)
r |= Us == WHITE ? db[index(Them, bksq, pop_lsb(&b), psq)] r |= Us == WHITE ? db[index(Them, ksq[Them] , pop_lsb(&b), psq)]
: db[index(Them, pop_lsb(&b), wksq, psq)]; : db[index(Them, pop_lsb(&b), ksq[Them] , psq)];
if (Us == WHITE && rank_of(psq) < RANK_7)
{
Square s = psq + DELTA_N;
r |= db[index(BLACK, bksq, wksq, s)]; // Single push
if (rank_of(s) == RANK_3 && s != wksq && s != bksq)
r |= db[index(BLACK, bksq, wksq, s + DELTA_N)]; // Double push
}
if (Us == WHITE) if (Us == WHITE)
return res = r & WIN ? WIN : r & UNKNOWN ? UNKNOWN : DRAW; {
else if (rank_of(psq) < RANK_7) // Single push
return res = r & DRAW ? DRAW : r & UNKNOWN ? UNKNOWN : WIN; r |= db[index(Them, ksq[Them], ksq[Us], psq + NORTH)];
if ( rank_of(psq) == RANK_2 // Double push
&& psq + NORTH != ksq[Us]
&& psq + NORTH != ksq[Them])
r |= db[index(Them, ksq[Them], ksq[Us], psq + NORTH + NORTH)];
}
return result = r & Good ? Good : r & UNKNOWN ? UNKNOWN : Bad;
} }
} // namespace } // namespace
+128 -203
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -18,227 +19,158 @@
*/ */
#include <algorithm> #include <algorithm>
#include <cstring>
#include <iostream>
#include "bitboard.h" #include "bitboard.h"
#include "bitcount.h"
#include "misc.h" #include "misc.h"
#include "rkiss.h"
CACHE_LINE_ALIGNMENT uint8_t PopCnt16[1 << 16];
int SquareDistance[SQUARE_NB][SQUARE_NB];
Bitboard RMasks[SQUARE_NB];
Bitboard RMagics[SQUARE_NB];
Bitboard* RAttacks[SQUARE_NB];
unsigned RShifts[SQUARE_NB];
Bitboard BMasks[SQUARE_NB];
Bitboard BMagics[SQUARE_NB];
Bitboard* BAttacks[SQUARE_NB];
unsigned BShifts[SQUARE_NB];
Bitboard SquareBB[SQUARE_NB]; Bitboard SquareBB[SQUARE_NB];
Bitboard FileBB[FILE_NB]; Bitboard FileBB[FILE_NB];
Bitboard RankBB[RANK_NB]; Bitboard RankBB[RANK_NB];
Bitboard AdjacentFilesBB[FILE_NB]; Bitboard AdjacentFilesBB[FILE_NB];
Bitboard InFrontBB[COLOR_NB][RANK_NB]; Bitboard ForwardRanksBB[COLOR_NB][RANK_NB];
Bitboard StepAttacksBB[PIECE_NB][SQUARE_NB];
Bitboard BetweenBB[SQUARE_NB][SQUARE_NB]; Bitboard BetweenBB[SQUARE_NB][SQUARE_NB];
Bitboard DistanceRingsBB[SQUARE_NB][8]; Bitboard LineBB[SQUARE_NB][SQUARE_NB];
Bitboard ForwardBB[COLOR_NB][SQUARE_NB]; Bitboard DistanceRingBB[SQUARE_NB][8];
Bitboard ForwardFileBB[COLOR_NB][SQUARE_NB];
Bitboard PassedPawnMask[COLOR_NB][SQUARE_NB]; Bitboard PassedPawnMask[COLOR_NB][SQUARE_NB];
Bitboard PawnAttackSpan[COLOR_NB][SQUARE_NB]; Bitboard PawnAttackSpan[COLOR_NB][SQUARE_NB];
Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB]; Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB];
Bitboard PawnAttacks[COLOR_NB][SQUARE_NB];
int SquareDistance[SQUARE_NB][SQUARE_NB]; Magic RookMagics[SQUARE_NB];
Magic BishopMagics[SQUARE_NB];
namespace { namespace {
// De Bruijn sequences. See chessprogramming.wikispaces.com/BitScan Bitboard RookTable[0x19000]; // To store rook attacks
const uint64_t DeBruijn_64 = 0x3F79D71B4CB0A89ULL; Bitboard BishopTable[0x1480]; // To store bishop attacks
const uint32_t DeBruijn_32 = 0x783A9B23;
CACHE_LINE_ALIGNMENT void init_magics(Bitboard table[], Magic magics[], Direction directions[]);
int MS1BTable[256]; // popcount16() counts the non-zero bits using SWAR-Popcount algorithm
Square BSFTable[SQUARE_NB];
Bitboard RTable[0x19000]; // Storage space for rook attacks
Bitboard BTable[0x1480]; // Storage space for bishop attacks
typedef unsigned (Fn)(Square, Bitboard); unsigned popcount16(unsigned u) {
u -= (u >> 1) & 0x5555U;
void init_magics(Bitboard table[], Bitboard* attacks[], Bitboard magics[], u = ((u >> 2) & 0x3333U) + (u & 0x3333U);
Bitboard masks[], unsigned shifts[], Square deltas[], Fn index); u = ((u >> 4) + u) & 0x0F0FU;
return (u * 0x0101U) >> 8;
FORCE_INLINE unsigned bsf_index(Bitboard b) {
// Matt Taylor's folding for 32 bit systems, extended to 64 bits by Kim Walisch
b ^= (b - 1);
return Is64Bit ? (b * DeBruijn_64) >> 58
: ((unsigned(b) ^ unsigned(b >> 32)) * DeBruijn_32) >> 26;
} }
} }
/// lsb()/msb() finds the least/most significant bit in a nonzero bitboard.
/// pop_lsb() finds and clears the least significant bit in a nonzero bitboard.
#ifndef USE_BSFQ /// Bitboards::pretty() returns an ASCII representation of a bitboard suitable
/// to be printed to standard output. Useful for debugging.
Square lsb(Bitboard b) { return BSFTable[bsf_index(b)]; } const std::string Bitboards::pretty(Bitboard b) {
Square pop_lsb(Bitboard* b) { std::string s = "+---+---+---+---+---+---+---+---+\n";
Bitboard bb = *b; for (Rank r = RANK_8; r >= RANK_1; --r)
*b = bb & (bb - 1);
return BSFTable[bsf_index(bb)];
}
Square msb(Bitboard b) {
unsigned b32;
int result = 0;
if (b > 0xFFFFFFFF)
{ {
b >>= 32; for (File f = FILE_A; f <= FILE_H; ++f)
result = 32; s += b & make_square(f, r) ? "| X " : "| ";
s += "|\n+---+---+---+---+---+---+---+---+\n";
} }
b32 = unsigned(b); return s;
if (b32 > 0xFFFF)
{
b32 >>= 16;
result += 16;
}
if (b32 > 0xFF)
{
b32 >>= 8;
result += 8;
}
return (Square)(result + MS1BTable[b32]);
}
#endif // ifndef USE_BSFQ
/// Bitboards::print() prints a bitboard in an easily readable format to the
/// standard output. This is sometimes useful for debugging.
void Bitboards::print(Bitboard b) {
sync_cout;
for (Rank rank = RANK_8; rank >= RANK_1; rank--)
{
std::cout << "+---+---+---+---+---+---+---+---+" << '\n';
for (File file = FILE_A; file <= FILE_H; file++)
std::cout << "| " << (b & (file | rank) ? "X " : " ");
std::cout << "|\n";
}
std::cout << "+---+---+---+---+---+---+---+---+" << sync_endl;
} }
/// Bitboards::init() initializes various bitboard arrays. It is called during /// Bitboards::init() initializes various bitboard tables. It is called at
/// program initialization. /// startup and relies on global objects to be already zero-initialized.
void Bitboards::init() { void Bitboards::init() {
for (int k = 0, i = 0; i < 8; i++) for (unsigned i = 0; i < (1 << 16); ++i)
while (k < (2 << i)) PopCnt16[i] = (uint8_t) popcount16(i);
MS1BTable[k++] = i;
for (int i = 0; i < 64; i++) for (Square s = SQ_A1; s <= SQ_H8; ++s)
BSFTable[bsf_index(1ULL << i)] = Square(i); SquareBB[s] = (1ULL << s);
for (Square s = SQ_A1; s <= SQ_H8; s++) for (File f = FILE_A; f <= FILE_H; ++f)
SquareBB[s] = 1ULL << s; FileBB[f] = f > FILE_A ? FileBB[f - 1] << 1 : FileABB;
FileBB[FILE_A] = FileABB; for (Rank r = RANK_1; r <= RANK_8; ++r)
RankBB[RANK_1] = Rank1BB; RankBB[r] = r > RANK_1 ? RankBB[r - 1] << 8 : Rank1BB;
for (int i = 1; i < 8; i++) for (File f = FILE_A; f <= FILE_H; ++f)
{
FileBB[i] = FileBB[i - 1] << 1;
RankBB[i] = RankBB[i - 1] << 8;
}
for (File f = FILE_A; f <= FILE_H; f++)
AdjacentFilesBB[f] = (f > FILE_A ? FileBB[f - 1] : 0) | (f < FILE_H ? FileBB[f + 1] : 0); AdjacentFilesBB[f] = (f > FILE_A ? FileBB[f - 1] : 0) | (f < FILE_H ? FileBB[f + 1] : 0);
for (Rank r = RANK_1; r < RANK_8; r++) for (Rank r = RANK_1; r < RANK_8; ++r)
InFrontBB[WHITE][r] = ~(InFrontBB[BLACK][r + 1] = InFrontBB[BLACK][r] | RankBB[r]); ForwardRanksBB[WHITE][r] = ~(ForwardRanksBB[BLACK][r + 1] = ForwardRanksBB[BLACK][r] | RankBB[r]);
for (Color c = WHITE; c <= BLACK; c++) for (Color c = WHITE; c <= BLACK; ++c)
for (Square s = SQ_A1; s <= SQ_H8; s++) for (Square s = SQ_A1; s <= SQ_H8; ++s)
{ {
ForwardBB[c][s] = InFrontBB[c][rank_of(s)] & FileBB[file_of(s)]; ForwardFileBB [c][s] = ForwardRanksBB[c][rank_of(s)] & FileBB[file_of(s)];
PawnAttackSpan[c][s] = InFrontBB[c][rank_of(s)] & AdjacentFilesBB[file_of(s)]; PawnAttackSpan[c][s] = ForwardRanksBB[c][rank_of(s)] & AdjacentFilesBB[file_of(s)];
PassedPawnMask[c][s] = ForwardBB[c][s] | PawnAttackSpan[c][s]; PassedPawnMask[c][s] = ForwardFileBB [c][s] | PawnAttackSpan[c][s];
} }
for (Square s1 = SQ_A1; s1 <= SQ_H8; s1++) for (Square s1 = SQ_A1; s1 <= SQ_H8; ++s1)
for (Square s2 = SQ_A1; s2 <= SQ_H8; s2++) for (Square s2 = SQ_A1; s2 <= SQ_H8; ++s2)
{
SquareDistance[s1][s2] = std::max(file_distance(s1, s2), rank_distance(s1, s2));
if (s1 != s2) if (s1 != s2)
DistanceRingsBB[s1][SquareDistance[s1][s2] - 1] |= s2; {
SquareDistance[s1][s2] = std::max(distance<File>(s1, s2), distance<Rank>(s1, s2));
DistanceRingBB[s1][SquareDistance[s1][s2]] |= s2;
} }
int steps[][9] = { {}, { 7, 9 }, { 17, 15, 10, 6, -6, -10, -15, -17 }, int steps[][5] = { {}, { 7, 9 }, { 6, 10, 15, 17 }, {}, {}, {}, { 1, 7, 8, 9 } };
{}, {}, {}, { 9, 7, -7, -9, 8, 1, -1, -8 } };
for (Color c = WHITE; c <= BLACK; c++) for (Color c = WHITE; c <= BLACK; ++c)
for (PieceType pt = PAWN; pt <= KING; pt++) for (PieceType pt : { PAWN, KNIGHT, KING })
for (Square s = SQ_A1; s <= SQ_H8; s++) for (Square s = SQ_A1; s <= SQ_H8; ++s)
for (int k = 0; steps[pt][k]; k++) for (int i = 0; steps[pt][i]; ++i)
{ {
Square to = s + Square(c == WHITE ? steps[pt][k] : -steps[pt][k]); Square to = s + Direction(c == WHITE ? steps[pt][i] : -steps[pt][i]);
if (is_ok(to) && square_distance(s, to) < 3) if (is_ok(to) && distance(s, to) < 3)
StepAttacksBB[make_piece(c, pt)][s] |= to; {
if (pt == PAWN)
PawnAttacks[c][s] |= to;
else
PseudoAttacks[pt][s] |= to;
}
} }
Square RDeltas[] = { DELTA_N, DELTA_E, DELTA_S, DELTA_W }; Direction RookDirections[] = { NORTH, EAST, SOUTH, WEST };
Square BDeltas[] = { DELTA_NE, DELTA_SE, DELTA_SW, DELTA_NW }; Direction BishopDirections[] = { NORTH_EAST, SOUTH_EAST, SOUTH_WEST, NORTH_WEST };
init_magics(RTable, RAttacks, RMagics, RMasks, RShifts, RDeltas, magic_index<ROOK>); init_magics(RookTable, RookMagics, RookDirections);
init_magics(BTable, BAttacks, BMagics, BMasks, BShifts, BDeltas, magic_index<BISHOP>); init_magics(BishopTable, BishopMagics, BishopDirections);
for (Square s = SQ_A1; s <= SQ_H8; s++) for (Square s1 = SQ_A1; s1 <= SQ_H8; ++s1)
{ {
PseudoAttacks[QUEEN][s] = PseudoAttacks[BISHOP][s] = attacks_bb<BISHOP>(s, 0); PseudoAttacks[QUEEN][s1] = PseudoAttacks[BISHOP][s1] = attacks_bb<BISHOP>(s1, 0);
PseudoAttacks[QUEEN][s] |= PseudoAttacks[ ROOK][s] = attacks_bb< ROOK>(s, 0); PseudoAttacks[QUEEN][s1] |= PseudoAttacks[ ROOK][s1] = attacks_bb< ROOK>(s1, 0);
for (PieceType pt : { BISHOP, ROOK })
for (Square s2 = SQ_A1; s2 <= SQ_H8; ++s2)
{
if (!(PseudoAttacks[pt][s1] & s2))
continue;
LineBB[s1][s2] = (attacks_bb(pt, s1, 0) & attacks_bb(pt, s2, 0)) | s1 | s2;
BetweenBB[s1][s2] = attacks_bb(pt, s1, SquareBB[s2]) & attacks_bb(pt, s2, SquareBB[s1]);
} }
for (Square s1 = SQ_A1; s1 <= SQ_H8; s1++)
for (Square s2 = SQ_A1; s2 <= SQ_H8; s2++)
if (PseudoAttacks[QUEEN][s1] & s2)
{
Square delta = (s2 - s1) / square_distance(s1, s2);
for (Square s = s1 + delta; s != s2; s += delta)
BetweenBB[s1][s2] |= s;
} }
} }
namespace { namespace {
Bitboard sliding_attack(Square deltas[], Square sq, Bitboard occupied) { Bitboard sliding_attack(Direction directions[], Square sq, Bitboard occupied) {
Bitboard attack = 0; Bitboard attack = 0;
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; ++i)
for (Square s = sq + deltas[i]; for (Square s = sq + directions[i];
is_ok(s) && square_distance(s, s - deltas[i]) == 1; is_ok(s) && distance(s, s - directions[i]) == 1;
s += deltas[i]) s += directions[i])
{ {
attack |= s; attack |= s;
@@ -250,38 +182,21 @@ namespace {
} }
Bitboard pick_random(RKISS& rk, int booster) {
// Values s1 and s2 are used to rotate the candidate magic of a
// quantity known to be the optimal to quickly find the magics.
int s1 = booster & 63, s2 = (booster >> 6) & 63;
Bitboard m = rk.rand<Bitboard>();
m = (m >> s1) | (m << (64 - s1));
m &= rk.rand<Bitboard>();
m = (m >> s2) | (m << (64 - s2));
return m & rk.rand<Bitboard>();
}
// init_magics() computes all rook and bishop attacks at startup. Magic // init_magics() computes all rook and bishop attacks at startup. Magic
// bitboards are used to look up attacks of sliding pieces. As a reference see // bitboards are used to look up attacks of sliding pieces. As a reference see
// chessprogramming.wikispaces.com/Magic+Bitboards. In particular, here we // chessprogramming.wikispaces.com/Magic+Bitboards. In particular, here we
// use the so called "fancy" approach. // use the so called "fancy" approach.
void init_magics(Bitboard table[], Bitboard* attacks[], Bitboard magics[], void init_magics(Bitboard table[], Magic magics[], Direction directions[]) {
Bitboard masks[], unsigned shifts[], Square deltas[], Fn index) {
// Optimal PRNG seeds to pick the correct magics in the shortest time
int seeds[][RANK_NB] = { { 8977, 44560, 54343, 38998, 5731, 95205, 104912, 17020 },
{ 728, 10316, 55013, 32803, 12281, 15100, 16645, 255 } };
int MagicBoosters[][8] = { { 3191, 2184, 1310, 3618, 2091, 1308, 2452, 3996 },
{ 1059, 3608, 605, 3234, 3326, 38, 2029, 3043 } };
RKISS rk;
Bitboard occupancy[4096], reference[4096], edges, b; Bitboard occupancy[4096], reference[4096], edges, b;
int i, size, booster; int epoch[4096] = {}, cnt = 0, size = 0;
// attacks[s] is a pointer to the beginning of the attacks table for square 's' for (Square s = SQ_A1; s <= SQ_H8; ++s)
attacks[SQ_A1] = table;
for (Square s = SQ_A1; s <= SQ_H8; s++)
{ {
// Board edges are not considered in the relevant occupancies // Board edges are not considered in the relevant occupancies
edges = ((Rank1BB | Rank8BB) & ~rank_bb(s)) | ((FileABB | FileHBB) & ~file_bb(s)); edges = ((Rank1BB | Rank8BB) & ~rank_bb(s)) | ((FileABB | FileHBB) & ~file_bb(s));
@@ -291,49 +206,59 @@ namespace {
// all the attacks for each possible subset of the mask and so is 2 power // all the attacks for each possible subset of the mask and so is 2 power
// the number of 1s of the mask. Hence we deduce the size of the shift to // the number of 1s of the mask. Hence we deduce the size of the shift to
// apply to the 64 or 32 bits word to get the index. // apply to the 64 or 32 bits word to get the index.
masks[s] = sliding_attack(deltas, s, 0) & ~edges; Magic& m = magics[s];
shifts[s] = (Is64Bit ? 64 : 32) - popcount<Max15>(masks[s]); m.mask = sliding_attack(directions, s, 0) & ~edges;
m.shift = (Is64Bit ? 64 : 32) - popcount(m.mask);
// Set the offset for the attacks table of the square. We have individual
// table sizes for each square with "Fancy Magic Bitboards".
m.attacks = s == SQ_A1 ? table : magics[s - 1].attacks + size;
// Use Carry-Rippler trick to enumerate all subsets of masks[s] and // Use Carry-Rippler trick to enumerate all subsets of masks[s] and
// store the corresponding sliding attack bitboard in reference[]. // store the corresponding sliding attack bitboard in reference[].
b = size = 0; b = size = 0;
do { do {
occupancy[size] = b; occupancy[size] = b;
reference[size++] = sliding_attack(deltas, s, b); reference[size] = sliding_attack(directions, s, b);
b = (b - masks[s]) & masks[s];
if (HasPext)
m.attacks[pext(b, m.mask)] = reference[size];
size++;
b = (b - m.mask) & m.mask;
} while (b); } while (b);
// Set the offset for the table of the next square. We have individual if (HasPext)
// table sizes for each square with "Fancy Magic Bitboards". continue;
if (s < SQ_H8)
attacks[s + 1] = attacks[s] + size;
booster = MagicBoosters[Is64Bit][rank_of(s)]; PRNG rng(seeds[Is64Bit][rank_of(s)]);
// Find a magic for square 's' picking up an (almost) random number // Find a magic for square 's' picking up an (almost) random number
// until we find the one that passes the verification test. // until we find the one that passes the verification test.
do { for (int i = 0; i < size; )
do magics[s] = pick_random(rk, booster); {
while (popcount<Max15>((magics[s] * masks[s]) >> 56) < 6); for (m.magic = 0; popcount((m.magic * m.mask) >> 56) < 6; )
m.magic = rng.sparse_rand<Bitboard>();
std::memset(attacks[s], 0, size * sizeof(Bitboard));
// A good magic must map every possible occupancy to an index that // A good magic must map every possible occupancy to an index that
// looks up the correct sliding attack in the attacks[s] database. // looks up the correct sliding attack in the attacks[s] database.
// Note that we build up the database for square 's' as a side // Note that we build up the database for square 's' as a side
// effect of verifying the magic. // effect of verifying the magic. Keep track of the attempt count
for (i = 0; i < size; i++) // and save it in epoch[], little speed-up trick to avoid resetting
// m.attacks[] after every failed attempt.
for (++cnt, i = 0; i < size; ++i)
{ {
Bitboard& attack = attacks[s][index(s, occupancy[i])]; unsigned idx = m.index(occupancy[i]);
if (attack && attack != reference[i]) if (epoch[idx] < cnt)
break; {
epoch[idx] = cnt;
assert(reference[i] != 0); m.attacks[idx] = reference[i];
}
attack = reference[i]; else if (m.attacks[idx] != reference[i])
break;
}
} }
} while (i != size);
} }
} }
} }
+236 -174
View File
@@ -1,14 +1,14 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or the Free Software Foundation, either version 3 of the License, or
(at your option) any later version. (at your option) any later version.
Stockfish is distributed in the hope that it will be useful, Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
@@ -21,123 +21,122 @@
#ifndef BITBOARD_H_INCLUDED #ifndef BITBOARD_H_INCLUDED
#define BITBOARD_H_INCLUDED #define BITBOARD_H_INCLUDED
#include <string>
#include "types.h" #include "types.h"
namespace Bitbases {
void init();
bool probe(Square wksq, Square wpsq, Square bksq, Color us);
}
namespace Bitboards { namespace Bitboards {
void init(); void init();
void print(Bitboard b); const std::string pretty(Bitboard b);
} }
namespace Bitbases { constexpr Bitboard AllSquares = ~Bitboard(0);
constexpr Bitboard DarkSquares = 0xAA55AA55AA55AA55ULL;
void init_kpk(); constexpr Bitboard FileABB = 0x0101010101010101ULL;
bool probe_kpk(Square wksq, Square wpsq, Square bksq, Color us); constexpr Bitboard FileBBB = FileABB << 1;
constexpr Bitboard FileCBB = FileABB << 2;
constexpr Bitboard FileDBB = FileABB << 3;
constexpr Bitboard FileEBB = FileABB << 4;
constexpr Bitboard FileFBB = FileABB << 5;
constexpr Bitboard FileGBB = FileABB << 6;
constexpr Bitboard FileHBB = FileABB << 7;
} constexpr Bitboard Rank1BB = 0xFF;
constexpr Bitboard Rank2BB = Rank1BB << (8 * 1);
constexpr Bitboard Rank3BB = Rank1BB << (8 * 2);
constexpr Bitboard Rank4BB = Rank1BB << (8 * 3);
constexpr Bitboard Rank5BB = Rank1BB << (8 * 4);
constexpr Bitboard Rank6BB = Rank1BB << (8 * 5);
constexpr Bitboard Rank7BB = Rank1BB << (8 * 6);
constexpr Bitboard Rank8BB = Rank1BB << (8 * 7);
const Bitboard FileABB = 0x0101010101010101ULL; extern int SquareDistance[SQUARE_NB][SQUARE_NB];
const Bitboard FileBBB = FileABB << 1;
const Bitboard FileCBB = FileABB << 2;
const Bitboard FileDBB = FileABB << 3;
const Bitboard FileEBB = FileABB << 4;
const Bitboard FileFBB = FileABB << 5;
const Bitboard FileGBB = FileABB << 6;
const Bitboard FileHBB = FileABB << 7;
const Bitboard Rank1BB = 0xFF;
const Bitboard Rank2BB = Rank1BB << (8 * 1);
const Bitboard Rank3BB = Rank1BB << (8 * 2);
const Bitboard Rank4BB = Rank1BB << (8 * 3);
const Bitboard Rank5BB = Rank1BB << (8 * 4);
const Bitboard Rank6BB = Rank1BB << (8 * 5);
const Bitboard Rank7BB = Rank1BB << (8 * 6);
const Bitboard Rank8BB = Rank1BB << (8 * 7);
CACHE_LINE_ALIGNMENT
extern Bitboard RMasks[SQUARE_NB];
extern Bitboard RMagics[SQUARE_NB];
extern Bitboard* RAttacks[SQUARE_NB];
extern unsigned RShifts[SQUARE_NB];
extern Bitboard BMasks[SQUARE_NB];
extern Bitboard BMagics[SQUARE_NB];
extern Bitboard* BAttacks[SQUARE_NB];
extern unsigned BShifts[SQUARE_NB];
extern Bitboard SquareBB[SQUARE_NB]; extern Bitboard SquareBB[SQUARE_NB];
extern Bitboard FileBB[FILE_NB]; extern Bitboard FileBB[FILE_NB];
extern Bitboard RankBB[RANK_NB]; extern Bitboard RankBB[RANK_NB];
extern Bitboard AdjacentFilesBB[FILE_NB]; extern Bitboard AdjacentFilesBB[FILE_NB];
extern Bitboard InFrontBB[COLOR_NB][RANK_NB]; extern Bitboard ForwardRanksBB[COLOR_NB][RANK_NB];
extern Bitboard StepAttacksBB[PIECE_NB][SQUARE_NB];
extern Bitboard BetweenBB[SQUARE_NB][SQUARE_NB]; extern Bitboard BetweenBB[SQUARE_NB][SQUARE_NB];
extern Bitboard DistanceRingsBB[SQUARE_NB][8]; extern Bitboard LineBB[SQUARE_NB][SQUARE_NB];
extern Bitboard ForwardBB[COLOR_NB][SQUARE_NB]; extern Bitboard DistanceRingBB[SQUARE_NB][8];
extern Bitboard ForwardFileBB[COLOR_NB][SQUARE_NB];
extern Bitboard PassedPawnMask[COLOR_NB][SQUARE_NB]; extern Bitboard PassedPawnMask[COLOR_NB][SQUARE_NB];
extern Bitboard PawnAttackSpan[COLOR_NB][SQUARE_NB]; extern Bitboard PawnAttackSpan[COLOR_NB][SQUARE_NB];
extern Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB]; extern Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB];
extern Bitboard PawnAttacks[COLOR_NB][SQUARE_NB];
extern int SquareDistance[SQUARE_NB][SQUARE_NB];
const Bitboard DarkSquares = 0xAA55AA55AA55AA55ULL; /// Magic holds all magic bitboards relevant data for a single square
struct Magic {
Bitboard mask;
Bitboard magic;
Bitboard* attacks;
unsigned shift;
// Compute the attack's index using the 'magic bitboards' approach
unsigned index(Bitboard occupied) const {
if (HasPext)
return unsigned(pext(occupied, mask));
if (Is64Bit)
return unsigned(((occupied & mask) * magic) >> shift);
unsigned lo = unsigned(occupied) & unsigned(mask);
unsigned hi = unsigned(occupied >> 32) & unsigned(mask >> 32);
return (lo * unsigned(magic) ^ hi * unsigned(magic >> 32)) >> shift;
}
};
extern Magic RookMagics[SQUARE_NB];
extern Magic BishopMagics[SQUARE_NB];
/// Overloads of bitwise operators between a Bitboard and a Square for testing /// Overloads of bitwise operators between a Bitboard and a Square for testing
/// whether a given bit is set in a bitboard, and for setting and clearing bits. /// whether a given bit is set in a bitboard, and for setting and clearing bits.
inline Bitboard operator&(Bitboard b, Square s) { inline Bitboard operator&(Bitboard b, Square s) {
assert(s >= SQ_A1 && s <= SQ_H8);
return b & SquareBB[s]; return b & SquareBB[s];
} }
inline Bitboard& operator|=(Bitboard& b, Square s) {
return b |= SquareBB[s];
}
inline Bitboard& operator^=(Bitboard& b, Square s) {
return b ^= SquareBB[s];
}
inline Bitboard operator|(Bitboard b, Square s) { inline Bitboard operator|(Bitboard b, Square s) {
assert(s >= SQ_A1 && s <= SQ_H8);
return b | SquareBB[s]; return b | SquareBB[s];
} }
inline Bitboard operator^(Bitboard b, Square s) { inline Bitboard operator^(Bitboard b, Square s) {
assert(s >= SQ_A1 && s <= SQ_H8);
return b ^ SquareBB[s]; return b ^ SquareBB[s];
} }
inline bool more_than_one(Bitboard b) { inline Bitboard& operator|=(Bitboard& b, Square s) {
assert(s >= SQ_A1 && s <= SQ_H8);
return b |= SquareBB[s];
}
inline Bitboard& operator^=(Bitboard& b, Square s) {
assert(s >= SQ_A1 && s <= SQ_H8);
return b ^= SquareBB[s];
}
constexpr bool more_than_one(Bitboard b) {
return b & (b - 1); return b & (b - 1);
} }
inline int square_distance(Square s1, Square s2) { /// rank_bb() and file_bb() return a bitboard representing all the squares on
return SquareDistance[s1][s2]; /// the given file or rank.
}
inline int file_distance(Square s1, Square s2) {
return abs(file_of(s1) - file_of(s2));
}
inline int rank_distance(Square s1, Square s2) {
return abs(rank_of(s1) - rank_of(s2));
}
/// shift_bb() moves bitboard one step along direction Delta. Mainly for pawns.
template<Square Delta>
inline Bitboard shift_bb(Bitboard b) {
return Delta == DELTA_N ? b << 8 : Delta == DELTA_S ? b >> 8
: Delta == DELTA_NE ? (b & ~FileHBB) << 9 : Delta == DELTA_SE ? (b & ~FileHBB) >> 7
: Delta == DELTA_NW ? (b & ~FileABB) << 7 : Delta == DELTA_SW ? (b & ~FileABB) >> 9
: 0;
}
/// rank_bb() and file_bb() take a file or a square as input and return
/// a bitboard representing all squares on the given file or rank.
inline Bitboard rank_bb(Rank r) { inline Bitboard rank_bb(Rank r) {
return RankBB[r]; return RankBB[r];
@@ -156,167 +155,230 @@ inline Bitboard file_bb(Square s) {
} }
/// adjacent_files_bb() takes a file as input and returns a bitboard representing /// shift() moves a bitboard one step along direction D (mainly for pawns)
/// all squares on the adjacent files.
template<Direction D>
constexpr Bitboard shift(Bitboard b) {
return D == NORTH ? b << 8 : D == SOUTH ? b >> 8
: D == EAST ? (b & ~FileHBB) << 1 : D == WEST ? (b & ~FileABB) >> 1
: D == NORTH_EAST ? (b & ~FileHBB) << 9 : D == NORTH_WEST ? (b & ~FileABB) << 7
: D == SOUTH_EAST ? (b & ~FileHBB) >> 7 : D == SOUTH_WEST ? (b & ~FileABB) >> 9
: 0;
}
/// pawn_attacks_bb() returns the pawn attacks for the given color from the
/// squares in the given bitboard.
template<Color C>
constexpr Bitboard pawn_attacks_bb(Bitboard b) {
return C == WHITE ? shift<NORTH_WEST>(b) | shift<NORTH_EAST>(b)
: shift<SOUTH_WEST>(b) | shift<SOUTH_EAST>(b);
}
/// adjacent_files_bb() returns a bitboard representing all the squares on the
/// adjacent files of the given one.
inline Bitboard adjacent_files_bb(File f) { inline Bitboard adjacent_files_bb(File f) {
return AdjacentFilesBB[f]; return AdjacentFilesBB[f];
} }
/// in_front_bb() takes a color and a rank as input, and returns a bitboard /// between_bb() returns a bitboard representing all the squares between the two
/// representing all the squares on all ranks in front of the rank, from the /// given ones. For instance, between_bb(SQ_C4, SQ_F7) returns a bitboard with
/// given color's point of view. For instance, in_front_bb(BLACK, RANK_3) will /// the bits for square d5 and e6 set. If s1 and s2 are not on the same rank, file
/// give all squares on ranks 1 and 2. /// or diagonal, 0 is returned.
inline Bitboard in_front_bb(Color c, Rank r) {
return InFrontBB[c][r];
}
/// between_bb() returns a bitboard representing all squares between two squares.
/// For instance, between_bb(SQ_C4, SQ_F7) returns a bitboard with the bits for
/// square d5 and e6 set. If s1 and s2 are not on the same line, file or diagonal,
/// 0 is returned.
inline Bitboard between_bb(Square s1, Square s2) { inline Bitboard between_bb(Square s1, Square s2) {
return BetweenBB[s1][s2]; return BetweenBB[s1][s2];
} }
/// forward_bb() takes a color and a square as input, and returns a bitboard /// forward_ranks_bb() returns a bitboard representing the squares on all the ranks
/// representing all squares along the line in front of the square, from the /// in front of the given one, from the point of view of the given color. For instance,
/// point of view of the given color. Definition of the table is: /// forward_ranks_bb(BLACK, SQ_D3) will return the 16 squares on ranks 1 and 2.
/// ForwardBB[c][s] = in_front_bb(c, s) & file_bb(s)
inline Bitboard forward_bb(Color c, Square s) { inline Bitboard forward_ranks_bb(Color c, Square s) {
return ForwardBB[c][s]; return ForwardRanksBB[c][rank_of(s)];
} }
/// pawn_attack_span() takes a color and a square as input, and returns a bitboard /// forward_file_bb() returns a bitboard representing all the squares along the line
/// representing all squares that can be attacked by a pawn of the given color /// in front of the given one, from the point of view of the given color:
/// when it moves along its file starting from the given square. Definition is: /// ForwardFileBB[c][s] = forward_ranks_bb(c, s) & file_bb(s)
/// PawnAttackSpan[c][s] = in_front_bb(c, s) & adjacent_files_bb(s);
inline Bitboard forward_file_bb(Color c, Square s) {
return ForwardFileBB[c][s];
}
/// pawn_attack_span() returns a bitboard representing all the squares that can be
/// attacked by a pawn of the given color when it moves along its file, starting
/// from the given square:
/// PawnAttackSpan[c][s] = forward_ranks_bb(c, s) & adjacent_files_bb(file_of(s));
inline Bitboard pawn_attack_span(Color c, Square s) { inline Bitboard pawn_attack_span(Color c, Square s) {
return PawnAttackSpan[c][s]; return PawnAttackSpan[c][s];
} }
/// passed_pawn_mask() takes a color and a square as input, and returns a /// passed_pawn_mask() returns a bitboard mask which can be used to test if a
/// bitboard mask which can be used to test if a pawn of the given color on /// pawn of the given color and on the given square is a passed pawn:
/// the given square is a passed pawn. Definition of the table is: /// PassedPawnMask[c][s] = pawn_attack_span(c, s) | forward_file_bb(c, s)
/// PassedPawnMask[c][s] = pawn_attack_span(c, s) | forward_bb(c, s)
inline Bitboard passed_pawn_mask(Color c, Square s) { inline Bitboard passed_pawn_mask(Color c, Square s) {
return PassedPawnMask[c][s]; return PassedPawnMask[c][s];
} }
/// squares_of_color() returns a bitboard representing all squares with the same /// aligned() returns true if the squares s1, s2 and s3 are aligned either on a
/// color of the given square. /// straight or on a diagonal line.
inline Bitboard squares_of_color(Square s) { inline bool aligned(Square s1, Square s2, Square s3) {
return DarkSquares & s ? DarkSquares : ~DarkSquares; return LineBB[s1][s2] & s3;
} }
/// squares_aligned() returns true if the squares s1, s2 and s3 are aligned /// distance() functions return the distance between x and y, defined as the
/// either on a straight or on a diagonal line. /// number of steps for a king in x to reach y. Works with squares, ranks, files.
inline bool squares_aligned(Square s1, Square s2, Square s3) { template<typename T> inline int distance(T x, T y) { return x < y ? y - x : x - y; }
return (BetweenBB[s1][s2] | BetweenBB[s1][s3] | BetweenBB[s2][s3]) template<> inline int distance<Square>(Square x, Square y) { return SquareDistance[x][y]; }
& ( SquareBB[s1] | SquareBB[s2] | SquareBB[s3]);
} template<typename T1, typename T2> inline int distance(T2 x, T2 y);
template<> inline int distance<File>(Square x, Square y) { return distance(file_of(x), file_of(y)); }
template<> inline int distance<Rank>(Square x, Square y) { return distance(rank_of(x), rank_of(y)); }
/// Functions for computing sliding attack bitboards. Function attacks_bb() takes /// attacks_bb() returns a bitboard representing all the squares attacked by a
/// a square and a bitboard of occupied squares as input, and returns a bitboard /// piece of type Pt (bishop or rook) placed on 's'.
/// representing all squares attacked by Pt (bishop or rook) on the given square.
template<PieceType Pt>
FORCE_INLINE unsigned magic_index(Square s, Bitboard occ) {
Bitboard* const Masks = Pt == ROOK ? RMasks : BMasks;
Bitboard* const Magics = Pt == ROOK ? RMagics : BMagics;
unsigned* const Shifts = Pt == ROOK ? RShifts : BShifts;
if (Is64Bit)
return unsigned(((occ & Masks[s]) * Magics[s]) >> Shifts[s]);
unsigned lo = unsigned(occ) & unsigned(Masks[s]);
unsigned hi = unsigned(occ >> 32) & unsigned(Masks[s] >> 32);
return (lo * unsigned(Magics[s]) ^ hi * unsigned(Magics[s] >> 32)) >> Shifts[s];
}
template<PieceType Pt> template<PieceType Pt>
inline Bitboard attacks_bb(Square s, Bitboard occ) { inline Bitboard attacks_bb(Square s, Bitboard occupied) {
return (Pt == ROOK ? RAttacks : BAttacks)[s][magic_index<Pt>(s, occ)];
const Magic& m = Pt == ROOK ? RookMagics[s] : BishopMagics[s];
return m.attacks[m.index(occupied)];
}
inline Bitboard attacks_bb(PieceType pt, Square s, Bitboard occupied) {
assert(pt != PAWN);
switch (pt)
{
case BISHOP: return attacks_bb<BISHOP>(s, occupied);
case ROOK : return attacks_bb< ROOK>(s, occupied);
case QUEEN : return attacks_bb<BISHOP>(s, occupied) | attacks_bb<ROOK>(s, occupied);
default : return PseudoAttacks[pt][s];
}
} }
/// lsb()/msb() finds the least/most significant bit in a nonzero bitboard. /// popcount() counts the number of non-zero bits in a bitboard
/// pop_lsb() finds and clears the least significant bit in a nonzero bitboard.
#ifdef USE_BSFQ inline int popcount(Bitboard b) {
# if defined(_MSC_VER) && !defined(__INTEL_COMPILER) #ifndef USE_POPCNT
FORCE_INLINE Square lsb(Bitboard b) { extern uint8_t PopCnt16[1 << 16];
unsigned long index; union { Bitboard bb; uint16_t u[4]; } v = { b };
_BitScanForward64(&index, b); return PopCnt16[v.u[0]] + PopCnt16[v.u[1]] + PopCnt16[v.u[2]] + PopCnt16[v.u[3]];
return (Square) index;
#elif defined(_MSC_VER) || defined(__INTEL_COMPILER)
return (int)_mm_popcnt_u64(b);
#else // Assumed gcc or compatible compiler
return __builtin_popcountll(b);
#endif
} }
FORCE_INLINE Square msb(Bitboard b) {
unsigned long index; /// lsb() and msb() return the least/most significant bit in a non-zero bitboard
_BitScanReverse64(&index, b);
return (Square) index; #if defined(__GNUC__) // GCC, Clang, ICC
inline Square lsb(Bitboard b) {
assert(b);
return Square(__builtin_ctzll(b));
} }
# elif defined(__arm__) inline Square msb(Bitboard b) {
assert(b);
FORCE_INLINE int lsb32(uint32_t v) { return Square(63 ^ __builtin_clzll(b));
__asm__("rbit %0, %1" : "=r"(v) : "r"(v));
return __builtin_clz(v);
} }
FORCE_INLINE Square msb(Bitboard b) { #elif defined(_MSC_VER) // MSVC
return (Square) (63 - __builtin_clzll(b));
#ifdef _WIN64 // MSVC, WIN64
inline Square lsb(Bitboard b) {
assert(b);
unsigned long idx;
_BitScanForward64(&idx, b);
return (Square) idx;
} }
FORCE_INLINE Square lsb(Bitboard b) { inline Square msb(Bitboard b) {
return (Square) (uint32_t(b) ? lsb32(uint32_t(b)) : 32 + lsb32(uint32_t(b >> 32))); assert(b);
unsigned long idx;
_BitScanReverse64(&idx, b);
return (Square) idx;
} }
# else #else // MSVC, WIN32
FORCE_INLINE Square lsb(Bitboard b) { // Assembly code by Heinz van Saanen inline Square lsb(Bitboard b) {
Bitboard index; assert(b);
__asm__("bsfq %1, %0": "=r"(index): "rm"(b) ); unsigned long idx;
return (Square) index;
if (b & 0xffffffff) {
_BitScanForward(&idx, int32_t(b));
return Square(idx);
} else {
_BitScanForward(&idx, int32_t(b >> 32));
return Square(idx + 32);
}
} }
FORCE_INLINE Square msb(Bitboard b) { inline Square msb(Bitboard b) {
Bitboard index; assert(b);
__asm__("bsrq %1, %0": "=r"(index): "rm"(b) ); unsigned long idx;
return (Square) index;
if (b >> 32) {
_BitScanReverse(&idx, int32_t(b >> 32));
return Square(idx + 32);
} else {
_BitScanReverse(&idx, int32_t(b));
return Square(idx);
}
} }
# endif #endif
FORCE_INLINE Square pop_lsb(Bitboard* b) { #else // Compiler is neither GCC nor MSVC compatible
#error "Compiler not supported."
#endif
/// pop_lsb() finds and clears the least significant bit in a non-zero bitboard
inline Square pop_lsb(Bitboard* b) {
const Square s = lsb(*b); const Square s = lsb(*b);
*b &= *b - 1; *b &= *b - 1;
return s; return s;
} }
#else // if defined(USE_BSFQ)
extern Square msb(Bitboard b); /// frontmost_sq() and backmost_sq() return the square corresponding to the
extern Square lsb(Bitboard b); /// most/least advanced bit relative to the given color.
extern Square pop_lsb(Bitboard* b);
#endif inline Square frontmost_sq(Color c, Bitboard b) { return c == WHITE ? msb(b) : lsb(b); }
inline Square backmost_sq(Color c, Bitboard b) { return c == WHITE ? lsb(b) : msb(b); }
#endif // #ifndef BITBOARD_H_INCLUDED #endif // #ifndef BITBOARD_H_INCLUDED
-105
View File
@@ -1,105 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef BITCOUNT_H_INCLUDED
#define BITCOUNT_H_INCLUDED
#include <cassert>
#include "types.h"
enum BitCountType {
CNT_64,
CNT_64_MAX15,
CNT_32,
CNT_32_MAX15,
CNT_HW_POPCNT
};
/// Determine at compile time the best popcount<> specialization according if
/// platform is 32 or 64 bits, to the maximum number of nonzero bits to count
/// and if hardware popcnt instruction is available.
const BitCountType Full = HasPopCnt ? CNT_HW_POPCNT : Is64Bit ? CNT_64 : CNT_32;
const BitCountType Max15 = HasPopCnt ? CNT_HW_POPCNT : Is64Bit ? CNT_64_MAX15 : CNT_32_MAX15;
/// popcount() counts the number of nonzero bits in a bitboard
template<BitCountType> inline int popcount(Bitboard);
template<>
inline int popcount<CNT_64>(Bitboard b) {
b -= (b >> 1) & 0x5555555555555555ULL;
b = ((b >> 2) & 0x3333333333333333ULL) + (b & 0x3333333333333333ULL);
b = ((b >> 4) + b) & 0x0F0F0F0F0F0F0F0FULL;
return (b * 0x0101010101010101ULL) >> 56;
}
template<>
inline int popcount<CNT_64_MAX15>(Bitboard b) {
b -= (b >> 1) & 0x5555555555555555ULL;
b = ((b >> 2) & 0x3333333333333333ULL) + (b & 0x3333333333333333ULL);
return (b * 0x1111111111111111ULL) >> 60;
}
template<>
inline int popcount<CNT_32>(Bitboard b) {
unsigned w = unsigned(b >> 32), v = unsigned(b);
v -= (v >> 1) & 0x55555555; // 0-2 in 2 bits
w -= (w >> 1) & 0x55555555;
v = ((v >> 2) & 0x33333333) + (v & 0x33333333); // 0-4 in 4 bits
w = ((w >> 2) & 0x33333333) + (w & 0x33333333);
v = ((v >> 4) + v + (w >> 4) + w) & 0x0F0F0F0F;
return (v * 0x01010101) >> 24;
}
template<>
inline int popcount<CNT_32_MAX15>(Bitboard b) {
unsigned w = unsigned(b >> 32), v = unsigned(b);
v -= (v >> 1) & 0x55555555; // 0-2 in 2 bits
w -= (w >> 1) & 0x55555555;
v = ((v >> 2) & 0x33333333) + (v & 0x33333333); // 0-4 in 4 bits
w = ((w >> 2) & 0x33333333) + (w & 0x33333333);
return ((v + w) * 0x11111111) >> 28;
}
template<>
inline int popcount<CNT_HW_POPCNT>(Bitboard b) {
#ifndef USE_POPCNT
assert(false);
return b != 0; // Avoid 'b not used' warning
#elif defined(_MSC_VER) && defined(__INTEL_COMPILER)
return _mm_popcnt_u64(b);
#elif defined(_MSC_VER)
return (int)__popcnt64(b);
#else
__asm__("popcnt %1, %0" : "=r" (b) : "r" (b));
return b;
#endif
}
#endif // #ifndef BITCOUNT_H_INCLUDED
-478
View File
@@ -1,478 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/*
The code in this file is based on the opening book code in PolyGlot
by Fabien Letouzey. PolyGlot is available under the GNU General
Public License, and can be downloaded from http://wbec-ridderkerk.nl
*/
#include <algorithm>
#include <cassert>
#include <iostream>
#include "book.h"
#include "misc.h"
#include "movegen.h"
using namespace std;
namespace {
// A Polyglot book is a series of "entries" of 16 bytes. All integers are
// stored in big-endian format, with highest byte first (regardless of size).
// The entries are ordered according to the key in ascending order.
struct Entry {
uint64_t key;
uint16_t move;
uint16_t count;
uint32_t learn;
};
// Random numbers from PolyGlot, used to compute book hash keys
const union {
Key PolyGlotRandoms[781];
struct {
Key psq[12][64]; // [piece][square]
Key castle[4]; // [castle right]
Key enpassant[8]; // [file]
Key turn;
} Zobrist;
} PG = {{
0x9D39247E33776D41ULL, 0x2AF7398005AAA5C7ULL, 0x44DB015024623547ULL,
0x9C15F73E62A76AE2ULL, 0x75834465489C0C89ULL, 0x3290AC3A203001BFULL,
0x0FBBAD1F61042279ULL, 0xE83A908FF2FB60CAULL, 0x0D7E765D58755C10ULL,
0x1A083822CEAFE02DULL, 0x9605D5F0E25EC3B0ULL, 0xD021FF5CD13A2ED5ULL,
0x40BDF15D4A672E32ULL, 0x011355146FD56395ULL, 0x5DB4832046F3D9E5ULL,
0x239F8B2D7FF719CCULL, 0x05D1A1AE85B49AA1ULL, 0x679F848F6E8FC971ULL,
0x7449BBFF801FED0BULL, 0x7D11CDB1C3B7ADF0ULL, 0x82C7709E781EB7CCULL,
0xF3218F1C9510786CULL, 0x331478F3AF51BBE6ULL, 0x4BB38DE5E7219443ULL,
0xAA649C6EBCFD50FCULL, 0x8DBD98A352AFD40BULL, 0x87D2074B81D79217ULL,
0x19F3C751D3E92AE1ULL, 0xB4AB30F062B19ABFULL, 0x7B0500AC42047AC4ULL,
0xC9452CA81A09D85DULL, 0x24AA6C514DA27500ULL, 0x4C9F34427501B447ULL,
0x14A68FD73C910841ULL, 0xA71B9B83461CBD93ULL, 0x03488B95B0F1850FULL,
0x637B2B34FF93C040ULL, 0x09D1BC9A3DD90A94ULL, 0x3575668334A1DD3BULL,
0x735E2B97A4C45A23ULL, 0x18727070F1BD400BULL, 0x1FCBACD259BF02E7ULL,
0xD310A7C2CE9B6555ULL, 0xBF983FE0FE5D8244ULL, 0x9F74D14F7454A824ULL,
0x51EBDC4AB9BA3035ULL, 0x5C82C505DB9AB0FAULL, 0xFCF7FE8A3430B241ULL,
0x3253A729B9BA3DDEULL, 0x8C74C368081B3075ULL, 0xB9BC6C87167C33E7ULL,
0x7EF48F2B83024E20ULL, 0x11D505D4C351BD7FULL, 0x6568FCA92C76A243ULL,
0x4DE0B0F40F32A7B8ULL, 0x96D693460CC37E5DULL, 0x42E240CB63689F2FULL,
0x6D2BDCDAE2919661ULL, 0x42880B0236E4D951ULL, 0x5F0F4A5898171BB6ULL,
0x39F890F579F92F88ULL, 0x93C5B5F47356388BULL, 0x63DC359D8D231B78ULL,
0xEC16CA8AEA98AD76ULL, 0x5355F900C2A82DC7ULL, 0x07FB9F855A997142ULL,
0x5093417AA8A7ED5EULL, 0x7BCBC38DA25A7F3CULL, 0x19FC8A768CF4B6D4ULL,
0x637A7780DECFC0D9ULL, 0x8249A47AEE0E41F7ULL, 0x79AD695501E7D1E8ULL,
0x14ACBAF4777D5776ULL, 0xF145B6BECCDEA195ULL, 0xDABF2AC8201752FCULL,
0x24C3C94DF9C8D3F6ULL, 0xBB6E2924F03912EAULL, 0x0CE26C0B95C980D9ULL,
0xA49CD132BFBF7CC4ULL, 0xE99D662AF4243939ULL, 0x27E6AD7891165C3FULL,
0x8535F040B9744FF1ULL, 0x54B3F4FA5F40D873ULL, 0x72B12C32127FED2BULL,
0xEE954D3C7B411F47ULL, 0x9A85AC909A24EAA1ULL, 0x70AC4CD9F04F21F5ULL,
0xF9B89D3E99A075C2ULL, 0x87B3E2B2B5C907B1ULL, 0xA366E5B8C54F48B8ULL,
0xAE4A9346CC3F7CF2ULL, 0x1920C04D47267BBDULL, 0x87BF02C6B49E2AE9ULL,
0x092237AC237F3859ULL, 0xFF07F64EF8ED14D0ULL, 0x8DE8DCA9F03CC54EULL,
0x9C1633264DB49C89ULL, 0xB3F22C3D0B0B38EDULL, 0x390E5FB44D01144BULL,
0x5BFEA5B4712768E9ULL, 0x1E1032911FA78984ULL, 0x9A74ACB964E78CB3ULL,
0x4F80F7A035DAFB04ULL, 0x6304D09A0B3738C4ULL, 0x2171E64683023A08ULL,
0x5B9B63EB9CEFF80CULL, 0x506AACF489889342ULL, 0x1881AFC9A3A701D6ULL,
0x6503080440750644ULL, 0xDFD395339CDBF4A7ULL, 0xEF927DBCF00C20F2ULL,
0x7B32F7D1E03680ECULL, 0xB9FD7620E7316243ULL, 0x05A7E8A57DB91B77ULL,
0xB5889C6E15630A75ULL, 0x4A750A09CE9573F7ULL, 0xCF464CEC899A2F8AULL,
0xF538639CE705B824ULL, 0x3C79A0FF5580EF7FULL, 0xEDE6C87F8477609DULL,
0x799E81F05BC93F31ULL, 0x86536B8CF3428A8CULL, 0x97D7374C60087B73ULL,
0xA246637CFF328532ULL, 0x043FCAE60CC0EBA0ULL, 0x920E449535DD359EULL,
0x70EB093B15B290CCULL, 0x73A1921916591CBDULL, 0x56436C9FE1A1AA8DULL,
0xEFAC4B70633B8F81ULL, 0xBB215798D45DF7AFULL, 0x45F20042F24F1768ULL,
0x930F80F4E8EB7462ULL, 0xFF6712FFCFD75EA1ULL, 0xAE623FD67468AA70ULL,
0xDD2C5BC84BC8D8FCULL, 0x7EED120D54CF2DD9ULL, 0x22FE545401165F1CULL,
0xC91800E98FB99929ULL, 0x808BD68E6AC10365ULL, 0xDEC468145B7605F6ULL,
0x1BEDE3A3AEF53302ULL, 0x43539603D6C55602ULL, 0xAA969B5C691CCB7AULL,
0xA87832D392EFEE56ULL, 0x65942C7B3C7E11AEULL, 0xDED2D633CAD004F6ULL,
0x21F08570F420E565ULL, 0xB415938D7DA94E3CULL, 0x91B859E59ECB6350ULL,
0x10CFF333E0ED804AULL, 0x28AED140BE0BB7DDULL, 0xC5CC1D89724FA456ULL,
0x5648F680F11A2741ULL, 0x2D255069F0B7DAB3ULL, 0x9BC5A38EF729ABD4ULL,
0xEF2F054308F6A2BCULL, 0xAF2042F5CC5C2858ULL, 0x480412BAB7F5BE2AULL,
0xAEF3AF4A563DFE43ULL, 0x19AFE59AE451497FULL, 0x52593803DFF1E840ULL,
0xF4F076E65F2CE6F0ULL, 0x11379625747D5AF3ULL, 0xBCE5D2248682C115ULL,
0x9DA4243DE836994FULL, 0x066F70B33FE09017ULL, 0x4DC4DE189B671A1CULL,
0x51039AB7712457C3ULL, 0xC07A3F80C31FB4B4ULL, 0xB46EE9C5E64A6E7CULL,
0xB3819A42ABE61C87ULL, 0x21A007933A522A20ULL, 0x2DF16F761598AA4FULL,
0x763C4A1371B368FDULL, 0xF793C46702E086A0ULL, 0xD7288E012AEB8D31ULL,
0xDE336A2A4BC1C44BULL, 0x0BF692B38D079F23ULL, 0x2C604A7A177326B3ULL,
0x4850E73E03EB6064ULL, 0xCFC447F1E53C8E1BULL, 0xB05CA3F564268D99ULL,
0x9AE182C8BC9474E8ULL, 0xA4FC4BD4FC5558CAULL, 0xE755178D58FC4E76ULL,
0x69B97DB1A4C03DFEULL, 0xF9B5B7C4ACC67C96ULL, 0xFC6A82D64B8655FBULL,
0x9C684CB6C4D24417ULL, 0x8EC97D2917456ED0ULL, 0x6703DF9D2924E97EULL,
0xC547F57E42A7444EULL, 0x78E37644E7CAD29EULL, 0xFE9A44E9362F05FAULL,
0x08BD35CC38336615ULL, 0x9315E5EB3A129ACEULL, 0x94061B871E04DF75ULL,
0xDF1D9F9D784BA010ULL, 0x3BBA57B68871B59DULL, 0xD2B7ADEEDED1F73FULL,
0xF7A255D83BC373F8ULL, 0xD7F4F2448C0CEB81ULL, 0xD95BE88CD210FFA7ULL,
0x336F52F8FF4728E7ULL, 0xA74049DAC312AC71ULL, 0xA2F61BB6E437FDB5ULL,
0x4F2A5CB07F6A35B3ULL, 0x87D380BDA5BF7859ULL, 0x16B9F7E06C453A21ULL,
0x7BA2484C8A0FD54EULL, 0xF3A678CAD9A2E38CULL, 0x39B0BF7DDE437BA2ULL,
0xFCAF55C1BF8A4424ULL, 0x18FCF680573FA594ULL, 0x4C0563B89F495AC3ULL,
0x40E087931A00930DULL, 0x8CFFA9412EB642C1ULL, 0x68CA39053261169FULL,
0x7A1EE967D27579E2ULL, 0x9D1D60E5076F5B6FULL, 0x3810E399B6F65BA2ULL,
0x32095B6D4AB5F9B1ULL, 0x35CAB62109DD038AULL, 0xA90B24499FCFAFB1ULL,
0x77A225A07CC2C6BDULL, 0x513E5E634C70E331ULL, 0x4361C0CA3F692F12ULL,
0xD941ACA44B20A45BULL, 0x528F7C8602C5807BULL, 0x52AB92BEB9613989ULL,
0x9D1DFA2EFC557F73ULL, 0x722FF175F572C348ULL, 0x1D1260A51107FE97ULL,
0x7A249A57EC0C9BA2ULL, 0x04208FE9E8F7F2D6ULL, 0x5A110C6058B920A0ULL,
0x0CD9A497658A5698ULL, 0x56FD23C8F9715A4CULL, 0x284C847B9D887AAEULL,
0x04FEABFBBDB619CBULL, 0x742E1E651C60BA83ULL, 0x9A9632E65904AD3CULL,
0x881B82A13B51B9E2ULL, 0x506E6744CD974924ULL, 0xB0183DB56FFC6A79ULL,
0x0ED9B915C66ED37EULL, 0x5E11E86D5873D484ULL, 0xF678647E3519AC6EULL,
0x1B85D488D0F20CC5ULL, 0xDAB9FE6525D89021ULL, 0x0D151D86ADB73615ULL,
0xA865A54EDCC0F019ULL, 0x93C42566AEF98FFBULL, 0x99E7AFEABE000731ULL,
0x48CBFF086DDF285AULL, 0x7F9B6AF1EBF78BAFULL, 0x58627E1A149BBA21ULL,
0x2CD16E2ABD791E33ULL, 0xD363EFF5F0977996ULL, 0x0CE2A38C344A6EEDULL,
0x1A804AADB9CFA741ULL, 0x907F30421D78C5DEULL, 0x501F65EDB3034D07ULL,
0x37624AE5A48FA6E9ULL, 0x957BAF61700CFF4EULL, 0x3A6C27934E31188AULL,
0xD49503536ABCA345ULL, 0x088E049589C432E0ULL, 0xF943AEE7FEBF21B8ULL,
0x6C3B8E3E336139D3ULL, 0x364F6FFA464EE52EULL, 0xD60F6DCEDC314222ULL,
0x56963B0DCA418FC0ULL, 0x16F50EDF91E513AFULL, 0xEF1955914B609F93ULL,
0x565601C0364E3228ULL, 0xECB53939887E8175ULL, 0xBAC7A9A18531294BULL,
0xB344C470397BBA52ULL, 0x65D34954DAF3CEBDULL, 0xB4B81B3FA97511E2ULL,
0xB422061193D6F6A7ULL, 0x071582401C38434DULL, 0x7A13F18BBEDC4FF5ULL,
0xBC4097B116C524D2ULL, 0x59B97885E2F2EA28ULL, 0x99170A5DC3115544ULL,
0x6F423357E7C6A9F9ULL, 0x325928EE6E6F8794ULL, 0xD0E4366228B03343ULL,
0x565C31F7DE89EA27ULL, 0x30F5611484119414ULL, 0xD873DB391292ED4FULL,
0x7BD94E1D8E17DEBCULL, 0xC7D9F16864A76E94ULL, 0x947AE053EE56E63CULL,
0xC8C93882F9475F5FULL, 0x3A9BF55BA91F81CAULL, 0xD9A11FBB3D9808E4ULL,
0x0FD22063EDC29FCAULL, 0xB3F256D8ACA0B0B9ULL, 0xB03031A8B4516E84ULL,
0x35DD37D5871448AFULL, 0xE9F6082B05542E4EULL, 0xEBFAFA33D7254B59ULL,
0x9255ABB50D532280ULL, 0xB9AB4CE57F2D34F3ULL, 0x693501D628297551ULL,
0xC62C58F97DD949BFULL, 0xCD454F8F19C5126AULL, 0xBBE83F4ECC2BDECBULL,
0xDC842B7E2819E230ULL, 0xBA89142E007503B8ULL, 0xA3BC941D0A5061CBULL,
0xE9F6760E32CD8021ULL, 0x09C7E552BC76492FULL, 0x852F54934DA55CC9ULL,
0x8107FCCF064FCF56ULL, 0x098954D51FFF6580ULL, 0x23B70EDB1955C4BFULL,
0xC330DE426430F69DULL, 0x4715ED43E8A45C0AULL, 0xA8D7E4DAB780A08DULL,
0x0572B974F03CE0BBULL, 0xB57D2E985E1419C7ULL, 0xE8D9ECBE2CF3D73FULL,
0x2FE4B17170E59750ULL, 0x11317BA87905E790ULL, 0x7FBF21EC8A1F45ECULL,
0x1725CABFCB045B00ULL, 0x964E915CD5E2B207ULL, 0x3E2B8BCBF016D66DULL,
0xBE7444E39328A0ACULL, 0xF85B2B4FBCDE44B7ULL, 0x49353FEA39BA63B1ULL,
0x1DD01AAFCD53486AULL, 0x1FCA8A92FD719F85ULL, 0xFC7C95D827357AFAULL,
0x18A6A990C8B35EBDULL, 0xCCCB7005C6B9C28DULL, 0x3BDBB92C43B17F26ULL,
0xAA70B5B4F89695A2ULL, 0xE94C39A54A98307FULL, 0xB7A0B174CFF6F36EULL,
0xD4DBA84729AF48ADULL, 0x2E18BC1AD9704A68ULL, 0x2DE0966DAF2F8B1CULL,
0xB9C11D5B1E43A07EULL, 0x64972D68DEE33360ULL, 0x94628D38D0C20584ULL,
0xDBC0D2B6AB90A559ULL, 0xD2733C4335C6A72FULL, 0x7E75D99D94A70F4DULL,
0x6CED1983376FA72BULL, 0x97FCAACBF030BC24ULL, 0x7B77497B32503B12ULL,
0x8547EDDFB81CCB94ULL, 0x79999CDFF70902CBULL, 0xCFFE1939438E9B24ULL,
0x829626E3892D95D7ULL, 0x92FAE24291F2B3F1ULL, 0x63E22C147B9C3403ULL,
0xC678B6D860284A1CULL, 0x5873888850659AE7ULL, 0x0981DCD296A8736DULL,
0x9F65789A6509A440ULL, 0x9FF38FED72E9052FULL, 0xE479EE5B9930578CULL,
0xE7F28ECD2D49EECDULL, 0x56C074A581EA17FEULL, 0x5544F7D774B14AEFULL,
0x7B3F0195FC6F290FULL, 0x12153635B2C0CF57ULL, 0x7F5126DBBA5E0CA7ULL,
0x7A76956C3EAFB413ULL, 0x3D5774A11D31AB39ULL, 0x8A1B083821F40CB4ULL,
0x7B4A38E32537DF62ULL, 0x950113646D1D6E03ULL, 0x4DA8979A0041E8A9ULL,
0x3BC36E078F7515D7ULL, 0x5D0A12F27AD310D1ULL, 0x7F9D1A2E1EBE1327ULL,
0xDA3A361B1C5157B1ULL, 0xDCDD7D20903D0C25ULL, 0x36833336D068F707ULL,
0xCE68341F79893389ULL, 0xAB9090168DD05F34ULL, 0x43954B3252DC25E5ULL,
0xB438C2B67F98E5E9ULL, 0x10DCD78E3851A492ULL, 0xDBC27AB5447822BFULL,
0x9B3CDB65F82CA382ULL, 0xB67B7896167B4C84ULL, 0xBFCED1B0048EAC50ULL,
0xA9119B60369FFEBDULL, 0x1FFF7AC80904BF45ULL, 0xAC12FB171817EEE7ULL,
0xAF08DA9177DDA93DULL, 0x1B0CAB936E65C744ULL, 0xB559EB1D04E5E932ULL,
0xC37B45B3F8D6F2BAULL, 0xC3A9DC228CAAC9E9ULL, 0xF3B8B6675A6507FFULL,
0x9FC477DE4ED681DAULL, 0x67378D8ECCEF96CBULL, 0x6DD856D94D259236ULL,
0xA319CE15B0B4DB31ULL, 0x073973751F12DD5EULL, 0x8A8E849EB32781A5ULL,
0xE1925C71285279F5ULL, 0x74C04BF1790C0EFEULL, 0x4DDA48153C94938AULL,
0x9D266D6A1CC0542CULL, 0x7440FB816508C4FEULL, 0x13328503DF48229FULL,
0xD6BF7BAEE43CAC40ULL, 0x4838D65F6EF6748FULL, 0x1E152328F3318DEAULL,
0x8F8419A348F296BFULL, 0x72C8834A5957B511ULL, 0xD7A023A73260B45CULL,
0x94EBC8ABCFB56DAEULL, 0x9FC10D0F989993E0ULL, 0xDE68A2355B93CAE6ULL,
0xA44CFE79AE538BBEULL, 0x9D1D84FCCE371425ULL, 0x51D2B1AB2DDFB636ULL,
0x2FD7E4B9E72CD38CULL, 0x65CA5B96B7552210ULL, 0xDD69A0D8AB3B546DULL,
0x604D51B25FBF70E2ULL, 0x73AA8A564FB7AC9EULL, 0x1A8C1E992B941148ULL,
0xAAC40A2703D9BEA0ULL, 0x764DBEAE7FA4F3A6ULL, 0x1E99B96E70A9BE8BULL,
0x2C5E9DEB57EF4743ULL, 0x3A938FEE32D29981ULL, 0x26E6DB8FFDF5ADFEULL,
0x469356C504EC9F9DULL, 0xC8763C5B08D1908CULL, 0x3F6C6AF859D80055ULL,
0x7F7CC39420A3A545ULL, 0x9BFB227EBDF4C5CEULL, 0x89039D79D6FC5C5CULL,
0x8FE88B57305E2AB6ULL, 0xA09E8C8C35AB96DEULL, 0xFA7E393983325753ULL,
0xD6B6D0ECC617C699ULL, 0xDFEA21EA9E7557E3ULL, 0xB67C1FA481680AF8ULL,
0xCA1E3785A9E724E5ULL, 0x1CFC8BED0D681639ULL, 0xD18D8549D140CAEAULL,
0x4ED0FE7E9DC91335ULL, 0xE4DBF0634473F5D2ULL, 0x1761F93A44D5AEFEULL,
0x53898E4C3910DA55ULL, 0x734DE8181F6EC39AULL, 0x2680B122BAA28D97ULL,
0x298AF231C85BAFABULL, 0x7983EED3740847D5ULL, 0x66C1A2A1A60CD889ULL,
0x9E17E49642A3E4C1ULL, 0xEDB454E7BADC0805ULL, 0x50B704CAB602C329ULL,
0x4CC317FB9CDDD023ULL, 0x66B4835D9EAFEA22ULL, 0x219B97E26FFC81BDULL,
0x261E4E4C0A333A9DULL, 0x1FE2CCA76517DB90ULL, 0xD7504DFA8816EDBBULL,
0xB9571FA04DC089C8ULL, 0x1DDC0325259B27DEULL, 0xCF3F4688801EB9AAULL,
0xF4F5D05C10CAB243ULL, 0x38B6525C21A42B0EULL, 0x36F60E2BA4FA6800ULL,
0xEB3593803173E0CEULL, 0x9C4CD6257C5A3603ULL, 0xAF0C317D32ADAA8AULL,
0x258E5A80C7204C4BULL, 0x8B889D624D44885DULL, 0xF4D14597E660F855ULL,
0xD4347F66EC8941C3ULL, 0xE699ED85B0DFB40DULL, 0x2472F6207C2D0484ULL,
0xC2A1E7B5B459AEB5ULL, 0xAB4F6451CC1D45ECULL, 0x63767572AE3D6174ULL,
0xA59E0BD101731A28ULL, 0x116D0016CB948F09ULL, 0x2CF9C8CA052F6E9FULL,
0x0B090A7560A968E3ULL, 0xABEEDDB2DDE06FF1ULL, 0x58EFC10B06A2068DULL,
0xC6E57A78FBD986E0ULL, 0x2EAB8CA63CE802D7ULL, 0x14A195640116F336ULL,
0x7C0828DD624EC390ULL, 0xD74BBE77E6116AC7ULL, 0x804456AF10F5FB53ULL,
0xEBE9EA2ADF4321C7ULL, 0x03219A39EE587A30ULL, 0x49787FEF17AF9924ULL,
0xA1E9300CD8520548ULL, 0x5B45E522E4B1B4EFULL, 0xB49C3B3995091A36ULL,
0xD4490AD526F14431ULL, 0x12A8F216AF9418C2ULL, 0x001F837CC7350524ULL,
0x1877B51E57A764D5ULL, 0xA2853B80F17F58EEULL, 0x993E1DE72D36D310ULL,
0xB3598080CE64A656ULL, 0x252F59CF0D9F04BBULL, 0xD23C8E176D113600ULL,
0x1BDA0492E7E4586EULL, 0x21E0BD5026C619BFULL, 0x3B097ADAF088F94EULL,
0x8D14DEDB30BE846EULL, 0xF95CFFA23AF5F6F4ULL, 0x3871700761B3F743ULL,
0xCA672B91E9E4FA16ULL, 0x64C8E531BFF53B55ULL, 0x241260ED4AD1E87DULL,
0x106C09B972D2E822ULL, 0x7FBA195410E5CA30ULL, 0x7884D9BC6CB569D8ULL,
0x0647DFEDCD894A29ULL, 0x63573FF03E224774ULL, 0x4FC8E9560F91B123ULL,
0x1DB956E450275779ULL, 0xB8D91274B9E9D4FBULL, 0xA2EBEE47E2FBFCE1ULL,
0xD9F1F30CCD97FB09ULL, 0xEFED53D75FD64E6BULL, 0x2E6D02C36017F67FULL,
0xA9AA4D20DB084E9BULL, 0xB64BE8D8B25396C1ULL, 0x70CB6AF7C2D5BCF0ULL,
0x98F076A4F7A2322EULL, 0xBF84470805E69B5FULL, 0x94C3251F06F90CF3ULL,
0x3E003E616A6591E9ULL, 0xB925A6CD0421AFF3ULL, 0x61BDD1307C66E300ULL,
0xBF8D5108E27E0D48ULL, 0x240AB57A8B888B20ULL, 0xFC87614BAF287E07ULL,
0xEF02CDD06FFDB432ULL, 0xA1082C0466DF6C0AULL, 0x8215E577001332C8ULL,
0xD39BB9C3A48DB6CFULL, 0x2738259634305C14ULL, 0x61CF4F94C97DF93DULL,
0x1B6BACA2AE4E125BULL, 0x758F450C88572E0BULL, 0x959F587D507A8359ULL,
0xB063E962E045F54DULL, 0x60E8ED72C0DFF5D1ULL, 0x7B64978555326F9FULL,
0xFD080D236DA814BAULL, 0x8C90FD9B083F4558ULL, 0x106F72FE81E2C590ULL,
0x7976033A39F7D952ULL, 0xA4EC0132764CA04BULL, 0x733EA705FAE4FA77ULL,
0xB4D8F77BC3E56167ULL, 0x9E21F4F903B33FD9ULL, 0x9D765E419FB69F6DULL,
0xD30C088BA61EA5EFULL, 0x5D94337FBFAF7F5BULL, 0x1A4E4822EB4D7A59ULL,
0x6FFE73E81B637FB3ULL, 0xDDF957BC36D8B9CAULL, 0x64D0E29EEA8838B3ULL,
0x08DD9BDFD96B9F63ULL, 0x087E79E5A57D1D13ULL, 0xE328E230E3E2B3FBULL,
0x1C2559E30F0946BEULL, 0x720BF5F26F4D2EAAULL, 0xB0774D261CC609DBULL,
0x443F64EC5A371195ULL, 0x4112CF68649A260EULL, 0xD813F2FAB7F5C5CAULL,
0x660D3257380841EEULL, 0x59AC2C7873F910A3ULL, 0xE846963877671A17ULL,
0x93B633ABFA3469F8ULL, 0xC0C0F5A60EF4CDCFULL, 0xCAF21ECD4377B28CULL,
0x57277707199B8175ULL, 0x506C11B9D90E8B1DULL, 0xD83CC2687A19255FULL,
0x4A29C6465A314CD1ULL, 0xED2DF21216235097ULL, 0xB5635C95FF7296E2ULL,
0x22AF003AB672E811ULL, 0x52E762596BF68235ULL, 0x9AEBA33AC6ECC6B0ULL,
0x944F6DE09134DFB6ULL, 0x6C47BEC883A7DE39ULL, 0x6AD047C430A12104ULL,
0xA5B1CFDBA0AB4067ULL, 0x7C45D833AFF07862ULL, 0x5092EF950A16DA0BULL,
0x9338E69C052B8E7BULL, 0x455A4B4CFE30E3F5ULL, 0x6B02E63195AD0CF8ULL,
0x6B17B224BAD6BF27ULL, 0xD1E0CCD25BB9C169ULL, 0xDE0C89A556B9AE70ULL,
0x50065E535A213CF6ULL, 0x9C1169FA2777B874ULL, 0x78EDEFD694AF1EEDULL,
0x6DC93D9526A50E68ULL, 0xEE97F453F06791EDULL, 0x32AB0EDB696703D3ULL,
0x3A6853C7E70757A7ULL, 0x31865CED6120F37DULL, 0x67FEF95D92607890ULL,
0x1F2B1D1F15F6DC9CULL, 0xB69E38A8965C6B65ULL, 0xAA9119FF184CCCF4ULL,
0xF43C732873F24C13ULL, 0xFB4A3D794A9A80D2ULL, 0x3550C2321FD6109CULL,
0x371F77E76BB8417EULL, 0x6BFA9AAE5EC05779ULL, 0xCD04F3FF001A4778ULL,
0xE3273522064480CAULL, 0x9F91508BFFCFC14AULL, 0x049A7F41061A9E60ULL,
0xFCB6BE43A9F2FE9BULL, 0x08DE8A1C7797DA9BULL, 0x8F9887E6078735A1ULL,
0xB5B4071DBFC73A66ULL, 0x230E343DFBA08D33ULL, 0x43ED7F5A0FAE657DULL,
0x3A88A0FBBCB05C63ULL, 0x21874B8B4D2DBC4FULL, 0x1BDEA12E35F6A8C9ULL,
0x53C065C6C8E63528ULL, 0xE34A1D250E7A8D6BULL, 0xD6B04D3B7651DD7EULL,
0x5E90277E7CB39E2DULL, 0x2C046F22062DC67DULL, 0xB10BB459132D0A26ULL,
0x3FA9DDFB67E2F199ULL, 0x0E09B88E1914F7AFULL, 0x10E8B35AF3EEAB37ULL,
0x9EEDECA8E272B933ULL, 0xD4C718BC4AE8AE5FULL, 0x81536D601170FC20ULL,
0x91B534F885818A06ULL, 0xEC8177F83F900978ULL, 0x190E714FADA5156EULL,
0xB592BF39B0364963ULL, 0x89C350C893AE7DC1ULL, 0xAC042E70F8B383F2ULL,
0xB49B52E587A1EE60ULL, 0xFB152FE3FF26DA89ULL, 0x3E666E6F69AE2C15ULL,
0x3B544EBE544C19F9ULL, 0xE805A1E290CF2456ULL, 0x24B33C9D7ED25117ULL,
0xE74733427B72F0C1ULL, 0x0A804D18B7097475ULL, 0x57E3306D881EDB4FULL,
0x4AE7D6A36EB5DBCBULL, 0x2D8D5432157064C8ULL, 0xD1E649DE1E7F268BULL,
0x8A328A1CEDFE552CULL, 0x07A3AEC79624C7DAULL, 0x84547DDC3E203C94ULL,
0x990A98FD5071D263ULL, 0x1A4FF12616EEFC89ULL, 0xF6F7FD1431714200ULL,
0x30C05B1BA332F41CULL, 0x8D2636B81555A786ULL, 0x46C9FEB55D120902ULL,
0xCCEC0A73B49C9921ULL, 0x4E9D2827355FC492ULL, 0x19EBB029435DCB0FULL,
0x4659D2B743848A2CULL, 0x963EF2C96B33BE31ULL, 0x74F85198B05A2E7DULL,
0x5A0F544DD2B1FB18ULL, 0x03727073C2E134B1ULL, 0xC7F6AA2DE59AEA61ULL,
0x352787BAA0D7C22FULL, 0x9853EAB63B5E0B35ULL, 0xABBDCDD7ED5C0860ULL,
0xCF05DAF5AC8D77B0ULL, 0x49CAD48CEBF4A71EULL, 0x7A4C10EC2158C4A6ULL,
0xD9E92AA246BF719EULL, 0x13AE978D09FE5557ULL, 0x730499AF921549FFULL,
0x4E4B705B92903BA4ULL, 0xFF577222C14F0A3AULL, 0x55B6344CF97AAFAEULL,
0xB862225B055B6960ULL, 0xCAC09AFBDDD2CDB4ULL, 0xDAF8E9829FE96B5FULL,
0xB5FDFC5D3132C498ULL, 0x310CB380DB6F7503ULL, 0xE87FBB46217A360EULL,
0x2102AE466EBB1148ULL, 0xF8549E1A3AA5E00DULL, 0x07A69AFDCC42261AULL,
0xC4C118BFE78FEAAEULL, 0xF9F4892ED96BD438ULL, 0x1AF3DBE25D8F45DAULL,
0xF5B4B0B0D2DEEEB4ULL, 0x962ACEEFA82E1C84ULL, 0x046E3ECAAF453CE9ULL,
0xF05D129681949A4CULL, 0x964781CE734B3C84ULL, 0x9C2ED44081CE5FBDULL,
0x522E23F3925E319EULL, 0x177E00F9FC32F791ULL, 0x2BC60A63A6F3B3F2ULL,
0x222BBFAE61725606ULL, 0x486289DDCC3D6780ULL, 0x7DC7785B8EFDFC80ULL,
0x8AF38731C02BA980ULL, 0x1FAB64EA29A2DDF7ULL, 0xE4D9429322CD065AULL,
0x9DA058C67844F20CULL, 0x24C0E332B70019B0ULL, 0x233003B5A6CFE6ADULL,
0xD586BD01C5C217F6ULL, 0x5E5637885F29BC2BULL, 0x7EBA726D8C94094BULL,
0x0A56A5F0BFE39272ULL, 0xD79476A84EE20D06ULL, 0x9E4C1269BAA4BF37ULL,
0x17EFEE45B0DEE640ULL, 0x1D95B0A5FCF90BC6ULL, 0x93CBE0B699C2585DULL,
0x65FA4F227A2B6D79ULL, 0xD5F9E858292504D5ULL, 0xC2B5A03F71471A6FULL,
0x59300222B4561E00ULL, 0xCE2F8642CA0712DCULL, 0x7CA9723FBB2E8988ULL,
0x2785338347F2BA08ULL, 0xC61BB3A141E50E8CULL, 0x150F361DAB9DEC26ULL,
0x9F6A419D382595F4ULL, 0x64A53DC924FE7AC9ULL, 0x142DE49FFF7A7C3DULL,
0x0C335248857FA9E7ULL, 0x0A9C32D5EAE45305ULL, 0xE6C42178C4BBB92EULL,
0x71F1CE2490D20B07ULL, 0xF1BCC3D275AFE51AULL, 0xE728E8C83C334074ULL,
0x96FBF83A12884624ULL, 0x81A1549FD6573DA5ULL, 0x5FA7867CAF35E149ULL,
0x56986E2EF3ED091BULL, 0x917F1DD5F8886C61ULL, 0xD20D8C88C8FFE65FULL,
0x31D71DCE64B2C310ULL, 0xF165B587DF898190ULL, 0xA57E6339DD2CF3A0ULL,
0x1EF6E6DBB1961EC9ULL, 0x70CC73D90BC26E24ULL, 0xE21A6B35DF0C3AD7ULL,
0x003A93D8B2806962ULL, 0x1C99DED33CB890A1ULL, 0xCF3145DE0ADD4289ULL,
0xD0E4427A5514FB72ULL, 0x77C621CC9FB3A483ULL, 0x67A34DAC4356550BULL,
0xF8D626AAAF278509ULL
}};
// polyglot_key() returns the PolyGlot hash key of the given position
Key polyglot_key(const Position& pos) {
Key key = 0;
Bitboard b = pos.pieces();
while (b)
{
Square s = pop_lsb(&b);
Piece p = pos.piece_on(s);
// PolyGlot pieces are: BP = 0, WP = 1, BN = 2, ... BK = 10, WK = 11
key ^= PG.Zobrist.psq[2 * (type_of(p) - 1) + (color_of(p) == WHITE)][s];
}
b = pos.can_castle(ALL_CASTLES);
while (b)
key ^= PG.Zobrist.castle[pop_lsb(&b)];
if (pos.ep_square() != SQ_NONE)
key ^= PG.Zobrist.enpassant[file_of(pos.ep_square())];
if (pos.side_to_move() == WHITE)
key ^= PG.Zobrist.turn;
return key;
}
} // namespace
PolyglotBook::PolyglotBook() : rkiss(Time::now() % 10000) {}
PolyglotBook::~PolyglotBook() { if (is_open()) close(); }
/// operator>>() reads sizeof(T) chars from the file's binary byte stream and
/// converts them in a number of type T. A Polyglot book stores numbers in
/// big-endian format.
template<typename T> PolyglotBook& PolyglotBook::operator>>(T& n) {
n = 0;
for (size_t i = 0; i < sizeof(T); i++)
n = T((n << 8) + ifstream::get());
return *this;
}
template<> PolyglotBook& PolyglotBook::operator>>(Entry& e) {
return *this >> e.key >> e.move >> e.count >> e.learn;
}
/// open() tries to open a book file with the given name after closing any
/// exsisting one.
bool PolyglotBook::open(const char* fName) {
if (is_open()) // Cannot close an already closed file
close();
ifstream::open(fName, ifstream::in | ifstream::binary);
fileName = is_open() ? fName : "";
ifstream::clear(); // Reset any error flag to allow retry ifstream::open()
return !fileName.empty();
}
/// probe() tries to find a book move for the given position. If no move is
/// found returns MOVE_NONE. If pickBest is true returns always the highest
/// rated move, otherwise randomly chooses one, based on the move score.
Move PolyglotBook::probe(const Position& pos, const string& fName, bool pickBest) {
if (fileName != fName && !open(fName.c_str()))
return MOVE_NONE;
Entry e;
uint16_t best = 0;
unsigned sum = 0;
Move move = MOVE_NONE;
Key key = polyglot_key(pos);
seekg(find_first(key) * sizeof(Entry), ios_base::beg);
while (*this >> e, e.key == key && good())
{
best = max(best, e.count);
sum += e.count;
// Choose book move according to its score. If a move has a very
// high score it has higher probability to be choosen than a move
// with lower score. Note that first entry is always chosen.
if ( (sum && rkiss.rand<unsigned>() % sum < e.count)
|| (pickBest && e.count == best))
move = Move(e.move);
}
if (!move)
return MOVE_NONE;
// A PolyGlot book move is encoded as follows:
//
// bit 0- 5: destination square (from 0 to 63)
// bit 6-11: origin square (from 0 to 63)
// bit 12-14: promotion piece (from KNIGHT == 1 to QUEEN == 4)
//
// Castling moves follow "king captures rook" representation. So in case book
// move is a promotion we have to convert to our representation, in all the
// other cases we can directly compare with a Move after having masked out
// the special Move's flags (bit 14-15) that are not supported by PolyGlot.
int pt = (move >> 12) & 7;
if (pt)
move = make<PROMOTION>(from_sq(move), to_sq(move), PieceType(pt + 1));
// Add 'special move' flags and verify it is legal
for (MoveList<LEGAL> it(pos); *it; ++it)
if (move == (*it ^ type_of(*it)))
return *it;
return MOVE_NONE;
}
/// find_first() takes a book key as input, and does a binary search through
/// the book file for the given key. Returns the index of the leftmost book
/// entry with the same key as the input.
size_t PolyglotBook::find_first(Key key) {
seekg(0, ios::end); // Move pointer to end, so tellg() gets file's size
size_t low = 0, mid, high = (size_t)tellg() / sizeof(Entry) - 1;
Entry e;
assert(low <= high);
while (low < high && good())
{
mid = (low + high) / 2;
assert(mid >= low && mid < high);
seekg(mid * sizeof(Entry), ios_base::beg);
*this >> e;
if (key <= e.key)
high = mid;
else
low = mid + 1;
}
assert(low == high);
return low;
}
-45
View File
@@ -1,45 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef BOOK_H_INCLUDED
#define BOOK_H_INCLUDED
#include <fstream>
#include <string>
#include "position.h"
#include "rkiss.h"
class PolyglotBook : private std::ifstream {
public:
PolyglotBook();
~PolyglotBook();
Move probe(const Position& pos, const std::string& fName, bool pickBest);
private:
template<typename T> PolyglotBook& operator>>(T& n);
bool open(const char* fName);
size_t find_first(Key key);
RKISS rkiss;
std::string fileName;
};
#endif // #ifndef BOOK_H_INCLUDED
+362 -555
View File
File diff suppressed because it is too large Load Diff
+66 -46
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -21,21 +22,20 @@
#define ENDGAME_H_INCLUDED #define ENDGAME_H_INCLUDED
#include <map> #include <map>
#include <memory>
#include <string> #include <string>
#include <type_traits>
#include <utility>
#include "position.h" #include "position.h"
#include "types.h" #include "types.h"
/// EndgameType lists all supported endgames /// EndgameCode lists all supported endgame functions by corresponding codes
enum EndgameType { enum EndgameCode {
// Evaluation functions EVALUATION_FUNCTIONS,
KK, // K vs K
KBK, // KB vs K
KNK, // KN vs K
KNNK, // KNN vs K KNNK, // KNN vs K
KXK, // Generic "mate lone king" eval KXK, // Generic "mate lone king" eval
KBNK, // KBN vs K KBNK, // KBN vs K
@@ -45,18 +45,14 @@ enum EndgameType {
KRKN, // KR vs KN KRKN, // KR vs KN
KQKP, // KQ vs KP KQKP, // KQ vs KP
KQKR, // KQ vs KR KQKR, // KQ vs KR
KBBKN, // KBB vs KN
KmmKm, // K and two minors vs K and one or two minors
SCALING_FUNCTIONS,
// Scaling functions KBPsK, // KB and pawns vs K
SCALE_FUNS, KQKRPs, // KQ vs KR and pawns
KBPsK, // KB+pawns vs K
KQKRPs, // KQ vs KR+pawns
KRPKR, // KRP vs KR KRPKR, // KRP vs KR
KRPKB, // KRP vs KB
KRPPKRP, // KRPP vs KRP KRPPKRP, // KRPP vs KRP
KPsK, // King and pawns vs king KPsK, // K and pawns vs K
KBPKB, // KBP vs KB KBPKB, // KBP vs KB
KBPPKB, // KBPP vs KB KBPPKB, // KBPP vs KB
KBPKN, // KBP vs KN KBPKN, // KBP vs KN
@@ -66,60 +62,84 @@ enum EndgameType {
}; };
/// Endgame functions can be of two types according if return a Value or a /// Endgame functions can be of two types depending on whether they return a
/// ScaleFactor. Type eg_fun<int>::type equals to either ScaleFactor or Value /// Value or a ScaleFactor.
/// depending if the template parameter is 0 or 1.
template<int> struct eg_fun { typedef Value type; }; template<EndgameCode E> using
template<> struct eg_fun<1> { typedef ScaleFactor type; }; eg_type = typename std::conditional<(E < SCALING_FUNCTIONS), Value, ScaleFactor>::type;
/// Base and derived templates for endgame evaluation and scaling functions /// Base and derived functors for endgame evaluation and scaling functions
template<typename T> template<typename T>
struct EndgameBase { struct EndgameBase {
virtual ~EndgameBase() {} explicit EndgameBase(Color c) : strongSide(c), weakSide(~c) {}
virtual Color color() const = 0; virtual ~EndgameBase() = default;
virtual T operator()(const Position&) const = 0; virtual T operator()(const Position&) const = 0;
const Color strongSide, weakSide;
}; };
template<EndgameType E, typename T = typename eg_fun<(E > SCALE_FUNS)>::type> template<EndgameCode E, typename T = eg_type<E>>
struct Endgame : public EndgameBase<T> { struct Endgame : public EndgameBase<T> {
explicit Endgame(Color c) : strongerSide(c), weakerSide(~c) {} explicit Endgame(Color c) : EndgameBase<T>(c) {}
Color color() const { return strongerSide; } T operator()(const Position&) const override;
T operator()(const Position&) const;
private:
Color strongerSide, weakerSide;
}; };
/// Endgames class stores in two std::map the pointers to endgame evaluation /// The Endgames class stores the pointers to endgame evaluation and scaling
/// and scaling base objects. Then we use polymorphism to invoke the actual /// base objects in two std::map. We use polymorphism to invoke the actual
/// endgame function calling its operator() that is virtual. /// endgame function by calling its virtual operator().
class Endgames { class Endgames {
typedef std::map<Key, EndgameBase<eg_fun<0>::type>*> M1; template<typename T> using Ptr = std::unique_ptr<EndgameBase<T>>;
typedef std::map<Key, EndgameBase<eg_fun<1>::type>*> M2; template<typename T> using Map = std::map<Key, Ptr<T>>;
M1 m1; template<typename T>
M2 m2; Map<T>& map() {
return std::get<std::is_same<T, ScaleFactor>::value>(maps);
}
M1& map(M1::mapped_type) { return m1; } template<EndgameCode E, typename T = eg_type<E>>
M2& map(M2::mapped_type) { return m2; } void add(const std::string& code) {
template<EndgameType E> void add(const std::string& code); StateInfo st;
map<T>()[Position().set(code, WHITE, &st).material_key()] = Ptr<T>(new Endgame<E>(WHITE));
map<T>()[Position().set(code, BLACK, &st).material_key()] = Ptr<T>(new Endgame<E>(BLACK));
}
std::pair<Map<Value>, Map<ScaleFactor>> maps;
public: public:
Endgames(); Endgames() {
~Endgames();
template<typename T> T probe(Key key, T& eg) add<KPK>("KPK");
{ return eg = map(eg).count(key) ? map(eg)[key] : NULL; } add<KNNK>("KNNK");
add<KBNK>("KBNK");
add<KRKP>("KRKP");
add<KRKB>("KRKB");
add<KRKN>("KRKN");
add<KQKP>("KQKP");
add<KQKR>("KQKR");
add<KNPK>("KNPK");
add<KNPKB>("KNPKB");
add<KRPKR>("KRPKR");
add<KRPKB>("KRPKB");
add<KBPKB>("KBPKB");
add<KBPKN>("KBPKN");
add<KBPPKB>("KBPPKB");
add<KRPPKRP>("KRPPKRP");
}
template<typename T>
const EndgameBase<T>* probe(Key key) {
return map<T>().count(key) ? map<T>()[key].get() : nullptr;
}
}; };
#endif // #ifndef ENDGAME_H_INCLUDED #endif // #ifndef ENDGAME_H_INCLUDED
+715 -981
View File
File diff suppressed because it is too large Load Diff
+8 -4
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,16 +21,19 @@
#ifndef EVALUATE_H_INCLUDED #ifndef EVALUATE_H_INCLUDED
#define EVALUATE_H_INCLUDED #define EVALUATE_H_INCLUDED
#include <string>
#include "types.h" #include "types.h"
class Position; class Position;
namespace Eval { namespace Eval {
extern void init(); constexpr Value Tempo = Value(20); // Must be visible to search
extern Value evaluate(const Position& pos, Value& margin);
extern std::string trace(const Position& pos);
std::string trace(const Position& pos);
Value evaluate(const Position& pos);
} }
#endif // #ifndef EVALUATE_H_INCLUDED #endif // #ifndef EVALUATE_H_INCLUDED
+16 -15
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -18,35 +19,35 @@
*/ */
#include <iostream> #include <iostream>
#include <string>
#include "bitboard.h" #include "bitboard.h"
#include "evaluate.h"
#include "position.h" #include "position.h"
#include "search.h" #include "search.h"
#include "thread.h" #include "thread.h"
#include "tt.h" #include "tt.h"
#include "ucioption.h" #include "uci.h"
#include "syzygy/tbprobe.h"
namespace PSQT {
void init();
}
int main(int argc, char* argv[]) { int main(int argc, char* argv[]) {
std::cout << engine_info() << std::endl; std::cout << engine_info() << std::endl;
UCI::init(Options); UCI::init(Options);
PSQT::init();
Bitboards::init(); Bitboards::init();
Position::init(); Position::init();
Bitbases::init_kpk(); Bitbases::init();
Search::init(); Search::init();
Eval::init(); Pawns::init();
Threads.init(); Threads.set(Options["Threads"]);
TT.set_size(Options["Hash"]); Search::clear(); // After threads are up
std::string args; UCI::loop(argc, argv);
for (int i = 1; i < argc; i++) Threads.set(0);
args += std::string(argv[i]) + " "; return 0;
UCI::loop(args);
Threads.exit();
} }
+106 -176
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -19,52 +20,41 @@
#include <algorithm> // For std::min #include <algorithm> // For std::min
#include <cassert> #include <cassert>
#include <cstring> #include <cstring> // For std::memset
#include "material.h" #include "material.h"
#include "thread.h"
using namespace std; using namespace std;
namespace { namespace {
// Values modified by Joona Kiiski // Polynomial material imbalance parameters
const Value MidgameLimit = Value(15581);
const Value EndgameLimit = Value(3998);
// Scale factors used when one side has no more pawns
const int NoPawnsSF[4] = { 6, 12, 32 };
// Polynomial material balance parameters
const Value RedundantQueen = Value(320);
const Value RedundantRook = Value(554);
constexpr int QuadraticOurs[][PIECE_TYPE_NB] = {
// OUR PIECES
// pair pawn knight bishop rook queen // pair pawn knight bishop rook queen
const int LinearCoefficients[6] = { 1617, -162, -1172, -190, 105, 26 }; {1438 }, // Bishop pair
{ 40, 38 }, // Pawn
const int QuadraticCoefficientsSameColor[][PIECE_TYPE_NB] = { { 32, 255, -62 }, // Knight OUR PIECES
// pair pawn knight bishop rook queen { 0, 104, 4, 0 }, // Bishop
{ 7 }, // Bishop pair { -26, -2, 47, 105, -208 }, // Rook
{ 39, 2 }, // Pawn {-189, 24, 117, 133, -134, -6 } // Queen
{ 35, 271, -4 }, // Knight
{ 7, 105, 4, 7 }, // Bishop
{ -27, -2, 46, 100, 56 }, // Rook
{ 58, 29, 83, 148, -3, -25 } // Queen
}; };
const int QuadraticCoefficientsOppositeColor[][PIECE_TYPE_NB] = { constexpr int QuadraticTheirs[][PIECE_TYPE_NB] = {
// THEIR PIECES // THEIR PIECES
// pair pawn knight bishop rook queen // pair pawn knight bishop rook queen
{ 41 }, // Bishop pair { 0 }, // Bishop pair
{ 37, 41 }, // Pawn { 36, 0 }, // Pawn
{ 10, 62, 41 }, // Knight OUR PIECES { 9, 63, 0 }, // Knight OUR PIECES
{ 57, 64, 39, 41 }, // Bishop { 59, 65, 42, 0 }, // Bishop
{ 50, 40, 23, -22, 41 }, // Rook { 46, 39, 24, -24, 0 }, // Rook
{ 106, 101, 3, 151, 171, 41 } // Queen { 97, 100, -42, 137, 268, 0 } // Queen
}; };
// Endgame evaluation and scaling functions accessed direcly and not through // Endgame evaluation and scaling functions are accessed directly and not through
// the function maps because correspond to more then one material hash key. // the function maps because they correspond to more than one material hash key.
Endgame<KmmKm> EvaluateKmmKm[] = { Endgame<KmmKm>(WHITE), Endgame<KmmKm>(BLACK) };
Endgame<KXK> EvaluateKXK[] = { Endgame<KXK>(WHITE), Endgame<KXK>(BLACK) }; Endgame<KXK> EvaluateKXK[] = { Endgame<KXK>(WHITE), Endgame<KXK>(BLACK) };
Endgame<KBPsK> ScaleKBPsK[] = { Endgame<KBPsK>(WHITE), Endgame<KBPsK>(BLACK) }; Endgame<KBPsK> ScaleKBPsK[] = { Endgame<KBPsK>(WHITE), Endgame<KBPsK>(BLACK) };
@@ -72,163 +62,128 @@ namespace {
Endgame<KPsK> ScaleKPsK[] = { Endgame<KPsK>(WHITE), Endgame<KPsK>(BLACK) }; Endgame<KPsK> ScaleKPsK[] = { Endgame<KPsK>(WHITE), Endgame<KPsK>(BLACK) };
Endgame<KPKP> ScaleKPKP[] = { Endgame<KPKP>(WHITE), Endgame<KPKP>(BLACK) }; Endgame<KPKP> ScaleKPKP[] = { Endgame<KPKP>(WHITE), Endgame<KPKP>(BLACK) };
// Helper templates used to detect a given material distribution // Helper used to detect a given material distribution
template<Color Us> bool is_KXK(const Position& pos) { bool is_KXK(const Position& pos, Color us) {
const Color Them = (Us == WHITE ? BLACK : WHITE); return !more_than_one(pos.pieces(~us))
return !pos.count<PAWN>(Them) && pos.non_pawn_material(us) >= RookValueMg;
&& pos.non_pawn_material(Them) == VALUE_ZERO
&& pos.non_pawn_material(Us) >= RookValueMg;
} }
template<Color Us> bool is_KBPsKs(const Position& pos) { bool is_KBPsK(const Position& pos, Color us) {
return pos.non_pawn_material(Us) == BishopValueMg return pos.non_pawn_material(us) == BishopValueMg
&& pos.count<BISHOP>(Us) == 1 && pos.count<BISHOP>(us) == 1
&& pos.count<PAWN >(Us) >= 1; && pos.count<PAWN >(us) >= 1;
} }
template<Color Us> bool is_KQKRPs(const Position& pos) { bool is_KQKRPs(const Position& pos, Color us) {
const Color Them = (Us == WHITE ? BLACK : WHITE); return !pos.count<PAWN>(us)
return !pos.count<PAWN>(Us) && pos.non_pawn_material(us) == QueenValueMg
&& pos.non_pawn_material(Us) == QueenValueMg && pos.count<QUEEN>(us) == 1
&& pos.count<QUEEN>(Us) == 1 && pos.count<ROOK>(~us) == 1
&& pos.count<ROOK>(Them) == 1 && pos.count<PAWN>(~us) >= 1;
&& pos.count<PAWN>(Them) >= 1;
} }
/// imbalance() calculates imbalance comparing piece count of each /// imbalance() calculates the imbalance by comparing the piece count of each
/// piece type for both colors. /// piece type for both colors.
template<Color Us> template<Color Us>
int imbalance(const int pieceCount[][PIECE_TYPE_NB]) { int imbalance(const int pieceCount[][PIECE_TYPE_NB]) {
const Color Them = (Us == WHITE ? BLACK : WHITE); constexpr Color Them = (Us == WHITE ? BLACK : WHITE);
int pt1, pt2, pc, v; int bonus = 0;
int value = 0;
// Redundancy of major pieces, formula based on Kaufman's paper // Second-degree polynomial material imbalance, by Tord Romstad
// "The Evaluation of Material Imbalances in Chess" for (int pt1 = NO_PIECE_TYPE; pt1 <= QUEEN; ++pt1)
if (pieceCount[Us][ROOK] > 0)
value -= RedundantRook * (pieceCount[Us][ROOK] - 1)
+ RedundantQueen * pieceCount[Us][QUEEN];
// Second-degree polynomial material imbalance by Tord Romstad
for (pt1 = NO_PIECE_TYPE; pt1 <= QUEEN; pt1++)
{ {
pc = pieceCount[Us][pt1]; if (!pieceCount[Us][pt1])
if (!pc)
continue; continue;
v = LinearCoefficients[pt1]; int v = 0;
for (pt2 = NO_PIECE_TYPE; pt2 <= pt1; pt2++) for (int pt2 = NO_PIECE_TYPE; pt2 <= pt1; ++pt2)
v += QuadraticCoefficientsSameColor[pt1][pt2] * pieceCount[Us][pt2] v += QuadraticOurs[pt1][pt2] * pieceCount[Us][pt2]
+ QuadraticCoefficientsOppositeColor[pt1][pt2] * pieceCount[Them][pt2]; + QuadraticTheirs[pt1][pt2] * pieceCount[Them][pt2];
value += pc * v; bonus += pieceCount[Us][pt1] * v;
} }
return value;
return bonus;
} }
} // namespace } // namespace
namespace Material { namespace Material {
/// Material::probe() takes a position object as input, looks up a MaterialEntry /// Material::probe() looks up the current position's material configuration in
/// object, and returns a pointer to it. If the material configuration is not /// the material hash table. It returns a pointer to the Entry if the position
/// already present in the table, it is computed and stored there, so we don't /// is found. Otherwise a new Entry is computed and stored there, so we don't
/// have to recompute everything when the same material configuration occurs again. /// have to recompute all when the same material configuration occurs again.
Entry* probe(const Position& pos, Table& entries, Endgames& endgames) { Entry* probe(const Position& pos) {
Key key = pos.material_key(); Key key = pos.material_key();
Entry* e = entries[key]; Entry* e = pos.this_thread()->materialTable[key];
// If e->key matches the position's material hash key, it means that we
// have analysed this material configuration before, and we can simply
// return the information we found the last time instead of recomputing it.
if (e->key == key) if (e->key == key)
return e; return e;
std::memset(e, 0, sizeof(Entry)); std::memset(e, 0, sizeof(Entry));
e->key = key; e->key = key;
e->factor[WHITE] = e->factor[BLACK] = (uint8_t)SCALE_FACTOR_NORMAL; e->factor[WHITE] = e->factor[BLACK] = (uint8_t)SCALE_FACTOR_NORMAL;
e->gamePhase = game_phase(pos);
// Let's look if we have a specialized evaluation function for this
// particular material configuration. First we look for a fixed
// configuration one, then a generic one if previous search failed.
if (endgames.probe(key, e->evaluationFunction))
return e;
if (is_KXK<WHITE>(pos))
{
e->evaluationFunction = &EvaluateKXK[WHITE];
return e;
}
if (is_KXK<BLACK>(pos))
{
e->evaluationFunction = &EvaluateKXK[BLACK];
return e;
}
if (!pos.pieces(PAWN) && !pos.pieces(ROOK) && !pos.pieces(QUEEN))
{
// Minor piece endgame with at least one minor piece per side and
// no pawns. Note that the case KmmK is already handled by KXK.
assert((pos.pieces(WHITE, KNIGHT) | pos.pieces(WHITE, BISHOP)));
assert((pos.pieces(BLACK, KNIGHT) | pos.pieces(BLACK, BISHOP)));
if ( pos.count<BISHOP>(WHITE) + pos.count<KNIGHT>(WHITE) <= 2
&& pos.count<BISHOP>(BLACK) + pos.count<KNIGHT>(BLACK) <= 2)
{
e->evaluationFunction = &EvaluateKmmKm[pos.side_to_move()];
return e;
}
}
// OK, we didn't find any special evaluation function for the current
// material configuration. Is there a suitable scaling function?
//
// We face problems when there are several conflicting applicable
// scaling functions and we need to decide which one to use.
EndgameBase<ScaleFactor>* sf;
if (endgames.probe(key, sf))
{
e->scalingFunction[sf->color()] = sf;
return e;
}
// Generic scaling functions that refer to more then one material
// distribution. Should be probed after the specialized ones.
// Note that these ones don't return after setting the function.
if (is_KBPsKs<WHITE>(pos))
e->scalingFunction[WHITE] = &ScaleKBPsK[WHITE];
if (is_KBPsKs<BLACK>(pos))
e->scalingFunction[BLACK] = &ScaleKBPsK[BLACK];
if (is_KQKRPs<WHITE>(pos))
e->scalingFunction[WHITE] = &ScaleKQKRPs[WHITE];
else if (is_KQKRPs<BLACK>(pos))
e->scalingFunction[BLACK] = &ScaleKQKRPs[BLACK];
Value npm_w = pos.non_pawn_material(WHITE); Value npm_w = pos.non_pawn_material(WHITE);
Value npm_b = pos.non_pawn_material(BLACK); Value npm_b = pos.non_pawn_material(BLACK);
Value npm = std::max(EndgameLimit, std::min(npm_w + npm_b, MidgameLimit));
if (npm_w + npm_b == VALUE_ZERO) // Map total non-pawn material into [PHASE_ENDGAME, PHASE_MIDGAME]
e->gamePhase = Phase(((npm - EndgameLimit) * PHASE_MIDGAME) / (MidgameLimit - EndgameLimit));
// Let's look if we have a specialized evaluation function for this particular
// material configuration. Firstly we look for a fixed configuration one, then
// for a generic one if the previous search failed.
if ((e->evaluationFunction = pos.this_thread()->endgames.probe<Value>(key)) != nullptr)
return e;
for (Color c = WHITE; c <= BLACK; ++c)
if (is_KXK(pos, c))
{
e->evaluationFunction = &EvaluateKXK[c];
return e;
}
// OK, we didn't find any special evaluation function for the current material
// configuration. Is there a suitable specialized scaling function?
const EndgameBase<ScaleFactor>* sf;
if ((sf = pos.this_thread()->endgames.probe<ScaleFactor>(key)) != nullptr)
{
e->scalingFunction[sf->strongSide] = sf; // Only strong color assigned
return e;
}
// We didn't find any specialized scaling function, so fall back on generic
// ones that refer to more than one material distribution. Note that in this
// case we don't return after setting the function.
for (Color c = WHITE; c <= BLACK; ++c)
{
if (is_KBPsK(pos, c))
e->scalingFunction[c] = &ScaleKBPsK[c];
else if (is_KQKRPs(pos, c))
e->scalingFunction[c] = &ScaleKQKRPs[c];
}
if (npm_w + npm_b == VALUE_ZERO && pos.pieces(PAWN)) // Only pawns on the board
{ {
if (!pos.count<PAWN>(BLACK)) if (!pos.count<PAWN>(BLACK))
{ {
assert(pos.count<PAWN>(WHITE) >= 2); assert(pos.count<PAWN>(WHITE) >= 2);
e->scalingFunction[WHITE] = &ScaleKPsK[WHITE]; e->scalingFunction[WHITE] = &ScaleKPsK[WHITE];
} }
else if (!pos.count<PAWN>(WHITE)) else if (!pos.count<PAWN>(WHITE))
{ {
assert(pos.count<PAWN>(BLACK) >= 2); assert(pos.count<PAWN>(BLACK) >= 2);
e->scalingFunction[BLACK] = &ScaleKPsK[BLACK]; e->scalingFunction[BLACK] = &ScaleKPsK[BLACK];
} }
else if (pos.count<PAWN>(WHITE) == 1 && pos.count<PAWN>(BLACK) == 1) else if (pos.count<PAWN>(WHITE) == 1 && pos.count<PAWN>(BLACK) == 1)
@@ -240,30 +195,19 @@ Entry* probe(const Position& pos, Table& entries, Endgames& endgames) {
} }
} }
// No pawns makes it difficult to win, even with a material advantage // Zero or just one pawn makes it difficult to win, even with a small material
// advantage. This catches some trivial draws like KK, KBK and KNK and gives a
// drawish scale factor for cases such as KRKBP and KmmKm (except for KBBKN).
if (!pos.count<PAWN>(WHITE) && npm_w - npm_b <= BishopValueMg) if (!pos.count<PAWN>(WHITE) && npm_w - npm_b <= BishopValueMg)
{ e->factor[WHITE] = uint8_t(npm_w < RookValueMg ? SCALE_FACTOR_DRAW :
e->factor[WHITE] = (uint8_t) npm_b <= BishopValueMg ? 4 : 14);
(npm_w == npm_b || npm_w < RookValueMg ? 0 : NoPawnsSF[std::min(pos.count<BISHOP>(WHITE), 2)]);
}
if (!pos.count<PAWN>(BLACK) && npm_b - npm_w <= BishopValueMg) if (!pos.count<PAWN>(BLACK) && npm_b - npm_w <= BishopValueMg)
{ e->factor[BLACK] = uint8_t(npm_b < RookValueMg ? SCALE_FACTOR_DRAW :
e->factor[BLACK] = (uint8_t) npm_w <= BishopValueMg ? 4 : 14);
(npm_w == npm_b || npm_b < RookValueMg ? 0 : NoPawnsSF[std::min(pos.count<BISHOP>(BLACK), 2)]);
}
// Compute the space weight
if (npm_w + npm_b >= 2 * QueenValueMg + 4 * RookValueMg + 2 * KnightValueMg)
{
int minorPieceCount = pos.count<KNIGHT>(WHITE) + pos.count<BISHOP>(WHITE)
+ pos.count<KNIGHT>(BLACK) + pos.count<BISHOP>(BLACK);
e->spaceWeight = make_score(minorPieceCount * minorPieceCount, 0);
}
// Evaluate the material imbalance. We use PIECE_TYPE_NONE as a place holder // Evaluate the material imbalance. We use PIECE_TYPE_NONE as a place holder
// for the bishop pair "extended piece", this allow us to be more flexible // for the bishop pair "extended piece", which allows us to be more flexible
// in defining bishop pair bonuses. // in defining bishop pair bonuses.
const int pieceCount[COLOR_NB][PIECE_TYPE_NB] = { const int pieceCount[COLOR_NB][PIECE_TYPE_NB] = {
{ pos.count<BISHOP>(WHITE) > 1, pos.count<PAWN>(WHITE), pos.count<KNIGHT>(WHITE), { pos.count<BISHOP>(WHITE) > 1, pos.count<PAWN>(WHITE), pos.count<KNIGHT>(WHITE),
@@ -271,22 +215,8 @@ Entry* probe(const Position& pos, Table& entries, Endgames& endgames) {
{ pos.count<BISHOP>(BLACK) > 1, pos.count<PAWN>(BLACK), pos.count<KNIGHT>(BLACK), { pos.count<BISHOP>(BLACK) > 1, pos.count<PAWN>(BLACK), pos.count<KNIGHT>(BLACK),
pos.count<BISHOP>(BLACK) , pos.count<ROOK>(BLACK), pos.count<QUEEN >(BLACK) } }; pos.count<BISHOP>(BLACK) , pos.count<ROOK>(BLACK), pos.count<QUEEN >(BLACK) } };
e->value = (int16_t)((imbalance<WHITE>(pieceCount) - imbalance<BLACK>(pieceCount)) / 16); e->value = int16_t((imbalance<WHITE>(pieceCount) - imbalance<BLACK>(pieceCount)) / 16);
return e; return e;
} }
/// Material::game_phase() calculates the phase given the current
/// position. Because the phase is strictly a function of the material, it
/// is stored in MaterialEntry.
Phase game_phase(const Position& pos) {
Value npm = pos.non_pawn_material(WHITE) + pos.non_pawn_material(BLACK);
return npm >= MidgameLimit ? PHASE_MIDGAME
: npm <= EndgameLimit ? PHASE_ENDGAME
: Phase(((npm - EndgameLimit) * 128) / (MidgameLimit - EndgameLimit));
}
} // namespace Material } // namespace Material
+26 -30
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -28,50 +29,45 @@
namespace Material { namespace Material {
/// Material::Entry contains various information about a material configuration. /// Material::Entry contains various information about a material configuration.
/// It contains a material balance evaluation, a function pointer to a special /// It contains a material imbalance evaluation, a function pointer to a special
/// endgame evaluation function (which in most cases is NULL, meaning that the /// endgame evaluation function (which in most cases is NULL, meaning that the
/// standard evaluation function will be used), and "scale factors". /// standard evaluation function will be used), and scale factors.
/// ///
/// The scale factors are used to scale the evaluation score up or down. /// The scale factors are used to scale the evaluation score up or down. For
/// For instance, in KRB vs KR endgames, the score is scaled down by a factor /// instance, in KRB vs KR endgames, the score is scaled down by a factor of 4,
/// of 4, which will result in scores of absolute value less than one pawn. /// which will result in scores of absolute value less than one pawn.
struct Entry { struct Entry {
Score material_value() const { return make_score(value, value); } Score imbalance() const { return make_score(value, value); }
Score space_weight() const { return spaceWeight; }
Phase game_phase() const { return gamePhase; } Phase game_phase() const { return gamePhase; }
bool specialized_eval_exists() const { return evaluationFunction != NULL; } bool specialized_eval_exists() const { return evaluationFunction != nullptr; }
Value evaluate(const Position& p) const { return (*evaluationFunction)(p); } Value evaluate(const Position& pos) const { return (*evaluationFunction)(pos); }
ScaleFactor scale_factor(const Position& pos, Color c) const;
// scale_factor takes a position and a color as input and returns a scale factor
// for the given color. We have to provide the position in addition to the color
// because the scale factor may also be a function which should be applied to
// the position. For instance, in KBP vs K endgames, the scaling function looks
// for rook pawns and wrong-colored bishops.
ScaleFactor scale_factor(const Position& pos, Color c) const {
ScaleFactor sf = scalingFunction[c] ? (*scalingFunction[c])(pos)
: SCALE_FACTOR_NONE;
return sf != SCALE_FACTOR_NONE ? sf : ScaleFactor(factor[c]);
}
Key key; Key key;
const EndgameBase<Value>* evaluationFunction;
const EndgameBase<ScaleFactor>* scalingFunction[COLOR_NB]; // Could be one for each
// side (e.g. KPKP, KBPsKs)
int16_t value; int16_t value;
uint8_t factor[COLOR_NB]; uint8_t factor[COLOR_NB];
EndgameBase<Value>* evaluationFunction;
EndgameBase<ScaleFactor>* scalingFunction[COLOR_NB];
Score spaceWeight;
Phase gamePhase; Phase gamePhase;
}; };
typedef HashTable<Entry, 8192> Table; typedef HashTable<Entry, 8192> Table;
Entry* probe(const Position& pos, Table& entries, Endgames& endgames); Entry* probe(const Position& pos);
Phase game_phase(const Position& pos);
/// Material::scale_factor takes a position and a color as input, and } // namespace Material
/// returns a scale factor for the given color. We have to provide the
/// position in addition to the color, because the scale factor need not
/// to be a constant: It can also be a function which should be applied to
/// the position. For instance, in KBP vs K endgames, a scaling function
/// which checks for draws with rook pawns and wrong-colored bishops.
inline ScaleFactor Entry::scale_factor(const Position& pos, Color c) const {
return !scalingFunction[c] || (*scalingFunction[c])(pos) == SCALE_FACTOR_NONE
? ScaleFactor(factor[c]) : (*scalingFunction[c])(pos);
}
}
#endif // #ifndef MATERIAL_H_INCLUDED #endif // #ifndef MATERIAL_H_INCLUDED
+224 -106
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,19 +18,103 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#ifdef _WIN32
#if _WIN32_WINNT < 0x0601
#undef _WIN32_WINNT
#define _WIN32_WINNT 0x0601 // Force to include needed API prototypes
#endif
#ifndef NOMINMAX
#define NOMINMAX
#endif
#include <windows.h>
// The needed Windows API for processor groups could be missed from old Windows
// versions, so instead of calling them directly (forcing the linker to resolve
// the calls at compile time), try to load them at runtime. To do this we need
// first to define the corresponding function pointers.
extern "C" {
typedef bool(*fun1_t)(LOGICAL_PROCESSOR_RELATIONSHIP,
PSYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX, PDWORD);
typedef bool(*fun2_t)(USHORT, PGROUP_AFFINITY);
typedef bool(*fun3_t)(HANDLE, CONST GROUP_AFFINITY*, PGROUP_AFFINITY);
}
#endif
#include <fstream>
#include <iomanip> #include <iomanip>
#include <iostream> #include <iostream>
#include <sstream> #include <sstream>
#include <vector>
#include "misc.h" #include "misc.h"
#include "thread.h" #include "thread.h"
using namespace std; using namespace std;
/// Version number. If Version is left empty, then compile date, in the namespace {
/// format DD-MM-YY, is shown in engine_info.
static const string Version = "4";
/// Version number. If Version is left empty, then compile date in the format
/// DD-MM-YY and show in engine_info.
const string Version = "10";
/// Our fancy logging facility. The trick here is to replace cin.rdbuf() and
/// cout.rdbuf() with two Tie objects that tie cin and cout to a file stream. We
/// can toggle the logging of std::cout and std:cin at runtime whilst preserving
/// usual I/O functionality, all without changing a single line of code!
/// Idea from http://groups.google.com/group/comp.lang.c++/msg/1d941c0f26ea0d81
struct Tie: public streambuf { // MSVC requires split streambuf for cin and cout
Tie(streambuf* b, streambuf* l) : buf(b), logBuf(l) {}
int sync() override { return logBuf->pubsync(), buf->pubsync(); }
int overflow(int c) override { return log(buf->sputc((char)c), "<< "); }
int underflow() override { return buf->sgetc(); }
int uflow() override { return log(buf->sbumpc(), ">> "); }
streambuf *buf, *logBuf;
int log(int c, const char* prefix) {
static int last = '\n'; // Single log file
if (last == '\n')
logBuf->sputn(prefix, 3);
return last = logBuf->sputc((char)c);
}
};
class Logger {
Logger() : in(cin.rdbuf(), file.rdbuf()), out(cout.rdbuf(), file.rdbuf()) {}
~Logger() { start(""); }
ofstream file;
Tie in, out;
public:
static void start(const std::string& fname) {
static Logger l;
if (!fname.empty() && !l.file.is_open())
{
l.file.open(fname, ifstream::out);
cin.rdbuf(&l.in);
cout.rdbuf(&l.out);
}
else if (fname.empty() && l.file.is_open())
{
cout.rdbuf(l.out.buf);
cin.rdbuf(l.in.buf);
l.file.close();
}
}
};
} // namespace
/// engine_info() returns the full name of the current Stockfish version. This /// engine_info() returns the full name of the current Stockfish version. This
/// will be either "Stockfish <Tag> DD-MM-YY" (where DD-MM-YY is the date when /// will be either "Stockfish <Tag> DD-MM-YY" (where DD-MM-YY is the date when
@@ -40,32 +125,31 @@ const string engine_info(bool to_uci) {
const string months("Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec"); const string months("Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec");
string month, day, year; string month, day, year;
stringstream s, date(__DATE__); // From compiler, format is "Sep 21 2008" stringstream ss, date(__DATE__); // From compiler, format is "Sep 21 2008"
s << "Stockfish " << Version << setfill('0'); ss << "Stockfish " << Version << setfill('0');
if (Version.empty()) if (Version.empty())
{ {
date >> month >> day >> year; date >> month >> day >> year;
s << setw(2) << day << setw(2) << (1 + months.find(month) / 4) << year.substr(2); ss << setw(2) << day << setw(2) << (1 + months.find(month) / 4) << year.substr(2);
} }
s << (Is64Bit ? " 64" : "") ss << (Is64Bit ? " 64" : "")
<< (HasPopCnt ? " SSE4.2" : "") << (HasPext ? " BMI2" : (HasPopCnt ? " POPCNT" : ""))
<< (to_uci ? "\nid author ": " by ") << (to_uci ? "\nid author ": " by ")
<< "Tord Romstad, Marco Costalba and Joona Kiiski"; << "T. Romstad, M. Costalba, J. Kiiski, G. Linscott";
return s.str(); return ss.str();
} }
/// Debug functions used mainly to collect run-time statistics /// Debug functions used mainly to collect run-time statistics
static int64_t hits[2], means[2];
static uint64_t hits[2], means[2]; void dbg_hit_on(bool b) { ++hits[0]; if (b) ++hits[1]; }
void dbg_hit_on(bool c, bool b) { if (c) dbg_hit_on(b); }
void dbg_hit_on(bool b) { hits[0]++; if (b) hits[1]++; } void dbg_mean_of(int v) { ++means[0]; means[1] += v; }
void dbg_hit_on_c(bool c, bool b) { if (c) dbg_hit_on(b); }
void dbg_mean_of(int v) { means[0]++; means[1] += v; }
void dbg_print() { void dbg_print() {
@@ -75,79 +159,21 @@ void dbg_print() {
if (means[0]) if (means[0])
cerr << "Total " << means[0] << " Mean " cerr << "Total " << means[0] << " Mean "
<< (float)means[1] / means[0] << endl; << (double)means[1] / means[0] << endl;
} }
/// Our fancy logging facility. The trick here is to replace cin.rdbuf() and /// Used to serialize access to std::cout to avoid multiple threads writing at
/// cout.rdbuf() with two Tie objects that tie cin and cout to a file stream. We
/// can toggle the logging of std::cout and std:cin at runtime while preserving
/// usual i/o functionality and without changing a single line of code!
/// Idea from http://groups.google.com/group/comp.lang.c++/msg/1d941c0f26ea0d81
struct Tie: public streambuf { // MSVC requires splitted streambuf for cin and cout
Tie(streambuf* b, ofstream* f) : buf(b), file(f) {}
int sync() { return file->rdbuf()->pubsync(), buf->pubsync(); }
int overflow(int c) { return log(buf->sputc((char)c), "<< "); }
int underflow() { return buf->sgetc(); }
int uflow() { return log(buf->sbumpc(), ">> "); }
streambuf* buf;
ofstream* file;
int log(int c, const char* prefix) {
static int last = '\n';
if (last == '\n')
file->rdbuf()->sputn(prefix, 3);
return last = file->rdbuf()->sputc((char)c);
}
};
class Logger {
Logger() : in(cin.rdbuf(), &file), out(cout.rdbuf(), &file) {}
~Logger() { start(false); }
ofstream file;
Tie in, out;
public:
static void start(bool b) {
static Logger l;
if (b && !l.file.is_open())
{
l.file.open("io_log.txt", ifstream::out | ifstream::app);
cin.rdbuf(&l.in);
cout.rdbuf(&l.out);
}
else if (!b && l.file.is_open())
{
cout.rdbuf(l.out.buf);
cin.rdbuf(l.in.buf);
l.file.close();
}
}
};
/// Used to serialize access to std::cout to avoid multiple threads to write at
/// the same time. /// the same time.
std::ostream& operator<<(std::ostream& os, SyncCout sc) { std::ostream& operator<<(std::ostream& os, SyncCout sc) {
static Mutex m; static Mutex m;
if (sc == io_lock) if (sc == IO_LOCK)
m.lock(); m.lock();
if (sc == io_unlock) if (sc == IO_UNLOCK)
m.unlock(); m.unlock();
return os; return os;
@@ -155,50 +181,142 @@ std::ostream& operator<<(std::ostream& os, SyncCout sc) {
/// Trampoline helper to avoid moving Logger to misc.h /// Trampoline helper to avoid moving Logger to misc.h
void start_logger(bool b) { Logger::start(b); } void start_logger(const std::string& fname) { Logger::start(fname); }
/// timed_wait() waits for msec milliseconds. It is mainly an helper to wrap /// prefetch() preloads the given address in L1/L2 cache. This is a non-blocking
/// conversion from milliseconds to struct timespec, as used by pthreads. /// function that doesn't stall the CPU waiting for data to be loaded from memory,
/// which can be quite slow.
void timed_wait(WaitCondition& sleepCond, Lock& sleepLock, int msec) {
#ifdef _WIN32
int tm = msec;
#else
timespec ts, *tm = &ts;
uint64_t ms = Time::now() + msec;
ts.tv_sec = ms / 1000;
ts.tv_nsec = (ms % 1000) * 1000000LL;
#endif
cond_timedwait(sleepCond, sleepLock, tm);
}
/// prefetch() preloads the given address in L1/L2 cache. This is a non
/// blocking function and do not stalls the CPU waiting for data to be
/// loaded from memory, that can be quite slow.
#ifdef NO_PREFETCH #ifdef NO_PREFETCH
void prefetch(char*) {} void prefetch(void*) {}
#else #else
void prefetch(char* addr) { void prefetch(void* addr) {
# if defined(__INTEL_COMPILER) # if defined(__INTEL_COMPILER)
// This hack prevents prefetches to be optimized away by // This hack prevents prefetches from being optimized away by
// Intel compiler. Both MSVC and gcc seems not affected. // Intel compiler. Both MSVC and gcc seem not be affected by this.
__asm__ (""); __asm__ ("");
# endif # endif
# if defined(__INTEL_COMPILER) || defined(_MSC_VER) # if defined(__INTEL_COMPILER) || defined(_MSC_VER)
_mm_prefetch(addr, _MM_HINT_T0); _mm_prefetch((char*)addr, _MM_HINT_T0);
# else # else
__builtin_prefetch(addr); __builtin_prefetch(addr);
# endif # endif
} }
#endif #endif
void prefetch2(void* addr) {
prefetch(addr);
prefetch((uint8_t*)addr + 64);
}
namespace WinProcGroup {
#ifndef _WIN32
void bindThisThread(size_t) {}
#else
/// best_group() retrieves logical processor information using Windows specific
/// API and returns the best group id for the thread with index idx. Original
/// code from Texel by Peter Österlund.
int best_group(size_t idx) {
int threads = 0;
int nodes = 0;
int cores = 0;
DWORD returnLength = 0;
DWORD byteOffset = 0;
// Early exit if the needed API is not available at runtime
HMODULE k32 = GetModuleHandle("Kernel32.dll");
auto fun1 = (fun1_t)(void(*)())GetProcAddress(k32, "GetLogicalProcessorInformationEx");
if (!fun1)
return -1;
// First call to get returnLength. We expect it to fail due to null buffer
if (fun1(RelationAll, nullptr, &returnLength))
return -1;
// Once we know returnLength, allocate the buffer
SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX *buffer, *ptr;
ptr = buffer = (SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX*)malloc(returnLength);
// Second call, now we expect to succeed
if (!fun1(RelationAll, buffer, &returnLength))
{
free(buffer);
return -1;
}
while (ptr->Size > 0 && byteOffset + ptr->Size <= returnLength)
{
if (ptr->Relationship == RelationNumaNode)
nodes++;
else if (ptr->Relationship == RelationProcessorCore)
{
cores++;
threads += (ptr->Processor.Flags == LTP_PC_SMT) ? 2 : 1;
}
byteOffset += ptr->Size;
ptr = (SYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX*)(((char*)ptr) + ptr->Size);
}
free(buffer);
std::vector<int> groups;
// Run as many threads as possible on the same node until core limit is
// reached, then move on filling the next node.
for (int n = 0; n < nodes; n++)
for (int i = 0; i < cores / nodes; i++)
groups.push_back(n);
// In case a core has more than one logical processor (we assume 2) and we
// have still threads to allocate, then spread them evenly across available
// nodes.
for (int t = 0; t < threads - cores; t++)
groups.push_back(t % nodes);
// If we still have more threads than the total number of logical processors
// then return -1 and let the OS to decide what to do.
return idx < groups.size() ? groups[idx] : -1;
}
/// bindThisThread() set the group affinity of the current thread
void bindThisThread(size_t idx) {
// Use only local variables to be thread-safe
int group = best_group(idx);
if (group == -1)
return;
// Early exit if the needed API are not available at runtime
HMODULE k32 = GetModuleHandle("Kernel32.dll");
auto fun2 = (fun2_t)(void(*)())GetProcAddress(k32, "GetNumaNodeProcessorMaskEx");
auto fun3 = (fun3_t)(void(*)())GetProcAddress(k32, "SetThreadGroupAffinity");
if (!fun2 || !fun3)
return;
GROUP_AFFINITY affinity;
if (fun2(group, &affinity))
fun3(GetCurrentThread(), &affinity, nullptr);
}
#endif
} // namespace WinProcGroup
+71 -25
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,49 +21,94 @@
#ifndef MISC_H_INCLUDED #ifndef MISC_H_INCLUDED
#define MISC_H_INCLUDED #define MISC_H_INCLUDED
#include <fstream> #include <cassert>
#include <chrono>
#include <ostream>
#include <string> #include <string>
#include <vector> #include <vector>
#include "types.h" #include "types.h"
extern const std::string engine_info(bool to_uci = false); const std::string engine_info(bool to_uci = false);
extern void timed_wait(WaitCondition&, Lock&, int); void prefetch(void* addr);
extern void prefetch(char* addr); void prefetch2(void* addr);
extern void start_logger(bool b); void start_logger(const std::string& fname);
extern void dbg_hit_on(bool b); void dbg_hit_on(bool b);
extern void dbg_hit_on_c(bool c, bool b); void dbg_hit_on(bool c, bool b);
extern void dbg_mean_of(int v); void dbg_mean_of(int v);
extern void dbg_print(); void dbg_print();
typedef std::chrono::milliseconds::rep TimePoint; // A value in milliseconds
struct Log : public std::ofstream { static_assert(sizeof(TimePoint) == sizeof(int64_t), "TimePoint should be 64 bits");
Log(const std::string& f = "log.txt") : std::ofstream(f.c_str(), std::ios::out | std::ios::app) {}
~Log() { if (is_open()) close(); }
};
inline TimePoint now() {
namespace Time { return std::chrono::duration_cast<std::chrono::milliseconds>
typedef int64_t point; (std::chrono::steady_clock::now().time_since_epoch()).count();
inline point now() { return system_time_to_msec(); }
} }
template<class Entry, int Size> template<class Entry, int Size>
struct HashTable { struct HashTable {
HashTable() : e(Size, Entry()) {} Entry* operator[](Key key) { return &table[(uint32_t)key & (Size - 1)]; }
Entry* operator[](Key k) { return &e[(uint32_t)k & (Size - 1)]; }
private: private:
std::vector<Entry> e; std::vector<Entry> table = std::vector<Entry>(Size);
}; };
enum SyncCout { io_lock, io_unlock }; enum SyncCout { IO_LOCK, IO_UNLOCK };
std::ostream& operator<<(std::ostream&, SyncCout); std::ostream& operator<<(std::ostream&, SyncCout);
#define sync_cout std::cout << io_lock #define sync_cout std::cout << IO_LOCK
#define sync_endl std::endl << io_unlock #define sync_endl std::endl << IO_UNLOCK
/// xorshift64star Pseudo-Random Number Generator
/// This class is based on original code written and dedicated
/// to the public domain by Sebastiano Vigna (2014).
/// It has the following characteristics:
///
/// - Outputs 64-bit numbers
/// - Passes Dieharder and SmallCrush test batteries
/// - Does not require warm-up, no zeroland to escape
/// - Internal state is a single 64-bit integer
/// - Period is 2^64 - 1
/// - Speed: 1.60 ns/call (Core i7 @3.40GHz)
///
/// For further analysis see
/// <http://vigna.di.unimi.it/ftp/papers/xorshift.pdf>
class PRNG {
uint64_t s;
uint64_t rand64() {
s ^= s >> 12, s ^= s << 25, s ^= s >> 27;
return s * 2685821657736338717LL;
}
public:
PRNG(uint64_t seed) : s(seed) { assert(seed); }
template<typename T> T rand() { return T(rand64()); }
/// Special generator used to fast init magic numbers.
/// Output values only have 1/8th of their bits set on average.
template<typename T> T sparse_rand()
{ return T(rand64() & rand64() & rand64()); }
};
/// Under Windows it is not possible for a process to run on more than one
/// logical processor group. This usually means to be limited to use max 64
/// cores. To overcome this, some special platform specific API should be
/// called to set group affinity for each thread. Original code from Texel by
/// Peter Österlund.
namespace WinProcGroup {
void bindThisThread(size_t idx);
}
#endif // #ifndef MISC_H_INCLUDED #endif // #ifndef MISC_H_INCLUDED
+166 -167
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -22,99 +23,87 @@
#include "movegen.h" #include "movegen.h"
#include "position.h" #include "position.h"
/// Simple macro to wrap a very common while loop, no facny, no flexibility,
/// hardcoded names 'mlist' and 'from'.
#define SERIALIZE(b) while (b) (mlist++)->move = make_move(from, pop_lsb(&b))
/// Version used for pawns, where the 'from' square is given as a delta from the 'to' square
#define SERIALIZE_PAWNS(b, d) while (b) { Square to = pop_lsb(&b); \
(mlist++)->move = make_move(to - (d), to); }
namespace { namespace {
template<CastlingSide Side, bool Checks, bool Chess960> template<Color Us, CastlingSide Cs, bool Checks, bool Chess960>
ExtMove* generate_castle(const Position& pos, ExtMove* mlist, Color us) { ExtMove* generate_castling(const Position& pos, ExtMove* moveList) {
if (pos.castle_impeded(us, Side) || !pos.can_castle(make_castle_right(us, Side))) constexpr CastlingRight Cr = Us | Cs;
return mlist; constexpr bool KingSide = (Cr == WHITE_OO || Cr == BLACK_OO);
if (pos.castling_impeded(Cr) || !pos.can_castle(Cr))
return moveList;
// After castling, the rook and king final positions are the same in Chess960 // After castling, the rook and king final positions are the same in Chess960
// as they would be in standard chess. // as they would be in standard chess.
Square kfrom = pos.king_square(us); Square kfrom = pos.square<KING>(Us);
Square rfrom = pos.castle_rook_square(us, Side); Square rfrom = pos.castling_rook_square(Cr);
Square kto = relative_square(us, Side == KING_SIDE ? SQ_G1 : SQ_C1); Square kto = relative_square(Us, KingSide ? SQ_G1 : SQ_C1);
Bitboard enemies = pos.pieces(~us); Bitboard enemies = pos.pieces(~Us);
assert(!pos.checkers()); assert(!pos.checkers());
const int K = Chess960 ? kto > kfrom ? -1 : 1 const Direction step = Chess960 ? kto > kfrom ? WEST : EAST
: Side == KING_SIDE ? -1 : 1; : KingSide ? WEST : EAST;
for (Square s = kto; s != kfrom; s += (Square)K) for (Square s = kto; s != kfrom; s += step)
if (pos.attackers_to(s) & enemies) if (pos.attackers_to(s) & enemies)
return mlist; return moveList;
// Because we generate only legal castling moves we need to verify that // Because we generate only legal castling moves we need to verify that
// when moving the castling rook we do not discover some hidden checker. // when moving the castling rook we do not discover some hidden checker.
// For instance an enemy queen in SQ_A1 when castling rook is in SQ_B1. // For instance an enemy queen in SQ_A1 when castling rook is in SQ_B1.
if (Chess960 && (pos.attackers_to(kto, pos.pieces() ^ rfrom) & enemies)) if (Chess960 && (attacks_bb<ROOK>(kto, pos.pieces() ^ rfrom) & pos.pieces(~Us, ROOK, QUEEN)))
return mlist; return moveList;
(mlist++)->move = make<CASTLE>(kfrom, rfrom); Move m = make<CASTLING>(kfrom, rfrom);
if (Checks && !pos.move_gives_check((mlist - 1)->move, CheckInfo(pos))) if (Checks && !pos.gives_check(m))
mlist--; return moveList;
return mlist; *moveList++ = m;
return moveList;
} }
template<GenType Type, Square Delta> template<GenType Type, Direction D>
inline ExtMove* generate_promotions(ExtMove* mlist, Bitboard pawnsOn7, ExtMove* make_promotions(ExtMove* moveList, Square to, Square ksq) {
Bitboard target, const CheckInfo* ci) {
Bitboard b = shift_bb<Delta>(pawnsOn7) & target;
while (b)
{
Square to = pop_lsb(&b);
if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS) if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
(mlist++)->move = make<PROMOTION>(to - Delta, to, QUEEN); *moveList++ = make<PROMOTION>(to - D, to, QUEEN);
if (Type == QUIETS || Type == EVASIONS || Type == NON_EVASIONS) if (Type == QUIETS || Type == EVASIONS || Type == NON_EVASIONS)
{ {
(mlist++)->move = make<PROMOTION>(to - Delta, to, ROOK); *moveList++ = make<PROMOTION>(to - D, to, ROOK);
(mlist++)->move = make<PROMOTION>(to - Delta, to, BISHOP); *moveList++ = make<PROMOTION>(to - D, to, BISHOP);
(mlist++)->move = make<PROMOTION>(to - Delta, to, KNIGHT); *moveList++ = make<PROMOTION>(to - D, to, KNIGHT);
} }
// Knight-promotion is the only one that can give a direct check not // Knight promotion is the only promotion that can give a direct check
// already included in the queen-promotion. // that's not already included in the queen promotion.
if (Type == QUIET_CHECKS && (StepAttacksBB[W_KNIGHT][to] & ci->ksq)) if (Type == QUIET_CHECKS && (PseudoAttacks[KNIGHT][to] & ksq))
(mlist++)->move = make<PROMOTION>(to - Delta, to, KNIGHT); *moveList++ = make<PROMOTION>(to - D, to, KNIGHT);
else else
(void)ci; // Silence a warning under MSVC (void)ksq; // Silence a warning under MSVC
}
return mlist; return moveList;
} }
template<Color Us, GenType Type> template<Color Us, GenType Type>
ExtMove* generate_pawn_moves(const Position& pos, ExtMove* mlist, ExtMove* generate_pawn_moves(const Position& pos, ExtMove* moveList, Bitboard target) {
Bitboard target, const CheckInfo* ci) {
// Compute our parametrized parameters at compile time, named according to // Compute our parametrized parameters at compile time, named according to
// the point of view of white side. // the point of view of white side.
const Color Them = (Us == WHITE ? BLACK : WHITE); constexpr Color Them = (Us == WHITE ? BLACK : WHITE);
const Bitboard TRank8BB = (Us == WHITE ? Rank8BB : Rank1BB); constexpr Bitboard TRank8BB = (Us == WHITE ? Rank8BB : Rank1BB);
const Bitboard TRank7BB = (Us == WHITE ? Rank7BB : Rank2BB); constexpr Bitboard TRank7BB = (Us == WHITE ? Rank7BB : Rank2BB);
const Bitboard TRank3BB = (Us == WHITE ? Rank3BB : Rank6BB); constexpr Bitboard TRank3BB = (Us == WHITE ? Rank3BB : Rank6BB);
const Square Up = (Us == WHITE ? DELTA_N : DELTA_S); constexpr Direction Up = (Us == WHITE ? NORTH : SOUTH);
const Square Right = (Us == WHITE ? DELTA_NE : DELTA_SW); constexpr Direction UpRight = (Us == WHITE ? NORTH_EAST : SOUTH_WEST);
const Square Left = (Us == WHITE ? DELTA_NW : DELTA_SE); constexpr Direction UpLeft = (Us == WHITE ? NORTH_WEST : SOUTH_EAST);
Bitboard b1, b2, dc1, dc2, emptySquares; Bitboard emptySquares;
Bitboard pawnsOn7 = pos.pieces(Us, PAWN) & TRank7BB; Bitboard pawnsOn7 = pos.pieces(Us, PAWN) & TRank7BB;
Bitboard pawnsNotOn7 = pos.pieces(Us, PAWN) & ~TRank7BB; Bitboard pawnsNotOn7 = pos.pieces(Us, PAWN) & ~TRank7BB;
@@ -127,8 +116,8 @@ namespace {
{ {
emptySquares = (Type == QUIETS || Type == QUIET_CHECKS ? target : ~pos.pieces()); emptySquares = (Type == QUIETS || Type == QUIET_CHECKS ? target : ~pos.pieces());
b1 = shift_bb<Up>(pawnsNotOn7) & emptySquares; Bitboard b1 = shift<Up>(pawnsNotOn7) & emptySquares;
b2 = shift_bb<Up>(b1 & TRank3BB) & emptySquares; Bitboard b2 = shift<Up>(b1 & TRank3BB) & emptySquares;
if (Type == EVASIONS) // Consider only blocking squares if (Type == EVASIONS) // Consider only blocking squares
{ {
@@ -138,25 +127,37 @@ namespace {
if (Type == QUIET_CHECKS) if (Type == QUIET_CHECKS)
{ {
b1 &= pos.attacks_from<PAWN>(ci->ksq, Them); Square ksq = pos.square<KING>(Them);
b2 &= pos.attacks_from<PAWN>(ci->ksq, Them);
b1 &= pos.attacks_from<PAWN>(ksq, Them);
b2 &= pos.attacks_from<PAWN>(ksq, Them);
// Add pawn pushes which give discovered check. This is possible only // Add pawn pushes which give discovered check. This is possible only
// if the pawn is not on the same file as the enemy king, because we // if the pawn is not on the same file as the enemy king, because we
// don't generate captures. Note that a possible discovery check // don't generate captures. Note that a possible discovery check
// promotion has been already generated among captures. // promotion has been already generated amongst the captures.
if (pawnsNotOn7 & ci->dcCandidates) Bitboard dcCandidates = pos.blockers_for_king(Them);
if (pawnsNotOn7 & dcCandidates)
{ {
dc1 = shift_bb<Up>(pawnsNotOn7 & ci->dcCandidates) & emptySquares & ~file_bb(ci->ksq); Bitboard dc1 = shift<Up>(pawnsNotOn7 & dcCandidates) & emptySquares & ~file_bb(ksq);
dc2 = shift_bb<Up>(dc1 & TRank3BB) & emptySquares; Bitboard dc2 = shift<Up>(dc1 & TRank3BB) & emptySquares;
b1 |= dc1; b1 |= dc1;
b2 |= dc2; b2 |= dc2;
} }
} }
SERIALIZE_PAWNS(b1, Up); while (b1)
SERIALIZE_PAWNS(b2, Up + Up); {
Square to = pop_lsb(&b1);
*moveList++ = make_move(to - Up, to);
}
while (b2)
{
Square to = pop_lsb(&b2);
*moveList++ = make_move(to - Up - Up, to);
}
} }
// Promotions and underpromotions // Promotions and underpromotions
@@ -168,19 +169,39 @@ namespace {
if (Type == EVASIONS) if (Type == EVASIONS)
emptySquares &= target; emptySquares &= target;
mlist = generate_promotions<Type, Right>(mlist, pawnsOn7, enemies, ci); Bitboard b1 = shift<UpRight>(pawnsOn7) & enemies;
mlist = generate_promotions<Type, Left >(mlist, pawnsOn7, enemies, ci); Bitboard b2 = shift<UpLeft >(pawnsOn7) & enemies;
mlist = generate_promotions<Type, Up>(mlist, pawnsOn7, emptySquares, ci); Bitboard b3 = shift<Up >(pawnsOn7) & emptySquares;
Square ksq = pos.square<KING>(Them);
while (b1)
moveList = make_promotions<Type, UpRight>(moveList, pop_lsb(&b1), ksq);
while (b2)
moveList = make_promotions<Type, UpLeft >(moveList, pop_lsb(&b2), ksq);
while (b3)
moveList = make_promotions<Type, Up >(moveList, pop_lsb(&b3), ksq);
} }
// Standard and en-passant captures // Standard and en-passant captures
if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS) if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
{ {
b1 = shift_bb<Right>(pawnsNotOn7) & enemies; Bitboard b1 = shift<UpRight>(pawnsNotOn7) & enemies;
b2 = shift_bb<Left >(pawnsNotOn7) & enemies; Bitboard b2 = shift<UpLeft >(pawnsNotOn7) & enemies;
SERIALIZE_PAWNS(b1, Right); while (b1)
SERIALIZE_PAWNS(b2, Left); {
Square to = pop_lsb(&b1);
*moveList++ = make_move(to - UpRight, to);
}
while (b2)
{
Square to = pop_lsb(&b2);
*moveList++ = make_move(to - UpLeft, to);
}
if (pos.ep_square() != SQ_NONE) if (pos.ep_square() != SQ_NONE)
{ {
@@ -190,90 +211,90 @@ namespace {
// is the double pushed pawn and so is in the target. Otherwise this // is the double pushed pawn and so is in the target. Otherwise this
// is a discovery check and we are forced to do otherwise. // is a discovery check and we are forced to do otherwise.
if (Type == EVASIONS && !(target & (pos.ep_square() - Up))) if (Type == EVASIONS && !(target & (pos.ep_square() - Up)))
return mlist; return moveList;
b1 = pawnsNotOn7 & pos.attacks_from<PAWN>(pos.ep_square(), Them); b1 = pawnsNotOn7 & pos.attacks_from<PAWN>(pos.ep_square(), Them);
assert(b1); assert(b1);
while (b1) while (b1)
(mlist++)->move = make<ENPASSANT>(pop_lsb(&b1), pos.ep_square()); *moveList++ = make<ENPASSANT>(pop_lsb(&b1), pos.ep_square());
} }
} }
return mlist; return moveList;
} }
template<PieceType Pt, bool Checks> FORCE_INLINE template<PieceType Pt, bool Checks>
ExtMove* generate_moves(const Position& pos, ExtMove* mlist, Color us, ExtMove* generate_moves(const Position& pos, ExtMove* moveList, Color us,
Bitboard target, const CheckInfo* ci) { Bitboard target) {
assert(Pt != KING && Pt != PAWN); assert(Pt != KING && Pt != PAWN);
const Square* pl = pos.list<Pt>(us); const Square* pl = pos.squares<Pt>(us);
for (Square from = *pl; from != SQ_NONE; from = *++pl) for (Square from = *pl; from != SQ_NONE; from = *++pl)
{ {
if (Checks) if (Checks)
{ {
if ( (Pt == BISHOP || Pt == ROOK || Pt == QUEEN) if ( (Pt == BISHOP || Pt == ROOK || Pt == QUEEN)
&& !(PseudoAttacks[Pt][from] & target & ci->checkSq[Pt])) && !(PseudoAttacks[Pt][from] & target & pos.check_squares(Pt)))
continue; continue;
if (unlikely(ci->dcCandidates) && (ci->dcCandidates & from)) if (pos.blockers_for_king(~us) & from)
continue; continue;
} }
Bitboard b = pos.attacks_from<Pt>(from) & target; Bitboard b = pos.attacks_from<Pt>(from) & target;
if (Checks) if (Checks)
b &= ci->checkSq[Pt]; b &= pos.check_squares(Pt);
SERIALIZE(b); while (b)
*moveList++ = make_move(from, pop_lsb(&b));
} }
return mlist; return moveList;
} }
template<Color Us, GenType Type> FORCE_INLINE template<Color Us, GenType Type>
ExtMove* generate_all(const Position& pos, ExtMove* mlist, Bitboard target, ExtMove* generate_all(const Position& pos, ExtMove* moveList, Bitboard target) {
const CheckInfo* ci = NULL) {
const bool Checks = Type == QUIET_CHECKS; constexpr bool Checks = Type == QUIET_CHECKS;
mlist = generate_pawn_moves<Us, Type>(pos, mlist, target, ci); moveList = generate_pawn_moves<Us, Type>(pos, moveList, target);
mlist = generate_moves<KNIGHT, Checks>(pos, mlist, Us, target, ci); moveList = generate_moves<KNIGHT, Checks>(pos, moveList, Us, target);
mlist = generate_moves<BISHOP, Checks>(pos, mlist, Us, target, ci); moveList = generate_moves<BISHOP, Checks>(pos, moveList, Us, target);
mlist = generate_moves< ROOK, Checks>(pos, mlist, Us, target, ci); moveList = generate_moves< ROOK, Checks>(pos, moveList, Us, target);
mlist = generate_moves< QUEEN, Checks>(pos, mlist, Us, target, ci); moveList = generate_moves< QUEEN, Checks>(pos, moveList, Us, target);
if (Type != QUIET_CHECKS && Type != EVASIONS) if (Type != QUIET_CHECKS && Type != EVASIONS)
{ {
Square from = pos.king_square(Us); Square ksq = pos.square<KING>(Us);
Bitboard b = pos.attacks_from<KING>(from) & target; Bitboard b = pos.attacks_from<KING>(ksq) & target;
SERIALIZE(b); while (b)
*moveList++ = make_move(ksq, pop_lsb(&b));
} }
if (Type != CAPTURES && Type != EVASIONS && pos.can_castle(Us)) if (Type != CAPTURES && Type != EVASIONS && pos.can_castle(Us))
{ {
if (pos.is_chess960()) if (pos.is_chess960())
{ {
mlist = generate_castle< KING_SIDE, Checks, true>(pos, mlist, Us); moveList = generate_castling<Us, KING_SIDE, Checks, true>(pos, moveList);
mlist = generate_castle<QUEEN_SIDE, Checks, true>(pos, mlist, Us); moveList = generate_castling<Us, QUEEN_SIDE, Checks, true>(pos, moveList);
} }
else else
{ {
mlist = generate_castle< KING_SIDE, Checks, false>(pos, mlist, Us); moveList = generate_castling<Us, KING_SIDE, Checks, false>(pos, moveList);
mlist = generate_castle<QUEEN_SIDE, Checks, false>(pos, mlist, Us); moveList = generate_castling<Us, QUEEN_SIDE, Checks, false>(pos, moveList);
} }
} }
return mlist; return moveList;
} }
} // namespace } // namespace
@@ -287,7 +308,7 @@ namespace {
/// non-captures. Returns a pointer to the end of the move list. /// non-captures. Returns a pointer to the end of the move list.
template<GenType Type> template<GenType Type>
ExtMove* generate(const Position& pos, ExtMove* mlist) { ExtMove* generate(const Position& pos, ExtMove* moveList) {
assert(Type == CAPTURES || Type == QUIETS || Type == NON_EVASIONS); assert(Type == CAPTURES || Type == QUIETS || Type == NON_EVASIONS);
assert(!pos.checkers()); assert(!pos.checkers());
@@ -298,8 +319,8 @@ ExtMove* generate(const Position& pos, ExtMove* mlist) {
: Type == QUIETS ? ~pos.pieces() : Type == QUIETS ? ~pos.pieces()
: Type == NON_EVASIONS ? ~pos.pieces(us) : 0; : Type == NON_EVASIONS ? ~pos.pieces(us) : 0;
return us == WHITE ? generate_all<WHITE, Type>(pos, mlist, target) return us == WHITE ? generate_all<WHITE, Type>(pos, moveList, target)
: generate_all<BLACK, Type>(pos, mlist, target); : generate_all<BLACK, Type>(pos, moveList, target);
} }
// Explicit template instantiations // Explicit template instantiations
@@ -311,13 +332,12 @@ template ExtMove* generate<NON_EVASIONS>(const Position&, ExtMove*);
/// generate<QUIET_CHECKS> generates all pseudo-legal non-captures and knight /// generate<QUIET_CHECKS> generates all pseudo-legal non-captures and knight
/// underpromotions that give check. Returns a pointer to the end of the move list. /// underpromotions that give check. Returns a pointer to the end of the move list.
template<> template<>
ExtMove* generate<QUIET_CHECKS>(const Position& pos, ExtMove* mlist) { ExtMove* generate<QUIET_CHECKS>(const Position& pos, ExtMove* moveList) {
assert(!pos.checkers()); assert(!pos.checkers());
Color us = pos.side_to_move(); Color us = pos.side_to_move();
CheckInfo ci(pos); Bitboard dc = pos.blockers_for_king(~us) & pos.pieces(us);
Bitboard dc = ci.dcCandidates;
while (dc) while (dc)
{ {
@@ -325,99 +345,78 @@ ExtMove* generate<QUIET_CHECKS>(const Position& pos, ExtMove* mlist) {
PieceType pt = type_of(pos.piece_on(from)); PieceType pt = type_of(pos.piece_on(from));
if (pt == PAWN) if (pt == PAWN)
continue; // Will be generated togheter with direct checks continue; // Will be generated together with direct checks
Bitboard b = pos.attacks_from(Piece(pt), from) & ~pos.pieces(); Bitboard b = pos.attacks_from(pt, from) & ~pos.pieces();
if (pt == KING) if (pt == KING)
b &= ~PseudoAttacks[QUEEN][ci.ksq]; b &= ~PseudoAttacks[QUEEN][pos.square<KING>(~us)];
SERIALIZE(b); while (b)
*moveList++ = make_move(from, pop_lsb(&b));
} }
return us == WHITE ? generate_all<WHITE, QUIET_CHECKS>(pos, mlist, ~pos.pieces(), &ci) return us == WHITE ? generate_all<WHITE, QUIET_CHECKS>(pos, moveList, ~pos.pieces())
: generate_all<BLACK, QUIET_CHECKS>(pos, mlist, ~pos.pieces(), &ci); : generate_all<BLACK, QUIET_CHECKS>(pos, moveList, ~pos.pieces());
} }
/// generate<EVASIONS> generates all pseudo-legal check evasions when the side /// generate<EVASIONS> generates all pseudo-legal check evasions when the side
/// to move is in check. Returns a pointer to the end of the move list. /// to move is in check. Returns a pointer to the end of the move list.
template<> template<>
ExtMove* generate<EVASIONS>(const Position& pos, ExtMove* mlist) { ExtMove* generate<EVASIONS>(const Position& pos, ExtMove* moveList) {
assert(pos.checkers()); assert(pos.checkers());
int checkersCnt = 0;
Color us = pos.side_to_move(); Color us = pos.side_to_move();
Square ksq = pos.king_square(us), from = ksq /* For SERIALIZE */, checksq; Square ksq = pos.square<KING>(us);
Bitboard sliderAttacks = 0; Bitboard sliderAttacks = 0;
Bitboard b = pos.checkers(); Bitboard sliders = pos.checkers() & ~pos.pieces(KNIGHT, PAWN);
assert(pos.checkers()); // Find all the squares attacked by slider checkers. We will remove them from
// the king evasions in order to skip known illegal moves, which avoids any
// Find squares attacked by slider checkers, we will remove them from the king // useless legality checks later on.
// evasions so to skip known illegal moves avoiding useless legality check later. while (sliders)
do
{ {
checkersCnt++; Square checksq = pop_lsb(&sliders);
checksq = pop_lsb(&b); sliderAttacks |= LineBB[checksq][ksq] ^ checksq;
assert(color_of(pos.piece_on(checksq)) == ~us);
switch (type_of(pos.piece_on(checksq)))
{
case BISHOP: sliderAttacks |= PseudoAttacks[BISHOP][checksq]; break;
case ROOK: sliderAttacks |= PseudoAttacks[ROOK][checksq]; break;
case QUEEN:
// If queen and king are far or not on a diagonal line we can safely
// remove all the squares attacked in the other direction becuase are
// not reachable by the king anyway.
if (between_bb(ksq, checksq) || !(PseudoAttacks[BISHOP][checksq] & ksq))
sliderAttacks |= PseudoAttacks[QUEEN][checksq];
// Otherwise we need to use real rook attacks to check if king is safe
// to move in the other direction. For example: king in B2, queen in A1
// a knight in B1, and we can safely move to C1.
else
sliderAttacks |= PseudoAttacks[BISHOP][checksq] | pos.attacks_from<ROOK>(checksq);
default:
break;
} }
} while (b);
// Generate evasions for king, capture and non capture moves // Generate evasions for king, capture and non capture moves
b = pos.attacks_from<KING>(ksq) & ~pos.pieces(us) & ~sliderAttacks; Bitboard b = pos.attacks_from<KING>(ksq) & ~pos.pieces(us) & ~sliderAttacks;
SERIALIZE(b); while (b)
*moveList++ = make_move(ksq, pop_lsb(&b));
if (checkersCnt > 1) if (more_than_one(pos.checkers()))
return mlist; // Double check, only a king move can save the day return moveList; // Double check, only a king move can save the day
// Generate blocking evasions or captures of the checking piece // Generate blocking evasions or captures of the checking piece
Square checksq = lsb(pos.checkers());
Bitboard target = between_bb(checksq, ksq) | checksq; Bitboard target = between_bb(checksq, ksq) | checksq;
return us == WHITE ? generate_all<WHITE, EVASIONS>(pos, mlist, target) return us == WHITE ? generate_all<WHITE, EVASIONS>(pos, moveList, target)
: generate_all<BLACK, EVASIONS>(pos, mlist, target); : generate_all<BLACK, EVASIONS>(pos, moveList, target);
} }
/// generate<LEGAL> generates all the legal moves in the given position /// generate<LEGAL> generates all the legal moves in the given position
template<> template<>
ExtMove* generate<LEGAL>(const Position& pos, ExtMove* mlist) { ExtMove* generate<LEGAL>(const Position& pos, ExtMove* moveList) {
ExtMove *end, *cur = mlist; Color us = pos.side_to_move();
Bitboard pinned = pos.pinned_pieces(); Bitboard pinned = pos.blockers_for_king(us) & pos.pieces(us);
Square ksq = pos.king_square(pos.side_to_move()); Square ksq = pos.square<KING>(us);
ExtMove* cur = moveList;
end = pos.checkers() ? generate<EVASIONS>(pos, mlist) moveList = pos.checkers() ? generate<EVASIONS >(pos, moveList)
: generate<NON_EVASIONS>(pos, mlist); : generate<NON_EVASIONS>(pos, moveList);
while (cur != end) while (cur != moveList)
if ( (pinned || from_sq(cur->move) == ksq || type_of(cur->move) == ENPASSANT) if ( (pinned || from_sq(*cur) == ksq || type_of(*cur) == ENPASSANT)
&& !pos.pl_move_is_legal(cur->move, pinned)) && !pos.legal(*cur))
cur->move = (--end)->move; *cur = (--moveList)->move;
else else
cur++; ++cur;
return end; return moveList;
} }
+31 -14
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,8 +21,12 @@
#ifndef MOVEGEN_H_INCLUDED #ifndef MOVEGEN_H_INCLUDED
#define MOVEGEN_H_INCLUDED #define MOVEGEN_H_INCLUDED
#include <algorithm>
#include "types.h" #include "types.h"
class Position;
enum GenType { enum GenType {
CAPTURES, CAPTURES,
QUIETS, QUIETS,
@@ -31,28 +36,40 @@ enum GenType {
LEGAL LEGAL
}; };
class Position; struct ExtMove {
Move move;
int value;
operator Move() const { return move; }
void operator=(Move m) { move = m; }
// Inhibit unwanted implicit conversions to Move
// with an ambiguity that yields to a compile error.
operator float() const = delete;
};
inline bool operator<(const ExtMove& f, const ExtMove& s) {
return f.value < s.value;
}
template<GenType> template<GenType>
ExtMove* generate(const Position& pos, ExtMove* mlist); ExtMove* generate(const Position& pos, ExtMove* moveList);
/// The MoveList struct is a simple wrapper around generate(), sometimes comes /// The MoveList struct is a simple wrapper around generate(). It sometimes comes
/// handy to use this class instead of the low level generate() function. /// in handy to use this class instead of the low level generate() function.
template<GenType T> template<GenType T>
struct MoveList { struct MoveList {
explicit MoveList(const Position& pos) : cur(mlist), last(generate<T>(pos, mlist)) { last->move = MOVE_NONE; } explicit MoveList(const Position& pos) : last(generate<T>(pos, moveList)) {}
void operator++() { cur++; } const ExtMove* begin() const { return moveList; }
Move operator*() const { return cur->move; } const ExtMove* end() const { return last; }
size_t size() const { return last - mlist; } size_t size() const { return last - moveList; }
bool contains(Move m) const { bool contains(Move move) const {
for (const ExtMove* it(mlist); it != last; ++it) if (it->move == m) return true; return std::find(begin(), end(), move) != end();
return false;
} }
private: private:
ExtMove mlist[MAX_MOVES]; ExtMove moveList[MAX_MOVES], *last;
ExtMove *cur, *last;
}; };
#endif // #ifndef MOVEGEN_H_INCLUDED #endif // #ifndef MOVEGEN_H_INCLUDED
+188 -296
View File
@@ -1,14 +1,14 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or the Free Software Foundation, either version 3 of the License, or
(at your option) any later version. (at your option) any later version.
Stockfish is distributed in the hope that it will be useful, Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
@@ -21,357 +21,249 @@
#include <cassert> #include <cassert>
#include "movepick.h" #include "movepick.h"
#include "thread.h"
namespace { namespace {
enum Stages { enum Stages {
MAIN_SEARCH, CAPTURES_S1, KILLERS_S1, QUIETS_1_S1, QUIETS_2_S1, BAD_CAPTURES_S1, MAIN_TT, CAPTURE_INIT, GOOD_CAPTURE, REFUTATION, QUIET_INIT, QUIET, BAD_CAPTURE,
EVASION, EVASIONS_S2, EVASION_TT, EVASION_INIT, EVASION,
QSEARCH_0, CAPTURES_S3, QUIET_CHECKS_S3, PROBCUT_TT, PROBCUT_INIT, PROBCUT,
QSEARCH_1, CAPTURES_S4, QSEARCH_TT, QCAPTURE_INIT, QCAPTURE, QCHECK_INIT, QCHECK
PROBCUT, CAPTURES_S5,
RECAPTURE, CAPTURES_S6,
STOP
}; };
// Our insertion sort, guaranteed to be stable, as is needed // Helper filter used with select()
void insertion_sort(ExtMove* begin, ExtMove* end) const auto Any = [](){ return true; };
{
ExtMove tmp, *p, *q;
for (p = begin + 1; p < end; ++p) // partial_insertion_sort() sorts moves in descending order up to and including
// a given limit. The order of moves smaller than the limit is left unspecified.
void partial_insertion_sort(ExtMove* begin, ExtMove* end, int limit) {
for (ExtMove *sortedEnd = begin, *p = begin + 1; p < end; ++p)
if (p->value >= limit)
{ {
tmp = *p; ExtMove tmp = *p, *q;
for (q = p; q != begin && *(q-1) < tmp; --q) *p = *++sortedEnd;
*q = *(q-1); for (q = sortedEnd; q != begin && *(q - 1) < tmp; --q)
*q = *(q - 1);
*q = tmp; *q = tmp;
} }
} }
// Unary predicate used by std::partition to split positive scores from remaining } // namespace
// ones so to sort separately the two sets, and with the second sort delayed.
inline bool has_positive_score(const ExtMove& ms) { return ms.score > 0; }
// Picks and moves to the front the best move in the range [begin, end),
// it is faster than sorting all the moves in advance when moves are few, as
// normally are the possible captures.
inline ExtMove* pick_best(ExtMove* begin, ExtMove* end)
{
std::swap(*begin, *std::max_element(begin, end));
return begin;
}
}
/// Constructors of the MovePicker class. As arguments we pass information /// Constructors of the MovePicker class. As arguments we pass information
/// to help it to return the presumably good moves first, to decide which /// to help it to return the (presumably) good moves first, to decide which
/// moves to return (in the quiescence search, for instance, we only want to /// moves to return (in the quiescence search, for instance, we only want to
/// search captures, promotions and some checks) and about how important good /// search captures, promotions, and some checks) and how important good move
/// move ordering is at the current node. /// ordering is at the current node.
MovePicker::MovePicker(const Position& p, Move ttm, Depth d, const HistoryStats& h, /// MovePicker constructor for the main search
Move* cm, Search::Stack* s) : pos(p), history(h), depth(d) { MovePicker::MovePicker(const Position& p, Move ttm, Depth d, const ButterflyHistory* mh,
const CapturePieceToHistory* cph, const PieceToHistory** ch, Move cm, Move* killers)
: pos(p), mainHistory(mh), captureHistory(cph), continuationHistory(ch),
refutations{{killers[0], 0}, {killers[1], 0}, {cm, 0}}, depth(d) {
assert(d > DEPTH_ZERO); assert(d > DEPTH_ZERO);
cur = end = moves; stage = pos.checkers() ? EVASION_TT : MAIN_TT;
endBadCaptures = moves + MAX_MOVES - 1; ttMove = ttm && pos.pseudo_legal(ttm) ? ttm : MOVE_NONE;
countermoves = cm; stage += (ttMove == MOVE_NONE);
ss = s;
if (p.checkers())
stage = EVASION;
else
stage = MAIN_SEARCH;
ttMove = (ttm && pos.is_pseudo_legal(ttm) ? ttm : MOVE_NONE);
end += (ttMove != MOVE_NONE);
} }
MovePicker::MovePicker(const Position& p, Move ttm, Depth d, const HistoryStats& h, /// MovePicker constructor for quiescence search
Square sq) : pos(p), history(h), cur(moves), end(moves) { MovePicker::MovePicker(const Position& p, Move ttm, Depth d, const ButterflyHistory* mh,
const CapturePieceToHistory* cph, const PieceToHistory** ch, Square rs)
: pos(p), mainHistory(mh), captureHistory(cph), continuationHistory(ch), recaptureSquare(rs), depth(d) {
assert(d <= DEPTH_ZERO); assert(d <= DEPTH_ZERO);
if (p.checkers()) stage = pos.checkers() ? EVASION_TT : QSEARCH_TT;
stage = EVASION; ttMove = ttm
&& pos.pseudo_legal(ttm)
else if (d > DEPTH_QS_NO_CHECKS) && (depth > DEPTH_QS_RECAPTURES || to_sq(ttm) == recaptureSquare) ? ttm : MOVE_NONE;
stage = QSEARCH_0; stage += (ttMove == MOVE_NONE);
else if (d > DEPTH_QS_RECAPTURES)
{
stage = QSEARCH_1;
// Skip TT move if is not a capture or a promotion, this avoids qsearch
// tree explosion due to a possible perpetual check or similar rare cases
// when TT table is full.
if (ttm && !pos.is_capture_or_promotion(ttm))
ttm = MOVE_NONE;
}
else
{
stage = RECAPTURE;
recaptureSquare = sq;
ttm = MOVE_NONE;
}
ttMove = (ttm && pos.is_pseudo_legal(ttm) ? ttm : MOVE_NONE);
end += (ttMove != MOVE_NONE);
} }
MovePicker::MovePicker(const Position& p, Move ttm, const HistoryStats& h, PieceType pt) /// MovePicker constructor for ProbCut: we generate captures with SEE greater
: pos(p), history(h), cur(moves), end(moves) { /// than or equal to the given threshold.
MovePicker::MovePicker(const Position& p, Move ttm, Value th, const CapturePieceToHistory* cph)
: pos(p), captureHistory(cph), threshold(th) {
assert(!pos.checkers()); assert(!pos.checkers());
stage = PROBCUT; stage = PROBCUT_TT;
ttMove = ttm
// In ProbCut we generate only captures better than parent's captured piece && pos.pseudo_legal(ttm)
captureThreshold = PieceValue[MG][pt]; && pos.capture(ttm)
ttMove = (ttm && pos.is_pseudo_legal(ttm) ? ttm : MOVE_NONE); && pos.see_ge(ttm, threshold) ? ttm : MOVE_NONE;
stage += (ttMove == MOVE_NONE);
if (ttMove && (!pos.is_capture(ttMove) || pos.see(ttMove) <= captureThreshold))
ttMove = MOVE_NONE;
end += (ttMove != MOVE_NONE);
} }
/// MovePicker::score() assigns a numerical value to each move in a list, used
/// for sorting. Captures are ordered by Most Valuable Victim (MVV), preferring
/// captures with a good history. Quiets moves are ordered using the histories.
template<GenType Type>
void MovePicker::score() {
/// score() assign a numerical move ordering score to each move in a move list. static_assert(Type == CAPTURES || Type == QUIETS || Type == EVASIONS, "Wrong type");
/// The moves with highest scores will be picked first.
template<>
void MovePicker::score<CAPTURES>() {
// Winning and equal captures in the main search are ordered by MVV/LVA.
// Suprisingly, this appears to perform slightly better than SEE based
// move ordering. The reason is probably that in a position with a winning
// capture, capturing a more valuable (but sufficiently defended) piece
// first usually doesn't hurt. The opponent will have to recapture, and
// the hanging piece will still be hanging (except in the unusual cases
// where it is possible to recapture with the hanging piece). Exchanging
// big pieces before capturing a hanging piece probably helps to reduce
// the subtree size.
// In main search we want to push captures with negative SEE values to
// badCaptures[] array, but instead of doing it now we delay till when
// the move has been picked up in pick_move_from_list(), this way we save
// some SEE calls in case we get a cutoff (idea from Pablo Vazquez).
Move m;
for (ExtMove* it = moves; it != end; ++it) for (auto& m : *this)
if (Type == CAPTURES)
m.value = PieceValue[MG][pos.piece_on(to_sq(m))]
+ (*captureHistory)[pos.moved_piece(m)][to_sq(m)][type_of(pos.piece_on(to_sq(m)))] / 8;
else if (Type == QUIETS)
m.value = (*mainHistory)[pos.side_to_move()][from_to(m)]
+ (*continuationHistory[0])[pos.moved_piece(m)][to_sq(m)]
+ (*continuationHistory[1])[pos.moved_piece(m)][to_sq(m)]
+ (*continuationHistory[3])[pos.moved_piece(m)][to_sq(m)];
else // Type == EVASIONS
{ {
m = it->move; if (pos.capture(m))
it->score = PieceValue[MG][pos.piece_on(to_sq(m))] m.value = PieceValue[MG][pos.piece_on(to_sq(m))]
- type_of(pos.piece_moved(m)); - Value(type_of(pos.moved_piece(m)));
if (type_of(m) == PROMOTION)
it->score += PieceValue[MG][promotion_type(m)] - PieceValue[MG][PAWN];
else if (type_of(m) == ENPASSANT)
it->score += PieceValue[MG][PAWN];
}
}
template<>
void MovePicker::score<QUIETS>() {
Move m;
for (ExtMove* it = moves; it != end; ++it)
{
m = it->move;
it->score = history[pos.piece_moved(m)][to_sq(m)];
}
}
template<>
void MovePicker::score<EVASIONS>() {
// Try good captures ordered by MVV/LVA, then non-captures if destination square
// is not under attack, ordered by history value, then bad-captures and quiet
// moves with a negative SEE. This last group is ordered by the SEE score.
Move m;
int seeScore;
for (ExtMove* it = moves; it != end; ++it)
{
m = it->move;
if ((seeScore = pos.see_sign(m)) < 0)
it->score = seeScore - HistoryStats::Max; // At the bottom
else if (pos.is_capture(m))
it->score = PieceValue[MG][pos.piece_on(to_sq(m))]
- type_of(pos.piece_moved(m)) + HistoryStats::Max;
else else
it->score = history[pos.piece_moved(m)][to_sq(m)]; m.value = (*mainHistory)[pos.side_to_move()][from_to(m)]
+ (*continuationHistory[0])[pos.moved_piece(m)][to_sq(m)]
- (1 << 28);
} }
} }
/// MovePicker::select() returns the next move satisfying a predicate function.
/// It never returns the TT move.
template<MovePicker::PickType T, typename Pred>
Move MovePicker::select(Pred filter) {
/// generate_next() generates, scores and sorts the next bunch of moves, when while (cur < endMoves)
/// there are no more moves to try for the current phase.
void MovePicker::generate_next() {
cur = moves;
switch (++stage) {
case CAPTURES_S1: case CAPTURES_S3: case CAPTURES_S4: case CAPTURES_S5: case CAPTURES_S6:
end = generate<CAPTURES>(pos, moves);
score<CAPTURES>();
return;
case KILLERS_S1:
cur = killers;
end = cur + 2;
killers[0].move = ss->killers[0];
killers[1].move = ss->killers[1];
killers[2].move = killers[3].move = MOVE_NONE;
// Be sure countermoves are different from killers
for (int i = 0; i < 2; i++)
if (countermoves[i] != cur->move && countermoves[i] != (cur+1)->move)
(end++)->move = countermoves[i];
if (countermoves[1] && countermoves[1] == countermoves[0]) // Due to SMP races
killers[3].move = MOVE_NONE;
return;
case QUIETS_1_S1:
endQuiets = end = generate<QUIETS>(pos, moves);
score<QUIETS>();
end = std::partition(cur, end, has_positive_score);
insertion_sort(cur, end);
return;
case QUIETS_2_S1:
cur = end;
end = endQuiets;
if (depth >= 3 * ONE_PLY)
insertion_sort(cur, end);
return;
case BAD_CAPTURES_S1:
// Just pick them in reverse order to get MVV/LVA ordering
cur = moves + MAX_MOVES - 1;
end = endBadCaptures;
return;
case EVASIONS_S2:
end = generate<EVASIONS>(pos, moves);
if (end > moves + 1)
score<EVASIONS>();
return;
case QUIET_CHECKS_S3:
end = generate<QUIET_CHECKS>(pos, moves);
return;
case EVASION: case QSEARCH_0: case QSEARCH_1: case PROBCUT: case RECAPTURE:
stage = STOP;
case STOP:
end = cur + 1; // Avoid another next_phase() call
return;
default:
assert(false);
}
}
/// next_move() is the most important method of the MovePicker class. It returns
/// a new pseudo legal move every time is called, until there are no more moves
/// left. It picks the move with the biggest score from a list of generated moves
/// taking care not returning the ttMove if has already been searched previously.
template<>
Move MovePicker::next_move<false>() {
Move move;
while (true)
{ {
while (cur == end) if (T == Best)
generate_next(); std::swap(*cur, *std::max_element(cur, endMoves));
move = *cur++;
if (move != ttMove && filter())
return move;
}
return move = MOVE_NONE;
}
/// MovePicker::next_move() is the most important method of the MovePicker class. It
/// returns a new pseudo legal move every time it is called until there are no more
/// moves left, picking the move with the highest score from a list of generated moves.
Move MovePicker::next_move(bool skipQuiets) {
top:
switch (stage) { switch (stage) {
case MAIN_SEARCH: case EVASION: case QSEARCH_0: case QSEARCH_1: case PROBCUT: case MAIN_TT:
cur++; case EVASION_TT:
case QSEARCH_TT:
case PROBCUT_TT:
++stage;
return ttMove; return ttMove;
case CAPTURES_S1: case CAPTURE_INIT:
move = pick_best(cur++, end)->move; case PROBCUT_INIT:
if (move != ttMove) case QCAPTURE_INIT:
{ cur = endBadCaptures = moves;
if (pos.see_sign(move) >= 0) endMoves = generate<CAPTURES>(pos, cur);
score<CAPTURES>();
++stage;
goto top;
case GOOD_CAPTURE:
if (select<Best>([&](){
return pos.see_ge(move, Value(-55 * (cur-1)->value / 1024)) ?
// Move losing capture to endBadCaptures to be tried later
true : (*endBadCaptures++ = move, false); }))
return move; return move;
// Losing capture, move it to the tail of the array // Prepare the pointers to loop over the refutations array
(endBadCaptures--)->move = move; cur = std::begin(refutations);
} endMoves = std::end(refutations);
break;
case KILLERS_S1: // If the countermove is the same as a killer, skip it
move = (cur++)->move; if ( refutations[0].move == refutations[2].move
if ( move != MOVE_NONE || refutations[1].move == refutations[2].move)
&& pos.is_pseudo_legal(move) --endMoves;
&& move != ttMove
&& !pos.is_capture(move)) ++stage;
/* fallthrough */
case REFUTATION:
if (select<Next>([&](){ return move != MOVE_NONE
&& !pos.capture(move)
&& pos.pseudo_legal(move); }))
return move; return move;
break; ++stage;
/* fallthrough */
case QUIETS_1_S1: case QUIETS_2_S1: case QUIET_INIT:
move = (cur++)->move; cur = endBadCaptures;
if ( move != ttMove endMoves = generate<QUIETS>(pos, cur);
&& move != killers[0].move
&& move != killers[1].move score<QUIETS>();
&& move != killers[2].move partial_insertion_sort(cur, endMoves, -4000 * depth / ONE_PLY);
&& move != killers[3].move) ++stage;
/* fallthrough */
case QUIET:
if ( !skipQuiets
&& select<Next>([&](){return move != refutations[0]
&& move != refutations[1]
&& move != refutations[2];}))
return move; return move;
break;
case BAD_CAPTURES_S1: // Prepare the pointers to loop over the bad captures
return (cur--)->move; cur = moves;
endMoves = endBadCaptures;
case EVASIONS_S2: case CAPTURES_S3: case CAPTURES_S4: ++stage;
move = pick_best(cur++, end)->move; /* fallthrough */
if (move != ttMove)
case BAD_CAPTURE:
return select<Next>(Any);
case EVASION_INIT:
cur = moves;
endMoves = generate<EVASIONS>(pos, cur);
score<EVASIONS>();
++stage;
/* fallthrough */
case EVASION:
return select<Best>(Any);
case PROBCUT:
return select<Best>([&](){ return pos.see_ge(move, threshold); });
case QCAPTURE:
if (select<Best>([&](){ return depth > DEPTH_QS_RECAPTURES
|| to_sq(move) == recaptureSquare; }))
return move; return move;
break;
case CAPTURES_S5: // If we did not find any move and we do not try checks, we have finished
move = pick_best(cur++, end)->move; if (depth != DEPTH_QS_CHECKS)
if (move != ttMove && pos.see(move) > captureThreshold)
return move;
break;
case CAPTURES_S6:
move = pick_best(cur++, end)->move;
if (to_sq(move) == recaptureSquare)
return move;
break;
case QUIET_CHECKS_S3:
move = (cur++)->move;
if (move != ttMove)
return move;
break;
case STOP:
return MOVE_NONE; return MOVE_NONE;
default: ++stage;
/* fallthrough */
case QCHECK_INIT:
cur = moves;
endMoves = generate<QUIET_CHECKS>(pos, cur);
++stage;
/* fallthrough */
case QCHECK:
return select<Next>(Any);
}
assert(false); assert(false);
} return MOVE_NONE; // Silence warning
}
} }
/// Version of next_move() to use at split point nodes where the move is grabbed
/// from the split point's shared MovePicker object. This function is not thread
/// safe so must be lock protected by the caller.
template<>
Move MovePicker::next_move<true>() { return ss->splitPoint->movePicker->next_move<false>(); }
+95 -53
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,56 +21,87 @@
#ifndef MOVEPICK_H_INCLUDED #ifndef MOVEPICK_H_INCLUDED
#define MOVEPICK_H_INCLUDED #define MOVEPICK_H_INCLUDED
#include <algorithm> // For std::max #include <array>
#include <cstring> // For std::memset #include <limits>
#include <type_traits>
#include "movegen.h" #include "movegen.h"
#include "position.h" #include "position.h"
#include "search.h"
#include "types.h" #include "types.h"
/// StatsEntry stores the stat table value. It is usually a number but could
/// be a move or even a nested history. We use a class instead of naked value
/// to directly call history update operator<<() on the entry so to use stats
/// tables at caller sites as simple multi-dim arrays.
template<typename T, int D>
class StatsEntry {
/// The Stats struct stores moves statistics. According to the template parameter T entry;
/// the class can store History, Gains and Countermoves. History records how often
/// different moves have been successful or unsuccessful during the current search
/// and is used for reduction and move ordering decisions. Gains records the move's
/// best evaluation gain from one ply to the next and is used for pruning decisions.
/// Countermoves store the move that refute a previous one. Entries are stored
/// according only to moving piece and destination square, hence two moves with
/// different origin but same destination and piece will be considered identical.
template<bool Gain, typename T>
struct Stats {
static const Value Max = Value(2000); public:
void operator=(const T& v) { entry = v; }
T* operator&() { return &entry; }
T* operator->() { return &entry; }
operator const T&() const { return entry; }
const T* operator[](Piece p) const { return table[p]; } void operator<<(int bonus) {
void clear() { std::memset(table, 0, sizeof(table)); } assert(abs(bonus) <= D); // Ensure range is [-D, D]
static_assert(D <= std::numeric_limits<T>::max(), "D overflows T");
void update(Piece p, Square to, Move m) { entry += bonus - entry * abs(bonus) / D;
if (m == table[p][to].first) assert(abs(entry) <= D);
return;
table[p][to].second = table[p][to].first;
table[p][to].first = m;
} }
void update(Piece p, Square to, Value v) {
if (Gain)
table[p][to] = std::max(v, table[p][to] - 1);
else if (abs(table[p][to] + v) < Max)
table[p][to] += v;
}
private:
T table[PIECE_NB][SQUARE_NB];
}; };
typedef Stats< true, Value> GainsStats; /// Stats is a generic N-dimensional array used to store various statistics.
typedef Stats<false, Value> HistoryStats; /// The first template parameter T is the base type of the array, the second
typedef Stats<false, std::pair<Move, Move> > CountermovesStats; /// template parameter D limits the range of updates in [-D, D] when we update
/// values with the << operator, while the last parameters (Size and Sizes)
/// encode the dimensions of the array.
template <typename T, int D, int Size, int... Sizes>
struct Stats : public std::array<Stats<T, D, Sizes...>, Size>
{
typedef Stats<T, D, Size, Sizes...> stats;
void fill(const T& v) {
// For standard-layout 'this' points to first struct member
assert(std::is_standard_layout<stats>::value);
typedef StatsEntry<T, D> entry;
entry* p = reinterpret_cast<entry*>(this);
std::fill(p, p + sizeof(*this) / sizeof(entry), v);
}
};
template <typename T, int D, int Size>
struct Stats<T, D, Size> : public std::array<StatsEntry<T, D>, Size> {};
/// In stats table, D=0 means that the template parameter is not used
enum StatsParams { NOT_USED = 0 };
/// ButterflyHistory records how often quiet moves have been successful or
/// unsuccessful during the current search, and is used for reduction and move
/// ordering decisions. It uses 2 tables (one for each color) indexed by
/// the move's from and to squares, see chessprogramming.wikispaces.com/Butterfly+Boards
typedef Stats<int16_t, 10692, COLOR_NB, int(SQUARE_NB) * int(SQUARE_NB)> ButterflyHistory;
/// CounterMoveHistory stores counter moves indexed by [piece][to] of the previous
/// move, see chessprogramming.wikispaces.com/Countermove+Heuristic
typedef Stats<Move, NOT_USED, PIECE_NB, SQUARE_NB> CounterMoveHistory;
/// CapturePieceToHistory is addressed by a move's [piece][to][captured piece type]
typedef Stats<int16_t, 10692, PIECE_NB, SQUARE_NB, PIECE_TYPE_NB> CapturePieceToHistory;
/// PieceToHistory is like ButterflyHistory but is addressed by a move's [piece][to]
typedef Stats<int16_t, 29952, PIECE_NB, SQUARE_NB> PieceToHistory;
/// ContinuationHistory is the combined history of a given pair of moves, usually
/// the current one given a previous one. The nested history table is based on
/// PieceToHistory instead of ButterflyBoards.
typedef Stats<PieceToHistory, NOT_USED, PIECE_NB, SQUARE_NB> ContinuationHistory;
/// MovePicker class is used to pick one pseudo legal move at a time from the /// MovePicker class is used to pick one pseudo legal move at a time from the
@@ -78,32 +110,42 @@ typedef Stats<false, std::pair<Move, Move> > CountermovesStats;
/// when MOVE_NONE is returned. In order to improve the efficiency of the alpha /// when MOVE_NONE is returned. In order to improve the efficiency of the alpha
/// beta algorithm, MovePicker attempts to return the moves which are most likely /// beta algorithm, MovePicker attempts to return the moves which are most likely
/// to get a cut-off first. /// to get a cut-off first.
class MovePicker { class MovePicker {
MovePicker& operator=(const MovePicker&); // Silence a warning under MSVC enum PickType { Next, Best };
public: public:
MovePicker(const Position&, Move, Depth, const HistoryStats&, Square); MovePicker(const MovePicker&) = delete;
MovePicker(const Position&, Move, const HistoryStats&, PieceType); MovePicker& operator=(const MovePicker&) = delete;
MovePicker(const Position&, Move, Depth, const HistoryStats&, Move*, Search::Stack*); MovePicker(const Position&, Move, Value, const CapturePieceToHistory*);
MovePicker(const Position&, Move, Depth, const ButterflyHistory*,
template<bool SpNode> Move next_move(); const CapturePieceToHistory*,
const PieceToHistory**,
Square);
MovePicker(const Position&, Move, Depth, const ButterflyHistory*,
const CapturePieceToHistory*,
const PieceToHistory**,
Move,
Move*);
Move next_move(bool skipQuiets = false);
private: private:
template<PickType T, typename Pred> Move select(Pred);
template<GenType> void score(); template<GenType> void score();
void generate_next(); ExtMove* begin() { return cur; }
ExtMove* end() { return endMoves; }
const Position& pos; const Position& pos;
const HistoryStats& history; const ButterflyHistory* mainHistory;
Search::Stack* ss; const CapturePieceToHistory* captureHistory;
Move* countermoves; const PieceToHistory** continuationHistory;
Depth depth;
Move ttMove; Move ttMove;
ExtMove killers[4]; ExtMove refutations[3], *cur, *endMoves, *endBadCaptures;
int stage;
Move move;
Square recaptureSquare; Square recaptureSquare;
int captureThreshold, stage; Value threshold;
ExtMove *cur, *end, *endQuiets, *endBadCaptures; Depth depth;
ExtMove moves[MAX_MOVES]; ExtMove moves[MAX_MOVES];
}; };
-263
View File
@@ -1,263 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include <cassert>
#include <iomanip>
#include <sstream>
#include <stack>
#include "movegen.h"
#include "notation.h"
#include "position.h"
using namespace std;
static const char* PieceToChar[COLOR_NB] = { " PNBRQK", " pnbrqk" };
/// score_to_uci() converts a value to a string suitable for use with the UCI
/// protocol specifications:
///
/// cp <x> The score from the engine's point of view in centipawns.
/// mate <y> Mate in y moves, not plies. If the engine is getting mated
/// use negative values for y.
string score_to_uci(Value v, Value alpha, Value beta) {
stringstream s;
if (abs(v) < VALUE_MATE_IN_MAX_PLY)
s << "cp " << v * 100 / int(PawnValueMg);
else
s << "mate " << (v > 0 ? VALUE_MATE - v + 1 : -VALUE_MATE - v) / 2;
s << (v >= beta ? " lowerbound" : v <= alpha ? " upperbound" : "");
return s.str();
}
/// move_to_uci() converts a move to a string in coordinate notation
/// (g1f3, a7a8q, etc.). The only special case is castling moves, where we print
/// in the e1g1 notation in normal chess mode, and in e1h1 notation in chess960
/// mode. Internally castle moves are always coded as "king captures rook".
const string move_to_uci(Move m, bool chess960) {
Square from = from_sq(m);
Square to = to_sq(m);
if (m == MOVE_NONE)
return "(none)";
if (m == MOVE_NULL)
return "0000";
if (type_of(m) == CASTLE && !chess960)
to = (to > from ? FILE_G : FILE_C) | rank_of(from);
string move = square_to_string(from) + square_to_string(to);
if (type_of(m) == PROMOTION)
move += PieceToChar[BLACK][promotion_type(m)]; // Lower case
return move;
}
/// move_from_uci() takes a position and a string representing a move in
/// simple coordinate notation and returns an equivalent legal Move if any.
Move move_from_uci(const Position& pos, string& str) {
if (str.length() == 5) // Junior could send promotion piece in uppercase
str[4] = char(tolower(str[4]));
for (MoveList<LEGAL> it(pos); *it; ++it)
if (str == move_to_uci(*it, pos.is_chess960()))
return *it;
return MOVE_NONE;
}
/// move_to_san() takes a position and a legal Move as input and returns its
/// short algebraic notation representation.
const string move_to_san(Position& pos, Move m) {
if (m == MOVE_NONE)
return "(none)";
if (m == MOVE_NULL)
return "(null)";
assert(MoveList<LEGAL>(pos).contains(m));
Bitboard others, b;
string san;
Color us = pos.side_to_move();
Square from = from_sq(m);
Square to = to_sq(m);
Piece pc = pos.piece_on(from);
PieceType pt = type_of(pc);
if (type_of(m) == CASTLE)
san = to > from ? "O-O" : "O-O-O";
else
{
if (pt != PAWN)
{
san = PieceToChar[WHITE][pt]; // Upper case
// Disambiguation if we have more then one piece of type 'pt' that can
// reach 'to' with a legal move.
others = b = (pos.attacks_from(pc, to) & pos.pieces(us, pt)) ^ from;
while (b)
{
Move move = make_move(pop_lsb(&b), to);
if (!pos.pl_move_is_legal(move, pos.pinned_pieces()))
others ^= from_sq(move);
}
if (others)
{
if (!(others & file_bb(from)))
san += file_to_char(file_of(from));
else if (!(others & rank_bb(from)))
san += rank_to_char(rank_of(from));
else
san += square_to_string(from);
}
}
else if (pos.is_capture(m))
san = file_to_char(file_of(from));
if (pos.is_capture(m))
san += 'x';
san += square_to_string(to);
if (type_of(m) == PROMOTION)
san += string("=") + PieceToChar[WHITE][promotion_type(m)];
}
if (pos.move_gives_check(m, CheckInfo(pos)))
{
StateInfo st;
pos.do_move(m, st);
san += MoveList<LEGAL>(pos).size() ? "+" : "#";
pos.undo_move(m);
}
return san;
}
/// pretty_pv() formats human-readable search information, typically to be
/// appended to the search log file. It uses the two helpers below to pretty
/// format time and score respectively.
static string time_to_string(int64_t msecs) {
const int MSecMinute = 1000 * 60;
const int MSecHour = 1000 * 60 * 60;
int64_t hours = msecs / MSecHour;
int64_t minutes = (msecs % MSecHour) / MSecMinute;
int64_t seconds = ((msecs % MSecHour) % MSecMinute) / 1000;
stringstream s;
if (hours)
s << hours << ':';
s << setfill('0') << setw(2) << minutes << ':' << setw(2) << seconds;
return s.str();
}
static string score_to_string(Value v) {
stringstream s;
if (v >= VALUE_MATE_IN_MAX_PLY)
s << "#" << (VALUE_MATE - v + 1) / 2;
else if (v <= VALUE_MATED_IN_MAX_PLY)
s << "-#" << (VALUE_MATE + v) / 2;
else
s << setprecision(2) << fixed << showpos << float(v) / PawnValueMg;
return s.str();
}
string pretty_pv(Position& pos, int depth, Value value, int64_t msecs, Move pv[]) {
const int64_t K = 1000;
const int64_t M = 1000000;
std::stack<StateInfo> st;
Move* m = pv;
string san, padding;
size_t length;
stringstream s;
s << setw(2) << depth
<< setw(8) << score_to_string(value)
<< setw(8) << time_to_string(msecs);
if (pos.nodes_searched() < M)
s << setw(8) << pos.nodes_searched() / 1 << " ";
else if (pos.nodes_searched() < K * M)
s << setw(7) << pos.nodes_searched() / K << "K ";
else
s << setw(7) << pos.nodes_searched() / M << "M ";
padding = string(s.str().length(), ' ');
length = padding.length();
while (*m != MOVE_NONE)
{
san = move_to_san(pos, *m);
if (length + san.length() > 80)
{
s << "\n" + padding;
length = padding.length();
}
s << san << ' ';
length += san.length() + 1;
st.push(StateInfo());
pos.do_move(*m++, st.top());
}
while (m != pv)
pos.undo_move(*--m);
return s.str();
}
-35
View File
@@ -1,35 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef NOTATION_H_INCLUDED
#define NOTATION_H_INCLUDED
#include <string>
#include "types.h"
class Position;
std::string score_to_uci(Value v, Value alpha = -VALUE_INFINITE, Value beta = VALUE_INFINITE);
Move move_from_uci(const Position& pos, std::string& str);
const std::string move_to_uci(Move m, bool chess960);
const std::string move_to_san(Position& pos, Move m);
std::string pretty_pv(Position& pos, int depth, Value score, int64_t msecs, Move pv[]);
#endif // #ifndef NOTATION_H_INCLUDED
+148 -166
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -21,90 +22,66 @@
#include <cassert> #include <cassert>
#include "bitboard.h" #include "bitboard.h"
#include "bitcount.h"
#include "pawns.h" #include "pawns.h"
#include "position.h" #include "position.h"
#include "thread.h"
namespace { namespace {
#define V Value #define V Value
#define S(mg, eg) make_score(mg, eg) #define S(mg, eg) make_score(mg, eg)
// Doubled pawn penalty by opposed flag and file // Pawn penalties
const Score Doubled[2][FILE_NB] = { constexpr Score Backward = S( 9, 24);
{ S(13, 43), S(20, 48), S(23, 48), S(23, 48), constexpr Score Doubled = S(11, 56);
S(23, 48), S(23, 48), S(20, 48), S(13, 43) }, constexpr Score Isolated = S( 5, 15);
{ S(13, 43), S(20, 48), S(23, 48), S(23, 48),
S(23, 48), S(23, 48), S(20, 48), S(13, 43) }};
// Isolated pawn penalty by opposed flag and file // Connected pawn bonus by opposed, phalanx, #support and rank
const Score Isolated[2][FILE_NB] = { Score Connected[2][2][3][RANK_NB];
{ S(37, 45), S(54, 52), S(60, 52), S(60, 52),
S(60, 52), S(60, 52), S(54, 52), S(37, 45) },
{ S(25, 30), S(36, 35), S(40, 35), S(40, 35),
S(40, 35), S(40, 35), S(36, 35), S(25, 30) }};
// Backward pawn penalty by opposed flag and file // Strength of pawn shelter for our king by [distance from edge][rank].
const Score Backward[2][FILE_NB] = { // RANK_1 = 0 is used for files where we have no pawn, or pawn is behind our king.
{ S(30, 42), S(43, 46), S(49, 46), S(49, 46), constexpr Value ShelterStrength[int(FILE_NB) / 2][RANK_NB] = {
S(49, 46), S(49, 46), S(43, 46), S(30, 42) }, { V( -6), V( 81), V( 93), V( 58), V( 39), V( 18), V( 25) },
{ S(20, 28), S(29, 31), S(33, 31), S(33, 31), { V(-43), V( 61), V( 35), V(-49), V(-29), V(-11), V( -63) },
S(33, 31), S(33, 31), S(29, 31), S(20, 28) }}; { V(-10), V( 75), V( 23), V( -2), V( 32), V( 3), V( -45) },
{ V(-39), V(-13), V(-29), V(-52), V(-48), V(-67), V(-166) }
// Pawn chain membership bonus by file
const Score ChainMember[FILE_NB] = {
S(11,-1), S(13,-1), S(13,-1), S(14,-1),
S(14,-1), S(13,-1), S(13,-1), S(11,-1)
}; };
// Candidate passed pawn bonus by rank // Danger of enemy pawns moving toward our king by [distance from edge][rank].
const Score CandidatePassed[RANK_NB] = { // RANK_1 = 0 is used for files where the enemy has no pawn, or their pawn
S( 0, 0), S( 6, 13), S(6,13), S(14,29), // is behind our king.
S(34,68), S(83,166), S(0, 0), S( 0, 0) constexpr Value UnblockedStorm[int(FILE_NB) / 2][RANK_NB] = {
{ V( 89), V(107), V(123), V(93), V(57), V( 45), V( 51) },
{ V( 44), V(-18), V(123), V(46), V(39), V( -7), V( 23) },
{ V( 4), V( 52), V(162), V(37), V( 7), V(-14), V( -2) },
{ V(-10), V(-14), V( 90), V(15), V( 2), V( -7), V(-16) }
}; };
// Weakness of our pawn shelter in front of the king indexed by [rank]
const Value ShelterWeakness[RANK_NB] =
{ V(100), V(0), V(27), V(73), V(92), V(101), V(101) };
// Danger of enemy pawns moving toward our king indexed by
// [no friendly pawn | pawn unblocked | pawn blocked][rank of enemy pawn]
const Value StormDanger[3][RANK_NB] = {
{ V( 0), V(64), V(128), V(51), V(26) },
{ V(26), V(32), V( 96), V(38), V(20) },
{ V( 0), V( 0), V( 64), V(25), V(13) }};
// Max bonus for king safety. Corresponds to start position with all the pawns
// in front of the king and no enemy pawn on the horizont.
const Value MaxSafetyBonus = V(263);
#undef S #undef S
#undef V #undef V
template<Color Us> template<Color Us>
Score evaluate(const Position& pos, Pawns::Entry* e) { Score evaluate(const Position& pos, Pawns::Entry* e) {
const Color Them = (Us == WHITE ? BLACK : WHITE); constexpr Color Them = (Us == WHITE ? BLACK : WHITE);
const Square Up = (Us == WHITE ? DELTA_N : DELTA_S); constexpr Direction Up = (Us == WHITE ? NORTH : SOUTH);
const Square Right = (Us == WHITE ? DELTA_NE : DELTA_SW);
const Square Left = (Us == WHITE ? DELTA_NW : DELTA_SE);
Bitboard b; Bitboard b, neighbours, stoppers, doubled, supported, phalanx;
Bitboard lever, leverPush;
Square s; Square s;
File f; bool opposed, backward;
Rank r; Score score = SCORE_ZERO;
bool passed, isolated, doubled, opposed, chain, backward, candidate; const Square* pl = pos.squares<PAWN>(Us);
Score value = SCORE_ZERO;
const Square* pl = pos.list<PAWN>(Us);
Bitboard ourPawns = pos.pieces(Us, PAWN); Bitboard ourPawns = pos.pieces( Us, PAWN);
Bitboard theirPawns = pos.pieces(Them, PAWN); Bitboard theirPawns = pos.pieces(Them, PAWN);
e->passedPawns[Us] = 0; e->passedPawns[Us] = e->pawnAttacksSpan[Us] = e->weakUnopposed[Us] = 0;
e->kingSquares[Us] = SQ_NONE;
e->semiopenFiles[Us] = 0xFF; e->semiopenFiles[Us] = 0xFF;
e->pawnAttacks[Us] = shift_bb<Right>(ourPawns) | shift_bb<Left>(ourPawns); e->kingSquares[Us] = SQ_NONE;
e->pawnsOnSquares[Us][BLACK] = popcount<Max15>(ourPawns & DarkSquares); e->pawnAttacks[Us] = pawn_attacks_bb<Us>(ourPawns);
e->pawnsOnSquares[Us][BLACK] = popcount(ourPawns & DarkSquares);
e->pawnsOnSquares[Us][WHITE] = pos.count<PAWN>(Us) - e->pawnsOnSquares[Us][BLACK]; e->pawnsOnSquares[Us][WHITE] = pos.count<PAWN>(Us) - e->pawnsOnSquares[Us][BLACK];
// Loop through all pawns of the current color and score each pawn // Loop through all pawns of the current color and score each pawn
@@ -112,170 +89,175 @@ namespace {
{ {
assert(pos.piece_on(s) == make_piece(Us, PAWN)); assert(pos.piece_on(s) == make_piece(Us, PAWN));
f = file_of(s); File f = file_of(s);
r = rank_of(s);
// This file cannot be semi-open
e->semiopenFiles[Us] &= ~(1 << f); e->semiopenFiles[Us] &= ~(1 << f);
e->pawnAttacksSpan[Us] |= pawn_attack_span(Us, s);
// Our rank plus previous one. Used for chain detection // Flag the pawn
b = rank_bb(r) | rank_bb(Us == WHITE ? r - Rank(1) : r + Rank(1)); opposed = theirPawns & forward_file_bb(Us, s);
stoppers = theirPawns & passed_pawn_mask(Us, s);
lever = theirPawns & PawnAttacks[Us][s];
leverPush = theirPawns & PawnAttacks[Us][s + Up];
doubled = ourPawns & (s - Up);
neighbours = ourPawns & adjacent_files_bb(f);
phalanx = neighbours & rank_bb(s);
supported = neighbours & rank_bb(s - Up);
// Flag the pawn as passed, isolated, doubled or member of a pawn // A pawn is backward when it is behind all pawns of the same color
// chain (but not the backward one). // on the adjacent files and cannot be safely advanced.
chain = ourPawns & adjacent_files_bb(f) & b; backward = !(ourPawns & pawn_attack_span(Them, s + Up))
isolated = !(ourPawns & adjacent_files_bb(f)); && (stoppers & (leverPush | (s + Up)));
doubled = ourPawns & forward_bb(Us, s);
opposed = theirPawns & forward_bb(Us, s);
passed = !(theirPawns & passed_pawn_mask(Us, s));
// Test for backward pawn
backward = false;
// If the pawn is passed, isolated, or member of a pawn chain it cannot
// be backward. If there are friendly pawns behind on adjacent files
// or if can capture an enemy pawn it cannot be backward either.
if ( !(passed | isolated | chain)
&& !(ourPawns & pawn_attack_span(Them, s))
&& !(pos.attacks_from<PAWN>(s, Us) & theirPawns))
{
// We now know that there are no friendly pawns beside or behind this
// pawn on adjacent files. We now check whether the pawn is
// backward by looking in the forward direction on the adjacent
// files, and seeing whether we meet a friendly or an enemy pawn first.
b = pos.attacks_from<PAWN>(s, Us);
// Note that we are sure to find something because pawn is not passed
// nor isolated, so loop is potentially infinite, but it isn't.
while (!(b & (ourPawns | theirPawns)))
b = shift_bb<Up>(b);
// The friendly pawn needs to be at least two ranks closer than the
// enemy pawn in order to help the potentially backward pawn advance.
backward = (b | shift_bb<Up>(b)) & theirPawns;
}
assert(opposed | passed | (pawn_attack_span(Us, s) & theirPawns));
// A not passed pawn is a candidate to become passed if it is free to
// advance and if the number of friendly pawns beside or behind this
// pawn on adjacent files is higher or equal than the number of
// enemy pawns in the forward direction on the adjacent files.
candidate = !(opposed | passed | backward | isolated)
&& (b = pawn_attack_span(Them, s + pawn_push(Us)) & ourPawns) != 0
&& popcount<Max15>(b) >= popcount<Max15>(pawn_attack_span(Us, s) & theirPawns);
// Passed pawns will be properly scored in evaluation because we need // Passed pawns will be properly scored in evaluation because we need
// full attack info to evaluate passed pawns. Only the frontmost passed // full attack info to evaluate them. Include also not passed pawns
// pawn on each file is considered a true passed pawn. // which could become passed after one or two pawn pushes when are
if (passed && !doubled) // not attacked more times than defended.
if ( !(stoppers ^ lever ^ leverPush)
&& popcount(supported) >= popcount(lever) - 1
&& popcount(phalanx) >= popcount(leverPush))
e->passedPawns[Us] |= s; e->passedPawns[Us] |= s;
// Score this pawn else if ( stoppers == SquareBB[s + Up]
if (isolated) && relative_rank(Us, s) >= RANK_5)
value -= Isolated[opposed][f]; {
b = shift<Up>(supported) & ~theirPawns;
if (doubled) while (b)
value -= Doubled[opposed][f]; if (!more_than_one(theirPawns & PawnAttacks[Us][pop_lsb(&b)]))
e->passedPawns[Us] |= s;
if (backward)
value -= Backward[opposed][f];
if (chain)
value += ChainMember[f];
if (candidate)
value += CandidatePassed[relative_rank(Us, s)];
} }
return value; // Score this pawn
if (supported | phalanx)
score += Connected[opposed][bool(phalanx)][popcount(supported)][relative_rank(Us, s)];
else if (!neighbours)
score -= Isolated, e->weakUnopposed[Us] += !opposed;
else if (backward)
score -= Backward, e->weakUnopposed[Us] += !opposed;
if (doubled && !supported)
score -= Doubled;
}
return score;
} }
} // namespace } // namespace
namespace Pawns { namespace Pawns {
/// probe() takes a position object as input, computes a Entry object, and returns /// Pawns::init() initializes some tables needed by evaluation. Instead of using
/// a pointer to it. The result is also stored in a hash table, so we don't have /// hard-coded tables, when makes sense, we prefer to calculate them with a formula
/// to recompute everything when the same pawn structure occurs again. /// to reduce independent parameters and to allow easier tuning and better insight.
Entry* probe(const Position& pos, Table& entries) { void init() {
static constexpr int Seed[RANK_NB] = { 0, 13, 24, 18, 65, 100, 175, 330 };
for (int opposed = 0; opposed <= 1; ++opposed)
for (int phalanx = 0; phalanx <= 1; ++phalanx)
for (int support = 0; support <= 2; ++support)
for (Rank r = RANK_2; r < RANK_8; ++r)
{
int v = 17 * support;
v += (Seed[r] + (phalanx ? (Seed[r + 1] - Seed[r]) / 2 : 0)) >> opposed;
Connected[opposed][phalanx][support][r] = make_score(v, v * (r - 2) / 4);
}
}
/// Pawns::probe() looks up the current position's pawns configuration in
/// the pawns hash table. It returns a pointer to the Entry if the position
/// is found. Otherwise a new Entry is computed and stored there, so we don't
/// have to recompute all when the same pawns configuration occurs again.
Entry* probe(const Position& pos) {
Key key = pos.pawn_key(); Key key = pos.pawn_key();
Entry* e = entries[key]; Entry* e = pos.this_thread()->pawnsTable[key];
if (e->key == key) if (e->key == key)
return e; return e;
e->key = key; e->key = key;
e->value = evaluate<WHITE>(pos, e) - evaluate<BLACK>(pos, e); e->scores[WHITE] = evaluate<WHITE>(pos, e);
e->scores[BLACK] = evaluate<BLACK>(pos, e);
e->openFiles = popcount(e->semiopenFiles[WHITE] & e->semiopenFiles[BLACK]);
e->asymmetry = popcount( (e->passedPawns[WHITE] | e->passedPawns[BLACK])
| (e->semiopenFiles[WHITE] ^ e->semiopenFiles[BLACK]));
return e; return e;
} }
/// Entry::shelter_storm() calculates shelter and storm penalties for the file /// Entry::evaluate_shelter() calculates the shelter bonus and the storm
/// the king is on, as well as the two adjacent files. /// penalty for a king, looking at the king file and the two closest files.
template<Color Us> template<Color Us>
Value Entry::shelter_storm(const Position& pos, Square ksq) { Value Entry::evaluate_shelter(const Position& pos, Square ksq) {
const Color Them = (Us == WHITE ? BLACK : WHITE); constexpr Color Them = (Us == WHITE ? BLACK : WHITE);
constexpr Direction Down = (Us == WHITE ? SOUTH : NORTH);
constexpr Bitboard BlockRanks = (Us == WHITE ? Rank1BB | Rank2BB : Rank8BB | Rank7BB);
Value safety = MaxSafetyBonus; Bitboard b = pos.pieces(PAWN) & ~forward_ranks_bb(Them, ksq);
Bitboard b = pos.pieces(PAWN) & (in_front_bb(Us, rank_of(ksq)) | rank_bb(ksq));
Bitboard ourPawns = b & pos.pieces(Us); Bitboard ourPawns = b & pos.pieces(Us);
Bitboard theirPawns = b & pos.pieces(Them); Bitboard theirPawns = b & pos.pieces(Them);
Rank rkUs, rkThem;
File kf = file_of(ksq);
kf = (kf == FILE_A) ? FILE_B : (kf == FILE_H) ? FILE_G : kf; Value safety = (shift<Down>(theirPawns) & (FileABB | FileHBB) & BlockRanks & ksq) ?
Value(374) : Value(5);
for (int f = kf - 1; f <= kf + 1; f++) File center = std::max(FILE_B, std::min(FILE_G, file_of(ksq)));
for (File f = File(center - 1); f <= File(center + 1); ++f)
{ {
b = ourPawns & FileBB[f]; b = ourPawns & file_bb(f);
rkUs = b ? relative_rank(Us, Us == WHITE ? lsb(b) : msb(b)) : RANK_1; int ourRank = b ? relative_rank(Us, backmost_sq(Us, b)) : 0;
safety -= ShelterWeakness[rkUs];
b = theirPawns & FileBB[f]; b = theirPawns & file_bb(f);
rkThem = b ? relative_rank(Us, Us == WHITE ? lsb(b) : msb(b)) : RANK_1; int theirRank = b ? relative_rank(Us, frontmost_sq(Them, b)) : 0;
safety -= StormDanger[rkUs == RANK_1 ? 0 : rkThem == rkUs + 1 ? 2 : 1][rkThem];
int d = std::min(f, ~f);
safety += ShelterStrength[d][ourRank];
safety -= (ourRank && (ourRank == theirRank - 1)) ? 66 * (theirRank == RANK_3)
: UnblockedStorm[d][theirRank];
} }
return safety; return safety;
} }
/// Entry::update_safety() calculates and caches a bonus for king safety. It is /// Entry::do_king_safety() calculates a bonus for king safety. It is called only
/// called only when king square changes, about 20% of total king_safety() calls. /// when king square changes, which is about 20% of total king_safety() calls.
template<Color Us> template<Color Us>
Score Entry::update_safety(const Position& pos, Square ksq) { Score Entry::do_king_safety(const Position& pos) {
Square ksq = pos.square<KING>(Us);
kingSquares[Us] = ksq; kingSquares[Us] = ksq;
castleRights[Us] = pos.can_castle(Us); castlingRights[Us] = pos.can_castle(Us);
minKPdistance[Us] = 0; int minKingPawnDistance = 0;
Bitboard pawns = pos.pieces(Us, PAWN); Bitboard pawns = pos.pieces(Us, PAWN);
if (pawns) if (pawns)
while (!(DistanceRingsBB[ksq][minKPdistance[Us]++] & pawns)) {} while (!(DistanceRingBB[ksq][++minKingPawnDistance] & pawns)) {}
if (relative_rank(Us, ksq) > RANK_4) Value bonus = evaluate_shelter<Us>(pos, ksq);
return kingSafety[Us] = make_score(0, -16 * minKPdistance[Us]);
Value bonus = shelter_storm<Us>(pos, ksq); // If we can castle use the bonus after the castling if it is bigger
if (pos.can_castle(Us | KING_SIDE))
bonus = std::max(bonus, evaluate_shelter<Us>(pos, relative_square(Us, SQ_G1)));
// If we can castle use the bonus after the castle if is bigger if (pos.can_castle(Us | QUEEN_SIDE))
if (pos.can_castle(make_castle_right(Us, KING_SIDE))) bonus = std::max(bonus, evaluate_shelter<Us>(pos, relative_square(Us, SQ_C1)));
bonus = std::max(bonus, shelter_storm<Us>(pos, relative_square(Us, SQ_G1)));
if (pos.can_castle(make_castle_right(Us, QUEEN_SIDE))) return make_score(bonus, -16 * minKingPawnDistance);
bonus = std::max(bonus, shelter_storm<Us>(pos, relative_square(Us, SQ_C1)));
return kingSafety[Us] = make_score(bonus, -16 * minKPdistance[Us]);
} }
// Explicit template instantiation // Explicit template instantiation
template Score Entry::update_safety<WHITE>(const Position& pos, Square ksq); template Score Entry::do_king_safety<WHITE>(const Position& pos);
template Score Entry::update_safety<BLACK>(const Position& pos, Square ksq); template Score Entry::do_king_safety<BLACK>(const Position& pos);
} // namespace Pawns } // namespace Pawns
+32 -24
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -26,53 +27,60 @@
namespace Pawns { namespace Pawns {
/// Pawns::Entry contains various information about a pawn structure. Currently, /// Pawns::Entry contains various information about a pawn structure. A lookup
/// it only includes a middle game and end game pawn structure evaluation, and a /// to the pawn hash table (performed by calling the probe function) returns a
/// bitboard of passed pawns. We may want to add further information in the future. /// pointer to an Entry object.
/// A lookup to the pawn hash table (performed by calling the probe function)
/// returns a pointer to an Entry object.
struct Entry { struct Entry {
Score pawns_value() const { return value; } Score pawn_score(Color c) const { return scores[c]; }
Bitboard pawn_attacks(Color c) const { return pawnAttacks[c]; } Bitboard pawn_attacks(Color c) const { return pawnAttacks[c]; }
Bitboard passed_pawns(Color c) const { return passedPawns[c]; } Bitboard passed_pawns(Color c) const { return passedPawns[c]; }
int pawns_on_same_color_squares(Color c, Square s) const { return pawnsOnSquares[c][!!(DarkSquares & s)]; } Bitboard pawn_attacks_span(Color c) const { return pawnAttacksSpan[c]; }
int semiopen(Color c, File f) const { return semiopenFiles[c] & (1 << int(f)); } int weak_unopposed(Color c) const { return weakUnopposed[c]; }
int semiopen_on_side(Color c, File f, bool left) const { int pawn_asymmetry() const { return asymmetry; }
int open_files() const { return openFiles; }
return semiopenFiles[c] & (left ? ((1 << int(f)) - 1) : ~((1 << int(f+1)) - 1)); int semiopen_file(Color c, File f) const {
return semiopenFiles[c] & (1 << f);
}
int pawns_on_same_color_squares(Color c, Square s) const {
return pawnsOnSquares[c][bool(DarkSquares & s)];
} }
template<Color Us> template<Color Us>
Score king_safety(const Position& pos, Square ksq) { Score king_safety(const Position& pos) {
return kingSquares[Us] == pos.square<KING>(Us) && castlingRights[Us] == pos.can_castle(Us)
return kingSquares[Us] == ksq && castleRights[Us] == pos.can_castle(Us) ? kingSafety[Us] : (kingSafety[Us] = do_king_safety<Us>(pos));
? kingSafety[Us] : update_safety<Us>(pos, ksq);
} }
template<Color Us> template<Color Us>
Score update_safety(const Position& pos, Square ksq); Score do_king_safety(const Position& pos);
template<Color Us> template<Color Us>
Value shelter_storm(const Position& pos, Square ksq); Value evaluate_shelter(const Position& pos, Square ksq);
Key key; Key key;
Score scores[COLOR_NB];
Bitboard passedPawns[COLOR_NB]; Bitboard passedPawns[COLOR_NB];
Bitboard pawnAttacks[COLOR_NB]; Bitboard pawnAttacks[COLOR_NB];
Bitboard pawnAttacksSpan[COLOR_NB];
Square kingSquares[COLOR_NB]; Square kingSquares[COLOR_NB];
int minKPdistance[COLOR_NB];
int castleRights[COLOR_NB];
Score value;
int semiopenFiles[COLOR_NB];
Score kingSafety[COLOR_NB]; Score kingSafety[COLOR_NB];
int pawnsOnSquares[COLOR_NB][COLOR_NB]; int weakUnopposed[COLOR_NB];
int castlingRights[COLOR_NB];
int semiopenFiles[COLOR_NB];
int pawnsOnSquares[COLOR_NB][COLOR_NB]; // [color][light/dark squares]
int asymmetry;
int openFiles;
}; };
typedef HashTable<Entry, 16384> Table; typedef HashTable<Entry, 16384> Table;
Entry* probe(const Position& pos, Table& entries); void init();
Entry* probe(const Position& pos);
} } // namespace Pawns
#endif // #ifndef PAWNS_H_INCLUDED #endif // #ifndef PAWNS_H_INCLUDED
-116
View File
@@ -1,116 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef PLATFORM_H_INCLUDED
#define PLATFORM_H_INCLUDED
#ifdef _MSC_VER
// Disable some silly and noisy warning from MSVC compiler
#pragma warning(disable: 4127) // Conditional expression is constant
#pragma warning(disable: 4146) // Unary minus operator applied to unsigned type
#pragma warning(disable: 4800) // Forcing value to bool 'true' or 'false'
#pragma warning(disable: 4996) // Function _ftime() may be unsafe
// MSVC does not support <inttypes.h>
typedef signed __int8 int8_t;
typedef unsigned __int8 uint8_t;
typedef signed __int16 int16_t;
typedef unsigned __int16 uint16_t;
typedef signed __int32 int32_t;
typedef unsigned __int32 uint32_t;
typedef signed __int64 int64_t;
typedef unsigned __int64 uint64_t;
#else
# include <inttypes.h>
#endif
#ifndef _WIN32 // Linux - Unix
# include <sys/time.h>
inline int64_t system_time_to_msec() {
timeval t;
gettimeofday(&t, NULL);
return t.tv_sec * 1000LL + t.tv_usec / 1000;
}
# include <pthread.h>
typedef pthread_mutex_t Lock;
typedef pthread_cond_t WaitCondition;
typedef pthread_t NativeHandle;
typedef void*(*pt_start_fn)(void*);
# define lock_init(x) pthread_mutex_init(&(x), NULL)
# define lock_grab(x) pthread_mutex_lock(&(x))
# define lock_release(x) pthread_mutex_unlock(&(x))
# define lock_destroy(x) pthread_mutex_destroy(&(x))
# define cond_destroy(x) pthread_cond_destroy(&(x))
# define cond_init(x) pthread_cond_init(&(x), NULL)
# define cond_signal(x) pthread_cond_signal(&(x))
# define cond_wait(x,y) pthread_cond_wait(&(x),&(y))
# define cond_timedwait(x,y,z) pthread_cond_timedwait(&(x),&(y),z)
# define thread_create(x,f,t) pthread_create(&(x),NULL,(pt_start_fn)f,t)
# define thread_join(x) pthread_join(x, NULL)
#else // Windows and MinGW
# include <sys/timeb.h>
inline int64_t system_time_to_msec() {
_timeb t;
_ftime(&t);
return t.time * 1000LL + t.millitm;
}
#ifndef NOMINMAX
# define NOMINMAX // disable macros min() and max()
#endif
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#undef WIN32_LEAN_AND_MEAN
#undef NOMINMAX
// We use critical sections on Windows to support Windows XP and older versions,
// unfortunatly cond_wait() is racy between lock_release() and WaitForSingleObject()
// but apart from this they have the same speed performance of SRW locks.
typedef CRITICAL_SECTION Lock;
typedef HANDLE WaitCondition;
typedef HANDLE NativeHandle;
// On Windows 95 and 98 parameter lpThreadId my not be null
inline DWORD* dwWin9xKludge() { static DWORD dw; return &dw; }
# define lock_init(x) InitializeCriticalSection(&(x))
# define lock_grab(x) EnterCriticalSection(&(x))
# define lock_release(x) LeaveCriticalSection(&(x))
# define lock_destroy(x) DeleteCriticalSection(&(x))
# define cond_init(x) { x = CreateEvent(0, FALSE, FALSE, 0); }
# define cond_destroy(x) CloseHandle(x)
# define cond_signal(x) SetEvent(x)
# define cond_wait(x,y) { lock_release(y); WaitForSingleObject(x, INFINITE); lock_grab(y); }
# define cond_timedwait(x,y,z) { lock_release(y); WaitForSingleObject(x,z); lock_grab(y); }
# define thread_create(x,f,t) (x = CreateThread(NULL,0,(LPTHREAD_START_ROUTINE)f,t,0,dwWin9xKludge()))
# define thread_join(x) { WaitForSingleObject(x, INFINITE); CloseHandle(x); }
#endif
#endif // #ifndef PLATFORM_H_INCLUDED
+670 -810
View File
File diff suppressed because it is too large Load Diff
+192 -220
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -21,86 +22,64 @@
#define POSITION_H_INCLUDED #define POSITION_H_INCLUDED
#include <cassert> #include <cassert>
#include <cstddef> #include <deque>
#include <memory> // For std::unique_ptr
#include <string>
#include "bitboard.h" #include "bitboard.h"
#include "types.h" #include "types.h"
/// The checkInfo struct is initialized at c'tor time and keeps info used /// StateInfo struct stores information needed to restore a Position object to
/// to detect if a move gives check. /// its previous state when we retract a move. Whenever a move is made on the
class Position; /// board (by calling Position::do_move), a StateInfo object must be passed.
struct Thread;
struct CheckInfo {
explicit CheckInfo(const Position&);
Bitboard dcCandidates;
Bitboard pinned;
Bitboard checkSq[PIECE_TYPE_NB];
Square ksq;
};
/// The StateInfo struct stores information we need to restore a Position
/// object to its previous state when we retract a move. Whenever a move
/// is made on the board (by calling Position::do_move), a StateInfo object
/// must be passed as a parameter.
struct StateInfo { struct StateInfo {
Key pawnKey, materialKey;
Value npMaterial[COLOR_NB]; // Copied when making a move
int castleRights, rule50, pliesFromNull; Key pawnKey;
Score psq; Key materialKey;
Value nonPawnMaterial[COLOR_NB];
int castlingRights;
int rule50;
int pliesFromNull;
Square epSquare; Square epSquare;
// Not copied when making a move (will be recomputed anyhow)
Key key; Key key;
Bitboard checkersBB; Bitboard checkersBB;
PieceType capturedType; Piece capturedPiece;
StateInfo* previous; StateInfo* previous;
Bitboard blockersForKing[COLOR_NB];
Bitboard pinners[COLOR_NB];
Bitboard checkSquares[PIECE_TYPE_NB];
}; };
/// A list to keep track of the position states along the setup moves (from the
/// When making a move the current StateInfo up to 'key' excluded is copied to /// start position to the position just before the search starts). Needed by
/// the new one. Here we calculate the quad words (64bits) needed to be copied. /// 'draw by repetition' detection. Use a std::deque because pointers to
const size_t StateCopySize64 = offsetof(StateInfo, key) / sizeof(uint64_t) + 1; /// elements are not invalidated upon list resizing.
typedef std::unique_ptr<std::deque<StateInfo>> StateListPtr;
/// The position data structure. A position consists of the following data: /// Position class stores information regarding the board representation as
/// /// pieces, side to move, hash keys, castling info, etc. Important methods are
/// * For each piece type, a bitboard representing the squares occupied /// do_move() and undo_move(), used by the search to update node info when
/// by pieces of that type. /// traversing the search tree.
/// * For each color, a bitboard representing the squares occupied by class Thread;
/// pieces of that color.
/// * A bitboard of all occupied squares.
/// * A bitboard of all checking pieces.
/// * A 64-entry array of pieces, indexed by the squares of the board.
/// * The current side to move.
/// * Information about the castling rights for both sides.
/// * The initial files of the kings and both pairs of rooks. This is
/// used to implement the Chess960 castling rules.
/// * The en passant square (which is SQ_NONE if no en passant capture is
/// possible).
/// * The squares of the kings for both sides.
/// * Hash keys for the position itself, the current pawn structure, and
/// the current material situation.
/// * Hash keys for all previous positions in the game for detecting
/// repetition draws.
/// * A counter for detecting 50 move rule draws.
class Position { class Position {
public: public:
Position() {}
Position(const Position& p, Thread* t) { *this = p; thisThread = t; }
Position(const std::string& f, bool c960, Thread* t) { set(f, c960, t); }
Position& operator=(const Position&);
static void init(); static void init();
// Text input/output Position() = default;
void set(const std::string& fen, bool isChess960, Thread* th); Position(const Position&) = delete;
Position& operator=(const Position&) = delete;
// FEN string input/output
Position& set(const std::string& fenStr, bool isChess960, StateInfo* si, Thread* th);
Position& set(const std::string& code, Color c, StateInfo* si);
const std::string fen() const; const std::string fen() const;
const std::string pretty(Move m = MOVE_NONE) const;
// Position representation // Position representation
Bitboard pieces() const; Bitboard pieces() const;
@@ -110,147 +89,132 @@ public:
Bitboard pieces(Color c, PieceType pt) const; Bitboard pieces(Color c, PieceType pt) const;
Bitboard pieces(Color c, PieceType pt1, PieceType pt2) const; Bitboard pieces(Color c, PieceType pt1, PieceType pt2) const;
Piece piece_on(Square s) const; Piece piece_on(Square s) const;
Square king_square(Color c) const;
Square ep_square() const; Square ep_square() const;
bool is_empty(Square s) const; bool empty(Square s) const;
template<PieceType Pt> int count(Color c) const; template<PieceType Pt> int count(Color c) const;
template<PieceType Pt> const Square* list(Color c) const; template<PieceType Pt> int count() const;
template<PieceType Pt> const Square* squares(Color c) const;
template<PieceType Pt> Square square(Color c) const;
// Castling // Castling
int can_castle(CastleRight f) const;
int can_castle(Color c) const; int can_castle(Color c) const;
bool castle_impeded(Color c, CastlingSide s) const; int can_castle(CastlingRight cr) const;
Square castle_rook_square(Color c, CastlingSide s) const; bool castling_impeded(CastlingRight cr) const;
Square castling_rook_square(CastlingRight cr) const;
// Checking // Checking
Bitboard checkers() const; Bitboard checkers() const;
Bitboard discovered_check_candidates() const; Bitboard blockers_for_king(Color c) const;
Bitboard pinned_pieces() const; Bitboard check_squares(PieceType pt) const;
// Attacks to/from a given square // Attacks to/from a given square
Bitboard attackers_to(Square s) const; Bitboard attackers_to(Square s) const;
Bitboard attackers_to(Square s, Bitboard occ) const; Bitboard attackers_to(Square s, Bitboard occupied) const;
Bitboard attacks_from(Piece p, Square s) const; Bitboard attacks_from(PieceType pt, Square s) const;
static Bitboard attacks_from(Piece p, Square s, Bitboard occ);
template<PieceType> Bitboard attacks_from(Square s) const; template<PieceType> Bitboard attacks_from(Square s) const;
template<PieceType> Bitboard attacks_from(Square s, Color c) const; template<PieceType> Bitboard attacks_from(Square s, Color c) const;
Bitboard slider_blockers(Bitboard sliders, Square s, Bitboard& pinners) const;
// Properties of moves // Properties of moves
bool move_gives_check(Move m, const CheckInfo& ci) const; bool legal(Move m) const;
bool pl_move_is_legal(Move m, Bitboard pinned) const; bool pseudo_legal(const Move m) const;
bool is_pseudo_legal(const Move m) const; bool capture(Move m) const;
bool is_capture(Move m) const; bool capture_or_promotion(Move m) const;
bool is_capture_or_promotion(Move m) const; bool gives_check(Move m) const;
bool is_passed_pawn_push(Move m) const; bool advanced_pawn_push(Move m) const;
Piece piece_moved(Move m) const; Piece moved_piece(Move m) const;
PieceType captured_piece_type() const; Piece captured_piece() const;
// Piece specific // Piece specific
bool pawn_is_passed(Color c, Square s) const; bool pawn_passed(Color c, Square s) const;
bool pawn_on_7th(Color c) const;
bool opposite_bishops() const; bool opposite_bishops() const;
bool bishop_pair(Color c) const;
// Doing and undoing moves // Doing and undoing moves
void do_move(Move m, StateInfo& st); void do_move(Move m, StateInfo& newSt);
void do_move(Move m, StateInfo& st, const CheckInfo& ci, bool moveIsCheck); void do_move(Move m, StateInfo& newSt, bool givesCheck);
void undo_move(Move m); void undo_move(Move m);
void do_null_move(StateInfo& st); void do_null_move(StateInfo& newSt);
void undo_null_move(); void undo_null_move();
// Static exchange evaluation // Static Exchange Evaluation
int see(Move m, int asymmThreshold = 0) const; bool see_ge(Move m, Value threshold = VALUE_ZERO) const;
int see_sign(Move m) const;
// Accessing hash keys // Accessing hash keys
Key key() const; Key key() const;
Key exclusion_key() const; Key key_after(Move m) const;
Key pawn_key() const;
Key material_key() const; Key material_key() const;
Key pawn_key() const;
// Incremental piece-square evaluation
Score psq_score() const;
Value non_pawn_material(Color c) const;
// Other properties of the position // Other properties of the position
Color side_to_move() const; Color side_to_move() const;
int game_ply() const; int game_ply() const;
bool is_chess960() const; bool is_chess960() const;
Thread* this_thread() const; Thread* this_thread() const;
int64_t nodes_searched() const; bool is_draw(int ply) const;
void set_nodes_searched(int64_t n); bool has_game_cycle(int ply) const;
bool is_draw() const; bool has_repeated() const;
int rule50_count() const;
Score psq_score() const;
Value non_pawn_material(Color c) const;
Value non_pawn_material() const;
// Position consistency check, for debugging // Position consistency check, for debugging
bool pos_is_ok(int* failedStep = NULL) const; bool pos_is_ok() const;
void flip(); void flip();
private: private:
// Initialization helpers (used while setting up a position) // Initialization helpers (used while setting up a position)
void clear(); void set_castling_right(Color c, Square rfrom);
void set_castle_right(Color c, Square rfrom); void set_state(StateInfo* si) const;
void set_check_info(StateInfo* si) const;
// Helper functions // Other helpers
void do_castle(Square kfrom, Square kto, Square rfrom, Square rto); void put_piece(Piece pc, Square s);
Bitboard hidden_checkers(Square ksq, Color c) const; void remove_piece(Piece pc, Square s);
void put_piece(Square s, Color c, PieceType pt); void move_piece(Piece pc, Square from, Square to);
void remove_piece(Square s, Color c, PieceType pt); template<bool Do>
void move_piece(Square from, Square to, Color c, PieceType pt); void do_castling(Color us, Square from, Square& to, Square& rfrom, Square& rto);
// Computing hash keys from scratch (for initialization and debugging) // Data members
Key compute_key() const;
Key compute_pawn_key() const;
Key compute_material_key() const;
// Computing incremental evaluation scores and material counts
Score compute_psq_score() const;
Value compute_non_pawn_material(Color c) const;
// Board and pieces
Piece board[SQUARE_NB]; Piece board[SQUARE_NB];
Bitboard byTypeBB[PIECE_TYPE_NB]; Bitboard byTypeBB[PIECE_TYPE_NB];
Bitboard byColorBB[COLOR_NB]; Bitboard byColorBB[COLOR_NB];
int pieceCount[COLOR_NB][PIECE_TYPE_NB]; int pieceCount[PIECE_NB];
Square pieceList[COLOR_NB][PIECE_TYPE_NB][16]; Square pieceList[PIECE_NB][16];
int index[SQUARE_NB]; int index[SQUARE_NB];
int castlingRightsMask[SQUARE_NB];
// Other info Square castlingRookSquare[CASTLING_RIGHT_NB];
int castleRightsMask[SQUARE_NB]; Bitboard castlingPath[CASTLING_RIGHT_NB];
Square castleRookSquare[COLOR_NB][CASTLING_SIDE_NB];
Bitboard castlePath[COLOR_NB][CASTLING_SIDE_NB];
StateInfo startState;
int64_t nodes;
int gamePly; int gamePly;
Color sideToMove; Color sideToMove;
Score psq;
Thread* thisThread; Thread* thisThread;
StateInfo* st; StateInfo* st;
int chess960; bool chess960;
}; };
inline int64_t Position::nodes_searched() const { namespace PSQT {
return nodes; extern Score psq[PIECE_NB][SQUARE_NB];
} }
inline void Position::set_nodes_searched(int64_t n) { extern std::ostream& operator<<(std::ostream& os, const Position& pos);
nodes = n;
inline Color Position::side_to_move() const {
return sideToMove;
}
inline bool Position::empty(Square s) const {
return board[s] == NO_PIECE;
} }
inline Piece Position::piece_on(Square s) const { inline Piece Position::piece_on(Square s) const {
return board[s]; return board[s];
} }
inline Piece Position::piece_moved(Move m) const { inline Piece Position::moved_piece(Move m) const {
return board[from_sq(m)]; return board[from_sq(m)];
} }
inline bool Position::is_empty(Square s) const {
return board[s] == NO_PIECE;
}
inline Color Position::side_to_move() const {
return sideToMove;
}
inline Bitboard Position::pieces() const { inline Bitboard Position::pieces() const {
return byTypeBB[ALL_PIECES]; return byTypeBB[ALL_PIECES];
} }
@@ -276,52 +240,57 @@ inline Bitboard Position::pieces(Color c, PieceType pt1, PieceType pt2) const {
} }
template<PieceType Pt> inline int Position::count(Color c) const { template<PieceType Pt> inline int Position::count(Color c) const {
return pieceCount[c][Pt]; return pieceCount[make_piece(c, Pt)];
} }
template<PieceType Pt> inline const Square* Position::list(Color c) const { template<PieceType Pt> inline int Position::count() const {
return pieceList[c][Pt]; return pieceCount[make_piece(WHITE, Pt)] + pieceCount[make_piece(BLACK, Pt)];
}
template<PieceType Pt> inline const Square* Position::squares(Color c) const {
return pieceList[make_piece(c, Pt)];
}
template<PieceType Pt> inline Square Position::square(Color c) const {
assert(pieceCount[make_piece(c, Pt)] == 1);
return pieceList[make_piece(c, Pt)][0];
} }
inline Square Position::ep_square() const { inline Square Position::ep_square() const {
return st->epSquare; return st->epSquare;
} }
inline Square Position::king_square(Color c) const { inline int Position::can_castle(CastlingRight cr) const {
return pieceList[c][KING][0]; return st->castlingRights & cr;
}
inline int Position::can_castle(CastleRight f) const {
return st->castleRights & f;
} }
inline int Position::can_castle(Color c) const { inline int Position::can_castle(Color c) const {
return st->castleRights & ((WHITE_OO | WHITE_OOO) << (2 * c)); return st->castlingRights & ((WHITE_OO | WHITE_OOO) << (2 * c));
} }
inline bool Position::castle_impeded(Color c, CastlingSide s) const { inline bool Position::castling_impeded(CastlingRight cr) const {
return byTypeBB[ALL_PIECES] & castlePath[c][s]; return byTypeBB[ALL_PIECES] & castlingPath[cr];
} }
inline Square Position::castle_rook_square(Color c, CastlingSide s) const { inline Square Position::castling_rook_square(CastlingRight cr) const {
return castleRookSquare[c][s]; return castlingRookSquare[cr];
} }
template<PieceType Pt> template<PieceType Pt>
inline Bitboard Position::attacks_from(Square s) const { inline Bitboard Position::attacks_from(Square s) const {
assert(Pt != PAWN);
return Pt == BISHOP || Pt == ROOK ? attacks_bb<Pt>(s, pieces()) return Pt == BISHOP || Pt == ROOK ? attacks_bb<Pt>(s, byTypeBB[ALL_PIECES])
: Pt == QUEEN ? attacks_from<ROOK>(s) | attacks_from<BISHOP>(s) : Pt == QUEEN ? attacks_from<ROOK>(s) | attacks_from<BISHOP>(s)
: StepAttacksBB[Pt][s]; : PseudoAttacks[Pt][s];
} }
template<> template<>
inline Bitboard Position::attacks_from<PAWN>(Square s, Color c) const { inline Bitboard Position::attacks_from<PAWN>(Square s, Color c) const {
return StepAttacksBB[make_piece(c, PAWN)][s]; return PawnAttacks[c][s];
} }
inline Bitboard Position::attacks_from(Piece p, Square s) const { inline Bitboard Position::attacks_from(PieceType pt, Square s) const {
return attacks_from(p, s, byTypeBB[ALL_PIECES]); return attacks_bb(pt, s, byTypeBB[ALL_PIECES]);
} }
inline Bitboard Position::attackers_to(Square s) const { inline Bitboard Position::attackers_to(Square s) const {
@@ -332,18 +301,23 @@ inline Bitboard Position::checkers() const {
return st->checkersBB; return st->checkersBB;
} }
inline Bitboard Position::discovered_check_candidates() const { inline Bitboard Position::blockers_for_king(Color c) const {
return hidden_checkers(king_square(~sideToMove), sideToMove); return st->blockersForKing[c];
} }
inline Bitboard Position::pinned_pieces() const { inline Bitboard Position::check_squares(PieceType pt) const {
return hidden_checkers(king_square(sideToMove), ~sideToMove); return st->checkSquares[pt];
} }
inline bool Position::pawn_is_passed(Color c, Square s) const { inline bool Position::pawn_passed(Color c, Square s) const {
return !(pieces(~c, PAWN) & passed_pawn_mask(c, s)); return !(pieces(~c, PAWN) & passed_pawn_mask(c, s));
} }
inline bool Position::advanced_pawn_push(Move m) const {
return type_of(moved_piece(m)) == PAWN
&& relative_rank(sideToMove, from_sq(m)) > RANK_4;
}
inline Key Position::key() const { inline Key Position::key() const {
return st->key; return st->key;
} }
@@ -357,103 +331,101 @@ inline Key Position::material_key() const {
} }
inline Score Position::psq_score() const { inline Score Position::psq_score() const {
return st->psq; return psq;
} }
inline Value Position::non_pawn_material(Color c) const { inline Value Position::non_pawn_material(Color c) const {
return st->npMaterial[c]; return st->nonPawnMaterial[c];
} }
inline bool Position::is_passed_pawn_push(Move m) const { inline Value Position::non_pawn_material() const {
return st->nonPawnMaterial[WHITE] + st->nonPawnMaterial[BLACK];
return type_of(piece_moved(m)) == PAWN
&& pawn_is_passed(sideToMove, to_sq(m));
} }
inline int Position::game_ply() const { inline int Position::game_ply() const {
return gamePly; return gamePly;
} }
inline int Position::rule50_count() const {
return st->rule50;
}
inline bool Position::opposite_bishops() const { inline bool Position::opposite_bishops() const {
return pieceCount[W_BISHOP] == 1
return pieceCount[WHITE][BISHOP] == 1 && pieceCount[B_BISHOP] == 1
&& pieceCount[BLACK][BISHOP] == 1 && opposite_colors(square<BISHOP>(WHITE), square<BISHOP>(BLACK));
&& opposite_colors(pieceList[WHITE][BISHOP][0], pieceList[BLACK][BISHOP][0]);
}
inline bool Position::bishop_pair(Color c) const {
return pieceCount[c][BISHOP] >= 2
&& opposite_colors(pieceList[c][BISHOP][0], pieceList[c][BISHOP][1]);
}
inline bool Position::pawn_on_7th(Color c) const {
return pieces(c, PAWN) & rank_bb(relative_rank(c, RANK_7));
} }
inline bool Position::is_chess960() const { inline bool Position::is_chess960() const {
return chess960; return chess960;
} }
inline bool Position::is_capture_or_promotion(Move m) const { inline bool Position::capture_or_promotion(Move m) const {
assert(is_ok(m)); assert(is_ok(m));
return type_of(m) ? type_of(m) != CASTLE : !is_empty(to_sq(m)); return type_of(m) != NORMAL ? type_of(m) != CASTLING : !empty(to_sq(m));
} }
inline bool Position::is_capture(Move m) const { inline bool Position::capture(Move m) const {
// Note that castle is coded as "king captures the rook"
assert(is_ok(m)); assert(is_ok(m));
return (!is_empty(to_sq(m)) && type_of(m) != CASTLE) || type_of(m) == ENPASSANT; // Castling is encoded as "king captures rook"
return (!empty(to_sq(m)) && type_of(m) != CASTLING) || type_of(m) == ENPASSANT;
} }
inline PieceType Position::captured_piece_type() const { inline Piece Position::captured_piece() const {
return st->capturedType; return st->capturedPiece;
} }
inline Thread* Position::this_thread() const { inline Thread* Position::this_thread() const {
return thisThread; return thisThread;
} }
inline void Position::put_piece(Square s, Color c, PieceType pt) { inline void Position::put_piece(Piece pc, Square s) {
board[s] = make_piece(c, pt); board[s] = pc;
byTypeBB[ALL_PIECES] |= s; byTypeBB[ALL_PIECES] |= s;
byTypeBB[pt] |= s; byTypeBB[type_of(pc)] |= s;
byColorBB[c] |= s; byColorBB[color_of(pc)] |= s;
index[s] = pieceCount[c][pt]++; index[s] = pieceCount[pc]++;
pieceList[c][pt][index[s]] = s; pieceList[pc][index[s]] = s;
pieceCount[make_piece(color_of(pc), ALL_PIECES)]++;
psq += PSQT::psq[pc][s];
} }
inline void Position::move_piece(Square from, Square to, Color c, PieceType pt) { inline void Position::remove_piece(Piece pc, Square s) {
// index[from] is not updated and becomes stale. This works as long
// as index[] is accessed just by known occupied squares.
Bitboard from_to_bb = SquareBB[from] ^ SquareBB[to];
byTypeBB[ALL_PIECES] ^= from_to_bb;
byTypeBB[pt] ^= from_to_bb;
byColorBB[c] ^= from_to_bb;
board[from] = NO_PIECE;
board[to] = make_piece(c, pt);
index[to] = index[from];
pieceList[c][pt][index[to]] = to;
}
inline void Position::remove_piece(Square s, Color c, PieceType pt) {
// WARNING: This is not a reversible operation. If we remove a piece in // WARNING: This is not a reversible operation. If we remove a piece in
// do_move() and then replace it in undo_move() we will put it at the end of // do_move() and then replace it in undo_move() we will put it at the end of
// the list and not in its original place, it means index[] and pieceList[] // the list and not in its original place, it means index[] and pieceList[]
// are not guaranteed to be invariant to a do_move() + undo_move() sequence. // are not invariant to a do_move() + undo_move() sequence.
byTypeBB[ALL_PIECES] ^= s; byTypeBB[ALL_PIECES] ^= s;
byTypeBB[pt] ^= s; byTypeBB[type_of(pc)] ^= s;
byColorBB[c] ^= s; byColorBB[color_of(pc)] ^= s;
/* board[s] = NO_PIECE; */ // Not needed, will be overwritten by capturing /* board[s] = NO_PIECE; Not needed, overwritten by the capturing one */
Square lastSquare = pieceList[c][pt][--pieceCount[c][pt]]; Square lastSquare = pieceList[pc][--pieceCount[pc]];
index[lastSquare] = index[s]; index[lastSquare] = index[s];
pieceList[c][pt][index[lastSquare]] = lastSquare; pieceList[pc][index[lastSquare]] = lastSquare;
pieceList[c][pt][pieceCount[c][pt]] = SQ_NONE; pieceList[pc][pieceCount[pc]] = SQ_NONE;
pieceCount[make_piece(color_of(pc), ALL_PIECES)]--;
psq -= PSQT::psq[pc][s];
}
inline void Position::move_piece(Piece pc, Square from, Square to) {
// index[from] is not updated and becomes stale. This works as long as index[]
// is accessed just by known occupied squares.
Bitboard fromTo = SquareBB[from] ^ SquareBB[to];
byTypeBB[ALL_PIECES] ^= fromTo;
byTypeBB[type_of(pc)] ^= fromTo;
byColorBB[color_of(pc)] ^= fromTo;
board[from] = NO_PIECE;
board[to] = pc;
index[to] = index[from];
pieceList[pc][index[to]] = to;
psq += PSQT::psq[pc][to] - PSQT::psq[pc][from];
}
inline void Position::do_move(Move m, StateInfo& newSt) {
do_move(m, newSt, gives_check(m));
} }
#endif // #ifndef POSITION_H_INCLUDED #endif // #ifndef POSITION_H_INCLUDED
+126
View File
@@ -0,0 +1,126 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include <algorithm>
#include "types.h"
Value PieceValue[PHASE_NB][PIECE_NB] = {
{ VALUE_ZERO, PawnValueMg, KnightValueMg, BishopValueMg, RookValueMg, QueenValueMg },
{ VALUE_ZERO, PawnValueEg, KnightValueEg, BishopValueEg, RookValueEg, QueenValueEg }
};
namespace PSQT {
#define S(mg, eg) make_score(mg, eg)
// Bonus[PieceType][Square / 2] contains Piece-Square scores. For each piece
// type on a given square a (middlegame, endgame) score pair is assigned. Table
// is defined for files A..D and white side: it is symmetric for black side and
// second half of the files.
constexpr Score Bonus[][RANK_NB][int(FILE_NB) / 2] = {
{ },
{ // Pawn
{ S( 0, 0), S( 0, 0), S( 0, 0), S( 0, 0) },
{ S(-11,-3), S( 7, -1), S( 7, 7), S(17, 2) },
{ S(-16,-2), S( -3, 2), S( 23, 6), S(23,-1) },
{ S(-14, 7), S( -7, -4), S( 20,-8), S(24, 2) },
{ S( -5,13), S( -2, 10), S( -1,-1), S(12,-8) },
{ S(-11,16), S(-12, 6), S( -2, 1), S( 4,16) },
{ S( -2, 1), S( 20,-12), S(-10, 6), S(-2,25) }
},
{ // Knight
{ S(-169,-105), S(-96,-74), S(-80,-46), S(-79,-18) },
{ S( -79, -70), S(-39,-56), S(-24,-15), S( -9, 6) },
{ S( -64, -38), S(-20,-33), S( 4, -5), S( 19, 27) },
{ S( -28, -36), S( 5, 0), S( 41, 13), S( 47, 34) },
{ S( -29, -41), S( 13,-20), S( 42, 4), S( 52, 35) },
{ S( -11, -51), S( 28,-38), S( 63,-17), S( 55, 19) },
{ S( -67, -64), S(-21,-45), S( 6,-37), S( 37, 16) },
{ S(-200, -98), S(-80,-89), S(-53,-53), S(-32,-16) }
},
{ // Bishop
{ S(-49,-58), S(- 7,-31), S(-10,-37), S(-34,-19) },
{ S(-24,-34), S( 9, -9), S( 15,-14), S( 1, 4) },
{ S( -9,-23), S( 22, 0), S( -3, -3), S( 12, 16) },
{ S( 4,-26), S( 9, -3), S( 18, -5), S( 40, 16) },
{ S( -8,-26), S( 27, -4), S( 13, -7), S( 30, 14) },
{ S(-17,-24), S( 14, -2), S( -6, 0), S( 6, 13) },
{ S(-19,-34), S(-13,-10), S( 7,-12), S(-11, 6) },
{ S(-47,-55), S( -7,-32), S(-17,-36), S(-29,-17) }
},
{ // Rook
{ S(-24, 0), S(-15, 3), S( -8, 0), S( 0, 3) },
{ S(-18,-7), S( -5,-5), S( -1,-5), S( 1,-1) },
{ S(-19, 6), S(-10,-7), S( 1, 3), S( 0, 3) },
{ S(-21, 0), S( -7, 4), S( -4,-2), S(-4, 1) },
{ S(-21,-7), S(-12, 5), S( -1,-5), S( 4,-7) },
{ S(-23, 3), S(-10, 2), S( 1,-1), S( 6, 3) },
{ S(-11,-1), S( 8, 7), S( 9,11), S(12,-1) },
{ S(-25, 6), S(-18, 4), S(-11, 6), S( 2, 2) }
},
{ // Queen
{ S( 3,-69), S(-5,-57), S(-5,-47), S( 4,-26) },
{ S( -3,-55), S( 5,-31), S( 8,-22), S(12, -4) },
{ S( -3,-39), S( 6,-18), S(13, -9), S( 7, 3) },
{ S( 4,-23), S( 5, -3), S( 9, 13), S( 8, 24) },
{ S( 0,-29), S(14, -6), S(12, 9), S( 5, 21) },
{ S( -4,-38), S(10,-18), S( 6,-12), S( 8, 1) },
{ S( -5,-50), S( 6,-27), S(10,-24), S( 8, -8) },
{ S( -2,-75), S(-2,-52), S( 1,-43), S(-2,-36) }
},
{ // King
{ S(272, 0), S(325, 41), S(273, 80), S(190, 93) },
{ S(277, 57), S(305, 98), S(241,138), S(183,131) },
{ S(198, 86), S(253,138), S(168,165), S(120,173) },
{ S(169,103), S(191,152), S(136,168), S(108,169) },
{ S(145, 98), S(176,166), S(112,197), S(69, 194) },
{ S(122, 87), S(159,164), S(85, 174), S(36, 189) },
{ S(87, 40), S(120, 99), S(64, 128), S(25, 141) },
{ S(64, 5), S(87, 60), S(49, 75), S(0, 75) }
}
};
#undef S
Score psq[PIECE_NB][SQUARE_NB];
// init() initializes piece-square tables: the white halves of the tables are
// copied from Bonus[] adding the piece value, then the black halves of the
// tables are initialized by flipping and changing the sign of the white scores.
void init() {
for (Piece pc = W_PAWN; pc <= W_KING; ++pc)
{
PieceValue[MG][~pc] = PieceValue[MG][pc];
PieceValue[EG][~pc] = PieceValue[EG][pc];
Score score = make_score(PieceValue[MG][pc], PieceValue[EG][pc]);
for (Square s = SQ_A1; s <= SQ_H8; ++s)
{
File f = std::min(file_of(s), ~file_of(s));
psq[ pc][ s] = score + Bonus[pc][rank_of(s)][f];
psq[~pc][~s] = -psq[pc][s];
}
}
}
} // namespace PSQT
-98
View File
@@ -1,98 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef PSQTAB_H_INCLUDED
#define PSQTAB_H_INCLUDED
#include "types.h"
#define S(mg, eg) make_score(mg, eg)
/// PSQT[PieceType][Square] contains Piece-Square scores. For each piece type on
/// a given square a (midgame, endgame) score pair is assigned. PSQT is defined
/// for white side, for black side the tables are symmetric.
static const Score PSQT[][SQUARE_NB] = {
{ },
{ // Pawn
S( 0, 0), S( 0, 0), S( 0, 0), S( 0, 0), S(0, 0), S( 0, 0), S( 0, 0), S( 0, 0),
S(-20,-8), S(-6,-8), S( 4,-8), S(14,-8), S(14,-8), S( 4,-8), S(-6,-8), S(-20,-8),
S(-20,-8), S(-6,-8), S( 9,-8), S(34,-8), S(34,-8), S( 9,-8), S(-6,-8), S(-20,-8),
S(-20,-8), S(-6,-8), S(17,-8), S(54,-8), S(54,-8), S(17,-8), S(-6,-8), S(-20,-8),
S(-20,-8), S(-6,-8), S(17,-8), S(34,-8), S(34,-8), S(17,-8), S(-6,-8), S(-20,-8),
S(-20,-8), S(-6,-8), S( 9,-8), S(14,-8), S(14,-8), S( 9,-8), S(-6,-8), S(-20,-8),
S(-20,-8), S(-6,-8), S( 4,-8), S(14,-8), S(14,-8), S( 4,-8), S(-6,-8), S(-20,-8),
S( 0, 0), S( 0, 0), S( 0, 0), S( 0, 0), S(0, 0), S( 0, 0), S( 0, 0), S( 0, 0)
},
{ // Knight
S(-135,-104), S(-107,-79), S(-80,-55), S(-67,-42), S(-67,-42), S(-80,-55), S(-107,-79), S(-135,-104),
S( -93, -79), S( -67,-55), S(-39,-30), S(-25,-17), S(-25,-17), S(-39,-30), S( -67,-55), S( -93, -79),
S( -53, -55), S( -25,-30), S( 1, -6), S( 13, 5), S( 13, 5), S( 1, -6), S( -25,-30), S( -53, -55),
S( -25, -42), S( 1,-17), S( 27, 5), S( 41, 18), S( 41, 18), S( 27, 5), S( 1,-17), S( -25, -42),
S( -11, -42), S( 13,-17), S( 41, 5), S( 55, 18), S( 55, 18), S( 41, 5), S( 13,-17), S( -11, -42),
S( -11, -55), S( 13,-30), S( 41, -6), S( 55, 5), S( 55, 5), S( 41, -6), S( 13,-30), S( -11, -55),
S( -53, -79), S( -25,-55), S( 1,-30), S( 13,-17), S( 13,-17), S( 1,-30), S( -25,-55), S( -53, -79),
S(-193,-104), S( -67,-79), S(-39,-55), S(-25,-42), S(-25,-42), S(-39,-55), S( -67,-79), S(-193,-104)
},
{ // Bishop
S(-40,-59), S(-40,-42), S(-35,-35), S(-30,-26), S(-30,-26), S(-35,-35), S(-40,-42), S(-40,-59),
S(-17,-42), S( 0,-26), S( -4,-18), S( 0,-11), S( 0,-11), S( -4,-18), S( 0,-26), S(-17,-42),
S(-13,-35), S( -4,-18), S( 8,-11), S( 4, -4), S( 4, -4), S( 8,-11), S( -4,-18), S(-13,-35),
S( -8,-26), S( 0,-11), S( 4, -4), S( 17, 4), S( 17, 4), S( 4, -4), S( 0,-11), S( -8,-26),
S( -8,-26), S( 0,-11), S( 4, -4), S( 17, 4), S( 17, 4), S( 4, -4), S( 0,-11), S( -8,-26),
S(-13,-35), S( -4,-18), S( 8,-11), S( 4, -4), S( 4, -4), S( 8,-11), S( -4,-18), S(-13,-35),
S(-17,-42), S( 0,-26), S( -4,-18), S( 0,-11), S( 0,-11), S( -4,-18), S( 0,-26), S(-17,-42),
S(-17,-59), S(-17,-42), S(-13,-35), S( -8,-26), S( -8,-26), S(-13,-35), S(-17,-42), S(-17,-59)
},
{ // Rook
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3),
S(-12, 3), S(-7, 3), S(-2, 3), S(2, 3), S(2, 3), S(-2, 3), S(-7, 3), S(-12, 3)
},
{ // Queen
S(8,-80), S(8,-54), S(8,-42), S(8,-30), S(8,-30), S(8,-42), S(8,-54), S(8,-80),
S(8,-54), S(8,-30), S(8,-18), S(8, -6), S(8, -6), S(8,-18), S(8,-30), S(8,-54),
S(8,-42), S(8,-18), S(8, -6), S(8, 6), S(8, 6), S(8, -6), S(8,-18), S(8,-42),
S(8,-30), S(8, -6), S(8, 6), S(8, 18), S(8, 18), S(8, 6), S(8, -6), S(8,-30),
S(8,-30), S(8, -6), S(8, 6), S(8, 18), S(8, 18), S(8, 6), S(8, -6), S(8,-30),
S(8,-42), S(8,-18), S(8, -6), S(8, 6), S(8, 6), S(8, -6), S(8,-18), S(8,-42),
S(8,-54), S(8,-30), S(8,-18), S(8, -6), S(8, -6), S(8,-18), S(8,-30), S(8,-54),
S(8,-80), S(8,-54), S(8,-42), S(8,-30), S(8,-30), S(8,-42), S(8,-54), S(8,-80)
},
{ // King
S(287, 18), S(311, 77), S(262,105), S(214,135), S(214,135), S(262,105), S(311, 77), S(287, 18),
S(262, 77), S(287,135), S(238,165), S(190,193), S(190,193), S(238,165), S(287,135), S(262, 77),
S(214,105), S(238,165), S(190,193), S(142,222), S(142,222), S(190,193), S(238,165), S(214,105),
S(190,135), S(214,193), S(167,222), S(119,251), S(119,251), S(167,222), S(214,193), S(190,135),
S(167,135), S(190,193), S(142,222), S( 94,251), S( 94,251), S(142,222), S(190,193), S(167,135),
S(142,105), S(167,165), S(119,193), S( 69,222), S( 69,222), S(119,193), S(167,165), S(142,105),
S(119, 77), S(142,135), S( 94,165), S( 46,193), S( 46,193), S( 94,165), S(142,135), S(119, 77),
S(94, 18), S(119, 77), S( 69,105), S( 21,135), S( 21,135), S( 69,105), S(119, 77), S( 94, 18)
}
};
#undef S
#endif // #ifndef PSQTAB_H_INCLUDED
-74
View File
@@ -1,74 +0,0 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
This file is based on original code by Heinz van Saanen and is
available under the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
*/
#ifndef RKISS_H_INCLUDED
#define RKISS_H_INCLUDED
#include "types.h"
/// RKISS is our pseudo random number generator (PRNG) used to compute hash keys.
/// George Marsaglia invented the RNG-Kiss-family in the early 90's. This is a
/// specific version that Heinz van Saanen derived from some public domain code
/// by Bob Jenkins. Following the feature list, as tested by Heinz.
///
/// - Quite platform independent
/// - Passes ALL dieharder tests! Here *nix sys-rand() e.g. fails miserably:-)
/// - ~12 times faster than my *nix sys-rand()
/// - ~4 times faster than SSE2-version of Mersenne twister
/// - Average cycle length: ~2^126
/// - 64 bit seed
/// - Return doubles with a full 53 bit mantissa
/// - Thread safe
class RKISS {
struct S { uint64_t a, b, c, d; } s; // Keep variables always together
uint64_t rotate(uint64_t x, uint64_t k) const {
return (x << k) | (x >> (64 - k));
}
uint64_t rand64() {
const uint64_t
e = s.a - rotate(s.b, 7);
s.a = s.b ^ rotate(s.c, 13);
s.b = s.c + rotate(s.d, 37);
s.c = s.d + e;
return s.d = e + s.a;
}
public:
RKISS(int seed = 73) {
s.a = 0xf1ea5eed;
s.b = s.c = s.d = 0xd4e12c77;
for (int i = 0; i < seed; i++) // Scramble a few rounds
rand64();
}
template<typename T> T rand() { return T(rand64()); }
};
#endif // #ifndef RKISS_H_INCLUDED
+1166 -1253
View File
File diff suppressed because it is too large Load Diff
+49 -53
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,92 +21,87 @@
#ifndef SEARCH_H_INCLUDED #ifndef SEARCH_H_INCLUDED
#define SEARCH_H_INCLUDED #define SEARCH_H_INCLUDED
#include <cstring>
#include <memory>
#include <stack>
#include <vector> #include <vector>
#include "misc.h" #include "misc.h"
#include "position.h" #include "movepick.h"
#include "types.h" #include "types.h"
struct SplitPoint; class Position;
namespace Search { namespace Search {
/// The Stack struct keeps track of the information we need to remember from /// Threshold used for countermoves based pruning
/// nodes shallower and deeper in the tree during the search. Each search thread constexpr int CounterMovePruneThreshold = 0;
/// has its own array of Stack objects, indexed by the current ply.
/// Stack struct keeps track of the information we need to remember from nodes
/// shallower and deeper in the tree during the search. Each search thread has
/// its own array of Stack objects, indexed by the current ply.
struct Stack { struct Stack {
SplitPoint* splitPoint; Move* pv;
PieceToHistory* continuationHistory;
int ply; int ply;
Move currentMove; Move currentMove;
Move excludedMove; Move excludedMove;
Move killers[2]; Move killers[2];
Depth reduction;
Value staticEval; Value staticEval;
Value evalMargin; int statScore;
int skipNullMove; int moveCount;
int futilityMoveCount;
}; };
/// RootMove struct is used for moves at the root of the tree. For each root /// RootMove struct is used for moves at the root of the tree. For each root move
/// move we store a score, a node count, and a PV (really a refutation in the /// we store a score and a PV (really a refutation in the case of moves which
/// case of moves which fail low). Score is normally set at -VALUE_INFINITE for /// fail low). Score is normally set at -VALUE_INFINITE for all non-pv moves.
/// all non-pv moves.
struct RootMove { struct RootMove {
RootMove(Move m) : score(-VALUE_INFINITE), prevScore(-VALUE_INFINITE) { explicit RootMove(Move m) : pv(1, m) {}
pv.push_back(m); pv.push_back(MOVE_NONE); bool extract_ponder_from_tt(Position& pos);
bool operator==(const Move& m) const { return pv[0] == m; }
bool operator<(const RootMove& m) const { // Sort in descending order
return m.score != score ? m.score < score
: m.previousScore < previousScore;
} }
bool operator<(const RootMove& m) const { return score > m.score; } // Ascending sort Value score = -VALUE_INFINITE;
bool operator==(const Move& m) const { return pv[0] == m; } Value previousScore = -VALUE_INFINITE;
int selDepth = 0;
void extract_pv_from_tt(Position& pos); int tbRank;
void insert_pv_in_tt(Position& pos); Value tbScore;
Value score;
Value prevScore;
std::vector<Move> pv; std::vector<Move> pv;
}; };
typedef std::vector<RootMove> RootMoves;
/// The LimitsType struct stores information sent by GUI about available time
/// to search the current move, maximum depth/time, if we are in analysis mode /// LimitsType struct stores information sent by GUI about available time to
/// or if we have to ponder while is our opponent's side to move. /// search the current move, maximum depth/time, or if we are in analysis mode.
struct LimitsType { struct LimitsType {
LimitsType() { std::memset(this, 0, sizeof(LimitsType)); } LimitsType() { // Init explicitly due to broken value-initialization of non POD in MSVC
bool use_time_management() const { return !(mate | movetime | depth | nodes | infinite); } time[WHITE] = time[BLACK] = inc[WHITE] = inc[BLACK] = npmsec = movetime = TimePoint(0);
movestogo = depth = mate = perft = infinite = 0;
nodes = 0;
}
int time[COLOR_NB], inc[COLOR_NB], movestogo, depth, nodes, movetime, mate, infinite, ponder; bool use_time_management() const {
return !(mate | movetime | depth | nodes | perft | infinite);
}
std::vector<Move> searchmoves;
TimePoint time[COLOR_NB], inc[COLOR_NB], npmsec, movetime, startTime;
int movestogo, depth, mate, perft, infinite;
int64_t nodes;
}; };
/// The SignalsType struct stores volatile flags updated during the search
/// typically in an async fashion, for instance to stop the search by the GUI.
struct SignalsType {
bool stopOnPonderhit, firstRootMove, stop, failedLowAtRoot;
};
typedef std::auto_ptr<std::stack<StateInfo> > StateStackPtr;
extern volatile SignalsType Signals;
extern LimitsType Limits; extern LimitsType Limits;
extern std::vector<RootMove> RootMoves;
extern Position RootPos;
extern Color RootColor;
extern Time::point SearchTime;
extern StateStackPtr SetupStates;
extern void init(); void init();
extern size_t perft(Position& pos, Depth depth); void clear();
extern void think();
} // namespace Search } // namespace Search
File diff suppressed because it is too large Load Diff
+79
View File
@@ -0,0 +1,79 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (c) 2013 Ronald de Man
Copyright (C) 2016-2018 Marco Costalba, Lucas Braesch
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef TBPROBE_H
#define TBPROBE_H
#include <ostream>
#include "../search.h"
namespace Tablebases {
enum WDLScore {
WDLLoss = -2, // Loss
WDLBlessedLoss = -1, // Loss, but draw under 50-move rule
WDLDraw = 0, // Draw
WDLCursedWin = 1, // Win, but draw under 50-move rule
WDLWin = 2, // Win
WDLScoreNone = -1000
};
// Possible states after a probing operation
enum ProbeState {
FAIL = 0, // Probe failed (missing file table)
OK = 1, // Probe succesful
CHANGE_STM = -1, // DTZ should check the other side
ZEROING_BEST_MOVE = 2 // Best move zeroes DTZ (capture or pawn move)
};
extern int MaxCardinality;
void init(const std::string& paths);
WDLScore probe_wdl(Position& pos, ProbeState* result);
int probe_dtz(Position& pos, ProbeState* result);
bool root_probe(Position& pos, Search::RootMoves& rootMoves);
bool root_probe_wdl(Position& pos, Search::RootMoves& rootMoves);
void rank_root_moves(Position& pos, Search::RootMoves& rootMoves);
inline std::ostream& operator<<(std::ostream& os, const WDLScore v) {
os << (v == WDLLoss ? "Loss" :
v == WDLBlessedLoss ? "Blessed loss" :
v == WDLDraw ? "Draw" :
v == WDLCursedWin ? "Cursed win" :
v == WDLWin ? "Win" : "None");
return os;
}
inline std::ostream& operator<<(std::ostream& os, const ProbeState v) {
os << (v == FAIL ? "Failed" :
v == OK ? "Success" :
v == CHANGE_STM ? "Probed opponent side" :
v == ZEROING_BEST_MOVE ? "Best move zeroes DTZ" : "None");
return os;
}
}
#endif
+129 -320
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -23,372 +24,180 @@
#include "movegen.h" #include "movegen.h"
#include "search.h" #include "search.h"
#include "thread.h" #include "thread.h"
#include "ucioption.h" #include "uci.h"
#include "syzygy/tbprobe.h"
using namespace Search; #include "tt.h"
ThreadPool Threads; // Global object ThreadPool Threads; // Global object
namespace {
// start_routine() is the C function which is called when a new thread /// Thread constructor launches the thread and waits until it goes to sleep
// is launched. It is a wrapper to the virtual function idle_loop(). /// in idle_loop(). Note that 'searching' and 'exit' should be alredy set.
extern "C" { long start_routine(ThreadBase* th) { th->idle_loop(); return 0; } } Thread::Thread(size_t n) : idx(n), stdThread(&Thread::idle_loop, this) {
// Helpers to launch a thread after creation and joining before delete. Must be
// outside Thread c'tor and d'tor because object shall be fully initialized
// when start_routine (and hence virtual idle_loop) is called and when joining.
template<typename T> T* new_thread() {
T* th = new T();
thread_create(th->handle, start_routine, th); // Will go to sleep
return th;
}
void delete_thread(ThreadBase* th) {
th->exit = true; // Search must be already finished
th->notify_one();
thread_join(th->handle); // Wait for thread termination
delete th;
}
wait_for_search_finished();
} }
// ThreadBase::notify_one() wakes up the thread when there is some work to do /// Thread destructor wakes up the thread in idle_loop() and waits
/// for its termination. Thread should be already waiting.
void ThreadBase::notify_one() { Thread::~Thread() {
mutex.lock(); assert(!searching);
sleepCondition.notify_one();
mutex.unlock(); exit = true;
start_searching();
stdThread.join();
} }
// ThreadBase::wait_for() set the thread to sleep until condition 'b' turns true /// Thread::clear() reset histories, usually before a new game
void ThreadBase::wait_for(volatile const bool& b) { void Thread::clear() {
mutex.lock(); counterMoves.fill(MOVE_NONE);
while (!b) sleepCondition.wait(mutex); mainHistory.fill(0);
mutex.unlock(); captureHistory.fill(0);
for (auto& to : continuationHistory)
for (auto& h : to)
h->fill(0);
continuationHistory[NO_PIECE][0]->fill(Search::CounterMovePruneThreshold - 1);
}
/// Thread::start_searching() wakes up the thread that will start the search
void Thread::start_searching() {
std::lock_guard<Mutex> lk(mutex);
searching = true;
cv.notify_one(); // Wake up the thread in idle_loop()
} }
// Thread c'tor just inits data but does not launch any thread of execution that /// Thread::wait_for_search_finished() blocks on the condition variable
// instead will be started only upon c'tor returns. /// until the thread has finished searching.
Thread::Thread() /* : splitPoints() */ { // Value-initialization bug in MSVC void Thread::wait_for_search_finished() {
searching = false; std::unique_lock<Mutex> lk(mutex);
maxPly = splitPointsSize = 0; cv.wait(lk, [&]{ return !searching; });
activeSplitPoint = NULL;
activePosition = NULL;
idx = Threads.size();
} }
// TimerThread::idle_loop() is where the timer thread waits msec milliseconds /// Thread::idle_loop() is where the thread is parked, blocked on the
// and then calls check_time(). If msec is 0 thread sleeps until is woken up. /// condition variable, when it has no work to do.
extern void check_time();
void TimerThread::idle_loop() { void Thread::idle_loop() {
while (!exit) // If OS already scheduled us on a different group than 0 then don't overwrite
{ // the choice, eventually we are one of many one-threaded processes running on
mutex.lock(); // some Windows NUMA hardware, for instance in fishtest. To make it simple,
// just check if running threads are below a threshold, in this case all this
if (!exit) // NUMA machinery is not needed.
sleepCondition.wait_for(mutex, msec ? msec : INT_MAX); if (Options["Threads"] > 8)
WinProcGroup::bindThisThread(idx);
mutex.unlock();
if (msec)
check_time();
}
}
// MainThread::idle_loop() is where the main thread is parked waiting to be started
// when there is a new search. Main thread will launch all the slave threads.
void MainThread::idle_loop() {
while (true) while (true)
{ {
mutex.lock(); std::unique_lock<Mutex> lk(mutex);
searching = false;
thinking = false; cv.notify_one(); // Wake up anyone waiting for search finished
cv.wait(lk, [&]{ return searching; });
while (!thinking && !exit)
{
Threads.sleepCondition.notify_one(); // Wake up UI thread if needed
sleepCondition.wait(mutex);
}
mutex.unlock();
if (exit) if (exit)
return; return;
searching = true; lk.unlock();
Search::think(); search();
assert(searching);
searching = false;
} }
} }
/// ThreadPool::set() creates/destroys threads to match the requested number.
/// Created and launched threads will go immediately to sleep in idle_loop.
/// Upon resizing, threads are recreated to allow for binding if necessary.
// Thread::cutoff_occurred() checks whether a beta cutoff has occurred in the void ThreadPool::set(size_t requested) {
// current active split point, or in some ancestor of the split point.
bool Thread::cutoff_occurred() const { if (size() > 0) { // destroy any existing thread(s)
main()->wait_for_search_finished();
for (SplitPoint* sp = activeSplitPoint; sp; sp = sp->parentSplitPoint) while (size() > 0)
if (sp->cutoff) delete back(), pop_back();
return true; }
return false; if (requested > 0) { // create new thread(s)
} push_back(new MainThread(0));
// Thread::is_available_to() checks whether the thread is available to help the
// thread 'master' at a split point. An obvious requirement is that thread must
// be idle. With more than two threads, this is not sufficient: If the thread is
// the master of some split point, it is only available as a slave to the slaves
// which are busy searching the split point at the top of slaves split point
// stack (the "helpful master concept" in YBWC terminology).
bool Thread::is_available_to(const Thread* master) const {
if (searching)
return false;
// Make a local copy to be sure doesn't become zero under our feet while
// testing next condition and so leading to an out of bound access.
int size = splitPointsSize;
// No split points means that the thread is available as a slave for any
// other thread otherwise apply the "helpful master" concept if possible.
return !size || (splitPoints[size - 1].slavesMask & (1ULL << master->idx));
}
// init() is called at startup to create and launch requested threads, that will
// go immediately to sleep due to 'sleepWhileIdle' set to true. We cannot use
// a c'tor becuase Threads is a static object and we need a fully initialized
// engine at this point due to allocation of Endgames in Thread c'tor.
void ThreadPool::init() {
sleepWhileIdle = true;
timer = new_thread<TimerThread>();
push_back(new_thread<MainThread>());
read_uci_options();
}
// exit() cleanly terminates the threads before the program exits
void ThreadPool::exit() {
delete_thread(timer); // As first because check_time() accesses threads data
for (iterator it = begin(); it != end(); ++it)
delete_thread(*it);
}
// read_uci_options() updates internal threads parameters from the corresponding
// UCI options and creates/destroys threads to match the requested number. Thread
// objects are dynamically allocated to avoid creating in advance all possible
// threads, with included pawns and material tables, if only few are used.
void ThreadPool::read_uci_options() {
maxThreadsPerSplitPoint = Options["Max Threads per Split Point"];
minimumSplitDepth = Options["Min Split Depth"] * ONE_PLY;
size_t requested = Options["Threads"];
assert(requested > 0);
// Value 0 has a special meaning: We determine the optimal minimum split depth
// automatically. Anyhow the minimumSplitDepth should never be under 4 plies.
if (!minimumSplitDepth)
minimumSplitDepth = (requested < 8 ? 4 : 7) * ONE_PLY;
else
minimumSplitDepth = std::max(4 * ONE_PLY, minimumSplitDepth);
while (size() < requested) while (size() < requested)
push_back(new_thread<Thread>()); push_back(new Thread(size()));
clear();
while (size() > requested)
{
delete_thread(back());
pop_back();
}
}
// slave_available() tries to find an idle thread which is available as a slave
// for the thread 'master'.
Thread* ThreadPool::available_slave(const Thread* master) const {
for (const_iterator it = begin(); it != end(); ++it)
if ((*it)->is_available_to(master))
return *it;
return NULL;
}
// split() does the actual work of distributing the work at a node between
// several available threads. If it does not succeed in splitting the node
// (because no idle threads are available), the function immediately returns.
// If splitting is possible, a SplitPoint object is initialized with all the
// data that must be copied to the helper threads and then helper threads are
// told that they have been assigned work. This will cause them to instantly
// leave their idle loops and call search(). When all threads have returned from
// search() then split() returns.
template <bool Fake>
void Thread::split(Position& pos, const Stack* ss, Value alpha, Value beta, Value* bestValue,
Move* bestMove, Depth depth, Move threatMove, int moveCount,
MovePicker* movePicker, int nodeType, bool cutNode) {
assert(pos.pos_is_ok());
assert(*bestValue <= alpha && alpha < beta && beta <= VALUE_INFINITE);
assert(*bestValue > -VALUE_INFINITE);
assert(depth >= Threads.minimumSplitDepth);
assert(searching);
assert(splitPointsSize < MAX_SPLITPOINTS_PER_THREAD);
// Pick the next available split point from the split point stack
SplitPoint& sp = splitPoints[splitPointsSize];
sp.masterThread = this;
sp.parentSplitPoint = activeSplitPoint;
sp.slavesMask = 1ULL << idx;
sp.depth = depth;
sp.bestValue = *bestValue;
sp.bestMove = *bestMove;
sp.threatMove = threatMove;
sp.alpha = alpha;
sp.beta = beta;
sp.nodeType = nodeType;
sp.cutNode = cutNode;
sp.movePicker = movePicker;
sp.moveCount = moveCount;
sp.pos = &pos;
sp.nodes = 0;
sp.cutoff = false;
sp.ss = ss;
// Try to allocate available threads and ask them to start searching setting
// 'searching' flag. This must be done under lock protection to avoid concurrent
// allocation of the same slave by another master.
Threads.mutex.lock();
sp.mutex.lock();
splitPointsSize++;
activeSplitPoint = &sp;
activePosition = NULL;
size_t slavesCnt = 1; // This thread is always included
Thread* slave;
while ( (slave = Threads.available_slave(this)) != NULL
&& ++slavesCnt <= Threads.maxThreadsPerSplitPoint && !Fake)
{
sp.slavesMask |= 1ULL << slave->idx;
slave->activeSplitPoint = &sp;
slave->searching = true; // Slave leaves idle_loop()
slave->notify_one(); // Could be sleeping
} }
// Everything is set up. The master thread enters the idle loop, from which // Reallocate the hash with the new threadpool size
// it will instantly launch a search, because its 'searching' flag is set. TT.resize(Options["Hash"]);
// The thread will return from the idle loop when all slaves have finished }
// their work at this split point.
if (slavesCnt > 1 || Fake) /// ThreadPool::clear() sets threadPool data to initial values.
void ThreadPool::clear() {
for (Thread* th : *this)
th->clear();
main()->callsCnt = 0;
main()->previousScore = VALUE_INFINITE;
main()->previousTimeReduction = 1.0;
}
/// ThreadPool::start_thinking() wakes up main thread waiting in idle_loop() and
/// returns immediately. Main thread will wake up other threads and start the search.
void ThreadPool::start_thinking(Position& pos, StateListPtr& states,
const Search::LimitsType& limits, bool ponderMode) {
main()->wait_for_search_finished();
stopOnPonderhit = stop = false;
ponder = ponderMode;
Search::Limits = limits;
Search::RootMoves rootMoves;
for (const auto& m : MoveList<LEGAL>(pos))
if ( limits.searchmoves.empty()
|| std::count(limits.searchmoves.begin(), limits.searchmoves.end(), m))
rootMoves.emplace_back(m);
if (!rootMoves.empty())
Tablebases::rank_root_moves(pos, rootMoves);
// After ownership transfer 'states' becomes empty, so if we stop the search
// and call 'go' again without setting a new position states.get() == NULL.
assert(states.get() || setupStates.get());
if (states.get())
setupStates = std::move(states); // Ownership transfer, states is now empty
// We use Position::set() to set root position across threads. But there are
// some StateInfo fields (previous, pliesFromNull, capturedPiece) that cannot
// be deduced from a fen string, so set() clears them and to not lose the info
// we need to backup and later restore setupStates->back(). Note that setupStates
// is shared by threads but is accessed in read-only mode.
StateInfo tmp = setupStates->back();
for (Thread* th : *this)
{ {
sp.mutex.unlock(); th->nodes = th->tbHits = th->nmpMinPly = 0;
Threads.mutex.unlock(); th->rootDepth = th->completedDepth = DEPTH_ZERO;
th->rootMoves = rootMoves;
Thread::idle_loop(); // Force a call to base class idle_loop() th->rootPos.set(pos.fen(), pos.is_chess960(), &setupStates->back(), th);
// In helpful master concept a master can help only a sub-tree of its split
// point, and because here is all finished is not possible master is booked.
assert(!searching);
assert(!activePosition);
// We have returned from the idle loop, which means that all threads are
// finished. Note that setting 'searching' and decreasing splitPointsSize is
// done under lock protection to avoid a race with Thread::is_available_to().
Threads.mutex.lock();
sp.mutex.lock();
} }
searching = true; setupStates->back() = tmp;
splitPointsSize--;
activeSplitPoint = sp.parentSplitPoint;
activePosition = &pos;
pos.set_nodes_searched(pos.nodes_searched() + sp.nodes);
*bestMove = sp.bestMove;
*bestValue = sp.bestValue;
sp.mutex.unlock(); main()->start_searching();
Threads.mutex.unlock();
}
// Explicit template instantiations
template void Thread::split<false>(Position&, const Stack*, Value, Value, Value*, Move*, Depth, Move, int, MovePicker*, int, bool);
template void Thread::split< true>(Position&, const Stack*, Value, Value, Value*, Move*, Depth, Move, int, MovePicker*, int, bool);
// wait_for_think_finished() waits for main thread to go to sleep then returns
void ThreadPool::wait_for_think_finished() {
MainThread* t = main();
t->mutex.lock();
while (t->thinking) sleepCondition.wait(t->mutex);
t->mutex.unlock();
}
// start_thinking() wakes up the main thread sleeping in MainThread::idle_loop()
// so to start a new search, then returns immediately.
void ThreadPool::start_thinking(const Position& pos, const LimitsType& limits,
const std::vector<Move>& searchMoves, StateStackPtr& states) {
wait_for_think_finished();
SearchTime = Time::now(); // As early as possible
Signals.stopOnPonderhit = Signals.firstRootMove = false;
Signals.stop = Signals.failedLowAtRoot = false;
RootMoves.clear();
RootPos = pos;
Limits = limits;
if (states.get()) // If we don't set a new position, preserve current state
{
SetupStates = states; // Ownership transfer here
assert(!states.get());
}
for (MoveList<LEGAL> it(pos); *it; ++it)
if ( searchMoves.empty()
|| std::count(searchMoves.begin(), searchMoves.end(), *it))
RootMoves.push_back(RootMove(*it));
main()->thinking = true;
main()->notify_one(); // Starts main thread
} }
+68 -122
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,6 +21,10 @@
#ifndef THREAD_H_INCLUDED #ifndef THREAD_H_INCLUDED
#define THREAD_H_INCLUDED #define THREAD_H_INCLUDED
#include <atomic>
#include <condition_variable>
#include <mutex>
#include <thread>
#include <vector> #include <vector>
#include "material.h" #include "material.h"
@@ -27,150 +32,91 @@
#include "pawns.h" #include "pawns.h"
#include "position.h" #include "position.h"
#include "search.h" #include "search.h"
#include "thread_win32.h"
const int MAX_THREADS = 64; // Because SplitPoint::slavesMask is a uint64_t
const int MAX_SPLITPOINTS_PER_THREAD = 8;
struct Mutex {
Mutex() { lock_init(l); }
~Mutex() { lock_destroy(l); }
void lock() { lock_grab(l); }
void unlock() { lock_release(l); }
private:
friend struct ConditionVariable;
Lock l;
};
struct ConditionVariable {
ConditionVariable() { cond_init(c); }
~ConditionVariable() { cond_destroy(c); }
void wait(Mutex& m) { cond_wait(c, m.l); }
void wait_for(Mutex& m, int ms) { timed_wait(c, m.l, ms); }
void notify_one() { cond_signal(c); }
private:
WaitCondition c;
};
struct Thread;
struct SplitPoint {
// Const data after split point has been setup
const Position* pos;
const Search::Stack* ss;
Thread* masterThread;
Depth depth;
Value beta;
int nodeType;
Move threatMove;
bool cutNode;
// Const pointers to shared data
MovePicker* movePicker;
SplitPoint* parentSplitPoint;
// Shared data
Mutex mutex;
volatile uint64_t slavesMask;
volatile int64_t nodes;
volatile Value alpha;
volatile Value bestValue;
volatile Move bestMove;
volatile int moveCount;
volatile bool cutoff;
};
/// ThreadBase struct is the base of the hierarchy from where we derive all the /// Thread class keeps together all the thread-related stuff. We use
/// specialized thread classes. /// per-thread pawn and material hash tables so that once we get a
/// pointer to an entry its life time is unlimited and we don't have
/// to care about someone changing the entry under our feet.
struct ThreadBase { class Thread {
ThreadBase() : exit(false) {}
virtual ~ThreadBase() {}
virtual void idle_loop() = 0;
void notify_one();
void wait_for(volatile const bool& b);
Mutex mutex; Mutex mutex;
ConditionVariable sleepCondition; ConditionVariable cv;
NativeHandle handle; size_t idx;
volatile bool exit; bool exit = false, searching = true; // Set before starting std::thread
}; std::thread stdThread;
public:
explicit Thread(size_t);
virtual ~Thread();
virtual void search();
void clear();
void idle_loop();
void start_searching();
void wait_for_search_finished();
/// Thread struct keeps together all the thread related stuff like locks, state Pawns::Table pawnsTable;
/// and especially split points. We also use per-thread pawn and material hash
/// tables so that once we get a pointer to an entry its life time is unlimited
/// and we don't have to care about someone changing the entry under our feet.
struct Thread : public ThreadBase {
Thread();
virtual void idle_loop();
bool cutoff_occurred() const;
bool is_available_to(const Thread* master) const;
template <bool Fake>
void split(Position& pos, const Search::Stack* ss, Value alpha, Value beta, Value* bestValue, Move* bestMove,
Depth depth, Move threatMove, int moveCount, MovePicker* movePicker, int nodeType, bool cutNode);
SplitPoint splitPoints[MAX_SPLITPOINTS_PER_THREAD];
Material::Table materialTable; Material::Table materialTable;
Endgames endgames; Endgames endgames;
Pawns::Table pawnsTable; size_t pvIdx, pvLast;
Position* activePosition; int selDepth, nmpMinPly;
size_t idx; Color nmpColor;
int maxPly; std::atomic<uint64_t> nodes, tbHits;
SplitPoint* volatile activeSplitPoint;
volatile int splitPointsSize; Position rootPos;
volatile bool searching; Search::RootMoves rootMoves;
Depth rootDepth, completedDepth;
CounterMoveHistory counterMoves;
ButterflyHistory mainHistory;
CapturePieceToHistory captureHistory;
ContinuationHistory continuationHistory;
Score contempt;
}; };
/// MainThread and TimerThread are derived classes used to characterize the two /// MainThread is a derived class specific for main thread
/// special threads: the main one and the recurring timer.
struct MainThread : public Thread { struct MainThread : public Thread {
MainThread() : thinking(true) {} // Avoid a race with start_thinking()
virtual void idle_loop();
volatile bool thinking;
};
struct TimerThread : public ThreadBase { using Thread::Thread;
TimerThread() : msec(0) {}
virtual void idle_loop(); void search() override;
int msec; void check_time();
double bestMoveChanges, previousTimeReduction;
Value previousScore;
int callsCnt;
}; };
/// ThreadPool struct handles all the threads related stuff like init, starting, /// ThreadPool struct handles all the threads-related stuff like init, starting,
/// parking and, the most important, launching a slave thread at a split point. /// parking and, most importantly, launching a thread. All the access to threads
/// All the access to shared thread data is done through this class. /// is done through this class.
struct ThreadPool : public std::vector<Thread*> { struct ThreadPool : public std::vector<Thread*> {
void init(); // No c'tor and d'tor, threads rely on globals that should void start_thinking(Position&, StateListPtr&, const Search::LimitsType&, bool = false);
void exit(); // be initialized and valid during the whole thread lifetime. void clear();
void set(size_t);
MainThread* main() { return static_cast<MainThread*>((*this)[0]); } MainThread* main() const { return static_cast<MainThread*>(front()); }
void read_uci_options(); uint64_t nodes_searched() const { return accumulate(&Thread::nodes); }
Thread* available_slave(const Thread* master) const; uint64_t tb_hits() const { return accumulate(&Thread::tbHits); }
void wait_for_think_finished();
void start_thinking(const Position&, const Search::LimitsType&,
const std::vector<Move>&, Search::StateStackPtr&);
bool sleepWhileIdle; std::atomic_bool stop, ponder, stopOnPonderhit;
Depth minimumSplitDepth;
size_t maxThreadsPerSplitPoint; private:
Mutex mutex; StateListPtr setupStates;
ConditionVariable sleepCondition;
TimerThread* timer; uint64_t accumulate(std::atomic<uint64_t> Thread::* member) const {
uint64_t sum = 0;
for (Thread* th : *this)
sum += (th->*member).load(std::memory_order_relaxed);
return sum;
}
}; };
extern ThreadPool Threads; extern ThreadPool Threads;
+70
View File
@@ -0,0 +1,70 @@
/*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
Stockfish is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef THREAD_WIN32_H_INCLUDED
#define THREAD_WIN32_H_INCLUDED
/// STL thread library used by mingw and gcc when cross compiling for Windows
/// relies on libwinpthread. Currently libwinpthread implements mutexes directly
/// on top of Windows semaphores. Semaphores, being kernel objects, require kernel
/// mode transition in order to lock or unlock, which is very slow compared to
/// interlocked operations (about 30% slower on bench test). To work around this
/// issue, we define our wrappers to the low level Win32 calls. We use critical
/// sections to support Windows XP and older versions. Unfortunately, cond_wait()
/// is racy between unlock() and WaitForSingleObject() but they have the same
/// speed performance as the SRW locks.
#include <condition_variable>
#include <mutex>
#if defined(_WIN32) && !defined(_MSC_VER)
#ifndef NOMINMAX
# define NOMINMAX // Disable macros min() and max()
#endif
#define WIN32_LEAN_AND_MEAN
#include <windows.h>
#undef WIN32_LEAN_AND_MEAN
#undef NOMINMAX
/// Mutex and ConditionVariable struct are wrappers of the low level locking
/// machinery and are modeled after the corresponding C++11 classes.
struct Mutex {
Mutex() { InitializeCriticalSection(&cs); }
~Mutex() { DeleteCriticalSection(&cs); }
void lock() { EnterCriticalSection(&cs); }
void unlock() { LeaveCriticalSection(&cs); }
private:
CRITICAL_SECTION cs;
};
typedef std::condition_variable_any ConditionVariable;
#else // Default case: use STL classes
typedef std::mutex Mutex;
typedef std::condition_variable ConditionVariable;
#endif
#endif // #ifndef THREAD_WIN32_H_INCLUDED
+88 -117
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,146 +18,116 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#include <cmath>
#include <algorithm> #include <algorithm>
#include <cfloat>
#include <cmath>
#include "search.h" #include "search.h"
#include "timeman.h" #include "timeman.h"
#include "ucioption.h" #include "uci.h"
TimeManagement Time; // Our global time management object
namespace { namespace {
/// Constants
const int MoveHorizon = 50; // Plan time management at most this many moves ahead
const float MaxRatio = 7.0f; // When in trouble, we can step over reserved time with this ratio
const float StealRatio = 0.33f; // However we must not steal time from remaining moves over this ratio
// MoveImportance[] is based on naive statistical analysis of "how many games are still undecided
// after n half-moves". Game is considered "undecided" as long as neither side has >275cp advantage.
// Data was extracted from CCRL game database with some simple filtering criteria.
const int MoveImportance[512] = {
7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780, 7780,
7780, 7780, 7780, 7780, 7778, 7778, 7776, 7776, 7776, 7773, 7770, 7768, 7766, 7763, 7757, 7751,
7743, 7735, 7724, 7713, 7696, 7689, 7670, 7656, 7627, 7605, 7571, 7549, 7522, 7493, 7462, 7425,
7385, 7350, 7308, 7272, 7230, 7180, 7139, 7094, 7055, 7010, 6959, 6902, 6841, 6778, 6705, 6651,
6569, 6508, 6435, 6378, 6323, 6253, 6152, 6085, 5995, 5931, 5859, 5794, 5717, 5646, 5544, 5462,
5364, 5282, 5172, 5078, 4988, 4901, 4831, 4764, 4688, 4609, 4536, 4443, 4365, 4293, 4225, 4155,
4085, 4005, 3927, 3844, 3765, 3693, 3634, 3560, 3479, 3404, 3331, 3268, 3207, 3146, 3077, 3011,
2947, 2894, 2828, 2776, 2727, 2676, 2626, 2589, 2538, 2490, 2442, 2394, 2345, 2302, 2243, 2192,
2156, 2115, 2078, 2043, 2004, 1967, 1922, 1893, 1845, 1809, 1772, 1736, 1702, 1674, 1640, 1605,
1566, 1536, 1509, 1479, 1452, 1423, 1388, 1362, 1332, 1304, 1289, 1266, 1250, 1228, 1206, 1180,
1160, 1134, 1118, 1100, 1080, 1068, 1051, 1034, 1012, 1001, 980, 960, 945, 934, 916, 900, 888,
878, 865, 852, 828, 807, 787, 770, 753, 744, 731, 722, 706, 700, 683, 676, 671, 664, 652, 641,
634, 627, 613, 604, 591, 582, 568, 560, 552, 540, 534, 529, 519, 509, 495, 484, 474, 467, 460,
450, 438, 427, 419, 410, 406, 399, 394, 387, 382, 377, 372, 366, 359, 353, 348, 343, 337, 333,
328, 321, 315, 309, 303, 298, 293, 287, 284, 281, 277, 273, 265, 261, 255, 251, 247, 241, 240,
235, 229, 218, 217, 213, 212, 208, 206, 197, 193, 191, 189, 185, 184, 180, 177, 172, 170, 170,
170, 166, 163, 159, 158, 156, 155, 151, 146, 141, 138, 136, 132, 130, 128, 125, 123, 122, 118,
118, 118, 117, 115, 114, 108, 107, 105, 105, 105, 102, 97, 97, 95, 94, 93, 91, 88, 86, 83, 80,
80, 79, 79, 79, 78, 76, 75, 72, 72, 71, 70, 68, 65, 63, 61, 61, 59, 59, 59, 58, 56, 55, 54, 54,
52, 49, 48, 48, 48, 48, 45, 45, 45, 44, 43, 41, 41, 41, 41, 40, 40, 38, 37, 36, 34, 34, 34, 33,
31, 29, 29, 29, 28, 28, 28, 28, 28, 28, 28, 27, 27, 27, 27, 27, 24, 24, 23, 23, 22, 21, 20, 20,
19, 19, 19, 19, 19, 18, 18, 18, 18, 17, 17, 17, 17, 17, 16, 16, 15, 15, 14, 14, 14, 12, 12, 11,
9, 9, 9, 9, 9, 9, 9, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8,
8, 8, 8, 8, 7, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 5, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4,
4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 3, 3, 2, 2, 2, 2,
2, 1, 1, 1, 1, 1, 1, 1 };
int move_importance(int ply) { return MoveImportance[std::min(ply, 511)]; }
/// Function Prototypes
enum TimeType { OptimumTime, MaxTime }; enum TimeType { OptimumTime, MaxTime };
template<TimeType> constexpr int MoveHorizon = 50; // Plan time management at most this many moves ahead
int remaining(int myTime, int movesToGo, int fullMoveNumber, int slowMover); constexpr double MaxRatio = 7.3; // When in trouble, we can step over reserved time with this ratio
} constexpr double StealRatio = 0.34; // However we must not steal time from remaining moves over this ratio
void TimeManager::pv_instability(int curChanges, int prevChanges) { // move_importance() is a skew-logistic function based on naive statistical
// analysis of "how many games are still undecided after n half-moves". Game
// is considered "undecided" as long as neither side has >275cp advantage.
// Data was extracted from the CCRL game database with some simple filtering criteria.
unstablePVExtraTime = curChanges * (optimumSearchTime / 2) double move_importance(int ply) {
+ prevChanges * (optimumSearchTime / 3);
} constexpr double XScale = 6.85;
constexpr double XShift = 64.5;
constexpr double Skew = 0.171;
return pow((1 + exp((ply - XShift) / XScale)), -Skew) + DBL_MIN; // Ensure non-zero
}
template<TimeType T>
TimePoint remaining(TimePoint myTime, int movesToGo, int ply, TimePoint slowMover) {
constexpr double TMaxRatio = (T == OptimumTime ? 1.0 : MaxRatio);
constexpr double TStealRatio = (T == OptimumTime ? 0.0 : StealRatio);
double moveImportance = (move_importance(ply) * slowMover) / 100.0;
double otherMovesImportance = 0.0;
for (int i = 1; i < movesToGo; ++i)
otherMovesImportance += move_importance(ply + 2 * i);
double ratio1 = (TMaxRatio * moveImportance) / (TMaxRatio * moveImportance + otherMovesImportance);
double ratio2 = (moveImportance + TStealRatio * otherMovesImportance) / (moveImportance + otherMovesImportance);
return TimePoint(myTime * std::min(ratio1, ratio2)); // Intel C++ asks for an explicit cast
}
} // namespace
void TimeManager::init(const Search::LimitsType& limits, int currentPly, Color us) /// init() is called at the beginning of the search and calculates the allowed
{ /// thinking time out of the time control and current game ply. We support four
/* We support four different kind of time controls: /// different kinds of time controls, passed in 'limits':
///
/// inc == 0 && movestogo == 0 means: x basetime [sudden death!]
/// inc == 0 && movestogo != 0 means: x moves in y minutes
/// inc > 0 && movestogo == 0 means: x basetime + z increment
/// inc > 0 && movestogo != 0 means: x moves in y minutes + z increment
increment == 0 && movesToGo == 0 means: x basetime [sudden death!] void TimeManagement::init(Search::LimitsType& limits, Color us, int ply) {
increment == 0 && movesToGo != 0 means: x moves in y minutes
increment > 0 && movesToGo == 0 means: x basetime + z increment
increment > 0 && movesToGo != 0 means: x moves in y minutes + z increment
Time management is adjusted by following UCI parameters: TimePoint minThinkingTime = Options["Minimum Thinking Time"];
TimePoint moveOverhead = Options["Move Overhead"];
TimePoint slowMover = Options["Slow Mover"];
TimePoint npmsec = Options["nodestime"];
TimePoint hypMyTime;
emergencyMoveHorizon: Be prepared to always play at least this many moves // If we have to play in 'nodes as time' mode, then convert from time
emergencyBaseTime : Always attempt to keep at least this much time (in ms) at clock // to nodes, and use resulting values in time management formulas.
emergencyMoveTime : Plus attempt to keep at least this much time for each remaining emergency move // WARNING: to avoid time losses, the given npmsec (nodes per millisecond)
minThinkingTime : No matter what, use at least this much thinking before doing the move // must be much lower than the real engine speed.
*/ if (npmsec)
int hypMTG, hypMyTime, t1, t2;
// Read uci parameters
int emergencyMoveHorizon = Options["Emergency Move Horizon"];
int emergencyBaseTime = Options["Emergency Base Time"];
int emergencyMoveTime = Options["Emergency Move Time"];
int minThinkingTime = Options["Minimum Thinking Time"];
int slowMover = Options["Slow Mover"];
// Initialize to maximum values but unstablePVExtraTime that is reset
unstablePVExtraTime = 0;
optimumSearchTime = maximumSearchTime = limits.time[us];
// We calculate optimum time usage for different hypothetic "moves to go"-values and choose the
// minimum of calculated search time values. Usually the greatest hypMTG gives the minimum values.
for (hypMTG = 1; hypMTG <= (limits.movestogo ? std::min(limits.movestogo, MoveHorizon) : MoveHorizon); hypMTG++)
{ {
// Calculate thinking time for hypothetic "moves to go"-value if (!availableNodes) // Only once at game start
availableNodes = npmsec * limits.time[us]; // Time is in msec
// Convert from milliseconds to nodes
limits.time[us] = TimePoint(availableNodes);
limits.inc[us] *= npmsec;
limits.npmsec = npmsec;
}
startTime = limits.startTime;
optimumTime = maximumTime = std::max(limits.time[us], minThinkingTime);
const int maxMTG = limits.movestogo ? std::min(limits.movestogo, MoveHorizon) : MoveHorizon;
// We calculate optimum time usage for different hypothetical "moves to go" values
// and choose the minimum of calculated search time values. Usually the greatest
// hypMTG gives the minimum values.
for (int hypMTG = 1; hypMTG <= maxMTG; ++hypMTG)
{
// Calculate thinking time for hypothetical "moves to go"-value
hypMyTime = limits.time[us] hypMyTime = limits.time[us]
+ limits.inc[us] * (hypMTG - 1) + limits.inc[us] * (hypMTG - 1)
- emergencyBaseTime - moveOverhead * (2 + std::min(hypMTG, 40));
- emergencyMoveTime * std::min(hypMTG, emergencyMoveHorizon);
hypMyTime = std::max(hypMyTime, 0); hypMyTime = std::max(hypMyTime, TimePoint(0));
t1 = minThinkingTime + remaining<OptimumTime>(hypMyTime, hypMTG, currentPly, slowMover); TimePoint t1 = minThinkingTime + remaining<OptimumTime>(hypMyTime, hypMTG, ply, slowMover);
t2 = minThinkingTime + remaining<MaxTime>(hypMyTime, hypMTG, currentPly, slowMover); TimePoint t2 = minThinkingTime + remaining<MaxTime >(hypMyTime, hypMTG, ply, slowMover);
optimumSearchTime = std::min(optimumSearchTime, t1); optimumTime = std::min(t1, optimumTime);
maximumSearchTime = std::min(maximumSearchTime, t2); maximumTime = std::min(t2, maximumTime);
} }
if (Options["Ponder"]) if (Options["Ponder"])
optimumSearchTime += optimumSearchTime / 4; optimumTime += optimumTime / 4;
// Make sure that maxSearchTime is not over absoluteMaxSearchTime
optimumSearchTime = std::min(optimumSearchTime, maximumSearchTime);
}
namespace {
template<TimeType T>
int remaining(int myTime, int movesToGo, int currentPly, int slowMover)
{
const float TMaxRatio = (T == OptimumTime ? 1 : MaxRatio);
const float TStealRatio = (T == OptimumTime ? 0 : StealRatio);
int thisMoveImportance = move_importance(currentPly) * slowMover / 100;
int otherMovesImportance = 0;
for (int i = 1; i < movesToGo; i++)
otherMovesImportance += move_importance(currentPly + 2 * i);
float ratio1 = (TMaxRatio * thisMoveImportance) / float(TMaxRatio * thisMoveImportance + otherMovesImportance);
float ratio2 = (thisMoveImportance + TStealRatio * otherMovesImportance) / float(thisMoveImportance + otherMovesImportance);
return int(floor(myTime * std::min(ratio1, ratio2)));
}
} }
+21 -11
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,20 +21,29 @@
#ifndef TIMEMAN_H_INCLUDED #ifndef TIMEMAN_H_INCLUDED
#define TIMEMAN_H_INCLUDED #define TIMEMAN_H_INCLUDED
/// The TimeManager class computes the optimal time to think depending on the #include "misc.h"
/// maximum available time, the move game number and other parameters. #include "search.h"
#include "thread.h"
class TimeManager { /// The TimeManagement class computes the optimal time to think depending on
/// the maximum available time, the game move number and other parameters.
class TimeManagement {
public: public:
void init(const Search::LimitsType& limits, int currentPly, Color us); void init(Search::LimitsType& limits, Color us, int ply);
void pv_instability(int curChanges, int prevChanges); TimePoint optimum() const { return optimumTime; }
int available_time() const { return optimumSearchTime + unstablePVExtraTime; } TimePoint maximum() const { return maximumTime; }
int maximum_time() const { return maximumSearchTime; } TimePoint elapsed() const { return Search::Limits.npmsec ?
TimePoint(Threads.nodes_searched()) : now() - startTime; }
int64_t availableNodes; // When in 'nodes as time' mode
private: private:
int optimumSearchTime; TimePoint startTime;
int maximumSearchTime; TimePoint optimumTime;
int unstablePVExtraTime; TimePoint maximumTime;
}; };
extern TimeManagement Time;
#endif // #ifndef TIMEMAN_H_INCLUDED #endif // #ifndef TIMEMAN_H_INCLUDED
+101 -66
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,31 +18,50 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#include <cstring> #include <cstring> // For std::memset
#include <iostream> #include <iostream>
#include <thread>
#include "bitboard.h" #include "bitboard.h"
#include "misc.h"
#include "tt.h" #include "tt.h"
#include "uci.h"
TranspositionTable TT; // Our global transposition table TranspositionTable TT; // Our global transposition table
/// TTEntry::save saves a TTEntry
void TTEntry::save(Key k, Value v, Bound b, Depth d, Move m, Value ev) {
/// TranspositionTable::set_size() sets the size of the transposition table, assert(d / ONE_PLY * ONE_PLY == d);
// Preserve any existing move for the same position
if (m || (k >> 48) != key16)
move16 = (uint16_t)m;
// Overwrite less valuable entries
if ( (k >> 48) != key16
|| d / ONE_PLY > depth8 - 4
|| b == BOUND_EXACT)
{
key16 = (uint16_t)(k >> 48);
value16 = (int16_t)v;
eval16 = (int16_t)ev;
genBound8 = (uint8_t)(TT.generation8 | b);
depth8 = (int8_t)(d / ONE_PLY);
}
}
/// TranspositionTable::resize() sets the size of the transposition table,
/// measured in megabytes. Transposition table consists of a power of 2 number /// measured in megabytes. Transposition table consists of a power of 2 number
/// of clusters and each cluster consists of ClusterSize number of TTEntry. /// of clusters and each cluster consists of ClusterSize number of TTEntry.
void TranspositionTable::set_size(size_t mbSize) { void TranspositionTable::resize(size_t mbSize) {
assert(msb((mbSize << 20) / sizeof(TTEntry)) < 32); clusterCount = mbSize * 1024 * 1024 / sizeof(Cluster);
uint32_t size = ClusterSize << msb((mbSize << 20) / sizeof(TTEntry[ClusterSize]));
if (hashMask == size - ClusterSize)
return;
hashMask = size - ClusterSize;
free(mem); free(mem);
mem = calloc(size * sizeof(TTEntry) + CACHE_LINE_SIZE - 1, 1); mem = malloc(clusterCount * sizeof(Cluster) + CacheLineSize - 1);
if (!mem) if (!mem)
{ {
@@ -50,72 +70,87 @@ void TranspositionTable::set_size(size_t mbSize) {
exit(EXIT_FAILURE); exit(EXIT_FAILURE);
} }
table = (TTEntry*)((uintptr_t(mem) + CACHE_LINE_SIZE - 1) & ~(CACHE_LINE_SIZE - 1)); table = (Cluster*)((uintptr_t(mem) + CacheLineSize - 1) & ~(CacheLineSize - 1));
clear();
} }
/// TranspositionTable::clear() overwrites the entire transposition table /// TranspositionTable::clear() initializes the entire transposition table to zero,
/// with zeroes. It is called whenever the table is resized, or when the // in a multi-threaded way.
/// user asks the program to clear the table (from the UCI interface).
void TranspositionTable::clear() { void TranspositionTable::clear() {
std::memset(table, 0, (hashMask + ClusterSize) * sizeof(TTEntry)); std::vector<std::thread> threads;
}
for (size_t idx = 0; idx < Options["Threads"]; idx++)
/// TranspositionTable::probe() looks up the current position in the
/// transposition table. Returns a pointer to the TTEntry or NULL if
/// position is not found.
const TTEntry* TranspositionTable::probe(const Key key) const {
const TTEntry* tte = first_entry(key);
uint32_t key32 = key >> 32;
for (unsigned i = 0; i < ClusterSize; i++, tte++)
if (tte->key() == key32)
return tte;
return NULL;
}
/// TranspositionTable::store() writes a new entry containing position key and
/// valuable information of current position. The lowest order bits of position
/// key are used to decide on which cluster the position will be placed.
/// When a new entry is written and there are no empty entries available in cluster,
/// it replaces the least valuable of entries. A TTEntry t1 is considered to be
/// more valuable than a TTEntry t2 if t1 is from the current search and t2 is from
/// a previous search, or if the depth of t1 is bigger than the depth of t2.
void TranspositionTable::store(const Key key, Value v, Bound b, Depth d, Move m, Value statV, Value evalM) {
int c1, c2, c3;
TTEntry *tte, *replace;
uint32_t key32 = key >> 32; // Use the high 32 bits as key inside the cluster
tte = replace = first_entry(key);
for (unsigned i = 0; i < ClusterSize; i++, tte++)
{ {
if (!tte->key() || tte->key() == key32) // Empty or overwrite old threads.emplace_back([this, idx]() {
{
if (!m)
m = tte->move(); // Preserve any existing ttMove
replace = tte; // Thread binding gives faster search on systems with a first-touch policy
break; if (Options["Threads"] > 8)
WinProcGroup::bindThisThread(idx);
// Each thread will zero its part of the hash table
const size_t stride = clusterCount / Options["Threads"],
start = stride * idx,
len = idx != Options["Threads"] - 1 ?
stride : clusterCount - start;
std::memset(&table[start], 0, len * sizeof(Cluster));
});
} }
// Implement replace strategy for (std::thread& th: threads)
c1 = (replace->generation() == generation ? 2 : 0); th.join();
c2 = (tte->generation() == generation || tte->bound() == BOUND_EXACT ? -2 : 0); }
c3 = (tte->depth() < replace->depth() ? 1 : 0);
if (c1 + c2 + c3 > 0) /// TranspositionTable::probe() looks up the current position in the transposition
replace = tte; /// table. It returns true and a pointer to the TTEntry if the position is found.
/// Otherwise, it returns false and a pointer to an empty or least valuable TTEntry
/// to be replaced later. The replace value of an entry is calculated as its depth
/// minus 8 times its relative age. TTEntry t1 is considered more valuable than
/// TTEntry t2 if its replace value is greater than that of t2.
TTEntry* TranspositionTable::probe(const Key key, bool& found) const {
TTEntry* const tte = first_entry(key);
const uint16_t key16 = key >> 48; // Use the high 16 bits as key inside the cluster
for (int i = 0; i < ClusterSize; ++i)
if (!tte[i].key16 || tte[i].key16 == key16)
{
tte[i].genBound8 = uint8_t(generation8 | tte[i].bound()); // Refresh
return found = (bool)tte[i].key16, &tte[i];
} }
replace->save(key32, v, b, d, m, generation, statV, evalM); // Find an entry to be replaced according to the replacement strategy
TTEntry* replace = tte;
for (int i = 1; i < ClusterSize; ++i)
// Due to our packed storage format for generation and its cyclic
// nature we add 259 (256 is the modulus plus 3 to keep the lowest
// two bound bits from affecting the result) to calculate the entry
// age correctly even after generation8 overflows into the next cycle.
if ( replace->depth8 - ((259 + generation8 - replace->genBound8) & 0xFC) * 2
> tte[i].depth8 - ((259 + generation8 - tte[i].genBound8) & 0xFC) * 2)
replace = &tte[i];
return found = false, replace;
}
/// TranspositionTable::hashfull() returns an approximation of the hashtable
/// occupation during a search. The hash is x permill full, as per UCI protocol.
int TranspositionTable::hashfull() const {
int cnt = 0;
for (int i = 0; i < 1000 / ClusterSize; i++)
{
const TTEntry* tte = &table[i].entry[0];
for (int j = 0; j < ClusterSize; j++)
if ((tte[j].genBound8 & 0xFC) == generation8)
cnt++;
}
return cnt;
} }
+49 -66
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -23,96 +24,78 @@
#include "misc.h" #include "misc.h"
#include "types.h" #include "types.h"
/// The TTEntry is the 128 bit transposition table entry, defined as below: /// TTEntry struct is the 10 bytes transposition table entry, defined as below:
/// ///
/// key: 32 bit /// key 16 bit
/// move: 16 bit /// move 16 bit
/// bound type: 8 bit /// value 16 bit
/// generation: 8 bit /// eval value 16 bit
/// value: 16 bit /// generation 6 bit
/// depth: 16 bit /// bound type 2 bit
/// static value: 16 bit /// depth 8 bit
/// static margin: 16 bit
struct TTEntry { struct TTEntry {
void save(uint32_t k, Value v, Bound b, Depth d, Move m, int g, Value ev, Value em) { Move move() const { return (Move )move16; }
key32 = (uint32_t)k;
move16 = (uint16_t)m;
bound8 = (uint8_t)b;
generation8 = (uint8_t)g;
value16 = (int16_t)v;
depth16 = (int16_t)d;
evalValue = (int16_t)ev;
evalMargin = (int16_t)em;
}
void set_generation(uint8_t g) { generation8 = g; }
uint32_t key() const { return key32; }
Depth depth() const { return (Depth)depth16; }
Move move() const { return (Move)move16; }
Value value() const { return (Value)value16; } Value value() const { return (Value)value16; }
Bound bound() const { return (Bound)bound8; } Value eval() const { return (Value)eval16; }
int generation() const { return (int)generation8; } Depth depth() const { return (Depth)(depth8 * int(ONE_PLY)); }
Value eval_value() const { return (Value)evalValue; } Bound bound() const { return (Bound)(genBound8 & 0x3); }
Value eval_margin() const { return (Value)evalMargin; } void save(Key k, Value v, Bound b, Depth d, Move m, Value ev);
private: private:
uint32_t key32; friend class TranspositionTable;
uint16_t key16;
uint16_t move16; uint16_t move16;
uint8_t bound8, generation8; int16_t value16;
int16_t value16, depth16, evalValue, evalMargin; int16_t eval16;
uint8_t genBound8;
int8_t depth8;
}; };
/// A TranspositionTable consists of a power of 2 number of clusters and each /// A TranspositionTable consists of a power of 2 number of clusters and each
/// cluster consists of ClusterSize number of TTEntry. Each non-empty entry /// cluster consists of ClusterSize number of TTEntry. Each non-empty entry
/// contains information of exactly one position. Size of a cluster shall not be /// contains information of exactly one position. The size of a cluster should
/// bigger than a cache line size. In case it is less, it should be padded to /// divide the size of a cache line size, to ensure that clusters never cross
/// guarantee always aligned accesses. /// cache lines. This ensures best cache performance, as the cacheline is
/// prefetched, as soon as possible.
class TranspositionTable { class TranspositionTable {
static const unsigned ClusterSize = 4; // A cluster is 64 Bytes static constexpr int CacheLineSize = 64;
static constexpr int ClusterSize = 3;
struct Cluster {
TTEntry entry[ClusterSize];
char padding[2]; // Align to a divisor of the cache line size
};
static_assert(CacheLineSize % sizeof(Cluster) == 0, "Cluster size incorrect");
public: public:
~TranspositionTable() { free(mem); } ~TranspositionTable() { free(mem); }
void new_search() { generation++; } void new_search() { generation8 += 4; } // Lower 2 bits are used by Bound
TTEntry* probe(const Key key, bool& found) const;
const TTEntry* probe(const Key key) const; int hashfull() const;
TTEntry* first_entry(const Key key) const; void resize(size_t mbSize);
void refresh(const TTEntry* tte) const;
void set_size(size_t mbSize);
void clear(); void clear();
void store(const Key key, Value v, Bound type, Depth d, Move m, Value statV, Value kingD);
// The 32 lowest order bits of the key are used to get the index of the cluster
TTEntry* first_entry(const Key key) const {
return &table[(uint32_t(key) * uint64_t(clusterCount)) >> 32].entry[0];
}
private: private:
uint32_t hashMask; friend struct TTEntry;
TTEntry* table;
size_t clusterCount;
Cluster* table;
void* mem; void* mem;
uint8_t generation; // Size must be not bigger than TTEntry::generation8 uint8_t generation8; // Size must be not bigger than TTEntry::genBound8
}; };
extern TranspositionTable TT; extern TranspositionTable TT;
/// TranspositionTable::first_entry() returns a pointer to the first entry of
/// a cluster given a position. The lowest order bits of the key are used to
/// get the index of the cluster.
inline TTEntry* TranspositionTable::first_entry(const Key key) const {
return table + ((uint32_t)key & hashMask);
}
/// TranspositionTable::refresh() updates the 'generation' value of the TTEntry
/// to avoid aging. Normally called after a TT hit.
inline void TranspositionTable::refresh(const TTEntry* tte) const {
const_cast<TTEntry*>(tte)->set_generation(generation);
}
#endif // #ifndef TT_H_INCLUDED #endif // #ifndef TT_H_INCLUDED
+223 -208
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -20,90 +21,101 @@
#ifndef TYPES_H_INCLUDED #ifndef TYPES_H_INCLUDED
#define TYPES_H_INCLUDED #define TYPES_H_INCLUDED
/// For Linux and OSX configuration is done automatically using Makefile. To get /// When compiling with provided Makefile (e.g. for Linux and OSX), configuration
/// started type 'make help'. /// is done automatically. To get started type 'make help'.
/// ///
/// For Windows, part of the configuration is detected automatically, but some /// When Makefile is not used (e.g. with Microsoft Visual Studio) some switches
/// switches need to be set manually: /// need to be set manually:
/// ///
/// -DNDEBUG | Disable debugging mode. Use always. /// -DNDEBUG | Disable debugging mode. Always use this for release.
/// ///
/// -DNO_PREFETCH | Disable use of prefetch asm-instruction. A must if you want /// -DNO_PREFETCH | Disable use of prefetch asm-instruction. You may need this to
/// | the executable to run on some very old machines. /// | run on some very old machines.
/// ///
/// -DUSE_POPCNT | Add runtime support for use of popcnt asm-instruction. Works /// -DUSE_POPCNT | Add runtime support for use of popcnt asm-instruction. Works
/// | only in 64-bit mode. For compiling requires hardware with /// | only in 64-bit mode and requires hardware with popcnt support.
/// | popcnt support. ///
/// -DUSE_PEXT | Add runtime support for use of pext asm-instruction. Works
/// | only in 64-bit mode and requires hardware with pext support.
#include <cassert> #include <cassert>
#include <cctype> #include <cctype>
#include <climits> #include <climits>
#include <cstdint>
#include <cstdlib> #include <cstdlib>
#include "platform.h" #if defined(_MSC_VER)
// Disable some silly and noisy warning from MSVC compiler
#pragma warning(disable: 4127) // Conditional expression is constant
#pragma warning(disable: 4146) // Unary minus operator applied to unsigned type
#pragma warning(disable: 4800) // Forcing value to bool 'true' or 'false'
#endif
#define unlikely(x) (x) // For code annotation purposes /// Predefined macros hell:
///
/// __GNUC__ Compiler is gcc, Clang or Intel on Linux
/// __INTEL_COMPILER Compiler is Intel
/// _MSC_VER Compiler is MSVC or Intel on Windows
/// _WIN32 Building on Windows (any)
/// _WIN64 Building on Windows 64 bit
#if defined(_WIN64) && !defined(IS_64BIT) #if defined(_WIN64) && defined(_MSC_VER) // No Makefile used
# include <intrin.h> // MSVC popcnt and bsfq instrinsics # include <intrin.h> // Microsoft header for _BitScanForward64()
# define IS_64BIT # define IS_64BIT
# define USE_BSFQ
#endif #endif
#if defined(USE_POPCNT) && defined(_MSC_VER) && defined(__INTEL_COMPILER) #if defined(USE_POPCNT) && (defined(__INTEL_COMPILER) || defined(_MSC_VER))
# include <nmmintrin.h> // Intel header for _mm_popcnt_u64() intrinsic # include <nmmintrin.h> // Intel and Microsoft header for _mm_popcnt_u64()
#endif #endif
# if !defined(NO_PREFETCH) && (defined(__INTEL_COMPILER) || defined(_MSC_VER)) #if !defined(NO_PREFETCH) && (defined(__INTEL_COMPILER) || defined(_MSC_VER))
# include <xmmintrin.h> // Intel and Microsoft header for _mm_prefetch() # include <xmmintrin.h> // Intel and Microsoft header for _mm_prefetch()
# endif
#define CACHE_LINE_SIZE 64
#if defined(_MSC_VER) || defined(__INTEL_COMPILER)
# define CACHE_LINE_ALIGNMENT __declspec(align(CACHE_LINE_SIZE))
#else
# define CACHE_LINE_ALIGNMENT __attribute__ ((aligned(CACHE_LINE_SIZE)))
#endif #endif
#ifdef _MSC_VER #if defined(USE_PEXT)
# define FORCE_INLINE __forceinline # include <immintrin.h> // Header for _pext_u64() intrinsic
#elif defined(__GNUC__) # define pext(b, m) _pext_u64(b, m)
# define FORCE_INLINE inline __attribute__((always_inline))
#else #else
# define FORCE_INLINE inline # define pext(b, m) 0
#endif #endif
#ifdef USE_POPCNT #ifdef USE_POPCNT
const bool HasPopCnt = true; constexpr bool HasPopCnt = true;
#else #else
const bool HasPopCnt = false; constexpr bool HasPopCnt = false;
#endif
#ifdef USE_PEXT
constexpr bool HasPext = true;
#else
constexpr bool HasPext = false;
#endif #endif
#ifdef IS_64BIT #ifdef IS_64BIT
const bool Is64Bit = true; constexpr bool Is64Bit = true;
#else #else
const bool Is64Bit = false; constexpr bool Is64Bit = false;
#endif #endif
typedef uint64_t Key; typedef uint64_t Key;
typedef uint64_t Bitboard; typedef uint64_t Bitboard;
const int MAX_MOVES = 192; constexpr int MAX_MOVES = 256;
const int MAX_PLY = 100; constexpr int MAX_PLY = 128;
const int MAX_PLY_PLUS_6 = MAX_PLY + 6;
/// A move needs 16 bits to be stored /// A move needs 16 bits to be stored
/// ///
/// bit 0- 5: destination square (from 0 to 63) /// bit 0- 5: destination square (from 0 to 63)
/// bit 6-11: origin square (from 0 to 63) /// bit 6-11: origin square (from 0 to 63)
/// bit 12-13: promotion piece type - 2 (from KNIGHT-2 to QUEEN-2) /// bit 12-13: promotion piece type - 2 (from KNIGHT-2 to QUEEN-2)
/// bit 14-15: special move flag: promotion (1), en passant (2), castle (3) /// bit 14-15: special move flag: promotion (1), en passant (2), castling (3)
/// NOTE: EN-PASSANT bit is set only when a pawn can be captured
/// ///
/// Special cases are MOVE_NONE and MOVE_NULL. We can sneak these in because in /// Special cases are MOVE_NONE and MOVE_NULL. We can sneak these in because in
/// any normal move destination square is always different from origin square /// any normal move destination square is always different from origin square
/// while MOVE_NONE and MOVE_NULL have the same origin and destination square. /// while MOVE_NONE and MOVE_NULL have the same origin and destination square.
enum Move { enum Move : int {
MOVE_NONE, MOVE_NONE,
MOVE_NULL = 65 MOVE_NULL = 65
}; };
@@ -112,23 +124,25 @@ enum MoveType {
NORMAL, NORMAL,
PROMOTION = 1 << 14, PROMOTION = 1 << 14,
ENPASSANT = 2 << 14, ENPASSANT = 2 << 14,
CASTLE = 3 << 14 CASTLING = 3 << 14
}; };
enum CastleRight { // Defined as in PolyGlot book hash key enum Color {
CASTLES_NONE, WHITE, BLACK, COLOR_NB = 2
};
enum CastlingSide {
KING_SIDE, QUEEN_SIDE, CASTLING_SIDE_NB = 2
};
enum CastlingRight {
NO_CASTLING,
WHITE_OO, WHITE_OO,
WHITE_OOO = WHITE_OO << 1, WHITE_OOO = WHITE_OO << 1,
BLACK_OO = WHITE_OO << 2, BLACK_OO = WHITE_OO << 2,
BLACK_OOO = WHITE_OO << 3, BLACK_OOO = WHITE_OO << 3,
ALL_CASTLES = WHITE_OO | WHITE_OOO | BLACK_OO | BLACK_OOO, ANY_CASTLING = WHITE_OO | WHITE_OOO | BLACK_OO | BLACK_OOO,
CASTLE_RIGHT_NB = 16 CASTLING_RIGHT_NB = 16
};
enum CastlingSide {
KING_SIDE,
QUEEN_SIDE,
CASTLING_SIDE_NB = 2
}; };
enum Phase { enum Phase {
@@ -151,25 +165,24 @@ enum Bound {
BOUND_EXACT = BOUND_UPPER | BOUND_LOWER BOUND_EXACT = BOUND_UPPER | BOUND_LOWER
}; };
enum Value { enum Value : int {
VALUE_ZERO = 0, VALUE_ZERO = 0,
VALUE_DRAW = 0, VALUE_DRAW = 0,
VALUE_KNOWN_WIN = 15000, VALUE_KNOWN_WIN = 10000,
VALUE_MATE = 30000, VALUE_MATE = 32000,
VALUE_INFINITE = 30001, VALUE_INFINITE = 32001,
VALUE_NONE = 30002, VALUE_NONE = 32002,
VALUE_MATE_IN_MAX_PLY = VALUE_MATE - MAX_PLY, VALUE_MATE_IN_MAX_PLY = VALUE_MATE - 2 * MAX_PLY,
VALUE_MATED_IN_MAX_PLY = -VALUE_MATE + MAX_PLY, VALUE_MATED_IN_MAX_PLY = -VALUE_MATE + 2 * MAX_PLY,
VALUE_ENSURE_INTEGER_SIZE_P = INT_MAX, PawnValueMg = 136, PawnValueEg = 208,
VALUE_ENSURE_INTEGER_SIZE_N = INT_MIN, KnightValueMg = 782, KnightValueEg = 865,
BishopValueMg = 830, BishopValueEg = 918,
RookValueMg = 1289, RookValueEg = 1378,
QueenValueMg = 2529, QueenValueEg = 2687,
PawnValueMg = 198, PawnValueEg = 258, MidgameLimit = 15258, EndgameLimit = 3915
KnightValueMg = 817, KnightValueEg = 846,
BishopValueMg = 836, BishopValueEg = 857,
RookValueMg = 1270, RookValueEg = 1278,
QueenValueMg = 2521, QueenValueEg = 2558
}; };
enum PieceType { enum PieceType {
@@ -185,23 +198,24 @@ enum Piece {
PIECE_NB = 16 PIECE_NB = 16
}; };
enum Color { extern Value PieceValue[PHASE_NB][PIECE_NB];
WHITE, BLACK, NO_COLOR, COLOR_NB = 2
};
enum Depth { enum Depth : int {
ONE_PLY = 2, ONE_PLY = 1,
DEPTH_ZERO = 0 * ONE_PLY, DEPTH_ZERO = 0 * ONE_PLY,
DEPTH_QS_CHECKS = -1 * ONE_PLY, DEPTH_QS_CHECKS = 0 * ONE_PLY,
DEPTH_QS_NO_CHECKS = -2 * ONE_PLY, DEPTH_QS_NO_CHECKS = -1 * ONE_PLY,
DEPTH_QS_RECAPTURES = -7 * ONE_PLY, DEPTH_QS_RECAPTURES = -5 * ONE_PLY,
DEPTH_NONE = -127 * ONE_PLY DEPTH_NONE = -6 * ONE_PLY,
DEPTH_MAX = MAX_PLY * ONE_PLY
}; };
enum Square { static_assert(!(ONE_PLY & (ONE_PLY - 1)), "ONE_PLY is not a power of 2");
enum Square : int {
SQ_A1, SQ_B1, SQ_C1, SQ_D1, SQ_E1, SQ_F1, SQ_G1, SQ_H1, SQ_A1, SQ_B1, SQ_C1, SQ_D1, SQ_E1, SQ_F1, SQ_G1, SQ_H1,
SQ_A2, SQ_B2, SQ_C2, SQ_D2, SQ_E2, SQ_F2, SQ_G2, SQ_H2, SQ_A2, SQ_B2, SQ_C2, SQ_D2, SQ_E2, SQ_F2, SQ_G2, SQ_H2,
SQ_A3, SQ_B3, SQ_C3, SQ_D3, SQ_E3, SQ_F3, SQ_G3, SQ_H3, SQ_A3, SQ_B3, SQ_C3, SQ_D3, SQ_E3, SQ_F3, SQ_G3, SQ_H3,
@@ -212,178 +226,190 @@ enum Square {
SQ_A8, SQ_B8, SQ_C8, SQ_D8, SQ_E8, SQ_F8, SQ_G8, SQ_H8, SQ_A8, SQ_B8, SQ_C8, SQ_D8, SQ_E8, SQ_F8, SQ_G8, SQ_H8,
SQ_NONE, SQ_NONE,
SQUARE_NB = 64, SQUARE_NB = 64
DELTA_N = 8,
DELTA_E = 1,
DELTA_S = -8,
DELTA_W = -1,
DELTA_NN = DELTA_N + DELTA_N,
DELTA_NE = DELTA_N + DELTA_E,
DELTA_SE = DELTA_S + DELTA_E,
DELTA_SS = DELTA_S + DELTA_S,
DELTA_SW = DELTA_S + DELTA_W,
DELTA_NW = DELTA_N + DELTA_W
}; };
enum File { enum Direction : int {
NORTH = 8,
EAST = 1,
SOUTH = -NORTH,
WEST = -EAST,
NORTH_EAST = NORTH + EAST,
SOUTH_EAST = SOUTH + EAST,
SOUTH_WEST = SOUTH + WEST,
NORTH_WEST = NORTH + WEST
};
enum File : int {
FILE_A, FILE_B, FILE_C, FILE_D, FILE_E, FILE_F, FILE_G, FILE_H, FILE_NB FILE_A, FILE_B, FILE_C, FILE_D, FILE_E, FILE_F, FILE_G, FILE_H, FILE_NB
}; };
enum Rank { enum Rank : int {
RANK_1, RANK_2, RANK_3, RANK_4, RANK_5, RANK_6, RANK_7, RANK_8, RANK_NB RANK_1, RANK_2, RANK_3, RANK_4, RANK_5, RANK_6, RANK_7, RANK_8, RANK_NB
}; };
/// Score enum keeps a midgame and an endgame value in a single integer (enum), /// Score enum stores a middlegame and an endgame value in a single integer (enum).
/// first LSB 16 bits are used to store endgame value, while upper bits are used /// The least significant 16 bits are used to store the middlegame value and the
/// for midgame value. Compiler is free to choose the enum type as long as can /// upper 16 bits are used to store the endgame value. We have to take care to
/// keep its data, so ensure Score to be an integer type. /// avoid left-shifting a signed int to avoid undefined behavior.
enum Score { enum Score : int { SCORE_ZERO };
SCORE_ZERO,
SCORE_ENSURE_INTEGER_SIZE_P = INT_MAX,
SCORE_ENSURE_INTEGER_SIZE_N = INT_MIN
};
inline Score make_score(int mg, int eg) { return Score((mg << 16) + eg); } constexpr Score make_score(int mg, int eg) {
return Score((int)((unsigned int)eg << 16) + mg);
/// Extracting the signed lower and upper 16 bits it not so trivial because
/// according to the standard a simple cast to short is implementation defined
/// and so is a right shift of a signed integer.
inline Value mg_value(Score s) { return Value(((s + 0x8000) & ~0xffff) / 0x10000); }
/// On Intel 64 bit we have a small speed regression with the standard conforming
/// version, so use a faster code in this case that, although not 100% standard
/// compliant it seems to work for Intel and MSVC.
#if defined(IS_64BIT) && (!defined(__GNUC__) || defined(__INTEL_COMPILER))
inline Value eg_value(Score s) { return Value(int16_t(s & 0xffff)); }
#else
inline Value eg_value(Score s) {
return Value((int)(unsigned(s) & 0x7fffu) - (int)(unsigned(s) & 0x8000u));
} }
#endif /// Extracting the signed lower and upper 16 bits is not so trivial because
/// according to the standard a simple cast to short is implementation defined
/// and so is a right shift of a signed integer.
inline Value eg_value(Score s) {
union { uint16_t u; int16_t s; } eg = { uint16_t(unsigned(s + 0x8000) >> 16) };
return Value(eg.s);
}
#define ENABLE_SAFE_OPERATORS_ON(T) \ inline Value mg_value(Score s) {
inline T operator+(const T d1, const T d2) { return T(int(d1) + int(d2)); } \ union { uint16_t u; int16_t s; } mg = { uint16_t(unsigned(s)) };
inline T operator-(const T d1, const T d2) { return T(int(d1) - int(d2)); } \ return Value(mg.s);
inline T operator*(int i, const T d) { return T(i * int(d)); } \ }
inline T operator*(const T d, int i) { return T(int(d) * i); } \
inline T operator-(const T d) { return T(-int(d)); } \
inline T& operator+=(T& d1, const T d2) { d1 = d1 + d2; return d1; } \
inline T& operator-=(T& d1, const T d2) { d1 = d1 - d2; return d1; } \
inline T& operator*=(T& d, int i) { d = T(int(d) * i); return d; }
#define ENABLE_OPERATORS_ON(T) ENABLE_SAFE_OPERATORS_ON(T) \ #define ENABLE_BASE_OPERATORS_ON(T) \
inline T operator++(T& d, int) { d = T(int(d) + 1); return d; } \ constexpr T operator+(T d1, T d2) { return T(int(d1) + int(d2)); } \
inline T operator--(T& d, int) { d = T(int(d) - 1); return d; } \ constexpr T operator-(T d1, T d2) { return T(int(d1) - int(d2)); } \
inline T operator/(const T d, int i) { return T(int(d) / i); } \ constexpr T operator-(T d) { return T(-int(d)); } \
inline T& operator/=(T& d, int i) { d = T(int(d) / i); return d; } inline T& operator+=(T& d1, T d2) { return d1 = d1 + d2; } \
inline T& operator-=(T& d1, T d2) { return d1 = d1 - d2; }
ENABLE_OPERATORS_ON(Value) #define ENABLE_INCR_OPERATORS_ON(T) \
ENABLE_OPERATORS_ON(PieceType) inline T& operator++(T& d) { return d = T(int(d) + 1); } \
ENABLE_OPERATORS_ON(Piece) inline T& operator--(T& d) { return d = T(int(d) - 1); }
ENABLE_OPERATORS_ON(Color)
ENABLE_OPERATORS_ON(Depth)
ENABLE_OPERATORS_ON(Square)
ENABLE_OPERATORS_ON(File)
ENABLE_OPERATORS_ON(Rank)
/// Added operators for adding integers to a Value #define ENABLE_FULL_OPERATORS_ON(T) \
inline Value operator+(Value v, int i) { return Value(int(v) + i); } ENABLE_BASE_OPERATORS_ON(T) \
inline Value operator-(Value v, int i) { return Value(int(v) - i); } ENABLE_INCR_OPERATORS_ON(T) \
constexpr T operator*(int i, T d) { return T(i * int(d)); } \
constexpr T operator*(T d, int i) { return T(int(d) * i); } \
constexpr T operator/(T d, int i) { return T(int(d) / i); } \
constexpr int operator/(T d1, T d2) { return int(d1) / int(d2); } \
inline T& operator*=(T& d, int i) { return d = T(int(d) * i); } \
inline T& operator/=(T& d, int i) { return d = T(int(d) / i); }
ENABLE_SAFE_OPERATORS_ON(Score) ENABLE_FULL_OPERATORS_ON(Value)
ENABLE_FULL_OPERATORS_ON(Depth)
ENABLE_FULL_OPERATORS_ON(Direction)
ENABLE_INCR_OPERATORS_ON(PieceType)
ENABLE_INCR_OPERATORS_ON(Piece)
ENABLE_INCR_OPERATORS_ON(Color)
ENABLE_INCR_OPERATORS_ON(Square)
ENABLE_INCR_OPERATORS_ON(File)
ENABLE_INCR_OPERATORS_ON(Rank)
ENABLE_BASE_OPERATORS_ON(Score)
#undef ENABLE_FULL_OPERATORS_ON
#undef ENABLE_INCR_OPERATORS_ON
#undef ENABLE_BASE_OPERATORS_ON
/// Additional operators to add integers to a Value
constexpr Value operator+(Value v, int i) { return Value(int(v) + i); }
constexpr Value operator-(Value v, int i) { return Value(int(v) - i); }
inline Value& operator+=(Value& v, int i) { return v = v + i; }
inline Value& operator-=(Value& v, int i) { return v = v - i; }
/// Additional operators to add a Direction to a Square
constexpr Square operator+(Square s, Direction d) { return Square(int(s) + int(d)); }
constexpr Square operator-(Square s, Direction d) { return Square(int(s) - int(d)); }
inline Square& operator+=(Square& s, Direction d) { return s = s + d; }
inline Square& operator-=(Square& s, Direction d) { return s = s - d; }
/// Only declared but not defined. We don't want to multiply two scores due to /// Only declared but not defined. We don't want to multiply two scores due to
/// a very high risk of overflow. So user should explicitly convert to integer. /// a very high risk of overflow. So user should explicitly convert to integer.
inline Score operator*(Score s1, Score s2); Score operator*(Score, Score) = delete;
/// Division of a Score must be handled separately for each term /// Division of a Score must be handled separately for each term
inline Score operator/(Score s, int i) { inline Score operator/(Score s, int i) {
return make_score(mg_value(s) / i, eg_value(s) / i); return make_score(mg_value(s) / i, eg_value(s) / i);
} }
#undef ENABLE_OPERATORS_ON /// Multiplication of a Score by an integer. We check for overflow in debug mode.
#undef ENABLE_SAFE_OPERATORS_ON inline Score operator*(Score s, int i) {
extern Value PieceValue[PHASE_NB][PIECE_NB]; Score result = Score(int(s) * i);
struct ExtMove { assert(eg_value(result) == (i * eg_value(s)));
Move move; assert(mg_value(result) == (i * mg_value(s)));
int score; assert((i == 0) || (result / i) == s);
};
inline bool operator<(const ExtMove& f, const ExtMove& s) { return result;
return f.score < s.score;
} }
inline Color operator~(Color c) { constexpr Color operator~(Color c) {
return Color(c ^ 1); return Color(c ^ BLACK); // Toggle color
} }
inline Square operator~(Square s) { constexpr Square operator~(Square s) {
return Square(s ^ 56); // Vertical flip SQ_A1 -> SQ_A8 return Square(s ^ SQ_A8); // Vertical flip SQ_A1 -> SQ_A8
} }
inline Square operator|(File f, Rank r) { constexpr File operator~(File f) {
return Square((r << 3) | f); return File(f ^ FILE_H); // Horizontal flip FILE_A -> FILE_H
} }
inline Value mate_in(int ply) { constexpr Piece operator~(Piece pc) {
return Piece(pc ^ 8); // Swap color of piece B_KNIGHT -> W_KNIGHT
}
constexpr CastlingRight operator|(Color c, CastlingSide s) {
return CastlingRight(WHITE_OO << ((s == QUEEN_SIDE) + 2 * c));
}
constexpr Value mate_in(int ply) {
return VALUE_MATE - ply; return VALUE_MATE - ply;
} }
inline Value mated_in(int ply) { constexpr Value mated_in(int ply) {
return -VALUE_MATE + ply; return -VALUE_MATE + ply;
} }
inline Piece make_piece(Color c, PieceType pt) { constexpr Square make_square(File f, Rank r) {
return Piece((c << 3) | pt); return Square((r << 3) + f);
} }
inline CastleRight make_castle_right(Color c, CastlingSide s) { constexpr Piece make_piece(Color c, PieceType pt) {
return CastleRight(WHITE_OO << ((s == QUEEN_SIDE) + 2 * c)); return Piece((c << 3) + pt);
} }
inline PieceType type_of(Piece p) { constexpr PieceType type_of(Piece pc) {
return PieceType(p & 7); return PieceType(pc & 7);
} }
inline Color color_of(Piece p) { inline Color color_of(Piece pc) {
assert(p != NO_PIECE); assert(pc != NO_PIECE);
return Color(p >> 3); return Color(pc >> 3);
} }
inline bool is_ok(Square s) { constexpr bool is_ok(Square s) {
return s >= SQ_A1 && s <= SQ_H8; return s >= SQ_A1 && s <= SQ_H8;
} }
inline File file_of(Square s) { constexpr File file_of(Square s) {
return File(s & 7); return File(s & 7);
} }
inline Rank rank_of(Square s) { constexpr Rank rank_of(Square s) {
return Rank(s >> 3); return Rank(s >> 3);
} }
inline Square mirror(Square s) { constexpr Square relative_square(Color c, Square s) {
return Square(s ^ 7); // Horizontal flip SQ_A1 -> SQ_H1
}
inline Square relative_square(Color c, Square s) {
return Square(s ^ (c * 56)); return Square(s ^ (c * 56));
} }
inline Rank relative_rank(Color c, Rank r) { constexpr Rank relative_rank(Color c, Rank r) {
return Rank(r ^ (c * 7)); return Rank(r ^ (c * 7));
} }
inline Rank relative_rank(Color c, Square s) { constexpr Rank relative_rank(Color c, Square s) {
return relative_rank(c, rank_of(s)); return relative_rank(c, rank_of(s));
} }
@@ -392,52 +418,41 @@ inline bool opposite_colors(Square s1, Square s2) {
return ((s >> 3) ^ s) & 1; return ((s >> 3) ^ s) & 1;
} }
inline char file_to_char(File f, bool tolower = true) { constexpr Direction pawn_push(Color c) {
return char(f - FILE_A + (tolower ? 'a' : 'A')); return c == WHITE ? NORTH : SOUTH;
} }
inline char rank_to_char(Rank r) { constexpr Square from_sq(Move m) {
return char(r - RANK_1 + '1');
}
inline Square pawn_push(Color c) {
return c == WHITE ? DELTA_N : DELTA_S;
}
inline Square from_sq(Move m) {
return Square((m >> 6) & 0x3F); return Square((m >> 6) & 0x3F);
} }
inline Square to_sq(Move m) { constexpr Square to_sq(Move m) {
return Square(m & 0x3F); return Square(m & 0x3F);
} }
inline MoveType type_of(Move m) { constexpr int from_to(Move m) {
return m & 0xFFF;
}
constexpr MoveType type_of(Move m) {
return MoveType(m & (3 << 14)); return MoveType(m & (3 << 14));
} }
inline PieceType promotion_type(Move m) { constexpr PieceType promotion_type(Move m) {
return PieceType(((m >> 12) & 3) + 2); return PieceType(((m >> 12) & 3) + KNIGHT);
} }
inline Move make_move(Square from, Square to) { constexpr Move make_move(Square from, Square to) {
return Move(to | (from << 6)); return Move((from << 6) + to);
} }
template<MoveType T> template<MoveType T>
inline Move make(Square from, Square to, PieceType pt = KNIGHT) { constexpr Move make(Square from, Square to, PieceType pt = KNIGHT) {
return Move(to | (from << 6) | T | ((pt - KNIGHT) << 12)); return Move(T + ((pt - KNIGHT) << 12) + (from << 6) + to);
} }
inline bool is_ok(Move m) { constexpr bool is_ok(Move m) {
return from_sq(m) != to_sq(m); // Catches also MOVE_NULL and MOVE_NONE return from_sq(m) != to_sq(m); // Catch MOVE_NULL and MOVE_NONE
}
#include <string>
inline const std::string square_to_string(Square s) {
char ch[] = { file_to_char(file_of(s)), rank_to_char(rank_of(s)), 0 };
return ch;
} }
#endif // #ifndef TYPES_H_INCLUDED #endif // #ifndef TYPES_H_INCLUDED
+206 -109
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,121 +18,37 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#include <iomanip> #include <cassert>
#include <iostream> #include <iostream>
#include <sstream> #include <sstream>
#include <string> #include <string>
#include "evaluate.h" #include "evaluate.h"
#include "notation.h" #include "movegen.h"
#include "position.h" #include "position.h"
#include "search.h" #include "search.h"
#include "thread.h" #include "thread.h"
#include "ucioption.h" #include "timeman.h"
#include "tt.h"
#include "uci.h"
#include "syzygy/tbprobe.h"
using namespace std; using namespace std;
extern void benchmark(const Position& pos, istream& is); extern vector<string> setup_bench(const Position&, istream&);
namespace { namespace {
// FEN string of the initial position, normal chess // FEN string of the initial position, normal chess
const char* StartFEN = "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"; const char* StartFEN = "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1";
// Keep track of position keys along the setup moves (from start position to the
// position just before to start searching). Needed by repetition draw detection.
Search::StateStackPtr SetupStates;
void setoption(istringstream& up);
void position(Position& pos, istringstream& up);
void go(const Position& pos, istringstream& up);
}
/// Wait for a command from the user, parse this text string as an UCI command,
/// and call the appropriate functions. Also intercepts EOF from stdin to ensure
/// that we exit gracefully if the GUI dies unexpectedly. In addition to the UCI
/// commands, the function also supports a few debug commands.
void UCI::loop(const string& args) {
Position pos(StartFEN, false, Threads.main()); // The root position
string token, cmd = args;
do {
if (args.empty() && !getline(cin, cmd)) // Block here waiting for input
cmd = "quit";
istringstream is(cmd);
is >> skipws >> token;
if (token == "quit" || token == "stop" || token == "ponderhit")
{
// GUI sends 'ponderhit' to tell us to ponder on the same move the
// opponent has played. In case Signals.stopOnPonderhit is set we are
// waiting for 'ponderhit' to stop the search (for instance because we
// already ran out of time), otherwise we should continue searching but
// switching from pondering to normal search.
if (token != "ponderhit" || Search::Signals.stopOnPonderhit)
{
Search::Signals.stop = true;
Threads.main()->notify_one(); // Could be sleeping
}
else
Search::Limits.ponder = false;
}
else if (token == "perft" && (is >> token)) // Read perft depth
{
stringstream ss;
ss << Options["Hash"] << " "
<< Options["Threads"] << " " << token << " current perft";
benchmark(pos, ss);
}
else if (token == "key")
sync_cout << hex << uppercase << setfill('0')
<< "position key: " << setw(16) << pos.key()
<< "\nmaterial key: " << setw(16) << pos.material_key()
<< "\npawn key: " << setw(16) << pos.pawn_key()
<< dec << sync_endl;
else if (token == "uci")
sync_cout << "id name " << engine_info(true)
<< "\n" << Options
<< "\nuciok" << sync_endl;
else if (token == "eval")
{
Search::RootColor = pos.side_to_move(); // Ensure it is set
sync_cout << Eval::trace(pos) << sync_endl;
}
else if (token == "ucinewgame") { /* Avoid returning "Unknown command" */ }
else if (token == "go") go(pos, is);
else if (token == "position") position(pos, is);
else if (token == "setoption") setoption(is);
else if (token == "flip") pos.flip();
else if (token == "bench") benchmark(pos, is);
else if (token == "d") sync_cout << pos.pretty() << sync_endl;
else if (token == "isready") sync_cout << "readyok" << sync_endl;
else
sync_cout << "Unknown command: " << cmd << sync_endl;
} while (token != "quit" && args.empty()); // Args have one-shot behaviour
Threads.wait_for_think_finished(); // Cannot quit while search is running
}
namespace {
// position() is called when engine receives the "position" UCI command. // position() is called when engine receives the "position" UCI command.
// The function sets up the position described in the given fen string ("fen") // The function sets up the position described in the given FEN string ("fen")
// or the starting position ("startpos") and then makes the moves given in the // or the starting position ("startpos") and then makes the moves given in the
// following move list ("moves"). // following move list ("moves").
void position(Position& pos, istringstream& is) { void position(Position& pos, istringstream& is, StateListPtr& states) {
Move m; Move m;
string token, fen; string token, fen;
@@ -149,14 +66,14 @@ namespace {
else else
return; return;
pos.set(fen, Options["UCI_Chess960"], Threads.main()); states = StateListPtr(new std::deque<StateInfo>(1)); // Drop old and create a new one
SetupStates = Search::StateStackPtr(new std::stack<StateInfo>()); pos.set(fen, Options["UCI_Chess960"], &states->back(), Threads.main());
// Parse move list (if any) // Parse move list (if any)
while (is >> token && (m = move_from_uci(pos, token)) != MOVE_NONE) while (is >> token && (m = UCI::to_move(pos, token)) != MOVE_NONE)
{ {
SetupStates->push(StateInfo()); states->emplace_back();
pos.do_move(m, SetupStates->top()); pos.do_move(m, states->back());
} }
} }
@@ -172,11 +89,11 @@ namespace {
// Read option name (can contain spaces) // Read option name (can contain spaces)
while (is >> token && token != "value") while (is >> token && token != "value")
name += string(" ", !name.empty()) + token; name += (name.empty() ? "" : " ") + token;
// Read option value (can contain spaces) // Read option value (can contain spaces)
while (is >> token) while (is >> token)
value += string(" ", !value.empty()) + token; value += (value.empty() ? "" : " ") + token;
if (Options.count(name)) if (Options.count(name))
Options[name] = value; Options[name] = value;
@@ -186,20 +103,21 @@ namespace {
// go() is called when engine receives the "go" UCI command. The function sets // go() is called when engine receives the "go" UCI command. The function sets
// the thinking time and other parameters from the input string, and starts // the thinking time and other parameters from the input string, then starts
// the search. // the search.
void go(const Position& pos, istringstream& is) { void go(Position& pos, istringstream& is, StateListPtr& states) {
Search::LimitsType limits; Search::LimitsType limits;
vector<Move> searchMoves;
string token; string token;
bool ponderMode = false;
limits.startTime = now(); // As early as possible!
while (is >> token) while (is >> token)
{
if (token == "searchmoves") if (token == "searchmoves")
while (is >> token) while (is >> token)
searchMoves.push_back(move_from_uci(pos, token)); limits.searchmoves.push_back(UCI::to_move(pos, token));
else if (token == "wtime") is >> limits.time[WHITE]; else if (token == "wtime") is >> limits.time[WHITE];
else if (token == "btime") is >> limits.time[BLACK]; else if (token == "btime") is >> limits.time[BLACK];
@@ -210,10 +128,189 @@ namespace {
else if (token == "nodes") is >> limits.nodes; else if (token == "nodes") is >> limits.nodes;
else if (token == "movetime") is >> limits.movetime; else if (token == "movetime") is >> limits.movetime;
else if (token == "mate") is >> limits.mate; else if (token == "mate") is >> limits.mate;
else if (token == "infinite") limits.infinite = true; else if (token == "perft") is >> limits.perft;
else if (token == "ponder") limits.ponder = true; else if (token == "infinite") limits.infinite = 1;
else if (token == "ponder") ponderMode = true;
Threads.start_thinking(pos, states, limits, ponderMode);
} }
Threads.start_thinking(pos, limits, searchMoves, SetupStates);
// bench() is called when engine receives the "bench" command. Firstly
// a list of UCI commands is setup according to bench parameters, then
// it is run one by one printing a summary at the end.
void bench(Position& pos, istream& args, StateListPtr& states) {
string token;
uint64_t num, nodes = 0, cnt = 1;
vector<string> list = setup_bench(pos, args);
num = count_if(list.begin(), list.end(), [](string s) { return s.find("go ") == 0; });
TimePoint elapsed = now();
for (const auto& cmd : list)
{
istringstream is(cmd);
is >> skipws >> token;
if (token == "go")
{
cerr << "\nPosition: " << cnt++ << '/' << num << endl;
go(pos, is, states);
Threads.main()->wait_for_search_finished();
nodes += Threads.nodes_searched();
} }
else if (token == "setoption") setoption(is);
else if (token == "position") position(pos, is, states);
else if (token == "ucinewgame") Search::clear();
}
elapsed = now() - elapsed + 1; // Ensure positivity to avoid a 'divide by zero'
dbg_print(); // Just before exiting
cerr << "\n==========================="
<< "\nTotal time (ms) : " << elapsed
<< "\nNodes searched : " << nodes
<< "\nNodes/second : " << 1000 * nodes / elapsed << endl;
}
} // namespace
/// UCI::loop() waits for a command from stdin, parses it and calls the appropriate
/// function. Also intercepts EOF from stdin to ensure gracefully exiting if the
/// GUI dies unexpectedly. When called with some command line arguments, e.g. to
/// run 'bench', once the command is executed the function returns immediately.
/// In addition to the UCI ones, also some additional debug commands are supported.
void UCI::loop(int argc, char* argv[]) {
Position pos;
string token, cmd;
StateListPtr states(new std::deque<StateInfo>(1));
auto uiThread = std::make_shared<Thread>(0);
pos.set(StartFEN, false, &states->back(), uiThread.get());
for (int i = 1; i < argc; ++i)
cmd += std::string(argv[i]) + " ";
do {
if (argc == 1 && !getline(cin, cmd)) // Block here waiting for input or EOF
cmd = "quit";
istringstream is(cmd);
token.clear(); // Avoid a stale if getline() returns empty or blank line
is >> skipws >> token;
// The GUI sends 'ponderhit' to tell us the user has played the expected move.
// So 'ponderhit' will be sent if we were told to ponder on the same move the
// user has played. We should continue searching but switch from pondering to
// normal search. In case Threads.stopOnPonderhit is set we are waiting for
// 'ponderhit' to stop the search, for instance if max search depth is reached.
if ( token == "quit"
|| token == "stop"
|| (token == "ponderhit" && Threads.stopOnPonderhit))
Threads.stop = true;
else if (token == "ponderhit")
Threads.ponder = false; // Switch to normal search
else if (token == "uci")
sync_cout << "id name " << engine_info(true)
<< "\n" << Options
<< "\nuciok" << sync_endl;
else if (token == "setoption") setoption(is);
else if (token == "go") go(pos, is, states);
else if (token == "position") position(pos, is, states);
else if (token == "ucinewgame") Search::clear();
else if (token == "isready") sync_cout << "readyok" << sync_endl;
// Additional custom non-UCI commands, mainly for debugging
else if (token == "flip") pos.flip();
else if (token == "bench") bench(pos, is, states);
else if (token == "d") sync_cout << pos << sync_endl;
else if (token == "eval") sync_cout << Eval::trace(pos) << sync_endl;
else
sync_cout << "Unknown command: " << cmd << sync_endl;
} while (token != "quit" && argc == 1); // Command line args are one-shot
}
/// UCI::value() converts a Value to a string suitable for use with the UCI
/// protocol specification:
///
/// cp <x> The score from the engine's point of view in centipawns.
/// mate <y> Mate in y moves, not plies. If the engine is getting mated
/// use negative values for y.
string UCI::value(Value v) {
assert(-VALUE_INFINITE < v && v < VALUE_INFINITE);
stringstream ss;
if (abs(v) < VALUE_MATE - MAX_PLY)
ss << "cp " << v * 100 / PawnValueEg;
else
ss << "mate " << (v > 0 ? VALUE_MATE - v + 1 : -VALUE_MATE - v) / 2;
return ss.str();
}
/// UCI::square() converts a Square to a string in algebraic notation (g1, a7, etc.)
std::string UCI::square(Square s) {
return std::string{ char('a' + file_of(s)), char('1' + rank_of(s)) };
}
/// UCI::move() converts a Move to a string in coordinate notation (g1f3, a7a8q).
/// The only special case is castling, where we print in the e1g1 notation in
/// normal chess mode, and in e1h1 notation in chess960 mode. Internally all
/// castling moves are always encoded as 'king captures rook'.
string UCI::move(Move m, bool chess960) {
Square from = from_sq(m);
Square to = to_sq(m);
if (m == MOVE_NONE)
return "(none)";
if (m == MOVE_NULL)
return "0000";
if (type_of(m) == CASTLING && !chess960)
to = make_square(to > from ? FILE_G : FILE_C, rank_of(from));
string move = UCI::square(from) + UCI::square(to);
if (type_of(m) == PROMOTION)
move += " pnbrqk"[promotion_type(m)];
return move;
}
/// UCI::to_move() converts a string representing a move in coordinate notation
/// (g1f3, a7a8q) to the corresponding legal Move, if any.
Move UCI::to_move(const Position& pos, string& str) {
if (str.length() == 5) // Junior could send promotion piece in uppercase
str[4] = char(tolower(str[4]));
for (const auto& m : MoveList<LEGAL>(pos))
if (str == UCI::move(m, pos.is_chess960()))
return m;
return MOVE_NONE;
} }
+26 -13
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -17,12 +18,16 @@
along with this program. If not, see <http://www.gnu.org/licenses/>. along with this program. If not, see <http://www.gnu.org/licenses/>.
*/ */
#ifndef UCIOPTION_H_INCLUDED #ifndef UCI_H_INCLUDED
#define UCIOPTION_H_INCLUDED #define UCI_H_INCLUDED
#include <map> #include <map>
#include <string> #include <string>
#include "types.h"
class Position;
namespace UCI { namespace UCI {
class Option; class Option;
@@ -38,17 +43,20 @@ typedef std::map<std::string, Option, CaseInsensitiveLess> OptionsMap;
/// Option class implements an option as defined by UCI protocol /// Option class implements an option as defined by UCI protocol
class Option { class Option {
typedef void (Fn)(const Option&); typedef void (*OnChange)(const Option&);
public: public:
Option(Fn* = NULL); Option(OnChange = nullptr);
Option(bool v, Fn* = NULL); Option(bool v, OnChange = nullptr);
Option(const char* v, Fn* = NULL); Option(const char* v, OnChange = nullptr);
Option(int v, int min, int max, Fn* = NULL); Option(double v, int minv, int maxv, OnChange = nullptr);
Option(const char* v, const char* cur, OnChange = nullptr);
Option& operator=(const std::string& v); Option& operator=(const std::string&);
operator int() const; void operator<<(const Option&);
operator double() const;
operator std::string() const; operator std::string() const;
bool operator==(const char*) const;
private: private:
friend std::ostream& operator<<(std::ostream&, const OptionsMap&); friend std::ostream& operator<<(std::ostream&, const OptionsMap&);
@@ -56,14 +64,19 @@ private:
std::string defaultValue, currentValue, type; std::string defaultValue, currentValue, type;
int min, max; int min, max;
size_t idx; size_t idx;
Fn* on_change; OnChange on_change;
}; };
void init(OptionsMap&); void init(OptionsMap&);
void loop(const std::string&); void loop(int argc, char* argv[]);
std::string value(Value v);
std::string square(Square s);
std::string move(Move m, bool chess960);
std::string pv(const Position& pos, Depth depth, Value alpha, Value beta);
Move to_move(const Position& pos, std::string& str);
} // namespace UCI } // namespace UCI
extern UCI::OptionsMap Options; extern UCI::OptionsMap Options;
#endif // #ifndef UCIOPTION_H_INCLUDED #endif // #ifndef UCI_H_INCLUDED
+75 -62
View File
@@ -1,7 +1,8 @@
/* /*
Stockfish, a UCI chess playing engine derived from Glaurung 2.1 Stockfish, a UCI chess playing engine derived from Glaurung 2.1
Copyright (C) 2004-2008 Tord Romstad (Glaurung author) Copyright (C) 2004-2008 Tord Romstad (Glaurung author)
Copyright (C) 2008-2013 Marco Costalba, Joona Kiiski, Tord Romstad Copyright (C) 2008-2015 Marco Costalba, Joona Kiiski, Tord Romstad
Copyright (C) 2015-2019 Marco Costalba, Joona Kiiski, Gary Linscott, Tord Romstad
Stockfish is free software: you can redistribute it and/or modify Stockfish is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by it under the terms of the GNU General Public License as published by
@@ -19,14 +20,14 @@
#include <algorithm> #include <algorithm>
#include <cassert> #include <cassert>
#include <cstdlib> #include <ostream>
#include <sstream>
#include "evaluate.h"
#include "misc.h" #include "misc.h"
#include "search.h"
#include "thread.h" #include "thread.h"
#include "tt.h" #include "tt.h"
#include "ucioption.h" #include "uci.h"
#include "syzygy/tbprobe.h"
using std::string; using std::string;
@@ -35,57 +36,47 @@ UCI::OptionsMap Options; // Global object
namespace UCI { namespace UCI {
/// 'On change' actions, triggered by an option's value change /// 'On change' actions, triggered by an option's value change
void on_clear_hash(const Option&) { Search::clear(); }
void on_hash_size(const Option& o) { TT.resize(o); }
void on_logger(const Option& o) { start_logger(o); } void on_logger(const Option& o) { start_logger(o); }
void on_eval(const Option&) { Eval::init(); } void on_threads(const Option& o) { Threads.set(o); }
void on_threads(const Option&) { Threads.read_uci_options(); } void on_tb_path(const Option& o) { Tablebases::init(o); }
void on_hash_size(const Option& o) { TT.set_size(o); }
void on_clear_hash(const Option&) { TT.clear(); }
/// Our case insensitive less() function as required by UCI protocol /// Our case insensitive less() function as required by UCI protocol
bool ci_less(char c1, char c2) { return tolower(c1) < tolower(c2); }
bool CaseInsensitiveLess::operator() (const string& s1, const string& s2) const { bool CaseInsensitiveLess::operator() (const string& s1, const string& s2) const {
return std::lexicographical_compare(s1.begin(), s1.end(), s2.begin(), s2.end(), ci_less);
return std::lexicographical_compare(s1.begin(), s1.end(), s2.begin(), s2.end(),
[](char c1, char c2) { return tolower(c1) < tolower(c2); });
} }
/// init() initializes the UCI options to their hard coded default values /// init() initializes the UCI options to their hard-coded default values
void init(OptionsMap& o) { void init(OptionsMap& o) {
o["Write Debug Log"] = Option(false, on_logger); // at most 2^32 clusters.
o["Write Search Log"] = Option(false); constexpr int MaxHashMB = Is64Bit ? 131072 : 2048;
o["Search Log Filename"] = Option("SearchLog.txt");
o["Book File"] = Option("book.bin"); o["Debug Log File"] << Option("", on_logger);
o["Best Book Move"] = Option(false); o["Contempt"] << Option(24, -100, 100);
o["Contempt Factor"] = Option(0, -50, 50); o["Analysis Contempt"] << Option("Both var Off var White var Black var Both", "Both");
o["Mobility (Midgame)"] = Option(100, 0, 200, on_eval); o["Threads"] << Option(1, 1, 512, on_threads);
o["Mobility (Endgame)"] = Option(100, 0, 200, on_eval); o["Hash"] << Option(16, 1, MaxHashMB, on_hash_size);
o["Pawn Structure (Midgame)"] = Option(100, 0, 200, on_eval); o["Clear Hash"] << Option(on_clear_hash);
o["Pawn Structure (Endgame)"] = Option(100, 0, 200, on_eval); o["Ponder"] << Option(false);
o["Passed Pawns (Midgame)"] = Option(100, 0, 200, on_eval); o["MultiPV"] << Option(1, 1, 500);
o["Passed Pawns (Endgame)"] = Option(100, 0, 200, on_eval); o["Skill Level"] << Option(20, 0, 20);
o["Space"] = Option(100, 0, 200, on_eval); o["Move Overhead"] << Option(30, 0, 5000);
o["Aggressiveness"] = Option(100, 0, 200, on_eval); o["Minimum Thinking Time"] << Option(20, 0, 5000);
o["Cowardice"] = Option(100, 0, 200, on_eval); o["Slow Mover"] << Option(84, 10, 1000);
o["Min Split Depth"] = Option(0, 0, 12, on_threads); o["nodestime"] << Option(0, 0, 10000);
o["Max Threads per Split Point"] = Option(5, 4, 8, on_threads); o["UCI_Chess960"] << Option(false);
o["Threads"] = Option(1, 1, MAX_THREADS, on_threads); o["UCI_AnalyseMode"] << Option(false);
o["Idle Threads Sleep"] = Option(false); o["SyzygyPath"] << Option("<empty>", on_tb_path);
o["Hash"] = Option(32, 1, 8192, on_hash_size); o["SyzygyProbeDepth"] << Option(1, 1, 100);
o["Clear Hash"] = Option(on_clear_hash); o["Syzygy50MoveRule"] << Option(true);
o["Ponder"] = Option(true); o["SyzygyProbeLimit"] << Option(7, 0, 7);
o["OwnBook"] = Option(false);
o["MultiPV"] = Option(1, 1, 500);
o["Skill Level"] = Option(20, 0, 20);
o["Emergency Move Horizon"] = Option(40, 0, 50);
o["Emergency Base Time"] = Option(200, 0, 30000);
o["Emergency Move Time"] = Option(70, 0, 5000);
o["Minimum Thinking Time"] = Option(20, 0, 5000);
o["Slow Mover"] = Option(100, 10, 1000);
o["UCI_Chess960"] = Option(false);
o["UCI_AnalyseMode"] = Option(false, on_eval);
} }
@@ -94,43 +85,48 @@ void init(OptionsMap& o) {
std::ostream& operator<<(std::ostream& os, const OptionsMap& om) { std::ostream& operator<<(std::ostream& os, const OptionsMap& om) {
for (size_t idx = 0; idx < om.size(); idx++) for (size_t idx = 0; idx < om.size(); ++idx)
for (OptionsMap::const_iterator it = om.begin(); it != om.end(); ++it) for (const auto& it : om)
if (it->second.idx == idx) if (it.second.idx == idx)
{ {
const Option& o = it->second; const Option& o = it.second;
os << "\noption name " << it->first << " type " << o.type; os << "\noption name " << it.first << " type " << o.type;
if (o.type != "button") if (o.type == "string" || o.type == "check" || o.type == "combo")
os << " default " << o.defaultValue; os << " default " << o.defaultValue;
if (o.type == "spin") if (o.type == "spin")
os << " min " << o.min << " max " << o.max; os << " default " << int(stof(o.defaultValue))
<< " min " << o.min
<< " max " << o.max;
break; break;
} }
return os; return os;
} }
/// Option c'tors and conversion operators /// Option class constructors and conversion operators
Option::Option(const char* v, Fn* f) : type("string"), min(0), max(0), idx(Options.size()), on_change(f) Option::Option(const char* v, OnChange f) : type("string"), min(0), max(0), on_change(f)
{ defaultValue = currentValue = v; } { defaultValue = currentValue = v; }
Option::Option(bool v, Fn* f) : type("check"), min(0), max(0), idx(Options.size()), on_change(f) Option::Option(bool v, OnChange f) : type("check"), min(0), max(0), on_change(f)
{ defaultValue = currentValue = (v ? "true" : "false"); } { defaultValue = currentValue = (v ? "true" : "false"); }
Option::Option(Fn* f) : type("button"), min(0), max(0), idx(Options.size()), on_change(f) Option::Option(OnChange f) : type("button"), min(0), max(0), on_change(f)
{} {}
Option::Option(int v, int minv, int maxv, Fn* f) : type("spin"), min(minv), max(maxv), idx(Options.size()), on_change(f) Option::Option(double v, int minv, int maxv, OnChange f) : type("spin"), min(minv), max(maxv), on_change(f)
{ std::ostringstream ss; ss << v; defaultValue = currentValue = ss.str(); } { defaultValue = currentValue = std::to_string(v); }
Option::Option(const char* v, const char* cur, OnChange f) : type("combo"), min(0), max(0), on_change(f)
{ defaultValue = v; currentValue = cur; }
Option::operator int() const { Option::operator double() const {
assert(type == "check" || type == "spin"); assert(type == "check" || type == "spin");
return (type == "spin" ? atoi(currentValue.c_str()) : currentValue == "true"); return (type == "spin" ? stof(currentValue) : currentValue == "true");
} }
Option::operator std::string() const { Option::operator std::string() const {
@@ -138,6 +134,23 @@ Option::operator std::string() const {
return currentValue; return currentValue;
} }
bool Option::operator==(const char* s) const {
assert(type == "combo");
return !CaseInsensitiveLess()(currentValue, s)
&& !CaseInsensitiveLess()(s, currentValue);
}
/// operator<<() inits options and assigns idx in the correct printing order
void Option::operator<<(const Option& o) {
static size_t insert_order = 0;
*this = o;
idx = insert_order++;
}
/// operator=() updates currentValue and triggers on_change() action. It's up to /// operator=() updates currentValue and triggers on_change() action. It's up to
/// the GUI to check for option's limits, but we could receive the new value from /// the GUI to check for option's limits, but we could receive the new value from
@@ -149,14 +162,14 @@ Option& Option::operator=(const string& v) {
if ( (type != "button" && v.empty()) if ( (type != "button" && v.empty())
|| (type == "check" && v != "true" && v != "false") || (type == "check" && v != "true" && v != "false")
|| (type == "spin" && (atoi(v.c_str()) < min || atoi(v.c_str()) > max))) || (type == "spin" && (stof(v) < min || stof(v) > max)))
return *this; return *this;
if (type != "button") if (type != "button")
currentValue = v; currentValue = v;
if (on_change) if (on_change)
(*on_change)(*this); on_change(*this);
return *this; return *this;
} }
+144
View File
@@ -0,0 +1,144 @@
#!/bin/bash
# check for errors under valgrind or sanitizers.
error()
{
echo "instrumented testing failed on line $1"
exit 1
}
trap 'error ${LINENO}' ERR
# define suitable post and prefixes for testing options
case $1 in
--valgrind)
echo "valgrind testing started"
prefix=''
exeprefix='valgrind --error-exitcode=42'
postfix='1>/dev/null'
threads="1"
;;
--valgrind-thread)
echo "valgrind-thread testing started"
prefix=''
exeprefix='valgrind --error-exitcode=42'
postfix='1>/dev/null'
threads="2"
;;
--sanitizer-undefined)
echo "sanitizer-undefined testing started"
prefix='!'
exeprefix=''
postfix='2>&1 | grep "runtime error:"'
threads="1"
;;
--sanitizer-thread)
echo "sanitizer-thread testing started"
prefix='!'
exeprefix=''
postfix='2>&1 | grep "WARNING: ThreadSanitizer:"'
threads="2"
cat << EOF > tsan.supp
race:TTEntry::move
race:TTEntry::depth
race:TTEntry::bound
race:TTEntry::save
race:TTEntry::value
race:TTEntry::eval
race:TranspositionTable::probe
race:TranspositionTable::hashfull
EOF
export TSAN_OPTIONS="suppressions=./tsan.supp"
;;
*)
echo "unknown testing started"
prefix=''
exeprefix=''
postfix=''
threads="1"
;;
esac
# simple command line testing
for args in "eval" \
"go nodes 1000" \
"go depth 10" \
"go movetime 1000" \
"go wtime 8000 btime 8000 winc 500 binc 500" \
"bench 128 $threads 10 default depth"
do
echo "$prefix $exeprefix ./stockfish $args $postfix"
eval "$prefix $exeprefix ./stockfish $args $postfix"
done
# more general testing, following an uci protocol exchange
cat << EOF > game.exp
set timeout 10
spawn $exeprefix ./stockfish
send "uci\n"
expect "uciok"
send "setoption name Threads value $threads\n"
send "ucinewgame\n"
send "position startpos\n"
send "go nodes 1000\n"
expect "bestmove"
send "position startpos moves e2e4 e7e6\n"
send "go nodes 1000\n"
expect "bestmove"
send "position fen 5rk1/1K4p1/8/8/3B4/8/8/8 b - - 0 1\n"
send "go depth 30\n"
expect "bestmove"
send "quit\n"
expect eof
# return error code of the spawned program, useful for valgrind
lassign [wait] pid spawnid os_error_flag value
exit \$value
EOF
#download TB as needed
if [ ! -d ../tests/syzygy ]; then
curl -sL https://api.github.com/repos/niklasf/python-chess/tarball/9b9aa13f9f36d08aadfabff872882f4ab1494e95 | tar -xzf -
mv niklasf-python-chess-9b9aa13 ../tests/syzygy
fi
cat << EOF > syzygy.exp
set timeout 240
spawn $exeprefix ./stockfish
send "uci\n"
send "setoption name SyzygyPath value ../tests/syzygy/\n"
expect "info string Found 35 tablebases" {} timeout {exit 1}
send "bench 128 1 10 default depth\n"
send "quit\n"
expect eof
# return error code of the spawned program, useful for valgrind
lassign [wait] pid spawnid os_error_flag value
exit \$value
EOF
for exp in game.exp syzygy.exp
do
echo "$prefix expect $exp $postfix"
eval "$prefix expect $exp $postfix"
rm $exp
done
rm -f tsan.supp
echo "instrumented testing OK"
Executable
+32
View File
@@ -0,0 +1,32 @@
#!/bin/bash
# verify perft numbers (positions from https://chessprogramming.wikispaces.com/Perft+Results)
error()
{
echo "perft testing failed on line $1"
exit 1
}
trap 'error ${LINENO}' ERR
echo "perft testing started"
cat << EOF > perft.exp
set timeout 10
lassign \$argv pos depth result
spawn ./stockfish
send "position \$pos\\ngo perft \$depth\\n"
expect "Nodes searched? \$result" {} timeout {exit 1}
send "quit\\n"
expect eof
EOF
expect perft.exp startpos 5 4865609 > /dev/null
expect perft.exp "fen r3k2r/p1ppqpb1/bn2pnp1/3PN3/1p2P3/2N2Q1p/PPPBBPPP/R3K2R w KQkq -" 5 193690690 > /dev/null
expect perft.exp "fen 8/2p5/3p4/KP5r/1R3p1k/8/4P1P1/8 w - -" 6 11030083 > /dev/null
expect perft.exp "fen r3k2r/Pppp1ppp/1b3nbN/nP6/BBP1P3/q4N2/Pp1P2PP/R2Q1RK1 w kq - 0 1" 5 15833292 > /dev/null
expect perft.exp "fen rnbq1k1r/pp1Pbppp/2p5/8/2B5/8/PPP1NnPP/RNBQK2R w KQ - 1 8" 5 89941194 > /dev/null
expect perft.exp "fen r4rk1/1pp1qppp/p1np1n2/2b1p1B1/2B1P1b1/P1NP1N2/1PP1QPPP/R4RK1 w - - 0 10" 5 164075551 > /dev/null
rm perft.exp
echo "perft testing OK"
+61
View File
@@ -0,0 +1,61 @@
#!/bin/bash
# verify reproducible search
error()
{
echo "reprosearch testing failed on line $1"
exit 1
}
trap 'error ${LINENO}' ERR
echo "reprosearch testing started"
# repeat two short games, separated by ucinewgame.
# with go nodes $nodes they should result in exactly
# the same node count for each iteration.
cat << EOF > repeat.exp
set timeout 10
spawn ./stockfish
lassign \$argv nodes
send "uci\n"
expect "uciok"
send "ucinewgame\n"
send "position startpos\n"
send "go nodes \$nodes\n"
expect "bestmove"
send "position startpos moves e2e4 e7e6\n"
send "go nodes \$nodes\n"
expect "bestmove"
send "ucinewgame\n"
send "position startpos\n"
send "go nodes \$nodes\n"
expect "bestmove"
send "position startpos moves e2e4 e7e6\n"
send "go nodes \$nodes\n"
expect "bestmove"
send "quit\n"
expect eof
EOF
# to increase the likelyhood of finding a non-reproducible case,
# the allowed number of nodes are varied systematically
for i in `seq 1 20`
do
nodes=$((100*3**i/2**i))
echo "reprosearch testing with $nodes nodes"
# each line should appear exactly an even number of times
expect repeat.exp $nodes 2>&1 | grep -o "nodes [0-9]*" | sort | uniq -c | awk '{if ($1%2!=0) exit(1)}'
done
rm repeat.exp
echo "reprosearch testing OK"
+31
View File
@@ -0,0 +1,31 @@
#!/bin/bash
# obtain and optionally verify Bench / signature
# if no reference is given, the output is deliberately limited to just the signature
error()
{
echo "running bench for signature failed on line $1"
exit 1
}
trap 'error ${LINENO}' ERR
# obtain
signature=`./stockfish bench 2>&1 | grep "Nodes searched : " | awk '{print $4}'`
if [ $# -gt 0 ]; then
# compare to given reference
if [ "$1" != "$signature" ]; then
if [ -z "$signature" ]; then
echo "No signature obtained from bench. Code crashed or assert triggered ?"
else
echo "signature mismatch: reference $1 obtained: $signature ."
fi
exit 1
else
echo "signature OK: $signature"
fi
else
# just report signature
echo $signature
fi