From e8caa6640df15b1823d5d4b94e759d52923769e6 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Thu, 8 Dec 2022 20:33:32 +0100
Subject: [PATCH 001/192] Restore development version

No functional change
---
 src/misc.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index 2d86969f..c7fa0e9a 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -68,7 +68,7 @@ namespace Stockfish {
 namespace {
 
 /// Version number or dev.
-const string version = "15.1";
+const string version = "dev";
 
 /// 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

From 9fc203a3d05262ac74bfca8b4618155e18d76003 Mon Sep 17 00:00:00 2001
From: ppigazzini <pasquale.pigazzini@gmail.com>
Date: Thu, 8 Dec 2022 18:32:30 +0100
Subject: [PATCH 002/192] Set the right PATH for ARM compiler and build tests
 in CI

Fix for the GitHub upgrade:
https://github.com/actions/runner-images/issues/5879
that broke our ARM workflows because it changed the value of
the ANDROID_NDK_HOME variable referenced in our PATH.

closes https://github.com/official-stockfish/Stockfish/pull/4267

No functional change
---
 .github/workflows/stockfish_test.yml | 43 ++++++++++++----------------
 1 file changed, 19 insertions(+), 24 deletions(-)

diff --git a/.github/workflows/stockfish_test.yml b/.github/workflows/stockfish_test.yml
index 46b4e26f..e4e6205f 100644
--- a/.github/workflows/stockfish_test.yml
+++ b/.github/workflows/stockfish_test.yml
@@ -35,7 +35,7 @@ jobs:
               os: ubuntu-20.04,
               compiler: aarch64-linux-android21-clang++,
               comp: ndk,
-              run_armv8_tests: false,
+              run_armv8_tests: true,
               shell: 'bash {0}'
             }
           - {
@@ -43,7 +43,7 @@ jobs:
               os: ubuntu-20.04,
               compiler: armv7a-linux-androideabi21-clang++,
               comp: ndk,
-              run_armv7_tests: false,
+              run_armv7_tests: true,
               shell: 'bash {0}'
             }
           - {
@@ -92,15 +92,6 @@ jobs:
               msys_env: 'clang-x86_64-clang',
               shell: 'msys2 {0}'
             }
-        exclude:
-          - config: 
-              {
-                name: "Ubuntu 20.04 NDK armv7"
-              }
-          - config:
-              {
-                name: "Ubuntu 20.04 NDK armv8"
-              }
     defaults:
       run:
         working-directory: src
@@ -120,8 +111,8 @@ jobs:
         if: runner.os == 'Windows'
         uses: msys2/setup-msys2@v2
         with:
-          msystem: ${{matrix.config.msys_sys}}
-          install: mingw-w64-${{matrix.config.msys_env}} make git expect
+          msystem: ${{ matrix.config.msys_sys }}
+          install: mingw-w64-${{ matrix.config.msys_env }} make git expect
 
       - name: Download the used network from the fishtest framework
         run: |
@@ -134,7 +125,14 @@ jobs:
 
       - name: Check compiler
         run: |
-          export PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin
+          if [ $COMP == ndk ]; then
+            ANDROID_ROOT=/usr/local/lib/android
+            ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
+            SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
+            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
+            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          fi
           $COMPILER -v
 
       - name: Test help target
@@ -233,9 +231,8 @@ jobs:
           ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
           SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=${ANDROID_SDK_ROOT}/ndk-bundle
-          ln -sfn $ANDROID_SDK_ROOT/ndk/21.4.7075529 $ANDROID_NDK_ROOT
-          export PATH=$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv8 build
@@ -250,9 +247,8 @@ jobs:
           ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
           SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=${ANDROID_SDK_ROOT}/ndk-bundle
-          ln -sfn $ANDROID_SDK_ROOT/ndk/21.4.7075529 $ANDROID_NDK_ROOT
-          export PATH=$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv7 build
@@ -265,9 +261,8 @@ jobs:
           ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
           SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=${ANDROID_SDK_ROOT}/ndk-bundle
-          ln -sfn $ANDROID_SDK_ROOT/ndk/21.4.7075529 $ANDROID_NDK_ROOT
-          export PATH=$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv7-neon build
@@ -281,4 +276,4 @@ jobs:
           make clean
           make -j2 ARCH=x86-64-modern build
           ../tests/perft.sh
-          ../tests/reprosearch.sh
\ No newline at end of file
+          ../tests/reprosearch.sh

From 98965c139df1483a3d684ee8bc7a60dc4b95efa1 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Fri, 2 Dec 2022 18:23:28 +0300
Subject: [PATCH 003/192] doEvenDeeperSearch + tuning

Credit for the main idea of doEvenDeeperSearch goes to Vizvezdenec,
tuning by FauziAkram: Expansion of existing logic of doDeeperSearch -
if value from LMR is really really good do full depth search not
1 ply deeper but rather 2 instead.

Passed STC:
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 330048 W: 87672 L: 86942 D: 155434
Ptnml(0-2): 1012, 36739, 88912, 37229, 1132
https://tests.stockfishchess.org/tests/view/638a1cadd2b9c924c4c621d2

Passed LTC:
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 216696 W: 57891 L: 57240 D: 101565
Ptnml(0-2): 72, 21221, 65152, 21790, 113
https://tests.stockfishchess.org/tests/view/638c7d52a971f1f096c68fe2

closes https://github.com/official-stockfish/Stockfish/pull/4256

Bench: 3461830
---
 src/evaluate.cpp | 14 +++++++-------
 src/search.cpp   |  7 ++++---
 2 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 87412b81..93e665df 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   else
   {
       int nnueComplexity;
-      int scale = 1064 + 106 * pos.non_pawn_material() / 5120;
+      int scale = 1076 + 96 * pos.non_pawn_material() / 5120;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];
@@ -1071,21 +1071,21 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  416 * nnueComplexity
-                        + 424 * abs(psq - nnue)
+      nnueComplexity = (  412 * nnueComplexity
+                        + 428 * abs(psq - nnue)
                         + (optimism  > 0 ? int(optimism) * int(psq - nnue) : 0)
-                        ) / 1024;
+                        ) / 1026;
 
       // Return hybrid NNUE complexity to caller
       if (complexity)
           *complexity = nnueComplexity;
 
-      optimism = optimism * (269 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 754)) / 1024;
+      optimism = optimism * (278 + nnueComplexity) / 256;
+      v = (nnue * scale + optimism * (scale - 755)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling
-  v = v * (195 - pos.rule50_count()) / 211;
+  v = v * (197 - pos.rule50_count()) / 214;
 
   // Guarantee evaluation does not hit the tablebase range
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
diff --git a/src/search.cpp b/src/search.cpp
index c8163d1f..343c098a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min((12 * d + 282) * d - 349 , 1594);
+    return std::min((12 * d + 282) * d - 349 , 1480);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -1172,7 +1172,7 @@ moves_loop: // When in check, search starts here
                          - 4433;
 
           // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-          r -= ss->statScore / (13628 + 4000 * (depth > 7 && depth < 19));
+          r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
 
           // In general we want to cap the LMR depth search at newDepth, but when
           // reduction is negative, we allow this move a limited search extension
@@ -1187,9 +1187,10 @@ moves_loop: // When in check, search starts here
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
               const bool doDeeperSearch = value > (alpha + 64 + 11 * (newDepth - d));
+              const bool doEvenDeeperSearch = value > alpha + 582;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
-              newDepth += doDeeperSearch - doShallowerSearch;
+              newDepth += doDeeperSearch - doShallowerSearch + doEvenDeeperSearch;
 
               if (newDepth > d)
                   value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth, !cutNode);

From cb0c7a98485fbef4e5d6ed5f5b08201113ce0b4e Mon Sep 17 00:00:00 2001
From: Guenther Demetz <guenther.demetz@wuerth-phoenix.com>
Date: Tue, 6 Dec 2022 19:09:33 +0100
Subject: [PATCH 004/192] Correctly output lowerbound/upperbound scores

fixes the lowerbound/upperbound output by avoiding
scores outside the alpha,beta bracket. Since SF search
uses fail-soft we can't simply take the returned value
as score.

closes https://github.com/official-stockfish/Stockfish/pull/4259

No functional change
---
 src/search.cpp | 14 ++++++++++----
 src/search.h   |  1 +
 2 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 343c098a..04f73e1c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1245,10 +1245,16 @@ moves_loop: // When in check, search starts here
           // PV move or new best move?
           if (moveCount == 1 || value > alpha)
           {
-              rm.score = value;
+              rm.score =  rm.uciScore = value;
               rm.selDepth = thisThread->selDepth;
-              rm.scoreLowerbound = value >= beta;
-              rm.scoreUpperbound = value <= alpha;
+              if (value >= beta) {
+                 rm.scoreLowerbound = true;
+                 rm.uciScore = beta;
+              }
+              else if (value <= alpha) {
+                 rm.scoreUpperbound = true;
+                 rm.uciScore = alpha;
+              }
               rm.pv.resize(1);
 
               assert((ss+1)->pv);
@@ -1841,7 +1847,7 @@ string UCI::pv(const Position& pos, Depth depth) {
           continue;
 
       Depth d = updated ? depth : std::max(1, depth - 1);
-      Value v = updated ? rootMoves[i].score : rootMoves[i].previousScore;
+      Value v = updated ? rootMoves[i].uciScore : rootMoves[i].previousScore;
 
       if (v == -VALUE_INFINITE)
           v = VALUE_ZERO;
diff --git a/src/search.h b/src/search.h
index 60f2762a..b620202d 100644
--- a/src/search.h
+++ b/src/search.h
@@ -71,6 +71,7 @@ struct RootMove {
   Value score = -VALUE_INFINITE;
   Value previousScore = -VALUE_INFINITE;
   Value averageScore = -VALUE_INFINITE;
+  Value uciScore = -VALUE_INFINITE;
   bool scoreLowerbound = false;
   bool scoreUpperbound = false;
   int selDepth = 0;

From 74fb936dbd49c61bf3352febd1c57a68888100d0 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Tue, 6 Dec 2022 21:32:42 +0900
Subject: [PATCH 005/192] Invoke .depend only on build targets

Add a constraint so that the dependency build only occurs when users
actually run build tasks.

This fixes a bug on some systems where gcc/g++ is not available.

closes https://github.com/official-stockfish/Stockfish/pull/4255

No functional change
---
 src/Makefile | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index 917bd5c0..0c98391b 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -512,7 +512,7 @@ endif
 
 ### Sometimes gcc is really clang
 ifeq ($(COMP),gcc)
-	gccversion = $(shell $(CXX) --version)
+	gccversion = $(shell $(CXX) --version 2>/dev/null)
 	gccisclang = $(findstring clang,$(gccversion))
 	ifneq ($(gccisclang),)
 		profile_make = clang-profile-make
@@ -1006,4 +1006,6 @@ icc-profile-use:
 .depend: $(SRCS)
 	-@$(CXX) $(DEPENDFLAGS) -MM $(SRCS) > $@ 2> /dev/null
 
+ifneq (, $(filter $(MAKECMDGOALS), build profile-build))
 -include .depend
+endif

From aa603cfeeb6e902bcf996758515170b996ec1fb6 Mon Sep 17 00:00:00 2001
From: ppigazzini <pasquale.pigazzini@gmail.com>
Date: Thu, 8 Dec 2022 20:50:32 +0100
Subject: [PATCH 006/192] GitHub Action: upload ARM artifacts

And some clean up in other files.

closes https://github.com/official-stockfish/Stockfish/pull/4269

No functional change
---
 .github/workflows/stockfish.yml              |   3 +
 .github/workflows/stockfish_arm_binaries.yml | 117 +++++++++++++++++++
 .github/workflows/stockfish_binaries.yml     |   7 +-
 .github/workflows/stockfish_compile_test.yml |   5 +-
 .github/workflows/stockfish_sanitizers.yml   |   5 +-
 .github/workflows/stockfish_test.yml         |  20 ++--
 6 files changed, 137 insertions(+), 20 deletions(-)
 create mode 100644 .github/workflows/stockfish_arm_binaries.yml

diff --git a/.github/workflows/stockfish.yml b/.github/workflows/stockfish.yml
index 07ecfc07..6345b27c 100644
--- a/.github/workflows/stockfish.yml
+++ b/.github/workflows/stockfish.yml
@@ -19,3 +19,6 @@ jobs:
   Binaries:
     if: github.ref == 'refs/heads/master'
     uses: ./.github/workflows/stockfish_binaries.yml
+  ARM_Binaries:
+    if: github.ref == 'refs/heads/master'
+    uses: ./.github/workflows/stockfish_arm_binaries.yml
diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
new file mode 100644
index 00000000..cf3ae710
--- /dev/null
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -0,0 +1,117 @@
+name: Stockfish
+on:
+  workflow_call:
+jobs:
+  Stockfish:
+    name: ${{ matrix.config.name }} ${{ matrix.binaries }}
+    runs-on: ${{ matrix.config.os }}
+    env:
+      COMPILER: ${{ matrix.config.compiler }}
+      COMP: ${{ matrix.config.comp }}
+      EMU: ${{ matrix.config.emu }}
+      EXT: ${{ matrix.config.ext }}
+      OS: ${{ matrix.config.os }}
+      BINARY: ${{ matrix.binaries }}
+    strategy:
+      matrix:
+        config:
+          - {
+              name: "Android NDK aarch64",
+              os: ubuntu-20.04,
+              compiler: aarch64-linux-android21-clang++,
+              emu: qemu-aarch64,
+              comp: ndk,
+              shell: 'bash {0}'
+            }
+          - {
+              name: "Android NDK arm",
+              os: ubuntu-20.04,
+              compiler: armv7a-linux-androideabi21-clang++,
+              emu: qemu-arm,
+              comp: ndk,
+              shell: 'bash {0}'
+            }
+        binaries:
+          - armv8
+          - armv7
+          - armv7-neon
+        exclude:
+          - binaries: armv8
+            config: {compiler: armv7a-linux-androideabi21-clang++}
+          - binaries: armv7
+            config: {compiler: aarch64-linux-android21-clang++}
+          - binaries: armv7-neon
+            config: {compiler: aarch64-linux-android21-clang++}
+    defaults:
+      run:
+        working-directory: src
+        shell: ${{ matrix.config.shell }}
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      - name: Download required linux packages
+        if: runner.os == 'Linux'
+        run: |
+          sudo apt update
+          sudo apt install qemu-user
+
+      - name: Download the used network from the fishtest framework
+        run: |
+          make net
+
+      - name: Check compiler
+        run: |
+          if [ $COMP == ndk ]; then
+            ANDROID_ROOT=/usr/local/lib/android
+            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
+            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
+            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          fi
+          $COMPILER -v
+
+      - name: Test help target
+        run: |
+          make help
+
+      # Compile profile guided builds
+
+      - name: Compile ${{ matrix.binaries }} build
+        run: |
+          if [ $COMP == ndk ]; then
+            ANDROID_ROOT=/usr/local/lib/android
+            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
+            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
+            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
+            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+            export LDFLAGS="-static -Wno-unused-command-line-argument"
+          fi
+          make clean
+          make -j2 profile-build ARCH=$BINARY COMP=$COMP WINE_PATH=$EMU
+          make strip ARCH=$BINARY COMP=$COMP
+          mv ./stockfish$EXT ../stockfish-android-$BINARY$EXT
+
+      - name: Remove non src files
+        run: rm -f *.o .depend *.nnue
+
+      - name: Create tar archive.
+        run: |
+          cd ..
+          mkdir stockfish
+          cp -r src stockfish/
+          cp stockfish-android-$BINARY$EXT stockfish/
+          cp "Top CPU Contributors.txt" stockfish/
+          cp Copying.txt stockfish/
+          cp AUTHORS stockfish/
+          tar -cvf stockfish-android-$BINARY.tar stockfish
+
+      - name: Upload binaries
+        uses: actions/upload-artifact@v3
+        with:
+          name: stockfish-android-${{ matrix.binaries }}
+          path: |
+            stockfish-android-${{ matrix.binaries }}.tar
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 0b205ded..1fa123fa 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -63,8 +63,8 @@ jobs:
         if: runner.os == 'Windows'
         uses: msys2/setup-msys2@v2
         with:
-          msystem: ${{matrix.config.msys_sys}}
-          install: mingw-w64-${{matrix.config.msys_env}} make git expect
+          msystem: ${{ matrix.config.msys_sys }}
+          install: mingw-w64-${{ matrix.config.msys_env }} make
 
       - name: Download the used network from the fishtest framework
         run: |
@@ -72,7 +72,6 @@ jobs:
 
       - name: Check compiler
         run: |
-          export PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin
           $COMPILER -v
 
       - name: Test help target
@@ -85,7 +84,7 @@ jobs:
         run: |
           make clean
           make -j2 profile-build ARCH=$BINARY COMP=$COMP
-          strip ./stockfish$EXT
+          make strip ARCH=$BINARY COMP=$COMP
           mv ./stockfish$EXT ../stockfish-$OS-$BINARY$EXT
 
       - name: Remove non src files
diff --git a/.github/workflows/stockfish_compile_test.yml b/.github/workflows/stockfish_compile_test.yml
index 63136737..8467f52d 100644
--- a/.github/workflows/stockfish_compile_test.yml
+++ b/.github/workflows/stockfish_compile_test.yml
@@ -72,7 +72,7 @@ jobs:
         uses: msys2/setup-msys2@v2
         with:
           msystem: ${{matrix.config.msys_sys}}
-          install: mingw-w64-${{matrix.config.msys_env}} make git expect
+          install: mingw-w64-${{matrix.config.msys_env}} make
 
       - name: Download the used network from the fishtest framework
         run: |
@@ -80,7 +80,6 @@ jobs:
 
       - name: Check compiler
         run: |
-          export PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin
           $COMPILER -v
 
       - name: Test help target
@@ -112,4 +111,4 @@ jobs:
       - name: Compile x86-64-vnni256 build
         run: |
           make clean
-          make -j2 ARCH=x86-64-vnni256 build
\ No newline at end of file
+          make -j2 ARCH=x86-64-vnni256 build
diff --git a/.github/workflows/stockfish_sanitizers.yml b/.github/workflows/stockfish_sanitizers.yml
index 61eaf0c9..b74c2f97 100644
--- a/.github/workflows/stockfish_sanitizers.yml
+++ b/.github/workflows/stockfish_sanitizers.yml
@@ -52,7 +52,7 @@ jobs:
       - name: Download required linux packages
         run: |
           sudo apt update
-          sudo apt install expect valgrind g++-multilib qemu-user
+          sudo apt install expect valgrind g++-multilib
 
       - name: Download the used network from the fishtest framework
         run: |
@@ -60,7 +60,6 @@ jobs:
 
       - name: Check compiler
         run: |
-          export PATH=$PATH:$ANDROID_NDK_HOME/toolchains/llvm/prebuilt/linux-x86_64/bin
           $COMPILER -v
 
       - name: Test help target
@@ -74,4 +73,4 @@ jobs:
           export CXXFLAGS="-O1 -fno-inline"
           make clean
           make -j2 ARCH=x86-64-modern ${{ matrix.sanitizers.make_option }} debug=yes optimize=no build > /dev/null
-          ../tests/instrumented.sh --${{ matrix.sanitizers.instrumented_option }}
\ No newline at end of file
+          ../tests/instrumented.sh --${{ matrix.sanitizers.instrumented_option }}
diff --git a/.github/workflows/stockfish_test.yml b/.github/workflows/stockfish_test.yml
index e4e6205f..3cb89d39 100644
--- a/.github/workflows/stockfish_test.yml
+++ b/.github/workflows/stockfish_test.yml
@@ -31,7 +31,7 @@ jobs:
               shell: 'bash {0}'
             }
           - {
-              name: "Ubuntu 20.04 NDK armv8",
+              name: "Android NDK aarch64",
               os: ubuntu-20.04,
               compiler: aarch64-linux-android21-clang++,
               comp: ndk,
@@ -39,7 +39,7 @@ jobs:
               shell: 'bash {0}'
             }
           - {
-              name: "Ubuntu 20.04 NDK armv7",
+              name: "Android NDK arm",
               os: ubuntu-20.04,
               compiler: armv7a-linux-androideabi21-clang++,
               comp: ndk,
@@ -127,8 +127,8 @@ jobs:
         run: |
           if [ $COMP == ndk ]; then
             ANDROID_ROOT=/usr/local/lib/android
-            ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
-            SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
+            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
             echo "y" | $SDKMANAGER "ndk;21.4.7075529"
             ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
             export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
@@ -228,8 +228,8 @@ jobs:
         if: ${{ matrix.config.run_armv8_tests }}
         run: |
           ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
-          SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
+          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
           ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
           export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
@@ -244,8 +244,8 @@ jobs:
         if: ${{ matrix.config.run_armv7_tests }}
         run: |
           ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
-          SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
+          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
           ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
           export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
@@ -258,8 +258,8 @@ jobs:
         if: ${{ matrix.config.run_armv7_tests }}
         run: |
           ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=${ANDROID_ROOT}/sdk
-          SDKMANAGER=${ANDROID_SDK_ROOT}/cmdline-tools/latest/bin/sdkmanager
+          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
           echo "y" | $SDKMANAGER "ndk;21.4.7075529"
           ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
           export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH

From 9d3fd011f1bc9ef6f3a3091aae10634b31e0032c Mon Sep 17 00:00:00 2001
From: Alfredo Menezes <lomalfredo8@gmail.com>
Date: Fri, 9 Dec 2022 12:11:43 -0300
Subject: [PATCH 007/192] Extend all moves at low depth if ttMove is doubly
 extended

If ttMove is doubly extended, we allow a depth growth of the remaining moves.
The idea is to get a more realistic score comparison, because of the depth
difference. We take some care to avoid this extension for high depths,
in order to avoid the cost, since the search result is supposed
to be more accurate in this case.

This pull request includes some small cleanups.

STC:
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 60256 W: 16189 L: 15848 D: 28219
Ptnml(0-2): 182, 6546, 16330, 6889, 181
https://tests.stockfishchess.org/tests/view/639109a1792a529ae8f27777

LTC:
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 106232 W: 28487 L: 28053 D: 49692
Ptnml(0-2): 46, 10224, 32145, 10652, 49
https://tests.stockfishchess.org/tests/view/63914cba792a529ae8f282ee

closes https://github.com/official-stockfish/Stockfish/pull/4271

Bench: 3622368
---
 src/search.cpp | 6 ++++--
 src/types.h    | 3 +++
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 04f73e1c..ec7cff54 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1067,7 +1067,10 @@ moves_loop: // When in check, search starts here
                   if (  !PvNode
                       && value < singularBeta - 25
                       && ss->doubleExtensions <= 9)
+                  {
                       extension = 2;
+                      depth += depth < 12;
+                  }
               }
 
               // Multi-cut pruning
@@ -1296,7 +1299,7 @@ moves_loop: // When in check, search starts here
                       && depth < 6
                       && beta  <  VALUE_KNOWN_WIN
                       && alpha > -VALUE_KNOWN_WIN)
-                     depth -= 1;
+                      depth -= 1;
 
                   assert(depth > 0);
               }
@@ -1521,7 +1524,6 @@ moves_loop: // When in check, search starts here
           &&  futilityBase > -VALUE_KNOWN_WIN
           &&  type_of(move) != PROMOTION)
       {
-
           if (moveCount > 2)
               continue;
 
diff --git a/src/types.h b/src/types.h
index c2087c6c..29c16ce7 100644
--- a/src/types.h
+++ b/src/types.h
@@ -186,6 +186,9 @@ enum Value : int {
   VALUE_MATE_IN_MAX_PLY  =  VALUE_MATE - MAX_PLY,
   VALUE_MATED_IN_MAX_PLY = -VALUE_MATE_IN_MAX_PLY,
 
+  // In the code, we make the assumption that these values
+  // are such that non_pawn_material() can be used to uniquely
+  // identify the material on the board.
   PawnValueMg   = 126,   PawnValueEg   = 208,
   KnightValueMg = 781,   KnightValueEg = 854,
   BishopValueMg = 825,   BishopValueEg = 915,

From 44ecadee10111f028e28f47df6dfc9accd908293 Mon Sep 17 00:00:00 2001
From: Douglas Matos Gomes <dsmsgomes@gmail.com>
Date: Thu, 8 Dec 2022 21:40:07 -0300
Subject: [PATCH 008/192] Simplify redundant condition.

closes https://github.com/official-stockfish/Stockfish/pull/4270

No functional change
---
 src/evaluate.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 93e665df..71c4e8d1 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1091,7 +1091,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
 
   // When not using NNUE, return classical complexity to caller
-  if (complexity && (!useNNUE || useClassical))
+  if (complexity && useClassical)
       *complexity = abs(v - psq);
 
   return v;

From aedf0251e6170a631b56a24dff31e83b9656933d Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Fri, 9 Dec 2022 17:56:55 +0100
Subject: [PATCH 009/192] CI workflows, install git on windows

ensures the SF dev version is reported correctly

No functional change
---
 .github/workflows/stockfish_binaries.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 1fa123fa..5ba4784e 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -64,7 +64,7 @@ jobs:
         uses: msys2/setup-msys2@v2
         with:
           msystem: ${{ matrix.config.msys_sys }}
-          install: mingw-w64-${{ matrix.config.msys_env }} make
+          install: mingw-w64-${{ matrix.config.msys_env }} make git
 
       - name: Download the used network from the fishtest framework
         run: |

From 3a30b478d20c16a357d1e538f1ce9428e7285736 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Fri, 9 Dec 2022 18:24:54 +0100
Subject: [PATCH 010/192] CI workflows, install git on windows

ensures the SF dev version is reported correctly

closes https://github.com/official-stockfish/Stockfish/pull/4272

No functional change
---
 .github/workflows/stockfish_arm_binaries.yml | 12 ++++++------
 .github/workflows/stockfish_binaries.yml     | 18 ++++++++----------
 .github/workflows/stockfish_compile_test.yml | 14 +++++++-------
 .github/workflows/stockfish_sanitizers.yml   | 12 ++++++------
 .github/workflows/stockfish_test.yml         |  9 +++++----
 5 files changed, 32 insertions(+), 33 deletions(-)

diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
index cf3ae710..ea738cef 100644
--- a/.github/workflows/stockfish_arm_binaries.yml
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -58,8 +58,7 @@ jobs:
           sudo apt install qemu-user
 
       - name: Download the used network from the fishtest framework
-        run: |
-          make net
+        run: make net
 
       - name: Check compiler
         run: |
@@ -74,8 +73,10 @@ jobs:
           $COMPILER -v
 
       - name: Test help target
-        run: |
-          make help
+        run: make help
+
+      - name: Check git
+        run: git --version
 
       # Compile profile guided builds
 
@@ -113,5 +114,4 @@ jobs:
         uses: actions/upload-artifact@v3
         with:
           name: stockfish-android-${{ matrix.binaries }}
-          path: |
-            stockfish-android-${{ matrix.binaries }}.tar
+          path: stockfish-android-${{ matrix.binaries }}.tar
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 5ba4784e..57535296 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -56,8 +56,7 @@ jobs:
 
       - name: Download required linux packages
         if: runner.os == 'Linux'
-        run: |
-          sudo apt update
+        run: sudo apt update
 
       - name: Setup msys and install required packages
         if: runner.os == 'Windows'
@@ -67,16 +66,16 @@ jobs:
           install: mingw-w64-${{ matrix.config.msys_env }} make git
 
       - name: Download the used network from the fishtest framework
-        run: |
-          make net
+        run: make net
 
       - name: Check compiler
-        run: |
-          $COMPILER -v
+        run: $COMPILER -v
 
       - name: Test help target
-        run: |
-          make help
+        run: make help
+
+      - name: Check git
+        run: git --version
 
       # Compile profile guided builds
 
@@ -105,5 +104,4 @@ jobs:
         uses: actions/upload-artifact@v3
         with:
           name: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}
-          path: |
-            stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}.tar
+          path: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}.tar
diff --git a/.github/workflows/stockfish_compile_test.yml b/.github/workflows/stockfish_compile_test.yml
index 8467f52d..eeb4229c 100644
--- a/.github/workflows/stockfish_compile_test.yml
+++ b/.github/workflows/stockfish_compile_test.yml
@@ -72,19 +72,19 @@ jobs:
         uses: msys2/setup-msys2@v2
         with:
           msystem: ${{matrix.config.msys_sys}}
-          install: mingw-w64-${{matrix.config.msys_env}} make
+          install: mingw-w64-${{matrix.config.msys_env}} make git
 
       - name: Download the used network from the fishtest framework
-        run: |
-          make net
+        run: make net
 
       - name: Check compiler
-        run: |
-          $COMPILER -v
+        run: $COMPILER -v
 
       - name: Test help target
-        run: |
-          make help
+        run: make help
+
+      - name: Check git
+        run: git --version
 
       # x86-64 with newer extensions tests
 
diff --git a/.github/workflows/stockfish_sanitizers.yml b/.github/workflows/stockfish_sanitizers.yml
index b74c2f97..fa679330 100644
--- a/.github/workflows/stockfish_sanitizers.yml
+++ b/.github/workflows/stockfish_sanitizers.yml
@@ -55,16 +55,16 @@ jobs:
           sudo apt install expect valgrind g++-multilib
 
       - name: Download the used network from the fishtest framework
-        run: |
-          make net
+        run: make net
 
       - name: Check compiler
-        run: |
-          $COMPILER -v
+        run: $COMPILER -v
 
       - name: Test help target
-        run: |
-          make help
+        run: make help
+
+      - name: Check git
+        run: git --version
 
       # Sanitizers
 
diff --git a/.github/workflows/stockfish_test.yml b/.github/workflows/stockfish_test.yml
index 3cb89d39..953f6820 100644
--- a/.github/workflows/stockfish_test.yml
+++ b/.github/workflows/stockfish_test.yml
@@ -115,8 +115,7 @@ jobs:
           install: mingw-w64-${{ matrix.config.msys_env }} make git expect
 
       - name: Download the used network from the fishtest framework
-        run: |
-          make net
+        run: make net
 
       - name: Extract the bench number from the commit history
         run: |
@@ -136,8 +135,10 @@ jobs:
           $COMPILER -v
 
       - name: Test help target
-        run: |
-          make help
+        run: make help
+
+      - name: Check git
+        run: git --version
 
       # x86-32 tests
 

From 8f817ef0824e4d940128f5701573f74819f50da5 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Fri, 9 Dec 2022 21:48:03 +0100
Subject: [PATCH 011/192] Fix lower/upper bounds output

Commit cb0c7a98485fbef4e5d6ed5f5b08201113ce0b4e doesnt reset the lower/upper bounds back to false.

fixes #4273
closes https://github.com/official-stockfish/Stockfish/pull/4274

No functional change
---
 src/search.cpp | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index ec7cff54..b2a5b940 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1250,6 +1250,8 @@ moves_loop: // When in check, search starts here
           {
               rm.score =  rm.uciScore = value;
               rm.selDepth = thisThread->selDepth;
+              rm.scoreLowerbound = rm.scoreUpperbound = false;
+
               if (value >= beta) {
                  rm.scoreLowerbound = true;
                  rm.uciScore = beta;

From 955edf1d1d4f5643b450b1ee1e95dc3f094e1884 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Mon, 12 Dec 2022 08:12:10 +0100
Subject: [PATCH 012/192] Revert "doEvenDeeperSearch + tuning"

This reverts commit 98965c139df1483a3d684ee8bc7a60dc4b95efa1.

The increase of depth could lead to search explosions,
most visible with TB.

fixes https://github.com/official-stockfish/Stockfish/issues/4276
closes https://github.com/official-stockfish/Stockfish/pull/4256

Bench: 3872306
---
 src/evaluate.cpp | 14 +++++++-------
 src/search.cpp   |  7 +++----
 2 files changed, 10 insertions(+), 11 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 71c4e8d1..6a0fae9e 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   else
   {
       int nnueComplexity;
-      int scale = 1076 + 96 * pos.non_pawn_material() / 5120;
+      int scale = 1064 + 106 * pos.non_pawn_material() / 5120;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];
@@ -1071,21 +1071,21 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  412 * nnueComplexity
-                        + 428 * abs(psq - nnue)
+      nnueComplexity = (  416 * nnueComplexity
+                        + 424 * abs(psq - nnue)
                         + (optimism  > 0 ? int(optimism) * int(psq - nnue) : 0)
-                        ) / 1026;
+                        ) / 1024;
 
       // Return hybrid NNUE complexity to caller
       if (complexity)
           *complexity = nnueComplexity;
 
-      optimism = optimism * (278 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 755)) / 1024;
+      optimism = optimism * (269 + nnueComplexity) / 256;
+      v = (nnue * scale + optimism * (scale - 754)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling
-  v = v * (197 - pos.rule50_count()) / 214;
+  v = v * (195 - pos.rule50_count()) / 211;
 
   // Guarantee evaluation does not hit the tablebase range
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
diff --git a/src/search.cpp b/src/search.cpp
index b2a5b940..b58a344a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min((12 * d + 282) * d - 349 , 1480);
+    return std::min((12 * d + 282) * d - 349 , 1594);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -1175,7 +1175,7 @@ moves_loop: // When in check, search starts here
                          - 4433;
 
           // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-          r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
+          r -= ss->statScore / (13628 + 4000 * (depth > 7 && depth < 19));
 
           // In general we want to cap the LMR depth search at newDepth, but when
           // reduction is negative, we allow this move a limited search extension
@@ -1190,10 +1190,9 @@ moves_loop: // When in check, search starts here
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
               const bool doDeeperSearch = value > (alpha + 64 + 11 * (newDepth - d));
-              const bool doEvenDeeperSearch = value > alpha + 582;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
-              newDepth += doDeeperSearch - doShallowerSearch + doEvenDeeperSearch;
+              newDepth += doDeeperSearch - doShallowerSearch;
 
               if (newDepth > d)
                   value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth, !cutNode);

From 310928e985a6d87bdd73542e2109f93c31e2cc41 Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Fri, 9 Dec 2022 18:50:06 -0800
Subject: [PATCH 013/192] Avoid truncated PV in the threaded case

strongly prefer to pick as bestThread those threads with a longer PV,
among those threads that all found the same bestmove.

extended discussion in #4244
closes https://github.com/official-stockfish/Stockfish/pull/4278

No functional change
---
 src/thread.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/thread.cpp b/src/thread.cpp
index b7471f60..e8723eb7 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -239,7 +239,8 @@ Thread* ThreadPool::get_best_thread() const {
                  || (   th->rootMoves[0].score > VALUE_TB_LOSS_IN_MAX_PLY
                      && (   votes[th->rootMoves[0].pv[0]] > votes[bestThread->rootMoves[0].pv[0]]
                          || (   votes[th->rootMoves[0].pv[0]] == votes[bestThread->rootMoves[0].pv[0]]
-                             && thread_value(th) > thread_value(bestThread)))))
+                             &&   thread_value(th) * int(th->rootMoves[0].pv.size() > 2)
+                                > thread_value(bestThread) * int(bestThread->rootMoves[0].pv.size() > 2)))))
             bestThread = th;
 
     return bestThread;

From 5fe1fa0210e25cecc9de5520da56173ef94b8e59 Mon Sep 17 00:00:00 2001
From: ppigazzini <pasquale.pigazzini@gmail.com>
Date: Sun, 11 Dec 2022 12:06:22 +0100
Subject: [PATCH 014/192] GitHub Actions: install NDK once and clean up yaml

Use Ubuntu 22.04 as runner for NDK to avoid a qemu bug with `profile-build`

closes https://github.com/official-stockfish/Stockfish/pull/4280

No functional change
---
 .github/workflows/stockfish_arm_binaries.yml |  56 +++---
 .github/workflows/stockfish_binaries.yml     |  42 ++---
 .github/workflows/stockfish_compile_test.yml |  80 ++++----
 .github/workflows/stockfish_sanitizers.yml   |  44 ++---
 .github/workflows/stockfish_test.yml         | 184 ++++++++-----------
 5 files changed, 177 insertions(+), 229 deletions(-)

diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
index ea738cef..a1b3cdab 100644
--- a/.github/workflows/stockfish_arm_binaries.yml
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -15,22 +15,18 @@ jobs:
     strategy:
       matrix:
         config:
-          - {
-              name: "Android NDK aarch64",
-              os: ubuntu-20.04,
-              compiler: aarch64-linux-android21-clang++,
-              emu: qemu-aarch64,
-              comp: ndk,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Android NDK arm",
-              os: ubuntu-20.04,
-              compiler: armv7a-linux-androideabi21-clang++,
-              emu: qemu-arm,
-              comp: ndk,
-              shell: 'bash {0}'
-            }
+          - name: Android NDK aarch64
+            os: ubuntu-22.04
+            compiler: aarch64-linux-android21-clang++
+            emu: qemu-aarch64
+            comp: ndk
+            shell: bash {0}
+          - name: Android NDK arm
+            os: ubuntu-22.04
+            compiler: armv7a-linux-androideabi21-clang++
+            emu: qemu-arm
+            comp: ndk
+            shell: bash {0}
         binaries:
           - armv8
           - armv7
@@ -57,18 +53,27 @@ jobs:
           sudo apt update
           sudo apt install qemu-user
 
+      - name: Install NDK
+        if: runner.os == 'Linux'
+        run: |
+          if [ $COMP == ndk ]; then
+            NDKV="21.4.7075529"
+            ANDROID_ROOT=/usr/local/lib/android
+            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
+            echo "y" | $SDKMANAGER "ndk;$NDKV"
+            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/$NDKV
+            ANDROID_NDK_BIN=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin
+            echo "ANDROID_NDK_BIN=$ANDROID_NDK_BIN" >> $GITHUB_ENV
+          fi
+
       - name: Download the used network from the fishtest framework
         run: make net
 
       - name: Check compiler
         run: |
           if [ $COMP == ndk ]; then
-            ANDROID_ROOT=/usr/local/lib/android
-            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+            export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
           fi
           $COMPILER -v
 
@@ -83,12 +88,7 @@ jobs:
       - name: Compile ${{ matrix.binaries }} build
         run: |
           if [ $COMP == ndk ]; then
-            ANDROID_ROOT=/usr/local/lib/android
-            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+            export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
             export LDFLAGS="-static -Wno-unused-command-line-argument"
           fi
           make clean
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 57535296..06b13a9e 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -14,30 +14,24 @@ jobs:
     strategy:
       matrix:
         config:
-          - {
-              name: "Ubuntu 20.04 GCC",
-              os: ubuntu-20.04,
-              compiler: g++,
-              comp: gcc,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "MacOS 12 Apple Clang",
-              os: macos-12,
-              compiler: clang++,
-              comp: clang,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 GCC x86_64",
-              os: windows-2022,
-              compiler: g++,
-              comp: mingw,
-              msys_sys: 'mingw64',
-              msys_env: 'x86_64-gcc',
-              shell: 'msys2 {0}',
-              ext: .exe
-            }
+          - name: Ubuntu 20.04 GCC
+            os: ubuntu-20.04
+            compiler: g++
+            comp: gcc
+            shell: bash {0}
+          - name: MacOS 12 Apple Clang
+            os: macos-12
+            compiler: clang++
+            comp: clang
+            shell: bash {0}
+          - name: Windows 2022 Mingw-w64 GCC x86_64
+            os: windows-2022
+            compiler: g++
+            comp: mingw
+            msys_sys: mingw64
+            msys_env: x86_64-gcc
+            shell: msys2 {0}
+            ext: .exe
         binaries:
           - x86-64
           - x86-64-modern
diff --git a/.github/workflows/stockfish_compile_test.yml b/.github/workflows/stockfish_compile_test.yml
index eeb4229c..c7280a85 100644
--- a/.github/workflows/stockfish_compile_test.yml
+++ b/.github/workflows/stockfish_compile_test.yml
@@ -11,52 +11,40 @@ jobs:
     strategy:
       matrix:
         config:
-          - {
-              name: "Ubuntu 20.04 GCC",
-              os: ubuntu-20.04,
-              compiler: g++,
-              comp: gcc,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Ubuntu 20.04 Clang",
-              os: ubuntu-20.04,
-              compiler: clang++,
-              comp: clang,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "MacOS 12 Apple Clang",
-              os: macos-12,
-              compiler: clang++,
-              comp: clang,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "MacOS 12 GCC 11",
-              os: macos-12,
-              compiler: g++-11,
-              comp: gcc,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 GCC x86_64",
-              os: windows-2022,
-              compiler: g++,
-              comp: mingw,
-              msys_sys: 'mingw64',
-              msys_env: 'x86_64-gcc',
-              shell: 'msys2 {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 Clang x86_64",
-              os: windows-2022,
-              compiler: clang++,
-              comp: clang,
-              msys_sys: 'clang64',
-              msys_env: 'clang-x86_64-clang',
-              shell: 'msys2 {0}'
-            }
+          - name: Ubuntu 20.04 GCC
+            os: ubuntu-20.04
+            compiler: g++
+            comp: gcc
+            shell: bash {0}
+          - name: Ubuntu 20.04 Clang
+            os: ubuntu-20.04
+            compiler: clang++
+            comp: clang
+            shell: bash {0}
+          - name: MacOS 12 Apple Clang
+            os: macos-12
+            compiler: clang++
+            comp: clang
+            shell: bash {0}
+          - name: MacOS 12 GCC 11
+            os: macos-12
+            compiler: g++-11
+            comp: gcc
+            shell: bash {0}
+          - name: Windows 2022 Mingw-w64 GCC x86_64
+            os: windows-2022
+            compiler: g++
+            comp: mingw
+            msys_sys: mingw64
+            msys_env: x86_64-gcc
+            shell: msys2 {0}
+          - name: Windows 2022 Mingw-w64 Clang x86_64
+            os: windows-2022
+            compiler: clang++
+            comp: clang
+            msys_sys: clang64
+            msys_env: clang-x86_64-clang
+            shell: msys2 {0}
 
     defaults:
       run:
diff --git a/.github/workflows/stockfish_sanitizers.yml b/.github/workflows/stockfish_sanitizers.yml
index fa679330..708c9227 100644
--- a/.github/workflows/stockfish_sanitizers.yml
+++ b/.github/workflows/stockfish_sanitizers.yml
@@ -12,34 +12,24 @@ jobs:
     strategy:
       matrix:
         config:
-          - {
-              name: "Ubuntu 20.04 GCC",
-              os: ubuntu-20.04,
-              compiler: g++,
-              comp: gcc,
-              shell: 'bash {0}'
-            }
+          - name: Ubuntu 20.04 GCC
+            os: ubuntu-20.04
+            compiler: g++
+            comp: gcc
+            shell: bash {0}
         sanitizers:
-          - {
-              name: Run with thread sanitizer,
-              make_option: sanitize=thread,
-              instrumented_option: sanitizer-thread
-            }
-          - {
-              name: Run with UB sanitizer,
-              make_option: sanitize=undefined,
-              instrumented_option: sanitizer-undefined
-            }
-          - {
-              name: Run under valgrind,
-              make_option: "",
-              instrumented_option: valgrind
-            }
-          - {
-              name: Run under valgrind-thread,
-              make_option: "",
-              instrumented_option: valgrind-thread
-            }
+          - name: Run with thread sanitizer
+            make_option: sanitize=thread
+            instrumented_option: sanitizer-thread
+          - name: Run with UB sanitizer
+            make_option: sanitize=undefined
+            instrumented_option: sanitizer-undefined
+          - name: Run under valgrind
+            make_option: ""
+            instrumented_option: valgrind
+          - name: Run under valgrind-thread
+            make_option: ""
+            instrumented_option: valgrind-thread
     defaults:
       run:
         working-directory: src
diff --git a/.github/workflows/stockfish_test.yml b/.github/workflows/stockfish_test.yml
index 953f6820..8c383fe7 100644
--- a/.github/workflows/stockfish_test.yml
+++ b/.github/workflows/stockfish_test.yml
@@ -12,86 +12,68 @@ jobs:
     strategy:
       matrix:
         config:
-          - {
-              name: "Ubuntu 20.04 GCC",
-              os: ubuntu-20.04,
-              compiler: g++,
-              comp: gcc,
-              run_32bit_tests: true,
-              run_64bit_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Ubuntu 20.04 Clang",
-              os: ubuntu-20.04,
-              compiler: clang++,
-              comp: clang,
-              run_32bit_tests: true,
-              run_64bit_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Android NDK aarch64",
-              os: ubuntu-20.04,
-              compiler: aarch64-linux-android21-clang++,
-              comp: ndk,
-              run_armv8_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Android NDK arm",
-              os: ubuntu-20.04,
-              compiler: armv7a-linux-androideabi21-clang++,
-              comp: ndk,
-              run_armv7_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "MacOS 12 Apple Clang",
-              os: macos-12,
-              compiler: clang++,
-              comp: clang,
-              run_64bit_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "MacOS 12 GCC 11",
-              os: macos-12,
-              compiler: g++-11,
-              comp: gcc,
-              run_64bit_tests: true,
-              shell: 'bash {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 GCC x86_64",
-              os: windows-2022,
-              compiler: g++,
-              comp: mingw,
-              run_64bit_tests: true,
-              msys_sys: 'mingw64',
-              msys_env: 'x86_64-gcc',
-              shell: 'msys2 {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 GCC i686",
-              os: windows-2022,
-              compiler: g++,
-              comp: mingw,
-              run_32bit_tests: true,
-              msys_sys: 'mingw32',
-              msys_env: 'i686-gcc',
-              shell: 'msys2 {0}'
-            }
-          - {
-              name: "Windows 2022 Mingw-w64 Clang x86_64",
-              os: windows-2022,
-              compiler: clang++,
-              comp: clang,
-              run_64bit_tests: true,
-              msys_sys: 'clang64',
-              msys_env: 'clang-x86_64-clang',
-              shell: 'msys2 {0}'
-            }
+          - name: Ubuntu 20.04 GCC
+            os: ubuntu-20.04
+            compiler: g++
+            comp: gcc
+            run_32bit_tests: true
+            run_64bit_tests: true
+            shell: bash {0}
+          - name: Ubuntu 20.04 Clang
+            os: ubuntu-20.04
+            compiler: clang++
+            comp: clang
+            run_32bit_tests: true
+            run_64bit_tests: true
+            shell: bash {0}
+          - name: Android NDK aarch64
+            os: ubuntu-22.04
+            compiler: aarch64-linux-android21-clang++
+            comp: ndk
+            run_armv8_tests: true
+            shell: bash {0}
+          - name: Android NDK arm
+            os: ubuntu-22.04
+            compiler: armv7a-linux-androideabi21-clang++
+            comp: ndk
+            run_armv7_tests: true
+            shell: bash {0}
+          - name: MacOS 12 Apple Clang
+            os: macos-12
+            compiler: clang++
+            comp: clang
+            run_64bit_tests: true
+            shell: bash {0}
+          - name: MacOS 12 GCC 11
+            os: macos-12
+            compiler: g++-11
+            comp: gcc
+            run_64bit_tests: true
+            shell: bash {0}
+          - name: Windows 2022 Mingw-w64 GCC x86_64
+            os: windows-2022
+            compiler: g++
+            comp: mingw
+            run_64bit_tests: true
+            msys_sys: mingw64
+            msys_env: x86_64-gcc
+            shell: msys2 {0}
+          - name: Windows 2022 Mingw-w64 GCC i686
+            os: windows-2022
+            compiler: g++
+            comp: mingw
+            run_32bit_tests: true
+            msys_sys: mingw32
+            msys_env: i686-gcc
+            shell: msys2 {0}
+          - name: Windows 2022 Mingw-w64 Clang x86_64
+            os: windows-2022
+            compiler: clang++
+            comp: clang
+            run_64bit_tests: true
+            msys_sys: clang64
+            msys_env: clang-x86_64-clang
+            shell: msys2 {0}
     defaults:
       run:
         working-directory: src
@@ -107,6 +89,20 @@ jobs:
           sudo apt update
           sudo apt install expect valgrind g++-multilib qemu-user
 
+      - name: Install NDK
+        if: runner.os == 'Linux'
+        run: |
+          if [ $COMP == ndk ]; then
+            NDKV="21.4.7075529"
+            ANDROID_ROOT=/usr/local/lib/android
+            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
+            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
+            echo "y" | $SDKMANAGER "ndk;$NDKV"
+            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/$NDKV
+            ANDROID_NDK_BIN=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin
+            echo "ANDROID_NDK_BIN=$ANDROID_NDK_BIN" >> $GITHUB_ENV
+          fi
+
       - name: Setup msys and install required packages
         if: runner.os == 'Windows'
         uses: msys2/setup-msys2@v2
@@ -125,12 +121,7 @@ jobs:
       - name: Check compiler
         run: |
           if [ $COMP == ndk ]; then
-            ANDROID_ROOT=/usr/local/lib/android
-            ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-            SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-            echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-            ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-            export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+            export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
           fi
           $COMPILER -v
 
@@ -228,12 +219,7 @@ jobs:
       - name: Test armv8 build
         if: ${{ matrix.config.run_armv8_tests }}
         run: |
-          ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-          echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv8 build
@@ -244,12 +230,7 @@ jobs:
       - name: Test armv7 build
         if: ${{ matrix.config.run_armv7_tests }}
         run: |
-          ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-          echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv7 build
@@ -258,12 +239,7 @@ jobs:
       - name: Test armv7-neon build
         if: ${{ matrix.config.run_armv7_tests }}
         run: |
-          ANDROID_ROOT=/usr/local/lib/android
-          ANDROID_SDK_ROOT=$ANDROID_ROOT/sdk
-          SDKMANAGER=$ANDROID_SDK_ROOT/cmdline-tools/latest/bin/sdkmanager
-          echo "y" | $SDKMANAGER "ndk;21.4.7075529"
-          ANDROID_NDK_ROOT=$ANDROID_SDK_ROOT/ndk/21.4.7075529
-          export PATH=$ANDROID_NDK_ROOT/toolchains/llvm/prebuilt/linux-x86_64/bin:$PATH
+          export PATH=${{ env.ANDROID_NDK_BIN }}:$PATH
           export LDFLAGS="-static -Wno-unused-command-line-argument"
           make clean
           make -j2 ARCH=armv7-neon build

From 7cf93f8b7101e5cf5df3f7921861f27b3beb9525 Mon Sep 17 00:00:00 2001
From: VoyagerOne <excelgeek@gmail.com>
Date: Sun, 11 Dec 2022 12:36:13 -0500
Subject: [PATCH 015/192] Simplify Capture Scoring

The parameters are now in one place for easier tuning.
New formula is very similar to current.

STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 48176 W: 12819 L: 12616 D: 22741
Ptnml(0-2): 139, 5316, 13001, 5467, 165

LTC:
LLR: 2.97 (-2.94,2.94) <-1.75,0.25>
Total: 176752 W: 47364 L: 47304 D: 82084
Ptnml(0-2): 83, 17302, 53536, 17382, 73
https://tests.stockfishchess.org/tests/view/638ec7d068532fcbf79dfa15

closes https://github.com/official-stockfish/Stockfish/pull/4281

Bench: 3410998
---
 src/movepick.cpp | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/src/movepick.cpp b/src/movepick.cpp
index 188d6bd8..564adc28 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -100,7 +100,7 @@ MovePicker::MovePicker(const Position& p, Move ttm, Value th, const CapturePiece
 
 /// 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.
+/// captures with a good history. Quiets moves are ordered using the history tables.
 template<GenType Type>
 void MovePicker::score() {
 
@@ -123,8 +123,8 @@ void MovePicker::score() {
 
   for (auto& m : *this)
       if constexpr (Type == CAPTURES)
-          m.value =  6 * int(PieceValue[MG][pos.piece_on(to_sq(m))])
-                   +     (*captureHistory)[pos.moved_piece(m)][to_sq(m)][type_of(pos.piece_on(to_sq(m)))];
+          m.value =  (7 * int(PieceValue[MG][pos.piece_on(to_sq(m))])
+                   +     (*captureHistory)[pos.moved_piece(m)][to_sq(m)][type_of(pos.piece_on(to_sq(m)))]) / 16;
 
       else if constexpr (Type == QUIETS)
           m.value =  2 * (*mainHistory)[pos.side_to_move()][from_to(m)]
@@ -197,7 +197,7 @@ top:
 
   case GOOD_CAPTURE:
       if (select<Next>([&](){
-                       return pos.see_ge(*cur, Value(-69 * cur->value / 1024)) ?
+                       return pos.see_ge(*cur, Value(-cur->value)) ?
                               // Move losing capture to endBadCaptures to be tried later
                               true : (*endBadCaptures++ = *cur, false); }))
           return *(cur - 1);

From 3659a9fda0096cac091458df2c259e5636bc9106 Mon Sep 17 00:00:00 2001
From: NguyenPham <phhnguyen@gmail.com>
Date: Thu, 15 Dec 2022 17:29:23 +1100
Subject: [PATCH 016/192] Fixed the help of Makefile

make profile-build more prominent, adjust comments

closes https://github.com/official-stockfish/Stockfish/pull/4284

No functional change
---
 src/Makefile | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index 0c98391b..da81ceb4 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -747,9 +747,9 @@ help:
 	@echo "Supported targets:"
 	@echo ""
 	@echo "help                    > Display architecture details"
-	@echo "build                   > Standard build"
+	@echo "profile-build           > standard build with profile-guided optimization"
+	@echo "build                   > skip profile-guided optimization"
 	@echo "net                     > Download the default nnue net"
-	@echo "profile-build           > Faster build (with profile-guided optimization)"
 	@echo "strip                   > Strip executable"
 	@echo "install                 > Install executable"
 	@echo "clean                   > Clean up"
@@ -789,14 +789,15 @@ help:
 	@echo "icc                     > Intel compiler"
 	@echo "ndk                     > Google NDK to cross-compile for Android"
 	@echo ""
-	@echo "Simple examples. If you don't know what to do, you likely want to run: "
+	@echo "Simple examples. If you don't know what to do, you likely want to run one of: "
 	@echo ""
-	@echo "make -j build ARCH=x86-64  (A portable, slow compile for 64-bit systems)"
-	@echo "make -j build ARCH=x86-32  (A portable, slow compile for 32-bit systems)"
+	@echo "make -j profile-build ARCH=x86-64-avx2    # typically a fast compile for common systems "
+	@echo "make -j profile-build ARCH=x86-64-modern  # A more portable compile for 64-bit systems "
+	@echo "make -j profile-build ARCH=x86-64         # A portable compile for 64-bit systems "
 	@echo ""
-	@echo "Advanced examples, for experienced users looking for performance: "
+	@echo "Advanced examples, for experienced users: "
 	@echo ""
-	@echo "make    help  ARCH=x86-64-bmi2"
+	@echo "make -j profile-build ARCH=x86-64-bmi2"
 	@echo "make -j profile-build ARCH=x86-64-bmi2 COMP=gcc COMPCXX=g++-9.0"
 	@echo "make -j build ARCH=x86-64-ssse3 COMP=clang"
 	@echo ""

From 726e90ccfaea85c5831ad4834a6a08f5bec3f2f2 Mon Sep 17 00:00:00 2001
From: PikaCat <73384062+PikaCat-OuO@users.noreply.github.com>
Date: Fri, 16 Dec 2022 10:28:31 +0800
Subject: [PATCH 017/192] Badge link fix

Fix the badge link issue mentioned in https://github.com/badges/shields/issues/8671

closes https://github.com/official-stockfish/Stockfish/pull/4285

No functional change
---
 AUTHORS   | 1 +
 README.md | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/AUTHORS b/AUTHORS
index 432d9b90..0c1c5529 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -163,6 +163,7 @@ Pasquale Pigazzini (ppigazzini)
 Patrick Jansen (mibere)
 Peter Schneider (pschneider1968)
 Peter Zsifkovits (CoffeeOne)
+PikaCat
 Praveen Kumar Tummala (praveentml)
 Rahul Dsilva (silversolver1)
 Ralph Stößer (Ralph Stoesser)
diff --git a/README.md b/README.md
index ac15fbfa..4cd5968e 100644
--- a/README.md
+++ b/README.md
@@ -369,7 +369,7 @@ For full details, read the copy of the GPL v3 found in the file named
 [website-link]:       https://stockfishchess.org
 [worker-link]:        https://github.com/glinscott/fishtest/wiki/Running-the-worker:-overview
 
-[build-badge]:        https://img.shields.io/github/workflow/status/official-stockfish/Stockfish/Stockfish?style=for-the-badge&label=stockfish&logo=github
+[build-badge]:        https://img.shields.io/github/actions/workflow/status/official-stockfish/Stockfish/stockfish.yml?branch=master&style=for-the-badge&label=stockfish&logo=github
 [commits-badge]:      https://img.shields.io/github/commits-since/official-stockfish/Stockfish/latest?style=for-the-badge
 [discord-badge]:      https://img.shields.io/discord/435943710472011776?style=for-the-badge&label=discord&logo=Discord
 [fishtest-badge]:     https://img.shields.io/website?style=for-the-badge&down_color=red&down_message=Offline&label=Fishtest&up_color=success&up_message=Online&url=https%3A%2F%2Ftests.stockfishchess.org%2Ftests%2Ffinished

From 39af98c807d236b6511b6e399caf40102398900c Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Sat, 17 Dec 2022 12:48:03 +0300
Subject: [PATCH 018/192] Reintroduce doEvenDeeperSearch

This patch is basically the same as a reverted patch
but now has some guarding against search being stuck - the same
way as we do with double extensions. This should help with
search explosions - albeit slowly but they eventually should be resolved.

passed STC:
https://tests.stockfishchess.org/tests/view/639733d0b4e52c95053f3485
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 514048 W: 136423 L: 135435 D: 242190
Ptnml(0-2): 1425, 56945, 139420, 57685, 1549

passed LTC:
https://tests.stockfishchess.org/tests/view/639ab79b93ed41c57eded5c3
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 113800 W: 30642 L: 30190 D: 52968
Ptnml(0-2): 53, 11092, 34178, 11504, 73

closes https://github.com/official-stockfish/Stockfish/pull/4287

bench 3611278
---
 src/evaluate.cpp | 12 ++++++------
 src/search.cpp   | 11 +++++++----
 2 files changed, 13 insertions(+), 10 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 6a0fae9e..7619398e 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   else
   {
       int nnueComplexity;
-      int scale = 1064 + 106 * pos.non_pawn_material() / 5120;
+      int scale = 1076 + 96 * pos.non_pawn_material() / 5120;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];
@@ -1071,8 +1071,8 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  416 * nnueComplexity
-                        + 424 * abs(psq - nnue)
+      nnueComplexity = (  412 * nnueComplexity
+                        + 428 * abs(psq - nnue)
                         + (optimism  > 0 ? int(optimism) * int(psq - nnue) : 0)
                         ) / 1024;
 
@@ -1080,12 +1080,12 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       if (complexity)
           *complexity = nnueComplexity;
 
-      optimism = optimism * (269 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 754)) / 1024;
+      optimism = optimism * (278 + nnueComplexity) / 256;
+      v = (nnue * scale + optimism * (scale - 755)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling
-  v = v * (195 - pos.rule50_count()) / 211;
+  v = v * (197 - pos.rule50_count()) / 214;
 
   // Guarantee evaluation does not hit the tablebase range
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
diff --git a/src/search.cpp b/src/search.cpp
index b58a344a..4507460a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min((12 * d + 282) * d - 349 , 1594);
+    return std::min((12 * d + 282) * d - 349 , 1480);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -1066,7 +1066,7 @@ moves_loop: // When in check, search starts here
                   // Avoid search explosion by limiting the number of double extensions
                   if (  !PvNode
                       && value < singularBeta - 25
-                      && ss->doubleExtensions <= 9)
+                      && ss->doubleExtensions <= 10)
                   {
                       extension = 2;
                       depth += depth < 12;
@@ -1175,7 +1175,7 @@ moves_loop: // When in check, search starts here
                          - 4433;
 
           // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-          r -= ss->statScore / (13628 + 4000 * (depth > 7 && depth < 19));
+          r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
 
           // In general we want to cap the LMR depth search at newDepth, but when
           // reduction is negative, we allow this move a limited search extension
@@ -1190,9 +1190,12 @@ moves_loop: // When in check, search starts here
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
               const bool doDeeperSearch = value > (alpha + 64 + 11 * (newDepth - d));
+              const bool doEvenDeeperSearch = value > alpha + 582 && ss->doubleExtensions <= 5;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
-              newDepth += doDeeperSearch - doShallowerSearch;
+              ss->doubleExtensions = ss->doubleExtensions + doEvenDeeperSearch;
+
+              newDepth += doDeeperSearch - doShallowerSearch + doEvenDeeperSearch;
 
               if (newDepth > d)
                   value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth, !cutNode);

From 3a32d3e00cca23648f10a40c5b322f2fd56dc0ae Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Mon, 12 Dec 2022 23:22:02 -0800
Subject: [PATCH 019/192] Don't reset increaseDepth back to true after it has
 been set to false

Resetting increaseDepth back to true each time on the very next iteration was not intended so this is a bug fix and a simplification.
See more discussion here #2482 (comment) Thanks to xoto10

STC: https://tests.stockfishchess.org/tests/view/6398c74693ed41c57ede7bfd
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 51128 W: 13543 L: 13220 D: 24365
Ptnml(0-2): 165, 5363, 14174, 5708, 154

LTC: https://tests.stockfishchess.org/tests/view/6399bcd393ed41c57edea750
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 290864 W: 77282 L: 77334 D: 136248
Ptnml(0-2): 107, 28127, 89029, 28049, 120

closes https://github.com/official-stockfish/Stockfish/pull/4288

bench: 3611278
---
 src/search.cpp | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 4507460a..d39a8ad2 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -487,12 +487,11 @@ void Thread::search() {
               else
                   Threads.stop = true;
           }
-          else if (   Threads.increaseDepth
-                   && !mainThread->ponder
+          else if (   !mainThread->ponder
                    && Time.elapsed() > totalTime * 0.53)
-                   Threads.increaseDepth = false;
+              Threads.increaseDepth = false;
           else
-                   Threads.increaseDepth = true;
+              Threads.increaseDepth = true;
       }
 
       mainThread->iterValue[iterIdx] = bestValue;

From 61ea1534ff7026009a3435575c7beee91534db83 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Mon, 19 Dec 2022 17:54:36 +0100
Subject: [PATCH 020/192] No error if net available but wget/curl missing

do not error out on missing wget/curl if these tools are not needed later on,
i.e. if the net is available already.

closes https://github.com/official-stockfish/Stockfish/pull/4291
closes https://github.com/official-stockfish/Stockfish/pull/4253

No functional change
---
 src/Makefile | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index da81ceb4..bcf0abdf 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -853,7 +853,7 @@ net:
 	$(eval nnuedownloadurl2 := https://github.com/official-stockfish/networks/raw/master/$(nnuenet))
 	$(eval curl_or_wget := $(shell if hash curl 2>/dev/null; then echo "curl -skL"; elif hash wget 2>/dev/null; then echo "wget -qO-"; fi))
 	@if [ "x$(curl_or_wget)" = "x" ]; then \
-	    echo "Automatic download failed: neither curl nor wget is installed. Install one of these tools or download the net manually"; exit 1; \
+	    echo "Neither curl nor wget is installed. Install one of these tools unless the net has been downloaded manually"; \
         fi
 	$(eval shasum_command := $(shell if hash shasum 2>/dev/null; then echo "shasum -a 256 "; elif hash sha256sum 2>/dev/null; then echo "sha256sum "; fi))
 	@if [ "x$(shasum_command)" = "x" ]; then \
@@ -864,7 +864,9 @@ net:
 	      echo "$(nnuenet) available."; \
 	   else \
 	      if [ "x$(curl_or_wget)" != "x" ]; then \
-	         echo "Downloading $${nnuedownloadurl}"; $(curl_or_wget) $${nnuedownloadurl} > $(nnuenet);\
+                 echo "Downloading $${nnuedownloadurl}"; $(curl_or_wget) $${nnuedownloadurl} > $(nnuenet);\
+              else \
+                 echo "No net found and download not possible"; exit 1;\
 	      fi; \
 	   fi; \
 	   if [ "x$(shasum_command)" != "x" ]; then \

From c2d507005c51145211a7963af7c41026bd759d08 Mon Sep 17 00:00:00 2001
From: Alfredo Menezes <lomalfredo8@gmail.com>
Date: Mon, 19 Dec 2022 16:07:09 -0300
Subject: [PATCH 021/192] Sometimes do a reduced search if LMR is skipped

If the node doesn't go through LMR and r is too big,
reduce search depth by one ply.

STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 664888 W: 176375 L: 175169 D: 313344
Ptnml(0-2): 1965, 73754, 179851, 74858, 2016
https://tests.stockfishchess.org/tests/view/6399414c93ed41c57ede8fb8

LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 150784 W: 40553 L: 40031 D: 70200
Ptnml(0-2): 76, 14668, 45387, 15180, 81
https://tests.stockfishchess.org/tests/view/639dee6e11c576d919dc2b38

closes https://github.com/official-stockfish/Stockfish/pull/4290

Bench: 3727508
---
 src/search.cpp | 96 +++++++++++++++++++++++++-------------------------
 1 file changed, 48 insertions(+), 48 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index d39a8ad2..7f2f5284 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1120,6 +1120,52 @@ moves_loop: // When in check, search starts here
       // Step 16. Make the move
       pos.do_move(move, st, givesCheck);
 
+      Depth r = reduction(improving, depth, moveCount, delta, thisThread->rootDelta);
+
+      // Decrease reduction if position is or has been on the PV
+      // and node is not likely to fail low. (~3 Elo)
+      if (   ss->ttPv
+          && !likelyFailLow)
+          r -= 2;
+
+      // Decrease reduction if opponent's move count is high (~1 Elo)
+      if ((ss-1)->moveCount > 7)
+          r--;
+
+      // Increase reduction for cut nodes (~3 Elo)
+      if (cutNode)
+          r += 2;
+
+      // Increase reduction if ttMove is a capture (~3 Elo)
+      if (ttCapture)
+          r++;
+
+      // Decrease reduction for PvNodes based on depth
+      if (PvNode)
+          r -= 1 + 11 / (3 + depth);
+
+      // Decrease reduction if ttMove has been singularly extended (~1 Elo)
+      if (singularQuietLMR)
+          r--;
+
+      // Decrease reduction if we move a threatened piece (~1 Elo)
+      if (   depth > 9
+          && (mp.threatenedPieces & from_sq(move)))
+          r--;
+
+      // Increase reduction if next ply has a lot of fail high
+      if ((ss+1)->cutoffCnt > 3)
+          r++;
+
+      ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
+                     + (*contHist[0])[movedPiece][to_sq(move)]
+                     + (*contHist[1])[movedPiece][to_sq(move)]
+                     + (*contHist[3])[movedPiece][to_sq(move)]
+                     - 4433;
+
+      // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
+      r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
+
       // Step 17. Late moves reduction / extension (LMR, ~98 Elo)
       // We use various heuristics for the sons of a node after the first son has
       // been searched. In general we would like to reduce them, but there are many
@@ -1130,52 +1176,6 @@ moves_loop: // When in check, search starts here
               || !capture
               || (cutNode && (ss-1)->moveCount > 1)))
       {
-          Depth r = reduction(improving, depth, moveCount, delta, thisThread->rootDelta);
-
-          // Decrease reduction if position is or has been on the PV
-          // and node is not likely to fail low. (~3 Elo)
-          if (   ss->ttPv
-              && !likelyFailLow)
-              r -= 2;
-
-          // Decrease reduction if opponent's move count is high (~1 Elo)
-          if ((ss-1)->moveCount > 7)
-              r--;
-
-          // Increase reduction for cut nodes (~3 Elo)
-          if (cutNode)
-              r += 2;
-
-          // Increase reduction if ttMove is a capture (~3 Elo)
-          if (ttCapture)
-              r++;
-
-          // Decrease reduction for PvNodes based on depth
-          if (PvNode)
-              r -= 1 + 11 / (3 + depth);
-
-          // Decrease reduction if ttMove has been singularly extended (~1 Elo)
-          if (singularQuietLMR)
-              r--;
-
-          // Decrease reduction if we move a threatened piece (~1 Elo)
-          if (   depth > 9
-              && (mp.threatenedPieces & from_sq(move)))
-              r--;
-
-          // Increase reduction if next ply has a lot of fail high
-          if ((ss+1)->cutoffCnt > 3)
-              r++;
-
-          ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
-                         + (*contHist[0])[movedPiece][to_sq(move)]
-                         + (*contHist[1])[movedPiece][to_sq(move)]
-                         + (*contHist[3])[movedPiece][to_sq(move)]
-                         - 4433;
-
-          // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-          r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
-
           // In general we want to cap the LMR depth search at newDepth, but when
           // reduction is negative, we allow this move a limited search extension
           // beyond the first move depth. This may lead to hidden double extensions.
@@ -1209,10 +1209,10 @@ moves_loop: // When in check, search starts here
           }
       }
 
-      // Step 18. Full depth search when LMR is skipped
+      // Step 18. Full depth search when LMR is skipped. If expected reduction is high, reduce its depth by 1.
       else if (!PvNode || moveCount > 1)
       {
-              value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth, !cutNode);
+               value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 4), !cutNode);
       }
 
       // For PV nodes only, do a full PV search on the first move or after a fail

From 20b02264620397eacf4e43736ec7f7c48a9a7068 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Tue, 20 Dec 2022 16:01:05 +0900
Subject: [PATCH 022/192] Fix a dependency bug

Instead of allowing .depend for specific build-related targets, filter
non-build-related targets (i.e. help, clean) so that other targets can
normally execute .depend target.

closes https://github.com/official-stockfish/Stockfish/pull/4293

No functional change
---
 src/Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index bcf0abdf..500c1006 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -1009,6 +1009,6 @@ icc-profile-use:
 .depend: $(SRCS)
 	-@$(CXX) $(DEPENDFLAGS) -MM $(SRCS) > $@ 2> /dev/null
 
-ifneq (, $(filter $(MAKECMDGOALS), build profile-build))
+ifeq (, $(filter $(MAKECMDGOALS), help strip install clean net objclean profileclean config-sanity))
 -include .depend
 endif

From c620886181b5c0fb4a0c067b4173143a81819199 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Mon, 19 Dec 2022 16:10:22 -0500
Subject: [PATCH 023/192] Update default net to nn-335a9b2d8a80.nnue

Created by retraining the master net with a combination of:

    the previous best dataset (Leela-dfrc_n5000.binpack), with about half the dataset filtered using depth6 multipv2 search to throw away positions where either of the 2 best moves are captures
    Leela T80 Oct and Nov training data rescored with best moves, adding ~9.5 billion positions

Trained effectively the same way as the previous master net:

python3 easy_train.py \
  --experiment-name=leela-dfrc-filtered-T80-oct-nov \
  --training-dataset=/data/leela-dfrc-filtered-T80-oct-nov.binpack \
  --start-from-engine-test-net True \
  --gpus="0," \
  --start-lambda=1.0 \
  --end-lambda=0.75 \
  --gamma=0.995 \
  --lr=4.375e-4 \
  --tui=False \
  --seed=$RANDOM \
  --max_epoch=800 \
  --auto-exit-timeout-on-training-finished=900 \
  --network-testing-threads 20 \
  --num-workers 6

Local testing at a fixed 25k nodes:
experiments/experiment_leela-dfrc-filtered-T80-oct-nov/training/run_0/nn-epoch779.nnue
localElo: run_0/nn-epoch779.nnue : 4.7 +/- 3.1

The new Leela T80 part of the dataset was prepared by downloading test80 training data from all of Oct 2022 and Nov 2022, rescoring with syzygy 6-piece tablebases and ~600 GB of 7-piece tablebases, saving best moves to exported .plain files, removing all positions with castling flags, then converting to binpacks and using interleave_binpacks.py to merge them together. Scripts used in this data conversion process are available at:
https://github.com/linrock/lc0-data-converter

Filtering binpack data using depth6 multipv2 search was done by modifying transform.cpp in the tools branch:
https://github.com/linrock/Stockfish/tree/tools-filter-multipv2-no-rescore

Links for downloading the training data (total size: 338 GB) are available at:
https://robotmoon.com/nnue-training-data/

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 30544 W: 8244 L: 7947 D: 14353
Ptnml(0-2): 93, 3243, 8302, 3542, 92
https://tests.stockfishchess.org/tests/view/63a0d377264a0cf18f86f82b

Passed LTC:
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 32464 W: 8866 L: 8573 D: 15025
Ptnml(0-2): 19, 3054, 9794, 3345, 20
https://tests.stockfishchess.org/tests/view/63a10bc9fb452d3c44b1e016

closes https://github.com/official-stockfish/Stockfish/pull/4295

Bench 3554904
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index f5ac3263..d1398dd5 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-ad9b42354671.nnue"
+  #define EvalFileDefaultName   "nn-335a9b2d8a80.nnue"
 
   namespace NNUE {
 

From b2bd8699eccc149cdda1f7bf6e1eed42088ec827 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Tue, 20 Dec 2022 22:54:12 +0300
Subject: [PATCH 024/192] Update Elo estimates for terms in search

based on 25k games per term, using the UHO_XXL_+0.90_+1.19.epd book, at STC.

More detailed information in the PR.

closes https://github.com/official-stockfish/Stockfish/pull/4294

No functional change
---
 src/search.cpp | 48 ++++++++++++++++++++++++------------------------
 1 file changed, 24 insertions(+), 24 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 7f2f5284..77d23ac2 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -633,16 +633,16 @@ namespace {
         && ttValue != VALUE_NONE // Possible in case of TT access race
         && (tte->bound() & (ttValue >= beta ? BOUND_LOWER : BOUND_UPPER)))
     {
-        // If ttMove is quiet, update move sorting heuristics on TT hit (~1 Elo)
+        // If ttMove is quiet, update move sorting heuristics on TT hit (~2 Elo)
         if (ttMove)
         {
             if (ttValue >= beta)
             {
-                // Bonus for a quiet ttMove that fails high (~3 Elo)
+                // Bonus for a quiet ttMove that fails high (~2 Elo)
                 if (!ttCapture)
                     update_quiet_stats(pos, ss, ttMove, stat_bonus(depth));
 
-                // Extra penalty for early quiet moves of the previous ply (~0 Elo)
+                // Extra penalty for early quiet moves of the previous ply (~0 Elo on STC, ~2 Elo on LTC)
                 if ((ss-1)->moveCount <= 2 && !priorCapture)
                     update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, -stat_bonus(depth + 1));
             }
@@ -734,7 +734,7 @@ namespace {
         else // Fall back to (semi)classical complexity for TT hits, the NNUE complexity is lost
             complexity = abs(ss->staticEval - pos.psq_eg_stm());
 
-        // ttValue can be used as a better position evaluation (~4 Elo)
+        // ttValue can be used as a better position evaluation (~7 Elo)
         if (    ttValue != VALUE_NONE
             && (tte->bound() & (ttValue > eval ? BOUND_LOWER : BOUND_UPPER)))
             eval = ttValue;
@@ -750,7 +750,7 @@ namespace {
 
     thisThread->complexityAverage.update(complexity);
 
-    // Use static evaluation difference to improve quiet move ordering (~3 Elo)
+    // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
         int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1914, 1914);
@@ -766,7 +766,7 @@ namespace {
                   :                                    168;
     improving = improvement > 0;
 
-    // Step 7. Razoring.
+    // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
     if (eval < alpha - 369 - 254 * depth * depth)
@@ -776,7 +776,7 @@ namespace {
             return value;
     }
 
-    // Step 8. Futility pruning: child node (~25 Elo).
+    // Step 8. Futility pruning: child node (~40 Elo).
     // The depth condition is important for mate finding.
     if (   !ss->ttPv
         &&  depth < 8
@@ -785,7 +785,7 @@ namespace {
         &&  eval < 28031) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
         return eval;
 
-    // Step 9. Null move search with verification search (~22 Elo)
+    // Step 9. Null move search with verification search (~35 Elo)
     if (   !PvNode
         && (ss-1)->currentMove != MOVE_NULL
         && (ss-1)->statScore < 17139
@@ -837,7 +837,7 @@ namespace {
 
     probCutBeta = beta + 191 - 54 * improving;
 
-    // Step 10. ProbCut (~4 Elo)
+    // Step 10. ProbCut (~10 Elo)
     // If we have a good enough capture and a reduced search returns a value
     // much above beta, we can (almost) safely prune the previous move.
     if (   !PvNode
@@ -888,7 +888,7 @@ namespace {
     }
 
     // Step 11. If the position is not in TT, decrease depth by 3.
-    // Use qsearch if depth is equal or below zero (~4 Elo)
+    // Use qsearch if depth is equal or below zero (~9 Elo)
     if (    PvNode
         && !ttMove)
         depth -= 3;
@@ -903,7 +903,7 @@ namespace {
 
 moves_loop: // When in check, search starts here
 
-    // Step 12. A small Probcut idea, when we are in check (~0 Elo)
+    // Step 12. A small Probcut idea, when we are in check (~4 Elo)
     probCutBeta = beta + 417;
     if (   ss->inCheck
         && !PvNode
@@ -980,12 +980,12 @@ moves_loop: // When in check, search starts here
 
       Value delta = beta - alpha;
 
-      // Step 14. Pruning at shallow depth (~98 Elo). Depth conditions are important for mate finding.
+      // Step 14. Pruning at shallow depth (~120 Elo). Depth conditions are important for mate finding.
       if (  !rootNode
           && pos.non_pawn_material(us)
           && bestValue > VALUE_TB_LOSS_IN_MAX_PLY)
       {
-          // Skip quiet moves if movecount exceeds our FutilityMoveCount threshold (~7 Elo)
+          // Skip quiet moves if movecount exceeds our FutilityMoveCount threshold (~8 Elo)
           moveCountPruning = moveCount >= futility_move_count(improving, depth);
 
           // Reduced depth of the next LMR search
@@ -994,7 +994,7 @@ moves_loop: // When in check, search starts here
           if (   capture
               || givesCheck)
           {
-              // Futility pruning for captures (~0 Elo)
+              // Futility pruning for captures (~2 Elo)
               if (   !givesCheck
                   && !PvNode
                   && lmrDepth < 7
@@ -1003,7 +1003,7 @@ moves_loop: // When in check, search starts here
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 6 < alpha)
                   continue;
 
-              // SEE based pruning (~9 Elo)
+              // SEE based pruning (~11 Elo)
               if (!pos.see_ge(move, Value(-222) * depth))
                   continue;
           }
@@ -1020,23 +1020,23 @@ moves_loop: // When in check, search starts here
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
 
-              // Futility pruning: parent node (~9 Elo)
+              // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
                   && lmrDepth < 13
                   && ss->staticEval + 106 + 145 * lmrDepth + history / 52 <= alpha)
                   continue;
 
-              // Prune moves with negative SEE (~3 Elo)
+              // Prune moves with negative SEE (~4 Elo)
               if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
                   continue;
           }
       }
 
-      // Step 15. Extensions (~66 Elo)
+      // Step 15. Extensions (~100 Elo)
       // We take care to not overdo to avoid search getting stuck.
       if (ss->ply < thisThread->rootDepth * 2)
       {
-          // Singular extension search (~58 Elo). If all moves but one fail low on a
+          // Singular extension search (~94 Elo). If all moves but one fail low on a
           // search of (alpha-s, beta-s), and just one fails high on (alpha, beta),
           // then that move is singular and should be extended. To verify this we do
           // a reduced search on all the other moves but the ttMove and if the
@@ -1095,7 +1095,7 @@ moves_loop: // When in check, search starts here
                    && abs(ss->staticEval) > 82)
               extension = 1;
 
-          // Quiet ttMove extensions (~0 Elo)
+          // Quiet ttMove extensions (~1 Elo)
           else if (   PvNode
                    && move == ttMove
                    && move == ss->killers[0]
@@ -1166,7 +1166,7 @@ moves_loop: // When in check, search starts here
       // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
       r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
 
-      // Step 17. Late moves reduction / extension (LMR, ~98 Elo)
+      // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has
       // been searched. In general we would like to reduce them, but there are many
       // cases where we extend a son if it has good chances to be "interesting".
@@ -1462,7 +1462,7 @@ moves_loop: // When in check, search starts here
             if ((ss->staticEval = bestValue = tte->eval()) == VALUE_NONE)
                 ss->staticEval = bestValue = evaluate(pos);
 
-            // ttValue can be used as a better position evaluation (~7 Elo)
+            // ttValue can be used as a better position evaluation (~13 Elo)
             if (    ttValue != VALUE_NONE
                 && (tte->bound() & (ttValue > bestValue ? BOUND_LOWER : BOUND_UPPER)))
                 bestValue = ttValue;
@@ -1520,7 +1520,7 @@ moves_loop: // When in check, search starts here
 
       moveCount++;
 
-      // Futility pruning and moveCount pruning (~5 Elo)
+      // Futility pruning and moveCount pruning (~10 Elo)
       if (    bestValue > VALUE_TB_LOSS_IN_MAX_PLY
           && !givesCheck
           &&  to_sq(move) != prevSq
@@ -1559,7 +1559,7 @@ moves_loop: // When in check, search starts here
                                                                 [pos.moved_piece(move)]
                                                                 [to_sq(move)];
 
-      // Continuation history based pruning (~2 Elo)
+      // Continuation history based pruning (~3 Elo)
       if (   !capture
           && bestValue > VALUE_TB_LOSS_IN_MAX_PLY
           && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0

From 64656f8583dde88d558d5367b4cf8fc136c3214a Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Sat, 24 Dec 2022 13:05:24 +0800
Subject: [PATCH 025/192] Add double bonus for prior countermove fail low

Add a double extra bonus for particularly bad fail low cases. Original idea by Yoshie2000.

STC: https://tests.stockfishchess.org/tests/view/63a2f0d86b5bf07ac7fad543
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 146488 W: 38992 L: 38532 D: 68964
Ptnml(0-2): 385, 16036, 39965, 16450, 408

LTC: https://tests.stockfishchess.org/tests/view/63a3eaeb6b5bf07ac7fafdec
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 103992 W: 27853 L: 27423 D: 48716
Ptnml(0-2): 41, 10029, 31435, 10441, 50

closes https://github.com/official-stockfish/Stockfish/pull/4302

Bench: 3801857
---
 AUTHORS        | 1 +
 src/search.cpp | 9 +++++----
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 0c1c5529..70b500ea 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -145,6 +145,7 @@ Mira
 Miroslav Fontán (Hexik)
 Moez Jellouli (MJZ1977)
 Mohammed Li (tthsqe12)
+Muzhen J (XInTheDark)
 Nathan Rugg (nmrugg)
 Nick Pelling (nickpelling)
 Nicklas Persson (NicklasPersson)
diff --git a/src/search.cpp b/src/search.cpp
index 77d23ac2..0c8377d0 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1353,16 +1353,17 @@ moves_loop: // When in check, search starts here
                          quietsSearched, quietCount, capturesSearched, captureCount, depth);
 
     // Bonus for prior countermove that caused the fail low
-    else if (   (depth >= 5 || PvNode)
+    else if (   (depth >= 5 || PvNode || bestValue < alpha - 62 * depth)
              && !priorCapture)
     {
         //Assign extra bonus if current node is PvNode or cutNode
         //or fail low was really bad
         bool extraBonus =    PvNode
-                          || cutNode
-                          || bestValue < alpha - 62 * depth;
+                          || cutNode;
 
-        update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * (1 + extraBonus));
+        bool doubleExtraBonus = extraBonus && bestValue < alpha - 85 * depth;
+
+        update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * (1 + extraBonus + doubleExtraBonus));
     }
 
     if (PvNode)

From f09b391ceb5ea282538b721b069db67e3b6e098f Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Tue, 27 Dec 2022 13:33:26 +0100
Subject: [PATCH 026/192] Fix comparison with uninitialized variable

In both modified methods, the variable 'result' is checked to detect
whether the probe operation failed. However, the variable is not
initialized on all paths, so the check might test an uninitialized
value.

A test position (with TB) is given by:

position fen 3K1k2/R7/8/8/8/8/8/R6Q w - - 0 1 moves a1b1 f8g8 b1a1 g8f8 a1b1 f8g8 b1a1

This is now fixed by always initializing the variable.

closes https://github.com/official-stockfish/Stockfish/pull/4309

No functional change
---
 src/syzygy/tbprobe.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index f2de036d..4df495a8 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -1514,7 +1514,7 @@ int Tablebases::probe_dtz(Position& pos, ProbeState* result) {
 // A return value false indicates that not all probes were successful.
 bool Tablebases::root_probe(Position& pos, Search::RootMoves& rootMoves) {
 
-    ProbeState result;
+    ProbeState result = OK;
     StateInfo st;
 
     // Obtain 50-move counter for the root position
@@ -1593,7 +1593,7 @@ bool Tablebases::root_probe_wdl(Position& pos, Search::RootMoves& rootMoves) {
 
     static const int WDL_to_rank[] = { -MAX_DTZ, -MAX_DTZ + 101, 0, MAX_DTZ - 101, MAX_DTZ };
 
-    ProbeState result;
+    ProbeState result = OK;
     StateInfo st;
     WDLScore wdl;
 

From 258c13ba8cc8d054b803a04c433a766d0c5a1e8f Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Tue, 27 Dec 2022 10:31:24 +0100
Subject: [PATCH 027/192] Remove redundant extern modifier for function
 declarations

Functions have external linkage by default, so there's no need to
declare them extern.

closes https://github.com/official-stockfish/Stockfish/pull/4308

No functional change
---
 src/position.h | 2 +-
 src/psqt.h     | 2 +-
 src/uci.cpp    | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/position.h b/src/position.h
index 078ff5b7..90247cb1 100644
--- a/src/position.h
+++ b/src/position.h
@@ -204,7 +204,7 @@ private:
   bool chess960;
 };
 
-extern std::ostream& operator<<(std::ostream& os, const Position& pos);
+std::ostream& operator<<(std::ostream& os, const Position& pos);
 
 inline Color Position::side_to_move() const {
   return sideToMove;
diff --git a/src/psqt.h b/src/psqt.h
index 4ee0e379..bd78be90 100644
--- a/src/psqt.h
+++ b/src/psqt.h
@@ -30,7 +30,7 @@ namespace Stockfish::PSQT
 extern Score psq[PIECE_NB][SQUARE_NB];
 
 // Fill psqt array from a set of internally linked parameters
-extern void init();
+void init();
 
 } // namespace Stockfish::PSQT
 
diff --git a/src/uci.cpp b/src/uci.cpp
index 5d842d25..dc5f10e1 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -36,7 +36,7 @@ using namespace std;
 
 namespace Stockfish {
 
-extern vector<string> setup_bench(const Position&, istream&);
+vector<string> setup_bench(const Position&, istream&);
 
 namespace {
 

From be9bc420afa11314c886c0aede4c4ae3d76f8a50 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Thu, 29 Dec 2022 10:34:28 -0500
Subject: [PATCH 028/192] Update default net to nn-60fa44e376d9.nnue

Created by retraining the master net on the previous best dataset with additional filtering. No new data was added.

More of the Leela-dfrc_n5000.binpack part of the dataset was pre-filtered with depth6 multipv2 search to remove bestmove captures. About 93% of the previous Leela/SF data and 99% of the SF dfrc data was filtered. Unfiltered parts of the dataset were left out. The new Leela T80 oct+nov data is the same as before. All early game positions with ply count <= 28 were skipped during training by modifying the training data loader in nnue-pytorch.

Trained in a similar way as recent master nets, with a different nnue-pytorch branch for early ply skipping:

python3 easy_train.py \
  --experiment-name=leela93-dfrc99-filt-only-T80-oct-nov-skip28 \
  --training-dataset=/data/leela93-dfrc99-filt-only-T80-oct-nov.binpack \
  --start-from-engine-test-net True \
  --nnue-pytorch-branch=linrock/nnue-pytorch/misc-fixes-skip-ply-lteq-28 \
  --gpus="0," \
  --start-lambda=1.0 \
  --end-lambda=0.75 \
  --gamma=0.995 \
  --lr=4.375e-4 \
  --tui=False \
  --seed=$RANDOM \
  --max_epoch=800 \
  --network-testing-threads 20 \
  --num-workers 6

For the exact training data used: https://robotmoon.com/nnue-training-data/
Details about the previous best dataset: #4295

Local testing at a fixed 25k nodes:
experiment_leela93-dfrc99-filt-only-T80-oct-nov-skip28
Local Elo: run_0/nn-epoch779.nnue : 5.1 +/- 1.5

Passed STC
https://tests.stockfishchess.org/tests/view/63adb3acae97a464904fd4e8
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 36504 W: 9847 L: 9538 D: 17119
Ptnml(0-2): 108, 3981, 9784, 4252, 127

Passed LTC
https://tests.stockfishchess.org/tests/view/63ae0ae25bd1e5f27f13d884
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 36592 W: 10017 L: 9717 D: 16858
Ptnml(0-2): 17, 3461, 11037, 3767, 14

closes https://github.com/official-stockfish/Stockfish/pull/4314

bench 4015511
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index d1398dd5..4001166b 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-335a9b2d8a80.nnue"
+  #define EvalFileDefaultName   "nn-60fa44e376d9.nnue"
 
   namespace NNUE {
 

From b60f9cc4515cdfb657a0166abb29a60257cc59e1 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sun, 1 Jan 2023 14:45:08 +0100
Subject: [PATCH 029/192] Update copyright years

Happy New Year!

closes https://github.com/official-stockfish/Stockfish/pull/4315

No functional change
---
 src/Makefile                        | 2 +-
 src/benchmark.cpp                   | 2 +-
 src/bitbase.cpp                     | 2 +-
 src/bitboard.cpp                    | 2 +-
 src/bitboard.h                      | 2 +-
 src/endgame.cpp                     | 2 +-
 src/endgame.h                       | 2 +-
 src/evaluate.cpp                    | 2 +-
 src/evaluate.h                      | 2 +-
 src/main.cpp                        | 2 +-
 src/material.cpp                    | 2 +-
 src/material.h                      | 2 +-
 src/misc.cpp                        | 2 +-
 src/misc.h                          | 2 +-
 src/movegen.cpp                     | 2 +-
 src/movegen.h                       | 2 +-
 src/movepick.cpp                    | 2 +-
 src/movepick.h                      | 2 +-
 src/nnue/evaluate_nnue.cpp          | 2 +-
 src/nnue/evaluate_nnue.h            | 2 +-
 src/nnue/features/half_ka_v2_hm.cpp | 2 +-
 src/nnue/features/half_ka_v2_hm.h   | 2 +-
 src/nnue/layers/affine_transform.h  | 2 +-
 src/nnue/layers/clipped_relu.h      | 2 +-
 src/nnue/layers/simd.h              | 2 +-
 src/nnue/layers/sqr_clipped_relu.h  | 2 +-
 src/nnue/nnue_accumulator.h         | 2 +-
 src/nnue/nnue_architecture.h        | 2 +-
 src/nnue/nnue_common.h              | 2 +-
 src/nnue/nnue_feature_transformer.h | 2 +-
 src/pawns.cpp                       | 2 +-
 src/pawns.h                         | 2 +-
 src/position.cpp                    | 2 +-
 src/position.h                      | 2 +-
 src/psqt.cpp                        | 2 +-
 src/psqt.h                          | 2 +-
 src/search.cpp                      | 2 +-
 src/search.h                        | 2 +-
 src/syzygy/tbprobe.cpp              | 2 +-
 src/syzygy/tbprobe.h                | 2 +-
 src/thread.cpp                      | 2 +-
 src/thread.h                        | 2 +-
 src/thread_win32_osx.h              | 2 +-
 src/timeman.cpp                     | 2 +-
 src/timeman.h                       | 2 +-
 src/tt.cpp                          | 2 +-
 src/tt.h                            | 2 +-
 src/tune.cpp                        | 2 +-
 src/tune.h                          | 2 +-
 src/types.h                         | 2 +-
 src/uci.cpp                         | 2 +-
 src/uci.h                           | 2 +-
 src/ucioption.cpp                   | 2 +-
 53 files changed, 53 insertions(+), 53 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index 500c1006..d4f089ee 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -1,5 +1,5 @@
 # Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-# Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+# Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 #
 # Stockfish is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
diff --git a/src/benchmark.cpp b/src/benchmark.cpp
index e1c025ad..2abb9c8f 100644
--- a/src/benchmark.cpp
+++ b/src/benchmark.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/bitbase.cpp b/src/bitbase.cpp
index 84300baf..e21d1fe9 100644
--- a/src/bitbase.cpp
+++ b/src/bitbase.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/bitboard.cpp b/src/bitboard.cpp
index fd0ba235..0ed13fd0 100644
--- a/src/bitboard.cpp
+++ b/src/bitboard.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/bitboard.h b/src/bitboard.h
index 2b6e2a69..d4485fcb 100644
--- a/src/bitboard.h
+++ b/src/bitboard.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/endgame.cpp b/src/endgame.cpp
index e773e7a9..9021f242 100644
--- a/src/endgame.cpp
+++ b/src/endgame.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/endgame.h b/src/endgame.h
index e79f696f..c184cb3f 100644
--- a/src/endgame.h
+++ b/src/endgame.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 7619398e..3b6cbbcd 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/evaluate.h b/src/evaluate.h
index 4001166b..db9e6424 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/main.cpp b/src/main.cpp
index fad0ef84..c40e0fa3 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/material.cpp b/src/material.cpp
index 1567358a..7102f879 100644
--- a/src/material.cpp
+++ b/src/material.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/material.h b/src/material.h
index 3ca169ce..f6db85c4 100644
--- a/src/material.h
+++ b/src/material.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/misc.cpp b/src/misc.cpp
index c7fa0e9a..5bb8da69 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/misc.h b/src/misc.h
index 77b81d50..4c1150f8 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/movegen.cpp b/src/movegen.cpp
index c7a3c29b..a960f863 100644
--- a/src/movegen.cpp
+++ b/src/movegen.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/movegen.h b/src/movegen.h
index bbb35b39..b8df3e65 100644
--- a/src/movegen.h
+++ b/src/movegen.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/movepick.cpp b/src/movepick.cpp
index 564adc28..dbe67357 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/movepick.h b/src/movepick.h
index e4c4a5bf..90f60b8a 100644
--- a/src/movepick.h
+++ b/src/movepick.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index 4715fed0..8d720ccb 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/evaluate_nnue.h b/src/nnue/evaluate_nnue.h
index 2e4f1f50..9499f7d9 100644
--- a/src/nnue/evaluate_nnue.h
+++ b/src/nnue/evaluate_nnue.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/features/half_ka_v2_hm.cpp b/src/nnue/features/half_ka_v2_hm.cpp
index 7dbd3415..19ebb15f 100644
--- a/src/nnue/features/half_ka_v2_hm.cpp
+++ b/src/nnue/features/half_ka_v2_hm.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/features/half_ka_v2_hm.h b/src/nnue/features/half_ka_v2_hm.h
index a95d4328..78063c36 100644
--- a/src/nnue/features/half_ka_v2_hm.h
+++ b/src/nnue/features/half_ka_v2_hm.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index 461a7b83..710ab8a7 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/layers/clipped_relu.h b/src/nnue/layers/clipped_relu.h
index f94d3082..51e562da 100644
--- a/src/nnue/layers/clipped_relu.h
+++ b/src/nnue/layers/clipped_relu.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/layers/simd.h b/src/nnue/layers/simd.h
index 7b9e8fb2..aeab39c4 100644
--- a/src/nnue/layers/simd.h
+++ b/src/nnue/layers/simd.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/layers/sqr_clipped_relu.h b/src/nnue/layers/sqr_clipped_relu.h
index b603a277..df539b39 100644
--- a/src/nnue/layers/sqr_clipped_relu.h
+++ b/src/nnue/layers/sqr_clipped_relu.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/nnue_accumulator.h b/src/nnue/nnue_accumulator.h
index 600483b5..8eba4497 100644
--- a/src/nnue/nnue_accumulator.h
+++ b/src/nnue/nnue_accumulator.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/nnue_architecture.h b/src/nnue/nnue_architecture.h
index cac83730..c43a23c3 100644
--- a/src/nnue/nnue_architecture.h
+++ b/src/nnue/nnue_architecture.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/nnue_common.h b/src/nnue/nnue_common.h
index 17956189..12309d26 100644
--- a/src/nnue/nnue_common.h
+++ b/src/nnue/nnue_common.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index b6dd54d3..62f1615d 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/pawns.cpp b/src/pawns.cpp
index fdcfa022..0ccafd9e 100644
--- a/src/pawns.cpp
+++ b/src/pawns.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/pawns.h b/src/pawns.h
index af0370fc..95c9ea3c 100644
--- a/src/pawns.h
+++ b/src/pawns.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/position.cpp b/src/position.cpp
index 5befcaf2..e82425af 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/position.h b/src/position.h
index 90247cb1..3de24f22 100644
--- a/src/position.h
+++ b/src/position.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/psqt.cpp b/src/psqt.cpp
index ca5664c2..d3ebb20d 100644
--- a/src/psqt.cpp
+++ b/src/psqt.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/psqt.h b/src/psqt.h
index bd78be90..9630f446 100644
--- a/src/psqt.h
+++ b/src/psqt.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/search.cpp b/src/search.cpp
index 0c8377d0..dd240366 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/search.h b/src/search.h
index b620202d..48a0f7ce 100644
--- a/src/search.h
+++ b/src/search.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index 4df495a8..b7ba3240 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/syzygy/tbprobe.h b/src/syzygy/tbprobe.h
index 179c7572..159c6865 100644
--- a/src/syzygy/tbprobe.h
+++ b/src/syzygy/tbprobe.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/thread.cpp b/src/thread.cpp
index e8723eb7..7e71edf1 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/thread.h b/src/thread.h
index 5f0b2c3e..680da209 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/thread_win32_osx.h b/src/thread_win32_osx.h
index 77d1c3c7..01ff1c77 100644
--- a/src/thread_win32_osx.h
+++ b/src/thread_win32_osx.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/timeman.cpp b/src/timeman.cpp
index cab0d767..5c826b4f 100644
--- a/src/timeman.cpp
+++ b/src/timeman.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/timeman.h b/src/timeman.h
index a86f0769..3462b823 100644
--- a/src/timeman.h
+++ b/src/timeman.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/tt.cpp b/src/tt.cpp
index c7118aea..39f18d3d 100644
--- a/src/tt.cpp
+++ b/src/tt.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/tt.h b/src/tt.h
index 03fe3e14..3e335b44 100644
--- a/src/tt.h
+++ b/src/tt.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/tune.cpp b/src/tune.cpp
index a885845f..41f6664d 100644
--- a/src/tune.cpp
+++ b/src/tune.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/tune.h b/src/tune.h
index 75ab484a..f5b787af 100644
--- a/src/tune.h
+++ b/src/tune.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/types.h b/src/types.h
index 29c16ce7..8a021342 100644
--- a/src/types.h
+++ b/src/types.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/uci.cpp b/src/uci.cpp
index dc5f10e1..c49b9b78 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/uci.h b/src/uci.h
index 3b5a6764..bd3df323 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
diff --git a/src/ucioption.cpp b/src/ucioption.cpp
index 9fb48345..78711c18 100644
--- a/src/ucioption.cpp
+++ b/src/ucioption.cpp
@@ -1,6 +1,6 @@
 /*
   Stockfish, a UCI chess playing engine derived from Glaurung 2.1
-  Copyright (C) 2004-2022 The Stockfish developers (see AUTHORS file)
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
 
   Stockfish is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by

From a6fa683418c6a74491035601e16497851eea6be6 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Thu, 29 Dec 2022 10:34:28 -0500
Subject: [PATCH 030/192] Update default net to nn-a3dc078bafc7.nnue

This is a later epoch (epoch 859) from the same experiment run that trained yesterday's master net nn-60fa44e376d9.nnue (epoch 779). The experiment was manually paused around epoch 790 and unpaused with max epoch increased to 900 mainly to get more local elo data without letting the GPU idle.

nn-60fa44e376d9.nnue is from #4314
nn-335a9b2d8a80.nnue is from #4295

Local elo vs. nn-335a9b2d8a80.nnue at 25k nodes per move:
experiment_leela93-dfrc99-filt-only-T80-oct-nov-skip28
run_0/nn-epoch779.nnue (nn-60fa44e376d9.nnue) : 5.0 +/- 1.2
run_0/nn-epoch859.nnue (nn-a3dc078bafc7.nnue) : 5.6 +/- 1.6

Passed STC vs. nn-335a9b2d8a80.nnue
https://tests.stockfishchess.org/tests/view/63ae10495bd1e5f27f13d94f
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 37536 W: 10088 L: 9781 D: 17667
Ptnml(0-2): 110, 4006, 10223, 4325, 104

An LTC test vs. nn-335a9b2d8a80.nnue was paused due to nn-60fa44e376d9.nnue passing LTC first:
https://tests.stockfishchess.org/tests/view/63ae5d34331d5fca5113703b

Passed LTC vs. nn-60fa44e376d9.nnue
https://tests.stockfishchess.org/tests/view/63af1e41465d2b022dbce4e7
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 148704 W: 39672 L: 39155 D: 69877
Ptnml(0-2): 59, 14443, 44843, 14936, 71

closes https://github.com/official-stockfish/Stockfish/pull/4319

bench 3984365
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index db9e6424..0a00faa7 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-60fa44e376d9.nnue"
+  #define EvalFileDefaultName   "nn-a3dc078bafc7.nnue"
 
   namespace NNUE {
 

From fc5b59b88bae00b7e74bbad0de7c3c33136937cf Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Mon, 2 Jan 2023 17:19:41 +0300
Subject: [PATCH 031/192] Parameter Tweaks

This patch is a parameter tweak that passed both STC and LTC tests.

STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 80944 W: 21557 L: 21189 D: 38198
Ptnml(0-2): 192, 8883, 22028, 9103, 266
https://tests.stockfishchess.org/tests/view/63b07fe2d421d8f75795a03b

LTC:
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 30440 W: 8296 L: 8007 D: 14137
Ptnml(0-2): 6, 2893, 9143, 3162, 16
https://tests.stockfishchess.org/tests/view/63b167d02ab1290f961644db

closes https://github.com/official-stockfish/Stockfish/pull/4318

Bench: 4182223
---
 src/evaluate.cpp | 12 +++++-----
 src/search.cpp   | 60 ++++++++++++++++++++++++------------------------
 2 files changed, 36 insertions(+), 36 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 3b6cbbcd..d5cda3d8 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1056,7 +1056,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   // We use the much less accurate but faster Classical eval when the NNUE
   // option is set to false. Otherwise we use the NNUE eval unless the
   // PSQ advantage is decisive and several pieces remain. (~3 Elo)
-  bool useClassical = !useNNUE || (pos.count<ALL_PIECES>() > 7 && abs(psq) > 1760);
+  bool useClassical = !useNNUE || (pos.count<ALL_PIECES>() > 7 && abs(psq) > 1781);
 
   if (useClassical)
       v = Evaluation<NO_TRACE>(pos).value();
@@ -1071,8 +1071,8 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  412 * nnueComplexity
-                        + 428 * abs(psq - nnue)
+      nnueComplexity = (  406 * nnueComplexity
+                        + 424 * abs(psq - nnue)
                         + (optimism  > 0 ? int(optimism) * int(psq - nnue) : 0)
                         ) / 1024;
 
@@ -1080,12 +1080,12 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       if (complexity)
           *complexity = nnueComplexity;
 
-      optimism = optimism * (278 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 755)) / 1024;
+      optimism = optimism * (272 + nnueComplexity) / 256;
+      v = (nnue * scale + optimism * (scale - 748)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling
-  v = v * (197 - pos.rule50_count()) / 214;
+  v = v * (200 - pos.rule50_count()) / 214;
 
   // Guarantee evaluation does not hit the tablebase range
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
diff --git a/src/search.cpp b/src/search.cpp
index dd240366..d0ed44fa 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -63,7 +63,7 @@ namespace {
 
   // Futility margin
   Value futility_margin(Depth d, bool improving) {
-    return Value(165 * (d - improving));
+    return Value(158 * (d - improving));
   }
 
   // Reductions lookup table, initialized at startup
@@ -71,7 +71,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1642 - int(delta) * 1024 / int(rootDelta)) / 1024 + (!i && r > 916);
+    return (r + 1460 - int(delta) * 1024 / int(rootDelta)) / 1024 + (!i && r > 937);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min((12 * d + 282) * d - 349 , 1480);
+    return std::min((11 * d + 284) * d - 363 , 1650);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -307,7 +307,7 @@ void Thread::search() {
 
   multiPV = std::min(multiPV, rootMoves.size());
 
-  complexityAverage.set(155, 1);
+  complexityAverage.set(153, 1);
 
   optimism[us] = optimism[~us] = VALUE_ZERO;
 
@@ -351,12 +351,12 @@ void Thread::search() {
           if (rootDepth >= 4)
           {
               Value prev = rootMoves[pvIdx].averageScore;
-              delta = Value(10) + int(prev) * prev / 15620;
+              delta = Value(10) + int(prev) * prev / 15400;
               alpha = std::max(prev - delta,-VALUE_INFINITE);
               beta  = std::min(prev + delta, VALUE_INFINITE);
 
               // Adjust optimism based on root move's previousScore
-              int opt = 118 * prev / (std::abs(prev) + 169);
+              int opt = 116 * prev / (std::abs(prev) + 170);
               optimism[ us] = Value(opt);
               optimism[~us] = -optimism[us];
           }
@@ -753,7 +753,7 @@ namespace {
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
-        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1914, 1914);
+        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1940, 1940);
         thisThread->mainHistory[~us][from_to((ss-1)->currentMove)] << bonus;
     }
 
@@ -763,13 +763,13 @@ namespace {
     // margin and the improving flag are used in various pruning heuristics.
     improvement =   (ss-2)->staticEval != VALUE_NONE ? ss->staticEval - (ss-2)->staticEval
                   : (ss-4)->staticEval != VALUE_NONE ? ss->staticEval - (ss-4)->staticEval
-                  :                                    168;
+                  :                                    172;
     improving = improvement > 0;
 
     // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
-    if (eval < alpha - 369 - 254 * depth * depth)
+    if (eval < alpha - 394 - 255 * depth * depth)
     {
         value = qsearch<NonPV>(pos, ss, alpha - 1, alpha);
         if (value < alpha)
@@ -780,18 +780,18 @@ namespace {
     // The depth condition is important for mate finding.
     if (   !ss->ttPv
         &&  depth < 8
-        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 303 >= beta
+        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 304 >= beta
         &&  eval >= beta
-        &&  eval < 28031) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
+        &&  eval < 28580) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
         return eval;
 
     // Step 9. Null move search with verification search (~35 Elo)
     if (   !PvNode
         && (ss-1)->currentMove != MOVE_NULL
-        && (ss-1)->statScore < 17139
+        && (ss-1)->statScore < 18200
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 233 + complexity / 25
+        &&  ss->staticEval >= beta - 20 * depth - improvement / 14 + 235 + complexity / 24
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly || us != thisThread->nmpColor))
@@ -799,7 +799,7 @@ namespace {
         assert(eval - beta >= 0);
 
         // Null move dynamic reduction based on depth, eval and complexity of position
-        Depth R = std::min(int(eval - beta) / 168, 7) + depth / 3 + 4 - (complexity > 861);
+        Depth R = std::min(int(eval - beta) / 165, 6) + depth / 3 + 4 - (complexity > 800);
 
         ss->currentMove = MOVE_NULL;
         ss->continuationHistory = &thisThread->continuationHistory[0][0][NO_PIECE][0];
@@ -835,7 +835,7 @@ namespace {
         }
     }
 
-    probCutBeta = beta + 191 - 54 * improving;
+    probCutBeta = beta + 180 - 54 * improving;
 
     // Step 10. ProbCut (~10 Elo)
     // If we have a good enough capture and a reduced search returns a value
@@ -904,7 +904,7 @@ namespace {
 moves_loop: // When in check, search starts here
 
     // Step 12. A small Probcut idea, when we are in check (~4 Elo)
-    probCutBeta = beta + 417;
+    probCutBeta = beta + 402;
     if (   ss->inCheck
         && !PvNode
         && depth >= 2
@@ -999,12 +999,12 @@ moves_loop: // When in check, search starts here
                   && !PvNode
                   && lmrDepth < 7
                   && !ss->inCheck
-                  && ss->staticEval + 180 + 201 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
+                  && ss->staticEval + 185 + 203 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 6 < alpha)
                   continue;
 
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, Value(-222) * depth))
+              if (!pos.see_ge(move, Value(-220) * depth))
                   continue;
           }
           else
@@ -1015,7 +1015,7 @@ moves_loop: // When in check, search starts here
 
               // Continuation history based pruning (~2 Elo)
               if (   lmrDepth < 5
-                  && history < -3875 * (depth - 1))
+                  && history < -4180 * (depth - 1))
                   continue;
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
@@ -1023,11 +1023,11 @@ moves_loop: // When in check, search starts here
               // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
                   && lmrDepth < 13
-                  && ss->staticEval + 106 + 145 * lmrDepth + history / 52 <= alpha)
+                  && ss->staticEval + 103 + 136 * lmrDepth + history / 53 <= alpha)
                   continue;
 
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
+              if (!pos.see_ge(move, Value(-25 * lmrDepth * lmrDepth - 16 * lmrDepth)))
                   continue;
           }
       }
@@ -1092,14 +1092,14 @@ moves_loop: // When in check, search starts here
           // Check extensions (~1 Elo)
           else if (   givesCheck
                    && depth > 9
-                   && abs(ss->staticEval) > 82)
+                   && abs(ss->staticEval) > 78)
               extension = 1;
 
           // Quiet ttMove extensions (~1 Elo)
           else if (   PvNode
                    && move == ttMove
                    && move == ss->killers[0]
-                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5177)
+                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5600)
               extension = 1;
       }
 
@@ -1161,10 +1161,10 @@ moves_loop: // When in check, search starts here
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]
                      + (*contHist[3])[movedPiece][to_sq(move)]
-                     - 4433;
+                     - 4467;
 
       // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-      r -= ss->statScore / (13000 + 4152 * (depth > 7 && depth < 19));
+      r -= ss->statScore / (12800 + 4410 * (depth > 7 && depth < 19));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has
@@ -1188,7 +1188,7 @@ moves_loop: // When in check, search starts here
           {
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
-              const bool doDeeperSearch = value > (alpha + 64 + 11 * (newDepth - d));
+              const bool doDeeperSearch = value > (alpha + 66 + 11 * (newDepth - d));
               const bool doEvenDeeperSearch = value > alpha + 582 && ss->doubleExtensions <= 5;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
@@ -1353,7 +1353,7 @@ moves_loop: // When in check, search starts here
                          quietsSearched, quietCount, capturesSearched, captureCount, depth);
 
     // Bonus for prior countermove that caused the fail low
-    else if (   (depth >= 5 || PvNode || bestValue < alpha - 62 * depth)
+    else if (   (depth >= 5 || PvNode || bestValue < alpha - 65 * depth)
              && !priorCapture)
     {
         //Assign extra bonus if current node is PvNode or cutNode
@@ -1361,7 +1361,7 @@ moves_loop: // When in check, search starts here
         bool extraBonus =    PvNode
                           || cutNode;
 
-        bool doubleExtraBonus = extraBonus && bestValue < alpha - 85 * depth;
+        bool doubleExtraBonus = extraBonus && bestValue < alpha - 88 * depth;
 
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * (1 + extraBonus + doubleExtraBonus));
     }
@@ -1488,7 +1488,7 @@ moves_loop: // When in check, search starts here
         if (PvNode && bestValue > alpha)
             alpha = bestValue;
 
-        futilityBase = bestValue + 153;
+        futilityBase = bestValue + 158;
     }
 
     const PieceToHistory* contHist[] = { (ss-1)->continuationHistory, (ss-2)->continuationHistory,
@@ -1690,7 +1690,7 @@ moves_loop: // When in check, search starts here
 
     if (!pos.capture(bestMove))
     {
-        int bonus2 = bestValue > beta + 137 ? bonus1               // larger bonus
+        int bonus2 = bestValue > beta + 146 ? bonus1               // larger bonus
                                             : stat_bonus(depth);   // smaller bonus
 
         // Increase stats for the best move in case it was a quiet move

From 9fe9ff00823a0d1e989fcfc45410396e94345987 Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Tue, 27 Dec 2022 16:44:32 -0800
Subject: [PATCH 032/192] Fix stack initialization

This fixes a bug where on line 278 the Stack::staticEvals are
initialized to 0. However VALUE_NONE is defined to be 32002 so
this is a bug in master. It is probably due to the calculation
of improvement, where staticEval prior to rootPos can be accessed.

https://tests.stockfishchess.org/tests/view/63ab91cf39af998100ce1666
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 53736 W: 14285 L: 13955 D: 25496
Ptnml(0-2): 121, 5921, 14500, 6159, 167

https://tests.stockfishchess.org/tests/view/63b2af5ee28ed36c814bed52
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 33776 W: 9130 L: 8934 D: 15712
Ptnml(0-2): 14, 3240, 10185, 3434, 15

closes https://github.com/official-stockfish/Stockfish/pull/4320

Bench: 4068510
---
 src/search.cpp | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index d0ed44fa..d27b21d2 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -276,8 +276,11 @@ void Thread::search() {
   int iterIdx = 0;
 
   std::memset(ss-7, 0, 10 * sizeof(Stack));
-  for (int i = 7; i > 0; i--)
+  for (int i = 7; i > 0; --i)
+  {
       (ss-i)->continuationHistory = &this->continuationHistory[0][0][NO_PIECE][0]; // Use as a sentinel
+      (ss-i)->staticEval = VALUE_NONE;
+  }
 
   for (int i = 0; i <= MAX_PLY + 2; ++i)
       (ss+i)->ply = i;

From ea0f34120fbc1ab7aa8047a1319714b87a1f3b87 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Sat, 24 Dec 2022 09:43:22 -0600
Subject: [PATCH 033/192] Late countermove bonus: remove "extraBonus &&"

passed stc: https://tests.stockfishchess.org/tests/view/63a71e409c0589b83751dc25
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 432480 W: 113846 L: 114055 D: 204579
Ptnml(0-2): 1164, 48205, 117701, 48016, 1154

passed ltc: https://tests.stockfishchess.org/tests/view/63aba66639af998100ce1aa9
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 245344 W: 65309 L: 65317 D: 114718
Ptnml(0-2): 117, 24257, 73903, 24307, 88

closes https://github.com/official-stockfish/Stockfish/pull/4322

bench 4379218
---
 src/search.cpp | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index d27b21d2..55af6aba 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1359,14 +1359,9 @@ moves_loop: // When in check, search starts here
     else if (   (depth >= 5 || PvNode || bestValue < alpha - 65 * depth)
              && !priorCapture)
     {
-        //Assign extra bonus if current node is PvNode or cutNode
-        //or fail low was really bad
-        bool extraBonus =    PvNode
-                          || cutNode;
-
-        bool doubleExtraBonus = extraBonus && bestValue < alpha - 88 * depth;
-
-        update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * (1 + extraBonus + doubleExtraBonus));
+        // Extra bonuses for PV/Cut nodes or bad fail lows
+        int bonus = 1 + (PvNode || cutNode) + (bestValue < alpha - 88 * depth);
+        update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 
     if (PvNode)

From 4101893a28b4d34262ed1fba10f2365e964b6537 Mon Sep 17 00:00:00 2001
From: candirufish <38038147+candirufish@users.noreply.github.com>
Date: Mon, 9 Jan 2023 14:26:51 +0100
Subject: [PATCH 034/192] On step 18 increase reduction by 2 if not ttmove and
 cutnode

stc: https://tests.stockfishchess.org/tests/view/63babc9fcd3db0c8d399f723
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 43104 W: 11711 L: 11389 D: 20004
Ptnml(0-2): 211, 4518, 11793, 4798, 232

ltc: https://tests.stockfishchess.org/tests/view/63bb1857cd3db0c8d39a0661
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 127104 W: 33810 L: 33339 D: 59955
Ptnml(0-2): 39, 12155, 38702, 12608, 48

closes https://github.com/official-stockfish/Stockfish/pull/4334

Bench: 4035725
---
 src/search.cpp | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index 55af6aba..f5eeb23b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1215,6 +1215,10 @@ moves_loop: // When in check, search starts here
       // Step 18. Full depth search when LMR is skipped. If expected reduction is high, reduce its depth by 1.
       else if (!PvNode || moveCount > 1)
       {
+               // Increase reduction for cut nodes and not ttMove (~1 Elo)
+               if (!ttMove && cutNode)
+                         r += 2;
+
                value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 4), !cutNode);
       }
 

From fcee83810a5bcf774d3e13cf9e65ccf3b29e3319 Mon Sep 17 00:00:00 2001
From: Jake Senne <26696846+w1wwwwww@users.noreply.github.com>
Date: Tue, 3 Jan 2023 15:16:45 -0600
Subject: [PATCH 035/192] Only close file if already open

Ensures that the tablebase file is only closed if already open.

Fixes #4268
Closes https://github.com/official-stockfish/Stockfish/pull/4321

No functional change
---
 AUTHORS                | 1 +
 src/syzygy/tbprobe.cpp | 9 +++------
 2 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 70b500ea..998399b9 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -92,6 +92,7 @@ Ivan Ivec (IIvec)
 Jacques B. (Timshel)
 Jan Ondruš (hxim)
 Jared Kish (Kurtbusch, kurt22i)
+Jake Senne (w1wwwwww)
 Jarrod Torriero (DU-jdto)
 Jean Gauthier (OuaisBla)
 Jean-Francois Romang (jromang)
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index b7ba3240..cc5e2852 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -199,13 +199,10 @@ public:
         }
     }
 
-    // Memory map the file and check it. File should be already open and will be
-    // closed after mapping.
+    // Memory map the file and check it.
     uint8_t* map(void** baseAddress, uint64_t* mapping, TBType type) {
-
-        assert(is_open());
-
-        close(); // Need to re-open to get native file descriptor
+        if (is_open())
+            close(); // Need to re-open to get native file descriptor
 
 #ifndef _WIN32
         struct stat statbuf;

From 31acd6bab70f4661316986c2c93163d39736fd61 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Tue, 27 Dec 2022 21:06:10 +0100
Subject: [PATCH 036/192] Warn if a global function has no previous declaration

If a global function has no previous declaration, either the declaration
is missing in the corresponding header file or the function should be
declared static. Static functions are local to the translation unit,
which allows the compiler to apply some optimizations earlier (when
compiling the translation unit rather than during link-time
optimization).

The commit enables the warning for gcc, clang, and mingw. It also fixes
the reported warnings by declaring the functions static or by adding a
header file (benchmark.h).

closes https://github.com/official-stockfish/Stockfish/pull/4325

No functional change
---
 src/Makefile               |  6 +++---
 src/benchmark.cpp          |  2 ++
 src/benchmark.h            | 34 ++++++++++++++++++++++++++++++++++
 src/evaluate.cpp           | 10 +++++-----
 src/misc.cpp               |  2 +-
 src/nnue/evaluate_nnue.cpp | 10 +++++-----
 src/uci.cpp                |  3 +--
 src/ucioption.cpp          | 14 +++++++-------
 8 files changed, 58 insertions(+), 23 deletions(-)
 create mode 100644 src/benchmark.h

diff --git a/src/Makefile b/src/Makefile
index d4f089ee..30c1be5e 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -366,7 +366,7 @@ endif
 ifeq ($(COMP),gcc)
 	comp=gcc
 	CXX=g++
-	CXXFLAGS += -pedantic -Wextra -Wshadow
+	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-declarations
 
 	ifeq ($(arch),$(filter $(arch),armv7 armv8 riscv64))
 		ifeq ($(OS),Android)
@@ -410,7 +410,7 @@ ifeq ($(COMP),mingw)
 			CXX=i686-w64-mingw32-c++-posix
 		endif
 	endif
-	CXXFLAGS += -pedantic -Wextra -Wshadow
+	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-declarations
 endif
 
 ifeq ($(COMP),icc)
@@ -426,7 +426,7 @@ ifeq ($(COMP),clang)
 		CXX=x86_64-w64-mingw32-clang++
 	endif
 
-	CXXFLAGS += -pedantic -Wextra -Wshadow
+	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-prototypes
 
 	ifeq ($(filter $(KERNEL),Darwin OpenBSD FreeBSD),)
 	ifeq ($(target_windows),)
diff --git a/src/benchmark.cpp b/src/benchmark.cpp
index 2abb9c8f..a1ad0550 100644
--- a/src/benchmark.cpp
+++ b/src/benchmark.cpp
@@ -16,6 +16,8 @@
   along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */
 
+#include "benchmark.h"
+
 #include <fstream>
 #include <iostream>
 #include <istream>
diff --git a/src/benchmark.h b/src/benchmark.h
new file mode 100644
index 00000000..64acf833
--- /dev/null
+++ b/src/benchmark.h
@@ -0,0 +1,34 @@
+/*
+  Stockfish, a UCI chess playing engine derived from Glaurung 2.1
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
+
+  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 BENCHMARK_H_INCLUDED
+#define BENCHMARK_H_INCLUDED
+
+#include <iosfwd>
+#include <string>
+#include <vector>
+
+namespace Stockfish {
+
+class Position;
+
+std::vector<std::string> setup_bench(const Position&, std::istream&);
+
+} // namespace Stockfish
+
+#endif // #ifndef BENCHMARK_H_INCLUDED
diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index d5cda3d8..3aae9f1c 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -159,24 +159,24 @@ namespace Trace {
 
   Score scores[TERM_NB][COLOR_NB];
 
-  double to_cp(Value v) { return double(v) / UCI::NormalizeToPawnValue; }
+  static double to_cp(Value v) { return double(v) / UCI::NormalizeToPawnValue; }
 
-  void add(int idx, Color c, Score s) {
+  static void add(int idx, Color c, Score s) {
     scores[idx][c] = s;
   }
 
-  void add(int idx, Score w, Score b = SCORE_ZERO) {
+  static void add(int idx, Score w, Score b = SCORE_ZERO) {
     scores[idx][WHITE] = w;
     scores[idx][BLACK] = b;
   }
 
-  std::ostream& operator<<(std::ostream& os, Score s) {
+  static std::ostream& operator<<(std::ostream& os, Score s) {
     os << std::setw(5) << to_cp(mg_value(s)) << " "
        << std::setw(5) << to_cp(eg_value(s));
     return os;
   }
 
-  std::ostream& operator<<(std::ostream& os, Term t) {
+  static std::ostream& operator<<(std::ostream& os, Term t) {
 
     if (t == MATERIAL || t == IMBALANCE || t == WINNABLE || t == TOTAL)
         os << " ----  ----"    << " | " << " ----  ----";
diff --git a/src/misc.cpp b/src/misc.cpp
index 5bb8da69..1cb515c3 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -517,7 +517,7 @@ void bindThisThread(size_t) {}
 /// API and returns the best node id for the thread with index idx. Original
 /// code from Texel by Peter Österlund.
 
-int best_node(size_t idx) {
+static int best_node(size_t idx) {
 
   int threads = 0;
   int nodes = 0;
diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index 8d720ccb..a06c1978 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -83,7 +83,7 @@ namespace Stockfish::Eval::NNUE {
   }  // namespace Detail
 
   // Initialize the evaluation function parameters
-  void initialize() {
+  static void initialize() {
 
     Detail::initialize(featureTransformer);
     for (std::size_t i = 0; i < LayerStacks; ++i)
@@ -91,7 +91,7 @@ namespace Stockfish::Eval::NNUE {
   }
 
   // Read network header
-  bool read_header(std::istream& stream, std::uint32_t* hashValue, std::string* desc)
+  static bool read_header(std::istream& stream, std::uint32_t* hashValue, std::string* desc)
   {
     std::uint32_t version, size;
 
@@ -105,7 +105,7 @@ namespace Stockfish::Eval::NNUE {
   }
 
   // Write network header
-  bool write_header(std::ostream& stream, std::uint32_t hashValue, const std::string& desc)
+  static bool write_header(std::ostream& stream, std::uint32_t hashValue, const std::string& desc)
   {
     write_little_endian<std::uint32_t>(stream, Version);
     write_little_endian<std::uint32_t>(stream, hashValue);
@@ -115,7 +115,7 @@ namespace Stockfish::Eval::NNUE {
   }
 
   // Read network parameters
-  bool read_parameters(std::istream& stream) {
+  static bool read_parameters(std::istream& stream) {
 
     std::uint32_t hashValue;
     if (!read_header(stream, &hashValue, &netDescription)) return false;
@@ -127,7 +127,7 @@ namespace Stockfish::Eval::NNUE {
   }
 
   // Write network parameters
-  bool write_parameters(std::ostream& stream) {
+  static bool write_parameters(std::ostream& stream) {
 
     if (!write_header(stream, HashValue, netDescription)) return false;
     if (!Detail::write_parameters(stream, *featureTransformer)) return false;
diff --git a/src/uci.cpp b/src/uci.cpp
index c49b9b78..eb158e72 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -22,6 +22,7 @@
 #include <sstream>
 #include <string>
 
+#include "benchmark.h"
 #include "evaluate.h"
 #include "movegen.h"
 #include "position.h"
@@ -36,8 +37,6 @@ using namespace std;
 
 namespace Stockfish {
 
-vector<string> setup_bench(const Position&, istream&);
-
 namespace {
 
   // FEN string for the initial position in standard chess
diff --git a/src/ucioption.cpp b/src/ucioption.cpp
index 78711c18..b4ce70b4 100644
--- a/src/ucioption.cpp
+++ b/src/ucioption.cpp
@@ -38,13 +38,13 @@ UCI::OptionsMap Options; // Global object
 namespace UCI {
 
 /// '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(size_t(o)); }
-void on_logger(const Option& o) { start_logger(o); }
-void on_threads(const Option& o) { Threads.set(size_t(o)); }
-void on_tb_path(const Option& o) { Tablebases::init(o); }
-void on_use_NNUE(const Option& ) { Eval::NNUE::init(); }
-void on_eval_file(const Option& ) { Eval::NNUE::init(); }
+static void on_clear_hash(const Option&) { Search::clear(); }
+static void on_hash_size(const Option& o) { TT.resize(size_t(o)); }
+static void on_logger(const Option& o) { start_logger(o); }
+static void on_threads(const Option& o) { Threads.set(size_t(o)); }
+static void on_tb_path(const Option& o) { Tablebases::init(o); }
+static void on_use_NNUE(const Option&) { Eval::NNUE::init(); }
+static void on_eval_file(const Option&) { Eval::NNUE::init(); }
 
 /// Our case insensitive less() function as required by UCI protocol
 bool CaseInsensitiveLess::operator() (const string& s1, const string& s2) const {

From 5a88c5bb9b3e5ee431ac85abb8981b1571b68b2d Mon Sep 17 00:00:00 2001
From: Stefano Di Martino <stefano.d@posteo.de>
Date: Sat, 7 Jan 2023 01:08:30 +0100
Subject: [PATCH 037/192] Modernize code base a little bit

Removed sprintf() which generated a warning, because of security reasons.
Replace NULL with nullptr
Replace typedef with using
Do not inherit from std::vector. Use composition instead.
optimize mutex-unlocking

closes https://github.com/official-stockfish/Stockfish/pull/4327

No functional change
---
 AUTHORS                    |  1 +
 src/material.h             |  2 +-
 src/misc.cpp               |  8 ++++----
 src/nnue/evaluate_nnue.cpp | 31 ++++++++++++-------------------
 src/thread.cpp             | 38 +++++++++++++++++++-------------------
 src/thread.h               | 14 +++++++++++---
 src/thread_win32_osx.h     |  6 +++---
 7 files changed, 51 insertions(+), 49 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 998399b9..87fed7b8 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -192,6 +192,7 @@ Shawn Varghese (xXH4CKST3RXx)
 Siad Daboul (Topologist)
 Stefan Geschwentner (locutus2)
 Stefano Cardanobile (Stefano80)
+Stefano Di Martino (StefanoD)
 Steinar Gunderson (sesse)
 Stéphane Nicolet (snicolet)
 Syine Mineta (MinetaS)
diff --git a/src/material.h b/src/material.h
index f6db85c4..73c83100 100644
--- a/src/material.h
+++ b/src/material.h
@@ -28,7 +28,7 @@ namespace Stockfish::Material {
 
 /// Material::Entry contains various information about a material configuration.
 /// 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 nullptr, meaning that the
 /// standard evaluation function will be used), and scale factors.
 ///
 /// The scale factors are used to scale the evaluation score up or down. For
diff --git a/src/misc.cpp b/src/misc.cpp
index 1cb515c3..be7abba5 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -413,7 +413,7 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
   if (!OpenProcessToken(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
       return nullptr;
 
-  if (LookupPrivilegeValue(NULL, SE_LOCK_MEMORY_NAME, &luid))
+  if (LookupPrivilegeValue(nullptr, SE_LOCK_MEMORY_NAME, &luid))
   {
       TOKEN_PRIVILEGES tp { };
       TOKEN_PRIVILEGES prevTp { };
@@ -432,10 +432,10 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
           // Round up size to full pages and allocate
           allocSize = (allocSize + largePageSize - 1) & ~size_t(largePageSize - 1);
           mem = VirtualAlloc(
-              NULL, allocSize, MEM_RESERVE | MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
+              nullptr, allocSize, MEM_RESERVE | MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
 
           // Privilege no longer needed, restore previous state
-          AdjustTokenPrivileges(hProcessToken, FALSE, &prevTp, 0, NULL, NULL);
+          AdjustTokenPrivileges(hProcessToken, FALSE, &prevTp, 0, nullptr, nullptr);
       }
   }
 
@@ -453,7 +453,7 @@ void* aligned_large_pages_alloc(size_t allocSize) {
 
   // Fall back to regular, page aligned, allocation if necessary
   if (!mem)
-      mem = VirtualAlloc(NULL, allocSize, MEM_RESERVE | MEM_COMMIT, PAGE_READWRITE);
+      mem = VirtualAlloc(nullptr, allocSize, MEM_RESERVE | MEM_COMMIT, PAGE_READWRITE);
 
   return mem;
 }
diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index a06c1978..06281da0 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -26,7 +26,6 @@
 
 #include "../evaluate.h"
 #include "../position.h"
-#include "../misc.h"
 #include "../uci.h"
 #include "../types.h"
 
@@ -245,14 +244,15 @@ namespace Stockfish::Eval::NNUE {
   }
 
 
-  // format_cp_aligned_dot() converts a Value into (centi)pawns and writes it in a buffer,
-  // always keeping two decimals. The buffer must have capacity for at least 7 chars.
-  static void format_cp_aligned_dot(Value v, char* buffer) {
+  // format_cp_aligned_dot() converts a Value into (centi)pawns, always keeping two decimals.
+  static void format_cp_aligned_dot(Value v, std::stringstream &stream) {
+    const double cp = 1.0 * std::abs(int(v)) / UCI::NormalizeToPawnValue;
 
-    buffer[0] = (v < 0 ? '-' : v > 0 ? '+' : ' ');
-
-    double cp = 1.0 * std::abs(int(v)) / UCI::NormalizeToPawnValue;
-    sprintf(&buffer[1], "%6.2f", cp);
+    stream << (v < 0 ? '-' : v > 0 ? '+' : ' ')
+           << std::setiosflags(std::ios::fixed)
+           << std::setw(6)
+           << std::setprecision(2)
+           << cp;
   }
 
 
@@ -332,17 +332,10 @@ namespace Stockfish::Eval::NNUE {
 
     for (std::size_t bucket = 0; bucket < LayerStacks; ++bucket)
     {
-      char buffer[3][8];
-      std::memset(buffer, '\0', sizeof(buffer));
-
-      format_cp_aligned_dot(t.psqt[bucket], buffer[0]);
-      format_cp_aligned_dot(t.positional[bucket], buffer[1]);
-      format_cp_aligned_dot(t.psqt[bucket] + t.positional[bucket], buffer[2]);
-
-      ss <<  "|  " << bucket    << "        "
-         << " |  " << buffer[0] << "  "
-         << " |  " << buffer[1] << "  "
-         << " |  " << buffer[2] << "  "
+      ss <<  "|  " << bucket    << "        ";
+      ss << " |  "; format_cp_aligned_dot(t.psqt[bucket], ss); ss << "  "
+         << " |  "; format_cp_aligned_dot(t.positional[bucket], ss); ss << "  "
+         << " |  "; format_cp_aligned_dot(t.psqt[bucket] + t.positional[bucket], ss); ss << "  "
          << " |";
       if (bucket == t.correctBucket)
           ss << " <-- this bucket is used";
diff --git a/src/thread.cpp b/src/thread.cpp
index 7e71edf1..ca1a7c85 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -73,9 +73,9 @@ void Thread::clear() {
 /// Thread::start_searching() wakes up the thread that will start the search
 
 void Thread::start_searching() {
-
-  std::lock_guard<std::mutex> lk(mutex);
+  mutex.lock();
   searching = true;
+  mutex.unlock(); // Unlock before notifying saves a few CPU-cycles
   cv.notify_one(); // Wake up the thread in idle_loop()
 }
 
@@ -125,20 +125,20 @@ void Thread::idle_loop() {
 
 void ThreadPool::set(size_t requested) {
 
-  if (size() > 0)   // destroy any existing thread(s)
+  if (threads.size() > 0)   // destroy any existing thread(s)
   {
       main()->wait_for_search_finished();
 
-      while (size() > 0)
-          delete back(), pop_back();
+      while (threads.size() > 0)
+          delete threads.back(), threads.pop_back();
   }
 
   if (requested > 0)   // create new thread(s)
   {
-      push_back(new MainThread(0));
+      threads.push_back(new MainThread(0));
 
-      while (size() < requested)
-          push_back(new Thread(size()));
+      while (threads.size() < requested)
+          threads.push_back(new Thread(threads.size()));
       clear();
 
       // Reallocate the hash with the new threadpool size
@@ -154,7 +154,7 @@ void ThreadPool::set(size_t requested) {
 
 void ThreadPool::clear() {
 
-  for (Thread* th : *this)
+  for (Thread* th : threads)
       th->clear();
 
   main()->callsCnt = 0;
@@ -187,7 +187,7 @@ void ThreadPool::start_thinking(Position& pos, StateListPtr& states,
       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.
+  // and call 'go' again without setting a new position states.get() == nullptr.
   assert(states.get() || setupStates.get());
 
   if (states.get())
@@ -198,7 +198,7 @@ void ThreadPool::start_thinking(Position& pos, StateListPtr& states,
   // be deduced from a fen string, so set() clears them and they are set from
   // setupStates->back() later. The rootState is per thread, earlier states are shared
   // since they are read-only.
-  for (Thread* th : *this)
+  for (Thread* th : threads)
   {
       th->nodes = th->tbHits = th->nmpMinPly = th->bestMoveChanges = 0;
       th->rootDepth = th->completedDepth = 0;
@@ -212,12 +212,12 @@ void ThreadPool::start_thinking(Position& pos, StateListPtr& states,
 
 Thread* ThreadPool::get_best_thread() const {
 
-    Thread* bestThread = front();
+    Thread* bestThread = threads.front();
     std::map<Move, int64_t> votes;
     Value minScore = VALUE_NONE;
 
     // Find minimum score of all threads
-    for (Thread* th: *this)
+    for (Thread* th: threads)
         minScore = std::min(minScore, th->rootMoves[0].score);
 
     // Vote according to score and depth, and select the best thread
@@ -225,10 +225,10 @@ Thread* ThreadPool::get_best_thread() const {
             return (th->rootMoves[0].score - minScore + 14) * int(th->completedDepth);
         };
 
-    for (Thread* th : *this)
+    for (Thread* th : threads)
         votes[th->rootMoves[0].pv[0]] += thread_value(th);
 
-    for (Thread* th : *this)
+    for (Thread* th : threads)
         if (abs(bestThread->rootMoves[0].score) >= VALUE_TB_WIN_IN_MAX_PLY)
         {
             // Make sure we pick the shortest mate / TB conversion or stave off mate the longest
@@ -251,8 +251,8 @@ Thread* ThreadPool::get_best_thread() const {
 
 void ThreadPool::start_searching() {
 
-    for (Thread* th : *this)
-        if (th != front())
+    for (Thread* th : threads)
+        if (th != threads.front())
             th->start_searching();
 }
 
@@ -261,8 +261,8 @@ void ThreadPool::start_searching() {
 
 void ThreadPool::wait_for_search_finished() const {
 
-    for (Thread* th : *this)
-        if (th != front())
+    for (Thread* th : threads)
+        if (th != threads.front())
             th->wait_for_search_finished();
 }
 
diff --git a/src/thread.h b/src/thread.h
index 680da209..7566322c 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -101,13 +101,13 @@ struct MainThread : public Thread {
 /// parking and, most importantly, launching a thread. All the access to threads
 /// is done through this class.
 
-struct ThreadPool : public std::vector<Thread*> {
+struct ThreadPool {
 
   void start_thinking(Position&, StateListPtr&, const Search::LimitsType&, bool = false);
   void clear();
   void set(size_t);
 
-  MainThread* main()        const { return static_cast<MainThread*>(front()); }
+  MainThread* main()        const { return static_cast<MainThread*>(threads.front()); }
   uint64_t nodes_searched() const { return accumulate(&Thread::nodes); }
   uint64_t tb_hits()        const { return accumulate(&Thread::tbHits); }
   Thread* get_best_thread() const;
@@ -116,13 +116,21 @@ struct ThreadPool : public std::vector<Thread*> {
 
   std::atomic_bool stop, increaseDepth;
 
+  auto cbegin() const noexcept { return threads.cbegin(); }
+  auto begin() noexcept { return threads.begin(); }
+  auto end() noexcept { return threads.end(); }
+  auto cend() const noexcept { return threads.cend(); }
+  auto size() const noexcept { return threads.size(); }
+  auto empty() const noexcept { return threads.empty(); }
+
 private:
   StateListPtr setupStates;
+  std::vector<Thread*> threads;
 
   uint64_t accumulate(std::atomic<uint64_t> Thread::* member) const {
 
     uint64_t sum = 0;
-    for (Thread* th : *this)
+    for (Thread* th : threads)
         sum += (th->*member).load(std::memory_order_relaxed);
     return sum;
   }
diff --git a/src/thread_win32_osx.h b/src/thread_win32_osx.h
index 01ff1c77..330a8341 100644
--- a/src/thread_win32_osx.h
+++ b/src/thread_win32_osx.h
@@ -41,7 +41,7 @@ void* start_routine(void* ptr)
    P* p = reinterpret_cast<P*>(ptr);
    (p->first->*(p->second))(); // Call member function pointer
    delete p;
-   return NULL;
+   return nullptr;
 }
 
 class NativeThread {
@@ -56,7 +56,7 @@ public:
     pthread_attr_setstacksize(attr, TH_STACK_SIZE);
     pthread_create(&thread, attr, start_routine<T>, new P(obj, fun));
   }
-  void join() { pthread_join(thread, NULL); }
+  void join() { pthread_join(thread, nullptr); }
 };
 
 } // namespace Stockfish
@@ -65,7 +65,7 @@ public:
 
 namespace Stockfish {
 
-typedef std::thread NativeThread;
+using NativeThread = std::thread;
 
 } // namespace Stockfish
 

From e9e7a7b83f78b5c7d29f69083589b449d9b52390 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sat, 7 Jan 2023 14:53:59 +0100
Subject: [PATCH 038/192] Replace some std::string occurrences with
 std::string_view

std::string_view is more lightweight than std::string. Furthermore,
std::string_view variables can be declared constexpr.

closes https://github.com/official-stockfish/Stockfish/pull/4328

No functional change
---
 src/misc.cpp               | 9 +++++----
 src/nnue/evaluate_nnue.cpp | 7 ++++---
 src/position.cpp           | 3 ++-
 src/syzygy/tbprobe.cpp     | 7 ++++---
 4 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index be7abba5..b651972b 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -41,12 +41,13 @@ typedef WORD(*fun5_t)();
 }
 #endif
 
+#include <cstdlib>
 #include <fstream>
 #include <iomanip>
 #include <iostream>
 #include <sstream>
+#include <string_view>
 #include <vector>
-#include <cstdlib>
 
 #if defined(__linux__) && !defined(__ANDROID__)
 #include <stdlib.h>
@@ -68,7 +69,7 @@ namespace Stockfish {
 namespace {
 
 /// Version number or dev.
-const string version = "dev";
+constexpr string_view version = "dev";
 
 /// 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
@@ -151,13 +152,13 @@ string engine_info(bool to_uci) {
   stringstream ss;
   ss << "Stockfish " << version << setfill('0');
 
-  if (version == "dev")
+  if constexpr (version == "dev")
   {
       ss << "-";
       #ifdef GIT_DATE
       ss << GIT_DATE;
       #else
-      const string months("Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec");
+      constexpr string_view months("Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec");
       string month, day, year;
       stringstream date(__DATE__); // From compiler, format is "Sep 21 2008"
 
diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index 06281da0..f132de71 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -18,11 +18,12 @@
 
 // Code for calculating NNUE evaluation function
 
+#include <fstream>
+#include <iomanip>
 #include <iostream>
 #include <set>
 #include <sstream>
-#include <iomanip>
-#include <fstream>
+#include <string_view>
 
 #include "../evaluate.h"
 #include "../position.h"
@@ -210,7 +211,7 @@ namespace Stockfish::Eval::NNUE {
     return t;
   }
 
-  static const std::string PieceToChar(" PNBRQK  pnbrqk");
+  constexpr std::string_view PieceToChar(" PNBRQK  pnbrqk");
 
 
   // format_cp_compact() converts a Value into (centi)pawns and writes it in a buffer.
diff --git a/src/position.cpp b/src/position.cpp
index e82425af..cfd98f68 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -22,6 +22,7 @@
 #include <cstring> // For std::memset, std::memcmp
 #include <iomanip>
 #include <sstream>
+#include <string_view>
 
 #include "bitboard.h"
 #include "misc.h"
@@ -46,7 +47,7 @@ namespace Zobrist {
 
 namespace {
 
-const string PieceToChar(" PNBRQK  pnbrqk");
+constexpr std::string_view PieceToChar(" PNBRQK  pnbrqk");
 
 constexpr Piece Pieces[] = { W_PAWN, W_KNIGHT, W_BISHOP, W_ROOK, W_QUEEN, W_KING,
                              B_PAWN, B_KNIGHT, B_BISHOP, B_ROOK, B_QUEEN, B_KING };
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index cc5e2852..bbfd819d 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -24,9 +24,10 @@
 #include <fstream>
 #include <iostream>
 #include <list>
-#include <sstream>
-#include <type_traits>
 #include <mutex>
+#include <sstream>
+#include <string_view>
+#include <type_traits>
 
 #include "../bitboard.h"
 #include "../movegen.h"
@@ -70,7 +71,7 @@ enum TBFlag { STM = 1, Mapped = 2, WinPlies = 4, LossPlies = 8, Wide = 16, Singl
 inline WDLScore operator-(WDLScore d) { return WDLScore(-int(d)); }
 inline Square operator^(Square s, int i) { return Square(int(s) ^ i); }
 
-const std::string PieceToChar = " PNBRQK  pnbrqk";
+constexpr std::string_view PieceToChar = " PNBRQK  pnbrqk";
 
 int MapPawns[SQUARE_NB];
 int MapB1H1H7[SQUARE_NB];

From 4f4e652ecaf8d42a9bd1092f72c3704435ddba12 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Fri, 6 Jan 2023 18:13:21 +0100
Subject: [PATCH 039/192] Avoid unnecessary string copies

closes https://github.com/official-stockfish/Stockfish/pull/4326

also fixes typo, closes https://github.com/official-stockfish/Stockfish/pull/4332

No functional change
---
 src/evaluate.cpp                   | 2 +-
 src/nnue/layers/sqr_clipped_relu.h | 2 +-
 src/uci.cpp                        | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 3aae9f1c..6db977a4 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -89,7 +89,7 @@ namespace Eval {
     vector<string> dirs = { "<internal>" , "" , CommandLine::binaryDirectory };
     #endif
 
-    for (string directory : dirs)
+    for (const string& directory : dirs)
         if (currentEvalFileName != eval_file)
         {
             if (directory != "<internal>")
diff --git a/src/nnue/layers/sqr_clipped_relu.h b/src/nnue/layers/sqr_clipped_relu.h
index df539b39..3fbb243c 100644
--- a/src/nnue/layers/sqr_clipped_relu.h
+++ b/src/nnue/layers/sqr_clipped_relu.h
@@ -106,7 +106,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
       for (IndexType i = Start; i < InputDimensions; ++i) {
         output[i] = static_cast<OutputType>(
-            // realy should be /127 but we need to make it fast
+            // really should be /127 but we need to make it fast
             // needs to be accounted for in the trainer
             std::max(0ll, std::min(127ll, (((long long)input[i] * input[i]) >> (2 * WeightScaleBits)) / 128)));
       }
diff --git a/src/uci.cpp b/src/uci.cpp
index eb158e72..30c1fa0c 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -160,7 +160,7 @@ namespace {
     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 || s.find("eval") == 0; });
+    num = count_if(list.begin(), list.end(), [](const string& s) { return s.find("go ") == 0 || s.find("eval") == 0; });
 
     TimePoint elapsed = now();
 

From 3d2381d76d7bf9686ef0e0671f60c3b885a7058a Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Wed, 11 Jan 2023 10:07:56 -0500
Subject: [PATCH 040/192] Update default net to nn-1e7ca356472e.nnue

Created by retraining the master net on a dataset composed of:

* The Leela-dfrc_n5000.binpack dataset filtered with depth6 multipv2 search to remove positions with only one good move, in addition to removing positions where either of the two best moves are captures
* The same Leela T80 oct+nov 2022 training data used in recent best datasets
* Additional Leela training data from T60 nov+dec 2021 and T79 apr+may 2022

Trained with end lambda 0.7 and started with max epoch 800. All positions with ply <= 28 were skipped:

```
python easy_train.py \
  --experiment-name leela95-dfrc96-mpv-eval-fonly-T80octnov-T79aprmayT60novdec-12tb7p-sk28-lambda7 \
  --training-dataset /data/leela95-dfrc96-mpv-eval-fonly-T80octnov-T79aprmayT60novdec-12tb7p.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes-skip-ply-lteq-28 \
  --start-from-engine-test-net True \
  --gpus "0," \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --gamma 0.995 \
  --lr 4.375e-4 \
  --tui False \
  --seed $RANDOM \
  --max_epoch 800
```

Around epoch 780, training was manually paused and max epoch increased to 920 before resuming.

During depth6 multipv2 data filtering, positions were considered to have only one good move if the score of the best move was significantly better than the 2nd best move in a way that changes the outcome of the game:

* the best move leads to a significant advantage while the 2nd best move equalizes or loses
* the best move is about equal while the 2nd best move loses

The modified stockfish branch and exact score thresholds used for filtering are at:
https://github.com/linrock/Stockfish/tree/tools-filter-multipv2-eval-diff/src/filter

About 95% of the Leela portion and 96% of the DFRC portion of the Leela-dfrc_n5000.binpack dataset was filtered. Unfiltered parts of the dataset were left out.

The additional Leela training data from T60 nov+dec 2021 and T79 apr+may 2022 was WDL-rescored with about 12TB of syzygy 7-piece tablebases where the material difference is less than around 6 pawns. Best moves were exported to .plain data files during data conversion with the lc0 rescorer.

The exact training data can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move
experiment_leela95-dfrc96-mpv-eval-fonly-T80octnov-T79aprmayT60novdec-12tb7p-sk28-lambda7
run_0/nn-epoch899.nnue : 3.8 +/- 1.6

Passed STC
https://tests.stockfishchess.org/tests/view/63bed1f540aa064159b9c89b
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 103344 W: 27392 L: 26991 D: 48961
Ptnml(0-2): 333, 11223, 28099, 11744, 273

Passed LTC
https://tests.stockfishchess.org/tests/view/63c010415705810de2deb3ec
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 21712 W: 5891 L: 5619 D: 10202
Ptnml(0-2): 12, 2022, 6511, 2304, 7

closes https://github.com/official-stockfish/Stockfish/pull/4338

bench 4106793
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 0a00faa7..643e8540 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-a3dc078bafc7.nnue"
+  #define EvalFileDefaultName   "nn-1e7ca356472e.nnue"
 
   namespace NNUE {
 

From da5bcec481dc22ad2de2e2c8e5c6e8304a373445 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sat, 14 Jan 2023 16:09:31 +0100
Subject: [PATCH 041/192] Fix asm modifiers in add_dpbusd_epi32x2
 implementations

The accumulator should be an earlyclobber because it is written before
all input operands are read. Otherwise, the asm code computes a wrong
result if the accumulator shares a register with one of the other input
operands (which happens if we pass in the same expression for the
accumulator and the operand).

Closes https://github.com/official-stockfish/Stockfish/pull/4339

No functional change
---
 src/nnue/layers/simd.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/nnue/layers/simd.h b/src/nnue/layers/simd.h
index aeab39c4..231f7891 100644
--- a/src/nnue/layers/simd.h
+++ b/src/nnue/layers/simd.h
@@ -153,7 +153,7 @@ namespace Stockfish::Simd {
       asm(
         "vpdpbusd %[b0], %[a0], %[acc]\n\t"
         "vpdpbusd %[b1], %[a1], %[acc]\n\t"
-        : [acc]"+v"(acc)
+        : [acc]"+&v"(acc)
         : [a0]"v"(a0), [b0]"vm"(b0), [a1]"v"(a1), [b1]"vm"(b1)
       );
 #   else
@@ -249,7 +249,7 @@ namespace Stockfish::Simd {
       asm(
         VNNI_PREFIX "vpdpbusd %[b0], %[a0], %[acc]\n\t"
         VNNI_PREFIX "vpdpbusd %[b1], %[a1], %[acc]\n\t"
-        : [acc]"+v"(acc)
+        : [acc]"+&v"(acc)
         : [a0]"v"(a0), [b0]"vm"(b0), [a1]"v"(a1), [b1]"vm"(b1)
       );
 #   else

From a08b8d4e9711c20acedbfe17d618c3c384b339ec Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sun, 15 Jan 2023 00:23:24 +0100
Subject: [PATCH 042/192] Update UCI_Elo parameterization

The old parameterization (https://github.com/official-stockfish/Stockfish/pull/2225/files) has now become quite inaccurate.
This updates the formula based on updated results with master. The formula is based on a fit of the Elo results for games
played between master at various skill levels, and various versions of the Stash engine, which have been ranked at CCRL.

```
   # PLAYER             :  RATING  ERROR  POINTS  PLAYED   (%)
   1 master-skill-19    :  3191.1   40.4   940.0    1707    55
   2 master-skill-18    :  3170.3   39.3  1343.0    2519    53
   3 master-skill-17    :  3141.3   37.8  2282.0    4422    52
   4 master-skill-16    :  3111.2   37.1  2773.0    5423    51
   5 master-skill-15    :  3069.5   37.2  2728.5    5386    51
   6 master-skill-14    :  3024.8   36.1  2702.0    5339    51
   7 master-skill-13    :  2972.9   35.4  2645.5    5263    50
   8 master-skill-12    :  2923.1   35.0  2653.5    5165    51
   9 master-skill-11    :  2855.5   33.6  2524.0    5081    50
  10 master-skill-10    :  2788.3   32.0  2724.5    5511    49
  11 stash-bot-v25.0    :  2744.0   31.5  1952.5    3840    51
  12 master-skill-9     :  2702.8   30.5  2670.0    5018    53
  13 master-skill-8     :  2596.2   28.5  2669.5    4975    54
  14 stash-bot-v21.0    :  2561.2   30.0  1338.0    3366    40
  15 master-skill-7     :  2499.5   28.5  1934.0    4178    46
  16 stash-bot-v20.0    :  2452.6   27.7  1606.5    3378    48
  17 stash-bot-v19.0    :  2425.3   26.7  1787.0    3365    53
  18 master-skill-6     :  2363.2   26.4  2510.5    4379    57
  19 stash-bot-v17.0    :  2280.7   25.4  2209.0    4378    50
  20 master-skill-5     :  2203.7   25.3  2859.5    5422    53
  21 stash-bot-v15.3    :  2200.0   25.4  1757.0    4383    40
  22 stash-bot-v14      :  2145.9   25.5  2890.0    5167    56
  23 stash-bot-v13      :  2042.7   25.8  2263.5    4363    52
  24 stash-bot-v12      :  1963.4   25.8  1769.5    4210    42
  25 master-skill-4     :  1922.9   25.9  2690.0    5399    50
  26 stash-bot-v11      :  1873.0   26.3  2203.5    4335    51
  27 stash-bot-v10      :  1783.8   27.8  2568.5    4301    60
  28 master-skill-3     :  1742.3   27.8  1909.5    4439    43
  29 master-skill-2     :  1608.4   29.4  2064.5    4389    47
  30 stash-bot-v9       :  1582.6   30.2  2130.0    4230    50
  31 master-skill-1     :  1467.6   31.3  2015.5    4244    47
  32 stash-bot-v8       :  1452.8   31.5  1953.5    3780    52
  33 master-skill-0     :  1320.1   32.9   651.5    2083    31
```

Skill 0 .. 19, now covers CCRL Blitz Elo from 1320 to 3190, approximately.
Indeed, the Elo of stash in this analysis is only to within +- 100 Elo of CCRL,
probably because it depends quite a bit on the opponent pool.

To obtain a skill level for a given Elo number, the above data is fit as a 3rd
degree polynomial Skill(Elo). A quick test confirms the correspondence to the above table:

```
Score of master-elo-2721 vs stash-bot-v21.0: 51 - 16 - 19  [0.703] 86
Elo difference: 150.1 +/- 70.2, LOS: 100.0 %, DrawRatio: 22.1 %
```

closes https://github.com/official-stockfish/Stockfish/pull/4341

No functional change.
---
 README.md         | 2 +-
 src/search.cpp    | 5 ++++-
 src/ucioption.cpp | 2 +-
 3 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 4cd5968e..ca90d5d4 100644
--- a/README.md
+++ b/README.md
@@ -102,7 +102,7 @@ change them via a chess GUI. This is a list of available UCI options in Stockfis
 
   * #### UCI_Elo
     If enabled by UCI_LimitStrength, aim for an engine strength of the given Elo.
-    This Elo rating has been calibrated at a time control of 60s+0.6s and anchored to CCRL 40/4.
+    This Elo rating has been calibrated at a time control of 120s+1.0s and anchored to +- 100 Elo to CCRL Blitz.
 
   * #### Skill Level
     Lower the Skill Level in order to make Stockfish play weaker (see also UCI_LimitStrength).
diff --git a/src/search.cpp b/src/search.cpp
index f5eeb23b..346fd6c3 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -96,7 +96,10 @@ namespace {
   struct Skill {
     Skill(int skill_level, int uci_elo) {
         if (uci_elo)
-            level = std::clamp(std::pow((uci_elo - 1346.6) / 143.4, 1 / 0.806), 0.0, 20.0);
+        {
+            double e = double(uci_elo - 1320) / (3190 - 1320);
+            level = std::clamp((((37.2473 * e - 40.8525) * e + 22.2943) * e - 0.311438), 0.0, 19.0);
+        }
         else
             level = double(skill_level);
     }
diff --git a/src/ucioption.cpp b/src/ucioption.cpp
index b4ce70b4..39933ea5 100644
--- a/src/ucioption.cpp
+++ b/src/ucioption.cpp
@@ -73,7 +73,7 @@ void init(OptionsMap& o) {
   o["UCI_Chess960"]          << Option(false);
   o["UCI_AnalyseMode"]       << Option(false);
   o["UCI_LimitStrength"]     << Option(false);
-  o["UCI_Elo"]               << Option(1350, 1350, 2850);
+  o["UCI_Elo"]               << Option(1320, 1320, 3190);
   o["UCI_ShowWDL"]           << Option(false);
   o["SyzygyPath"]            << Option("<empty>", on_tb_path);
   o["SyzygyProbeDepth"]      << Option(1, 1, 100);

From 734315ff3099564c80737ad2be8121008870d28b Mon Sep 17 00:00:00 2001
From: Stephen Touset <stephen@touset.org>
Date: Mon, 16 Jan 2023 14:25:47 -0800
Subject: [PATCH 043/192] Remove precomputed SquareBB

Bit-shifting is a single instruction, and should be faster than an array lookup
on supported architectures. Besides (ever so slightly) speeding up the
conversion of a square into a bitboard, we may see minor general performance
improvements due to preserving more of the CPU's existing cache.

passed STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 47280 W: 12469 L: 12271 D: 22540
Ptnml(0-2): 128, 4893, 13402, 5087, 130
https://tests.stockfishchess.org/tests/view/63c5cfe618c20f4929c5fe46

Small speedup locally:

```
Result of  20 runs
==================
base (./stockfish.master       ) =    1752135  +/- 10943
test (./stockfish.patch        ) =    1763939  +/- 10818
diff                             =     +11804  +/- 4731

speedup        = +0.0067
P(speedup > 0) =  1.0000

CPU: 16 x AMD Ryzen 9 3950X 16-Core Processor
```

Closes https://github.com/official-stockfish/Stockfish/pull/4343

Bench: 4106793
---
 src/bitboard.cpp | 4 ----
 src/bitboard.h   | 3 +--
 2 files changed, 1 insertion(+), 6 deletions(-)

diff --git a/src/bitboard.cpp b/src/bitboard.cpp
index 0ed13fd0..fd5c3c22 100644
--- a/src/bitboard.cpp
+++ b/src/bitboard.cpp
@@ -27,7 +27,6 @@ namespace Stockfish {
 uint8_t PopCnt16[1 << 16];
 uint8_t SquareDistance[SQUARE_NB][SQUARE_NB];
 
-Bitboard SquareBB[SQUARE_NB];
 Bitboard LineBB[SQUARE_NB][SQUARE_NB];
 Bitboard BetweenBB[SQUARE_NB][SQUARE_NB];
 Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB];
@@ -82,9 +81,6 @@ void Bitboards::init() {
   for (unsigned i = 0; i < (1 << 16); ++i)
       PopCnt16[i] = uint8_t(std::bitset<16>(i).count());
 
-  for (Square s = SQ_A1; s <= SQ_H8; ++s)
-      SquareBB[s] = (1ULL << s);
-
   for (Square s1 = SQ_A1; s1 <= SQ_H8; ++s1)
       for (Square s2 = SQ_A1; s2 <= SQ_H8; ++s2)
           SquareDistance[s1][s2] = std::max(distance<File>(s1, s2), distance<Rank>(s1, s2));
diff --git a/src/bitboard.h b/src/bitboard.h
index d4485fcb..42fd0e97 100644
--- a/src/bitboard.h
+++ b/src/bitboard.h
@@ -74,7 +74,6 @@ constexpr Bitboard KingFlank[FILE_NB] = {
 extern uint8_t PopCnt16[1 << 16];
 extern uint8_t SquareDistance[SQUARE_NB][SQUARE_NB];
 
-extern Bitboard SquareBB[SQUARE_NB];
 extern Bitboard BetweenBB[SQUARE_NB][SQUARE_NB];
 extern Bitboard LineBB[SQUARE_NB][SQUARE_NB];
 extern Bitboard PseudoAttacks[PIECE_TYPE_NB][SQUARE_NB];
@@ -108,7 +107,7 @@ extern Magic BishopMagics[SQUARE_NB];
 
 inline Bitboard square_bb(Square s) {
   assert(is_ok(s));
-  return SquareBB[s];
+  return (1ULL << s);
 }
 
 

From a2038c1a011552c3d56d3b318780f7f5eadaf05d Mon Sep 17 00:00:00 2001
From: Jonathan <jonathan.mcdermid1@ucalgary.ca>
Date: Tue, 17 Jan 2023 21:30:50 -0700
Subject: [PATCH 044/192] apply if constexpr to additional instances

as a form of documentation, and a hint to the compiler.

closes https://github.com/official-stockfish/Stockfish/pull/4345

No functional change
---
 AUTHORS          |  1 +
 src/evaluate.cpp | 10 +++++-----
 src/movegen.cpp  | 16 ++++++++--------
 src/movepick.cpp |  2 +-
 4 files changed, 15 insertions(+), 14 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 87fed7b8..42ba5930 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -101,6 +101,7 @@ Jerry Donald Watson (jerrydonaldwatson)
 jjoshua2
 Jonathan Calovski (Mysseno)
 Jonathan Buladas Dumale (SFisGOD)
+Jonathan McDermid (jonathanmcdermid)
 Joost VandeVondele (vondele)
 Jörg Oster (joergoster)
 Joseph Ellis (jhellis3)
diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 6db977a4..8683182c 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -388,10 +388,10 @@ namespace {
   template<Tracing T> template<Color Us, PieceType Pt>
   Score Evaluation<T>::pieces() {
 
-    constexpr Color     Them = ~Us;
-    constexpr Direction Down = -pawn_push(Us);
-    constexpr Bitboard OutpostRanks = (Us == WHITE ? Rank4BB | Rank5BB | Rank6BB
-                                                   : Rank5BB | Rank4BB | Rank3BB);
+    constexpr Color Them = ~Us;
+    [[maybe_unused]] constexpr Direction Down = -pawn_push(Us);
+    [[maybe_unused]] constexpr Bitboard OutpostRanks = (Us == WHITE ? Rank4BB | Rank5BB | Rank6BB
+                                                                    : Rank5BB | Rank4BB | Rank3BB);
     Bitboard b1 = pos.pieces(Us, Pt);
     Bitboard b, bb;
     Score score = SCORE_ZERO;
@@ -430,7 +430,7 @@ namespace {
         int mob = popcount(b & mobilityArea[Us]);
         mobility[Us] += MobilityBonus[Pt - 2][mob];
 
-        if (Pt == BISHOP || Pt == KNIGHT)
+        if constexpr (Pt == BISHOP || Pt == KNIGHT)
         {
             // Bonus if the piece is on an outpost square or can reach one
             // Bonus for knights (UncontestedOutpost) if few relevant targets
diff --git a/src/movegen.cpp b/src/movegen.cpp
index a960f863..47154164 100644
--- a/src/movegen.cpp
+++ b/src/movegen.cpp
@@ -26,12 +26,12 @@ namespace Stockfish {
 namespace {
 
   template<GenType Type, Direction D>
-  ExtMove* make_promotions(ExtMove* moveList, Square to) {
+  ExtMove* make_promotions(ExtMove* moveList, [[maybe_unused]] Square to) {
 
-    if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
+    if constexpr (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
         *moveList++ = make<PROMOTION>(to - D, to, QUEEN);
 
-    if (Type == QUIETS || Type == EVASIONS || Type == NON_EVASIONS)
+    if constexpr (Type == QUIETS || Type == EVASIONS || Type == NON_EVASIONS)
     {
         *moveList++ = make<PROMOTION>(to - D, to, ROOK);
         *moveList++ = make<PROMOTION>(to - D, to, BISHOP);
@@ -60,18 +60,18 @@ namespace {
     Bitboard pawnsNotOn7 = pos.pieces(Us, PAWN) & ~TRank7BB;
 
     // Single and double pawn pushes, no promotions
-    if (Type != CAPTURES)
+    if constexpr (Type != CAPTURES)
     {
         Bitboard b1 = shift<Up>(pawnsNotOn7)   & emptySquares;
         Bitboard b2 = shift<Up>(b1 & TRank3BB) & emptySquares;
 
-        if (Type == EVASIONS) // Consider only blocking squares
+        if constexpr (Type == EVASIONS) // Consider only blocking squares
         {
             b1 &= target;
             b2 &= target;
         }
 
-        if (Type == QUIET_CHECKS)
+        if constexpr (Type == QUIET_CHECKS)
         {
             // To make a quiet check, you either make a direct check by pushing a pawn
             // or push a blocker pawn that is not on the same file as the enemy king.
@@ -102,7 +102,7 @@ namespace {
         Bitboard b2 = shift<UpLeft >(pawnsOn7) & enemies;
         Bitboard b3 = shift<Up     >(pawnsOn7) & emptySquares;
 
-        if (Type == EVASIONS)
+        if constexpr (Type == EVASIONS)
             b3 &= target;
 
         while (b1)
@@ -116,7 +116,7 @@ namespace {
     }
 
     // Standard and en passant captures
-    if (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
+    if constexpr (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
     {
         Bitboard b1 = shift<UpRight>(pawnsNotOn7) & enemies;
         Bitboard b2 = shift<UpLeft >(pawnsNotOn7) & enemies;
diff --git a/src/movepick.cpp b/src/movepick.cpp
index dbe67357..65155a73 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -158,7 +158,7 @@ Move MovePicker::select(Pred filter) {
 
   while (cur < endMoves)
   {
-      if (T == Best)
+      if constexpr (T == Best)
           std::swap(*cur, *std::max_element(cur, endMoves));
 
       if (*cur != ttMove && filter())

From 596a528c6a9ace6fb1a8407c86d972d96653418d Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Fri, 20 Jan 2023 10:41:46 -0500
Subject: [PATCH 045/192] Update default net to nn-bc24c101ada0.nnue

Created by retraining the master net with Leela T78 data from Aug+Sep 2022 added to the previous best dataset. Trained with end lambda 0.7 and started with max epoch 800. All positions with ply <= 28 were skipped:

```
python easy_train.py \
  --experiment-name leela95-dfrc96-filt-only-T80octnov-T60novdecT78augsepT79aprmay-12tb7p-sk28-lambda7 \
  --training-dataset /data/leela95-dfrc96-filt-only-T80octnov-T60novdecT78augsepT79aprmay-12tb7p.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes-skip-ply-lteq-28 \
  --start-from-engine-test-net True \
  --gpus "0," \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --gamma 0.995 \
  --lr 4.375e-4 \
  --tui False \
  --seed $RANDOM \
  --max_epoch 800
```

Around epoch 750, training was manually paused and max epoch increased to 950 before resuming. The additional Leela training data from T78 was prepared in the same way as the previous best dataset.

The exact training data used can be found at:
https://robotmoon.com/nnue-training-data/

While the local elo ratings during this experiment were much lower than in recent master nets, several later epochs had a consistent elo above zero, and this was hypothesized to represent potential strength at slower time controls.

Local elo at 25k nodes per move
leela95-dfrc96-filt-only-T80octnov-T60novdecT78augsepT79aprmay-12tb7p-sk28-lambda7
nn-epoch819.nnue : 0.4 +/- 1.1 (nn-bc24c101ada0.nnue)
nn-epoch799.nnue : 0.3 +/- 1.2
nn-epoch759.nnue : 0.3 +/- 1.1
nn-epoch839.nnue : 0.2 +/- 1.4

Passed STC
https://tests.stockfishchess.org/tests/view/63cabf6f0eefe8694a0c6013
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 41608 W: 11161 L: 10848 D: 19599
Ptnml(0-2): 116, 4496, 11281, 4781, 130

Passed LTC
https://tests.stockfishchess.org/tests/view/63cb1856344bb01c191af263
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 76760 W: 20517 L: 20137 D: 36106
Ptnml(0-2): 34, 7435, 23070, 7799, 42

closes https://github.com/official-stockfish/Stockfish/pull/4351

bench 3941848
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 643e8540..f7ecaac9 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-1e7ca356472e.nnue"
+  #define EvalFileDefaultName   "nn-bc24c101ada0.nnue"
 
   namespace NNUE {
 

From 3dd0a7a7cd9daf306cffc795c343459de6d6f54b Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Fri, 20 Jan 2023 03:02:25 -0600
Subject: [PATCH 046/192] `stat_bonus`: replace quadratic with nearly identical
 line

passed stc: https://tests.stockfishchess.org/tests/view/63ca58c90eefe8694a0c4eac
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 287960 W: 76146 L: 76201 D: 135613
Ptnml(0-2): 947, 31890, 78307, 31943, 893

passed ltc: https://tests.stockfishchess.org/tests/view/63cc8a51344bb01c191b30f0
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 73784 W: 19559 L: 19402 D: 34823
Ptnml(0-2): 33, 7171, 22327, 7328, 33

closes https://github.com/official-stockfish/Stockfish/pull/4352

bench 3990490
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 346fd6c3..b9ba3811 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min((11 * d + 284) * d - 363 , 1650);
+    return std::min(350 * d - 400, 1650);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness

From d3860f8d5efa4d9df726fa605b24689d6a829c7e Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Wed, 25 Jan 2023 08:12:40 +0300
Subject: [PATCH 047/192] Rebalance usage of history heuristics in pruning

This patch has multiple effects:

* history heuristics sum in futility pruning now can't exceed some negative value so futility pruning for moves with negative histories should become slightly less aggressive;

* history heuristics are now used in SEE pruning for quiet moves;

Passed STC:
https://tests.stockfishchess.org/tests/view/63cde339c93e8828d0f02e3a
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 88424 W: 23681 L: 23303 D: 41440
Ptnml(0-2): 258, 9559, 24219, 9899, 277

Passed LTC:
https://tests.stockfishchess.org/tests/view/63ce9009c93e8828d0f04e4f
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 79536 W: 21223 L: 20843 D: 37470
Ptnml(0-2): 22, 7599, 24146, 7979, 22

closes https://github.com/official-stockfish/Stockfish/pull/4355

Bench: 4208265
---
 src/search.cpp | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index b9ba3811..4b2deadd 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1026,12 +1026,17 @@ moves_loop: // When in check, search starts here
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
 
+              lmrDepth += history / 7208;
+              lmrDepth = std::max(lmrDepth, -2);
+
               // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
                   && lmrDepth < 13
-                  && ss->staticEval + 103 + 136 * lmrDepth + history / 53 <= alpha)
+                  && ss->staticEval + 103 + 136 * lmrDepth <= alpha)
                   continue;
 
+              lmrDepth = std::max(lmrDepth, 0);
+
               // Prune moves with negative SEE (~4 Elo)
               if (!pos.see_ge(move, Value(-25 * lmrDepth * lmrDepth - 16 * lmrDepth)))
                   continue;

From def296670dbde6cfad446a735f37f25cfe9df6a2 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Mon, 23 Jan 2023 19:32:26 +0100
Subject: [PATCH 048/192] Fixed UCI TB win values

This patch results in search values for a TB win/loss to be reported in a way that does not change with normalization, i.e. will be consistent over time.

A value of 200.00 pawns is now reported upon entering a TB won position. Values smaller than 200.00 relate to the distance in plies from the root to the probed position position,
with 1 cp being 1 ply distance.

closes https://github.com/official-stockfish/Stockfish/pull/4353

No functional change
---
 src/uci.cpp | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/src/uci.cpp b/src/uci.cpp
index 30c1fa0c..2afd2de7 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -314,8 +314,13 @@ string UCI::value(Value v) {
 
   stringstream ss;
 
-  if (abs(v) < VALUE_MATE_IN_MAX_PLY)
+  if (abs(v) < VALUE_TB_WIN_IN_MAX_PLY)
       ss << "cp " << v * 100 / NormalizeToPawnValue;
+  else if (abs(v) < VALUE_MATE_IN_MAX_PLY)
+  {
+      const int ply = VALUE_MATE_IN_MAX_PLY - 1 - std::abs(v);  // recompute ss->ply
+      ss << "cp " << (v > 0 ? 20000 - ply : -20000 + ply);
+  }
   else
       ss << "mate " << (v > 0 ? VALUE_MATE - v + 1 : -VALUE_MATE - v) / 2;
 

From 2167942b6eab54bafb6aed96d7360c74fec95358 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Wed, 25 Jan 2023 22:25:18 +0100
Subject: [PATCH 049/192] Simplify functions to read/write network parameters

closes https://github.com/official-stockfish/Stockfish/pull/4358

No functional change
---
 src/nnue/nnue_architecture.h | 24 +++++++++++-------------
 1 file changed, 11 insertions(+), 13 deletions(-)

diff --git a/src/nnue/nnue_architecture.h b/src/nnue/nnue_architecture.h
index c43a23c3..508f3aae 100644
--- a/src/nnue/nnue_architecture.h
+++ b/src/nnue/nnue_architecture.h
@@ -72,22 +72,20 @@ struct Network
 
   // Read network parameters
   bool read_parameters(std::istream& stream) {
-    if (!fc_0.read_parameters(stream)) return false;
-    if (!ac_0.read_parameters(stream)) return false;
-    if (!fc_1.read_parameters(stream)) return false;
-    if (!ac_1.read_parameters(stream)) return false;
-    if (!fc_2.read_parameters(stream)) return false;
-    return true;
+    return   fc_0.read_parameters(stream)
+          && ac_0.read_parameters(stream)
+          && fc_1.read_parameters(stream)
+          && ac_1.read_parameters(stream)
+          && fc_2.read_parameters(stream);
   }
 
-  // Read network parameters
+  // Write network parameters
   bool write_parameters(std::ostream& stream) const {
-    if (!fc_0.write_parameters(stream)) return false;
-    if (!ac_0.write_parameters(stream)) return false;
-    if (!fc_1.write_parameters(stream)) return false;
-    if (!ac_1.write_parameters(stream)) return false;
-    if (!fc_2.write_parameters(stream)) return false;
-    return true;
+    return   fc_0.write_parameters(stream)
+          && ac_0.write_parameters(stream)
+          && fc_1.write_parameters(stream)
+          && ac_1.write_parameters(stream)
+          && fc_2.write_parameters(stream);
   }
 
   std::int32_t propagate(const TransformedFeatureType* transformedFeatures)

From 2f67409506e65a47f038055de834462b4a707ccd Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Wed, 25 Jan 2023 22:04:02 +0100
Subject: [PATCH 050/192] Remove redundant const qualifiers

The const qualifiers are already implied by the constexpr qualifiers.

closes https://github.com/official-stockfish/Stockfish/pull/4359

No functional change
---
 src/nnue/layers/affine_transform.h | 36 +++++++++++++++---------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index 710ab8a7..363b4916 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -179,33 +179,33 @@ namespace Stockfish::Eval::NNUE::Layers {
     static_assert(PaddedInputDimensions >= LargeInputSize, "Something went wrong. This specialization should not have been chosen.");
 
 #if defined (USE_AVX512)
-    static constexpr const IndexType InputSimdWidth = 64;
-    static constexpr const IndexType MaxNumOutputRegs = 16;
+    static constexpr IndexType InputSimdWidth = 64;
+    static constexpr IndexType MaxNumOutputRegs = 16;
 #elif defined (USE_AVX2)
-    static constexpr const IndexType InputSimdWidth = 32;
-    static constexpr const IndexType MaxNumOutputRegs = 8;
+    static constexpr IndexType InputSimdWidth = 32;
+    static constexpr IndexType MaxNumOutputRegs = 8;
 #elif defined (USE_SSSE3)
-    static constexpr const IndexType InputSimdWidth = 16;
-    static constexpr const IndexType MaxNumOutputRegs = 8;
+    static constexpr IndexType InputSimdWidth = 16;
+    static constexpr IndexType MaxNumOutputRegs = 8;
 #elif defined (USE_NEON)
-    static constexpr const IndexType InputSimdWidth = 8;
-    static constexpr const IndexType MaxNumOutputRegs = 8;
+    static constexpr IndexType InputSimdWidth = 8;
+    static constexpr IndexType MaxNumOutputRegs = 8;
 #else
     // The fallback implementation will not have permuted weights.
     // We define these to avoid a lot of ifdefs later.
-    static constexpr const IndexType InputSimdWidth = 1;
-    static constexpr const IndexType MaxNumOutputRegs = 1;
+    static constexpr IndexType InputSimdWidth = 1;
+    static constexpr IndexType MaxNumOutputRegs = 1;
 #endif
 
     // A big block is a region in the weight matrix of the size [PaddedInputDimensions, NumOutputRegs].
     // A small block is a region of size [InputSimdWidth, 1]
 
-    static constexpr const IndexType NumOutputRegs = std::min(MaxNumOutputRegs, OutputDimensions);
-    static constexpr const IndexType SmallBlockSize = InputSimdWidth;
-    static constexpr const IndexType BigBlockSize = NumOutputRegs * PaddedInputDimensions;
-    static constexpr const IndexType NumSmallBlocksInBigBlock = BigBlockSize / SmallBlockSize;
-    static constexpr const IndexType NumSmallBlocksPerOutput = PaddedInputDimensions / SmallBlockSize;
-    static constexpr const IndexType NumBigBlocks = OutputDimensions / NumOutputRegs;
+    static constexpr IndexType NumOutputRegs = std::min(MaxNumOutputRegs, OutputDimensions);
+    static constexpr IndexType SmallBlockSize = InputSimdWidth;
+    static constexpr IndexType BigBlockSize = NumOutputRegs * PaddedInputDimensions;
+    static constexpr IndexType NumSmallBlocksInBigBlock = BigBlockSize / SmallBlockSize;
+    static constexpr IndexType NumSmallBlocksPerOutput = PaddedInputDimensions / SmallBlockSize;
+    static constexpr IndexType NumBigBlocks = OutputDimensions / NumOutputRegs;
 
     static_assert(OutputDimensions % NumOutputRegs == 0);
 
@@ -396,8 +396,8 @@ namespace Stockfish::Eval::NNUE::Layers {
     static_assert(PaddedInputDimensions < LargeInputSize, "Something went wrong. This specialization should not have been chosen.");
 
 #if defined (USE_SSSE3)
-    static constexpr const IndexType OutputSimdWidth = SimdWidth / 4;
-    static constexpr const IndexType InputSimdWidth = SimdWidth;
+    static constexpr IndexType OutputSimdWidth = SimdWidth / 4;
+    static constexpr IndexType InputSimdWidth = SimdWidth;
 #endif
 
     // Hash value embedded in the evaluation file

From d4d1cec29631f041adeec98adc5893b5c6a54969 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Sun, 15 Jan 2023 04:08:33 -0600
Subject: [PATCH 051/192] Remove `previousDepth` in favor of `completedDepth +
 2`

Beyond the simplification, this could be considered a bugfix from a certain point of view.
However, the effect is very subtle and essentially impossible for users to notice.
5372f81cc8 added about 2 Elo at LTC, but only for second and later `go` commands; now, with
this patch, the first `go` command will also benefit from that gain. Games under time
controls are unaffected (as per the tests).

STC: https://tests.stockfishchess.org/tests/view/63c3d291330c0d3d051d48a8
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 473792 W: 124858 L: 125104 D: 223830
Ptnml(0-2): 1338, 49653, 135063, 49601, 1241

LTC: https://tests.stockfishchess.org/tests/view/63c8cd56a83c702aac083bc9
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 290728 W: 76926 L: 76978 D: 136824
Ptnml(0-2): 106, 27987, 89221, 27953, 97

closes https://github.com/official-stockfish/Stockfish/pull/4361

bench 4208265
---
 src/search.cpp | 5 +----
 src/thread.cpp | 1 -
 src/thread.h   | 2 +-
 3 files changed, 2 insertions(+), 6 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 4b2deadd..c748f1ff 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -242,9 +242,6 @@ void MainThread::search() {
   bestPreviousScore = bestThread->rootMoves[0].score;
   bestPreviousAverageScore = bestThread->rootMoves[0].averageScore;
 
-  for (Thread* th : Threads)
-    th->previousDepth = bestThread->completedDepth;
-
   // Send again PV info if we have a new best thread
   if (bestThread != this)
       sync_cout << UCI::pv(bestThread->rootPos, bestThread->completedDepth) << sync_endl;
@@ -1053,7 +1050,7 @@ moves_loop: // When in check, search starts here
           // a reduced search on all the other moves but the ttMove and if the
           // result is lower than ttValue minus a margin, then we will extend the ttMove.
           if (   !rootNode
-              &&  depth >= 4 - (thisThread->previousDepth > 24) + 2 * (PvNode && tte->is_pv())
+              &&  depth >= 4 - (thisThread->completedDepth > 22) + 2 * (PvNode && tte->is_pv())
               &&  move == ttMove
               && !excludedMove // Avoid recursive singular search
            /* &&  ttValue != VALUE_NONE Already implicit in the next condition */
diff --git a/src/thread.cpp b/src/thread.cpp
index ca1a7c85..c680393e 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -60,7 +60,6 @@ void Thread::clear() {
   counterMoves.fill(MOVE_NONE);
   mainHistory.fill(0);
   captureHistory.fill(0);
-  previousDepth = 0;
 
   for (bool inCheck : { false, true })
       for (StatsType c : { NoCaptures, Captures })
diff --git a/src/thread.h b/src/thread.h
index 7566322c..46cdb11c 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -69,7 +69,7 @@ public:
   Position rootPos;
   StateInfo rootState;
   Search::RootMoves rootMoves;
-  Depth rootDepth, completedDepth, previousDepth;
+  Depth rootDepth, completedDepth;
   Value rootDelta;
   CounterMoveHistory counterMoves;
   ButterflyHistory mainHistory;

From e4e61cd9cc953e6bc17070da84639d53b7514709 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Thu, 19 Jan 2023 09:49:42 +0900
Subject: [PATCH 052/192] Remove maxNextDepth

This patch allows full PV search to have double extensions as well when
extension == 1 && doDeeperSearch && doEvenDeeperSearch && !doShallowerSearch
is true, which is extremely rare to occur.

Passed non-regression STC (master 3d2381d):
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 370824 W: 97835 L: 97974 D: 175015
Ptnml(0-2): 1073, 38814, 105731, 38767, 1027
https://tests.stockfishchess.org/tests/view/63c89416a83c702aac08314c

Passed non-regression LTC (master 3d2381d):
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 306048 W: 81173 L: 81237 D: 143638
Ptnml(0-2): 117, 27977, 96901, 27911, 118
https://tests.stockfishchess.org/tests/view/63cc4e84344bb01c191b2658

Bench: 4208265
---
 src/search.cpp | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index c748f1ff..41096c9c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -522,7 +522,6 @@ namespace {
 
     constexpr bool PvNode = nodeType != NonPV;
     constexpr bool rootNode = nodeType == Root;
-    const Depth maxNextDepth = rootNode ? depth : depth + 1;
 
     // Check if we have an upcoming move which draws by repetition, or
     // if the opponent had an alternative move earlier to this position.
@@ -1235,8 +1234,7 @@ moves_loop: // When in check, search starts here
           (ss+1)->pv = pv;
           (ss+1)->pv[0] = MOVE_NONE;
 
-          value = -search<PV>(pos, ss+1, -beta, -alpha,
-                              std::min(maxNextDepth, newDepth), false);
+          value = -search<PV>(pos, ss+1, -beta, -alpha, newDepth, false);
       }
 
       // Step 19. Undo move

From 0827e00f10709a4475ece44f0588277fc8cdcd9d Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Wed, 1 Feb 2023 15:44:53 +0800
Subject: [PATCH 053/192] Decrease reduction for killer moves with good history

If move is a main killer and we have a good history, decrease reduction.

STC: https://tests.stockfishchess.org/tests/view/63d38b37721fe2bff693069a
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 46688 W: 12542 L: 12222 D: 21924
Ptnml(0-2): 126, 5013, 12769, 5287, 149

LTC: https://tests.stockfishchess.org/tests/view/63d471e2bde6e5f3cb4be5d3
LLR: 2.93 (-2.94,2.94) <0.50,2.50>
Total: 130976 W: 35033 L: 34555 D: 61388
Ptnml(0-2): 38, 12551, 39833, 13027, 39

closes https://github.com/official-stockfish/Stockfish/pull/4369

Bench: 4069938
---
 src/search.cpp | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index 41096c9c..7d618a65 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1164,6 +1164,11 @@ moves_loop: // When in check, search starts here
       if ((ss+1)->cutoffCnt > 3)
           r++;
 
+      // Decrease reduction if move is a killer and we have a good history
+      if (move == ss->killers[0]
+          && (*contHist[0])[movedPiece][to_sq(move)] >= 3600)
+          r--;
+
       ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]

From 7fc0f589d601cb013f995ff44a49b5d2ae6bb253 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Thu, 2 Feb 2023 00:08:23 +0900
Subject: [PATCH 054/192] Add -Wconditional-uninitialized when using Clang

Add -Wconditional-uninitialized as it is not controlled by -Wall.

closes https://github.com/official-stockfish/Stockfish/pull/4371

No functional change
---
 src/Makefile | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index 30c1be5e..775c72c3 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -426,7 +426,8 @@ ifeq ($(COMP),clang)
 		CXX=x86_64-w64-mingw32-clang++
 	endif
 
-	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-prototypes
+	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-prototypes \
+	            -Wconditional-uninitialized
 
 	ifeq ($(filter $(KERNEL),Darwin OpenBSD FreeBSD),)
 	ifeq ($(target_windows),)

From 3589bd008a3470336b3587e3a292a4cd6b02bf6b Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Thu, 2 Feb 2023 08:41:32 +0100
Subject: [PATCH 055/192] Update WLD model

update the WLD model with about 400M positions extracted from recent LTC games after the net updates.
This ensures that the 50% win rate is again at 1.0 eval.

closes https://github.com/official-stockfish/Stockfish/pull/4373

No functional change.
---
 src/uci.cpp | 4 ++--
 src/uci.h   | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/uci.cpp b/src/uci.cpp
index 2afd2de7..3883b3d3 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -206,8 +206,8 @@ namespace {
      // The coefficients of a third-order polynomial fit is based on the fishtest data
      // for two parameters that need to transform eval to the argument of a logistic
      // function.
-     constexpr double as[] = {  -0.58270499,    2.68512549,   15.24638015,  344.49745382};
-     constexpr double bs[] = {  -2.65734562,   15.96509799,  -20.69040836,   73.61029937 };
+     constexpr double as[] = {   0.33677609,   -4.30175627,   33.08810557,  365.60223431};
+     constexpr double bs[] = {  -2.50471102,   14.23235405,  -14.33066859,   71.42705250 };
 
      // Enforce that NormalizeToPawnValue corresponds to a 50% win rate at ply 64
      static_assert(UCI::NormalizeToPawnValue == int(as[0] + as[1] + as[2] + as[3]));
diff --git a/src/uci.h b/src/uci.h
index bd3df323..5d8ccf1a 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -35,7 +35,7 @@ namespace UCI {
 // the win_rate_model() such that Stockfish outputs an advantage of
 // "100 centipawns" for a position if the engine has a 50% probability to win
 // from this position in selfplay at fishtest LTC time control.
-const int NormalizeToPawnValue = 361;
+const int NormalizeToPawnValue = 394;
 
 class Option;
 

From da8513f0eae439f526ab0a905a3433bdb44bcf66 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Thu, 2 Feb 2023 13:05:54 +0300
Subject: [PATCH 056/192] Do less SEE pruning in qsearch

Current master prunes all moves with negative SEE values in qsearch.
This patch sets constant negative threshold thus allowing some moves with negative SEE values to be searched.
Value of threshold is completely arbitrary and can be tweaked - also it as function of depth can be tried.
Original idea by author of Alexandria engine.

Passed STC
https://tests.stockfishchess.org/tests/view/63d79a59a67dd929a5564976
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 34864 W: 9392 L: 9086 D: 16386
Ptnml(0-2): 113, 3742, 9429, 4022, 126

Passed LTC
https://tests.stockfishchess.org/tests/view/63d8074aa67dd929a5565bc2
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 91616 W: 24532 L: 24126 D: 42958
Ptnml(0-2): 32, 8840, 27662, 9238, 36

closes https://github.com/official-stockfish/Stockfish/pull/4376

Bench: 4010877
---
 src/search.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 7d618a65..b9ca3961 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1556,9 +1556,9 @@ moves_loop: // When in check, search starts here
           }
       }
 
-      // Do not search moves with negative SEE values (~5 Elo)
+      // Do not search moves with bad enough SEE values (~5 Elo)
       if (    bestValue > VALUE_TB_LOSS_IN_MAX_PLY
-          && !pos.see_ge(move))
+          && !pos.see_ge(move, Value(-108)))
           continue;
 
       // Speculative prefetch as early as possible

From 5a30b087c3a3c4435e680e6c61082d5432fa0ace Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Wed, 25 Jan 2023 07:32:02 +0900
Subject: [PATCH 057/192] Expand statistics tools for engine development

This patch adds more debugging slots up to 32 per type and provide tools
to calculate standard deviation and Pearson's correlation coefficient.

However, due to slot being 0 at default, dbg_hit_on(c, b) has to be removed.

Initial idea from snicolet/Stockfish@d8ab604

closes https://github.com/official-stockfish/Stockfish/pull/4354

No functional change
---
 src/misc.cpp | 94 ++++++++++++++++++++++++++++++++++++++++++++++------
 src/misc.h   |  7 ++--
 2 files changed, 88 insertions(+), 13 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index b651972b..7d848d32 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -41,6 +41,7 @@ typedef WORD(*fun5_t)();
 }
 #endif
 
+#include <cmath>
 #include <cstdlib>
 #include <fstream>
 #include <iomanip>
@@ -299,21 +300,94 @@ std::string compiler_info() {
 
 
 /// Debug functions used mainly to collect run-time statistics
-static std::atomic<int64_t> hits[2], means[2];
+constexpr int MaxDebugSlots = 32;
 
-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_mean_of(int v) { ++means[0]; means[1] += v; }
+namespace {
+
+template<size_t N>
+struct DebugInfo {
+    std::atomic<int64_t> data[N] = { 0 };
+
+    constexpr inline std::atomic<int64_t>& operator[](int index) { return data[index]; }
+};
+
+DebugInfo<2> hit[MaxDebugSlots];
+DebugInfo<2> mean[MaxDebugSlots];
+DebugInfo<3> stdev[MaxDebugSlots];
+DebugInfo<6> correl[MaxDebugSlots];
+
+}  // namespace
+
+void dbg_hit_on(bool cond, int slot) {
+
+    ++hit[slot][0];
+    if (cond)
+        ++hit[slot][1];
+}
+
+void dbg_mean_of(int64_t value, int slot) {
+
+    ++mean[slot][0];
+    mean[slot][1] += value;
+}
+
+void dbg_stdev_of(int64_t value, int slot) {
+
+    ++stdev[slot][0];
+    stdev[slot][1] += value;
+    stdev[slot][2] += value * value;
+}
+
+void dbg_correl_of(int64_t value1, int64_t value2, int slot) {
+
+    ++correl[slot][0];
+    correl[slot][1] += value1;
+    correl[slot][2] += value1 * value1;
+    correl[slot][3] += value2;
+    correl[slot][4] += value2 * value2;
+    correl[slot][5] += value1 * value2;
+}
 
 void dbg_print() {
 
-  if (hits[0])
-      cerr << "Total " << hits[0] << " Hits " << hits[1]
-           << " hit rate (%) " << 100 * hits[1] / hits[0] << endl;
+    int64_t n;
+    auto E   = [&n](int64_t x) { return double(x) / n; };
+    auto sqr = [](double x) { return x * x; };
 
-  if (means[0])
-      cerr << "Total " << means[0] << " Mean "
-           << (double)means[1] / means[0] << endl;
+    for (int i = 0; i < MaxDebugSlots; ++i)
+        if ((n = hit[i][0]))
+            std::cerr << "Hit #" << i
+                      << ": Total " << n << " Hits " << hit[i][1]
+                      << " Hit Rate (%) " << 100.0 * E(hit[i][1])
+                      << std::endl;
+
+    for (int i = 0; i < MaxDebugSlots; ++i)
+        if ((n = mean[i][0]))
+        {
+            std::cerr << "Mean #" << i
+                      << ": Total " << n << " Mean " << E(mean[i][1])
+                      << std::endl;
+        }
+
+    for (int i = 0; i < MaxDebugSlots; ++i)
+        if ((n = stdev[i][0]))
+        {
+            double r = sqrtl(E(stdev[i][2]) - sqr(E(stdev[i][1])));
+            std::cerr << "Stdev #" << i
+                      << ": Total " << n << " Stdev " << r
+                      << std::endl;
+        }
+
+    for (int i = 0; i < MaxDebugSlots; ++i)
+        if ((n = correl[i][0]))
+        {
+            double r = (E(correl[i][5]) - E(correl[i][1]) * E(correl[i][3]))
+                       / (  sqrtl(E(correl[i][2]) - sqr(E(correl[i][1])))
+                          * sqrtl(E(correl[i][4]) - sqr(E(correl[i][3]))));
+            std::cerr << "Correl. #" << i
+                      << ": Total " << n << " Coefficient " << r
+                      << std::endl;
+        }
 }
 
 
diff --git a/src/misc.h b/src/misc.h
index 4c1150f8..9761da8a 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -39,9 +39,10 @@ void std_aligned_free(void* ptr);
 void* aligned_large_pages_alloc(size_t size); // memory aligned by page size, min alignment: 4096 bytes
 void aligned_large_pages_free(void* mem); // nop if mem == nullptr
 
-void dbg_hit_on(bool b);
-void dbg_hit_on(bool c, bool b);
-void dbg_mean_of(int v);
+void dbg_hit_on(bool cond, int slot = 0);
+void dbg_mean_of(int64_t value, int slot = 0);
+void dbg_stdev_of(int64_t value, int slot = 0);
+void dbg_correl_of(int64_t value1, int64_t value2, int slot = 0);
 void dbg_print();
 
 typedef std::chrono::milliseconds::rep TimePoint; // A value in milliseconds

From 1cdc0f78bd937637128ad10f5168cdb80390f6fb Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 3 Feb 2023 10:52:14 +0300
Subject: [PATCH 058/192] Simplify usage of optimism in complexity

This patch removes one condition in optimism usage in complexity, now negative optimism also impacts it.

Passed STC:
https://tests.stockfishchess.org/tests/view/63d34f43721fe2bff692fb12
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 204920 W: 54343 L: 54309 D: 96268
Ptnml(0-2): 598, 22648, 55897, 22756, 561

Passed LTC:
https://tests.stockfishchess.org/tests/view/63d612a2a67dd929a556075c
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 200712 W: 53207 L: 53172 D: 94333
Ptnml(0-2): 58, 19664, 60901, 19651, 82

closes https://github.com/official-stockfish/Stockfish/pull/4377

bench 4204964
---
 src/evaluate.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 8683182c..6d5a8a0c 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1073,7 +1073,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
       // Blend nnue complexity with (semi)classical complexity
       nnueComplexity = (  406 * nnueComplexity
                         + 424 * abs(psq - nnue)
-                        + (optimism  > 0 ? int(optimism) * int(psq - nnue) : 0)
+                        + int(optimism) * int(psq - nnue)
                         ) / 1024;
 
       // Return hybrid NNUE complexity to caller

From d2f79ff0e0efc33797120f59355c2a5571b4ab80 Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Fri, 3 Feb 2023 18:54:30 +0800
Subject: [PATCH 059/192] Remove reduced LMR capture bonus

In LMR, simplify away the reduced capture bonus (i.e. if (capture) bonus /= 6).

Non-regression STC: https://tests.stockfishchess.org/tests/view/63da1da9bbadd17b3787dced
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 28152 W: 7521 L: 7296 D: 13335
Ptnml(0-2): 76, 3069, 7568, 3280, 83

Non-regression LTC: https://tests.stockfishchess.org/tests/view/63da6ad4bbadd17b3787e98c
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 52472 W: 14120 L: 13941 D: 24411
Ptnml(0-2): 16, 5071, 15887, 5242, 20

closes https://github.com/official-stockfish/Stockfish/pull/4378

Bench: 4034016
---
 src/search.cpp | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index b9ca3961..f6bc0aa9 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1214,9 +1214,6 @@ moves_loop: // When in check, search starts here
               int bonus = value > alpha ?  stat_bonus(newDepth)
                                         : -stat_bonus(newDepth);
 
-              if (capture)
-                  bonus /= 6;
-
               update_continuation_histories(ss, movedPiece, to_sq(move), bonus);
           }
       }

From 8d3457a9966f8c744ab7f8536be408196ccd8af9 Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@PHXL0356.wp.lan>
Date: Fri, 3 Feb 2023 17:57:19 +0100
Subject: [PATCH 060/192] Improve excluded move logic

PR consists of 2 improvements on nodes with excludeMove:

1. Remove xoring the posKey with make_key(excludedMove)

   Since we never call tte->save anymore with excludedMove,
   the unique left purpose of the xoring was to avoid a TT hit.
   Nevertheless on a normal bench run this produced ~25 false positives
   (key collisions)
   To avoid that we now forbid early TT cutoff's with excludeMove
   Maybe these accesses to TT with xored key caused useless misses
   in the CPU caches (L1, L2 ...)
   Now doing the probe with the same key as the enclosing search does,
   should hit the CPU cache.

2. Don't probe Tablebases with excludedMove.

   This can't be tested on fishtest, but it's obvious that
   tablebases don't deliver any information about suboptimal moves.

Side note:
   Very surprisingly it looks like we cannot use static eval's from
   TT since they slightly differ over time due to changing optimism.
   Attempts to use static eval's from TT did loose about 13 ELO.
   This is something about to investigate.

LTC: https://tests.stockfishchess.org/tests/view/63dc0f8de9d4cdfbe672d0c6
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 44736 W: 12046 L: 11733 D: 20957
Ptnml(0-2): 12, 4212, 13617, 4505, 22

An analogue of this passed STC & LTC
see PR #4374 (thanks Dubslow for reviewing!)

closes https://github.com/official-stockfish/Stockfish/pull/4380

Bench: 4758694
---
 src/search.cpp | 26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index f6bc0aa9..30a08cb7 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -615,22 +615,24 @@ namespace {
     if (!rootNode)
         (ss+2)->statScore = 0;
 
-    // Step 4. Transposition table lookup. We don't want the score of a partial
-    // search to overwrite a previous full search TT value, so we use a different
-    // position key in case of an excluded move.
+    // Step 4. Transposition table lookup.
     excludedMove = ss->excludedMove;
-    posKey = excludedMove == MOVE_NONE ? pos.key() : pos.key() ^ make_key(excludedMove);
+    posKey = pos.key();
     tte = TT.probe(posKey, ss->ttHit);
     ttValue = ss->ttHit ? value_from_tt(tte->value(), ss->ply, pos.rule50_count()) : VALUE_NONE;
     ttMove =  rootNode ? thisThread->rootMoves[thisThread->pvIdx].pv[0]
             : ss->ttHit    ? tte->move() : MOVE_NONE;
     ttCapture = ttMove && pos.capture(ttMove);
+
+    // At this point, if excluded, skip straight to step 6, static eval. However,
+    // to save indentation, we list the condition in all code between here and there.
     if (!excludedMove)
         ss->ttPv = PvNode || (ss->ttHit && tte->is_pv());
 
     // At non-PV nodes we check for an early TT cutoff
     if (  !PvNode
         && ss->ttHit
+        && !excludedMove
         && tte->depth() > depth - (tte->bound() == BOUND_EXACT)
         && ttValue != VALUE_NONE // Possible in case of TT access race
         && (tte->bound() & (ttValue >= beta ? BOUND_LOWER : BOUND_UPPER)))
@@ -664,7 +666,7 @@ namespace {
     }
 
     // Step 5. Tablebases probe
-    if (!rootNode && TB::Cardinality)
+    if (!rootNode && !excludedMove && TB::Cardinality)
     {
         int piecesCount = pos.count<ALL_PIECES>();
 
@@ -727,6 +729,12 @@ namespace {
         complexity = 0;
         goto moves_loop;
     }
+    else if (excludedMove) {
+        // excludeMove implies that we had a ttHit on the containing non-excluded search with ss->staticEval filled from TT
+        // However static evals from the TT aren't good enough (-13 elo), presumably due to changing optimism context
+        // Recalculate value with current optimism (without updating thread avgComplexity)
+        ss->staticEval = eval = evaluate(pos, &complexity);
+    }
     else if (ss->ttHit)
     {
         // Never assume anything about values stored in TT
@@ -735,6 +743,7 @@ namespace {
             ss->staticEval = eval = evaluate(pos, &complexity);
         else // Fall back to (semi)classical complexity for TT hits, the NNUE complexity is lost
             complexity = abs(ss->staticEval - pos.psq_eg_stm());
+        thisThread->complexityAverage.update(complexity);
 
         // ttValue can be used as a better position evaluation (~7 Elo)
         if (    ttValue != VALUE_NONE
@@ -744,14 +753,12 @@ namespace {
     else
     {
         ss->staticEval = eval = evaluate(pos, &complexity);
+        thisThread->complexityAverage.update(complexity);
 
         // Save static evaluation into transposition table
-        if (!excludedMove)
-            tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
+        tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
     }
 
-    thisThread->complexityAverage.update(complexity);
-
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
@@ -1061,6 +1068,7 @@ moves_loop: // When in check, search starts here
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;
+              // the search with excludedMove will update ss->staticEval
               value = search<NonPV>(pos, ss, singularBeta - 1, singularBeta, singularDepth, cutNode);
               ss->excludedMove = MOVE_NONE;
 

From d5817a5896a8d93c4560b405da576a02aaa0c08a Mon Sep 17 00:00:00 2001
From: Disservin <disservin.social@gmail.com>
Date: Sat, 4 Feb 2023 14:16:58 +0100
Subject: [PATCH 061/192] remove unnecessary variable

pinned already has to be true for the bitwise &

closes https://github.com/official-stockfish/Stockfish/pull/4381

No functional change
---
 src/movegen.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/movegen.cpp b/src/movegen.cpp
index 47154164..255dce04 100644
--- a/src/movegen.cpp
+++ b/src/movegen.cpp
@@ -264,7 +264,7 @@ ExtMove* generate<LEGAL>(const Position& pos, ExtMove* moveList) {
   moveList = pos.checkers() ? generate<EVASIONS    >(pos, moveList)
                             : generate<NON_EVASIONS>(pos, moveList);
   while (cur != moveList)
-      if (  ((pinned && pinned & from_sq(*cur)) || from_sq(*cur) == ksq || type_of(*cur) == EN_PASSANT)
+      if (  ((pinned & from_sq(*cur)) || from_sq(*cur) == ksq || type_of(*cur) == EN_PASSANT)
           && !pos.legal(*cur))
           *cur = (--moveList)->move;
       else

From 8f843633db3faaf447cc191cbaed9f5ddfd374bd Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Sat, 4 Feb 2023 23:46:44 +0300
Subject: [PATCH 062/192] Cleanup and reorder in qsearch

This patch is a simplification / code normalisation in qsearch.

Adds steps in comments the same way we have in search;

Makes a separate "pruning" stage instead of heuristics randomly being spread over qsearch code;
Reorders pruning heuristics from least taxing ones to more taxing ones;
Removes repeated check for best value not being mated, instead uses 1 check - thus removes some lines of code.
Moves prefetch and move setup after pruning - makes no sense to do them if move will actually get pruned.

Passed non-regression test:
https://tests.stockfishchess.org/tests/view/63dd2c5ff9a50a69252c1413
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 113504 W: 29898 L: 29770 D: 53836
Ptnml(0-2): 287, 11861, 32327, 11991, 286

https://github.com/official-stockfish/Stockfish/pull/4382

Non-functional change.
---
 src/search.cpp | 54 ++++++++++++++++++++++++++++----------------------
 1 file changed, 30 insertions(+), 24 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 30a08cb7..aa87948b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1427,6 +1427,7 @@ moves_loop: // When in check, search starts here
     bool pvHit, givesCheck, capture;
     int moveCount;
 
+    // Step 1. Initialize node
     if (PvNode)
     {
         (ss+1)->pv = pv;
@@ -1438,7 +1439,7 @@ moves_loop: // When in check, search starts here
     ss->inCheck = pos.checkers();
     moveCount = 0;
 
-    // Check for an immediate draw or maximum ply reached
+    // Step 2. Check for an immediate draw or maximum ply reached
     if (   pos.is_draw(ss->ply)
         || ss->ply >= MAX_PLY)
         return (ss->ply >= MAX_PLY && !ss->inCheck) ? evaluate(pos) : VALUE_DRAW;
@@ -1450,13 +1451,14 @@ moves_loop: // When in check, search starts here
     // only two types of depth in TT: DEPTH_QS_CHECKS or DEPTH_QS_NO_CHECKS.
     ttDepth = ss->inCheck || depth >= DEPTH_QS_CHECKS ? DEPTH_QS_CHECKS
                                                   : DEPTH_QS_NO_CHECKS;
-    // Transposition table lookup
+    // Step 3. Transposition table lookup
     posKey = pos.key();
     tte = TT.probe(posKey, ss->ttHit);
     ttValue = ss->ttHit ? value_from_tt(tte->value(), ss->ply, pos.rule50_count()) : VALUE_NONE;
     ttMove = ss->ttHit ? tte->move() : MOVE_NONE;
     pvHit = ss->ttHit && tte->is_pv();
 
+    // At non-PV nodes we check for an early TT cutoff
     if (  !PvNode
         && ss->ttHit
         && tte->depth() >= ttDepth
@@ -1464,7 +1466,7 @@ moves_loop: // When in check, search starts here
         && (tte->bound() & (ttValue >= beta ? BOUND_LOWER : BOUND_UPPER)))
         return ttValue;
 
-    // Evaluate the position statically
+    // Step 4. Static evaluation of the position
     if (ss->inCheck)
     {
         ss->staticEval = VALUE_NONE;
@@ -1522,7 +1524,8 @@ moves_loop: // When in check, search starts here
 
     int quietCheckEvasions = 0;
 
-    // Loop through the moves until no moves remain or a beta cutoff occurs
+    // Step 5. Loop through all pseudo-legal moves until no moves remain
+    // or a beta cutoff occurs.
     while ((move = mp.next_move()) != MOVE_NONE)
     {
       assert(is_ok(move));
@@ -1536,9 +1539,11 @@ moves_loop: // When in check, search starts here
 
       moveCount++;
 
+    // Step 6. Pruning.
+    if (bestValue > VALUE_TB_LOSS_IN_MAX_PLY)
+    {
       // Futility pruning and moveCount pruning (~10 Elo)
-      if (    bestValue > VALUE_TB_LOSS_IN_MAX_PLY
-          && !givesCheck
+      if (   !givesCheck
           &&  to_sq(move) != prevSq
           &&  futilityBase > -VALUE_KNOWN_WIN
           &&  type_of(move) != PROMOTION)
@@ -1561,43 +1566,43 @@ moves_loop: // When in check, search starts here
           }
       }
 
-      // Do not search moves with bad enough SEE values (~5 Elo)
-      if (    bestValue > VALUE_TB_LOSS_IN_MAX_PLY
-          && !pos.see_ge(move, Value(-108)))
+      // We prune after 2nd quiet check evasion where being 'in check' is implicitly checked through the counter
+      // and being a 'quiet' apart from being a tt move is assumed after an increment because captures are pushed ahead.
+      if (quietCheckEvasions > 1)
+          break;
+
+      // Continuation history based pruning (~3 Elo)
+      if (   !capture
+          && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
+          && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
           continue;
 
+      // Do not search moves with bad enough SEE values (~5 Elo)
+      if (!pos.see_ge(move, Value(-108)))
+          continue;
+
+    }
+
       // Speculative prefetch as early as possible
       prefetch(TT.first_entry(pos.key_after(move)));
 
+      // Update the current move
       ss->currentMove = move;
       ss->continuationHistory = &thisThread->continuationHistory[ss->inCheck]
                                                                 [capture]
                                                                 [pos.moved_piece(move)]
                                                                 [to_sq(move)];
 
-      // Continuation history based pruning (~3 Elo)
-      if (   !capture
-          && bestValue > VALUE_TB_LOSS_IN_MAX_PLY
-          && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
-          && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
-          continue;
-
-      // We prune after 2nd quiet check evasion where being 'in check' is implicitly checked through the counter
-      // and being a 'quiet' apart from being a tt move is assumed after an increment because captures are pushed ahead.
-      if (   bestValue > VALUE_TB_LOSS_IN_MAX_PLY
-          && quietCheckEvasions > 1)
-          break;
-
       quietCheckEvasions += !capture && ss->inCheck;
 
-      // Make and search the move
+      // Step 7. Make and search the move
       pos.do_move(move, st, givesCheck);
       value = -qsearch<nodeType>(pos, ss+1, -beta, -alpha, depth - 1);
       pos.undo_move(move);
 
       assert(value > -VALUE_INFINITE && value < VALUE_INFINITE);
 
-      // Check for a new best move
+      // Step 8. Check for a new best move
       if (value > bestValue)
       {
           bestValue = value;
@@ -1617,6 +1622,7 @@ moves_loop: // When in check, search starts here
        }
     }
 
+    // Step 9. Check for mate
     // All legal moves have been searched. A special case: if we're in check
     // and no legal moves were found, it is checkmate.
     if (ss->inCheck && bestValue == -VALUE_INFINITE)

From e25bcaed3cb69406cec4cd3d212cdf5232234949 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Fri, 3 Feb 2023 18:18:55 -0600
Subject: [PATCH 063/192] Update `complexityAverage` in all branches of static
 eval

STC: https://tests.stockfishchess.org/tests/view/63dda49573223e7f52ad0f8c
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 117416 W: 31173 L: 31049 D: 55194
Ptnml(0-2): 290, 12246, 33533, 12328, 311

LTC: https://tests.stockfishchess.org/tests/view/63dfa90873223e7f52ad69b8
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 44416 W: 11924 L: 11744 D: 20748
Ptnml(0-2): 5, 4036, 13968, 4172, 27

closes https://github.com/official-stockfish/Stockfish/pull/4385

bench 4758694
---
 src/search.cpp | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index aa87948b..2eed74b8 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -743,7 +743,6 @@ namespace {
             ss->staticEval = eval = evaluate(pos, &complexity);
         else // Fall back to (semi)classical complexity for TT hits, the NNUE complexity is lost
             complexity = abs(ss->staticEval - pos.psq_eg_stm());
-        thisThread->complexityAverage.update(complexity);
 
         // ttValue can be used as a better position evaluation (~7 Elo)
         if (    ttValue != VALUE_NONE
@@ -753,11 +752,10 @@ namespace {
     else
     {
         ss->staticEval = eval = evaluate(pos, &complexity);
-        thisThread->complexityAverage.update(complexity);
-
         // Save static evaluation into transposition table
         tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
     }
+    thisThread->complexityAverage.update(complexity);
 
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)

From 05dea2ca4657dec10637bb53c4ad583f680e0677 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Sun, 5 Feb 2023 13:00:30 -0500
Subject: [PATCH 064/192] Update default net to nn-1337b1adec5b.nnue

Created by retraining the master net on a dataset composed of:

* Most of the previous best dataset filtered to remove positions likely having only one good move
* Adding training data from Leela T77 dec2021 rescored with 16tb of 7-piece tablebases

Trained with end lambda 0.7 and max epoch 900. Positions with ply <= 28 were removed from most of the previous best dataset before training began. A new nnue-pytorch trainer param for skipping early plies was used to skip plies <= 24 in the unfiltered and additional Leela T77 parts of the dataset.

```
python easy_train.py \
  --experiment-name leela96-dfrc99-T80octnovT79aprmayT60novdec-eval-filt-v2-T78augsep-12tb-T77dec-16tb-lambda7-sk24 \
  --training-dataset /data/leela96-dfrc99-T80octnovT79aprmayT60novdec-eval-filt-v2-T78augsep-12tb-T77dec-16tb.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/easy-train-early-fen-skipping \
  --early-fen-skipping 24 \
  --gpus "0," \
  --start-from-engine-test-net True \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --gamma 0.995 \
  --lr 4.375e-4 \
  --tui False \
  --seed $RANDOM \
  --max_epoch 900
```

The depth6 multipv2 search filtering method is the same as the one used for filtering recent best datasets, with a lower eval difference threshold to remove slightly more positions than before. These parts of the dataset were filtered:

* 96% of T60T70wIsRightFarseerT60T74T75T76.binpack
* 99% of dfrc_n5000.binpack
* T80 oct + nov 2022 data, no positions with castling flags, rescored with ~600gb 7p tablebases
* T79 apr + may 2022 data, rescored with 12tb 7p tablebases
* T60 nov + dec 2021 data, rescored with 12tb 7p tablebases

These parts of the dataset were not filtered. Positions with ply <= 24 were skipped during training:

* T78 aug + sep 2022 data, rescored with 12tb 7p tablebases
* 84% of T77 dec 2021 data, rescored with 16tb 7p tablebases

The code and exact evaluation thresholds used for data filtering can be found at:
https://github.com/linrock/Stockfish/tree/tools-filter-multipv2-eval-diff-t2/src/filter

The exact training data used can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch859.nnue : 3.5 +/ 1.2

Passed STC:
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
https://tests.stockfishchess.org/tests/view/63dfeefc73223e7f52ad769f
Total: 219744 W: 58572 L: 58002 D: 103170
Ptnml(0-2): 609, 24446, 59284, 24832, 701

Passed LTC:
https://tests.stockfishchess.org/tests/view/63e268fc73223e7f52ade7b6
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 91256 W: 24528 L: 24121 D: 42607
Ptnml(0-2): 48, 8863, 27390, 9288, 39

closes https://github.com/official-stockfish/Stockfish/pull/4387

bench 3841998
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index f7ecaac9..cdea2ab2 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-bc24c101ada0.nnue"
+  #define EvalFileDefaultName   "nn-1337b1adec5b.nnue"
 
   namespace NNUE {
 

From e5f6d71b96b5149e5e1df30721e1870abdb218ce Mon Sep 17 00:00:00 2001
From: borg323 <borg323@users.noreply.github.com>
Date: Thu, 9 Feb 2023 21:14:59 +0200
Subject: [PATCH 065/192] Fix build on arm windows

avoids the use of _mm_malloc on arm windows.

fixes #4379
closes https://github.com/official-stockfish/Stockfish/pull/4388

No functional change
---
 AUTHORS      | 1 +
 src/misc.cpp | 8 ++++++--
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 42ba5930..634de4a3 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -35,6 +35,7 @@ Ben Chaney (Chaneybenjamini)
 Ben Koshy (BKSpurgeon)
 Bill Henry (VoyagerOne)
 Bojun Guo (noobpwnftw, Nooby)
+borg323
 Boštjan Mejak (PedanticHacker)
 braich
 Brian Sheppard (SapphireBrand, briansheppard-toast)
diff --git a/src/misc.cpp b/src/misc.cpp
index 7d848d32..e65faab9 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -448,8 +448,10 @@ void* std_aligned_alloc(size_t alignment, size_t size) {
 #if defined(POSIXALIGNEDALLOC)
   void *mem;
   return posix_memalign(&mem, alignment, size) ? nullptr : mem;
-#elif defined(_WIN32)
+#elif defined(_WIN32) && !defined(_M_ARM) && !defined(_M_ARM64)
   return _mm_malloc(size, alignment);
+#elif defined(_WIN32)
+  return _aligned_malloc(size, alignment);
 #else
   return std::aligned_alloc(alignment, size);
 #endif
@@ -459,8 +461,10 @@ void std_aligned_free(void* ptr) {
 
 #if defined(POSIXALIGNEDALLOC)
   free(ptr);
-#elif defined(_WIN32)
+#elif defined(_WIN32) && !defined(_M_ARM) && !defined(_M_ARM64)
   _mm_free(ptr);
+#elif defined(_WIN32)
+  _aligned_free(ptr);
 #else
   free(ptr);
 #endif

From 852330ee5060e42a42c1cddd85b82e28d09f4229 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Sat, 11 Feb 2023 17:14:16 +0100
Subject: [PATCH 066/192] update cuckoo link

use webarchive to link to the cycle detection paper by Kervinck.

closes https://github.com/official-stockfish/Stockfish/pull/4389

No functional change
---
 src/position.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/position.cpp b/src/position.cpp
index cfd98f68..37aa2e9e 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -97,7 +97,7 @@ std::ostream& operator<<(std::ostream& os, const Position& pos) {
 
 // Marcel van Kervinck's cuckoo algorithm for fast detection of "upcoming repetition"
 // situations. Description of the algorithm in the following paper:
-// https://marcelk.net/2013-04-06/paper/upcoming-rep-v2.pdf
+// http://web.archive.org/web/20201107002606/https://marcelk.net/2013-04-06/paper/upcoming-rep-v2.pdf
 
 // First and second hash functions for indexing the cuckoo tables
 inline int H1(Key h) { return h & 0x1fff; }

From 2c36d1e7e7374b8babb3cc503c0bc07ceb83dbf8 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Mon, 13 Feb 2023 11:54:59 +0900
Subject: [PATCH 067/192] Fix overflow in add_dpbusd_epi32x2

This patch fixes 16bit overflow in *_add_dpbusd_epi32x2 functions,
that can be triggered in rare cases depending on the NNUE weights.

While the code leads to some slowdown on affected architectures
(most notably avx2), the fix is simpler than some of the other
options discussed in
https://github.com/official-stockfish/Stockfish/pull/4394

Code suggested by Sopel97.

Result of "bench 4096 1 30 default depth nnue":

| Architecture        | master    | patch (gcc) | patch (clang) |
|---------------------|-----------|-------------|---------------|
| x86-64-vnni512      | 762122798 | 762122798   | 762122798     |
| x86-64-avx512       | 769723503 | 762122798   | 762122798     |
| x86-64-bmi2         | 769723503 | 762122798   | 762122798     |
| x86-64-ssse3        | 769723503 | 762122798   | 762122798     |
| x86-64              | 762122798 | 762122798   | 762122798     |

Following architectures will experience ~4% slowdown due to an
additional instruction in the middle of hot path:

* x86-64-avx512
* x86-64-bmi2
* x86-64-avx2
* x86-64-sse41-popcnt (x86-64-modern)
* x86-64-ssse3
* x86-32-sse41-popcnt

This patch clearly loses Elo against master with both STC and LTC.

Failed non-regression STC (256bit fix only):
LLR: -2.95 (-2.94,2.94) <-1.75,0.25>
Total: 33528 W: 8769 L: 9049 D: 15710
Ptnml(0-2): 96, 3616, 9600, 3376, 76
https://tests.stockfishchess.org/tests/view/63e6a5b44299542b1e26a485

60+0.6 @ 30000 games:
Elo: -1.67 +-1.7 (95%) LOS: 2.8%
Total: 30000 W: 7848 L: 7992 D: 14160
Ptnml(0-2): 12, 2847, 9436, 2683, 22
nElo: -3.84 +-3.9 (95%) PairsRatio: 0.95
https://tests.stockfishchess.org/tests/view/63e7ac716d0e1db55f35a660

However, a test against nn-a3dc078bafc7.nnue, which is the latest "safe"
network not causing the bug, passed with regular bounds.

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 160456 W: 42658 L: 42175 D: 75623
Ptnml(0-2): 487, 17638, 43469, 18173, 461
https://tests.stockfishchess.org/tests/view/63e89836d62a5d02b0fa82c8

closes https://github.com/official-stockfish/Stockfish/pull/4391
closes https://github.com/official-stockfish/Stockfish/pull/4394

No functional change
---
 src/nnue/layers/simd.h | 33 ++++++++++++++++++---------------
 1 file changed, 18 insertions(+), 15 deletions(-)

diff --git a/src/nnue/layers/simd.h b/src/nnue/layers/simd.h
index 231f7891..381e7a68 100644
--- a/src/nnue/layers/simd.h
+++ b/src/nnue/layers/simd.h
@@ -165,18 +165,19 @@ namespace Stockfish::Simd {
       __m512i tmp0 = _mm512_maddubs_epi16(a0, b0);
       __m512i tmp1 = _mm512_maddubs_epi16(a1, b1);
       asm(
-          "vpaddsw     %[tmp0], %[tmp1], %[tmp0]\n\t"
           "vpmaddwd    %[tmp0], %[ones], %[tmp0]\n\t"
+          "vpmaddwd    %[tmp1], %[ones], %[tmp1]\n\t"
+          "vpaddd      %[tmp0], %[tmp1], %[tmp0]\n\t"
           "vpaddd      %[acc], %[tmp0], %[acc]\n\t"
-          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0)
-          : [tmp1]"v"(tmp1), [ones]"v"(_mm512_set1_epi16(1))
+          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0), [tmp1]"+&v"(tmp1)
+          : [ones]"v"(_mm512_set1_epi16(1))
       );
 #   else
       __m512i product0 = _mm512_maddubs_epi16(a0, b0);
       __m512i product1 = _mm512_maddubs_epi16(a1, b1);
-      product0 = _mm512_adds_epi16(product0, product1);
       product0 = _mm512_madd_epi16(product0, _mm512_set1_epi16(1));
-      acc = _mm512_add_epi32(acc, product0);
+      product1 = _mm512_madd_epi16(product1, _mm512_set1_epi16(1));
+      acc = _mm512_add_epi32(acc, _mm512_add_epi32(product0, product1));
 #   endif
 # endif
     }
@@ -261,18 +262,19 @@ namespace Stockfish::Simd {
       __m256i tmp0 = _mm256_maddubs_epi16(a0, b0);
       __m256i tmp1 = _mm256_maddubs_epi16(a1, b1);
       asm(
-          "vpaddsw     %[tmp0], %[tmp1], %[tmp0]\n\t"
           "vpmaddwd    %[tmp0], %[ones], %[tmp0]\n\t"
+          "vpmaddwd    %[tmp1], %[ones], %[tmp1]\n\t"
+          "vpaddd      %[tmp0], %[tmp1], %[tmp0]\n\t"
           "vpaddd      %[acc], %[tmp0], %[acc]\n\t"
-          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0)
-          : [tmp1]"v"(tmp1), [ones]"v"(_mm256_set1_epi16(1))
+          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0), [tmp1]"+&v"(tmp1)
+          : [ones]"v"(_mm256_set1_epi16(1))
       );
 #   else
       __m256i product0 = _mm256_maddubs_epi16(a0, b0);
       __m256i product1 = _mm256_maddubs_epi16(a1, b1);
-      product0 = _mm256_adds_epi16(product0, product1);
       product0 = _mm256_madd_epi16(product0, _mm256_set1_epi16(1));
-      acc = _mm256_add_epi32(acc, product0);
+      product1 = _mm256_madd_epi16(product1, _mm256_set1_epi16(1));
+      acc = _mm256_add_epi32(acc, _mm256_add_epi32(product0, product1));
 #   endif
 # endif
     }
@@ -326,18 +328,19 @@ namespace Stockfish::Simd {
       __m128i tmp0 = _mm_maddubs_epi16(a0, b0);
       __m128i tmp1 = _mm_maddubs_epi16(a1, b1);
       asm(
-          "paddsw     %[tmp1], %[tmp0]\n\t"
           "pmaddwd    %[ones], %[tmp0]\n\t"
+          "pmaddwd    %[ones], %[tmp1]\n\t"
+          "paddd      %[tmp1], %[tmp0]\n\t"
           "paddd      %[tmp0], %[acc]\n\t"
-          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0)
-          : [tmp1]"v"(tmp1), [ones]"v"(_mm_set1_epi16(1))
+          : [acc]"+v"(acc), [tmp0]"+&v"(tmp0), [tmp1]"+&v"(tmp1)
+          : [ones]"v"(_mm_set1_epi16(1))
       );
 #   else
       __m128i product0 = _mm_maddubs_epi16(a0, b0);
       __m128i product1 = _mm_maddubs_epi16(a1, b1);
-      product0 = _mm_adds_epi16(product0, product1);
       product0 = _mm_madd_epi16(product0, _mm_set1_epi16(1));
-      acc = _mm_add_epi32(acc, product0);
+      product1 = _mm_madd_epi16(product1, _mm_set1_epi16(1));
+      acc = _mm_add_epi32(acc, _mm_add_epi32(product0, product1));
 #   endif
     }
 

From 29c1e072b669c2257e4b48094391e7dc39fb31a5 Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Sun, 12 Feb 2023 17:33:19 -0800
Subject: [PATCH 068/192] Simplify nnueComplexity calculation.

further simplification after https://github.com/official-stockfish/Stockfish/pull/4377

STC https://tests.stockfishchess.org/tests/view/63e02a3773223e7f52ad8190
LLR: 2.97 (-2.94,2.94) <-1.75,0.25>
Total: 359072 W: 94605 L: 94733 D: 169734
Ptnml(0-2): 994, 39874, 97958, 39686, 1024

LTC https://tests.stockfishchess.org/tests/view/63e3fd12b5f425d71f77002a
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 248424 W: 66020 L: 66030 D: 116374
Ptnml(0-2): 113, 24653, 74689, 24645, 112

closes https://github.com/official-stockfish/Stockfish/pull/4390

bench: 4098325
---
 src/evaluate.cpp | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 6d5a8a0c..080d412b 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1072,8 +1072,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
 
       // Blend nnue complexity with (semi)classical complexity
       nnueComplexity = (  406 * nnueComplexity
-                        + 424 * abs(psq - nnue)
-                        + int(optimism) * int(psq - nnue)
+                        + (424 + optimism) * abs(psq - nnue)
                         ) / 1024;
 
       // Return hybrid NNUE complexity to caller

From 085cace4574bf561472d8d3d3afe50c2c536b4e3 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Sun, 22 Jan 2023 19:45:46 -0600
Subject: [PATCH 069/192] Simplify late countermove bonus condition

STC: https://tests.stockfishchess.org/tests/view/63d53ac6a67dd929a555e1e2
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 216096 W: 56862 L: 56839 D: 102395
Ptnml(0-2): 648, 24033, 58650, 24082, 635

LTC: https://tests.stockfishchess.org/tests/view/63d7f9a6a67dd929a5565991
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 808512 W: 214060 L: 214610 D: 379842
Ptnml(0-2): 301, 79448, 245293, 78928, 286

closes https://github.com/official-stockfish/Stockfish/pull/4392

Bench: 4283297
---
 src/search.cpp | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 2eed74b8..eccb97fd 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1371,11 +1371,10 @@ moves_loop: // When in check, search starts here
                          quietsSearched, quietCount, capturesSearched, captureCount, depth);
 
     // Bonus for prior countermove that caused the fail low
-    else if (   (depth >= 5 || PvNode || bestValue < alpha - 65 * depth)
-             && !priorCapture)
+    else if (!priorCapture)
     {
         // Extra bonuses for PV/Cut nodes or bad fail lows
-        int bonus = 1 + (PvNode || cutNode) + (bestValue < alpha - 88 * depth);
+        int bonus = (depth > 4) + (PvNode || cutNode) + (bestValue < alpha - 88 * depth);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 

From 037ef3e18dc7f5455cc671995ae38d5b4d1fce4a Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Tue, 14 Feb 2023 18:04:17 -0600
Subject: [PATCH 070/192] Remove one `reduction` call

even though bench is unchanged to depth 28,
due to adjusting depth in singular extensions this might be functional.

STC: https://tests.stockfishchess.org/tests/view/63ec21affe833123fef34153
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 195712 W: 51625 L: 51581 D: 92506
Ptnml(0-2): 504, 20527, 55779, 20513, 533

LTC: https://tests.stockfishchess.org/tests/view/63ed3487fe833123fef375ed
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 32176 W: 8631 L: 8442 D: 15103
Ptnml(0-2): 5, 2794, 10309, 2967, 13

closes https://github.com/official-stockfish/Stockfish/pull/4395

Bench 4283297
---
 src/search.cpp | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index eccb97fd..6ca2cfa5 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -987,6 +987,8 @@ moves_loop: // When in check, search starts here
 
       Value delta = beta - alpha;
 
+      Depth r = reduction(improving, depth, moveCount, delta, thisThread->rootDelta);
+
       // Step 14. Pruning at shallow depth (~120 Elo). Depth conditions are important for mate finding.
       if (  !rootNode
           && pos.non_pawn_material(us)
@@ -996,7 +998,7 @@ moves_loop: // When in check, search starts here
           moveCountPruning = moveCount >= futility_move_count(improving, depth);
 
           // Reduced depth of the next LMR search
-          int lmrDepth = std::max(newDepth - reduction(improving, depth, moveCount, delta, thisThread->rootDelta), 0);
+          int lmrDepth = std::max(newDepth - r, 0);
 
           if (   capture
               || givesCheck)
@@ -1133,8 +1135,6 @@ moves_loop: // When in check, search starts here
       // Step 16. Make the move
       pos.do_move(move, st, givesCheck);
 
-      Depth r = reduction(improving, depth, moveCount, delta, thisThread->rootDelta);
-
       // Decrease reduction if position is or has been on the PV
       // and node is not likely to fail low. (~3 Elo)
       if (   ss->ttPv

From b4ad3a3c4b68f9c8736f444aeb3364f833247fdc Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Tue, 21 Feb 2023 22:18:17 +0100
Subject: [PATCH 071/192] Add support for ARM dot product instructions

The sdot instruction computes (and accumulates) a signed dot product,
which is quite handy for Stockfish's NNUE code. The instruction is
optional for Armv8.2 and Armv8.3, and mandatory for Armv8.4 and above.

The commit adds a new 'arm-dotprod' architecture with enabled dot
product support. It also enables dot product support for the existing
'apple-silicon' architecture, which is at least Armv8.5.

The following local speed test was performed on an Apple M1 with
ARCH=apple-silicon. I had to remove CPU pinning from the benchmark
script. However, the results were still consistent: Checking both
binaries against themselves reported a speedup of +0.0000 and +0.0005,
respectively.

```
Result of 100 runs
==================
base (...ish.037ef3e1) =    1917997  +/- 7152
test (...fish.dotprod) =    2159682  +/- 9066
diff                   =    +241684  +/- 2923

speedup        = +0.1260
P(speedup > 0) =  1.0000

CPU: 10 x arm
Hyperthreading: off
```

Fixes #4193

closes https://github.com/official-stockfish/Stockfish/pull/4400

No functional change
---
 src/Makefile                       | 65 +++++++++++++++++++-----------
 src/nnue/layers/affine_transform.h | 24 +++++++++++
 src/nnue/layers/simd.h             | 13 ++++++
 3 files changed, 78 insertions(+), 24 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index 775c72c3..3d6432fd 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -69,32 +69,33 @@ VPATH = syzygy:nnue:nnue/features
 ### Section 2. High-level Configuration
 ### ==========================================================================
 #
-# flag                --- Comp switch      --- Description
+# flag                --- Comp switch        --- Description
 # ----------------------------------------------------------------------------
 #
-# debug = yes/no      --- -DNDEBUG         --- Enable/Disable debug mode
+# debug = yes/no      --- -DNDEBUG           --- Enable/Disable debug mode
 # sanitize = none/<sanitizer> ... (-fsanitize )
-#                     --- ( undefined )    --- enable undefined behavior checks
-#                     --- ( thread    )    --- enable threading error checks
-#                     --- ( address   )    --- enable memory access checks
-#                     --- ...etc...        --- see compiler documentation for supported sanitizers
-# optimize = yes/no   --- (-O3/-fast etc.) --- Enable/Disable optimizations
-# arch = (name)       --- (-arch)          --- Target architecture
-# bits = 64/32        --- -DIS_64BIT       --- 64-/32-bit operating system
-# prefetch = yes/no   --- -DUSE_PREFETCH   --- Use prefetch asm-instruction
-# popcnt = yes/no     --- -DUSE_POPCNT     --- Use popcnt asm-instruction
-# pext = yes/no       --- -DUSE_PEXT       --- Use pext x86_64 asm-instruction
-# sse = yes/no        --- -msse            --- Use Intel Streaming SIMD Extensions
-# mmx = yes/no        --- -mmmx            --- Use Intel MMX instructions
-# sse2 = yes/no       --- -msse2           --- Use Intel Streaming SIMD Extensions 2
-# ssse3 = yes/no      --- -mssse3          --- Use Intel Supplemental Streaming SIMD Extensions 3
-# sse41 = yes/no      --- -msse4.1         --- Use Intel Streaming SIMD Extensions 4.1
-# avx2 = yes/no       --- -mavx2           --- Use Intel Advanced Vector Extensions 2
-# avxvnni = yes/no    --- -mavxvnni        --- Use Intel Vector Neural Network Instructions AVX
-# avx512 = yes/no     --- -mavx512bw       --- Use Intel Advanced Vector Extensions 512
-# vnni256 = yes/no    --- -mavx512vnni     --- Use Intel Vector Neural Network Instructions 256
-# vnni512 = yes/no    --- -mavx512vnni     --- Use Intel Vector Neural Network Instructions 512
-# neon = yes/no       --- -DUSE_NEON       --- Use ARM SIMD architecture
+#                     --- ( undefined )      --- enable undefined behavior checks
+#                     --- ( thread    )      --- enable threading error checks
+#                     --- ( address   )      --- enable memory access checks
+#                     --- ...etc...          --- see compiler documentation for supported sanitizers
+# optimize = yes/no   --- (-O3/-fast etc.)   --- Enable/Disable optimizations
+# arch = (name)       --- (-arch)            --- Target architecture
+# bits = 64/32        --- -DIS_64BIT         --- 64-/32-bit operating system
+# prefetch = yes/no   --- -DUSE_PREFETCH     --- Use prefetch asm-instruction
+# popcnt = yes/no     --- -DUSE_POPCNT       --- Use popcnt asm-instruction
+# pext = yes/no       --- -DUSE_PEXT         --- Use pext x86_64 asm-instruction
+# sse = yes/no        --- -msse              --- Use Intel Streaming SIMD Extensions
+# mmx = yes/no        --- -mmmx              --- Use Intel MMX instructions
+# sse2 = yes/no       --- -msse2             --- Use Intel Streaming SIMD Extensions 2
+# ssse3 = yes/no      --- -mssse3            --- Use Intel Supplemental Streaming SIMD Extensions 3
+# sse41 = yes/no      --- -msse4.1           --- Use Intel Streaming SIMD Extensions 4.1
+# avx2 = yes/no       --- -mavx2             --- Use Intel Advanced Vector Extensions 2
+# avxvnni = yes/no    --- -mavxvnni          --- Use Intel Vector Neural Network Instructions AVX
+# avx512 = yes/no     --- -mavx512bw         --- Use Intel Advanced Vector Extensions 512
+# vnni256 = yes/no    --- -mavx512vnni       --- Use Intel Vector Neural Network Instructions 256
+# vnni512 = yes/no    --- -mavx512vnni       --- Use Intel Vector Neural Network Instructions 512
+# neon = yes/no       --- -DUSE_NEON         --- Use ARM SIMD architecture
+# dotprod = yes/no    --- -DUSE_NEON_DOTPROD --- Use ARM advanced SIMD Int8 dot product instructions
 #
 # 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
@@ -116,7 +117,7 @@ ifeq ($(ARCH), $(filter $(ARCH), \
                  x86-64-vnni512 x86-64-vnni256 x86-64-avx512 x86-64-avxvnni x86-64-bmi2 \
                  x86-64-avx2 x86-64-sse41-popcnt x86-64-modern x86-64-ssse3 x86-64-sse3-popcnt \
                  x86-64 x86-32-sse41-popcnt x86-32-sse2 x86-32 ppc-64 ppc-32 e2k \
-                 armv7 armv7-neon armv8 apple-silicon general-64 general-32 riscv64))
+                 armv7 armv7-neon armv8 armv8-dotprod apple-silicon general-64 general-32 riscv64))
    SUPPORTED_ARCH=true
 else
    SUPPORTED_ARCH=false
@@ -140,6 +141,7 @@ avx512 = no
 vnni256 = no
 vnni512 = no
 neon = no
+dotprod = no
 arm_version = 0
 STRIP = strip
 
@@ -308,11 +310,21 @@ ifeq ($(ARCH),armv8)
 	arm_version = 8
 endif
 
+ifeq ($(ARCH),armv8-dotprod)
+	arch = armv8
+	prefetch = yes
+	popcnt = yes
+	neon = yes
+	dotprod = yes
+	arm_version = 8
+endif
+
 ifeq ($(ARCH),apple-silicon)
 	arch = arm64
 	prefetch = yes
 	popcnt = yes
 	neon = yes
+	dotprod = yes
 	arm_version = 8
 endif
 
@@ -675,6 +687,10 @@ ifeq ($(neon),yes)
 	endif
 endif
 
+ifeq ($(dotprod),yes)
+	CXXFLAGS += -march=armv8.2-a+dotprod -DUSE_NEON_DOTPROD
+endif
+
 ### 3.7 pext
 ifeq ($(pext),yes)
 	CXXFLAGS += -DUSE_PEXT
@@ -776,6 +792,7 @@ help:
 	@echo "armv7                   > ARMv7 32-bit"
 	@echo "armv7-neon              > ARMv7 32-bit with popcnt and neon"
 	@echo "armv8                   > ARMv8 64-bit with popcnt and neon"
+	@echo "armv8-dotprod           > ARMv8 64-bit with popcnt, neon and dot product support"
 	@echo "e2k                     > Elbrus 2000"
 	@echo "apple-silicon           > Apple silicon ARM64"
 	@echo "general-64              > unspecified 64-bit"
diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index 363b4916..63b58af3 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -72,6 +72,10 @@ namespace Stockfish::Eval::NNUE::Layers {
     const __m64 Zeros = _mm_setzero_si64();
     const auto inputVector = reinterpret_cast<const __m64*>(input);
 
+# elif defined(USE_NEON_DOTPROD)
+    constexpr IndexType NumChunks = ceil_to_multiple<IndexType>(InputDimensions, 16) / 16;
+    const auto inputVector = reinterpret_cast<const int8x16_t*>(input);
+
 # elif defined(USE_NEON)
     constexpr IndexType NumChunks = ceil_to_multiple<IndexType>(InputDimensions, 16) / 16;
     const auto inputVector = reinterpret_cast<const int8x8_t*>(input);
@@ -123,6 +127,14 @@ namespace Stockfish::Eval::NNUE::Layers {
       sum = _mm_add_pi32(sum, _mm_unpackhi_pi32(sum, sum));
       output[i] = _mm_cvtsi64_si32(sum);
 
+# elif defined(USE_NEON_DOTPROD)
+      int32x4_t sum = {biases[i]};
+      const auto row = reinterpret_cast<const int8x16_t*>(&weights[offset]);
+      for (IndexType j = 0; j < NumChunks; ++j) {
+        sum = vdotq_s32(sum, inputVector[j], row[j]);
+      }
+      output[i] = vaddvq_s32(sum);
+
 # elif defined(USE_NEON)
       int32x4_t sum = {biases[i]};
       const auto row = reinterpret_cast<const int8x8_t*>(&weights[offset]);
@@ -187,6 +199,9 @@ namespace Stockfish::Eval::NNUE::Layers {
 #elif defined (USE_SSSE3)
     static constexpr IndexType InputSimdWidth = 16;
     static constexpr IndexType MaxNumOutputRegs = 8;
+#elif defined (USE_NEON_DOTPROD)
+    static constexpr IndexType InputSimdWidth = 16;
+    static constexpr IndexType MaxNumOutputRegs = 8;
 #elif defined (USE_NEON)
     static constexpr IndexType InputSimdWidth = 8;
     static constexpr IndexType MaxNumOutputRegs = 8;
@@ -292,6 +307,15 @@ namespace Stockfish::Eval::NNUE::Layers {
       #define vec_add_dpbusd_32x2 Simd::m128_add_dpbusd_epi32x2
       #define vec_hadd Simd::m128_hadd
       #define vec_haddx4 Simd::m128_haddx4
+#elif defined (USE_NEON_DOTPROD)
+      using acc_vec_t = int32x4_t;
+      using bias_vec_t = int32x4_t;
+      using weight_vec_t = int8x16_t;
+      using in_vec_t = int8x16_t;
+      #define vec_zero {0}
+      #define vec_add_dpbusd_32x2 Simd::dotprod_m128_add_dpbusd_epi32x2
+      #define vec_hadd Simd::neon_m128_hadd
+      #define vec_haddx4 Simd::neon_m128_haddx4
 #elif defined (USE_NEON)
       using acc_vec_t = int32x4_t;
       using bias_vec_t = int32x4_t;
diff --git a/src/nnue/layers/simd.h b/src/nnue/layers/simd.h
index 381e7a68..22c51980 100644
--- a/src/nnue/layers/simd.h
+++ b/src/nnue/layers/simd.h
@@ -346,6 +346,19 @@ namespace Stockfish::Simd {
 
 #endif
 
+#if defined (USE_NEON_DOTPROD)
+
+    [[maybe_unused]] static void dotprod_m128_add_dpbusd_epi32x2(
+        int32x4_t& acc,
+        int8x16_t a0, int8x16_t b0,
+        int8x16_t a1, int8x16_t b1) {
+
+        acc = vdotq_s32(acc, a0, b0);
+        acc = vdotq_s32(acc, a1, b1);
+    }
+
+#endif
+
 #if defined (USE_NEON)
 
     [[maybe_unused]] static int neon_m128_reduce_add_epi32(int32x4_t s) {

From 77dfcbedce2861b2c6c5056d49e7a8731fea4256 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sun, 19 Feb 2023 11:25:10 +0100
Subject: [PATCH 072/192] Remove unused macros

closes https://github.com/official-stockfish/Stockfish/pull/4397

No functional change
---
 src/nnue/layers/affine_transform.h | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index 63b58af3..313b1568 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -480,18 +480,14 @@ namespace Stockfish::Eval::NNUE::Layers {
       #define vec_set_32 _mm256_set1_epi32
       #define vec_add_dpbusd_32 Simd::m256_add_dpbusd_epi32
       #define vec_add_dpbusd_32x2 Simd::m256_add_dpbusd_epi32x2
-      #define vec_add_dpbusd_32x4 Simd::m256_add_dpbusd_epi32x4
       #define vec_hadd Simd::m256_hadd
-      #define vec_haddx4 Simd::m256_haddx4
 #elif defined (USE_SSSE3)
       using vec_t = __m128i;
       #define vec_setzero _mm_setzero_si128
       #define vec_set_32 _mm_set1_epi32
       #define vec_add_dpbusd_32 Simd::m128_add_dpbusd_epi32
       #define vec_add_dpbusd_32x2 Simd::m128_add_dpbusd_epi32x2
-      #define vec_add_dpbusd_32x4 Simd::m128_add_dpbusd_epi32x4
       #define vec_hadd Simd::m128_hadd
-      #define vec_haddx4 Simd::m128_haddx4
 #endif
 
 #if defined (USE_SSSE3)
@@ -542,9 +538,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 # undef vec_set_32
 # undef vec_add_dpbusd_32
 # undef vec_add_dpbusd_32x2
-# undef vec_add_dpbusd_32x4
 # undef vec_hadd
-# undef vec_haddx4
 #else
       // Use old implementation for the other architectures.
       affine_transform_non_ssse3<

From 08385527dd470ece814ac85013802995a0e7f6ca Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Mon, 20 Feb 2023 20:02:55 +0100
Subject: [PATCH 073/192] Introduce a function to compute NNUE accumulator

This patch introduces `hint_common_parent_position()` to signal that potentially several child nodes will require an NNUE eval. By populating explicitly the accumulator, these subsequent evaluations can be performed more efficiently.

This was based on the observation that calculating the evaluation in an excluded move position yielded a significant Elo gain, even though the evaluation itself was already available (work by pb00067).

Sopel wrote the code to perform just the accumulator update. This PR is based on cleaned up code that

passed STC:
https://tests.stockfishchess.org/tests/view/63f62f9be74a12625bcd4aa0
 LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 110368 W: 29607 L: 29167 D: 51594
Ptnml(0-2): 41, 10551, 33572, 10967, 53

and in an the earlier (equivalent) version

passed STC:
https://tests.stockfishchess.org/tests/view/63f3c3fee74a12625bcce2a6
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 47552 W: 12786 L: 12467 D: 22299
Ptnml(0-2): 120, 5107, 12997, 5438, 114

passed LTC:
https://tests.stockfishchess.org/tests/view/63f45cc2e74a12625bccfa63
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 110368 W: 29607 L: 29167 D: 51594
Ptnml(0-2): 41, 10551, 33572, 10967, 53

closes https://github.com/official-stockfish/Stockfish/pull/4402

Bench: 3726250
---
 src/evaluate.h                      |   1 +
 src/nnue/evaluate_nnue.cpp          |   5 +
 src/nnue/evaluate_nnue.h            |   1 +
 src/nnue/nnue_feature_transformer.h | 425 +++++++++++++++++-----------
 src/search.cpp                      |   8 +-
 5 files changed, 266 insertions(+), 174 deletions(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index cdea2ab2..46f20259 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -45,6 +45,7 @@ namespace Eval {
 
     std::string trace(Position& pos);
     Value evaluate(const Position& pos, bool adjusted = false, int* complexity = nullptr);
+    void hint_common_parent_position(const Position& pos);
 
     void init();
     void verify();
diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index f132de71..f33aa3b8 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -136,6 +136,11 @@ namespace Stockfish::Eval::NNUE {
     return (bool)stream;
   }
 
+  void hint_common_parent_position(const Position& pos) {
+    if (Eval::useNNUE)
+        featureTransformer->hint_common_access(pos);
+  }
+
   // Evaluation function. Perform differential calculation.
   Value evaluate(const Position& pos, bool adjusted, int* complexity) {
 
diff --git a/src/nnue/evaluate_nnue.h b/src/nnue/evaluate_nnue.h
index 9499f7d9..15638cae 100644
--- a/src/nnue/evaluate_nnue.h
+++ b/src/nnue/evaluate_nnue.h
@@ -31,6 +31,7 @@ namespace Stockfish::Eval::NNUE {
   constexpr std::uint32_t HashValue =
       FeatureTransformer::get_hash_value() ^ Network::get_hash_value();
 
+
   // Deleter for automating release of memory area
   template <typename T>
   struct AlignedDeleter {
diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index 62f1615d..13f1604f 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -25,6 +25,7 @@
 #include "nnue_architecture.h"
 
 #include <cstring> // std::memset()
+#include <utility> // std::pair
 
 namespace Stockfish::Eval::NNUE {
 
@@ -332,27 +333,16 @@ namespace Stockfish::Eval::NNUE {
 #endif
 
       return psqt;
+    } // end of function transform()
 
-   } // end of function transform()
-
-
+    void hint_common_access(const Position& pos) const {
+      hint_common_access_for_perspective<WHITE>(pos);
+      hint_common_access_for_perspective<BLACK>(pos);
+    }
 
    private:
     template<Color Perspective>
-    void update_accumulator(const Position& pos) const {
-
-      // The size must be enough to contain the largest possible update.
-      // That might depend on the feature set and generally relies on the
-      // feature set's update cost calculation to be correct and never
-      // allow updates with more added/removed features than MaxActiveDimensions.
-
-  #ifdef VECTOR
-      // Gcc-10.2 unnecessarily spills AVX2 registers if this array
-      // is defined in the VECTOR code below, once in each branch
-      vec_t acc[NumRegs];
-      psqt_vec_t psqt[NumPsqtRegs];
-  #endif
-
+    [[nodiscard]] std::pair<StateInfo*, StateInfo*> try_find_computed_accumulator(const Position& pos) const {
       // Look for a usable accumulator of an earlier position. We keep track
       // of the estimated gain in terms of features to be added/subtracted.
       StateInfo *st = pos.state(), *next = nullptr;
@@ -367,218 +357,313 @@ namespace Stockfish::Eval::NNUE {
         next = st;
         st = st->previous;
       }
+      return { st, next };
+    }
 
-      if (st->accumulator.computed[Perspective])
-      {
-        if (next == nullptr)
-          return;
+    // NOTE: The parameter states_to_update is an array of position states, ending with nullptr.
+    //       All states must be sequential, that is states_to_update[i] must either be reachable
+    //       by repeatedly applying ->previous from states_to_update[i+1] or states_to_update[i] == nullptr.
+    //       computed_st must be reachable by repeatadly applying ->previous on states_to_update[0], if not nullptr.
+    template<Color Perspective, size_t N>
+    void update_accumulator_incremetal(const Position& pos, StateInfo* computed_st, StateInfo* states_to_update[N]) const {
+      static_assert(N > 0);
+      assert(states_to_update[N-1] == nullptr);
 
-        // Update incrementally in two steps. First, we update the "next"
-        // accumulator. Then, we update the current accumulator (pos.state()).
-
-        // Gather all features to be updated.
-        const Square ksq = pos.square<KING>(Perspective);
-        FeatureSet::IndexList removed[2], added[2];
-        FeatureSet::append_changed_indices<Perspective>(
-          ksq, next->dirtyPiece, removed[0], added[0]);
-        for (StateInfo *st2 = pos.state(); st2 != next; st2 = st2->previous)
-          FeatureSet::append_changed_indices<Perspective>(
-            ksq, st2->dirtyPiece, removed[1], added[1]);
-
-        // Mark the accumulators as computed.
-        next->accumulator.computed[Perspective] = true;
-        pos.state()->accumulator.computed[Perspective] = true;
-
-        // Now update the accumulators listed in states_to_update[], where the last element is a sentinel.
-        StateInfo *states_to_update[3] =
-          { next, next == pos.state() ? nullptr : pos.state(), nullptr };
   #ifdef VECTOR
-        for (IndexType j = 0; j < HalfDimensions / TileHeight; ++j)
+      // Gcc-10.2 unnecessarily spills AVX2 registers if this array
+      // is defined in the VECTOR code below, once in each branch
+      vec_t acc[NumRegs];
+      psqt_vec_t psqt[NumPsqtRegs];
+  #endif
+
+      if (states_to_update[0] == nullptr)
+        return;
+
+      // Update incrementally going back through states_to_update.
+
+      // Gather all features to be updated.
+      const Square ksq = pos.square<KING>(Perspective);
+
+      // The size must be enough to contain the largest possible update.
+      // That might depend on the feature set and generally relies on the
+      // feature set's update cost calculation to be correct and never
+      // allow updates with more added/removed features than MaxActiveDimensions.
+      FeatureSet::IndexList removed[N-1], added[N-1];
+
+      {
+        int i = N-2; // last potential state to update. Skip last element because it must be nullptr.
+        while (states_to_update[i] == nullptr)
+          --i;
+
+        StateInfo *st2 = states_to_update[i];
+
+        for (; i >= 0; --i)
         {
-          // Load accumulator
-          auto accTile = reinterpret_cast<vec_t*>(
-            &st->accumulator.accumulation[Perspective][j * TileHeight]);
-          for (IndexType k = 0; k < NumRegs; ++k)
-            acc[k] = vec_load(&accTile[k]);
+          states_to_update[i]->accumulator.computed[Perspective] = true;
 
-          for (IndexType i = 0; states_to_update[i]; ++i)
-          {
-            // Difference calculation for the deactivated features
-            for (const auto index : removed[i])
-            {
-              const IndexType offset = HalfDimensions * index + j * TileHeight;
-              auto column = reinterpret_cast<const vec_t*>(&weights[offset]);
-              for (IndexType k = 0; k < NumRegs; ++k)
-                acc[k] = vec_sub_16(acc[k], column[k]);
-            }
+          StateInfo* end_state = i == 0 ? computed_st : states_to_update[i - 1];
 
-            // Difference calculation for the activated features
-            for (const auto index : added[i])
-            {
-              const IndexType offset = HalfDimensions * index + j * TileHeight;
-              auto column = reinterpret_cast<const vec_t*>(&weights[offset]);
-              for (IndexType k = 0; k < NumRegs; ++k)
-                acc[k] = vec_add_16(acc[k], column[k]);
-            }
-
-            // Store accumulator
-            accTile = reinterpret_cast<vec_t*>(
-              &states_to_update[i]->accumulator.accumulation[Perspective][j * TileHeight]);
-            for (IndexType k = 0; k < NumRegs; ++k)
-              vec_store(&accTile[k], acc[k]);
-          }
+          for (; st2 != end_state; st2 = st2->previous)
+            FeatureSet::append_changed_indices<Perspective>(
+              ksq, st2->dirtyPiece, removed[i], added[i]);
         }
+      }
 
-        for (IndexType j = 0; j < PSQTBuckets / PsqtTileHeight; ++j)
-        {
-          // Load accumulator
-          auto accTilePsqt = reinterpret_cast<psqt_vec_t*>(
-            &st->accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
-          for (std::size_t k = 0; k < NumPsqtRegs; ++k)
-            psqt[k] = vec_load_psqt(&accTilePsqt[k]);
+      StateInfo* st = computed_st;
 
-          for (IndexType i = 0; states_to_update[i]; ++i)
-          {
-            // Difference calculation for the deactivated features
-            for (const auto index : removed[i])
-            {
-              const IndexType offset = PSQTBuckets * index + j * PsqtTileHeight;
-              auto columnPsqt = reinterpret_cast<const psqt_vec_t*>(&psqtWeights[offset]);
-              for (std::size_t k = 0; k < NumPsqtRegs; ++k)
-                psqt[k] = vec_sub_psqt_32(psqt[k], columnPsqt[k]);
-            }
+      // Now update the accumulators listed in states_to_update[], where the last element is a sentinel.
+#ifdef VECTOR
+      for (IndexType j = 0; j < HalfDimensions / TileHeight; ++j)
+      {
+        // Load accumulator
+        auto accTile = reinterpret_cast<vec_t*>(
+          &st->accumulator.accumulation[Perspective][j * TileHeight]);
+        for (IndexType k = 0; k < NumRegs; ++k)
+          acc[k] = vec_load(&accTile[k]);
 
-            // Difference calculation for the activated features
-            for (const auto index : added[i])
-            {
-              const IndexType offset = PSQTBuckets * index + j * PsqtTileHeight;
-              auto columnPsqt = reinterpret_cast<const psqt_vec_t*>(&psqtWeights[offset]);
-              for (std::size_t k = 0; k < NumPsqtRegs; ++k)
-                psqt[k] = vec_add_psqt_32(psqt[k], columnPsqt[k]);
-            }
-
-            // Store accumulator
-            accTilePsqt = reinterpret_cast<psqt_vec_t*>(
-              &states_to_update[i]->accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
-            for (std::size_t k = 0; k < NumPsqtRegs; ++k)
-              vec_store_psqt(&accTilePsqt[k], psqt[k]);
-          }
-        }
-
-  #else
         for (IndexType i = 0; states_to_update[i]; ++i)
         {
-          std::memcpy(states_to_update[i]->accumulator.accumulation[Perspective],
-              st->accumulator.accumulation[Perspective],
-              HalfDimensions * sizeof(BiasType));
-
-          for (std::size_t k = 0; k < PSQTBuckets; ++k)
-            states_to_update[i]->accumulator.psqtAccumulation[Perspective][k] = st->accumulator.psqtAccumulation[Perspective][k];
-
-          st = states_to_update[i];
-
           // Difference calculation for the deactivated features
           for (const auto index : removed[i])
           {
-            const IndexType offset = HalfDimensions * index;
-
-            for (IndexType j = 0; j < HalfDimensions; ++j)
-              st->accumulator.accumulation[Perspective][j] -= weights[offset + j];
-
-            for (std::size_t k = 0; k < PSQTBuckets; ++k)
-              st->accumulator.psqtAccumulation[Perspective][k] -= psqtWeights[index * PSQTBuckets + k];
+            const IndexType offset = HalfDimensions * index + j * TileHeight;
+            auto column = reinterpret_cast<const vec_t*>(&weights[offset]);
+            for (IndexType k = 0; k < NumRegs; ++k)
+              acc[k] = vec_sub_16(acc[k], column[k]);
           }
 
           // Difference calculation for the activated features
           for (const auto index : added[i])
-          {
-            const IndexType offset = HalfDimensions * index;
-
-            for (IndexType j = 0; j < HalfDimensions; ++j)
-              st->accumulator.accumulation[Perspective][j] += weights[offset + j];
-
-            for (std::size_t k = 0; k < PSQTBuckets; ++k)
-              st->accumulator.psqtAccumulation[Perspective][k] += psqtWeights[index * PSQTBuckets + k];
-          }
-        }
-  #endif
-      }
-      else
-      {
-        // Refresh the accumulator
-        auto& accumulator = pos.state()->accumulator;
-        accumulator.computed[Perspective] = true;
-        FeatureSet::IndexList active;
-        FeatureSet::append_active_indices<Perspective>(pos, active);
-
-  #ifdef VECTOR
-        for (IndexType j = 0; j < HalfDimensions / TileHeight; ++j)
-        {
-          auto biasesTile = reinterpret_cast<const vec_t*>(
-              &biases[j * TileHeight]);
-          for (IndexType k = 0; k < NumRegs; ++k)
-            acc[k] = biasesTile[k];
-
-          for (const auto index : active)
           {
             const IndexType offset = HalfDimensions * index + j * TileHeight;
             auto column = reinterpret_cast<const vec_t*>(&weights[offset]);
-
-            for (unsigned k = 0; k < NumRegs; ++k)
+            for (IndexType k = 0; k < NumRegs; ++k)
               acc[k] = vec_add_16(acc[k], column[k]);
           }
 
-          auto accTile = reinterpret_cast<vec_t*>(
-              &accumulator.accumulation[Perspective][j * TileHeight]);
-          for (unsigned k = 0; k < NumRegs; k++)
+          // Store accumulator
+          accTile = reinterpret_cast<vec_t*>(
+            &states_to_update[i]->accumulator.accumulation[Perspective][j * TileHeight]);
+          for (IndexType k = 0; k < NumRegs; ++k)
             vec_store(&accTile[k], acc[k]);
         }
+      }
 
-        for (IndexType j = 0; j < PSQTBuckets / PsqtTileHeight; ++j)
+      for (IndexType j = 0; j < PSQTBuckets / PsqtTileHeight; ++j)
+      {
+        // Load accumulator
+        auto accTilePsqt = reinterpret_cast<psqt_vec_t*>(
+          &st->accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
+        for (std::size_t k = 0; k < NumPsqtRegs; ++k)
+          psqt[k] = vec_load_psqt(&accTilePsqt[k]);
+
+        for (IndexType i = 0; states_to_update[i]; ++i)
         {
-          for (std::size_t k = 0; k < NumPsqtRegs; ++k)
-            psqt[k] = vec_zero_psqt();
-
-          for (const auto index : active)
+          // Difference calculation for the deactivated features
+          for (const auto index : removed[i])
           {
             const IndexType offset = PSQTBuckets * index + j * PsqtTileHeight;
             auto columnPsqt = reinterpret_cast<const psqt_vec_t*>(&psqtWeights[offset]);
+            for (std::size_t k = 0; k < NumPsqtRegs; ++k)
+              psqt[k] = vec_sub_psqt_32(psqt[k], columnPsqt[k]);
+          }
 
+          // Difference calculation for the activated features
+          for (const auto index : added[i])
+          {
+            const IndexType offset = PSQTBuckets * index + j * PsqtTileHeight;
+            auto columnPsqt = reinterpret_cast<const psqt_vec_t*>(&psqtWeights[offset]);
             for (std::size_t k = 0; k < NumPsqtRegs; ++k)
               psqt[k] = vec_add_psqt_32(psqt[k], columnPsqt[k]);
           }
 
-          auto accTilePsqt = reinterpret_cast<psqt_vec_t*>(
-            &accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
+          // Store accumulator
+          accTilePsqt = reinterpret_cast<psqt_vec_t*>(
+            &states_to_update[i]->accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
           for (std::size_t k = 0; k < NumPsqtRegs; ++k)
             vec_store_psqt(&accTilePsqt[k], psqt[k]);
         }
+      }
 
-  #else
-        std::memcpy(accumulator.accumulation[Perspective], biases,
+#else
+      for (IndexType i = 0; states_to_update[i]; ++i)
+      {
+        std::memcpy(states_to_update[i]->accumulator.accumulation[Perspective],
+            st->accumulator.accumulation[Perspective],
             HalfDimensions * sizeof(BiasType));
 
         for (std::size_t k = 0; k < PSQTBuckets; ++k)
-          accumulator.psqtAccumulation[Perspective][k] = 0;
+          states_to_update[i]->accumulator.psqtAccumulation[Perspective][k] = st->accumulator.psqtAccumulation[Perspective][k];
 
-        for (const auto index : active)
+        st = states_to_update[i];
+
+        // Difference calculation for the deactivated features
+        for (const auto index : removed[i])
         {
           const IndexType offset = HalfDimensions * index;
 
           for (IndexType j = 0; j < HalfDimensions; ++j)
-            accumulator.accumulation[Perspective][j] += weights[offset + j];
+            st->accumulator.accumulation[Perspective][j] -= weights[offset + j];
 
           for (std::size_t k = 0; k < PSQTBuckets; ++k)
-            accumulator.psqtAccumulation[Perspective][k] += psqtWeights[index * PSQTBuckets + k];
+            st->accumulator.psqtAccumulation[Perspective][k] -= psqtWeights[index * PSQTBuckets + k];
+        }
+
+        // Difference calculation for the activated features
+        for (const auto index : added[i])
+        {
+          const IndexType offset = HalfDimensions * index;
+
+          for (IndexType j = 0; j < HalfDimensions; ++j)
+            st->accumulator.accumulation[Perspective][j] += weights[offset + j];
+
+          for (std::size_t k = 0; k < PSQTBuckets; ++k)
+            st->accumulator.psqtAccumulation[Perspective][k] += psqtWeights[index * PSQTBuckets + k];
         }
-  #endif
       }
+#endif
 
   #if defined(USE_MMX)
       _mm_empty();
   #endif
     }
 
+    template<Color Perspective>
+    void update_accumulator_refresh(const Position& pos) const {
+  #ifdef VECTOR
+      // Gcc-10.2 unnecessarily spills AVX2 registers if this array
+      // is defined in the VECTOR code below, once in each branch
+      vec_t acc[NumRegs];
+      psqt_vec_t psqt[NumPsqtRegs];
+  #endif
+
+      // Refresh the accumulator
+      // Could be extracted to a separate function because it's done in 2 places,
+      // but it's unclear if compilers would correctly handle register allocation.
+      auto& accumulator = pos.state()->accumulator;
+      accumulator.computed[Perspective] = true;
+      FeatureSet::IndexList active;
+      FeatureSet::append_active_indices<Perspective>(pos, active);
+
+#ifdef VECTOR
+      for (IndexType j = 0; j < HalfDimensions / TileHeight; ++j)
+      {
+        auto biasesTile = reinterpret_cast<const vec_t*>(
+            &biases[j * TileHeight]);
+        for (IndexType k = 0; k < NumRegs; ++k)
+          acc[k] = biasesTile[k];
+
+        for (const auto index : active)
+        {
+          const IndexType offset = HalfDimensions * index + j * TileHeight;
+          auto column = reinterpret_cast<const vec_t*>(&weights[offset]);
+
+          for (unsigned k = 0; k < NumRegs; ++k)
+            acc[k] = vec_add_16(acc[k], column[k]);
+        }
+
+        auto accTile = reinterpret_cast<vec_t*>(
+            &accumulator.accumulation[Perspective][j * TileHeight]);
+        for (unsigned k = 0; k < NumRegs; k++)
+          vec_store(&accTile[k], acc[k]);
+      }
+
+      for (IndexType j = 0; j < PSQTBuckets / PsqtTileHeight; ++j)
+      {
+        for (std::size_t k = 0; k < NumPsqtRegs; ++k)
+          psqt[k] = vec_zero_psqt();
+
+        for (const auto index : active)
+        {
+          const IndexType offset = PSQTBuckets * index + j * PsqtTileHeight;
+          auto columnPsqt = reinterpret_cast<const psqt_vec_t*>(&psqtWeights[offset]);
+
+          for (std::size_t k = 0; k < NumPsqtRegs; ++k)
+            psqt[k] = vec_add_psqt_32(psqt[k], columnPsqt[k]);
+        }
+
+        auto accTilePsqt = reinterpret_cast<psqt_vec_t*>(
+          &accumulator.psqtAccumulation[Perspective][j * PsqtTileHeight]);
+        for (std::size_t k = 0; k < NumPsqtRegs; ++k)
+          vec_store_psqt(&accTilePsqt[k], psqt[k]);
+      }
+
+#else
+      std::memcpy(accumulator.accumulation[Perspective], biases,
+          HalfDimensions * sizeof(BiasType));
+
+      for (std::size_t k = 0; k < PSQTBuckets; ++k)
+        accumulator.psqtAccumulation[Perspective][k] = 0;
+
+      for (const auto index : active)
+      {
+        const IndexType offset = HalfDimensions * index;
+
+        for (IndexType j = 0; j < HalfDimensions; ++j)
+          accumulator.accumulation[Perspective][j] += weights[offset + j];
+
+        for (std::size_t k = 0; k < PSQTBuckets; ++k)
+          accumulator.psqtAccumulation[Perspective][k] += psqtWeights[index * PSQTBuckets + k];
+      }
+#endif
+
+  #if defined(USE_MMX)
+      _mm_empty();
+  #endif
+    }
+
+    template<Color Perspective>
+    void hint_common_access_for_perspective(const Position& pos) const {
+
+      // Works like update_accumulator, but performs less work.
+      // Updates ONLY the accumulator for pos.
+
+      // Look for a usable accumulator of an earlier position. We keep track
+      // of the estimated gain in terms of features to be added/subtracted.
+      // Fast early exit.
+      if (pos.state()->accumulator.computed[Perspective])
+        return;
+
+      auto [oldest_st, _] = try_find_computed_accumulator<Perspective>(pos);
+
+      if (oldest_st->accumulator.computed[Perspective])
+      {
+        // Only update current position accumulator to minimize work.
+        StateInfo* states_to_update[2] = { pos.state(), nullptr };
+        update_accumulator_incremetal<Perspective, 2>(pos, oldest_st, states_to_update);
+      }
+      else
+      {
+        update_accumulator_refresh<Perspective>(pos);
+      }
+    }
+
+    template<Color Perspective>
+    void update_accumulator(const Position& pos) const {
+
+      auto [oldest_st, next] = try_find_computed_accumulator<Perspective>(pos);
+
+      if (oldest_st->accumulator.computed[Perspective])
+      {
+        if (next == nullptr)
+          return;
+
+        // Now update the accumulators listed in states_to_update[], where the last element is a sentinel.
+        // Currently we update 2 accumulators.
+        //     1. for the current position
+        //     2. the next accumulator after the computed one
+        // The heuristic may change in the future.
+        StateInfo *states_to_update[3] =
+          { next, next == pos.state() ? nullptr : pos.state(), nullptr };
+
+        update_accumulator_incremetal<Perspective, 3>(pos, oldest_st, states_to_update);
+      }
+      else
+      {
+        update_accumulator_refresh<Perspective>(pos);
+      }
+    }
+
     alignas(CacheLineSize) BiasType biases[HalfDimensions];
     alignas(CacheLineSize) WeightType weights[HalfDimensions * InputDimensions];
     alignas(CacheLineSize) PSQTWeightType psqtWeights[InputDimensions * PSQTBuckets];
diff --git a/src/search.cpp b/src/search.cpp
index 6ca2cfa5..5cb9750c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -730,10 +730,10 @@ namespace {
         goto moves_loop;
     }
     else if (excludedMove) {
-        // excludeMove implies that we had a ttHit on the containing non-excluded search with ss->staticEval filled from TT
-        // However static evals from the TT aren't good enough (-13 elo), presumably due to changing optimism context
-        // Recalculate value with current optimism (without updating thread avgComplexity)
-        ss->staticEval = eval = evaluate(pos, &complexity);
+        // Providing the hint that this node's accumulator will be used often brings significant Elo gain (13 elo)
+        Eval::NNUE::hint_common_parent_position(pos);
+        eval = ss->staticEval;
+        complexity = abs(ss->staticEval - pos.psq_eg_stm());
     }
     else if (ss->ttHit)
     {

From 69639d764bde566e524b8c2566119bf677cb2622 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Tue, 21 Feb 2023 11:17:59 -0500
Subject: [PATCH 074/192] Reintroduce nnue pawn scaling with lower lazy
 thresholds

Params found with the nevergrad TBPSA optimizer via nevergrad4sf modified to:

* use SPRT LLR with fishtest STC elo gainer bounds [0, 2] as the objective function
* increase the game batch size after each new optimal point is found

The params were the optimal point after TBPSA iteration 7 and 160 nevergrad evaluations with:

* initial batch size of 96 games per evaluation
* batch size increase of 64 games after each iteration
* a budget of 512 evaluations
* TC: fixed 1.5 million nodes per move, no time limit

nevergrad4sf enables optimizing stockfish params with TBPSA:
https://github.com/vondele/nevergrad4sf

Using pentanomial game results with smaller game batch sizes was inspired by:

Use of SPRT LLR calculated from pentanomial game results as the objective function was an experiment at maximizing the information from game batches to reduce the computational cost for TBPSA to converge on good parameters.

For the exact code used to find the params:
https://github.com/linrock/tuning-fork

Passed STC:
https://tests.stockfishchess.org/tests/view/63f4ef5ee74a12625bcd114a
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 66552 W: 17736 L: 17390 D: 31426
Ptnml(0-2): 164, 7229, 18166, 7531, 186

Passed LTC:
https://tests.stockfishchess.org/tests/view/63f56028e74a12625bcd2550
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 71264 W: 19150 L: 18787 D: 33327
Ptnml(0-2): 23, 6728, 21771, 7083, 27

closes https://github.com/official-stockfish/Stockfish/pull/4401

bench 3687580
---
 src/evaluate.cpp | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 080d412b..cf6f23ea 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -193,8 +193,8 @@ using namespace Trace;
 namespace {
 
   // Threshold for lazy and space evaluation
-  constexpr Value LazyThreshold1    =  Value(3631);
-  constexpr Value LazyThreshold2    =  Value(2084);
+  constexpr Value LazyThreshold1    =  Value(3622);
+  constexpr Value LazyThreshold2    =  Value(1962);
   constexpr Value SpaceThreshold    =  Value(11551);
 
   // KingAttackWeights[PieceType] contains king attack weights by piece type
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   else
   {
       int nnueComplexity;
-      int scale = 1076 + 96 * pos.non_pawn_material() / 5120;
+      int scale = 1001 + 5 * pos.count<PAWN>() + 61 * pos.non_pawn_material() / 4096;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];

From 29b5ad5deaf323f43019443b322090caec13f847 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Thu, 23 Feb 2023 19:01:51 +0100
Subject: [PATCH 075/192] Fix typo in method name

closes https://github.com/official-stockfish/Stockfish/pull/4404

No functional change
---
 src/nnue/nnue_feature_transformer.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index 13f1604f..b0d5743e 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -365,7 +365,7 @@ namespace Stockfish::Eval::NNUE {
     //       by repeatedly applying ->previous from states_to_update[i+1] or states_to_update[i] == nullptr.
     //       computed_st must be reachable by repeatadly applying ->previous on states_to_update[0], if not nullptr.
     template<Color Perspective, size_t N>
-    void update_accumulator_incremetal(const Position& pos, StateInfo* computed_st, StateInfo* states_to_update[N]) const {
+    void update_accumulator_incremental(const Position& pos, StateInfo* computed_st, StateInfo* states_to_update[N]) const {
       static_assert(N > 0);
       assert(states_to_update[N-1] == nullptr);
 
@@ -630,7 +630,7 @@ namespace Stockfish::Eval::NNUE {
       {
         // Only update current position accumulator to minimize work.
         StateInfo* states_to_update[2] = { pos.state(), nullptr };
-        update_accumulator_incremetal<Perspective, 2>(pos, oldest_st, states_to_update);
+        update_accumulator_incremental<Perspective, 2>(pos, oldest_st, states_to_update);
       }
       else
       {
@@ -656,7 +656,7 @@ namespace Stockfish::Eval::NNUE {
         StateInfo *states_to_update[3] =
           { next, next == pos.state() ? nullptr : pos.state(), nullptr };
 
-        update_accumulator_incremetal<Perspective, 3>(pos, oldest_st, states_to_update);
+        update_accumulator_incremental<Perspective, 3>(pos, oldest_st, states_to_update);
       }
       else
       {

From 472e726bff0d0e496dc8359cc071726a76317a72 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 24 Feb 2023 18:25:24 +0300
Subject: [PATCH 076/192] Search tuning at very long time control

This patch is a result of tuning session of approximately 100k games at 120+1.2.
Biggest changes are in extensions, stat bonus and depth reduction for nodes without a tt move.

Failed STC:
https://tests.stockfishchess.org/tests/view/63f72c72e74a12625bcd7938
LLR: -2.94 (-2.94,2.94) <0.00,2.00>
Total: 13872 W: 3535 L: 3769 D: 6568
Ptnml(0-2): 56, 1621, 3800, 1419, 40

Close to neutral at LTC:
https://tests.stockfishchess.org/tests/view/63f738f5e74a12625bcd7b8a
Elo: 0.80 +-1.2 (95%) LOS: 90.0%
Total: 60000 W: 16213 L: 16074 D: 27713
Ptnml(0-2): 24, 5718, 18379, 5853, 26
nElo: 1.82 +-2.8 (95%) PairsRatio: 1.02

Passed 180+1.8 VLTC:
https://tests.stockfishchess.org/tests/view/63f868f3e74a12625bcdb33e
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 15864 W: 4449 L: 4202 D: 7213
Ptnml(0-2): 1, 1301, 5083, 1544, 3

Passed 60+0.6 8 threads SMP VLTC:
https://tests.stockfishchess.org/tests/view/63f8a5d6e74a12625bcdbdb3
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 6288 W: 1821 L: 1604 D: 2863
Ptnml(0-2): 0, 402, 2123, 619, 0

closes https://github.com/official-stockfish/Stockfish/pull/4406

bench 4705194
---
 src/search.cpp | 102 ++++++++++++++++++++++++-------------------------
 1 file changed, 51 insertions(+), 51 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 5cb9750c..a41ea4fd 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -63,7 +63,7 @@ namespace {
 
   // Futility margin
   Value futility_margin(Depth d, bool improving) {
-    return Value(158 * (d - improving));
+    return Value(154 * (d - improving));
   }
 
   // Reductions lookup table, initialized at startup
@@ -71,7 +71,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1460 - int(delta) * 1024 / int(rootDelta)) / 1024 + (!i && r > 937);
+    return (r + 1449 - int(delta) * 1032 / int(rootDelta)) / 1024 + (!i && r > 941);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -81,7 +81,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min(350 * d - 400, 1650);
+    return std::min(340 * d - 470, 1855);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -161,7 +161,7 @@ namespace {
 void Search::init() {
 
   for (int i = 1; i < MAX_MOVES; ++i)
-      Reductions[i] = int((20.26 + std::log(Threads.size()) / 2) * std::log(i));
+      Reductions[i] = int((19.47 + std::log(Threads.size()) / 2) * std::log(i));
 }
 
 
@@ -354,12 +354,12 @@ void Thread::search() {
           if (rootDepth >= 4)
           {
               Value prev = rootMoves[pvIdx].averageScore;
-              delta = Value(10) + int(prev) * prev / 15400;
+              delta = Value(10) + int(prev) * prev / 16502;
               alpha = std::max(prev - delta,-VALUE_INFINITE);
               beta  = std::min(prev + delta, VALUE_INFINITE);
 
               // Adjust optimism based on root move's previousScore
-              int opt = 116 * prev / (std::abs(prev) + 170);
+              int opt = 120 * prev / (std::abs(prev) + 161);
               optimism[ us] = Value(opt);
               optimism[~us] = -optimism[us];
           }
@@ -462,16 +462,16 @@ void Thread::search() {
           && !Threads.stop
           && !mainThread->stopOnPonderhit)
       {
-          double fallingEval = (71 + 12 * (mainThread->bestPreviousAverageScore - bestValue)
-                                    +  6 * (mainThread->iterValue[iterIdx] - bestValue)) / 656.7;
+          double fallingEval = (69 + 13 * (mainThread->bestPreviousAverageScore - bestValue)
+                                    +  6 * (mainThread->iterValue[iterIdx] - bestValue)) / 619.6;
           fallingEval = std::clamp(fallingEval, 0.5, 1.5);
 
           // If the bestMove is stable over several iterations, reduce time accordingly
-          timeReduction = lastBestMoveDepth + 9 < completedDepth ? 1.37 : 0.65;
-          double reduction = (1.4 + mainThread->previousTimeReduction) / (2.15 * timeReduction);
-          double bestMoveInstability = 1 + 1.7 * totBestMoveChanges / Threads.size();
+          timeReduction = lastBestMoveDepth + 8 < completedDepth ? 1.57 : 0.65;
+          double reduction = (1.4 + mainThread->previousTimeReduction) / (2.08 * timeReduction);
+          double bestMoveInstability = 1 + 1.8 * totBestMoveChanges / Threads.size();
           int complexity = mainThread->complexityAverage.value();
-          double complexPosition = std::min(1.0 + (complexity - 261) / 1738.7, 1.5);
+          double complexPosition = std::min(1.03 + (complexity - 241) / 1552.0, 1.45);
 
           double totalTime = Time.optimum() * fallingEval * reduction * bestMoveInstability * complexPosition;
 
@@ -491,7 +491,7 @@ void Thread::search() {
                   Threads.stop = true;
           }
           else if (   !mainThread->ponder
-                   && Time.elapsed() > totalTime * 0.53)
+                   && Time.elapsed() > totalTime * 0.50)
               Threads.increaseDepth = false;
           else
               Threads.increaseDepth = true;
@@ -760,7 +760,7 @@ namespace {
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
-        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1940, 1940);
+        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1920, 1920);
         thisThread->mainHistory[~us][from_to((ss-1)->currentMove)] << bonus;
     }
 
@@ -770,13 +770,13 @@ namespace {
     // margin and the improving flag are used in various pruning heuristics.
     improvement =   (ss-2)->staticEval != VALUE_NONE ? ss->staticEval - (ss-2)->staticEval
                   : (ss-4)->staticEval != VALUE_NONE ? ss->staticEval - (ss-4)->staticEval
-                  :                                    172;
+                  :                                    156;
     improving = improvement > 0;
 
     // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
-    if (eval < alpha - 394 - 255 * depth * depth)
+    if (eval < alpha - 426 - 252 * depth * depth)
     {
         value = qsearch<NonPV>(pos, ss, alpha - 1, alpha);
         if (value < alpha)
@@ -786,19 +786,19 @@ namespace {
     // Step 8. Futility pruning: child node (~40 Elo).
     // The depth condition is important for mate finding.
     if (   !ss->ttPv
-        &&  depth < 8
-        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 304 >= beta
+        &&  depth < 9
+        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 280 >= beta
         &&  eval >= beta
-        &&  eval < 28580) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
+        &&  eval < 25128) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
         return eval;
 
     // Step 9. Null move search with verification search (~35 Elo)
     if (   !PvNode
         && (ss-1)->currentMove != MOVE_NULL
-        && (ss-1)->statScore < 18200
+        && (ss-1)->statScore < 18755
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 20 * depth - improvement / 14 + 235 + complexity / 24
+        &&  ss->staticEval >= beta - 19 * depth - improvement / 13 + 253 + complexity / 25
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly || us != thisThread->nmpColor))
@@ -806,7 +806,7 @@ namespace {
         assert(eval - beta >= 0);
 
         // Null move dynamic reduction based on depth, eval and complexity of position
-        Depth R = std::min(int(eval - beta) / 165, 6) + depth / 3 + 4 - (complexity > 800);
+        Depth R = std::min(int(eval - beta) / 168, 6) + depth / 3 + 4 - (complexity > 825);
 
         ss->currentMove = MOVE_NULL;
         ss->continuationHistory = &thisThread->continuationHistory[0][0][NO_PIECE][0];
@@ -842,7 +842,7 @@ namespace {
         }
     }
 
-    probCutBeta = beta + 180 - 54 * improving;
+    probCutBeta = beta + 186 - 54 * improving;
 
     // Step 10. ProbCut (~10 Elo)
     // If we have a good enough capture and a reduced search returns a value
@@ -904,14 +904,14 @@ namespace {
         return qsearch<PV>(pos, ss, alpha, beta);
 
     if (    cutNode
-        &&  depth >= 9
+        &&  depth >= 7
         && !ttMove)
         depth -= 2;
 
 moves_loop: // When in check, search starts here
 
     // Step 12. A small Probcut idea, when we are in check (~4 Elo)
-    probCutBeta = beta + 402;
+    probCutBeta = beta + 391;
     if (   ss->inCheck
         && !PvNode
         && depth >= 2
@@ -1006,14 +1006,14 @@ moves_loop: // When in check, search starts here
               // Futility pruning for captures (~2 Elo)
               if (   !givesCheck
                   && !PvNode
-                  && lmrDepth < 7
+                  && lmrDepth < 6
                   && !ss->inCheck
-                  && ss->staticEval + 185 + 203 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
-                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 6 < alpha)
+                  && ss->staticEval + 182 + 230 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
+                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, Value(-220) * depth))
+              if (!pos.see_ge(move, Value(-206) * depth))
                   continue;
           }
           else
@@ -1024,24 +1024,24 @@ moves_loop: // When in check, search starts here
 
               // Continuation history based pruning (~2 Elo)
               if (   lmrDepth < 5
-                  && history < -4180 * (depth - 1))
+                  && history < -4405 * (depth - 1))
                   continue;
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
 
-              lmrDepth += history / 7208;
+              lmrDepth += history / 7278;
               lmrDepth = std::max(lmrDepth, -2);
 
               // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
                   && lmrDepth < 13
-                  && ss->staticEval + 103 + 136 * lmrDepth <= alpha)
+                  && ss->staticEval + 103 + 138 * lmrDepth <= alpha)
                   continue;
 
               lmrDepth = std::max(lmrDepth, 0);
 
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-25 * lmrDepth * lmrDepth - 16 * lmrDepth)))
+              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
                   continue;
           }
       }
@@ -1056,7 +1056,7 @@ moves_loop: // When in check, search starts here
           // a reduced search on all the other moves but the ttMove and if the
           // result is lower than ttValue minus a margin, then we will extend the ttMove.
           if (   !rootNode
-              &&  depth >= 4 - (thisThread->completedDepth > 22) + 2 * (PvNode && tte->is_pv())
+              &&  depth >= 4 - (thisThread->completedDepth > 21) + 2 * (PvNode && tte->is_pv())
               &&  move == ttMove
               && !excludedMove // Avoid recursive singular search
            /* &&  ttValue != VALUE_NONE Already implicit in the next condition */
@@ -1064,7 +1064,7 @@ moves_loop: // When in check, search starts here
               && (tte->bound() & BOUND_LOWER)
               &&  tte->depth() >= depth - 3)
           {
-              Value singularBeta = ttValue - (3 + (ss->ttPv && !PvNode)) * depth;
+              Value singularBeta = ttValue - (2 + (ss->ttPv && !PvNode)) * depth;
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;
@@ -1083,7 +1083,7 @@ moves_loop: // When in check, search starts here
                       && ss->doubleExtensions <= 10)
                   {
                       extension = 2;
-                      depth += depth < 12;
+                      depth += depth < 13;
                   }
               }
 
@@ -1106,15 +1106,15 @@ moves_loop: // When in check, search starts here
 
           // Check extensions (~1 Elo)
           else if (   givesCheck
-                   && depth > 9
-                   && abs(ss->staticEval) > 78)
+                   && depth > 10
+                   && abs(ss->staticEval) > 88)
               extension = 1;
 
           // Quiet ttMove extensions (~1 Elo)
           else if (   PvNode
                    && move == ttMove
                    && move == ss->killers[0]
-                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5600)
+                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5705)
               extension = 1;
       }
 
@@ -1155,7 +1155,7 @@ moves_loop: // When in check, search starts here
 
       // Decrease reduction for PvNodes based on depth
       if (PvNode)
-          r -= 1 + 11 / (3 + depth);
+          r -= 1 + 12 / (3 + depth);
 
       // Decrease reduction if ttMove has been singularly extended (~1 Elo)
       if (singularQuietLMR)
@@ -1172,17 +1172,17 @@ moves_loop: // When in check, search starts here
 
       // Decrease reduction if move is a killer and we have a good history
       if (move == ss->killers[0]
-          && (*contHist[0])[movedPiece][to_sq(move)] >= 3600)
+          && (*contHist[0])[movedPiece][to_sq(move)] >= 3722)
           r--;
 
       ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]
                      + (*contHist[3])[movedPiece][to_sq(move)]
-                     - 4467;
+                     - 4182;
 
       // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
-      r -= ss->statScore / (12800 + 4410 * (depth > 7 && depth < 19));
+      r -= ss->statScore / (11791 + 3992 * (depth > 6 && depth < 19));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has
@@ -1206,8 +1206,8 @@ moves_loop: // When in check, search starts here
           {
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
-              const bool doDeeperSearch = value > (alpha + 66 + 11 * (newDepth - d));
-              const bool doEvenDeeperSearch = value > alpha + 582 && ss->doubleExtensions <= 5;
+              const bool doDeeperSearch = value > (alpha + 58 + 12 * (newDepth - d));
+              const bool doEvenDeeperSearch = value > alpha + 588 && ss->doubleExtensions <= 5;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
               ss->doubleExtensions = ss->doubleExtensions + doEvenDeeperSearch;
@@ -1318,8 +1318,8 @@ moves_loop: // When in check, search starts here
                   // Reduce other moves if we have found at least one score improvement
                   if (   depth > 1
                       && depth < 6
-                      && beta  <  VALUE_KNOWN_WIN
-                      && alpha > -VALUE_KNOWN_WIN)
+                      && beta  <  10534
+                      && alpha > -10534)
                       depth -= 1;
 
                   assert(depth > 0);
@@ -1374,7 +1374,7 @@ moves_loop: // When in check, search starts here
     else if (!priorCapture)
     {
         // Extra bonuses for PV/Cut nodes or bad fail lows
-        int bonus = (depth > 4) + (PvNode || cutNode) + (bestValue < alpha - 88 * depth);
+        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 97 * depth);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 
@@ -1502,7 +1502,7 @@ moves_loop: // When in check, search starts here
         if (PvNode && bestValue > alpha)
             alpha = bestValue;
 
-        futilityBase = bestValue + 158;
+        futilityBase = bestValue + 168;
     }
 
     const PieceToHistory* contHist[] = { (ss-1)->continuationHistory, (ss-2)->continuationHistory,
@@ -1575,7 +1575,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, Value(-108)))
+      if (!pos.see_ge(move, Value(-110)))
           continue;
 
     }
@@ -1708,7 +1708,7 @@ moves_loop: // When in check, search starts here
 
     if (!pos.capture(bestMove))
     {
-        int bonus2 = bestValue > beta + 146 ? bonus1               // larger bonus
+        int bonus2 = bestValue > beta + 153 ? bonus1               // larger bonus
                                             : stat_bonus(depth);   // smaller bonus
 
         // Increase stats for the best move in case it was a quiet move

From 98dafda6c8d432924821d085fd958c89bc8834c3 Mon Sep 17 00:00:00 2001
From: Alfredo Menezes <lomalfredo8@gmail.com>
Date: Mon, 27 Feb 2023 00:39:26 -0300
Subject: [PATCH 077/192] Simplify condition in step 15

Remove 'ttValue <= alpha' check for negative extension in singular search.
Also apply some small code style changes.

STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 127888 W: 33766 L: 33651 D: 60471
Ptnml(0-2): 303, 14082, 35089, 14137, 333
https://tests.stockfishchess.org/tests/view/63f79528e74a12625bcd8c05

LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 89048 W: 23924 L: 23782 D: 41342
Ptnml(0-2): 27, 8635, 27065, 8763, 34
https://tests.stockfishchess.org/tests/view/63f82177e74a12625bcda6f4

LTC (retest):
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 196360 W: 52514 L: 52475 D: 91371
Ptnml(0-2): 103, 19066, 59780, 19151, 80
https://tests.stockfishchess.org/tests/view/63f934bfe74a12625bcdd929

closes https://github.com/official-stockfish/Stockfish/pull/4407

Bench: 5310866
---
 src/search.cpp | 48 ++++++++++++++++++++++++++----------------------
 1 file changed, 26 insertions(+), 22 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index a41ea4fd..6ccc70cc 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -334,7 +334,7 @@ void Thread::search() {
       pvLast = 0;
 
       if (!Threads.increaseDepth)
-         searchAgainCounter++;
+          searchAgainCounter++;
 
       // MultiPV loop. We perform a full root search for each PV line
       for (pvIdx = 0; pvIdx < multiPV && !Threads.stop; ++pvIdx)
@@ -432,9 +432,10 @@ void Thread::search() {
       if (!Threads.stop)
           completedDepth = rootDepth;
 
-      if (rootMoves[0].pv[0] != lastBestMove) {
-         lastBestMove = rootMoves[0].pv[0];
-         lastBestMoveDepth = rootDepth;
+      if (rootMoves[0].pv[0] != lastBestMove)
+      {
+          lastBestMove = rootMoves[0].pv[0];
+          lastBestMoveDepth = rootDepth;
       }
 
       // Have we found a "mate in x"?
@@ -729,7 +730,8 @@ namespace {
         complexity = 0;
         goto moves_loop;
     }
-    else if (excludedMove) {
+    else if (excludedMove)
+    {
         // Providing the hint that this node's accumulator will be used often brings significant Elo gain (13 elo)
         Eval::NNUE::hint_common_parent_position(pos);
         eval = ss->staticEval;
@@ -755,6 +757,7 @@ namespace {
         // Save static evaluation into transposition table
         tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
     }
+
     thisThread->complexityAverage.update(complexity);
 
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
@@ -920,11 +923,9 @@ moves_loop: // When in check, search starts here
         && tte->depth() >= depth - 3
         && ttValue >= probCutBeta
         && abs(ttValue) <= VALUE_KNOWN_WIN
-        && abs(beta) <= VALUE_KNOWN_WIN
-       )
+        && abs(beta) <= VALUE_KNOWN_WIN)
         return probCutBeta;
 
-
     const PieceToHistory* contHist[] = { (ss-1)->continuationHistory, (ss-2)->continuationHistory,
                                           nullptr                   , (ss-4)->continuationHistory,
                                           nullptr                   , (ss-6)->continuationHistory };
@@ -1099,8 +1100,8 @@ moves_loop: // When in check, search starts here
               else if (ttValue >= beta)
                   extension = -2;
 
-              // If the eval of ttMove is less than alpha and value, we reduce it (negative extension)
-              else if (ttValue <= alpha && ttValue <= value)
+              // If the eval of ttMove is less than value, we reduce it (negative extension)
+              else if (ttValue <= value)
                   extension = -1;
           }
 
@@ -1227,11 +1228,11 @@ moves_loop: // When in check, search starts here
       // Step 18. Full depth search when LMR is skipped. If expected reduction is high, reduce its depth by 1.
       else if (!PvNode || moveCount > 1)
       {
-               // Increase reduction for cut nodes and not ttMove (~1 Elo)
-               if (!ttMove && cutNode)
-                         r += 2;
+          // Increase reduction for cut nodes and not ttMove (~1 Elo)
+          if (!ttMove && cutNode)
+              r += 2;
 
-               value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 4), !cutNode);
+          value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 4), !cutNode);
       }
 
       // For PV nodes only, do a full PV search on the first move or after a fail
@@ -1271,14 +1272,17 @@ moves_loop: // When in check, search starts here
               rm.selDepth = thisThread->selDepth;
               rm.scoreLowerbound = rm.scoreUpperbound = false;
 
-              if (value >= beta) {
-                 rm.scoreLowerbound = true;
-                 rm.uciScore = beta;
+              if (value >= beta)
+              {
+                  rm.scoreLowerbound = true;
+                  rm.uciScore = beta;
               }
-              else if (value <= alpha) {
-                 rm.scoreUpperbound = true;
-                 rm.uciScore = alpha;
+              else if (value <= alpha)
+              {
+                  rm.scoreUpperbound = true;
+                  rm.uciScore = alpha;
               }
+
               rm.pv.resize(1);
 
               assert((ss+1)->pv);
@@ -1447,7 +1451,8 @@ moves_loop: // When in check, search starts here
     // TT entry depth that we are going to use. Note that in qsearch we use
     // only two types of depth in TT: DEPTH_QS_CHECKS or DEPTH_QS_NO_CHECKS.
     ttDepth = ss->inCheck || depth >= DEPTH_QS_CHECKS ? DEPTH_QS_CHECKS
-                                                  : DEPTH_QS_NO_CHECKS;
+                                                      : DEPTH_QS_NO_CHECKS;
+
     // Step 3. Transposition table lookup
     posKey = pos.key();
     tte = TT.probe(posKey, ss->ttHit);
@@ -1577,7 +1582,6 @@ moves_loop: // When in check, search starts here
       // Do not search moves with bad enough SEE values (~5 Elo)
       if (!pos.see_ge(move, Value(-110)))
           continue;
-
     }
 
       // Speculative prefetch as early as possible

From 728b963614a765f5cb64c44a078169cca977750f Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@phxl0401.wp.lan>
Date: Sun, 26 Feb 2023 09:59:35 +0100
Subject: [PATCH 078/192] Use common_parent_position hint also at PVNodes TT
 hits.

Credits to Stefan Geschwentner (locutus2) showing that the hint
is useful on PvNodes. In contrast to his test,
this version avoids to use the hint when in check.
I believe checking positions aren't good candidates for the hint
because:
- evasion moves are rather few, so a checking pos. has much less childs
than a normal position
- if the king has to move the NNUE eval can't use incremental updates,
  so the child nodes have to do a full refresh anyway.

Passed STC:
https://tests.stockfishchess.org/tests/view/63f9c5b1e74a12625bcdf585
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 124472 W: 33268 L: 32846 D: 58358
Ptnml(0-2): 350, 12986, 35170, 13352, 378

closes https://github.com/official-stockfish/Stockfish/pull/4410

no functional change
---
 src/search.cpp | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index 6ccc70cc..206779ed 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -744,7 +744,11 @@ namespace {
         if (eval == VALUE_NONE)
             ss->staticEval = eval = evaluate(pos, &complexity);
         else // Fall back to (semi)classical complexity for TT hits, the NNUE complexity is lost
+        {
             complexity = abs(ss->staticEval - pos.psq_eg_stm());
+            if (PvNode)
+               Eval::NNUE::hint_common_parent_position(pos);
+        }
 
         // ttValue can be used as a better position evaluation (~7 Elo)
         if (    ttValue != VALUE_NONE

From ff5a6f8df196d61a0d9b1ebe54d84eeb9af20079 Mon Sep 17 00:00:00 2001
From: Stefan Geschwentner <stgeschwentner@gmail.com>
Date: Sun, 26 Feb 2023 09:56:54 +0100
Subject: [PATCH 079/192] NNUE accumulator update in probcut.

Call the recently added hint function for NNUE accumulator update after a failed probcut search.
In this case we already searched at least some captures and tt move which, however, is not sufficient for a cutoff.
So it seems we have a greater chance that the full search will also have no cutoff and hence all moves have to be searched.

STC: https://tests.stockfishchess.org/tests/view/63fa74a4e74a12625bce1823
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 70096 W: 18770 L: 18423 D: 32903
Ptnml(0-2): 191, 7342, 19654, 7651, 210

To be sure that we have no heavy interaction retest on top of #4410.

Rebased STC: https://tests.stockfishchess.org/tests/view/63fb2f62e74a12625bce3b03
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 137688 W: 36790 L: 36349 D: 64549
Ptnml(0-2): 397, 14373, 38919, 14702, 453

closes https://github.com/official-stockfish/Stockfish/pull/4411

No functional change
---
 src/search.cpp | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index 206779ed..cfb569b9 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -899,6 +899,8 @@ namespace {
                     return value;
                 }
             }
+
+        Eval::NNUE::hint_common_parent_position(pos);
     }
 
     // Step 11. If the position is not in TT, decrease depth by 3.

From 564456a6a824bfca26d6d9af5b35a055eb9fc6c2 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sun, 26 Feb 2023 19:42:31 +0100
Subject: [PATCH 080/192] Unify type alias declarations

The commit unifies the declaration of type aliases by replacing all
typedefs with corresponding using statements.

closing https://github.com/official-stockfish/Stockfish/pull/4412

No functional change
---
 src/material.h                      |  2 +-
 src/misc.cpp                        | 12 ++++++------
 src/misc.h                          |  4 ++--
 src/movepick.h                      | 14 +++++++-------
 src/nnue/nnue_feature_transformer.h | 20 ++++++++++----------
 src/pawns.h                         |  2 +-
 src/position.h                      |  2 +-
 src/search.h                        |  2 +-
 src/syzygy/tbprobe.cpp              |  4 ++--
 src/tune.h                          |  6 +++---
 src/types.h                         |  6 +++---
 src/uci.h                           |  4 ++--
 12 files changed, 39 insertions(+), 39 deletions(-)

diff --git a/src/material.h b/src/material.h
index 73c83100..9acf78f5 100644
--- a/src/material.h
+++ b/src/material.h
@@ -62,7 +62,7 @@ struct Entry {
   uint8_t factor[COLOR_NB];
 };
 
-typedef HashTable<Entry, 8192> Table;
+using Table = HashTable<Entry, 8192>;
 
 Entry* probe(const Position& pos);
 
diff --git a/src/misc.cpp b/src/misc.cpp
index e65faab9..c22126af 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -32,12 +32,12 @@
 // 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);
-typedef bool(*fun4_t)(USHORT, PGROUP_AFFINITY, USHORT, PUSHORT);
-typedef WORD(*fun5_t)();
+using fun1_t = bool(*)(LOGICAL_PROCESSOR_RELATIONSHIP,
+                       PSYSTEM_LOGICAL_PROCESSOR_INFORMATION_EX, PDWORD);
+using fun2_t = bool(*)(USHORT, PGROUP_AFFINITY);
+using fun3_t = bool(*)(HANDLE, CONST GROUP_AFFINITY*, PGROUP_AFFINITY);
+using fun4_t = bool(*)(USHORT, PGROUP_AFFINITY, USHORT, PUSHORT);
+using fun5_t = WORD(*)();
 }
 #endif
 
diff --git a/src/misc.h b/src/misc.h
index 9761da8a..c20a816e 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -45,7 +45,7 @@ void dbg_stdev_of(int64_t value, int slot = 0);
 void dbg_correl_of(int64_t value1, int64_t value2, int slot = 0);
 void dbg_print();
 
-typedef std::chrono::milliseconds::rep TimePoint; // A value in milliseconds
+using TimePoint = std::chrono::milliseconds::rep; // A value in milliseconds
 static_assert(sizeof(TimePoint) == sizeof(int64_t), "TimePoint should be 64 bits");
 inline TimePoint now() {
   return std::chrono::duration_cast<std::chrono::milliseconds>
@@ -165,7 +165,7 @@ public:
 
 inline uint64_t mul_hi64(uint64_t a, uint64_t b) {
 #if defined(__GNUC__) && defined(IS_64BIT)
-    __extension__ typedef unsigned __int128 uint128;
+    __extension__ using uint128 = unsigned __int128;
     return ((uint128)a * (uint128)b) >> 64;
 #else
     uint64_t aL = (uint32_t)a, aH = a >> 32;
diff --git a/src/movepick.h b/src/movepick.h
index 90f60b8a..b6c07378 100644
--- a/src/movepick.h
+++ b/src/movepick.h
@@ -62,14 +62,14 @@ public:
 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;
+  using stats = Stats<T, D, Size, Sizes...>;
 
   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;
+    using entry = StatsEntry<T, D>;
     entry* p = reinterpret_cast<entry*>(this);
     std::fill(p, p + sizeof(*this) / sizeof(entry), v);
   }
@@ -87,23 +87,23 @@ enum StatsType { NoCaptures, Captures };
 /// ordering decisions. It uses 2 tables (one for each color) indexed by
 /// the move's from and to squares, see www.chessprogramming.org/Butterfly_Boards
 /// (~11 elo)
-typedef Stats<int16_t, 7183, COLOR_NB, int(SQUARE_NB) * int(SQUARE_NB)> ButterflyHistory;
+using ButterflyHistory = Stats<int16_t, 7183, COLOR_NB, int(SQUARE_NB) * int(SQUARE_NB)>;
 
 /// CounterMoveHistory stores counter moves indexed by [piece][to] of the previous
 /// move, see www.chessprogramming.org/Countermove_Heuristic
-typedef Stats<Move, NOT_USED, PIECE_NB, SQUARE_NB> CounterMoveHistory;
+using CounterMoveHistory = Stats<Move, NOT_USED, PIECE_NB, SQUARE_NB>;
 
 /// 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;
+using CapturePieceToHistory = Stats<int16_t, 10692, PIECE_NB, SQUARE_NB, PIECE_TYPE_NB>;
 
 /// PieceToHistory is like ButterflyHistory but is addressed by a move's [piece][to]
-typedef Stats<int16_t, 29952, PIECE_NB, SQUARE_NB> PieceToHistory;
+using PieceToHistory = Stats<int16_t, 29952, PIECE_NB, SQUARE_NB>;
 
 /// 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.
 /// (~63 elo)
-typedef Stats<PieceToHistory, NOT_USED, PIECE_NB, SQUARE_NB> ContinuationHistory;
+using ContinuationHistory = Stats<PieceToHistory, NOT_USED, PIECE_NB, SQUARE_NB>;
 
 
 /// MovePicker class is used to pick one pseudo-legal move at a time from the
diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index b0d5743e..8087ea55 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -42,8 +42,8 @@ namespace Stockfish::Eval::NNUE {
     "Per feature PSQT values cannot be processed at granularity lower than 8 at a time.");
 
   #ifdef USE_AVX512
-  typedef __m512i vec_t;
-  typedef __m256i psqt_vec_t;
+  using vec_t = __m512i;
+  using psqt_vec_t = __m256i;
   #define vec_load(a) _mm512_load_si512(a)
   #define vec_store(a,b) _mm512_store_si512(a,b)
   #define vec_add_16(a,b) _mm512_add_epi16(a,b)
@@ -66,8 +66,8 @@ namespace Stockfish::Eval::NNUE {
   #define MaxChunkSize 64
 
   #elif USE_AVX2
-  typedef __m256i vec_t;
-  typedef __m256i psqt_vec_t;
+  using vec_t = __m256i;
+  using psqt_vec_t = __m256i;
   #define vec_load(a) _mm256_load_si256(a)
   #define vec_store(a,b) _mm256_store_si256(a,b)
   #define vec_add_16(a,b) _mm256_add_epi16(a,b)
@@ -90,8 +90,8 @@ namespace Stockfish::Eval::NNUE {
   #define MaxChunkSize 32
 
   #elif USE_SSE2
-  typedef __m128i vec_t;
-  typedef __m128i psqt_vec_t;
+  using vec_t = __m128i;
+  using psqt_vec_t = __m128i;
   #define vec_load(a) (*(a))
   #define vec_store(a,b) *(a)=(b)
   #define vec_add_16(a,b) _mm_add_epi16(a,b)
@@ -111,8 +111,8 @@ namespace Stockfish::Eval::NNUE {
   #define MaxChunkSize 16
 
   #elif USE_MMX
-  typedef __m64 vec_t;
-  typedef __m64 psqt_vec_t;
+  using vec_t = __m64;
+  using psqt_vec_t = __m64;
   #define vec_load(a) (*(a))
   #define vec_store(a,b) *(a)=(b)
   #define vec_add_16(a,b) _mm_add_pi16(a,b)
@@ -139,8 +139,8 @@ namespace Stockfish::Eval::NNUE {
   #define MaxChunkSize 8
 
   #elif USE_NEON
-  typedef int16x8_t vec_t;
-  typedef int32x4_t psqt_vec_t;
+  using vec_t = int16x8_t;
+  using psqt_vec_t = int32x4_t;
   #define vec_load(a) (*(a))
   #define vec_store(a,b) *(a)=(b)
   #define vec_add_16(a,b) vaddq_s16(a,b)
diff --git a/src/pawns.h b/src/pawns.h
index 95c9ea3c..d20e7c2e 100644
--- a/src/pawns.h
+++ b/src/pawns.h
@@ -61,7 +61,7 @@ struct Entry {
   int blockedCount;
 };
 
-typedef HashTable<Entry, 131072> Table;
+using Table = HashTable<Entry, 131072>;
 
 Entry* probe(const Position& pos);
 
diff --git a/src/position.h b/src/position.h
index 3de24f22..c82c7a8b 100644
--- a/src/position.h
+++ b/src/position.h
@@ -68,7 +68,7 @@ struct StateInfo {
 /// start position to the position just before the search starts). Needed by
 /// 'draw by repetition' detection. Use a std::deque because pointers to
 /// elements are not invalidated upon list resizing.
-typedef std::unique_ptr<std::deque<StateInfo>> StateListPtr;
+using StateListPtr = std::unique_ptr<std::deque<StateInfo>>;
 
 
 /// Position class stores information regarding the board representation as
diff --git a/src/search.h b/src/search.h
index 48a0f7ce..806e4be6 100644
--- a/src/search.h
+++ b/src/search.h
@@ -80,7 +80,7 @@ struct RootMove {
   std::vector<Move> pv;
 };
 
-typedef std::vector<RootMove> RootMoves;
+using RootMoves = std::vector<RootMove>;
 
 
 /// LimitsType struct stores information sent by GUI about available time to
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index bbfd819d..b594ac37 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -142,7 +142,7 @@ struct SparseEntry {
 
 static_assert(sizeof(SparseEntry) == 6, "SparseEntry must be 6 bytes");
 
-typedef uint16_t Sym; // Huffman symbol
+using Sym = uint16_t; // Huffman symbol
 
 struct LR {
     enum Side { Left, Right };
@@ -327,7 +327,7 @@ struct PairsData {
 // first access, when the corresponding file is memory mapped.
 template<TBType Type>
 struct TBTable {
-    typedef typename std::conditional<Type == WDL, WDLScore, int>::type Ret;
+    using Ret = typename std::conditional<Type == WDL, WDLScore, int>::type;
 
     static constexpr int Sides = Type == WDL ? 2 : 1;
 
diff --git a/src/tune.h b/src/tune.h
index f5b787af..440d950a 100644
--- a/src/tune.h
+++ b/src/tune.h
@@ -26,8 +26,8 @@
 
 namespace Stockfish {
 
-typedef std::pair<int, int> Range; // Option's min-max values
-typedef Range (RangeFun) (int);
+using Range = std::pair<int, int>; // Option's min-max values
+using RangeFun = Range (int);
 
 // Default Range function, to calculate Option's min-max values
 inline Range default_range(int v) {
@@ -75,7 +75,7 @@ struct SetRange {
 
 class Tune {
 
-  typedef void (PostUpdate) (); // Post-update function
+  using PostUpdate = void (); // Post-update function
 
   Tune() { read_results(); }
   Tune(const Tune&) = delete;
diff --git a/src/types.h b/src/types.h
index 8a021342..37ce343a 100644
--- a/src/types.h
+++ b/src/types.h
@@ -103,8 +103,8 @@ constexpr bool Is64Bit = true;
 constexpr bool Is64Bit = false;
 #endif
 
-typedef uint64_t Key;
-typedef uint64_t Bitboard;
+using Key = uint64_t;
+using Bitboard = uint64_t;
 
 constexpr int MAX_MOVES = 256;
 constexpr int MAX_PLY   = 246;
@@ -218,7 +218,7 @@ constexpr Value PieceValue[PHASE_NB][PIECE_NB] = {
     VALUE_ZERO, PawnValueEg, KnightValueEg, BishopValueEg, RookValueEg, QueenValueEg, VALUE_ZERO, VALUE_ZERO }
 };
 
-typedef int Depth;
+using Depth = int;
 
 enum : int {
   DEPTH_QS_CHECKS     =  0,
diff --git a/src/uci.h b/src/uci.h
index 5d8ccf1a..70e45acc 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -45,12 +45,12 @@ struct CaseInsensitiveLess {
 };
 
 /// The options container is defined as a std::map
-typedef std::map<std::string, Option, CaseInsensitiveLess> OptionsMap;
+using OptionsMap = std::map<std::string, Option, CaseInsensitiveLess>;
 
 /// The Option class implements each option as specified by the UCI protocol
 class Option {
 
-  typedef void (*OnChange)(const Option&);
+  using OnChange = void (*)(const Option&);
 
 public:
   Option(OnChange = nullptr);

From 6adbc6fa05bbd2a574a8c4c6a6ef2307280217ab Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Wed, 22 Feb 2023 05:45:43 -0600
Subject: [PATCH 081/192] Late counter bonus: boost underestimated moves

The idea here is very intuitive: since we've just proven that the move is good, then if it previously had poor stats, boost those stats more than otherwise.

Passed STC: https://tests.stockfishchess.org/tests/view/63fb504ce74a12625bce4154
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 21128 W: 5763 L: 5481 D: 9884
Ptnml(0-2): 52, 2212, 5759, 2484, 57

Passed LTC: https://tests.stockfishchess.org/tests/view/63fb7825e74a12625bce491b
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 91904 W: 24764 L: 24359 D: 42781
Ptnml(0-2): 45, 8735, 27984, 9146, 42

closes https://github.com/official-stockfish/Stockfish/pull/4415

bench 4318808
---
 src/search.cpp | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index cfb569b9..6585f7bc 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1383,8 +1383,7 @@ moves_loop: // When in check, search starts here
     // Bonus for prior countermove that caused the fail low
     else if (!priorCapture)
     {
-        // Extra bonuses for PV/Cut nodes or bad fail lows
-        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 97 * depth);
+        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 97 * depth) + ((ss-1)->moveCount > 10);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 

From 876906965b8d552866486c0e6eda1184fdb1d636 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Fri, 24 Feb 2023 09:16:55 -0500
Subject: [PATCH 082/192] Update default net to nn-52471d67216a.nnue

Created by retraining the master net with modifications to the previous best dataset:

* Improving T80 oct+nov 2022 endgame lambda accuracy by rescoring with 12-16tb of syzygy 7p tablebases
* Filtering T78 jun+jul+aug 2022 with d6pv2 search to remove positions with bestmove captures or one good move
* Adding T80 sep 2022 data, rescored with 16tb of 7p tablebases, unfiltered

Trained with max-epoch 900, end-lambda 0.7, and early-fen-skipping 28.

```
python3 easy_train.py \
  --experiment-name leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p-lambda7-sk28 \
  --training-dataset /data/leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/easy-train-early-fen-skipping \
  --early-fen-skipping 28 \
  --start-from-engine-test-net True \
  --gpus "0," \
  --max_epoch 900 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --gamma 0.995 \
  --lr 4.375e-4 \
  --tui False \
  --seed $RANDOM
```

Training data was rescored and d6pv2 filtered in the same way as recent best datasets.
For preparing the merged training dataset:

```
python3 interleave_binpacks.py \
  leela96-eval-filt-v2.binpack \
  dfrc99-eval-filt-v2.binpack \
  test80-oct2022-16tb7p-eval-filt-v2-d6.binpack \
  test80-nov2022-12tb7p-eval-filt-v2-d6.binpack \
  T79-apr2022-12tb7p-eval-filt-v2.binpack \
  T79-may2022-12tb7p-eval-filt-v2.binpack \
  test78-junjulaug2022-16tb7p-eval-filt-v2-d6.binpack \
  T60-nov2021-12tb7p-eval-filt-v2.binpack \
  T60-dec2021-12tb7p-eval-filt-v2.binpack \
  T78-sep2022-12tb7p.binpack \
  test77-dec2021-16gb7p.binpack \
  test80-sep2022-16tb7p.binpack \
  /data/leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p.binpack
```

Links for downloading the training data components can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch839.nnue : 0.6 +/- 1.4

Passed STC:
https://tests.stockfishchess.org/tests/view/63f9ab4be74a12625bcdf02e
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 84656 W: 22681 L: 22302 D: 39673
Ptnml(0-2): 271, 9343, 22734, 9696, 284

Passed LTC:
https://tests.stockfishchess.org/tests/view/63fa3833e74a12625bce0c0e
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 184664 W: 49933 L: 49344 D: 85387
Ptnml(0-2): 111, 17977, 55561, 18578, 105

closes https://github.com/official-stockfish/Stockfish/pull/4416

bench: 4814343
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 46f20259..5238cb81 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-1337b1adec5b.nnue"
+  #define EvalFileDefaultName   "nn-52471d67216a.nnue"
 
   namespace NNUE {
 

From 5c75c1c2fbb7bb4f0bf7c44fb855c415b788cbf7 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 24 Feb 2023 12:09:45 +0300
Subject: [PATCH 083/192] Fix duplicated moves generation in movepicker

in a some of cases movepicker returned some moves more than once which lead
to them being searched more than once. This bug was possible because of how
we use queen promotions - they are generated as a captures but are not
included in position function which checks if move is a capture. Thus if
any refutation (killer or countermove) was a queen promotion it was
searched twice - once as a capture and one as a refutation.

This patch affects various things, namely stats assignments for queen promotions
and other moves if best move is queen promotion,
also some heuristics in search and qsearch.

With this patch every queen promotion is now considered a capture.

After this patch number of found duplicated moves is 0 during normal 13 depth bench run.

Passed STC:
https://tests.stockfishchess.org/tests/view/63f77e01e74a12625bcd87d7
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 80920 W: 21455 L: 21289 D: 38176
Ptnml(0-2): 198, 8839, 22241, 8963, 219

Passed LTC:
https://tests.stockfishchess.org/tests/view/63f7e020e74a12625bcd9a76
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 89712 W: 23674 L: 23533 D: 42505
Ptnml(0-2): 24, 8737, 27202, 8860, 33

closes https://github.com/official-stockfish/Stockfish/pull/4405

bench 4681731
---
 src/movepick.cpp       |  6 +++---
 src/position.h         | 12 ++++++++----
 src/search.cpp         | 12 ++++++------
 src/syzygy/tbprobe.cpp |  4 ++--
 4 files changed, 19 insertions(+), 15 deletions(-)

diff --git a/src/movepick.cpp b/src/movepick.cpp
index 65155a73..36ee46b5 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -93,7 +93,7 @@ MovePicker::MovePicker(const Position& p, Move ttm, Value th, const CapturePiece
 {
   assert(!pos.checkers());
 
-  stage = PROBCUT_TT + !(ttm && pos.capture(ttm)
+  stage = PROBCUT_TT + !(ttm && pos.capture_stage(ttm)
                              && pos.pseudo_legal(ttm)
                              && pos.see_ge(ttm, threshold));
 }
@@ -141,7 +141,7 @@ void MovePicker::score() {
                    +     bool(pos.check_squares(type_of(pos.moved_piece(m))) & to_sq(m)) * 16384;
       else // Type == EVASIONS
       {
-          if (pos.capture(m))
+          if (pos.capture_stage(m))
               m.value =  PieceValue[MG][pos.piece_on(to_sq(m))]
                        - Value(type_of(pos.moved_piece(m)))
                        + (1 << 28);
@@ -216,7 +216,7 @@ top:
 
   case REFUTATION:
       if (select<Next>([&](){ return    *cur != MOVE_NONE
-                                    && !pos.capture(*cur)
+                                    && !pos.capture_stage(*cur)
                                     &&  pos.pseudo_legal(*cur); }))
           return *(cur - 1);
       ++stage;
diff --git a/src/position.h b/src/position.h
index c82c7a8b..485540ef 100644
--- a/src/position.h
+++ b/src/position.h
@@ -125,7 +125,7 @@ public:
   // Properties of moves
   bool legal(Move m) const;
   bool pseudo_legal(const Move m) const;
-  bool capture(Move m) const;
+  bool capture_stage(Move m) const;
   bool gives_check(Move m) const;
   Piece moved_piece(Move m) const;
   Piece captured_piece() const;
@@ -381,10 +381,14 @@ inline bool Position::is_chess960() const {
   return chess960;
 }
 
-inline bool Position::capture(Move m) const {
+// returns true if a move is generated from the capture stage
+// having also queen promotions covered, i.e. consistency with the capture stage move generation
+// is needed to avoid the generation of duplicate moves.
+inline bool Position::capture_stage(Move m) const {
   assert(is_ok(m));
-  // Castling is encoded as "king captures rook"
-  return (!empty(to_sq(m)) && type_of(m) != CASTLING) || type_of(m) == EN_PASSANT;
+  return     (!empty(to_sq(m)) && type_of(m) != CASTLING)
+          || (type_of(m) == PROMOTION && promotion_type(m) == QUEEN)
+          ||  type_of(m) == EN_PASSANT;
 }
 
 inline Piece Position::captured_piece() const {
diff --git a/src/search.cpp b/src/search.cpp
index 6585f7bc..fcdb8d67 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -623,7 +623,7 @@ namespace {
     ttValue = ss->ttHit ? value_from_tt(tte->value(), ss->ply, pos.rule50_count()) : VALUE_NONE;
     ttMove =  rootNode ? thisThread->rootMoves[thisThread->pvIdx].pv[0]
             : ss->ttHit    ? tte->move() : MOVE_NONE;
-    ttCapture = ttMove && pos.capture(ttMove);
+    ttCapture = ttMove && pos.capture_stage(ttMove);
 
     // At this point, if excluded, skip straight to step 6, static eval. However,
     // to save indentation, we list the condition in all code between here and there.
@@ -852,7 +852,7 @@ namespace {
     probCutBeta = beta + 186 - 54 * improving;
 
     // Step 10. ProbCut (~10 Elo)
-    // If we have a good enough capture and a reduced search returns a value
+    // If we have a good enough capture (or queen promotion) and a reduced search returns a value
     // much above beta, we can (almost) safely prune the previous move.
     if (   !PvNode
         &&  depth > 4
@@ -873,7 +873,7 @@ namespace {
         while ((move = mp.next_move()) != MOVE_NONE)
             if (move != excludedMove && pos.legal(move))
             {
-                assert(pos.capture(move) || promotion_type(move) == QUEEN);
+                assert(pos.capture_stage(move));
 
                 ss->currentMove = move;
                 ss->continuationHistory = &thisThread->continuationHistory[ss->inCheck]
@@ -985,7 +985,7 @@ moves_loop: // When in check, search starts here
           (ss+1)->pv = nullptr;
 
       extension = 0;
-      capture = pos.capture(move);
+      capture = pos.capture_stage(move);
       movedPiece = pos.moved_piece(move);
       givesCheck = pos.gives_check(move);
 
@@ -1542,7 +1542,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       givesCheck = pos.gives_check(move);
-      capture = pos.capture(move);
+      capture = pos.capture_stage(move);
 
       moveCount++;
 
@@ -1715,7 +1715,7 @@ moves_loop: // When in check, search starts here
     PieceType captured = type_of(pos.piece_on(to_sq(bestMove)));
     int bonus1 = stat_bonus(depth + 1);
 
-    if (!pos.capture(bestMove))
+    if (!pos.capture_stage(bestMove))
     {
         int bonus2 = bestValue > beta + 153 ? bonus1               // larger bonus
                                             : stat_bonus(depth);   // smaller bonus
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index b594ac37..2a9e1b68 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -1203,7 +1203,7 @@ WDLScore search(Position& pos, ProbeState* result) {
 
     for (const Move move : moveList)
     {
-        if (   !pos.capture(move)
+        if (   !pos.capture_stage(move)
             && (!CheckZeroingMoves || type_of(pos.moved_piece(move)) != PAWN))
             continue;
 
@@ -1472,7 +1472,7 @@ int Tablebases::probe_dtz(Position& pos, ProbeState* result) {
 
     for (const Move move : MoveList<LEGAL>(pos))
     {
-        bool zeroing = pos.capture(move) || type_of(pos.moved_piece(move)) == PAWN;
+        bool zeroing = pos.capture_stage(move) || type_of(pos.moved_piece(move)) == PAWN;
 
         pos.do_move(move, st);
 

From 5c589142ae5714ecb9ded29bec12c591a1ba8f3f Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Sat, 4 Mar 2023 16:34:34 +0100
Subject: [PATCH 084/192] Add wiki to artifacts

snapshot the wiki https://github.com/official-stockfish/stockfish/wiki as part of the artifacts generated.
This will allow future release to include the wiki pages as a form of documentation

closes https://github.com/official-stockfish/Stockfish/pull/4420

No functional change
---
 .github/workflows/stockfish_arm_binaries.yml | 7 +++++++
 .github/workflows/stockfish_binaries.yml     | 7 +++++++
 2 files changed, 14 insertions(+)

diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
index a1b3cdab..e088c441 100644
--- a/.github/workflows/stockfish_arm_binaries.yml
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -99,10 +99,17 @@ jobs:
       - name: Remove non src files
         run: rm -f *.o .depend *.nnue
 
+      - name: Download wiki
+        run: |
+          git clone https://github.com/official-stockfish/Stockfish.wiki.git ../wiki
+          cd ../wiki
+          rm -rf .git
+
       - name: Create tar archive.
         run: |
           cd ..
           mkdir stockfish
+          cp -r wiki stockfish/
           cp -r src stockfish/
           cp stockfish-android-$BINARY$EXT stockfish/
           cp "Top CPU Contributors.txt" stockfish/
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 06b13a9e..b22897cf 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -83,10 +83,17 @@ jobs:
       - name: Remove non src files
         run: rm -f *.o .depend *.nnue
 
+      - name: Download wiki
+        run: |
+          git clone https://github.com/official-stockfish/Stockfish.wiki.git ../wiki
+          cd ../wiki
+          rm -rf .git
+
       - name: Create tar archive.
         run: |
           cd ..
           mkdir stockfish
+          cp -r wiki stockfish/
           cp -r src stockfish/
           cp stockfish-$OS-$BINARY$EXT stockfish/
           cp "Top CPU Contributors.txt" stockfish/

From 3a634f5282700484445c0f66b21b35e96fcb947b Mon Sep 17 00:00:00 2001
From: dav1312 <63931154+dav1312@users.noreply.github.com>
Date: Sat, 4 Mar 2023 13:24:35 +0100
Subject: [PATCH 085/192] Update README.md

Update and simplify the readme, removing duplicated and outdated stuff and pointing to the new wiki

closes https://github.com/official-stockfish/Stockfish/pull/4421

No functional change
---
 README.md | 376 ++++++++++++------------------------------------------
 1 file changed, 81 insertions(+), 295 deletions(-)

diff --git a/README.md b/README.md
index ca90d5d4..1f462d31 100644
--- a/README.md
+++ b/README.md
@@ -2,6 +2,21 @@
 
   [![Stockfish][stockfish128-logo]][website-link]
 
+  <h3>Stockfish</h3>
+
+  A free and strong UCI chess engine.
+  <br>
+  <strong>[Explore Stockfish docs »][wiki-link]</strong>
+  <br>
+  <br>
+  [Report bug][issue-link]
+  ·
+  [Open a discussion][discussions-link]
+  ·
+  [Discord][discord-link]
+  ·
+  [Blog][website-blog-link]
+
   [![Build][build-badge]][build-link]
   [![License][license-badge]][license-link]
   <br>
@@ -16,19 +31,16 @@
 
 ## Overview
 
-[Stockfish][website-link] is a free, powerful UCI chess engine derived from
-Glaurung 2.1. Stockfish is not a complete chess program and requires a UCI-compatible
-graphical user interface (GUI) (e.g. XBoard with PolyGlot, Scid, Cute Chess, eboard,
-Arena, Sigma Chess, Shredder, Chess Partner or Fritz) in order to be used comfortably.
-Read the documentation for your GUI of choice for information about how to use
+[Stockfish][website-link] is a **free and strong UCI chess engine** derived from
+Glaurung 2.1 that analyzes chess positions and computes the optimal moves.
+
+Stockfish **does not include a graphical user interface** (GUI) that is required
+to display a chessboard and to make it easy to input moves. These GUIs are
+developed independently from Stockfish and are available online. **Read the
+documentation for your GUI** of choice for information about how to use
 Stockfish with it.
 
-The Stockfish engine features two evaluation functions for chess. The efficiently
-updatable neural network (NNUE) based evaluation is the default and by far the strongest.
-The classical evaluation based on handcrafted terms remains available. The strongest
-network is integrated in the binary and downloaded automatically during the build process.
-The NNUE evaluation benefits from the vector intrinsics available on most CPUs (sse2,
-avx2, neon, or similar).
+See also the Stockfish [documentation][wiki-usage-link] for further usage help.
 
 ## Files
 
@@ -36,338 +48,112 @@ This distribution of Stockfish consists of the following files:
 
   * [README.md][readme-link], the file you are currently reading.
 
-  * [Copying.txt][license-link], a text file containing the GNU General Public License
-    version 3.
+  * [Copying.txt][license-link], a text file containing the GNU General Public
+    License version 3.
 
   * [AUTHORS][authors-link], a text file with the list of authors for the project.
 
-  * [src][src-link], a subdirectory containing the full source code, including a Makefile
-    that can be used to compile Stockfish on Unix-like systems.
+  * [src][src-link], a subdirectory containing the full source code, including a
+    Makefile that can be used to compile Stockfish on Unix-like systems.
 
-  * a file with the .nnue extension, storing the neural network for the NNUE evaluation.
-    Binary distributions will have this file embedded.
+  * a file with the .nnue extension, storing the neural network for the NNUE
+    evaluation. Binary distributions will have this file embedded.
 
-## The UCI protocol and available options
+## The UCI protocol
 
-The Universal Chess Interface (UCI) is a standard protocol used to communicate with
-a chess engine, and is the recommended way to do so for typical graphical user interfaces
-(GUI) or chess tools. Stockfish implements the majority of its options as described
-in [the UCI protocol][uci-link].
+The [Universal Chess Interface][uci-link] (UCI) is a standard text-based protocol
+used to communicate with a chess engine and is the recommended way to do so for
+typical graphical user interfaces (GUI) or chess tools. Stockfish implements the
+majority of its options.
 
-Developers can see the default values for UCI options available in Stockfish by typing
-`./stockfish uci` in a terminal, but the majority of users will typically see them and
-change them via a chess GUI. This is a list of available UCI options in Stockfish:
+Developers can see the default values for the UCI options available in Stockfish
+by typing `./stockfish uci` in a terminal, but most users should typically use a
+chess GUI to interact with Stockfish.
 
-  * #### Threads
-    The number of CPU threads used for searching a position. For best performance, set
-    this equal to the number of CPU cores available.
+For more information on UCI or debug commands, see our [documentation][wiki-commands-link].
 
-  * #### Hash
-    The size of the hash table in MB. It is recommended to set Hash after setting Threads.
+## Compiling Stockfish
 
-  * #### Clear Hash
-    Clear the hash table.
+Stockfish has support for 32 or 64-bit CPUs, certain hardware instructions,
+big-endian machines such as Power PC, and other platforms.
 
-  * #### Ponder
-    Let Stockfish ponder its next move while the opponent is thinking.
-
-  * #### MultiPV
-    Output the N best lines (principal variations, PVs) when searching.
-    Leave at 1 for best performance.
-
-  * #### Use NNUE
-    Toggle between the NNUE and classical evaluation functions. If set to "true",
-    the network parameters must be available to load from file (see also EvalFile),
-    if they are not embedded in the binary.
-
-  * #### EvalFile
-    The name of the file of the NNUE evaluation parameters. Depending on the GUI the
-    filename might have to include the full path to the folder/directory that contains
-    the file. Other locations, such as the directory that contains the binary and the
-    working directory, are also searched.
-
-  * #### UCI_AnalyseMode
-    An option handled by your GUI.
-
-  * #### UCI_Chess960
-    An option handled by your GUI. If true, Stockfish will play Chess960.
-
-  * #### UCI_ShowWDL
-    If enabled, show approximate WDL statistics as part of the engine output.
-    These WDL numbers model expected game outcomes for a given evaluation and
-    game ply for engine self-play at fishtest LTC conditions (60+0.6s per game).
-
-  * #### UCI_LimitStrength
-    Enable weaker play aiming for an Elo rating as set by UCI_Elo. This option overrides Skill Level.
-
-  * #### UCI_Elo
-    If enabled by UCI_LimitStrength, aim for an engine strength of the given Elo.
-    This Elo rating has been calibrated at a time control of 120s+1.0s and anchored to +- 100 Elo to CCRL Blitz.
-
-  * #### Skill Level
-    Lower the Skill Level in order to make Stockfish play weaker (see also UCI_LimitStrength).
-    Internally, MultiPV is enabled, and with a certain probability depending on the Skill Level a
-    weaker move will be played.
-
-  * #### SyzygyPath
-    Path to the folders/directories storing the Syzygy tablebase files. Multiple
-    directories are to 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 HDD. It is recommended to verify all md5 checksums
-    of the downloaded tablebase files (`md5sum -c checksum.md5`) as corruption will
-    lead to engine crashes.
-
-  * #### SyzygyProbeDepth
-    Minimum remaining search depth for which a position is probed. Set this option
-    to a higher value to probe less aggressively if you experience too much slowdown
-    (in terms of nps) due to tablebase probing.
-
-  * #### Syzygy50MoveRule
-    Disable to let fifty-move rule draws detected by Syzygy tablebase probes count
-    as wins or losses. This is useful for ICCF correspondence games.
-
-  * #### SyzygyProbeLimit
-    Limit Syzygy tablebase probing to positions with at most this many pieces left
-    (including kings and pawns).
-
-  * #### Move Overhead
-    Assume a time delay of x ms due to network and GUI overheads. This is useful to
-    avoid losses on time in those cases.
-
-  * #### Slow Mover
-    Lower values will make Stockfish take less time in games, higher values will
-    make it think longer.
-
-  * #### nodestime
-    Tells the engine to use nodes searched instead of wall time to account for
-    elapsed time. Useful for engine testing.
-
-  * #### Debug Log File
-    Write all communication to and from the engine into a text file.
-
-For developers the following non-standard commands might be of interest, mainly useful for debugging:
-
-  * #### bench *ttSize threads limit fenFile limitType evalType*
-    Performs a standard benchmark using various options. The signature of a version
-    (standard node count) is obtained using all defaults. `bench` is currently
-    `bench 16 1 13 default depth mixed`.
-
-  * #### compiler
-    Give information about the compiler and environment used for building a binary.
-
-  * #### d
-    Display the current position, with ascii art and fen.
-
-  * #### eval
-    Return the evaluation of the current position.
-
-  * #### export_net [filename]
-    Exports the currently loaded network to a file.
-    If the currently loaded network is the embedded network and the filename
-    is not specified then the network is saved to the file matching the name
-    of the embedded network, as defined in evaluate.h.
-    If the currently loaded network is not the embedded network (some net set
-    through the UCI setoption) then the filename parameter is required and the
-    network is saved into that file.
-
-  * #### flip
-    Flips the side to move.
-
-
-## A note on classical evaluation versus NNUE evaluation
-
-Both approaches assign a value to a position that is used in alpha-beta (PVS) search
-to find the best move. The classical evaluation computes this value as a function
-of various chess concepts, handcrafted by experts, tested and tuned using fishtest.
-The NNUE evaluation computes this value with a neural network based on basic
-inputs (e.g. piece positions only). The network is optimized and trained
-on the evaluations of millions of positions at moderate search depth.
-
-The NNUE evaluation was first introduced in shogi, and ported to Stockfish afterward.
-It can be evaluated efficiently on CPUs, and exploits the fact that only parts
-of the neural network need to be updated after a typical chess move.
-[The nodchip repository][nodchip-link] provided the first version of the needed tools
-to train and develop the NNUE networks. Today, more advanced training tools are
-available in [the nnue-pytorch repository][pytorch-link], while data generation tools
-are available in [a dedicated branch][tools-link].
-
-On CPUs supporting modern vector instructions (avx2 and similar), the NNUE evaluation
-results in much stronger playing strength, even if the nodes per second computed by
-the engine is somewhat lower (roughly 80% of nps is typical).
-
-Notes:
-
-1) the NNUE evaluation depends on the Stockfish binary and the network parameter file
-(see the EvalFile UCI option). Not every parameter file is compatible with a given
-Stockfish binary, but the default value of the EvalFile UCI option is the name of a
-network that is guaranteed to be compatible with that binary.
-
-2) to use the NNUE evaluation, the additional data file with neural network parameters
-needs to be available. Normally, this file is already embedded in the binary or it can
-be downloaded. The filename for the default (recommended) net can be found as the default
-value of the `EvalFile` UCI option, with the format `nn-[SHA256 first 12 digits].nnue`
-(for instance, `nn-c157e0a5755b.nnue`). This file can be downloaded from
-```
-https://tests.stockfishchess.org/api/nn/[filename]
-```
-replacing `[filename]` as needed.
-
-## What to expect from the Syzygy tablebases?
-
-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 153.xx), this means
-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 this 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 the Syzygy tablebases 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 the Syzygy tablebases 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.
-
-## Large Pages
-
-Stockfish supports large pages on Linux and Windows. Large pages make
-the hash access more efficient, improving the engine speed, especially
-on large hash sizes. Typical increases are 5..10% in terms of nodes per
-second, but speed increases up to 30% have been measured. The support is
-automatic. Stockfish attempts to use large pages when available and
-will fall back to regular memory allocation when this is not the case.
-
-### Support on Linux
-
-Large page support on Linux is obtained by the Linux kernel
-transparent huge pages functionality. Typically, transparent huge pages
-are already enabled, and no configuration is needed.
-
-### Support on Windows
-
-The use of large pages requires "Lock Pages in Memory" privilege. See
-[Enable the Lock Pages in Memory Option (Windows)][lockpages-link]
-on how to enable this privilege, then run [RAMMap][rammap-link]
-to double-check that large pages are used. We suggest that you reboot
-your computer after you have enabled large pages, because long Windows
-sessions suffer from memory fragmentation, which may prevent Stockfish
-from getting large pages: a fresh session is better in this regard.
-
-## Compiling Stockfish yourself from the sources
-
-Stockfish has support for 32 or 64-bit CPUs, certain hardware
-instructions, big-endian machines such as Power PC, and other platforms.
-
-On Unix-like systems, it should be easy to compile Stockfish
-directly from the source code with the included Makefile in the folder
-`src`. In general it is recommended to run `make help` to see a list of make
-targets with corresponding descriptions.
+On Unix-like systems, it should be easy to compile Stockfish directly from the
+source code with the included Makefile in the folder `src`. In general, it is
+recommended to run `make help` to see a list of make targets with corresponding
+descriptions.
 
 ```
-    cd src
-    make help
-    make net
-    make build ARCH=x86-64-modern
+cd src
+make -j build ARCH=x86-64-modern
 ```
 
-When not using the Makefile to compile (for instance, with Microsoft MSVC) you
-need to manually set/unset some switches in the compiler command line; see
-file *types.h* for a quick reference.
+Detailed compilation instructions for all platforms can be found in our
+[documentation][wiki-compile-link].
 
-When reporting an issue or a bug, please tell us which Stockfish version
-and which compiler you used to create your executable. This information
-can be found by typing the following command in a console:
-
-```
-    ./stockfish compiler
-```
-
-## Understanding the code base and participating in the project
-
-Stockfish's improvement over the last decade has been a great community
-effort. There are a few ways to help contribute to its growth.
+## Contributing
 
 ### Donating hardware
 
-Improving Stockfish requires a massive amount of testing. You can donate
-your hardware resources by installing the [Fishtest Worker][worker-link]
-and view the current tests on [Fishtest][fishtest-link].
+Improving Stockfish requires a massive amount of testing. You can donate your
+hardware resources by installing the [Fishtest Worker][worker-link] and viewing
+the current tests on [Fishtest][fishtest-link].
 
 ### Improving the code
 
-If you want to help improve the code, there are several valuable resources:
-
-* [In this wiki,][programming-link] many techniques used in
+In the [chessprogramming wiki][programming-link], many techniques used in
 Stockfish are explained with a lot of background information.
+The [section on Stockfish][programmingsf-link] describes many features
+and techniques used by Stockfish. However, it is generic rather than
+focused on Stockfish's precise implementation.
 
-* [The section on Stockfish][programmingsf-link]
-describes many features and techniques used by Stockfish. However, it is
-generic rather than being focused on Stockfish's precise implementation.
-Nevertheless, a helpful resource.
-
-* The latest source can always be found on [GitHub][github-link].
-Discussions about Stockfish take place these days mainly in the [FishCooking][fishcooking-link]
-group and on the [Stockfish Discord channel][discord-link].
 The engine testing is done on [Fishtest][fishtest-link].
 If you want to help improve Stockfish, please read this [guideline][guideline-link]
 first, where the basics of Stockfish development are explained.
 
+Discussions about Stockfish take place these days mainly in the Stockfish
+[Discord server][discord-link]. This is also the best place to ask questions
+about the codebase and how to improve it.
 
 ## Terms of use
 
-Stockfish is free, and distributed under the **GNU General Public License version 3**
-(GPL v3). Essentially, this means you are free to do almost exactly
-what you want with the program, including distributing it among your
-friends, making it available for download from your website, selling
-it (either by itself or as part of some bigger software package), or
-using it as the starting point for a software project of your own.
+Stockfish is free and distributed under the
+[**GNU General Public License version 3**][license-link] (GPL v3). Essentially,
+this means you are free to do almost exactly what you want with the program,
+including distributing it among your friends, making it available for download
+from your website, selling it (either by itself or as part of some bigger
+software package), or using it as the starting point for a software project of
+your own.
 
-The only real limitation is that whenever you distribute Stockfish in
-some way, you MUST always include the license and the full source code
-(or a pointer to where the source code can be found) to generate the
-exact binary you are distributing. If you make any changes to the
-source code, these changes must also be made available under the GPL v3.
-
-For full details, read the copy of the GPL v3 found in the file named
-[*Copying.txt*][license-link].
+The only real limitation is that whenever you distribute Stockfish in some way,
+you MUST always include the license and the full source code (or a pointer to
+where the source code can be found) to generate the exact binary you are
+distributing. If you make any changes to the source code, these changes must
+also be made available under GPL v3.
 
 
 [authors-link]:       https://github.com/official-stockfish/Stockfish/blob/master/AUTHORS
 [build-link]:         https://github.com/official-stockfish/Stockfish/actions/workflows/stockfish.yml
 [commits-link]:       https://github.com/official-stockfish/Stockfish/commits/master
 [discord-link]:       https://discord.gg/GWDRS3kU6R
-[fishcooking-link]:   https://groups.google.com/g/fishcooking
+[issue-link]:         https://github.com/official-stockfish/Stockfish/issues/new?assignees=&labels=&template=BUG-REPORT.yml
+[discussions-link]:   https://github.com/official-stockfish/Stockfish/discussions/new
 [fishtest-link]:      https://tests.stockfishchess.org/tests
-[github-link]:        https://github.com/official-stockfish/Stockfish
 [guideline-link]:     https://github.com/glinscott/fishtest/wiki/Creating-my-first-test
 [license-link]:       https://github.com/official-stockfish/Stockfish/blob/master/Copying.txt
-[lockpages-link]:     https://docs.microsoft.com/en-us/sql/database-engine/configure-windows/enable-the-lock-pages-in-memory-option-windows
-[nodchip-link]:       https://github.com/nodchip/Stockfish
 [programming-link]:   https://www.chessprogramming.org/Main_Page
 [programmingsf-link]: https://www.chessprogramming.org/Stockfish
-[pytorch-link]:       https://github.com/glinscott/nnue-pytorch
-[rammap-link]:        https://docs.microsoft.com/en-us/sysinternals/downloads/rammap
 [readme-link]:        https://github.com/official-stockfish/Stockfish/blob/master/README.md
 [release-link]:       https://github.com/official-stockfish/Stockfish/releases/latest
 [src-link]:           https://github.com/official-stockfish/Stockfish/tree/master/src
 [stockfish128-logo]:  https://stockfishchess.org/images/logo/icon_128x128.png
-[tools-link]:         https://github.com/official-stockfish/Stockfish/tree/tools
-[uci-link]:           https://www.shredderchess.com/download/div/uci.zip
+[uci-link]:           https://backscattering.de/chess/uci/
 [website-link]:       https://stockfishchess.org
-[worker-link]:        https://github.com/glinscott/fishtest/wiki/Running-the-worker:-overview
+[website-blog-link]:  https://stockfishchess.org/blog/
+[wiki-link]:          https://github.com/official-stockfish/Stockfish/wiki
+[wiki-usage-link]:    https://github.com/official-stockfish/Stockfish/wiki/Download-and-usage
+[wiki-compile-link]:  https://github.com/official-stockfish/Stockfish/wiki/Compiling-from-source
+[wiki-commands-link]: https://github.com/official-stockfish/Stockfish/wiki/Commands
+[worker-link]:        https://github.com/glinscott/fishtest/wiki/Running-the-worker
 
 [build-badge]:        https://img.shields.io/github/actions/workflow/status/official-stockfish/Stockfish/stockfish.yml?branch=master&style=for-the-badge&label=stockfish&logo=github
 [commits-badge]:      https://img.shields.io/github/commits-since/official-stockfish/Stockfish/latest?style=for-the-badge

From cdec775a1555ef83cb9c878c42a11b7dde0a627b Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sat, 4 Mar 2023 18:14:23 +0100
Subject: [PATCH 086/192] Add CITATION.cff file

Make the stockfish software more easily citable, for example in academic papers.

fixes https://github.com/official-stockfish/Stockfish/issues/4419
closes https://github.com/official-stockfish/Stockfish/pull/4422

No functional change
---
 CITATION.cff | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)
 create mode 100644 CITATION.cff

diff --git a/CITATION.cff b/CITATION.cff
new file mode 100644
index 00000000..bc0889a8
--- /dev/null
+++ b/CITATION.cff
@@ -0,0 +1,23 @@
+# This CITATION.cff file was generated with cffinit.
+# Visit https://bit.ly/cffinit to generate yours today!
+
+cff-version: 1.2.0
+title: Stockfish
+message: >-
+  Please cite this software using the metadata from this
+  file.
+type: software
+authors:
+  - name: The Stockfish developers (see AUTHORS file)
+repository-code: 'https://github.com/official-stockfish/Stockfish'
+url: 'https://stockfishchess.org/'
+repository-artifact: 'https://stockfishchess.org/download/'
+abstract: Stockfish is a free and strong UCI chess engine.
+keywords:
+  - chess
+  - artificial intelligence (AI)
+  - tree search
+  - alpha-beta search
+  - neural networks (NN)
+  - efficiently updatable neural networks (NNUE)
+license: GPL-3.0

From 70dfa141d560c18cd1aa28884b7cd8ab0f094944 Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Sun, 5 Mar 2023 17:10:52 +0200
Subject: [PATCH 087/192] Clarify the description of the x86-64-vnni256 and
 x86-64-avxvnni architectures

Now it is clearly explained that "x86-64-vnni256" requires full
support of AVX512-VNNI, but only 256-bit operands are used.

closes https://github.com/official-stockfish/Stockfish/pull/4427

No functional change
---
 AUTHORS      | 1 +
 src/Makefile | 8 ++++----
 2 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 634de4a3..49f7009f 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -134,6 +134,7 @@ Matt Ginsberg (mattginsberg)
 Matthew Lai (matthewlai)
 Matthew Sullivan (Matt14916)
 Max A. (Disservin)
+Maxim Masiutin (maximmasiutin)
 Maxim Molchanov (Maxim)
 Michael An (man)
 Michael Byrne (MichaelB7)
diff --git a/src/Makefile b/src/Makefile
index 3d6432fd..774ba6ea 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -92,7 +92,7 @@ VPATH = syzygy:nnue:nnue/features
 # avx2 = yes/no       --- -mavx2             --- Use Intel Advanced Vector Extensions 2
 # avxvnni = yes/no    --- -mavxvnni          --- Use Intel Vector Neural Network Instructions AVX
 # avx512 = yes/no     --- -mavx512bw         --- Use Intel Advanced Vector Extensions 512
-# vnni256 = yes/no    --- -mavx512vnni       --- Use Intel Vector Neural Network Instructions 256
+# vnni256 = yes/no    --- -mavx256vnni       --- Use Intel Vector Neural Network Instructions 512 with 256bit operands
 # vnni512 = yes/no    --- -mavx512vnni       --- Use Intel Vector Neural Network Instructions 512
 # neon = yes/no       --- -DUSE_NEON         --- Use ARM SIMD architecture
 # dotprod = yes/no    --- -DUSE_NEON_DOTPROD --- Use ARM advanced SIMD Int8 dot product instructions
@@ -773,10 +773,10 @@ help:
 	@echo ""
 	@echo "Supported archs:"
 	@echo ""
-	@echo "x86-64-vnni512          > x86 64-bit with vnni support 512bit wide"
-	@echo "x86-64-vnni256          > x86 64-bit with vnni support 256bit wide"
+	@echo "x86-64-vnni512          > x86 64-bit with vnni 512bit support"
+	@echo "x86-64-vnni256          > x86 64-bit with vnni 512bit support, limit operands to 256bit wide"
 	@echo "x86-64-avx512           > x86 64-bit with avx512 support"
-	@echo "x86-64-avxvnni          > x86 64-bit with avxvnni support"
+	@echo "x86-64-avxvnni          > x86 64-bit with vnni 256bit support"
 	@echo "x86-64-bmi2             > x86 64-bit with bmi2 support"
 	@echo "x86-64-avx2             > x86 64-bit with avx2 support"
 	@echo "x86-64-sse41-popcnt     > x86 64-bit with sse41 and popcnt support"

From 6ce225bb4c31298b131714eff67b56de3b8ee78d Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sun, 5 Mar 2023 17:26:12 +0100
Subject: [PATCH 088/192] Fix TB after capture_stage fix

https://github.com/official-stockfish/Stockfish/commit/5c75c1c2fbb7bb4f0bf7c44fb855c415b788cbf7
introduced a capture_stage() function, but TB usage needs a pure capture() function.

closes https://github.com/official-stockfish/Stockfish/pull/4428

No functional change.
---
 src/position.h         | 11 ++++++++---
 src/syzygy/tbprobe.cpp |  4 ++--
 2 files changed, 10 insertions(+), 5 deletions(-)

diff --git a/src/position.h b/src/position.h
index 485540ef..1fc6b3b8 100644
--- a/src/position.h
+++ b/src/position.h
@@ -125,6 +125,7 @@ public:
   // Properties of moves
   bool legal(Move m) const;
   bool pseudo_legal(const Move m) const;
+  bool capture(Move m) const;
   bool capture_stage(Move m) const;
   bool gives_check(Move m) const;
   Piece moved_piece(Move m) const;
@@ -381,14 +382,18 @@ inline bool Position::is_chess960() const {
   return chess960;
 }
 
+inline bool Position::capture(Move m) const {
+  assert(is_ok(m));
+  return     (!empty(to_sq(m)) && type_of(m) != CASTLING)
+          ||  type_of(m) == EN_PASSANT;
+}
+
 // returns true if a move is generated from the capture stage
 // having also queen promotions covered, i.e. consistency with the capture stage move generation
 // is needed to avoid the generation of duplicate moves.
 inline bool Position::capture_stage(Move m) const {
   assert(is_ok(m));
-  return     (!empty(to_sq(m)) && type_of(m) != CASTLING)
-          || (type_of(m) == PROMOTION && promotion_type(m) == QUEEN)
-          ||  type_of(m) == EN_PASSANT;
+  return  capture(m) || (type_of(m) == PROMOTION && promotion_type(m) == QUEEN);
 }
 
 inline Piece Position::captured_piece() const {
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index 2a9e1b68..b594ac37 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -1203,7 +1203,7 @@ WDLScore search(Position& pos, ProbeState* result) {
 
     for (const Move move : moveList)
     {
-        if (   !pos.capture_stage(move)
+        if (   !pos.capture(move)
             && (!CheckZeroingMoves || type_of(pos.moved_piece(move)) != PAWN))
             continue;
 
@@ -1472,7 +1472,7 @@ int Tablebases::probe_dtz(Position& pos, ProbeState* result) {
 
     for (const Move move : MoveList<LEGAL>(pos))
     {
-        bool zeroing = pos.capture_stage(move) || type_of(pos.moved_piece(move)) == PAWN;
+        bool zeroing = pos.capture(move) || type_of(pos.moved_piece(move)) == PAWN;
 
         pos.do_move(move, st);
 

From 39da50ed23ee3f1bd32a58c8f02471faa9a9fd63 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Tue, 7 Mar 2023 01:54:20 +0300
Subject: [PATCH 089/192] Do more negative extensions

This patch does negatively extend transposition table move if singular search failed and tt value is not bigger than alpha.
Logic is close to what we had before recent simplification of negative extensions but uses or condition instead of and condition.

Passed STC:
https://tests.stockfishchess.org/tests/view/6404c8102644b62c33934607
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 119040 W: 31841 L: 31416 D: 55783
Ptnml(0-2): 356, 13070, 32292, 13397, 405

Passed LTC:
https://tests.stockfishchess.org/tests/view/6405abda2644b62c33937119
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 47216 W: 12816 L: 12496 D: 21904
Ptnml(0-2): 12, 4500, 14286, 4776, 34

closes https://github.com/official-stockfish/Stockfish/pull/4430

bench 4747020
---
 src/search.cpp | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index fcdb8d67..349d4389 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1109,6 +1109,10 @@ moves_loop: // When in check, search starts here
               // If the eval of ttMove is less than value, we reduce it (negative extension)
               else if (ttValue <= value)
                   extension = -1;
+
+              // If the eval of ttMove is less than alpha, we reduce it (negative extension)
+              else if (ttValue <= alpha)
+                  extension = -1;
           }
 
           // Check extensions (~1 Elo)

From a48573e15fbd18fecb087c9fc02c38a07cece68b Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Thu, 9 Mar 2023 18:33:13 -0600
Subject: [PATCH 090/192] More negative extensions on nonsingular nonpv nodes.

Following up the previous gainer also in this nonsingular node
section of code. Credit shared with @FauziAkram for realizing
this nonsingular node stuff had some potential, and @XInTheDark
for reminding us that !PvNodes better handle extensions/reductions
than Pv.

Passed STC: https://tests.stockfishchess.org/tests/view/640a7bb32644b62c339457c3
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 136776 W: 36598 L: 36149 D: 64029
Ptnml(0-2): 439, 14834, 37384, 15301, 430

Passed LTC: https://tests.stockfishchess.org/tests/view/640c43a02644b62c3394b23c
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 79536 W: 21363 L: 20984 D: 37189
Ptnml(0-2): 28, 7525, 24285, 7900, 30

closes https://github.com/official-stockfish/Stockfish/pull/4441

Bench: 4444953
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 349d4389..582e4457 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1104,7 +1104,7 @@ moves_loop: // When in check, search starts here
 
               // If the eval of ttMove is greater than beta, we reduce it (negative extension)
               else if (ttValue >= beta)
-                  extension = -2;
+                  extension = -2 - !PvNode;
 
               // If the eval of ttMove is less than value, we reduce it (negative extension)
               else if (ttValue <= value)

From 78532af9dc585a81730d5d28763cc9d5273f96d1 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Mon, 13 Mar 2023 15:57:58 +0300
Subject: [PATCH 091/192] Do more singular extensions

This patch continues trend of last VLTC tuning - as measured by dubslow
most of it gains was in lowering marging in calculation of singularBeta.
This patch is a manual adjustment on top of it - it lowers multiplier
of depth in calculation of singularBeta even further, from 2/3 to 1,5/2,5.

Was negative at STC:
https://tests.stockfishchess.org/tests/view/64089c632644b62c3393fc12
Elo: -2.49 +-1.7 (95%) LOS: 0.2%
Total: 40000 W: 10601 L: 10888 D: 18511
Ptnml(0-2): 123, 4580, 10875, 4305, 117
nElo: -5.03 +-3.4 (95%) PairsRatio: 0.94

Passed 180+1.8 SPRT:
https://tests.stockfishchess.org/tests/view/640096dae74a12625bcf3b33
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 160952 W: 43753 L: 43342 D: 73857
Ptnml(0-2): 25, 13984, 52039, 14411, 17

Passed 60+0.6 8 threads SPRT:
https://tests.stockfishchess.org/tests/view/640dca8e65775d3b539cb7f6
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 38824 W: 10825 L: 10554 D: 17445
Ptnml(0-2): 0, 2939, 13268, 3200, 5

closes https://github.com/official-stockfish/Stockfish/pull/4443

bench 4776866
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 582e4457..7eb4a0c5 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1071,7 +1071,7 @@ moves_loop: // When in check, search starts here
               && (tte->bound() & BOUND_LOWER)
               &&  tte->depth() >= depth - 3)
           {
-              Value singularBeta = ttValue - (2 + (ss->ttPv && !PvNode)) * depth;
+              Value singularBeta = ttValue - (3 + 2 * (ss->ttPv && !PvNode)) * depth / 2;
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;

From 55896a1384b022624a8c33b9f7f51b6b551eed50 Mon Sep 17 00:00:00 2001
From: Alfredo Menezes <lomalfredo8@gmail.com>
Date: Sun, 12 Mar 2023 22:29:07 -0300
Subject: [PATCH 092/192] Change mode of incbin.h

Keep incbin.h with the same mode as the other source files.

A mode diff might show up when working with patch files or sending the source code between devices.
This patch should fix such behaviour.

closes https://github.com/official-stockfish/Stockfish/pull/4442

No functional change
---
 src/incbin/incbin.h | 0
 1 file changed, 0 insertions(+), 0 deletions(-)
 mode change 100755 => 100644 src/incbin/incbin.h

diff --git a/src/incbin/incbin.h b/src/incbin/incbin.h
old mode 100755
new mode 100644

From d1e17989b51f220629e4e81a4bf413974f4b18e2 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Sat, 11 Mar 2023 22:08:35 +0100
Subject: [PATCH 093/192] Fix Makefile for clang 16

The clang 16 release will remove the -fexperimental-new-pass-manager
flag (see https://github.com/llvm/llvm-project/commit/69b2b7282e92a1b576b7bd26f3b16716a5027e8e).
Thus, the commit adapts the Makefile to use this flag only for older
clang versions.

closes https://github.com/official-stockfish/Stockfish/pull/4437

No functional change
---
 src/Makefile | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index 774ba6ea..e257bc63 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -584,7 +584,10 @@ ifeq ($(optimize),yes)
 	endif
 
 	ifeq ($(comp),clang)
-		CXXFLAGS += -fexperimental-new-pass-manager
+		clangmajorversion = $(shell $(CXX) -dumpversion 2>/dev/null | cut -f1 -d.)
+		ifeq ($(shell expr $(clangmajorversion) \< 16),1)
+			CXXFLAGS += -fexperimental-new-pass-manager
+		endif
 	endif
 endif
 

From 7077fbdd1481829a0a20b6975c4245609118b938 Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Sat, 11 Mar 2023 11:51:08 -0800
Subject: [PATCH 094/192] Remove redundant condition from capture_stage()

Change a non functional promotion check to an assert.

closes https://github.com/official-stockfish/Stockfish/pull/4436

No functional change
---
 src/position.cpp | 3 +--
 src/position.h   | 2 +-
 2 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/src/position.cpp b/src/position.cpp
index 37aa2e9e..632a40b5 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -569,8 +569,7 @@ bool Position::pseudo_legal(const Move m) const {
                         : MoveList<NON_EVASIONS>(*this).contains(m);
 
   // Is not a promotion, so promotion piece must be empty
-  if (promotion_type(m) - KNIGHT != NO_PIECE_TYPE)
-      return false;
+  assert(promotion_type(m) - KNIGHT == NO_PIECE_TYPE);
 
   // If the 'from' square is not occupied by a piece belonging to the side to
   // move, the move is obviously not legal.
diff --git a/src/position.h b/src/position.h
index 1fc6b3b8..cc606a5a 100644
--- a/src/position.h
+++ b/src/position.h
@@ -393,7 +393,7 @@ inline bool Position::capture(Move m) const {
 // is needed to avoid the generation of duplicate moves.
 inline bool Position::capture_stage(Move m) const {
   assert(is_ok(m));
-  return  capture(m) || (type_of(m) == PROMOTION && promotion_type(m) == QUEEN);
+  return  capture(m) || promotion_type(m) == QUEEN;
 }
 
 inline Piece Position::captured_piece() const {

From f0556dcbe3ba2fc804ab26d4552446602a75f064 Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@phxl0401.wp.lan>
Date: Wed, 1 Mar 2023 10:29:51 +0100
Subject: [PATCH 095/192] Small cleanups

remove some unneeded assignments, typos, incorrect comments, add authors entry.

closes https://github.com/official-stockfish/Stockfish/pull/4417

no functional change
---
 AUTHORS                             | 1 +
 src/nnue/nnue_feature_transformer.h | 2 +-
 src/position.cpp                    | 3 ---
 src/search.cpp                      | 2 --
 4 files changed, 2 insertions(+), 6 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 49f7009f..9b36111e 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -190,6 +190,7 @@ Sergei Antonov (saproj)
 Sergei Ivanov (svivanov72)
 Sergio Vieri (sergiovieri)
 sf-x
+Shahin M. Shahin (peregrine)
 Shane Booth (shane31)
 Shawn Varghese (xXH4CKST3RXx)
 Siad Daboul (Topologist)
diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index 8087ea55..a1888c7a 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -363,7 +363,7 @@ namespace Stockfish::Eval::NNUE {
     // NOTE: The parameter states_to_update is an array of position states, ending with nullptr.
     //       All states must be sequential, that is states_to_update[i] must either be reachable
     //       by repeatedly applying ->previous from states_to_update[i+1] or states_to_update[i] == nullptr.
-    //       computed_st must be reachable by repeatadly applying ->previous on states_to_update[0], if not nullptr.
+    //       computed_st must be reachable by repeatedly applying ->previous on states_to_update[0], if not nullptr.
     template<Color Perspective, size_t N>
     void update_accumulator_incremental(const Position& pos, StateInfo* computed_st, StateInfo* states_to_update[N]) const {
       static_assert(N > 0);
diff --git a/src/position.cpp b/src/position.cpp
index 632a40b5..aeac23a3 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -765,9 +765,6 @@ void Position::do_move(Move m, StateInfo& newSt, bool givesCheck) {
       // Update board and piece lists
       remove_piece(capsq);
 
-      if (type_of(m) == EN_PASSANT)
-          board[capsq] = NO_PIECE;
-
       // Update material hash key and prefetch access to materialTable
       k ^= Zobrist::psq[captured][capsq];
       st->materialKey ^= Zobrist::psq[captured][pieceCount[captured]];
diff --git a/src/search.cpp b/src/search.cpp
index 7eb4a0c5..d6571a14 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -601,7 +601,6 @@ namespace {
 
     assert(0 <= ss->ply && ss->ply < MAX_PLY);
 
-    (ss+1)->ttPv         = false;
     (ss+1)->excludedMove = bestMove = MOVE_NONE;
     (ss+2)->killers[0]   = (ss+2)->killers[1] = MOVE_NONE;
     (ss+2)->cutoffCnt    = 0;
@@ -1075,7 +1074,6 @@ moves_loop: // When in check, search starts here
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;
-              // the search with excludedMove will update ss->staticEval
               value = search<NonPV>(pos, ss, singularBeta - 1, singularBeta, singularDepth, cutNode);
               ss->excludedMove = MOVE_NONE;
 

From 515b66f18833ed87e97313d2ec4dfa4e2329d3df Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Sun, 12 Mar 2023 01:22:55 +0300
Subject: [PATCH 096/192] Fix null move issue

Fix altering for stats landing on B1 Square after a null move and fix considering counter-moves on A1 for root node.

fixes https://github.com/official-stockfish/Stockfish/issues/4333 by preventing calls to from_sq and to_sq functions over null-moves and none-moves.

closes https://github.com/official-stockfish/Stockfish/pull/4448

bench: 4980082
---
 src/search.cpp | 19 ++++++++++++-------
 src/types.h    | 10 ++++++----
 src/uci.cpp    |  6 +++---
 3 files changed, 21 insertions(+), 14 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index d6571a14..466e0d6f 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -605,7 +605,7 @@ namespace {
     (ss+2)->killers[0]   = (ss+2)->killers[1] = MOVE_NONE;
     (ss+2)->cutoffCnt    = 0;
     ss->doubleExtensions = (ss-1)->doubleExtensions;
-    Square prevSq        = to_sq((ss-1)->currentMove);
+    Square prevSq        = is_ok((ss-1)->currentMove) ? to_sq((ss-1)->currentMove) : SQ_NONE;
 
     // Initialize statScore to zero for the grandchildren of the current position.
     // So statScore is shared between all grandchildren and only the first grandchild
@@ -647,7 +647,7 @@ namespace {
                     update_quiet_stats(pos, ss, ttMove, stat_bonus(depth));
 
                 // Extra penalty for early quiet moves of the previous ply (~0 Elo on STC, ~2 Elo on LTC)
-                if ((ss-1)->moveCount <= 2 && !priorCapture)
+                if (prevSq != SQ_NONE && (ss-1)->moveCount <= 2 && !priorCapture)
                     update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, -stat_bonus(depth + 1));
             }
             // Penalty for a quiet ttMove that fails low (~1 Elo)
@@ -935,7 +935,7 @@ moves_loop: // When in check, search starts here
                                           nullptr                   , (ss-4)->continuationHistory,
                                           nullptr                   , (ss-6)->continuationHistory };
 
-    Move countermove = thisThread->counterMoves[pos.piece_on(prevSq)][prevSq];
+    Move countermove = prevSq != SQ_NONE ? thisThread->counterMoves[pos.piece_on(prevSq)][prevSq] : MOVE_NONE;
 
     MovePicker mp(pos, ttMove, depth, &thisThread->mainHistory,
                                       &captureHistory,
@@ -1383,7 +1383,7 @@ moves_loop: // When in check, search starts here
                          quietsSearched, quietCount, capturesSearched, captureCount, depth);
 
     // Bonus for prior countermove that caused the fail low
-    else if (!priorCapture)
+    else if (!priorCapture && prevSq != SQ_NONE)
     {
         int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 97 * depth) + ((ss-1)->moveCount > 10);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
@@ -1525,7 +1525,7 @@ moves_loop: // When in check, search starts here
     // to search the moves. Because the depth is <= 0 here, only captures,
     // queen promotions, and other checks (only if depth >= DEPTH_QS_CHECKS)
     // will be generated.
-    Square prevSq = to_sq((ss-1)->currentMove);
+    Square prevSq = (ss-1)->currentMove != MOVE_NULL ? to_sq((ss-1)->currentMove) : SQ_NONE;
     MovePicker mp(pos, ttMove, depth, &thisThread->mainHistory,
                                       &thisThread->captureHistory,
                                       contHist,
@@ -1714,7 +1714,8 @@ moves_loop: // When in check, search starts here
     Thread* thisThread = pos.this_thread();
     CapturePieceToHistory& captureHistory = thisThread->captureHistory;
     Piece moved_piece = pos.moved_piece(bestMove);
-    PieceType captured = type_of(pos.piece_on(to_sq(bestMove)));
+    PieceType captured;
+
     int bonus1 = stat_bonus(depth + 1);
 
     if (!pos.capture_stage(bestMove))
@@ -1733,12 +1734,16 @@ moves_loop: // When in check, search starts here
         }
     }
     else
+    {
         // Increase stats for the best move in case it was a capture move
+        captured = type_of(pos.piece_on(to_sq(bestMove)));
         captureHistory[moved_piece][to_sq(bestMove)][captured] << bonus1;
+    }
 
     // Extra penalty for a quiet early move that was not a TT move or
     // main killer move in previous ply when it gets refuted.
-    if (   ((ss-1)->moveCount == 1 + (ss-1)->ttHit || ((ss-1)->currentMove == (ss-1)->killers[0]))
+    if (   prevSq != SQ_NONE
+        && ((ss-1)->moveCount == 1 + (ss-1)->ttHit || ((ss-1)->currentMove == (ss-1)->killers[0]))
         && !pos.captured_piece())
             update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, -bonus1);
 
diff --git a/src/types.h b/src/types.h
index 37ce343a..06b0a059 100644
--- a/src/types.h
+++ b/src/types.h
@@ -416,6 +416,10 @@ inline Color color_of(Piece pc) {
   return Color(pc >> 3);
 }
 
+constexpr bool is_ok(Move m) {
+  return m != MOVE_NONE && m != MOVE_NULL;
+}
+
 constexpr bool is_ok(Square s) {
   return s >= SQ_A1 && s <= SQ_H8;
 }
@@ -445,10 +449,12 @@ constexpr Direction pawn_push(Color c) {
 }
 
 constexpr Square from_sq(Move m) {
+  assert(is_ok(m));
   return Square((m >> 6) & 0x3F);
 }
 
 constexpr Square to_sq(Move m) {
+  assert(is_ok(m));
   return Square(m & 0x3F);
 }
 
@@ -473,10 +479,6 @@ constexpr Move make(Square from, Square to, PieceType pt = KNIGHT) {
   return Move(T + ((pt - KNIGHT) << 12) + (from << 6) + to);
 }
 
-constexpr bool is_ok(Move m) {
-  return from_sq(m) != to_sq(m); // Catch MOVE_NULL and MOVE_NONE
-}
-
 /// Based on a congruential pseudo random number generator
 constexpr Key make_key(uint64_t seed) {
   return seed * 6364136223846793005ULL + 1442695040888963407ULL;
diff --git a/src/uci.cpp b/src/uci.cpp
index 3883b3d3..d3d99243 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -358,15 +358,15 @@ std::string UCI::square(Square s) {
 
 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";
 
+  Square from = from_sq(m);
+  Square to = to_sq(m);
+
   if (type_of(m) == CASTLING && !chess960)
       to = make_square(to > from ? FILE_G : FILE_C, rank_of(from));
 

From af4b62a593cc4fa6d7d34110c41301028a5c9695 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Mon, 13 Mar 2023 19:35:27 +0100
Subject: [PATCH 097/192] NNUE namespace cleanup

This patch moves the nnue namespace in the appropiate header that correspondes with the definition.
It also makes navigation a bit easier.

closes https://github.com/official-stockfish/Stockfish/pull/4445

No functional change
---
 src/evaluate.cpp         | 6 +++---
 src/evaluate.h           | 8 --------
 src/nnue/evaluate_nnue.h | 8 ++++++++
 src/search.cpp           | 1 +
 src/uci.cpp              | 1 +
 5 files changed, 13 insertions(+), 11 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index cf6f23ea..99b87300 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -36,7 +36,7 @@
 #include "timeman.h"
 #include "uci.h"
 #include "incbin/incbin.h"
-
+#include "nnue/evaluate_nnue.h"
 
 // Macro to embed the default efficiently updatable neural network (NNUE) file
 // data in the engine binary (using incbin.h, by Dale Weiler).
@@ -95,7 +95,7 @@ namespace Eval {
             if (directory != "<internal>")
             {
                 ifstream stream(directory + eval_file, ios::binary);
-                if (load_eval(eval_file, stream))
+                if (NNUE::load_eval(eval_file, stream))
                     currentEvalFileName = eval_file;
             }
 
@@ -111,7 +111,7 @@ namespace Eval {
                 (void) gEmbeddedNNUEEnd; // Silence warning on unused variable
 
                 istream stream(&buffer);
-                if (load_eval(eval_file, stream))
+                if (NNUE::load_eval(eval_file, stream))
                     currentEvalFileName = eval_file;
             }
         }
diff --git a/src/evaluate.h b/src/evaluate.h
index 5238cb81..3615fe6d 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -43,17 +43,9 @@ namespace Eval {
 
   namespace NNUE {
 
-    std::string trace(Position& pos);
-    Value evaluate(const Position& pos, bool adjusted = false, int* complexity = nullptr);
-    void hint_common_parent_position(const Position& pos);
-
     void init();
     void verify();
 
-    bool load_eval(std::string name, std::istream& stream);
-    bool save_eval(std::ostream& stream);
-    bool save_eval(const std::optional<std::string>& filename);
-
   } // namespace NNUE
 
 } // namespace Eval
diff --git a/src/nnue/evaluate_nnue.h b/src/nnue/evaluate_nnue.h
index 15638cae..b84bed8b 100644
--- a/src/nnue/evaluate_nnue.h
+++ b/src/nnue/evaluate_nnue.h
@@ -55,6 +55,14 @@ namespace Stockfish::Eval::NNUE {
   template <typename T>
   using LargePagePtr = std::unique_ptr<T, LargePageDeleter<T>>;
 
+  std::string trace(Position& pos);
+  Value evaluate(const Position& pos, bool adjusted = false, int* complexity = nullptr);
+  void hint_common_parent_position(const Position& pos);
+
+  bool load_eval(std::string name, std::istream& stream);
+  bool save_eval(std::ostream& stream);
+  bool save_eval(const std::optional<std::string>& filename);
+
 }  // namespace Stockfish::Eval::NNUE
 
 #endif // #ifndef NNUE_EVALUATE_NNUE_H_INCLUDED
diff --git a/src/search.cpp b/src/search.cpp
index 466e0d6f..17c1c28b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -34,6 +34,7 @@
 #include "tt.h"
 #include "uci.h"
 #include "syzygy/tbprobe.h"
+#include "nnue/evaluate_nnue.h"
 
 namespace Stockfish {
 
diff --git a/src/uci.cpp b/src/uci.cpp
index d3d99243..8f9684ee 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -32,6 +32,7 @@
 #include "tt.h"
 #include "uci.h"
 #include "syzygy/tbprobe.h"
+#include "nnue/evaluate_nnue.h"
 
 using namespace std;
 

From 02e4697055519ed206fa76e4ef9abb9f156cd1a0 Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@phxl0401.wp.lan>
Date: Mon, 13 Mar 2023 18:32:40 +0100
Subject: [PATCH 098/192] Remove 'si' StateInfo variable/parameter.

Since st is a member of position we don't need to pass it separately as
parameter.

While being there also remove some line in pos_is_ok, where
a copy of StateInfo was made by using default copy constructor and
then verified it's correctedness by doing a memcmp.
There is no point in doing that.

Passed non-regression test
https://tests.stockfishchess.org/tests/view/64098d562644b62c33942b35
LLR: 3.24 (-2.94,2.94) <-1.75,0.25>
Total: 548960 W: 145834 L: 146134 D: 256992
Ptnml(0-2): 1617, 57652, 156261, 57314, 1636

closes https://github.com/official-stockfish/Stockfish/pull/4444

No functional change
---
 src/position.cpp | 58 ++++++++++++++++++++++--------------------------
 src/position.h   |  4 ++--
 2 files changed, 28 insertions(+), 34 deletions(-)

diff --git a/src/position.cpp b/src/position.cpp
index aeac23a3..171193ec 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -282,7 +282,7 @@ Position& Position::set(const string& fenStr, bool isChess960, StateInfo* si, Th
 
   chess960 = isChess960;
   thisThread = th;
-  set_state(st);
+  set_state();
 
   assert(pos_is_ok());
 
@@ -313,19 +313,19 @@ void Position::set_castling_right(Color c, Square rfrom) {
 
 /// Position::set_check_info() sets king attacks to detect if a move gives check
 
-void Position::set_check_info(StateInfo* si) const {
+void Position::set_check_info() const {
 
-  si->blockersForKing[WHITE] = slider_blockers(pieces(BLACK), square<KING>(WHITE), si->pinners[BLACK]);
-  si->blockersForKing[BLACK] = slider_blockers(pieces(WHITE), square<KING>(BLACK), si->pinners[WHITE]);
+  st->blockersForKing[WHITE] = slider_blockers(pieces(BLACK), square<KING>(WHITE), st->pinners[BLACK]);
+  st->blockersForKing[BLACK] = slider_blockers(pieces(WHITE), square<KING>(BLACK), st->pinners[WHITE]);
 
   Square ksq = square<KING>(~sideToMove);
 
-  si->checkSquares[PAWN]   = pawn_attacks_bb(~sideToMove, ksq);
-  si->checkSquares[KNIGHT] = attacks_bb<KNIGHT>(ksq);
-  si->checkSquares[BISHOP] = attacks_bb<BISHOP>(ksq, pieces());
-  si->checkSquares[ROOK]   = attacks_bb<ROOK>(ksq, pieces());
-  si->checkSquares[QUEEN]  = si->checkSquares[BISHOP] | si->checkSquares[ROOK];
-  si->checkSquares[KING]   = 0;
+  st->checkSquares[PAWN]   = pawn_attacks_bb(~sideToMove, ksq);
+  st->checkSquares[KNIGHT] = attacks_bb<KNIGHT>(ksq);
+  st->checkSquares[BISHOP] = attacks_bb<BISHOP>(ksq, pieces());
+  st->checkSquares[ROOK]   = attacks_bb<ROOK>(ksq, pieces());
+  st->checkSquares[QUEEN]  = st->checkSquares[BISHOP] | st->checkSquares[ROOK];
+  st->checkSquares[KING]   = 0;
 }
 
 
@@ -334,39 +334,39 @@ void Position::set_check_info(StateInfo* si) const {
 /// The function is only used when a new position is set up, and to verify
 /// the correctness of the StateInfo data when running in debug mode.
 
-void Position::set_state(StateInfo* si) const {
+void Position::set_state() const {
 
-  si->key = si->materialKey = 0;
-  si->pawnKey = Zobrist::noPawns;
-  si->nonPawnMaterial[WHITE] = si->nonPawnMaterial[BLACK] = VALUE_ZERO;
-  si->checkersBB = attackers_to(square<KING>(sideToMove)) & pieces(~sideToMove);
+  st->key = st->materialKey = 0;
+  st->pawnKey = Zobrist::noPawns;
+  st->nonPawnMaterial[WHITE] = st->nonPawnMaterial[BLACK] = VALUE_ZERO;
+  st->checkersBB = attackers_to(square<KING>(sideToMove)) & pieces(~sideToMove);
 
-  set_check_info(si);
+  set_check_info();
 
   for (Bitboard b = pieces(); b; )
   {
       Square s = pop_lsb(b);
       Piece pc = piece_on(s);
-      si->key ^= Zobrist::psq[pc][s];
+      st->key ^= Zobrist::psq[pc][s];
 
       if (type_of(pc) == PAWN)
-          si->pawnKey ^= Zobrist::psq[pc][s];
+          st->pawnKey ^= Zobrist::psq[pc][s];
 
       else if (type_of(pc) != KING)
-          si->nonPawnMaterial[color_of(pc)] += PieceValue[MG][pc];
+          st->nonPawnMaterial[color_of(pc)] += PieceValue[MG][pc];
   }
 
-  if (si->epSquare != SQ_NONE)
-      si->key ^= Zobrist::enpassant[file_of(si->epSquare)];
+  if (st->epSquare != SQ_NONE)
+      st->key ^= Zobrist::enpassant[file_of(st->epSquare)];
 
   if (sideToMove == BLACK)
-      si->key ^= Zobrist::side;
+      st->key ^= Zobrist::side;
 
-  si->key ^= Zobrist::castling[si->castlingRights];
+  st->key ^= Zobrist::castling[st->castlingRights];
 
   for (Piece pc : Pieces)
       for (int cnt = 0; cnt < pieceCount[pc]; ++cnt)
-          si->materialKey ^= Zobrist::psq[pc][cnt];
+          st->materialKey ^= Zobrist::psq[pc][cnt];
 }
 
 
@@ -865,7 +865,7 @@ void Position::do_move(Move m, StateInfo& newSt, bool givesCheck) {
   sideToMove = ~sideToMove;
 
   // Update king attacks used for fast check detection
-  set_check_info(st);
+  set_check_info();
 
   // Calculate the repetition info. It is the ply distance from the previous
   // occurrence of the same position, negative in the 3-fold case, or zero
@@ -1017,7 +1017,7 @@ void Position::do_null_move(StateInfo& newSt) {
 
   sideToMove = ~sideToMove;
 
-  set_check_info(st);
+  set_check_info();
 
   st->repetition = 0;
 
@@ -1320,12 +1320,6 @@ bool Position::pos_is_ok() const {
           if (p1 != p2 && (pieces(p1) & pieces(p2)))
               assert(0 && "pos_is_ok: Bitboards");
 
-  StateInfo si = *st;
-  ASSERT_ALIGNED(&si, Eval::NNUE::CacheLineSize);
-
-  set_state(&si);
-  if (std::memcmp(&si, st, sizeof(StateInfo)))
-      assert(0 && "pos_is_ok: State");
 
   for (Piece pc : Pieces)
       if (   pieceCount[pc] != popcount(pieces(color_of(pc), type_of(pc)))
diff --git a/src/position.h b/src/position.h
index cc606a5a..d3483bcf 100644
--- a/src/position.h
+++ b/src/position.h
@@ -179,8 +179,8 @@ public:
 private:
   // Initialization helpers (used while setting up a position)
   void set_castling_right(Color c, Square rfrom);
-  void set_state(StateInfo* si) const;
-  void set_check_info(StateInfo* si) const;
+  void set_state() const;
+  void set_check_info() const;
 
   // Other helpers
   void move_piece(Square from, Square to);

From 24b37e4586ba610d331048446bd036bec5544c4f Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@phxl0401.wp.lan>
Date: Mon, 20 Mar 2023 08:56:44 +0100
Subject: [PATCH 099/192] Verified SEE pruning for capturing and checking
 moves.

Patch analyzes field after SEE exchanges concluded with a recapture by
the opponent:
if opponent Queen/Rook/King results under attack after the exchanges, we
consider the move sharp and don't prune it.

Important note:
By accident I forgot to adjust 'occupied' when the king takes part in
the exchanges.
As result of this a move is considered sharp too, when opponent king
apparently can evade check by recapturing.
Surprisingly this seems contribute to patch's strength.

STC:
https://tests.stockfishchess.org/tests/view/640b16132644b62c33947397
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 116400 W: 31239 L: 30817 D: 54344
Ptnml(0-2): 350, 12742, 31618, 13116, 374

LTC:
https://tests.stockfishchess.org/tests/view/640c88092644b62c3394c1c5
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 177600 W: 47988 L: 47421 D: 82191
Ptnml(0-2): 62, 16905, 54317, 17436, 80

closes https://github.com/official-stockfish/Stockfish/pull/4453

bench: 5012145
---
 src/movepick.cpp |  6 +++---
 src/movepick.h   |  1 +
 src/position.cpp | 15 +++++++--------
 src/position.h   |  2 +-
 src/search.cpp   | 30 +++++++++++++++++++++++++-----
 5 files changed, 37 insertions(+), 17 deletions(-)

diff --git a/src/movepick.cpp b/src/movepick.cpp
index 36ee46b5..855f2b1d 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -95,7 +95,7 @@ MovePicker::MovePicker(const Position& p, Move ttm, Value th, const CapturePiece
 
   stage = PROBCUT_TT + !(ttm && pos.capture_stage(ttm)
                              && pos.pseudo_legal(ttm)
-                             && pos.see_ge(ttm, threshold));
+                             && pos.see_ge(ttm, occupied, threshold));
 }
 
 /// MovePicker::score() assigns a numerical value to each move in a list, used
@@ -197,7 +197,7 @@ top:
 
   case GOOD_CAPTURE:
       if (select<Next>([&](){
-                       return pos.see_ge(*cur, Value(-cur->value)) ?
+                       return pos.see_ge(*cur, occupied, Value(-cur->value)) ?
                               // Move losing capture to endBadCaptures to be tried later
                               true : (*endBadCaptures++ = *cur, false); }))
           return *(cur - 1);
@@ -264,7 +264,7 @@ top:
       return select<Best>([](){ return true; });
 
   case PROBCUT:
-      return select<Next>([&](){ return pos.see_ge(*cur, threshold); });
+      return select<Next>([&](){ return pos.see_ge(*cur, occupied, threshold); });
 
   case QCAPTURE:
       if (select<Next>([&](){ return   depth > DEPTH_QS_RECAPTURES
diff --git a/src/movepick.h b/src/movepick.h
index b6c07378..725607b8 100644
--- a/src/movepick.h
+++ b/src/movepick.h
@@ -150,6 +150,7 @@ private:
   Value threshold;
   Depth depth;
   ExtMove moves[MAX_MOVES];
+  Bitboard occupied;
 };
 
 } // namespace Stockfish
diff --git a/src/position.cpp b/src/position.cpp
index 171193ec..ba6888eb 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -1062,7 +1062,7 @@ Key Position::key_after(Move m) const {
 /// SEE value of move is greater or equal to the given threshold. We'll use an
 /// algorithm similar to alpha-beta pruning with a null window.
 
-bool Position::see_ge(Move m, Value threshold) const {
+bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
 
   assert(is_ok(m));
 
@@ -1081,7 +1081,7 @@ bool Position::see_ge(Move m, Value threshold) const {
       return true;
 
   assert(color_of(piece_on(from)) == sideToMove);
-  Bitboard occupied = pieces() ^ from ^ to;
+  occupied = pieces() ^ from ^ to; // xoring to is important for pinned piece logic
   Color stm = sideToMove;
   Bitboard attackers = attackers_to(to, occupied);
   Bitboard stmAttackers, bb;
@@ -1112,45 +1112,44 @@ bool Position::see_ge(Move m, Value threshold) const {
       // the bitboard 'attackers' any X-ray attackers behind it.
       if ((bb = stmAttackers & pieces(PAWN)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = PawnValueMg - swap) < res)
               break;
 
-          occupied ^= least_significant_square_bb(bb);
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(KNIGHT)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = KnightValueMg - swap) < res)
               break;
-
-          occupied ^= least_significant_square_bb(bb);
       }
 
       else if ((bb = stmAttackers & pieces(BISHOP)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = BishopValueMg - swap) < res)
               break;
 
-          occupied ^= least_significant_square_bb(bb);
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(ROOK)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = RookValueMg - swap) < res)
               break;
 
-          occupied ^= least_significant_square_bb(bb);
           attackers |= attacks_bb<ROOK>(to, occupied) & pieces(ROOK, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(QUEEN)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = QueenValueMg - swap) < res)
               break;
 
-          occupied ^= least_significant_square_bb(bb);
           attackers |=  (attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN))
                       | (attacks_bb<ROOK  >(to, occupied) & pieces(ROOK  , QUEEN));
       }
diff --git a/src/position.h b/src/position.h
index d3483bcf..670b621c 100644
--- a/src/position.h
+++ b/src/position.h
@@ -144,7 +144,7 @@ public:
   void undo_null_move();
 
   // Static Exchange Evaluation
-  bool see_ge(Move m, Value threshold = VALUE_ZERO) const;
+  bool see_ge(Move m, Bitboard& occupied, Value threshold = VALUE_ZERO) const;
 
   // Accessing hash keys
   Key key() const;
diff --git a/src/search.cpp b/src/search.cpp
index 17c1c28b..7564c109 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1019,9 +1019,27 @@ moves_loop: // When in check, search starts here
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
+              Bitboard occupied;
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, Value(-206) * depth))
-                  continue;
+              if (!pos.see_ge(move, occupied, Value(-206) * depth))
+              {
+                  if (depth < 2 - capture)
+                      continue;
+                  // don't prune move if a heavy enemy piece (KQR) is under attack after the exchanges
+                  Bitboard leftEnemies = (pos.pieces(~us, QUEEN, ROOK) | pos.pieces(~us, KING)) & occupied;
+                  Bitboard attacks = 0;
+                  occupied |= to_sq(move);
+                  while (leftEnemies && !attacks)
+                  {
+                      Square sq = pop_lsb(leftEnemies);
+                      attacks |= pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
+                      // exclude Queen/Rook(s) which were already threatened before SEE
+                      if (attacks && (sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us))))
+                          attacks = 0;
+                  }
+                  if (!attacks)
+                      continue;
+              }
           }
           else
           {
@@ -1047,8 +1065,9 @@ moves_loop: // When in check, search starts here
 
               lmrDepth = std::max(lmrDepth, 0);
 
+              Bitboard occupied;
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
+              if (!pos.see_ge(move, occupied, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
                   continue;
           }
       }
@@ -1533,6 +1552,7 @@ moves_loop: // When in check, search starts here
                                       prevSq);
 
     int quietCheckEvasions = 0;
+    Bitboard occupied;
 
     // Step 5. Loop through all pseudo-legal moves until no moves remain
     // or a beta cutoff occurs.
@@ -1569,7 +1589,7 @@ moves_loop: // When in check, search starts here
               continue;
           }
 
-          if (futilityBase <= alpha && !pos.see_ge(move, VALUE_ZERO + 1))
+          if (futilityBase <= alpha && !pos.see_ge(move, occupied, VALUE_ZERO + 1))
           {
               bestValue = std::max(bestValue, futilityBase);
               continue;
@@ -1588,7 +1608,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, Value(-110)))
+      if (!pos.see_ge(move, occupied, Value(-110)))
           continue;
     }
 

From b973e40e45d75c3b3391141d149d4186494d4652 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Tue, 21 Mar 2023 23:58:25 +0300
Subject: [PATCH 100/192] Update Elo estimates for terms in search

Setting the Elo value of some functions which were not set before.
All tests run at 10+0.1 (STC), 25000 games (Same as #4294).
Book used: UHO_XXL_+0.90_+1.19.epd

Values are rounded to the nearest non-negative integer.

Test links:
https://tests.stockfishchess.org/tests/view/6419ab5b65775d3b539f46c6
https://tests.stockfishchess.org/tests/view/6419adb465775d3b539f4730
https://tests.stockfishchess.org/tests/view/6419ae9c65775d3b539f4756
https://tests.stockfishchess.org/tests/view/6419b03f65775d3b539f47a8
https://tests.stockfishchess.org/tests/view/6419b35d65775d3b539f4860
https://tests.stockfishchess.org/tests/view/6419b6b965775d3b539f48e6
https://tests.stockfishchess.org/tests/view/6419cade65775d3b539f4cd5
https://tests.stockfishchess.org/tests/view/6419cbb565775d3b539f4d01
https://tests.stockfishchess.org/tests/view/6419cc6965775d3b539f4d1e

closes https://github.com/official-stockfish/Stockfish/pull/4459

No functional change
---
 src/search.cpp | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 7564c109..b2983f66 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -732,7 +732,7 @@ namespace {
     }
     else if (excludedMove)
     {
-        // Providing the hint that this node's accumulator will be used often brings significant Elo gain (13 elo)
+        // Providing the hint that this node's accumulator will be used often brings significant Elo gain (13 Elo)
         Eval::NNUE::hint_common_parent_position(pos);
         eval = ss->staticEval;
         complexity = abs(ss->staticEval - pos.psq_eg_stm());
@@ -1120,15 +1120,15 @@ moves_loop: // When in check, search starts here
               else if (singularBeta >= beta)
                   return singularBeta;
 
-              // If the eval of ttMove is greater than beta, we reduce it (negative extension)
+              // If the eval of ttMove is greater than beta, we reduce it (negative extension) (~7 Elo)
               else if (ttValue >= beta)
                   extension = -2 - !PvNode;
 
-              // If the eval of ttMove is less than value, we reduce it (negative extension)
+              // If the eval of ttMove is less than value, we reduce it (negative extension) (~1 Elo)
               else if (ttValue <= value)
                   extension = -1;
 
-              // If the eval of ttMove is less than alpha, we reduce it (negative extension)
+              // If the eval of ttMove is less than alpha, we reduce it (negative extension) (~1 Elo)
               else if (ttValue <= alpha)
                   extension = -1;
           }
@@ -1182,7 +1182,7 @@ moves_loop: // When in check, search starts here
       if (ttCapture)
           r++;
 
-      // Decrease reduction for PvNodes based on depth
+      // Decrease reduction for PvNodes based on depth (~2 Elo)
       if (PvNode)
           r -= 1 + 12 / (3 + depth);
 
@@ -1195,11 +1195,11 @@ moves_loop: // When in check, search starts here
           && (mp.threatenedPieces & from_sq(move)))
           r--;
 
-      // Increase reduction if next ply has a lot of fail high
+      // Increase reduction if next ply has a lot of fail high (~5 Elo)
       if ((ss+1)->cutoffCnt > 3)
           r++;
 
-      // Decrease reduction if move is a killer and we have a good history
+      // Decrease reduction if move is a killer and we have a good history (~1 Elo)
       if (move == ss->killers[0]
           && (*contHist[0])[movedPiece][to_sq(move)] >= 3722)
           r--;
@@ -1210,7 +1210,7 @@ moves_loop: // When in check, search starts here
                      + (*contHist[3])[movedPiece][to_sq(move)]
                      - 4182;
 
-      // Decrease/increase reduction for moves with a good/bad history (~30 Elo)
+      // Decrease/increase reduction for moves with a good/bad history (~25 Elo)
       r -= ss->statScore / (11791 + 3992 * (depth > 6 && depth < 19));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
@@ -1347,7 +1347,7 @@ moves_loop: // When in check, search starts here
               {
                   alpha = value;
 
-                  // Reduce other moves if we have found at least one score improvement
+                  // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
                       && depth < 6
                       && beta  <  10534
@@ -1413,7 +1413,7 @@ moves_loop: // When in check, search starts here
         bestValue = std::min(bestValue, maxValue);
 
     // If no good move is found and the previous position was ttPv, then the previous
-    // opponent move is probably good and the new position is added to the search tree.
+    // opponent move is probably good and the new position is added to the search tree. (~7 Elo)
     if (bestValue <= alpha)
         ss->ttPv = ss->ttPv || ((ss-1)->ttPv && depth > 3);
 
@@ -1432,7 +1432,7 @@ moves_loop: // When in check, search starts here
 
   // qsearch() is the quiescence search function, which is called by the main search
   // function with zero depth, or recursively with further decreasing depth per call.
-  // (~155 elo)
+  // (~155 Elo)
   template <NodeType nodeType>
   Value qsearch(Position& pos, Stack* ss, Value alpha, Value beta, Depth depth) {
 

From 1b5738e0c958ac8d3d140a3d182b85f8c0c0cd2c Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 24 Mar 2023 07:45:22 +0300
Subject: [PATCH 101/192] Simplify statScore initialization

This patch simplifies initialization of statScore to "always set it up to 0" instead of setting it up to 0 two plies deeper.
Reason for why it was done in previous way partially was because of LMR usage of previous statScore which was simplified long time ago so it makes sense to make in more simple there.

Passed STC:
https://tests.stockfishchess.org/tests/view/641a86d1db43ab2ba6f7b31d
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 115648 W: 30895 L: 30764 D: 53989
Ptnml(0-2): 368, 12741, 31473, 12876, 366

Passed LTC:
https://tests.stockfishchess.org/tests/view/641b1c31db43ab2ba6f7d17a
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 175576 W: 47122 L: 47062 D: 81392
Ptnml(0-2): 91, 17077, 53390, 17141, 89

closes https://github.com/official-stockfish/Stockfish/pull/4460

bench 5081969
---
 src/search.cpp | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index b2983f66..d2358ea2 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -607,14 +607,7 @@ namespace {
     (ss+2)->cutoffCnt    = 0;
     ss->doubleExtensions = (ss-1)->doubleExtensions;
     Square prevSq        = is_ok((ss-1)->currentMove) ? to_sq((ss-1)->currentMove) : SQ_NONE;
-
-    // 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.
-    if (!rootNode)
-        (ss+2)->statScore = 0;
+    ss->statScore        = 0;
 
     // Step 4. Transposition table lookup.
     excludedMove = ss->excludedMove;

From 587bc647d7d14b53d8625c4446006e23a4acd82a Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Wed, 22 Mar 2023 16:50:55 +0900
Subject: [PATCH 102/192] Remove non_pawn_material in NNUE::evaluate

After "Use NNUE complexity in search, retune related parameters" commit,
the effect of non-pawn material adjustment has been nearly diminished.
This patch removes pos.non_pawn_material as a simplification, which
passed non-regression tests with both STC and LTC.

Passed non-regression STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 75152 W: 20030 L: 19856 D: 35266
Ptnml(0-2): 215, 8281, 20459, 8357, 264
https://tests.stockfishchess.org/tests/view/641ab471db43ab2ba6f7bc58

Passed non-regression LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 193864 W: 51870 L: 51829 D: 90165
Ptnml(0-2): 86, 18968, 58794, 18987, 97
https://tests.stockfishchess.org/tests/view/641b4fe6db43ab2ba6f7db96

closes https://github.com/official-stockfish/Stockfish/pull/4461

Bench: 5020718
---
 src/nnue/evaluate_nnue.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/nnue/evaluate_nnue.cpp b/src/nnue/evaluate_nnue.cpp
index f33aa3b8..329adfda 100644
--- a/src/nnue/evaluate_nnue.cpp
+++ b/src/nnue/evaluate_nnue.cpp
@@ -148,7 +148,7 @@ namespace Stockfish::Eval::NNUE {
     // overaligning stack variables with alignas() doesn't work correctly.
 
     constexpr uint64_t alignment = CacheLineSize;
-    int delta = 24 - pos.non_pawn_material() / 9560;
+    constexpr int delta = 24;
 
 #if defined(ALIGNAS_ON_STACK_VARIABLES_BROKEN)
     TransformedFeatureType transformedFeaturesUnaligned[

From 43108a619899af084e45224e8744ca668a9efed2 Mon Sep 17 00:00:00 2001
From: Sebastian Buchwald <UniQP@web.de>
Date: Tue, 28 Mar 2023 19:53:43 +0200
Subject: [PATCH 103/192] Reuse existing functions to read/write array of
 network parameters

closes https://github.com/official-stockfish/Stockfish/pull/4463

No functional change
---
 src/nnue/layers/affine_transform.h | 12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)

diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index 313b1568..f84f054e 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -256,8 +256,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     // Read network parameters
     bool read_parameters(std::istream& stream) {
-      for (IndexType i = 0; i < OutputDimensions; ++i)
-        biases[i] = read_little_endian<BiasType>(stream);
+      read_little_endian<BiasType>(stream, biases, OutputDimensions);
 
       for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
         weights[get_weight_index(i)] = read_little_endian<WeightType>(stream);
@@ -267,8 +266,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     // Write network parameters
     bool write_parameters(std::ostream& stream) const {
-      for (IndexType i = 0; i < OutputDimensions; ++i)
-          write_little_endian<BiasType>(stream, biases[i]);
+      write_little_endian<BiasType>(stream, biases, OutputDimensions);
 
       for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
         write_little_endian<WeightType>(stream, weights[get_weight_index(i)]);
@@ -452,8 +450,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     // Read network parameters
     bool read_parameters(std::istream& stream) {
-      for (IndexType i = 0; i < OutputDimensions; ++i)
-        biases[i] = read_little_endian<BiasType>(stream);
+      read_little_endian<BiasType>(stream, biases, OutputDimensions);
       for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
         weights[get_weight_index(i)] = read_little_endian<WeightType>(stream);
 
@@ -462,8 +459,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     // Write network parameters
     bool write_parameters(std::ostream& stream) const {
-      for (IndexType i = 0; i < OutputDimensions; ++i)
-        write_little_endian<BiasType>(stream, biases[i]);
+      write_little_endian<BiasType>(stream, biases, OutputDimensions);
 
       for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
         write_little_endian<WeightType>(stream, weights[get_weight_index(i)]);

From 37160c4b1632245d46d86cec7bd22b76f5a87531 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Mon, 27 Mar 2023 01:00:54 -0400
Subject: [PATCH 104/192] Update default net to nn-dabb1ed23026.nnue

Created by retraining the master net with these modifications:

* New filtering methods for existing data from T80 sep+oct2022, T79 apr2022, T78 jun+jul+aug+sep2022, T77 dec2021
* Adding new filtered data from T80 aug2022 and T78 apr+may2022
* Increasing early-fen-skipping from 28 to 30

```
python3 easy_train.py \
  --experiment-name leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3-sk30 \
  --training-dataset /data/leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes \
  --start-from-engine-test-net True \
  --early-fen-skipping 30 \
  --max_epoch 900 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --tui False \
  --gpus "0," \
  --seed $RANDOM
```

The v3 filtering used for data from T77dec 2021 differs from v2 filtering in that:

* To improve binpack compression, positions after ply 28 were skipped during training by setting position scores to VALUE_NONE (32002) instead of removing them entirely
* All early-game positions with ply <= 28 were removed to maximize binpack compression
* Only bestmove captures at d6pv2 search were skipped, not 2nd bestmove captures
* Binpack compression was repaired for the remaining positions by effectively replacing bestmoves with "played moves" to maintain contiguous sequences of positions in the training game data

After improving binpack compression, The T77 dec2021 data size was reduced from 95G to 19G.

The v6 filtering used for data from T80augsepoctT79aprT78aprtosep 2022 differs from v2 in that:

* All positions with only one legal move were removed
* Tighter score differences at d6pv2 search were used to remove more positions with only one good move than before
* d6pv2 search was not used to remove positions where the best 2 moves were captures

```
python3 interleave_binpacks.py \
  nn-547-dataset/leela96-eval-filt-v2.binpack \
  nn-547-dataset/dfrc99-eval-filt-v2.binpack \
  nn-547-dataset/test80-nov2022-12tb7p-eval-filt-v2-d6.binpack \
  nn-547-dataset/T79-may2022-12tb7p-eval-filt-v2.binpack \
  nn-547-dataset/T60-nov2021-12tb7p-eval-filt-v2.binpack \
  nn-547-dataset/T60-dec2021-12tb7p-eval-filt-v2.binpack \
  filt-v6/test80-aug2022-16tb7p-filter-v6.binpack \
  filt-v6/test80-sep2022-16tb7p-filter-v6.binpack \
  filt-v6/test80-oct2022-16tb7p-filter-v6.binpack \
  filt-v6/test79-apr2022-16tb7p-filter-v6.binpack \
  filt-v6/test78-aprmay2022-16tb7p-filter-v6.binpack \
  filt-v6/test78-junjulaug2022-16tb7p-filter-v6.binpack \
  filt-v6/test78-sep2022-16tb7p-filter-v6.binpack \
  filt-v3/test77-dec2021-16tb7p-filt-v3.binpack \
  /data/leela96-dfrc99-T80novT79mayT60novdec-v2-T80augsepoctT79aprT78aprtosep-v6-T77dec-v3.binpack
```

The code for the new data filtering methods is available at:
https://github.com/linrock/Stockfish/tree/nnue-data-v3/nnue-data

The code for giving hexword names to .nnue files is at:
https://github.com/linrock/nnue-namer

Links for downloading the training data components can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch779.nnue : 0.6 +/- 3.1

Passed STC:
https://tests.stockfishchess.org/tests/view/64212412db43ab2ba6f8efb0
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 82256 W: 22185 L: 21809 D: 38262
Ptnml(0-2): 286, 9065, 22067, 9407, 303

Passed LTC:
https://tests.stockfishchess.org/tests/view/64223726db43ab2ba6f91d6c
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 30840 W: 8437 L: 8149 D: 14254
Ptnml(0-2): 14, 2891, 9323, 3177, 15

closes https://github.com/official-stockfish/Stockfish/pull/4465

bench 5101970
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 3615fe6d..61846073 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-52471d67216a.nnue"
+  #define EvalFileDefaultName   "nn-dabb1ed23026.nnue"
 
   namespace NNUE {
 

From a9c26357deb01c764cd16ef4e61acb4f687cbd77 Mon Sep 17 00:00:00 2001
From: Miguel Lahoz <miguel_lahoz@protonmail.com>
Date: Tue, 28 Mar 2023 00:06:24 +0800
Subject: [PATCH 105/192] Clean up repetitive declarations for see_ge

The occupied bitboard is only used in one place and is otherwise thrown away.
To simplify use, see_ge function can instead be overloaded.
Repetitive declarations for occupied bitboard can be removed.

Passed non-regression test
https://tests.stockfishchess.org/tests/view/6421c286db43ab2ba6f908eb
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 48912 W: 13196 L: 13001 D: 22715
Ptnml(0-2): 146, 5003, 13967, 5190, 150

closes https://github.com/official-stockfish/Stockfish/pull/4469

No functional change.
---
 src/movepick.cpp | 6 +++---
 src/movepick.h   | 1 -
 src/position.cpp | 5 +++++
 src/position.h   | 1 +
 src/search.cpp   | 8 +++-----
 5 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/src/movepick.cpp b/src/movepick.cpp
index 855f2b1d..36ee46b5 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -95,7 +95,7 @@ MovePicker::MovePicker(const Position& p, Move ttm, Value th, const CapturePiece
 
   stage = PROBCUT_TT + !(ttm && pos.capture_stage(ttm)
                              && pos.pseudo_legal(ttm)
-                             && pos.see_ge(ttm, occupied, threshold));
+                             && pos.see_ge(ttm, threshold));
 }
 
 /// MovePicker::score() assigns a numerical value to each move in a list, used
@@ -197,7 +197,7 @@ top:
 
   case GOOD_CAPTURE:
       if (select<Next>([&](){
-                       return pos.see_ge(*cur, occupied, Value(-cur->value)) ?
+                       return pos.see_ge(*cur, Value(-cur->value)) ?
                               // Move losing capture to endBadCaptures to be tried later
                               true : (*endBadCaptures++ = *cur, false); }))
           return *(cur - 1);
@@ -264,7 +264,7 @@ top:
       return select<Best>([](){ return true; });
 
   case PROBCUT:
-      return select<Next>([&](){ return pos.see_ge(*cur, occupied, threshold); });
+      return select<Next>([&](){ return pos.see_ge(*cur, threshold); });
 
   case QCAPTURE:
       if (select<Next>([&](){ return   depth > DEPTH_QS_RECAPTURES
diff --git a/src/movepick.h b/src/movepick.h
index 725607b8..b6c07378 100644
--- a/src/movepick.h
+++ b/src/movepick.h
@@ -150,7 +150,6 @@ private:
   Value threshold;
   Depth depth;
   ExtMove moves[MAX_MOVES];
-  Bitboard occupied;
 };
 
 } // namespace Stockfish
diff --git a/src/position.cpp b/src/position.cpp
index ba6888eb..e6fdb511 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -1163,6 +1163,11 @@ bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
   return bool(res);
 }
 
+bool Position::see_ge(Move m, Value threshold) const {
+    Bitboard occupied;
+    return see_ge(m, occupied, threshold);
+}
+
 
 /// Position::is_draw() tests whether the position is drawn by 50-move rule
 /// or by repetition. It does not detect stalemates.
diff --git a/src/position.h b/src/position.h
index 670b621c..bb45c44a 100644
--- a/src/position.h
+++ b/src/position.h
@@ -145,6 +145,7 @@ public:
 
   // Static Exchange Evaluation
   bool see_ge(Move m, Bitboard& occupied, Value threshold = VALUE_ZERO) const;
+  bool see_ge(Move m, Value threshold = VALUE_ZERO) const;
 
   // Accessing hash keys
   Key key() const;
diff --git a/src/search.cpp b/src/search.cpp
index d2358ea2..ac74cdaf 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1058,9 +1058,8 @@ moves_loop: // When in check, search starts here
 
               lmrDepth = std::max(lmrDepth, 0);
 
-              Bitboard occupied;
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, occupied, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
+              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
                   continue;
           }
       }
@@ -1545,7 +1544,6 @@ moves_loop: // When in check, search starts here
                                       prevSq);
 
     int quietCheckEvasions = 0;
-    Bitboard occupied;
 
     // Step 5. Loop through all pseudo-legal moves until no moves remain
     // or a beta cutoff occurs.
@@ -1582,7 +1580,7 @@ moves_loop: // When in check, search starts here
               continue;
           }
 
-          if (futilityBase <= alpha && !pos.see_ge(move, occupied, VALUE_ZERO + 1))
+          if (futilityBase <= alpha && !pos.see_ge(move, VALUE_ZERO + 1))
           {
               bestValue = std::max(bestValue, futilityBase);
               continue;
@@ -1601,7 +1599,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, occupied, Value(-110)))
+      if (!pos.see_ge(move, Value(-110)))
           continue;
     }
 

From 3f01e3f41f11aa66befec2307a32ee023c699a2a Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Thu, 23 Mar 2023 13:35:34 +0300
Subject: [PATCH 106/192] Allow PvNode in futility pruning for captures.

Passed non-regression STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 148128 W: 39428 L: 39333 D: 69367
Ptnml(0-2): 492, 16326, 40315, 16457, 474
https://tests.stockfishchess.org/tests/view/641c2dbfdb43ab2ba6f804e8

Passed non-regression LTC:
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 376256 W: 100906 L: 101039 D: 174311
Ptnml(0-2): 186, 36697, 114494, 36566, 185
https://tests.stockfishchess.org/tests/view/641d33b2db43ab2ba6f83338

closes https://github.com/official-stockfish/Stockfish/pull/4470

bench: 4935616
---
 src/search.cpp | 1 -
 1 file changed, 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index ac74cdaf..5f95a1bd 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1005,7 +1005,6 @@ moves_loop: // When in check, search starts here
           {
               // Futility pruning for captures (~2 Elo)
               if (   !givesCheck
-                  && !PvNode
                   && lmrDepth < 6
                   && !ss->inCheck
                   && ss->staticEval + 182 + 230 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]

From 7a6fa34f5f9f0f193d9350cd58c82a8f98a6505d Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Sun, 12 Mar 2023 15:16:51 +0200
Subject: [PATCH 107/192] Improve compatibility

this makes it easier to compile under MSVC, even though we recommend gcc/clang for production compiles at the moment.

In Win32 API, by default, most null-terminated character strings arguments are of wchar_t (UTF16, formerly UCS16-LE) type, i.e. 2 bytes (at least) per character. So, src/misc.cpp should have proper type. Respectively, for src/syzygy/tbprobe.cpp, in Widows, file paths should be std::wstring rather than std::string. However, this requires a very big number of changes, since the config files are also keeping the 8-bit-per-character std::string strings. Therefore, just one change of using 8-byte-per-character CreateFileA make it compile under MSVC.

closes https://github.com/official-stockfish/Stockfish/pull/4438

No functional change
---
 src/misc.cpp           | 4 ++--
 src/syzygy/tbprobe.cpp | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index c22126af..6469c5cf 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -605,7 +605,7 @@ static int best_node(size_t idx) {
   DWORD byteOffset = 0;
 
   // Early exit if the needed API is not available at runtime
-  HMODULE k32 = GetModuleHandle("Kernel32.dll");
+  HMODULE k32 = GetModuleHandle(TEXT("Kernel32.dll"));
   auto fun1 = (fun1_t)(void(*)())GetProcAddress(k32, "GetLogicalProcessorInformationEx");
   if (!fun1)
       return -1;
@@ -675,7 +675,7 @@ void bindThisThread(size_t idx) {
       return;
 
   // Early exit if the needed API are not available at runtime
-  HMODULE k32 = GetModuleHandle("Kernel32.dll");
+  HMODULE k32 = GetModuleHandle(TEXT("Kernel32.dll"));
   auto fun2 = (fun2_t)(void(*)())GetProcAddress(k32, "GetNumaNodeProcessorMaskEx");
   auto fun3 = (fun3_t)(void(*)())GetProcAddress(k32, "SetThreadGroupAffinity");
   auto fun4 = (fun4_t)(void(*)())GetProcAddress(k32, "GetNumaNodeProcessorMask2");
diff --git a/src/syzygy/tbprobe.cpp b/src/syzygy/tbprobe.cpp
index b594ac37..9cb0bfdb 100644
--- a/src/syzygy/tbprobe.cpp
+++ b/src/syzygy/tbprobe.cpp
@@ -234,7 +234,7 @@ public:
         }
 #else
         // Note FILE_FLAG_RANDOM_ACCESS is only a hint to Windows and as such may get ignored.
-        HANDLE fd = CreateFile(fname.c_str(), GENERIC_READ, FILE_SHARE_READ, nullptr,
+        HANDLE fd = CreateFileA(fname.c_str(), GENERIC_READ, FILE_SHARE_READ, nullptr,
                                OPEN_EXISTING, FILE_FLAG_RANDOM_ACCESS, nullptr);
 
         if (fd == INVALID_HANDLE_VALUE)

From e8742bdab35c63253b5110d1861f27337e18f9fc Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Wed, 29 Mar 2023 12:43:36 +0300
Subject: [PATCH 108/192] Made advanced Windows API calls dynamically linked

Made advanced Windows API calls (those from Advapi32.dll) dynamically
linked to avoid link errors when compiling using
Intel icx compiler for Windows.

https://github.com/official-stockfish/Stockfish/pull/4467

No functional change
---
 src/misc.cpp | 28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index 6469c5cf..cac9dd94 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -38,6 +38,9 @@ using fun2_t = bool(*)(USHORT, PGROUP_AFFINITY);
 using fun3_t = bool(*)(HANDLE, CONST GROUP_AFFINITY*, PGROUP_AFFINITY);
 using fun4_t = bool(*)(USHORT, PGROUP_AFFINITY, USHORT, PUSHORT);
 using fun5_t = WORD(*)();
+using fun6_t = bool(*)(HANDLE, DWORD, PHANDLE);
+using fun7_t = bool(*)(LPCSTR, LPCSTR, PLUID);
+using fun8_t = bool(*)(HANDLE, BOOL, PTOKEN_PRIVILEGES, DWORD, PTOKEN_PRIVILEGES, PDWORD);
 }
 #endif
 
@@ -488,11 +491,26 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
   if (!largePageSize)
       return nullptr;
 
+  // Dynamically link OpenProcessToken, LookupPrivilegeValue and AdjustTokenPrivileges
+  HMODULE k32 = GetModuleHandle("Advapi32.dll");
+  auto fun6 = (fun6_t)(void(*)())GetProcAddress(k32, "OpenProcessToken");
+  if (!fun6)
+      return nullptr;
+  auto fun7 = (fun7_t)(void(*)())GetProcAddress(k32, "LookupPrivilegeValueA");
+  if (!fun7)
+      return nullptr;
+  auto fun8 = (fun8_t)(void(*)())GetProcAddress(k32, "AdjustTokenPrivileges");
+  if (!fun8)
+      return nullptr;
+            
+
   // We need SeLockMemoryPrivilege, so try to enable it for the process
-  if (!OpenProcessToken(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
+  // OpenProcessToken()
+  if (!fun6(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
       return nullptr;
 
-  if (LookupPrivilegeValue(nullptr, SE_LOCK_MEMORY_NAME, &luid))
+  // LookupPrivilegeValueA()
+  if (fun7(nullptr, SE_LOCK_MEMORY_NAME, &luid))
   {
       TOKEN_PRIVILEGES tp { };
       TOKEN_PRIVILEGES prevTp { };
@@ -504,7 +522,8 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
 
       // Try to enable SeLockMemoryPrivilege. Note that even if AdjustTokenPrivileges() succeeds,
       // we still need to query GetLastError() to ensure that the privileges were actually obtained.
-      if (AdjustTokenPrivileges(
+      // AdjustTokenPrivileges()
+      if (fun8(
               hProcessToken, FALSE, &tp, sizeof(TOKEN_PRIVILEGES), &prevTp, &prevTpLen) &&
           GetLastError() == ERROR_SUCCESS)
       {
@@ -514,7 +533,8 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
               nullptr, allocSize, MEM_RESERVE | MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
 
           // Privilege no longer needed, restore previous state
-          AdjustTokenPrivileges(hProcessToken, FALSE, &prevTp, 0, nullptr, nullptr);
+	  // AdjustTokenPrivileges ()
+          fun8(hProcessToken, FALSE, &prevTp, 0, nullptr, nullptr);
       }
   }
 

From c3c46feebba470dcbaa0a5a6ef83534091dffe6a Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Thu, 30 Mar 2023 17:46:15 +0800
Subject: [PATCH 109/192] Remove reduction for moving threatened piece

Simplify away "Decrease reduction if we move a threatened piece".

Running a dbg_hit_on() shows that this line is only called ~0.12% of the time.

Simplification STC: https://tests.stockfishchess.org/tests/view/641ec2dcdb43ab2ba6f88103
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 146128 W: 39168 L: 39070 D: 67890
Ptnml(0-2): 466, 16117, 39830, 16155, 496

Simplification LTC: https://tests.stockfishchess.org/tests/view/64200689db43ab2ba6f8bca8
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 248016 W: 66703 L: 66714 D: 114599
Ptnml(0-2): 105, 24202, 75406, 24189, 106

closes https://github.com/official-stockfish/Stockfish/pull/4471

Bench: 4961236
---
 src/search.cpp | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 5f95a1bd..2fcbc7df 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1181,11 +1181,6 @@ moves_loop: // When in check, search starts here
       if (singularQuietLMR)
           r--;
 
-      // Decrease reduction if we move a threatened piece (~1 Elo)
-      if (   depth > 9
-          && (mp.threatenedPieces & from_sq(move)))
-          r--;
-
       // Increase reduction if next ply has a lot of fail high (~5 Elo)
       if ((ss+1)->cutoffCnt > 3)
           r++;

From 66bf45b99e2061c1ba74f9975bc5059ac0121dfd Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sat, 1 Apr 2023 11:56:49 +0200
Subject: [PATCH 110/192] Stringify the git info passed

avoid escaping the string in the Makefile.

Alternative to https://github.com/official-stockfish/Stockfish/pull/4476

closes https://github.com/official-stockfish/Stockfish/pull/4481

No functional change.
---
 src/Makefile     | 4 ++--
 src/evaluate.cpp | 2 --
 src/misc.cpp     | 6 ++----
 src/misc.h       | 3 +++
 4 files changed, 7 insertions(+), 8 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index e257bc63..0b22fb4e 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -705,13 +705,13 @@ endif
 ### 3.7.1 Try to include git commit sha for versioning
 GIT_SHA = $(shell git rev-parse --short HEAD 2>/dev/null)
 ifneq ($(GIT_SHA), )
-	CXXFLAGS += -DGIT_SHA=\"$(GIT_SHA)\"
+	CXXFLAGS += -DGIT_SHA=$(GIT_SHA)
 endif
 
 ### 3.7.2 Try to include git commit date for versioning
 GIT_DATE = $(shell git show -s --date=format:'%Y%m%d' --format=%cd HEAD 2>/dev/null)
 ifneq ($(GIT_DATE), )
-	CXXFLAGS += -DGIT_DATE=\"$(GIT_DATE)\"
+	CXXFLAGS += -DGIT_DATE=$(GIT_DATE)
 endif
 
 ### 3.8 Link Time Optimization
diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 99b87300..12883fcc 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -82,8 +82,6 @@ namespace Eval {
         eval_file = EvalFileDefaultName;
 
     #if defined(DEFAULT_NNUE_DIRECTORY)
-    #define stringify2(x) #x
-    #define stringify(x) stringify2(x)
     vector<string> dirs = { "<internal>" , "" , CommandLine::binaryDirectory , stringify(DEFAULT_NNUE_DIRECTORY) };
     #else
     vector<string> dirs = { "<internal>" , "" , CommandLine::binaryDirectory };
diff --git a/src/misc.cpp b/src/misc.cpp
index cac9dd94..e36a04bc 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -160,7 +160,7 @@ string engine_info(bool to_uci) {
   {
       ss << "-";
       #ifdef GIT_DATE
-      ss << GIT_DATE;
+      ss << stringify(GIT_DATE);
       #else
       constexpr string_view months("Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec");
       string month, day, year;
@@ -173,7 +173,7 @@ string engine_info(bool to_uci) {
       ss << "-";
 
       #ifdef GIT_SHA
-      ss << GIT_SHA;
+      ss << stringify(GIT_SHA);
       #else
       ss << "nogit";
       #endif
@@ -190,8 +190,6 @@ string engine_info(bool to_uci) {
 
 std::string compiler_info() {
 
-  #define stringify2(x) #x
-  #define stringify(x) stringify2(x)
   #define make_version_string(major, minor, patch) stringify(major) "." stringify(minor) "." stringify(patch)
 
 /// Predefined macros hell:
diff --git a/src/misc.h b/src/misc.h
index c20a816e..d4965156 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -28,6 +28,9 @@
 
 #include "types.h"
 
+#define stringify2(x) #x
+#define stringify(x) stringify2(x)
+
 namespace Stockfish {
 
 std::string engine_info(bool to_uci = false);

From 38a80c0b47397dbdd9167ec1476dfd3c033020d6 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Thu, 30 Mar 2023 11:14:30 +0000
Subject: [PATCH 111/192] Simplify away complexityAverage

Instead of tracking the average of complexity values, calculate
complexity of root position at the beginning of the search and use it as
a scaling factor in time management.

Passed non-regression STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 58752 W: 15738 L: 15551 D: 27463
Ptnml(0-2): 164, 6194, 16478, 6371, 169
https://tests.stockfishchess.org/tests/view/6423010edb43ab2ba6f9424a

Passed non-regression LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 92872 W: 24865 L: 24729 D: 43278
Ptnml(0-2): 38, 8652, 28929, 8770, 47
https://tests.stockfishchess.org/tests/view/6423c1f0db43ab2ba6f9644f

closes https://github.com/official-stockfish/Stockfish/pull/4472

No functional change
---
 src/misc.h     | 26 --------------------------
 src/search.cpp | 17 ++++++++++-------
 src/thread.h   |  2 +-
 3 files changed, 11 insertions(+), 34 deletions(-)

diff --git a/src/misc.h b/src/misc.h
index d4965156..69d470c2 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -89,32 +89,6 @@ static inline const union { uint32_t i; char c[4]; } Le = { 0x01020304 };
 static inline const bool IsLittleEndian = (Le.c[0] == 4);
 
 
-// RunningAverage : a class to calculate a running average of a series of values.
-// For efficiency, all computations are done with integers.
-class RunningAverage {
-  public:
-
-      // Reset the running average to rational value p / q
-      void set(int64_t p, int64_t q)
-        { average = p * PERIOD * RESOLUTION / q; }
-
-      // Update average with value v
-      void update(int64_t v)
-        { average = RESOLUTION * v + (PERIOD - 1) * average / PERIOD; }
-
-      // Test if average is strictly greater than rational a / b
-      bool is_greater(int64_t a, int64_t b) const
-        { return b * average > a * (PERIOD * RESOLUTION); }
-
-      int64_t value() const
-        { return average / (PERIOD * RESOLUTION); }
-
-  private :
-      static constexpr int64_t PERIOD     = 4096;
-      static constexpr int64_t RESOLUTION = 1024;
-      int64_t average;
-};
-
 template <typename T, std::size_t MaxSize>
 class ValueList {
 
diff --git a/src/search.cpp b/src/search.cpp
index 2fcbc7df..3136046d 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -293,6 +293,15 @@ void Thread::search() {
 
   if (mainThread)
   {
+
+      int rootComplexity;
+      if (Eval::useNNUE)
+          Eval::NNUE::evaluate(rootPos, true, &rootComplexity);
+      else
+          Eval::evaluate(rootPos, &rootComplexity);
+
+      mainThread->complexity = std::min(1.03 + (rootComplexity - 241) / 1552.0, 1.45);
+
       if (mainThread->bestPreviousScore == VALUE_INFINITE)
           for (int i = 0; i < 4; ++i)
               mainThread->iterValue[i] = VALUE_ZERO;
@@ -311,8 +320,6 @@ void Thread::search() {
 
   multiPV = std::min(multiPV, rootMoves.size());
 
-  complexityAverage.set(153, 1);
-
   optimism[us] = optimism[~us] = VALUE_ZERO;
 
   int searchAgainCounter = 0;
@@ -472,10 +479,8 @@ void Thread::search() {
           timeReduction = lastBestMoveDepth + 8 < completedDepth ? 1.57 : 0.65;
           double reduction = (1.4 + mainThread->previousTimeReduction) / (2.08 * timeReduction);
           double bestMoveInstability = 1 + 1.8 * totBestMoveChanges / Threads.size();
-          int complexity = mainThread->complexityAverage.value();
-          double complexPosition = std::min(1.03 + (complexity - 241) / 1552.0, 1.45);
 
-          double totalTime = Time.optimum() * fallingEval * reduction * bestMoveInstability * complexPosition;
+          double totalTime = Time.optimum() * fallingEval * reduction * bestMoveInstability * mainThread->complexity;
 
           // Cap used time in case of a single legal move for a better viewer experience in tournaments
           // yielding correct scores and sufficiently fast moves.
@@ -755,8 +760,6 @@ namespace {
         tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
     }
 
-    thisThread->complexityAverage.update(complexity);
-
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
diff --git a/src/thread.h b/src/thread.h
index 46cdb11c..d6a48eca 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -60,7 +60,6 @@ public:
   Pawns::Table pawnsTable;
   Material::Table materialTable;
   size_t pvIdx, pvLast;
-  RunningAverage complexityAverage;
   std::atomic<uint64_t> nodes, tbHits, bestMoveChanges;
   int selDepth, nmpMinPly;
   Color nmpColor;
@@ -87,6 +86,7 @@ struct MainThread : public Thread {
   void search() override;
   void check_time();
 
+  double complexity;
   double previousTimeReduction;
   Value bestPreviousScore;
   Value bestPreviousAverageScore;

From bc50378ff1915a8ad6ac3e4946193c65e4cacb56 Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Fri, 31 Mar 2023 18:16:50 +0300
Subject: [PATCH 112/192] Replace deprecated icc with icx

Replace the deprecated Intel compiler icc with its newer icx variant.
This newer compiler is based on clang, and yields good performance.
As before, currently only linux is supported.

closes https://github.com/official-stockfish/Stockfish/pull/4478

No functional change
---
 src/Makefile | 70 +++++++++++++++++++++++++++++-----------------------
 1 file changed, 39 insertions(+), 31 deletions(-)

diff --git a/src/Makefile b/src/Makefile
index 0b22fb4e..abcf11b0 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -425,10 +425,11 @@ ifeq ($(COMP),mingw)
 	CXXFLAGS += -pedantic -Wextra -Wshadow -Wmissing-declarations
 endif
 
-ifeq ($(COMP),icc)
-	comp=icc
-	CXX=icpc
-	CXXFLAGS += -diag-disable 1476,10120 -Wcheck -Wabi -Wdeprecated -strict-ansi
+ifeq ($(COMP),icx)
+	comp=icx
+	CXX=icpx
+	CXXFLAGS += --intel -pedantic -Wextra -Wshadow -Wmissing-prototypes \
+		-Wconditional-uninitialized -Wabi -Wdeprecated
 endif
 
 ifeq ($(COMP),clang)
@@ -499,9 +500,9 @@ ifeq ($(COMP),ndk)
 	LDFLAGS += -static-libstdc++ -pie -lm -latomic
 endif
 
-ifeq ($(comp),icc)
-	profile_make = icc-profile-make
-	profile_use = icc-profile-use
+ifeq ($(comp),icx)
+	profile_make = icx-profile-make
+	profile_use = icx-profile-use
 else ifeq ($(comp),clang)
 	profile_make = clang-profile-make
 	profile_use = clang-profile-use
@@ -572,7 +573,7 @@ ifeq ($(optimize),yes)
 	endif
 
 	ifeq ($(KERNEL),Darwin)
-		ifeq ($(comp),$(filter $(comp),clang icc))
+		ifeq ($(comp),$(filter $(comp),clang icx))
 			CXXFLAGS += -mdynamic-no-pic
 		endif
 
@@ -608,8 +609,6 @@ endif
 ifeq ($(popcnt),yes)
 	ifeq ($(arch),$(filter $(arch),ppc64 armv7 armv8 arm64))
 		CXXFLAGS += -DUSE_POPCNT
-	else ifeq ($(comp),icc)
-		CXXFLAGS += -msse3 -DUSE_POPCNT
 	else
 		CXXFLAGS += -msse3 -mpopcnt -DUSE_POPCNT
 	endif
@@ -618,63 +617,63 @@ endif
 ### 3.6 SIMD architectures
 ifeq ($(avx2),yes)
 	CXXFLAGS += -DUSE_AVX2
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mavx2 -mbmi
 	endif
 endif
 
 ifeq ($(avxvnni),yes)
 	CXXFLAGS += -DUSE_VNNI -DUSE_AVXVNNI
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mavxvnni
 	endif
 endif
 
 ifeq ($(avx512),yes)
 	CXXFLAGS += -DUSE_AVX512
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mavx512f -mavx512bw
 	endif
 endif
 
 ifeq ($(vnni256),yes)
 	CXXFLAGS += -DUSE_VNNI
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -mprefer-vector-width=256
 	endif
 endif
 
 ifeq ($(vnni512),yes)
 	CXXFLAGS += -DUSE_VNNI
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
-		CXXFLAGS += -mavx512vnni -mavx512dq -mavx512vl
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
+		CXXFLAGS += -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -mprefer-vector-width=512
 	endif
 endif
 
 ifeq ($(sse41),yes)
 	CXXFLAGS += -DUSE_SSE41
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -msse4.1
 	endif
 endif
 
 ifeq ($(ssse3),yes)
 	CXXFLAGS += -DUSE_SSSE3
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mssse3
 	endif
 endif
 
 ifeq ($(sse2),yes)
 	CXXFLAGS += -DUSE_SSE2
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -msse2
 	endif
 endif
 
 ifeq ($(mmx),yes)
 	CXXFLAGS += -DUSE_MMX
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mmmx
 	endif
 endif
@@ -697,7 +696,7 @@ endif
 ### 3.7 pext
 ifeq ($(pext),yes)
 	CXXFLAGS += -DUSE_PEXT
-	ifeq ($(comp),$(filter $(comp),gcc clang mingw))
+	ifeq ($(comp),$(filter $(comp),gcc clang mingw icx))
 		CXXFLAGS += -mbmi2
 	endif
 endif
@@ -719,8 +718,11 @@ endif
 ### needs access to the optimization flags.
 ifeq ($(optimize),yes)
 ifeq ($(debug), no)
-	ifeq ($(comp),clang)
+	ifeq ($(comp),$(filter $(comp),clang icx))
 		CXXFLAGS += -flto=full
+		ifeq ($(comp),icx)
+			CXXFLAGS += -fwhole-program-vtables
+                endif
 		ifeq ($(target_windows),yes)
 			CXXFLAGS += -fuse-ld=lld
 		endif
@@ -807,7 +809,7 @@ help:
 	@echo "gcc                     > Gnu compiler (default)"
 	@echo "mingw                   > Gnu compiler with MinGW under Windows"
 	@echo "clang                   > LLVM Clang compiler"
-	@echo "icc                     > Intel compiler"
+	@echo "icx                     > Intel oneAPI DPC++/C++ Compiler"
 	@echo "ndk                     > Google NDK to cross-compile for Android"
 	@echo ""
 	@echo "Simple examples. If you don't know what to do, you likely want to run one of: "
@@ -833,8 +835,10 @@ endif
 
 
 .PHONY: help build profile-build strip install clean net objclean profileclean \
-        config-sanity icc-profile-use icc-profile-make gcc-profile-use gcc-profile-make \
-        clang-profile-use clang-profile-make FORCE
+	config-sanity \
+	icx-profile-use icx-profile-make \
+	gcc-profile-use gcc-profile-make \
+	clang-profile-use clang-profile-make FORCE
 
 build: net config-sanity
 	$(MAKE) ARCH=$(ARCH) COMP=$(COMP) all
@@ -949,7 +953,9 @@ config-sanity: net
 	@echo "vnni256: '$(vnni256)'"
 	@echo "vnni512: '$(vnni512)'"
 	@echo "neon: '$(neon)'"
+	@echo "dotprod: '$(dotprod)'"
 	@echo "arm_version: '$(arm_version)'"
+	@echo "target_windows: '$(target_windows)'"
 	@echo ""
 	@echo "Flags:"
 	@echo "CXX: $(CXX)"
@@ -978,7 +984,7 @@ config-sanity: net
 	@test "$(vnni256)" = "yes" || test "$(vnni256)" = "no"
 	@test "$(vnni512)" = "yes" || test "$(vnni512)" = "no"
 	@test "$(neon)" = "yes" || test "$(neon)" = "no"
-	@test "$(comp)" = "gcc" || test "$(comp)" = "icc" || test "$(comp)" = "mingw" || test "$(comp)" = "clang" \
+	@test "$(comp)" = "gcc" || test "$(comp)" = "icx" || test "$(comp)" = "mingw" || test "$(comp)" = "clang" \
 	|| test "$(comp)" = "armv7a-linux-androideabi16-clang"  || test "$(comp)" = "aarch64-linux-android21-clang"
 
 $(EXE): $(OBJS)
@@ -1016,15 +1022,17 @@ gcc-profile-use:
 	EXTRALDFLAGS='-lgcov' \
 	all
 
-icc-profile-make:
-	@mkdir -p profdir
+icx-profile-make:
 	$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
-	EXTRACXXFLAGS='-prof-gen=srcpos -prof_dir ./profdir' \
+	EXTRACXXFLAGS='-fprofile-instr-generate ' \
+	EXTRALDFLAGS=' -fprofile-instr-generate' \
 	all
 
-icc-profile-use:
+icx-profile-use:
+	$(XCRUN) llvm-profdata merge -output=stockfish.profdata *.profraw
 	$(MAKE) ARCH=$(ARCH) COMP=$(COMP) \
-	EXTRACXXFLAGS='-prof_use -prof_dir ./profdir' \
+	EXTRACXXFLAGS='-fprofile-instr-use=stockfish.profdata' \
+	EXTRALDFLAGS='-fprofile-use ' \
 	all
 
 .depend: $(SRCS)

From 6a6e32dfc80488dfdcd6c23e601063b47729e890 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Sat, 1 Apr 2023 15:22:53 +0300
Subject: [PATCH 113/192] Decrease Depth more for positions not in TT.

If the position is not in TT, decrease depth by 2
or by 4 if the TT entry for the current position was hit
and the stored depth is greater than or equal to the current depth.

Many thanks to Vizvezdenec as the main idea was his.

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 70664 W: 18995 L: 18639 D: 33030
Ptnml(0-2): 228, 7712, 19090, 8080, 222
https://tests.stockfishchess.org/tests/view/64258a8bdb43ab2ba6f9b682

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 85040 W: 23227 L: 22836 D: 38977
Ptnml(0-2): 26, 8115, 25867, 8466, 46
https://tests.stockfishchess.org/tests/view/64262057db43ab2ba6f9d0e7

closes https://github.com/official-stockfish/Stockfish/pull/4482

bench: 4380438
---
 src/search.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 3136046d..46cca50f 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -899,11 +899,11 @@ namespace {
         Eval::NNUE::hint_common_parent_position(pos);
     }
 
-    // Step 11. If the position is not in TT, decrease depth by 3.
+    // Step 11. If the position is not in TT, decrease depth by 2 (or by 4 if the TT entry for the current position was hit and the stored depth is greater than or equal to the current depth).
     // Use qsearch if depth is equal or below zero (~9 Elo)
     if (    PvNode
         && !ttMove)
-        depth -= 3;
+        depth -= 2 + 2 * (ss->ttHit &&  tte->depth() >= depth);
 
     if (depth <= 0)
         return qsearch<PV>(pos, ss, alpha, beta);

From 1fee996999364bedbd9ca4c29649d5c7321947c5 Mon Sep 17 00:00:00 2001
From: Miguel Lahoz <miguel_lahoz@protonmail.com>
Date: Sun, 2 Apr 2023 17:28:39 +0800
Subject: [PATCH 114/192] Remove unneeded bitboard from MP

Recent simplification has removed the need for an extra bitboard in MP struct.
Use a local variable instead.

STC: Passed Non-regression test
https://tests.stockfishchess.org/tests/view/64294ae677ff3301150cba16
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 64872 W: 17383 L: 17203 D: 30286
Ptnml(0-2): 179, 6675, 18546, 6859, 177

closes https://github.com/official-stockfish/Stockfish/pull/4490

No functional change.
---
 src/movepick.cpp | 3 +--
 src/movepick.h   | 2 --
 2 files changed, 1 insertion(+), 4 deletions(-)

diff --git a/src/movepick.cpp b/src/movepick.cpp
index 36ee46b5..6fbcb2c3 100644
--- a/src/movepick.cpp
+++ b/src/movepick.cpp
@@ -69,7 +69,6 @@ MovePicker::MovePicker(const Position& p, Move ttm, Depth d, const ButterflyHist
 
   stage = (pos.checkers() ? EVASION_TT : MAIN_TT) +
           !(ttm && pos.pseudo_legal(ttm));
-  threatenedPieces = 0;
 }
 
 /// MovePicker constructor for quiescence search
@@ -106,7 +105,7 @@ void MovePicker::score() {
 
   static_assert(Type == CAPTURES || Type == QUIETS || Type == EVASIONS, "Wrong type");
 
-  [[maybe_unused]] Bitboard threatenedByPawn, threatenedByMinor, threatenedByRook;
+  [[maybe_unused]] Bitboard threatenedByPawn, threatenedByMinor, threatenedByRook, threatenedPieces;
   if constexpr (Type == QUIETS)
   {
       Color us = pos.side_to_move();
diff --git a/src/movepick.h b/src/movepick.h
index b6c07378..0b44557f 100644
--- a/src/movepick.h
+++ b/src/movepick.h
@@ -131,8 +131,6 @@ public:
   MovePicker(const Position&, Move, Value, const CapturePieceToHistory*);
   Move next_move(bool skipQuiets = false);
 
-  Bitboard threatenedPieces;
-
 private:
   template<PickType T, typename Pred> Move select(Pred);
   template<GenType> void score();

From 77e2b915e1e4f2469a414712e52b469633fb3273 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Sat, 1 Apr 2023 18:48:31 -0500
Subject: [PATCH 115/192] Simplifiy TM's root complexity

Also requires moving optimism initialization, this is a very early `evaluate()` call.

STC: https://tests.stockfishchess.org/tests/view/6428c39677ff3301150ca0d7
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 51256 W: 13805 L: 13612 D: 23839
Ptnml(0-2): 145, 5283, 14592, 5450, 158

LTC: https://tests.stockfishchess.org/tests/view/64296ff377ff3301150cc519
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 106968 W: 28951 L: 28830 D: 49187
Ptnml(0-2): 47, 9746, 33789, 9843, 59

closes https://github.com/official-stockfish/Stockfish/pull/4492

no functional change
---
 src/search.cpp | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 46cca50f..becaee3a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -290,15 +290,13 @@ void Thread::search() {
 
   bestValue = delta = alpha = -VALUE_INFINITE;
   beta = VALUE_INFINITE;
+  optimism[us] = optimism[~us] = VALUE_ZERO;
 
   if (mainThread)
   {
 
       int rootComplexity;
-      if (Eval::useNNUE)
-          Eval::NNUE::evaluate(rootPos, true, &rootComplexity);
-      else
-          Eval::evaluate(rootPos, &rootComplexity);
+      Eval::evaluate(rootPos, &rootComplexity);
 
       mainThread->complexity = std::min(1.03 + (rootComplexity - 241) / 1552.0, 1.45);
 
@@ -320,8 +318,6 @@ void Thread::search() {
 
   multiPV = std::min(multiPV, rootMoves.size());
 
-  optimism[us] = optimism[~us] = VALUE_ZERO;
-
   int searchAgainCounter = 0;
 
   // Iterative deepening loop until requested to stop or the target depth is reached

From 9a42bbdf3163222db5e0fa764d48ca0a09a0dec2 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Tue, 4 Apr 2023 14:26:29 +0300
Subject: [PATCH 116/192] Parameters Tweak

Passed STC
LLR: 3.22 (-2.94,2.94) <0.00,2.00>
Total: 664048 W: 177526 L: 176301 D: 310221
Ptnml(0-2): 2002, 72968, 180891, 74129, 2034
https://tests.stockfishchess.org/tests/view/64219901db43ab2ba6f901fa

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 77576 W: 21125 L: 20750 D: 35701
Ptnml(0-2): 24, 7350, 23669, 7717, 28
https://tests.stockfishchess.org/tests/view/642abe3377ff3301150d3a16

closes https://github.com/official-stockfish/Stockfish/pull/4493

bench: 4522076
---
 src/search.cpp | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index becaee3a..749de792 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -72,7 +72,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1449 - int(delta) * 1032 / int(rootDelta)) / 1024 + (!i && r > 941);
+    return (r + 1449 - int(delta) * 1001 / int(rootDelta)) / 1024 + (!i && r > 941);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -82,7 +82,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min(340 * d - 470, 1855);
+    return std::min(341 * d - 470, 1855);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -1057,7 +1057,7 @@ moves_loop: // When in check, search starts here
               lmrDepth = std::max(lmrDepth, 0);
 
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 15 * lmrDepth)))
+              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 16 * lmrDepth)))
                   continue;
           }
       }
@@ -1193,10 +1193,10 @@ moves_loop: // When in check, search starts here
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]
                      + (*contHist[3])[movedPiece][to_sq(move)]
-                     - 4182;
+                     - 4082;
 
       // Decrease/increase reduction for moves with a good/bad history (~25 Elo)
-      r -= ss->statScore / (11791 + 3992 * (depth > 6 && depth < 19));
+      r -= ss->statScore / (11079 + 4626 * (depth > 6 && depth < 19));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has

From a2737d8bb5e480563823820fb12a8887d61c991e Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Wed, 5 Apr 2023 07:25:00 +0800
Subject: [PATCH 117/192] Simplify away piece count condition for useClassical

Simplify away the piece count condition for useClassical. In compensation, the psq requirement is increased by 15%.

Also updated the Elo estimate for useClassical, based on recent testing.

Simplification STC:
https://tests.stockfishchess.org/tests/view/642acbb577ff3301150d3ef5
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 51984 W: 13906 L: 13707 D: 24371
Ptnml(0-2): 150, 5638, 14227, 5817, 160

Simplification LTC:
https://tests.stockfishchess.org/tests/view/642b9c5777ff3301150d778a
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 119696 W: 32412 L: 32300 D: 54984
Ptnml(0-2): 53, 11529, 36567, 11651, 48

closes https://github.com/official-stockfish/Stockfish/pull/4494

Bench: 5089321
---
 src/evaluate.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 12883fcc..703cf869 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1053,8 +1053,8 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
 
   // We use the much less accurate but faster Classical eval when the NNUE
   // option is set to false. Otherwise we use the NNUE eval unless the
-  // PSQ advantage is decisive and several pieces remain. (~3 Elo)
-  bool useClassical = !useNNUE || (pos.count<ALL_PIECES>() > 7 && abs(psq) > 1781);
+  // PSQ advantage is decisive. (~4 Elo at STC, 1 Elo at LTC)
+  bool useClassical = !useNNUE || abs(psq) > 2048;
 
   if (useClassical)
       v = Evaluation<NO_TRACE>(pos).value();

From 510aca1ef62279dc35941fa45ed61fb9d3796f10 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Wed, 5 Apr 2023 05:48:34 +0300
Subject: [PATCH 118/192] Assign negative stat bonuses for quiet moves at Pv
 nodes

This patch assigns negative stat bonuses for quiet moves
at pv nodes which are searched at depth greater than
this node assumes, so are extended.

Passed STC:
https://tests.stockfishchess.org/tests/view/6426198bdb43ab2ba6f9cfa2
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 548944 W: 147287 L: 146254 D: 255403
Ptnml(0-2): 1662, 59772, 150671, 60605, 1762

Passed LTC:
https://tests.stockfishchess.org/tests/view/642be4f177ff3301150d892d
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 93352 W: 25400 L: 24994 D: 42958
Ptnml(0-2): 44, 8817, 28547, 9225, 43

closes https://github.com/official-stockfish/Stockfish/pull/4495

bench 5044536
---
 src/search.cpp | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 749de792..aa1a3e8c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -290,7 +290,7 @@ void Thread::search() {
 
   bestValue = delta = alpha = -VALUE_INFINITE;
   beta = VALUE_INFINITE;
-  optimism[us] = optimism[~us] = VALUE_ZERO;
+  optimism[WHITE] = optimism[BLACK] = VALUE_ZERO;
 
   if (mainThread)
   {
@@ -1257,6 +1257,9 @@ moves_loop: // When in check, search starts here
           (ss+1)->pv[0] = MOVE_NONE;
 
           value = -search<PV>(pos, ss+1, -beta, -alpha, newDepth, false);
+
+          if (moveCount > 1 && newDepth >= depth && !capture)
+              update_continuation_histories(ss, movedPiece, to_sq(move), -stat_bonus(newDepth));
       }
 
       // Step 19. Undo move

From 59f2085469a7dd96146905a5d8d0c1a5d987187d Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Thu, 6 Apr 2023 00:35:05 +0300
Subject: [PATCH 119/192] Depth Tweak and tuning

tunes reduction related parameters, and introduces more reduction on found good moves.

credit for this patch goes also to candirufish Yoshie2000 dubslow peregrineshahin Vizvezdenec

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 38424 W: 10346 L: 10040 D: 18038
Ptnml(0-2): 103, 4111, 10473, 4427, 98
https://tests.stockfishchess.org/tests/view/642ca74277ff3301150db511

Passed LTC:
LLR: 2.97 (-2.94,2.94) <0.50,2.50>
Total: 136968 W: 37151 L: 36660 D: 63157
Ptnml(0-2): 43, 13052, 41808, 13533, 48
https://tests.stockfishchess.org/tests/view/642d632377ff3301150dddbe

closes https://github.com/official-stockfish/Stockfish/pull/4499

bench: 3672914
---
 src/search.cpp | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index aa1a3e8c..fba9685b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -72,7 +72,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1449 - int(delta) * 1001 / int(rootDelta)) / 1024 + (!i && r > 941);
+    return (r + 1449 - int(delta) * 937 / int(rootDelta)) / 1024 + (!i && r > 941);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -82,7 +82,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min(341 * d - 470, 1855);
+    return std::min(341 * d - 470, 1710);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -775,7 +775,7 @@ namespace {
     // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
-    if (eval < alpha - 426 - 252 * depth * depth)
+    if (eval < alpha - 426 - 256 * depth * depth)
     {
         value = qsearch<NonPV>(pos, ss, alpha - 1, alpha);
         if (value < alpha)
@@ -797,7 +797,7 @@ namespace {
         && (ss-1)->statScore < 18755
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 19 * depth - improvement / 13 + 253 + complexity / 25
+        &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 253 + complexity / 25
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly || us != thisThread->nmpColor))
@@ -805,7 +805,7 @@ namespace {
         assert(eval - beta >= 0);
 
         // Null move dynamic reduction based on depth, eval and complexity of position
-        Depth R = std::min(int(eval - beta) / 168, 6) + depth / 3 + 4 - (complexity > 825);
+        Depth R = std::min(int(eval - beta) / 172, 6) + depth / 3 + 4 - (complexity > 825);
 
         ss->currentMove = MOVE_NULL;
         ss->continuationHistory = &thisThread->continuationHistory[0][0][NO_PIECE][0];
@@ -1333,16 +1333,18 @@ moves_loop: // When in check, search starts here
 
               if (PvNode && value < beta) // Update alpha! Always alpha < beta
               {
-                  alpha = value;
 
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
-                      && depth < 6
-                      && beta  <  10534
-                      && alpha > -10534)
-                      depth -= 1;
+                      && ((improving && complexity > 971) || (value < (5 * alpha + 75 * beta) / 87) || depth < 6)
+                      && beta  <  12535
+                      && value > -12535) {
+                      bool extraReduction = depth > 2 && alpha > -12535 && bestValue != -VALUE_INFINITE && (value - bestValue) > (7 * (beta - alpha)) / 8;
+                      depth -= 1 + extraReduction;
+                  }
 
                   assert(depth > 0);
+                  alpha = value;
               }
               else
               {

From b36d39de3d61b8f31c11d85233631aafaf760ee1 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sun, 9 Apr 2023 09:18:29 +0200
Subject: [PATCH 120/192] Fix rootComplexity calculation

The calculation of rootComplexity can't call eval when in check.
Doing so triggers an assert if compiled in debug mode when
the rootpos is evaluated using classical eval.

Fixes https://github.com/official-stockfish/Stockfish/issues/4512

Passed STC:
https://tests.stockfishchess.org/tests/view/6432697431feee5c6d306876
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 41096 W: 11017 L: 10815 D: 19264
Ptnml(0-2): 113, 4172, 11780, 4366, 117

Running LTC:
https://tests.stockfishchess.org/tests/view/6432974d31feee5c6d306fc0
LLR: 1.76 (-2.94,2.94) <-1.75,0.25>
Total: 73200 W: 19792 L: 19728 D: 33680
Ptnml(0-2): 24, 6659, 23182, 6699, 36

closes https://github.com/official-stockfish/Stockfish/pull/4515

No functional change
---
 src/evaluate.cpp |  2 ++
 src/search.cpp   | 10 ++++++----
 src/thread.cpp   |  1 +
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 703cf869..2d0df89c 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1048,6 +1048,8 @@ make_v:
 
 Value Eval::evaluate(const Position& pos, int* complexity) {
 
+  assert(!pos.checkers());
+
   Value v;
   Value psq = pos.psq_eg_stm();
 
diff --git a/src/search.cpp b/src/search.cpp
index fba9685b..5d54a15d 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -295,10 +295,12 @@ void Thread::search() {
   if (mainThread)
   {
 
-      int rootComplexity;
-      Eval::evaluate(rootPos, &rootComplexity);
-
-      mainThread->complexity = std::min(1.03 + (rootComplexity - 241) / 1552.0, 1.45);
+      if (!rootPos.checkers())
+      {
+          int rootComplexity;
+          Eval::evaluate(rootPos, &rootComplexity);
+          mainThread->complexity = std::min(1.03 + (rootComplexity - 241) / 1552.0, 1.45);
+      }
 
       if (mainThread->bestPreviousScore == VALUE_INFINITE)
           for (int i = 0; i < 4; ++i)
diff --git a/src/thread.cpp b/src/thread.cpp
index c680393e..202768c8 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -160,6 +160,7 @@ void ThreadPool::clear() {
   main()->bestPreviousScore = VALUE_INFINITE;
   main()->bestPreviousAverageScore = VALUE_INFINITE;
   main()->previousTimeReduction = 1.0;
+  main()->complexity = 1.0;
 }
 
 

From 5d258e168f7ea9019ed640ae2e56f04b26aea6a2 Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Sat, 1 Apr 2023 20:14:41 +0300
Subject: [PATCH 121/192] Fix linking / character types of windows API calls

ensures large pages can be allocated again.

closes https://github.com/official-stockfish/Stockfish/pull/4509

No functional change
---
 src/misc.cpp | 31 +++++++++++++++++--------------
 1 file changed, 17 insertions(+), 14 deletions(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index e36a04bc..f1554060 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -490,25 +490,29 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
       return nullptr;
 
   // Dynamically link OpenProcessToken, LookupPrivilegeValue and AdjustTokenPrivileges
-  HMODULE k32 = GetModuleHandle("Advapi32.dll");
-  auto fun6 = (fun6_t)(void(*)())GetProcAddress(k32, "OpenProcessToken");
+
+  HMODULE hAdvapi32 = GetModuleHandle(TEXT("advapi32.dll"));
+
+  if (!hAdvapi32)
+      hAdvapi32 = LoadLibrary(TEXT("advapi32.dll"));
+
+  auto fun6 = (fun6_t)(void(*)())GetProcAddress(hAdvapi32, "OpenProcessToken");
   if (!fun6)
       return nullptr;
-  auto fun7 = (fun7_t)(void(*)())GetProcAddress(k32, "LookupPrivilegeValueA");
+  auto fun7 = (fun7_t)(void(*)())GetProcAddress(hAdvapi32, "LookupPrivilegeValueA");
   if (!fun7)
       return nullptr;
-  auto fun8 = (fun8_t)(void(*)())GetProcAddress(k32, "AdjustTokenPrivileges");
+  auto fun8 = (fun8_t)(void(*)())GetProcAddress(hAdvapi32, "AdjustTokenPrivileges");
   if (!fun8)
       return nullptr;
-            
 
   // We need SeLockMemoryPrivilege, so try to enable it for the process
-  // OpenProcessToken()
-  if (!fun6(GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
-      return nullptr;
+  if (!fun6( // OpenProcessToken()
+      GetCurrentProcess(), TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY, &hProcessToken))
+          return nullptr;
 
-  // LookupPrivilegeValueA()
-  if (fun7(nullptr, SE_LOCK_MEMORY_NAME, &luid))
+  if (fun7( // LookupPrivilegeValue(nullptr, SE_LOCK_MEMORY_NAME, &luid)
+      nullptr, "SeLockMemoryPrivilege", &luid))
   {
       TOKEN_PRIVILEGES tp { };
       TOKEN_PRIVILEGES prevTp { };
@@ -520,8 +524,7 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
 
       // Try to enable SeLockMemoryPrivilege. Note that even if AdjustTokenPrivileges() succeeds,
       // we still need to query GetLastError() to ensure that the privileges were actually obtained.
-      // AdjustTokenPrivileges()
-      if (fun8(
+      if (fun8( // AdjustTokenPrivileges()
               hProcessToken, FALSE, &tp, sizeof(TOKEN_PRIVILEGES), &prevTp, &prevTpLen) &&
           GetLastError() == ERROR_SUCCESS)
       {
@@ -531,8 +534,8 @@ static void* aligned_large_pages_alloc_windows([[maybe_unused]] size_t allocSize
               nullptr, allocSize, MEM_RESERVE | MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE);
 
           // Privilege no longer needed, restore previous state
-	  // AdjustTokenPrivileges ()
-          fun8(hProcessToken, FALSE, &prevTp, 0, nullptr, nullptr);
+          fun8( // AdjustTokenPrivileges ()
+              hProcessToken, FALSE, &prevTp, 0, nullptr, nullptr);
       }
   }
 

From 6e63dd63a4f1e3074a9f5a8d7f64fdd0eba19c7e Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Fri, 7 Apr 2023 15:23:04 +0000
Subject: [PATCH 122/192] Use int conversion for Option class

The current implementation generates warnings on MSVC. However, we have
no real use cases for double-typed UCI option values now. Also parameter
tuning only accepts following three types:

  int, Value, Score

closes https://github.com/official-stockfish/Stockfish/pull/4505

No functional change
---
 src/tt.cpp        | 4 ++--
 src/uci.h         | 2 +-
 src/ucioption.cpp | 4 ++--
 3 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/src/tt.cpp b/src/tt.cpp
index 39f18d3d..3339c993 100644
--- a/src/tt.cpp
+++ b/src/tt.cpp
@@ -87,7 +87,7 @@ void TranspositionTable::clear() {
 
   std::vector<std::thread> threads;
 
-  for (size_t idx = 0; idx < Options["Threads"]; ++idx)
+  for (size_t idx = 0; idx < size_t(Options["Threads"]); ++idx)
   {
       threads.emplace_back([this, idx]() {
 
@@ -98,7 +98,7 @@ void TranspositionTable::clear() {
           // Each thread will zero its part of the hash table
           const size_t stride = size_t(clusterCount / Options["Threads"]),
                        start  = size_t(stride * idx),
-                       len    = idx != Options["Threads"] - 1 ?
+                       len    = idx != size_t(Options["Threads"]) - 1 ?
                                 stride : clusterCount - start;
 
           std::memset(&table[start], 0, len * sizeof(Cluster));
diff --git a/src/uci.h b/src/uci.h
index 70e45acc..9ca0ed36 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -61,7 +61,7 @@ public:
 
   Option& operator=(const std::string&);
   void operator<<(const Option&);
-  operator double() const;
+  operator int() const;
   operator std::string() const;
   bool operator==(const char*) const;
 
diff --git a/src/ucioption.cpp b/src/ucioption.cpp
index 39933ea5..f6342e5c 100644
--- a/src/ucioption.cpp
+++ b/src/ucioption.cpp
@@ -128,9 +128,9 @@ Option::Option(double v, int minv, int maxv, OnChange f) : type("spin"), min(min
 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 double() const {
+Option::operator int() const {
   assert(type == "check" || type == "spin");
-  return (type == "spin" ? stof(currentValue) : currentValue == "true");
+  return (type == "spin" ? std::stoi(currentValue) : currentValue == "true");
 }
 
 Option::operator std::string() const {

From a5643b89fda5060d4b40dff54afe02816c899dd4 Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Fri, 7 Apr 2023 14:30:59 +0000
Subject: [PATCH 123/192] Remove extraReduction

Since bestValue becomes value and beta - alpha is always non-negative,
extraReduction is always false, hence it has no effect.

This patch includes small changes to improve readability.

closes https://github.com/official-stockfish/Stockfish/pull/4505

No functional change
---
 src/search.cpp | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 5d54a15d..4463b42a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1335,15 +1335,14 @@ moves_loop: // When in check, search starts here
 
               if (PvNode && value < beta) // Update alpha! Always alpha < beta
               {
-
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
-                      && ((improving && complexity > 971) || (value < (5 * alpha + 75 * beta) / 87) || depth < 6)
+                      && (   (improving && complexity > 971)
+                          || value < (5 * alpha + 75 * beta) / 87
+                          || depth < 6)
                       && beta  <  12535
-                      && value > -12535) {
-                      bool extraReduction = depth > 2 && alpha > -12535 && bestValue != -VALUE_INFINITE && (value - bestValue) > (7 * (beta - alpha)) / 8;
-                      depth -= 1 + extraReduction;
-                  }
+                      && value > -12535)
+                      depth -= 1;
 
                   assert(depth > 0);
                   alpha = value;

From 2f2f45f9f47c1212f3229c22304456c9bad8f843 Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Thu, 6 Apr 2023 14:08:50 +0300
Subject: [PATCH 124/192] Made two specializations for affine transform easier
 to understand.

Added AVX-512 for the specialization for small inputs

closes https://github.com/official-stockfish/Stockfish/pull/4502

No functional change
---
 src/nnue/layers/affine_transform.h | 28 ++++++++++++++++------------
 1 file changed, 16 insertions(+), 12 deletions(-)

diff --git a/src/nnue/layers/affine_transform.h b/src/nnue/layers/affine_transform.h
index f84f054e..9e2f2f97 100644
--- a/src/nnue/layers/affine_transform.h
+++ b/src/nnue/layers/affine_transform.h
@@ -31,7 +31,7 @@
   This file contains the definition for a fully connected layer (aka affine transform).
   Two approaches are employed, depending on the sizes of the transform.
 
-  Approach 1:
+  Approach 1 (a specialization for large inputs):
     - used when the PaddedInputDimensions >= 128
     - uses AVX512 if possible
     - processes inputs in batches of 2*InputSimdWidth
@@ -42,9 +42,8 @@
       depends on the architecture (the amount of registers)
     - accumulate + hadd is used
 
-  Approach 2:
+  Approach 2 (a specialization for small inputs):
     - used when the PaddedInputDimensions < 128
-    - does not use AVX512
     - expected use-case is for when PaddedInputDimensions == 32 and InputDimensions <= 32.
       - that's why AVX512 is hard to implement
     - expected use-case is small layers
@@ -169,7 +168,7 @@ namespace Stockfish::Eval::NNUE::Layers {
   constexpr IndexType LargeInputSize = std::numeric_limits<IndexType>::max();
 #endif
 
-  // A specialization for large inputs.
+  // A specialization for large inputs
   template <IndexType InDims, IndexType OutDims>
   class AffineTransform<InDims, OutDims, std::enable_if_t<(ceil_to_multiple<IndexType>(InDims, MaxSimdWidth) >= LargeInputSize)>> {
    public:
@@ -188,7 +187,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     using OutputBuffer = OutputType[PaddedOutputDimensions];
 
-    static_assert(PaddedInputDimensions >= LargeInputSize, "Something went wrong. This specialization should not have been chosen.");
+    static_assert(PaddedInputDimensions >= LargeInputSize, "Something went wrong. This specialization (for large inputs) should not have been chosen.");
 
 #if defined (USE_AVX512)
     static constexpr IndexType InputSimdWidth = 64;
@@ -396,6 +395,7 @@ namespace Stockfish::Eval::NNUE::Layers {
     alignas(CacheLineSize) WeightType weights[OutputDimensions * PaddedInputDimensions];
   };
 
+  // A specialization for small inputs
   template <IndexType InDims, IndexType OutDims>
   class AffineTransform<InDims, OutDims, std::enable_if_t<(ceil_to_multiple<IndexType>(InDims, MaxSimdWidth) < LargeInputSize)>> {
    public:
@@ -415,12 +415,7 @@ namespace Stockfish::Eval::NNUE::Layers {
 
     using OutputBuffer = OutputType[PaddedOutputDimensions];
 
-    static_assert(PaddedInputDimensions < LargeInputSize, "Something went wrong. This specialization should not have been chosen.");
-
-#if defined (USE_SSSE3)
-    static constexpr IndexType OutputSimdWidth = SimdWidth / 4;
-    static constexpr IndexType InputSimdWidth = SimdWidth;
-#endif
+    static_assert(PaddedInputDimensions < LargeInputSize, "Something went wrong. This specialization (for small inputs) should not have been chosen.");
 
     // Hash value embedded in the evaluation file
     static constexpr std::uint32_t get_hash_value(std::uint32_t prevHash) {
@@ -470,7 +465,14 @@ namespace Stockfish::Eval::NNUE::Layers {
     const OutputType* propagate(
         const InputType* input, OutputType* output) const {
 
-#if defined (USE_AVX2)
+#if defined (USE_AVX512)
+      using vec_t = __m512i;
+      #define vec_setzero _mm512_setzero_si512
+      #define vec_set_32 _mm512_set1_epi32
+      #define vec_add_dpbusd_32 Simd::m512_add_dpbusd_epi32
+      #define vec_add_dpbusd_32x2 Simd::m512_add_dpbusd_epi32x2
+      #define vec_hadd Simd::m512_hadd
+#elif defined (USE_AVX2)
       using vec_t = __m256i;
       #define vec_setzero _mm256_setzero_si256
       #define vec_set_32 _mm256_set1_epi32
@@ -489,6 +491,8 @@ namespace Stockfish::Eval::NNUE::Layers {
 #if defined (USE_SSSE3)
       const auto inputVector = reinterpret_cast<const vec_t*>(input);
 
+      static constexpr IndexType OutputSimdWidth = sizeof(vec_t) / sizeof(OutputType);
+
       static_assert(OutputDimensions % OutputSimdWidth == 0 || OutputDimensions == 1);
 
       if constexpr (OutputDimensions % OutputSimdWidth == 0)

From 7a9f67747f23e837a8691ba9e6e4f0d1fdafff73 Mon Sep 17 00:00:00 2001
From: mstembera <MissingEmail@email>
Date: Tue, 4 Apr 2023 19:44:09 -0700
Subject: [PATCH 125/192] Reduce Position::pieces() overloads

Reduce the number of overloads for pieces() by using a more general template implementation.
Secondly simplify some code in search.cpp using the new general functionality.

TC https://tests.stockfishchess.org/tests/view/642ce27877ff3301150dc193
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 269640 W: 71775 L: 71809 D: 126056
Ptnml(0-2): 687, 27294, 78885, 27274, 680

closes https://github.com/official-stockfish/Stockfish/pull/4501

No functional change.
---
 src/position.h | 19 ++++++++-----------
 src/search.cpp | 12 +++++++-----
 2 files changed, 15 insertions(+), 16 deletions(-)

diff --git a/src/position.h b/src/position.h
index bb45c44a..a736f3e6 100644
--- a/src/position.h
+++ b/src/position.h
@@ -92,10 +92,9 @@ public:
 
   // Position representation
   Bitboard pieces(PieceType pt) const;
-  Bitboard pieces(PieceType pt1, PieceType pt2) const;
+  template<typename ...PieceTypes> Bitboard pieces(PieceType pt, PieceTypes... pts) const;
   Bitboard pieces(Color c) const;
-  Bitboard pieces(Color c, PieceType pt) const;
-  Bitboard pieces(Color c, PieceType pt1, PieceType pt2) const;
+  template<typename ...PieceTypes> Bitboard pieces(Color c, PieceTypes... pts) const;
   Piece piece_on(Square s) const;
   Square ep_square() const;
   bool empty(Square s) const;
@@ -229,20 +228,18 @@ inline Bitboard Position::pieces(PieceType pt = ALL_PIECES) const {
   return byTypeBB[pt];
 }
 
-inline Bitboard Position::pieces(PieceType pt1, PieceType pt2) const {
-  return pieces(pt1) | pieces(pt2);
+template<typename ...PieceTypes>
+inline Bitboard Position::pieces(PieceType pt, PieceTypes... pts) const {
+  return pieces(pt) | pieces(pts...);
 }
 
 inline Bitboard Position::pieces(Color c) const {
   return byColorBB[c];
 }
 
-inline Bitboard Position::pieces(Color c, PieceType pt) const {
-  return pieces(c) & pieces(pt);
-}
-
-inline Bitboard Position::pieces(Color c, PieceType pt1, PieceType pt2) const {
-  return pieces(c) & (pieces(pt1) | pieces(pt2));
+template<typename ...PieceTypes>
+inline Bitboard Position::pieces(Color c, PieceTypes... pts) const {
+  return pieces(c) & pieces(pts...);
 }
 
 template<PieceType Pt> inline int Position::count(Color c) const {
diff --git a/src/search.cpp b/src/search.cpp
index 4463b42a..4187117b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1018,16 +1018,18 @@ moves_loop: // When in check, search starts here
               {
                   if (depth < 2 - capture)
                       continue;
-                  // don't prune move if a heavy enemy piece (KQR) is under attack after the exchanges
-                  Bitboard leftEnemies = (pos.pieces(~us, QUEEN, ROOK) | pos.pieces(~us, KING)) & occupied;
+                  // Don't prune the move if opp. King/Queen/Rook is attacked by a slider after the exchanges.
+                  // Since in see_ge we don't update occupied when the king recaptures, we also don't prune the
+                  // move when the opp. King gets a discovered slider attack DURING the exchanges.
+                  Bitboard leftEnemies = pos.pieces(~us, ROOK, QUEEN, KING) & occupied;
                   Bitboard attacks = 0;
                   occupied |= to_sq(move);
                   while (leftEnemies && !attacks)
                   {
                       Square sq = pop_lsb(leftEnemies);
-                      attacks |= pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
-                      // exclude Queen/Rook(s) which were already threatened before SEE
-                      if (attacks && (sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us))))
+                      attacks = pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
+                      // Exclude Queen/Rook(s) which were already threatened before SEE
+                      if (attacks && sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us)))
                           attacks = 0;
                   }
                   if (!attacks)

From 1a64afb1c65591ccd374504c791eb27b762f6c8f Mon Sep 17 00:00:00 2001
From: Maxim Masiutin <maxim@masiutin.com>
Date: Sat, 1 Apr 2023 23:49:03 +0300
Subject: [PATCH 126/192] Do no initialize TM in all cases

Avoid doing full TM initialization if it won't be used, avoids division by zero.

closes https://github.com/official-stockfish/Stockfish/pull/4484

No functional change
---
 src/timeman.cpp | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/src/timeman.cpp b/src/timeman.cpp
index 5c826b4f..061de018 100644
--- a/src/timeman.cpp
+++ b/src/timeman.cpp
@@ -36,6 +36,12 @@ TimeManagement Time; // Our global time management object
 
 void TimeManagement::init(Search::LimitsType& limits, Color us, int ply) {
 
+  // if we have no time, no need to initialize TM, except for the start time,
+  // which is used by movetime.
+  startTime = limits.startTime;
+  if (limits.time[us] == 0)
+      return;
+
   TimePoint moveOverhead    = TimePoint(Options["Move Overhead"]);
   TimePoint slowMover       = TimePoint(Options["Slow Mover"]);
   TimePoint npmsec          = TimePoint(Options["nodestime"]);
@@ -59,8 +65,6 @@ void TimeManagement::init(Search::LimitsType& limits, Color us, int ply) {
       limits.npmsec = npmsec;
   }
 
-  startTime = limits.startTime;
-
   // Maximum move horizon of 50 moves
   int mtg = limits.movestogo ? std::min(limits.movestogo, 50) : 50;
 

From 7bd23d4d04d6644b6ccae8ea63cfc6646e4248dd Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Fri, 7 Apr 2023 12:33:28 -0400
Subject: [PATCH 127/192] Simplify away nnue scale pawn count multiplier

Removes 2x multipliers in nnue scale calculation along with the pawn count term that was recently reintroduced.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/64305bc720eb941419bdf72e
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 38008 W: 10234 L: 10021 D: 17753
Ptnml(0-2): 96, 4151, 10323, 4312, 122

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/6430b76a028b029b01ac9bfd
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 91232 W: 24686 L: 24547 D: 41999
Ptnml(0-2): 30, 8721, 27986, 8838, 41

closes https://github.com/official-stockfish/Stockfish/pull/4510

bench 4017320
---
 src/evaluate.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 2d0df89c..873dc5d2 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   else
   {
       int nnueComplexity;
-      int scale = 1001 + 5 * pos.count<PAWN>() + 61 * pos.non_pawn_material() / 4096;
+      int scale = 1001 + pos.non_pawn_material() / 64;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];

From 4ad2713e19cbf8db1e588c4d24d1fe4af5c6e917 Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Fri, 17 Mar 2023 00:04:41 +0300
Subject: [PATCH 128/192] Fix capturing underpromotions issue

Fix underpromotion captures are generated amongst quiets although dealt with as a capture_stage in search, this makes not skipping them when move count pruning kicks-in consistent with updating their histories amongst captures.

Passed STC:
https://tests.stockfishchess.org/tests/view/6415579f65775d3b539e7537
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 118896 W: 31678 L: 31553 D: 55665
Ptnml(0-2): 356, 12911, 32793, 13028, 360

Passed LTC:
https://tests.stockfishchess.org/tests/view/641633b965775d3b539e9e95
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 126800 W: 34255 L: 34148 D: 58397
Ptnml(0-2): 57, 12216, 38763, 12291, 73

see also discussion in https://github.com/official-stockfish/Stockfish/pull/4436
closes https://github.com/official-stockfish/Stockfish/pull/4452

bench: 3979409
---
 src/movegen.cpp | 18 +++++++++++++-----
 1 file changed, 13 insertions(+), 5 deletions(-)

diff --git a/src/movegen.cpp b/src/movegen.cpp
index 255dce04..6b28a52e 100644
--- a/src/movegen.cpp
+++ b/src/movegen.cpp
@@ -25,13 +25,21 @@ namespace Stockfish {
 
 namespace {
 
-  template<GenType Type, Direction D>
+  template<GenType Type, Direction D, bool Enemy>
   ExtMove* make_promotions(ExtMove* moveList, [[maybe_unused]] Square to) {
 
     if constexpr (Type == CAPTURES || Type == EVASIONS || Type == NON_EVASIONS)
+    {
         *moveList++ = make<PROMOTION>(to - D, to, QUEEN);
+        if constexpr (Enemy && Type == CAPTURES)
+        {
+            *moveList++ = make<PROMOTION>(to - D, to, ROOK);
+            *moveList++ = make<PROMOTION>(to - D, to, BISHOP);
+            *moveList++ = make<PROMOTION>(to - D, to, KNIGHT);
+        }
+    }
 
-    if constexpr (Type == QUIETS || Type == EVASIONS || Type == NON_EVASIONS)
+    if constexpr ((Type == QUIETS && !Enemy) || Type == EVASIONS || Type == NON_EVASIONS)
     {
         *moveList++ = make<PROMOTION>(to - D, to, ROOK);
         *moveList++ = make<PROMOTION>(to - D, to, BISHOP);
@@ -106,13 +114,13 @@ namespace {
             b3 &= target;
 
         while (b1)
-            moveList = make_promotions<Type, UpRight>(moveList, pop_lsb(b1));
+            moveList = make_promotions<Type, UpRight, true>(moveList, pop_lsb(b1));
 
         while (b2)
-            moveList = make_promotions<Type, UpLeft >(moveList, pop_lsb(b2));
+            moveList = make_promotions<Type, UpLeft, true>(moveList, pop_lsb(b2));
 
         while (b3)
-            moveList = make_promotions<Type, Up     >(moveList, pop_lsb(b3));
+            moveList = make_promotions<Type, Up,    false>(moveList, pop_lsb(b3));
     }
 
     // Standard and en passant captures

From f66c36277fe57d0ce4f10a4aeb5b41eb0cb9ebd1 Mon Sep 17 00:00:00 2001
From: Dubslow <bunslow@gmail.com>
Date: Wed, 5 Apr 2023 20:30:01 -0500
Subject: [PATCH 129/192] Remove nmpColor

no benefit seen, neither in game play nor for zugzwang test positions

STC: https://tests.stockfishchess.org/tests/view/642e293977ff3301150e9b55
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 161848 W: 43332 L: 43254 D: 75262
Ptnml(0-2): 418, 16987, 46058, 17021, 440

LTC: https://tests.stockfishchess.org/tests/view/642fea9420eb941419bde296
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 120208 W: 32529 L: 32418 D: 55261
Ptnml(0-2): 35, 11424, 37080, 11525, 40

closes https://github.com/official-stockfish/Stockfish/pull/4511

bench 3979409
---
 src/search.cpp | 5 ++---
 src/thread.h   | 1 -
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 4187117b..04299dbd 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -802,7 +802,7 @@ namespace {
         &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 253 + complexity / 25
         && !excludedMove
         &&  pos.non_pawn_material(us)
-        && (ss->ply >= thisThread->nmpMinPly || us != thisThread->nmpColor))
+        && (ss->ply >= thisThread->nmpMinPly))
     {
         assert(eval - beta >= 0);
 
@@ -830,9 +830,8 @@ namespace {
             assert(!thisThread->nmpMinPly); // Recursive verification is not allowed
 
             // Do verification search at high depths, with null move pruning disabled
-            // for us, until ply exceeds nmpMinPly.
+            // until ply exceeds nmpMinPly.
             thisThread->nmpMinPly = ss->ply + 3 * (depth-R) / 4;
-            thisThread->nmpColor = us;
 
             Value v = search<NonPV>(pos, ss, beta-1, beta, depth-R, false);
 
diff --git a/src/thread.h b/src/thread.h
index d6a48eca..3a114c79 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -62,7 +62,6 @@ public:
   size_t pvIdx, pvLast;
   std::atomic<uint64_t> nodes, tbHits, bestMoveChanges;
   int selDepth, nmpMinPly;
-  Color nmpColor;
   Value bestValue, optimism[COLOR_NB];
 
   Position rootPos;

From 9829bceda90d025a5f5d7c04457902413e367041 Mon Sep 17 00:00:00 2001
From: Stefan Geschwentner <stgeschwentner@gmail.com>
Date: Tue, 11 Apr 2023 01:20:29 +0200
Subject: [PATCH 130/192] Remove good killer reduction rule.

STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 246544 W: 65646 L: 65657 D: 115241
Ptnml(0-2): 706, 27350, 67138, 27405, 673
https://tests.stockfishchess.org/tests/view/642e253277ff3301150e9aa2

LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 165136 W: 44878 L: 44809 D: 75449
Ptnml(0-2): 64, 15991, 50378, 16082, 53
https://tests.stockfishchess.org/tests/view/6430db07028b029b01acd87f

closes https://github.com/official-stockfish/Stockfish/pull/4519

Bench: 3746080
---
 src/search.cpp | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 04299dbd..411befde 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1187,11 +1187,6 @@ moves_loop: // When in check, search starts here
       if ((ss+1)->cutoffCnt > 3)
           r++;
 
-      // Decrease reduction if move is a killer and we have a good history (~1 Elo)
-      if (move == ss->killers[0]
-          && (*contHist[0])[movedPiece][to_sq(move)] >= 3722)
-          r--;
-
       ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]

From acb0d204d56e16398c58822df2cc60b90ef1ae85 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Tue, 11 Apr 2023 19:53:36 +0300
Subject: [PATCH 131/192] Simplify stats assignment for Pv nodes

This patch is a simplification of my recent elo gainer.
Logically the Elo gainer didn't make much sense and this patch simplifies it into smth more logical.
Instead of assigning negative bonuses to all non-first moves that enter PV nodes
we assign positive bonuses in full depth search after LMR only for moves that
will result in a fail high - thus not assigning positive bonuses
for moves that will go to pv search - so doing "almost" the same as we do in master now for them.
Logic differs for some other moves, though, but this removes some lines of code.

Passed STC:
https://tests.stockfishchess.org/tests/view/642cf5cf77ff3301150dc5ec
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 409320 W: 109124 L: 109308 D: 190888
Ptnml(0-2): 1149, 45385, 111751, 45251, 1124

Passed LTC:
https://tests.stockfishchess.org/tests/view/642fe75d20eb941419bde200
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 260336 W: 70280 L: 70303 D: 119753
Ptnml(0-2): 99, 25236, 79528, 25199, 106

closes https://github.com/official-stockfish/Stockfish/pull/4522

Bench:  4286815
---
 src/search.cpp | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 411befde..390c6b1c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1229,8 +1229,9 @@ moves_loop: // When in check, search starts here
               if (newDepth > d)
                   value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth, !cutNode);
 
-              int bonus = value > alpha ?  stat_bonus(newDepth)
-                                        : -stat_bonus(newDepth);
+              int bonus = value <= alpha ? -stat_bonus(newDepth)
+                        : value >= beta  ?  stat_bonus(newDepth)
+                                         :  0;
 
               update_continuation_histories(ss, movedPiece, to_sq(move), bonus);
           }
@@ -1255,9 +1256,6 @@ moves_loop: // When in check, search starts here
           (ss+1)->pv[0] = MOVE_NONE;
 
           value = -search<PV>(pos, ss+1, -beta, -alpha, newDepth, false);
-
-          if (moveCount > 1 && newDepth >= depth && !capture)
-              update_continuation_histories(ss, movedPiece, to_sq(move), -stat_bonus(newDepth));
       }
 
       // Step 19. Undo move

From 96b6c0b36f28f0ef5fa58f48405db710542521df Mon Sep 17 00:00:00 2001
From: MinetaS <skdty87@gmail.com>
Date: Fri, 7 Apr 2023 14:49:05 +0000
Subject: [PATCH 132/192] Remove some conditions at PV improvement reduction

Non-regression STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 63664 W: 17007 L: 16823 D: 29834
Ptnml(0-2): 163, 6998, 17336, 7162, 173
https://tests.stockfishchess.org/tests/view/6430b124028b029b01ac99f2

Non-regression LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 90016 W: 24399 L: 24258 D: 41359
Ptnml(0-2): 52, 8672, 27405, 8841, 38
https://tests.stockfishchess.org/tests/view/64310e74028b029b01ad3131

closes https://github.com/official-stockfish/Stockfish/pull/4526

Bench: 3661938
---
 src/search.cpp | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 390c6b1c..ed81263a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1331,9 +1331,6 @@ moves_loop: // When in check, search starts here
               {
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
-                      && (   (improving && complexity > 971)
-                          || value < (5 * alpha + 75 * beta) / 87
-                          || depth < 6)
                       && beta  <  12535
                       && value > -12535)
                       depth -= 1;

From f9d9c69bc33dc7a17c28cd586d7e67c1bfff66f6 Mon Sep 17 00:00:00 2001
From: Torom <torsten.hellwig93@gmail.com>
Date: Thu, 13 Apr 2023 21:55:13 +0200
Subject: [PATCH 133/192] Set the length of GIT_SHA to 8 characters

Previously, the length of git commit hashes could vary depending on the git environment.

closes https://github.com/official-stockfish/Stockfish/pull/4527

No functional change
---
 src/Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/Makefile b/src/Makefile
index abcf11b0..82664618 100644
--- a/src/Makefile
+++ b/src/Makefile
@@ -702,7 +702,7 @@ ifeq ($(pext),yes)
 endif
 
 ### 3.7.1 Try to include git commit sha for versioning
-GIT_SHA = $(shell git rev-parse --short HEAD 2>/dev/null)
+GIT_SHA = $(shell git rev-parse HEAD 2>/dev/null | cut -c 1-8)
 ifneq ($(GIT_SHA), )
 	CXXFLAGS += -DGIT_SHA=$(GIT_SHA)
 endif

From c90dd38903206ede56fa73c15d7d2b366d56ebdb Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Fri, 14 Apr 2023 20:48:22 +0800
Subject: [PATCH 134/192] Simplify away complexity in evaluation

Simplification STC: https://tests.stockfishchess.org/tests/view/64394bc0605991a801b4f6f0
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 72360 W: 19313 L: 19138 D: 33909
Ptnml(0-2): 206, 7883, 19800, 8112, 179

Simplification LTC: https://tests.stockfishchess.org/tests/view/6439e788c233ce943b6bdac1
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 224992 W: 60665 L: 60654 D: 103673
Ptnml(0-2): 96, 21875, 68526, 21920, 79

closes https://github.com/official-stockfish/Stockfish/pull/4530

Bench: 3709369
---
 src/evaluate.cpp | 10 +---------
 src/evaluate.h   |  2 +-
 src/search.cpp   | 27 ++++++++-------------------
 src/thread.cpp   |  1 -
 src/thread.h     |  1 -
 5 files changed, 10 insertions(+), 31 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 873dc5d2..851ccfe1 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1046,7 +1046,7 @@ make_v:
 /// evaluate() is the evaluator for the outer world. It returns a static
 /// evaluation of the position from the point of view of the side to move.
 
-Value Eval::evaluate(const Position& pos, int* complexity) {
+Value Eval::evaluate(const Position& pos) {
 
   assert(!pos.checkers());
 
@@ -1075,10 +1075,6 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
                         + (424 + optimism) * abs(psq - nnue)
                         ) / 1024;
 
-      // Return hybrid NNUE complexity to caller
-      if (complexity)
-          *complexity = nnueComplexity;
-
       optimism = optimism * (272 + nnueComplexity) / 256;
       v = (nnue * scale + optimism * (scale - 748)) / 1024;
   }
@@ -1089,10 +1085,6 @@ Value Eval::evaluate(const Position& pos, int* complexity) {
   // Guarantee evaluation does not hit the tablebase range
   v = std::clamp(v, VALUE_TB_LOSS_IN_MAX_PLY + 1, VALUE_TB_WIN_IN_MAX_PLY - 1);
 
-  // When not using NNUE, return classical complexity to caller
-  if (complexity && useClassical)
-      *complexity = abs(v - psq);
-
   return v;
 }
 
diff --git a/src/evaluate.h b/src/evaluate.h
index 61846073..48076670 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -31,7 +31,7 @@ class Position;
 namespace Eval {
 
   std::string trace(Position& pos);
-  Value evaluate(const Position& pos, int* complexity = nullptr);
+  Value evaluate(const Position& pos);
 
   extern bool useNNUE;
   extern std::string currentEvalFileName;
diff --git a/src/search.cpp b/src/search.cpp
index ed81263a..3a7f85a8 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -294,14 +294,6 @@ void Thread::search() {
 
   if (mainThread)
   {
-
-      if (!rootPos.checkers())
-      {
-          int rootComplexity;
-          Eval::evaluate(rootPos, &rootComplexity);
-          mainThread->complexity = std::min(1.03 + (rootComplexity - 241) / 1552.0, 1.45);
-      }
-
       if (mainThread->bestPreviousScore == VALUE_INFINITE)
           for (int i = 0; i < 4; ++i)
               mainThread->iterValue[i] = VALUE_ZERO;
@@ -478,7 +470,7 @@ void Thread::search() {
           double reduction = (1.4 + mainThread->previousTimeReduction) / (2.08 * timeReduction);
           double bestMoveInstability = 1 + 1.8 * totBestMoveChanges / Threads.size();
 
-          double totalTime = Time.optimum() * fallingEval * reduction * bestMoveInstability * mainThread->complexity;
+          double totalTime = Time.optimum() * fallingEval * reduction * bestMoveInstability;
 
           // Cap used time in case of a single legal move for a better viewer experience in tournaments
           // yielding correct scores and sufficiently fast moves.
@@ -561,7 +553,7 @@ namespace {
     bool givesCheck, improving, priorCapture, singularQuietLMR;
     bool capture, moveCountPruning, ttCapture;
     Piece movedPiece;
-    int moveCount, captureCount, quietCount, improvement, complexity;
+    int moveCount, captureCount, quietCount, improvement;
 
     // Step 1. Initialize node
     Thread* thisThread = pos.this_thread();
@@ -723,7 +715,6 @@ namespace {
         ss->staticEval = eval = VALUE_NONE;
         improving = false;
         improvement = 0;
-        complexity = 0;
         goto moves_loop;
     }
     else if (excludedMove)
@@ -731,17 +722,15 @@ namespace {
         // Providing the hint that this node's accumulator will be used often brings significant Elo gain (13 Elo)
         Eval::NNUE::hint_common_parent_position(pos);
         eval = ss->staticEval;
-        complexity = abs(ss->staticEval - pos.psq_eg_stm());
     }
     else if (ss->ttHit)
     {
         // Never assume anything about values stored in TT
         ss->staticEval = eval = tte->eval();
         if (eval == VALUE_NONE)
-            ss->staticEval = eval = evaluate(pos, &complexity);
-        else // Fall back to (semi)classical complexity for TT hits, the NNUE complexity is lost
+            ss->staticEval = eval = evaluate(pos);
+        else
         {
-            complexity = abs(ss->staticEval - pos.psq_eg_stm());
             if (PvNode)
                Eval::NNUE::hint_common_parent_position(pos);
         }
@@ -753,7 +742,7 @@ namespace {
     }
     else
     {
-        ss->staticEval = eval = evaluate(pos, &complexity);
+        ss->staticEval = eval = evaluate(pos);
         // Save static evaluation into transposition table
         tte->save(posKey, VALUE_NONE, ss->ttPv, BOUND_NONE, DEPTH_NONE, MOVE_NONE, eval);
     }
@@ -799,15 +788,15 @@ namespace {
         && (ss-1)->statScore < 18755
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 253 + complexity / 25
+        &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 253
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly))
     {
         assert(eval - beta >= 0);
 
-        // Null move dynamic reduction based on depth, eval and complexity of position
-        Depth R = std::min(int(eval - beta) / 172, 6) + depth / 3 + 4 - (complexity > 825);
+        // Null move dynamic reduction based on depth and eval
+        Depth R = std::min(int(eval - beta) / 172, 6) + depth / 3 + 4;
 
         ss->currentMove = MOVE_NULL;
         ss->continuationHistory = &thisThread->continuationHistory[0][0][NO_PIECE][0];
diff --git a/src/thread.cpp b/src/thread.cpp
index 202768c8..c680393e 100644
--- a/src/thread.cpp
+++ b/src/thread.cpp
@@ -160,7 +160,6 @@ void ThreadPool::clear() {
   main()->bestPreviousScore = VALUE_INFINITE;
   main()->bestPreviousAverageScore = VALUE_INFINITE;
   main()->previousTimeReduction = 1.0;
-  main()->complexity = 1.0;
 }
 
 
diff --git a/src/thread.h b/src/thread.h
index 3a114c79..09bdb470 100644
--- a/src/thread.h
+++ b/src/thread.h
@@ -85,7 +85,6 @@ struct MainThread : public Thread {
   void search() override;
   void check_time();
 
-  double complexity;
   double previousTimeReduction;
   Value bestPreviousScore;
   Value bestPreviousAverageScore;

From 7b9b793fd544aa7a599b113a40533cde16de640b Mon Sep 17 00:00:00 2001
From: Guenther Demetz <guenther.demetz@wuerth-phoenix.com>
Date: Mon, 17 Apr 2023 11:38:26 +0200
Subject: [PATCH 135/192] Simplification of SEE verification logic

Use same logic for all handled pieces.
Don't prune the move if opponent King, Queen, Rook gets a discovered
attack while or after the exchanges.

remove an obsolete comment in position.cpp

Passed STC non regression:
https://tests.stockfishchess.org/tests/view/6437907594daa91835c290d0
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 107432 W: 28359 L: 28221 D: 50852
Ptnml(0-2): 298, 11724, 29524, 11882, 288

Passed LTC non-regression:
https://tests.stockfishchess.org/tests/view/6438ed2ebd1a5470263c51e8
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 236288 W: 63656 L: 63656 D: 108976
Ptnml(0-2): 99, 22960, 72011, 22990, 84

closes https://github.com/official-stockfish/Stockfish/pull/4533

bench: 3741125
---
 src/position.cpp | 3 +--
 src/search.cpp   | 8 +++-----
 2 files changed, 4 insertions(+), 7 deletions(-)

diff --git a/src/position.cpp b/src/position.cpp
index e6fdb511..2a9d798f 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -331,8 +331,7 @@ void Position::set_check_info() const {
 
 /// Position::set_state() computes the hash keys of the position, and other
 /// data that once computed is updated incrementally as moves are made.
-/// The function is only used when a new position is set up, and to verify
-/// the correctness of the StateInfo data when running in debug mode.
+/// The function is only used when a new position is set up
 
 void Position::set_state() const {
 
diff --git a/src/search.cpp b/src/search.cpp
index 3a7f85a8..5205fb57 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1006,17 +1006,15 @@ moves_loop: // When in check, search starts here
               {
                   if (depth < 2 - capture)
                       continue;
-                  // Don't prune the move if opp. King/Queen/Rook is attacked by a slider after the exchanges.
-                  // Since in see_ge we don't update occupied when the king recaptures, we also don't prune the
-                  // move when the opp. King gets a discovered slider attack DURING the exchanges.
-                  Bitboard leftEnemies = pos.pieces(~us, ROOK, QUEEN, KING) & occupied;
+                  // Don't prune the move if opp. King/Queen/Rook gets a discovered attack during or after the exchanges
+                  Bitboard leftEnemies = pos.pieces(~us, KING, QUEEN, ROOK);
                   Bitboard attacks = 0;
                   occupied |= to_sq(move);
                   while (leftEnemies && !attacks)
                   {
                       Square sq = pop_lsb(leftEnemies);
                       attacks = pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
-                      // Exclude Queen/Rook(s) which were already threatened before SEE
+                      // Exclude Queen/Rook(s) which were already threatened before SEE (opp King can't be in check when it's our turn)
                       if (attacks && sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us)))
                           attacks = 0;
                   }

From d64d4ac426e06cdd52249b0464d22f3cdb7fcf79 Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Fri, 14 Apr 2023 20:33:33 +0800
Subject: [PATCH 136/192] Simplify away depth condition for aspiration window
 adjust

Simplification STC:
https://tests.stockfishchess.org/tests/view/64351654596a20f264276ded
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 374664 W: 98942 L: 99089 D: 176633
Ptnml(0-2): 1049, 41767, 101878, 41558, 1080

Simplification LTC:
https://tests.stockfishchess.org/tests/view/6439499f605991a801b4f684
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 457880 W: 123021 L: 123233 D: 211626
Ptnml(0-2): 166, 44739, 139335, 44541, 159

closes https://github.com/official-stockfish/Stockfish/pull/4534

Bench: 3879281
---
 src/search.cpp | 23 +++++++++--------------
 1 file changed, 9 insertions(+), 14 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 5205fb57..e322a1c9 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -288,9 +288,7 @@ void Thread::search() {
 
   ss->pv = pv;
 
-  bestValue = delta = alpha = -VALUE_INFINITE;
-  beta = VALUE_INFINITE;
-  optimism[WHITE] = optimism[BLACK] = VALUE_ZERO;
+  bestValue = -VALUE_INFINITE;
 
   if (mainThread)
   {
@@ -349,18 +347,15 @@ void Thread::search() {
           selDepth = 0;
 
           // Reset aspiration window starting size
-          if (rootDepth >= 4)
-          {
-              Value prev = rootMoves[pvIdx].averageScore;
-              delta = Value(10) + int(prev) * prev / 16502;
-              alpha = std::max(prev - delta,-VALUE_INFINITE);
-              beta  = std::min(prev + delta, VALUE_INFINITE);
+          Value prev = rootMoves[pvIdx].averageScore;
+          delta = Value(10) + int(prev) * prev / 16502;
+          alpha = std::max(prev - delta,-VALUE_INFINITE);
+          beta  = std::min(prev + delta, VALUE_INFINITE);
 
-              // Adjust optimism based on root move's previousScore
-              int opt = 120 * prev / (std::abs(prev) + 161);
-              optimism[ us] = Value(opt);
-              optimism[~us] = -optimism[us];
-          }
+          // Adjust optimism based on root move's previousScore
+          int opt = 120 * prev / (std::abs(prev) + 161);
+          optimism[ us] = Value(opt);
+          optimism[~us] = -optimism[us];
 
           // Start with a small aspiration window and, in the case of a fail
           // high/low, re-search with a bigger window until we don't fail

From ba06c480a752458a8159db0c9110bd3b7e34145a Mon Sep 17 00:00:00 2001
From: Stefan Geschwentner <stgeschwentner@gmail.com>
Date: Fri, 21 Apr 2023 16:22:55 +0200
Subject: [PATCH 137/192] Less reduction for tt move.

This idea is a result of my second condition combination tuning for reductions:
https://tests.stockfishchess.org/tests/view/643ed5573806eca398f06d61

There were used two parameters per combination: one for the 'sign' of the first and the second condition in a combination. Values >= 50 indicate using a condition directly and values <= -50 means use the negation of a condition.

Each condition pair (X,Y) had two occurances dependent of the order of the two conditions:
- if X < Y the parameters used for more reduction
- if X > Y the parameters used for less reduction
- if X = Y then only one condition is present and A[X][X][0]/A[X][X][1] stands for using more/less reduction for only this condition.

The parameter pair A[7][2][0] (value = -94.70) and A[7][2][1] (value = 93.60) was one of the strongest signals with values near 100/-100.
Here condition nr. 7 was '(ss+1)->cutoffCnt > 3' and condition nr. 2 'move == ttMove'. For condition nr. 7 the negation is used because A[7][2][0] is negative.

This translates finally to less reduction (because 7 > 2) for tt moves if child cutoffs <= 3.

STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 65728 W: 17704 L: 17358 D: 30666
Ptnml(0-2): 184, 7092, 18008, 7354, 226
https://tests.stockfishchess.org/tests/view/643ff767ef2529086a7ed042

LTC:
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 139200 W: 37776 L: 37282 D: 64142
Ptnml(0-2): 58, 13241, 42509, 13733, 59
https://tests.stockfishchess.org/tests/view/6440bfa9ef2529086a7edbc7

closes https://github.com/official-stockfish/Stockfish/pull/4538

Bench: 3548023
---
 src/search.cpp | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/search.cpp b/src/search.cpp
index e322a1c9..366065b8 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1169,6 +1169,9 @@ moves_loop: // When in check, search starts here
       if ((ss+1)->cutoffCnt > 3)
           r++;
 
+      else if (move == ttMove)
+          r--;
+
       ss->statScore =  2 * thisThread->mainHistory[us][from_to(move)]
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]

From b22a1b10bbae2bb773afb50eba23dbf15e426365 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Bo=C5=A1tjan=20Mejak?= <bostjan.xperia@gmail.com>
Date: Tue, 11 Apr 2023 13:55:14 +0200
Subject: [PATCH 138/192] Update AUTHORS

Improved some comments in the AUTHORS file, sort contributors

closes https://github.com/official-stockfish/Stockfish/pull/4520

No functional change
---
 AUTHORS | 37 +++++++++++++++++--------------------
 1 file changed, 17 insertions(+), 20 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 9b36111e..b6723246 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -1,17 +1,15 @@
-# List of authors for Stockfish
-
-# Founders of the Stockfish project and fishtest infrastructure
+# Founders of the Stockfish project and Fishtest infrastructure
 Tord Romstad (romstad)
 Marco Costalba (mcostalba)
 Joona Kiiski (zamar)
 Gary Linscott (glinscott)
 
-# Authors and inventors of NNUE, training, NNUE port
+# Authors and inventors of NNUE, training, and NNUE port
 Yu Nasu (ynasu87)
 Motohiro Isozaki (yaneurao)
 Hisayori Noda (nodchip)
 
-# all other authors of the code in alphabetical order
+# All other authors of Stockfish code (in alphabetical order)
 Aditya (absimaldata)
 Adrian Petrescu (apetresc)
 Ajith Chandy Jose (ajithcj)
@@ -47,12 +45,12 @@ Chess13234
 Chris Cain (ceebo)
 clefrks
 Dale Weiler (graphitemaster)
-Dan Schmidt (dfannius)
 Daniel Axtens (daxtens)
 Daniel Dugovic (ddugovic)
+Dan Schmidt (dfannius)
 Dariusz Orzechowski (dorzechowski)
-David Zar
 David (dav1312)
+David Zar
 Daylen Yang (daylen)
 Deshawn Mohan-Smith (GoldenRare)
 Dieter Dobbelaere (ddobbelaere)
@@ -66,7 +64,6 @@ Eelco de Groot (KingDefender)
 Elvin Liu (solarlight2)
 erbsenzaehler
 Ernesto Gatti
-Linmiao Xu (linrock)
 Fabian Beuke (madnight)
 Fabian Fichter (ianfab)
 Fanael Linithien (Fanael)
@@ -83,30 +80,30 @@ Gontran Lemaire (gonlem)
 Goodkov Vasiliy Aleksandrovich (goodkov)
 Gregor Cramer
 GuardianRM
-Günther Demetz (pb00067, pb00068)
 Guy Vreuls (gvreuls)
+Günther Demetz (pb00067, pb00068)
 Henri Wiechers
 Hiraoka Takuya (HiraokaTakuya)
 homoSapiensSapiens
 Hongzhi Cheng
 Ivan Ivec (IIvec)
 Jacques B. (Timshel)
+Jake Senne (w1wwwwww)
 Jan Ondruš (hxim)
 Jared Kish (Kurtbusch, kurt22i)
-Jake Senne (w1wwwwww)
 Jarrod Torriero (DU-jdto)
-Jean Gauthier (OuaisBla)
 Jean-Francois Romang (jromang)
+Jean Gauthier (OuaisBla)
 Jekaa
 Jerry Donald Watson (jerrydonaldwatson)
 jjoshua2
-Jonathan Calovski (Mysseno)
 Jonathan Buladas Dumale (SFisGOD)
+Jonathan Calovski (Mysseno)
 Jonathan McDermid (jonathanmcdermid)
 Joost VandeVondele (vondele)
-Jörg Oster (joergoster)
 Joseph Ellis (jhellis3)
 Joseph R. Prostko
+Jörg Oster (joergoster)
 Julian Willemer (NightlyKing)
 jundery
 Justin Blanchard (UncombedCoconut)
@@ -120,6 +117,7 @@ Krystian Kuzniarek (kuzkry)
 Leonardo Ljubičić (ICCF World Champion)
 Leonid Pechenik (lp--)
 Liam Keegan (lkeegan)
+Linmiao Xu (linrock)
 Linus Arver (listx)
 loco-loco
 Lub van den Berg (ElbertoOne)
@@ -151,11 +149,11 @@ Moez Jellouli (MJZ1977)
 Mohammed Li (tthsqe12)
 Muzhen J (XInTheDark)
 Nathan Rugg (nmrugg)
-Nick Pelling (nickpelling)
+Nguyen Pham (nguyenpham)
 Nicklas Persson (NicklasPersson)
+Nick Pelling (nickpelling)
 Niklas Fiekas (niklasf)
 Nikolay Kostov (NikolayIT)
-Nguyen Pham (nguyenpham)
 Norman Schmidt (FireFather)
 notruck
 Ofek Shochat (OfekShochat, ghostway)
@@ -170,6 +168,7 @@ Peter Schneider (pschneider1968)
 Peter Zsifkovits (CoffeeOne)
 PikaCat
 Praveen Kumar Tummala (praveentml)
+Prokop Randáček (ProkopRandacek)
 Rahul Dsilva (silversolver1)
 Ralph Stößer (Ralph Stoesser)
 Raminder Singh
@@ -178,8 +177,8 @@ Reuven Peleg (R-Peleg)
 Richard Lloyd (Richard-Lloyd)
 Rodrigo Exterckötter Tjäder
 Rodrigo Roim (roim)
-Ron Britvich (Britvich)
 Ronald de Man (syzygy1, syzygy)
+Ron Britvich (Britvich)
 rqs
 Rui Coelho (ruicoelhopedro)
 Ryan Schmitt
@@ -200,13 +199,12 @@ Stefano Di Martino (StefanoD)
 Steinar Gunderson (sesse)
 Stéphane Nicolet (snicolet)
 Syine Mineta (MinetaS)
-Prokop Randáček (ProkopRandacek)
 Thanar2
 thaspel
 theo77186
+Tomasz Sobczyk (Sopel97)
 Tom Truscott
 Tom Vijlbrief (tomtor)
-Tomasz Sobczyk (Sopel97)
 Torsten Franz (torfranz, tfranzer)
 Torsten Hellwig (Torom)
 Tracey Emery (basepr1me)
@@ -217,8 +215,7 @@ Vince Negri (cuddlestmonkey)
 xefoci7612
 zz4032
 
-
 # Additionally, we acknowledge the authors and maintainers of fishtest,
-# an amazing and essential framework for the development of Stockfish!
+# an amazing and essential framework for Stockfish development!
 #
 # https://github.com/glinscott/fishtest/blob/master/AUTHORS

From c3ce2204083400267592dc088b8ad9e88aed56b1 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Sat, 15 Apr 2023 09:51:01 -0400
Subject: [PATCH 139/192] Created by retraining the master net with these
 changes to the dataset:

* Extending v6 filtering to data from T77 dec2021, T79 may2022, and T80 nov2022
* Reducing the number of duplicate positions, prioritizing position scores seen later in time
* Using a binpack minimizer to reduce the overall data size

Trained the same way as the previous master net, aside from the dataset changes:

```
python3 easy_train.py \
  --experiment-name leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovT79aprmayT78jantosepT77dec-v6dd \
  --training-dataset /data/leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovT79aprmayT78jantosepT77dec-v6dd.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes \
  --start-from-engine-test-net True \
  --early-fen-skipping 30 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --max_epoch 900 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --tui False \
  --gpus "0," \
  --seed $RANDOM
```

The new v6-dd filtering reduces duplicate positions by iterating over hourly data files within leela test runs, starting with the most recent, then keeping positions the first time they're seen and ignoring positions that are seen again. This ordering was done with the assumption that position scores seen later in time are generally more accurate than scores seen earlier in the test run. Positions are de-duplicated based on piece orientations, the first token in fen strings.

The binpack minimizer was run with default settings after first merging monthly data into single binpacks.

```
python3 interleave_binpacks.py \
  leela96-filt-v2.binpack \
  dfrc99-filt-v2.binpack \
  T60-nov2021-12tb7p-eval-filt-v2.binpack \
  T60-dec2021-12tb7p-eval-filt-v2.binpack \
  filt-v6/test80-aug2022-16tb7p-filter-v6.min-mar2023.binpack \
  filt-v6/test80-sep2022-16tb7p-filter-v6.min-mar2023.binpack \
  filt-v6-dd/test80-jun2022-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test80-jul2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-oct2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-nov2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test79-apr2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test79-may2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test78-jantomay2022-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test78-juntosep2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test77-dec2021-16tb7p-filter-v6-dd.binpack \
  /data/leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovT79aprmayT78jantosepT77dec-v6dd.binpack
```

The code for v6-dd filtering is available along with training data preparation scripts at:
https://github.com/linrock/nnue-data

Links for downloading the training data components:
https://robotmoon.com/nnue-training-data/

The binpack minimizer is from: #4447

Local elo at 25k nodes per move:
nn-epoch859.nnue : 1.2 +/- 2.6

Passed STC:
https://tests.stockfishchess.org/tests/view/643aad7db08900ff1bc5a832
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 565040 W: 150225 L: 149162 D: 265653
Ptnml(0-2): 1875, 62137, 153229, 63608, 1671

Passed LTC:
https://tests.stockfishchess.org/tests/view/643ecf2fa43cf30e719d2042
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 1014840 W: 274645 L: 272456 D: 467739
Ptnml(0-2): 515, 98565, 306970, 100956, 414

closes https://github.com/official-stockfish/Stockfish/pull/4545

bench 3476305
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 48076670..9dc45371 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-dabb1ed23026.nnue"
+  #define EvalFileDefaultName   "nn-1ceb1a57d117.nnue"
 
   namespace NNUE {
 

From 41f50b2c83a0ba36a2b9c507c1783e57c9b13485 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Fri, 21 Apr 2023 09:37:29 -0400
Subject: [PATCH 140/192] Update default net to nn-e1fb1ade4432.nnue

Created by retraining nn-dabb1ed23026.nnue with a dataset composed of:

* The previous best dataset (nn-1ceb1a57d117.nnue dataset)
* Adding de-duplicated T80 data from feb2023 and the last 10 days of jan2023, filtered with v6-dd

Initially trained with the same options as the recent master net (nn-1ceb1a57d117.nnue).
Around epoch 890, training was manually stopped and max epoch increased to 1000.

```
python3 easy_train.py \
  --experiment-name leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovjanfebT79aprmayT78jantosepT77dec-v6dd \
  --training-dataset /data/leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovjanfebT79aprmayT78jantosepT77dec-v6dd.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes \
  --start-from-engine-test-net True \
  --early-fen-skipping 30 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --max_epoch 900 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --tui False \
  --gpus "0," \
  --seed $RANDOM
```

The same v6-dd filtering and binpack minimizer was used for preparing the recent nn-1ceb1a57d117.nnue dataset.

```
python3 interleave_binpacks.py \
  leela96-filt-v2.binpack \
  dfrc99-filt-v2.binpack \
  T60-nov2021-12tb7p-eval-filt-v2.binpack \
  T60-dec2021-12tb7p-eval-filt-v2.binpack \
  filt-v6/test80-aug2022-16tb7p-filter-v6.min-mar2023.binpack \
  filt-v6/test80-sep2022-16tb7p-filter-v6.min-mar2023.binpack \
  filt-v6-dd/test80-jun2022-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test80-jul2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-oct2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-nov2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-jan2022-3of3-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test80-feb2023-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test79-apr2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test79-may2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test78-jantomay2022-16tb7p-filter-v6-dd.min-mar2023.binpack \
  filt-v6-dd/test78-juntosep2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test77-dec2021-16tb7p-filter-v6-dd.binpack \
  /data/leela96-dfrc99-T60novdec-v2-T80augsep-v6-T80junjuloctnovjanfebT79aprmayT78jantosepT77dec-v6dd.binpack
```

Links for downloading the training data components can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch919.nnue : 2.6 +/- 2.8

Passed STC vs. nn-dabb1ed23026.nnue
https://tests.stockfishchess.org/tests/view/644420df94ff3db5625f2af5
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 125960 W: 33898 L: 33464 D: 58598
Ptnml(0-2): 351, 13920, 34021, 14320, 368

Passed LTC vs. nn-1ceb1a57d117.nnue
https://tests.stockfishchess.org/tests/view/64469f128d30316529b3dc46
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 24544 W: 6817 L: 6542 D: 11185
Ptnml(0-2): 8, 2252, 7488, 2505, 19

closes https://github.com/official-stockfish/Stockfish/pull/4546

bench 3714847
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 9dc45371..f5db1c1e 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-1ceb1a57d117.nnue"
+  #define EvalFileDefaultName   "nn-e1fb1ade4432.nnue"
 
   namespace NNUE {
 

From 21d6b69f7c8d0c0a71fe627714913a59d39a3b57 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Sat, 29 Apr 2023 15:45:22 -0400
Subject: [PATCH 141/192] Update 7 eval and optimism params

Params found using spsa at 30+0.3 with this tuning config:

```
// evaluate.cpp
int nnueOptScaleBase = 1001;
int nnueComplexityMult = 406;
int nnueComplexityOptOffset = 424;
int evalOptComplexityOffset = 272;
int evalOptScaleOffset = 748;
TUNE(SetRange(801, 1201), nnueOptScaleBase);
TUNE(SetRange(306, 506), nnueComplexityMult);
TUNE(SetRange(324, 524), nnueComplexityOptOffset);
TUNE(SetRange(172, 372), evalOptComplexityOffset);
TUNE(SetRange(648, 848), evalOptScaleOffset);

// search.cpp
int searchOptBase = 120;
int searchOptDenom = 161;
TUNE(SetRange(20, 220), searchOptBase);
TUNE(SetRange(111, 211), searchOptDenom);
```

Passed STC:
https://tests.stockfishchess.org/tests/view/644dda8accf5e93df5e50cbe
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 136800 W: 36682 L: 36237 D: 63881
Ptnml(0-2): 353, 14910, 37492, 15229, 416

Passed LTC:
https://tests.stockfishchess.org/tests/view/644eaedb3f31c3bbe4a3d345
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 64548 W: 17624 L: 17272 D: 29652
Ptnml(0-2): 33, 6112, 19631, 6466, 32

closes https://github.com/official-stockfish/Stockfish/pull/4550

bench 3670343
---
 src/evaluate.cpp | 10 +++++-----
 src/search.cpp   |  2 +-
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 851ccfe1..d8f4e2e1 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos) {
   else
   {
       int nnueComplexity;
-      int scale = 1001 + pos.non_pawn_material() / 64;
+      int scale = 967 + pos.non_pawn_material() / 64;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];
@@ -1071,12 +1071,12 @@ Value Eval::evaluate(const Position& pos) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  406 * nnueComplexity
-                        + (424 + optimism) * abs(psq - nnue)
+      nnueComplexity = (  402 * nnueComplexity
+                        + (454 + optimism) * abs(psq - nnue)
                         ) / 1024;
 
-      optimism = optimism * (272 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 748)) / 1024;
+      optimism = optimism * (274 + nnueComplexity) / 256;
+      v = (nnue * scale + optimism * (scale - 791)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling
diff --git a/src/search.cpp b/src/search.cpp
index 366065b8..8ce9c56e 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -353,7 +353,7 @@ void Thread::search() {
           beta  = std::min(prev + delta, VALUE_INFINITE);
 
           // Adjust optimism based on root move's previousScore
-          int opt = 120 * prev / (std::abs(prev) + 161);
+          int opt = 102 * prev / (std::abs(prev) + 147);
           optimism[ us] = Value(opt);
           optimism[~us] = -optimism[us];
 

From 72d542f00026fd8437e6033c95802714e4cd45d1 Mon Sep 17 00:00:00 2001
From: candirufish <38038147+candirufish@users.noreply.github.com>
Date: Fri, 28 Apr 2023 19:05:56 +0200
Subject: [PATCH 142/192] Adjust reductions

Decrease further on cutNodes with tte->depth() >= depth + 3 condition.

LTC: https://tests.stockfishchess.org/tests/view/644dc84bccf5e93df5e50c13
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 155346 W: 42184 L: 41660 D: 71502
Ptnml(0-2): 59, 14765, 47504, 15283, 62

STC: https://tests.stockfishchess.org/tests/view/644d05de68e01d8194cd9bbb
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 121888 W: 32868 L: 32444 D: 56576
Ptnml(0-2): 332, 13273, 33343, 13631, 365

closes https://github.com/official-stockfish/Stockfish/pull/4552

bench: 3739675
---
 src/search.cpp | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 8ce9c56e..a6618c5b 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -884,7 +884,7 @@ namespace {
     // Use qsearch if depth is equal or below zero (~9 Elo)
     if (    PvNode
         && !ttMove)
-        depth -= 2 + 2 * (ss->ttHit &&  tte->depth() >= depth);
+        depth -= 2 + 2 * (ss->ttHit && tte->depth() >= depth);
 
     if (depth <= 0)
         return qsearch<PV>(pos, ss, alpha, beta);
@@ -1141,9 +1141,10 @@ moves_loop: // When in check, search starts here
 
       // Decrease reduction if position is or has been on the PV
       // and node is not likely to fail low. (~3 Elo)
+      // Decrease further on cutNodes. (~1 Elo)
       if (   ss->ttPv
           && !likelyFailLow)
-          r -= 2;
+          r -= cutNode && tte->depth() >= depth + 3 ? 3 : 2;
 
       // Decrease reduction if opponent's move count is high (~1 Elo)
       if ((ss-1)->moveCount > 7)

From 2429e162890478a48ab9b1cf0c431de9ebaf9429 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 5 May 2023 07:54:12 +0300
Subject: [PATCH 143/192] Reduce more if current node has a lot of refuted
 moves.

This patch refines idea of cutoff count - in master we reduce more if current node has at least 4 moves that are refuted by search, this patch increases this count by 1 if refutation happened without having a tt move.

Passed STC:
https://tests.stockfishchess.org/tests/view/645363c36206ee34ebf8191d
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 67616 W: 18220 L: 17874 D: 31522
Ptnml(0-2): 142, 7346, 18504, 7656, 160

Passed LTC:
https://tests.stockfishchess.org/tests/view/6453a0ea6206ee34ebf82796
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 195228 W: 52741 L: 52140 D: 90347
Ptnml(0-2): 53, 18718, 59482, 19297, 64

closes https://github.com/official-stockfish/Stockfish/pull/4556

bench 3448916
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index a6618c5b..a1f916d0 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1326,7 +1326,7 @@ moves_loop: // When in check, search starts here
               }
               else
               {
-                  ss->cutoffCnt++;
+                  ss->cutoffCnt += 1 + !ttMove;
                   assert(value >= beta); // Fail high
                   break;
               }

From 28442195c7d168a87221c6f1ae9ac51893427250 Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Wed, 3 May 2023 14:11:55 +0300
Subject: [PATCH 144/192] Clean up after "Simplify away complexity in
 evaluation"

closes https://github.com/official-stockfish/Stockfish/pull/4555

No functional change.
---
 src/search.cpp | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index a1f916d0..8d9dc04f 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -724,11 +724,8 @@ namespace {
         ss->staticEval = eval = tte->eval();
         if (eval == VALUE_NONE)
             ss->staticEval = eval = evaluate(pos);
-        else
-        {
-            if (PvNode)
-               Eval::NNUE::hint_common_parent_position(pos);
-        }
+        else if (PvNode)
+            Eval::NNUE::hint_common_parent_position(pos);
 
         // ttValue can be used as a better position evaluation (~7 Elo)
         if (    ttValue != VALUE_NONE

From 464ebdf127273db0ccb0084c881b255b880e922e Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Fri, 5 May 2023 23:32:56 +0300
Subject: [PATCH 145/192] Small cleanup

In search remove one condition check and reorder conditions. Removes some code.

Passed non-regression test:
https://tests.stockfishchess.org/tests/view/64548fa06206ee34ebf853ad
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 282976 W: 75327 L: 75374 D: 132275
Ptnml(0-2): 604, 29673, 80995, 29598, 618

closes https://github.com/official-stockfish/Stockfish/pull/4557

No functional change
---
 src/search.cpp | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 8d9dc04f..45fc1a7e 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1310,7 +1310,13 @@ moves_loop: // When in check, search starts here
               if (PvNode && !rootNode) // Update pv even in fail-high case
                   update_pv(ss->pv, move, (ss+1)->pv);
 
-              if (PvNode && value < beta) // Update alpha! Always alpha < beta
+              if (value >= beta)
+              {
+                  ss->cutoffCnt += 1 + !ttMove;
+                  assert(value >= beta); // Fail high
+                  break;
+              }
+              else
               {
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
@@ -1319,13 +1325,7 @@ moves_loop: // When in check, search starts here
                       depth -= 1;
 
                   assert(depth > 0);
-                  alpha = value;
-              }
-              else
-              {
-                  ss->cutoffCnt += 1 + !ttMove;
-                  assert(value >= beta); // Fail high
-                  break;
+                  alpha = value; // Update alpha! Always alpha < beta
               }
           }
       }

From 65e2150501b87e6ce00fae4e3f056444f39462fd Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Sun, 7 May 2023 23:33:04 +0300
Subject: [PATCH 146/192] Refine deeper post-lmr searches

This patch improves logic conditions for performing deeper searches after passed LMR.

Instead of exceeding alpha by some margin now it requires to exceed the
current best value - which may be lower than alpha (but never bigger since we
update alpha with bestvalue if it exceeds alpha).

Passed STC:
https://tests.stockfishchess.org/tests/view/6455f78008858de8313775b6
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 209344 W: 55993 L: 55448 D: 97903
Ptnml(0-2): 507, 22798, 57526, 23325, 516

Passed LTC:
https://tests.stockfishchess.org/tests/view/64572d46eb75932ccfebff97
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 66288 W: 17867 L: 17514 D: 30907
Ptnml(0-2): 21, 6240, 20269, 6593, 21

closes https://github.com/official-stockfish/Stockfish/pull/4559

bench 3808503
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 45fc1a7e..bc495c0e 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1201,7 +1201,7 @@ moves_loop: // When in check, search starts here
           {
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
-              const bool doDeeperSearch = value > (alpha + 58 + 12 * (newDepth - d));
+              const bool doDeeperSearch = value > (bestValue + 68 + 12 * (newDepth - d));
               const bool doEvenDeeperSearch = value > alpha + 588 && ss->doubleExtensions <= 5;
               const bool doShallowerSearch = value < bestValue + newDepth;
 

From 5f7b26aaa0d0bbdeb50ea6b17f049f167c1eb996 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Fri, 19 May 2023 23:02:27 +0200
Subject: [PATCH 147/192] Update WLD model

using data of May, recalibrate the WLD model.

closes https://github.com/official-stockfish/Stockfish/pull/4577

No functional change
---
 src/uci.cpp | 4 ++--
 src/uci.h   | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/uci.cpp b/src/uci.cpp
index 8f9684ee..523d551e 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -207,8 +207,8 @@ namespace {
      // The coefficients of a third-order polynomial fit is based on the fishtest data
      // for two parameters that need to transform eval to the argument of a logistic
      // function.
-     constexpr double as[] = {   0.33677609,   -4.30175627,   33.08810557,  365.60223431};
-     constexpr double bs[] = {  -2.50471102,   14.23235405,  -14.33066859,   71.42705250 };
+     constexpr double as[] = {   1.07390458,   -6.94334517,   31.95090161,  317.75424048};
+     constexpr double bs[] = {  -2.82843814,   16.64518180,  -19.74439200,   68.39499088 };
 
      // Enforce that NormalizeToPawnValue corresponds to a 50% win rate at ply 64
      static_assert(UCI::NormalizeToPawnValue == int(as[0] + as[1] + as[2] + as[3]));
diff --git a/src/uci.h b/src/uci.h
index 9ca0ed36..680d2d2c 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -35,7 +35,7 @@ namespace UCI {
 // the win_rate_model() such that Stockfish outputs an advantage of
 // "100 centipawns" for a position if the engine has a 50% probability to win
 // from this position in selfplay at fishtest LTC time control.
-const int NormalizeToPawnValue = 394;
+const int NormalizeToPawnValue = 343;
 
 class Option;
 

From f030a1c592eabfa602ff8560d365e516298363ce Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Tue, 16 May 2023 21:30:53 +0800
Subject: [PATCH 148/192] Search tuning at very long time control

Many search parameter changes, tuned
(https://tests.stockfishchess.org/tests/view/645e4c67d55cccb2e64220ff)
at ~300k games @ VLTC (120+1.2).

Failed STC:
https://tests.stockfishchess.org/tests/view/6465fcd77968ca827c1410c2
LLR: -2.95 (-2.94,2.94) <0.00,2.00>
Total: 33824 W: 8863 L: 9067 D: 15894
Ptnml(0-2): 89, 3833, 9266, 3641, 83

Neutral LTC:
https://tests.stockfishchess.org/tests/view/646385ce87f6567dd4df4e37
Elo: -0.48 +-1.2 (95%) LOS: 22.2%
Total: 60000 W: 16235 L: 16318 D: 27447
Ptnml(0-2): 27, 5831, 18366, 5750, 26
nElo: -1.08 +-2.8 (95%) PairsRatio: 0.99

Passed VLTC 180+1.8:
https://tests.stockfishchess.org/tests/view/646385f787f6567dd4df4e3e
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 126448 W: 34704 L: 34258 D: 57486
Ptnml(0-2): 9, 10970, 40825, 11406, 14

Passed VLTC SMP 60+0.6 8thread:
https://tests.stockfishchess.org/tests/view/646628de884ce93b65df2ac9
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 59456 W: 16791 L: 16487 D: 26178
Ptnml(0-2): 5, 4473, 20467, 4779, 4

closes https://github.com/official-stockfish/Stockfish/pull/4574

Bench: 3347573
---
 src/search.cpp | 94 +++++++++++++++++++++++++-------------------------
 1 file changed, 47 insertions(+), 47 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index bc495c0e..429db9a5 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -64,7 +64,7 @@ namespace {
 
   // Futility margin
   Value futility_margin(Depth d, bool improving) {
-    return Value(154 * (d - improving));
+    return Value(148 * (d - improving));
   }
 
   // Reductions lookup table, initialized at startup
@@ -72,7 +72,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1449 - int(delta) * 937 / int(rootDelta)) / 1024 + (!i && r > 941);
+    return (r + 1356 - int(delta) * 983 / int(rootDelta)) / 1024 + (!i && r > 901);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -82,7 +82,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min(341 * d - 470, 1710);
+    return std::min(337 * d - 497, 1632);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -162,7 +162,7 @@ namespace {
 void Search::init() {
 
   for (int i = 1; i < MAX_MOVES; ++i)
-      Reductions[i] = int((19.47 + std::log(Threads.size()) / 2) * std::log(i));
+      Reductions[i] = int((20.89 + std::log(Threads.size()) / 2) * std::log(i));
 }
 
 
@@ -348,12 +348,12 @@ void Thread::search() {
 
           // Reset aspiration window starting size
           Value prev = rootMoves[pvIdx].averageScore;
-          delta = Value(10) + int(prev) * prev / 16502;
+          delta = Value(11) + int(prev) * prev / 15368;
           alpha = std::max(prev - delta,-VALUE_INFINITE);
           beta  = std::min(prev + delta, VALUE_INFINITE);
 
           // Adjust optimism based on root move's previousScore
-          int opt = 102 * prev / (std::abs(prev) + 147);
+          int opt = 116 * prev / (std::abs(prev) + 143);
           optimism[ us] = Value(opt);
           optimism[~us] = -optimism[us];
 
@@ -742,7 +742,7 @@ namespace {
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
-        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1920, 1920);
+        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1717, 1717);
         thisThread->mainHistory[~us][from_to((ss-1)->currentMove)] << bonus;
     }
 
@@ -752,13 +752,13 @@ namespace {
     // margin and the improving flag are used in various pruning heuristics.
     improvement =   (ss-2)->staticEval != VALUE_NONE ? ss->staticEval - (ss-2)->staticEval
                   : (ss-4)->staticEval != VALUE_NONE ? ss->staticEval - (ss-4)->staticEval
-                  :                                    156;
+                  :                                    163;
     improving = improvement > 0;
 
     // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
-    if (eval < alpha - 426 - 256 * depth * depth)
+    if (eval < alpha - 467 - 266 * depth * depth)
     {
         value = qsearch<NonPV>(pos, ss, alpha - 1, alpha);
         if (value < alpha)
@@ -769,18 +769,18 @@ namespace {
     // The depth condition is important for mate finding.
     if (   !ss->ttPv
         &&  depth < 9
-        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 280 >= beta
+        &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 306 >= beta
         &&  eval >= beta
-        &&  eval < 25128) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
+        &&  eval < 22761) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
         return eval;
 
     // Step 9. Null move search with verification search (~35 Elo)
     if (   !PvNode
         && (ss-1)->currentMove != MOVE_NULL
-        && (ss-1)->statScore < 18755
+        && (ss-1)->statScore < 18404
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 20 * depth - improvement / 13 + 253
+        &&  ss->staticEval >= beta - 19 * depth - improvement / 13 + 257
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly))
@@ -823,13 +823,13 @@ namespace {
         }
     }
 
-    probCutBeta = beta + 186 - 54 * improving;
+    probCutBeta = beta + 174 - 60 * improving;
 
     // Step 10. ProbCut (~10 Elo)
     // If we have a good enough capture (or queen promotion) and a reduced search returns a value
     // much above beta, we can (almost) safely prune the previous move.
     if (   !PvNode
-        &&  depth > 4
+        &&  depth > 3
         &&  abs(beta) < VALUE_TB_WIN_IN_MAX_PLY
         // if value from transposition table is lower than probCutBeta, don't attempt probCut
         // there and in further interactions with transposition table cutoff depth is set to depth - 3
@@ -887,20 +887,20 @@ namespace {
         return qsearch<PV>(pos, ss, alpha, beta);
 
     if (    cutNode
-        &&  depth >= 7
+        &&  depth >= 8
         && !ttMove)
         depth -= 2;
 
 moves_loop: // When in check, search starts here
 
     // Step 12. A small Probcut idea, when we are in check (~4 Elo)
-    probCutBeta = beta + 391;
+    probCutBeta = beta + 430;
     if (   ss->inCheck
         && !PvNode
         && depth >= 2
         && ttCapture
         && (tte->bound() & BOUND_LOWER)
-        && tte->depth() >= depth - 3
+        && tte->depth() >= depth - 4
         && ttValue >= probCutBeta
         && abs(ttValue) <= VALUE_KNOWN_WIN
         && abs(beta) <= VALUE_KNOWN_WIN)
@@ -986,15 +986,15 @@ moves_loop: // When in check, search starts here
           {
               // Futility pruning for captures (~2 Elo)
               if (   !givesCheck
-                  && lmrDepth < 6
+                  && lmrDepth < 7
                   && !ss->inCheck
-                  && ss->staticEval + 182 + 230 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
-                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
+                  && ss->staticEval + 207 + 223 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
+                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] * 1078 / 7000 < alpha)
                   continue;
 
               Bitboard occupied;
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, occupied, Value(-206) * depth))
+              if (!pos.see_ge(move, occupied, Value(-205) * depth))
               {
                   if (depth < 2 - capture)
                       continue;
@@ -1022,24 +1022,24 @@ moves_loop: // When in check, search starts here
 
               // Continuation history based pruning (~2 Elo)
               if (   lmrDepth < 5
-                  && history < -4405 * (depth - 1))
+                  && history < -3792 * (depth - 1))
                   continue;
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
 
-              lmrDepth += history / 7278;
+              lmrDepth += history / 7019;
               lmrDepth = std::max(lmrDepth, -2);
 
               // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
-                  && lmrDepth < 13
-                  && ss->staticEval + 103 + 138 * lmrDepth <= alpha)
+                  && lmrDepth < 12
+                  && ss->staticEval + 111 + 136 * lmrDepth <= alpha)
                   continue;
 
               lmrDepth = std::max(lmrDepth, 0);
 
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-24 * lmrDepth * lmrDepth - 16 * lmrDepth)))
+              if (!pos.see_ge(move, Value(-27 * lmrDepth * lmrDepth - 33 * lmrDepth / 2)))
                   continue;
           }
       }
@@ -1054,7 +1054,7 @@ moves_loop: // When in check, search starts here
           // a reduced search on all the other moves but the ttMove and if the
           // result is lower than ttValue minus a margin, then we will extend the ttMove.
           if (   !rootNode
-              &&  depth >= 4 - (thisThread->completedDepth > 21) + 2 * (PvNode && tte->is_pv())
+              &&  depth >= 4 - (thisThread->completedDepth > 22) + 2 * (PvNode && tte->is_pv())
               &&  move == ttMove
               && !excludedMove // Avoid recursive singular search
            /* &&  ttValue != VALUE_NONE Already implicit in the next condition */
@@ -1062,7 +1062,7 @@ moves_loop: // When in check, search starts here
               && (tte->bound() & BOUND_LOWER)
               &&  tte->depth() >= depth - 3)
           {
-              Value singularBeta = ttValue - (3 + 2 * (ss->ttPv && !PvNode)) * depth / 2;
+              Value singularBeta = ttValue - (99 + 65 * (ss->ttPv && !PvNode)) * depth / 64;
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;
@@ -1076,8 +1076,8 @@ moves_loop: // When in check, search starts here
 
                   // Avoid search explosion by limiting the number of double extensions
                   if (  !PvNode
-                      && value < singularBeta - 25
-                      && ss->doubleExtensions <= 10)
+                      && value < singularBeta - 22
+                      && ss->doubleExtensions <= 11)
                   {
                       extension = 2;
                       depth += depth < 13;
@@ -1107,15 +1107,15 @@ moves_loop: // When in check, search starts here
 
           // Check extensions (~1 Elo)
           else if (   givesCheck
-                   && depth > 10
-                   && abs(ss->staticEval) > 88)
+                   && depth > 9
+                   && abs(ss->staticEval) > 87)
               extension = 1;
 
           // Quiet ttMove extensions (~1 Elo)
           else if (   PvNode
                    && move == ttMove
                    && move == ss->killers[0]
-                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5705)
+                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5480)
               extension = 1;
       }
 
@@ -1144,7 +1144,7 @@ moves_loop: // When in check, search starts here
           r -= cutNode && tte->depth() >= depth + 3 ? 3 : 2;
 
       // Decrease reduction if opponent's move count is high (~1 Elo)
-      if ((ss-1)->moveCount > 7)
+      if ((ss-1)->moveCount > 8)
           r--;
 
       // Increase reduction for cut nodes (~3 Elo)
@@ -1157,7 +1157,7 @@ moves_loop: // When in check, search starts here
 
       // Decrease reduction for PvNodes based on depth (~2 Elo)
       if (PvNode)
-          r -= 1 + 12 / (3 + depth);
+          r -= 1 + 11 / (3 + depth);
 
       // Decrease reduction if ttMove has been singularly extended (~1 Elo)
       if (singularQuietLMR)
@@ -1174,10 +1174,10 @@ moves_loop: // When in check, search starts here
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]
                      + (*contHist[3])[movedPiece][to_sq(move)]
-                     - 4082;
+                     - 3755;
 
       // Decrease/increase reduction for moves with a good/bad history (~25 Elo)
-      r -= ss->statScore / (11079 + 4626 * (depth > 6 && depth < 19));
+      r -= ss->statScore / (10445 + 4762 * (depth > 6 && depth < 21));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has
@@ -1201,8 +1201,8 @@ moves_loop: // When in check, search starts here
           {
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
-              const bool doDeeperSearch = value > (bestValue + 68 + 12 * (newDepth - d));
-              const bool doEvenDeeperSearch = value > alpha + 588 && ss->doubleExtensions <= 5;
+              const bool doDeeperSearch = value > (bestValue + 63 + 11 * (newDepth - d));
+              const bool doEvenDeeperSearch = value > alpha + 662 && ss->doubleExtensions <= 6;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
               ss->doubleExtensions = ss->doubleExtensions + doEvenDeeperSearch;
@@ -1227,7 +1227,7 @@ moves_loop: // When in check, search starts here
           if (!ttMove && cutNode)
               r += 2;
 
-          value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 4), !cutNode);
+          value = -search<NonPV>(pos, ss+1, -(alpha+1), -alpha, newDepth - (r > 3), !cutNode);
       }
 
       // For PV nodes only, do a full PV search on the first move or after a fail
@@ -1320,8 +1320,8 @@ moves_loop: // When in check, search starts here
               {
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   if (   depth > 1
-                      && beta  <  12535
-                      && value > -12535)
+                      && beta  <  14001
+                      && value > -12754)
                       depth -= 1;
 
                   assert(depth > 0);
@@ -1370,7 +1370,7 @@ moves_loop: // When in check, search starts here
     // Bonus for prior countermove that caused the fail low
     else if (!priorCapture && prevSq != SQ_NONE)
     {
-        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 97 * depth) + ((ss-1)->moveCount > 10);
+        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 100 * depth) + ((ss-1)->moveCount > 11);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 
@@ -1499,7 +1499,7 @@ moves_loop: // When in check, search starts here
         if (PvNode && bestValue > alpha)
             alpha = bestValue;
 
-        futilityBase = bestValue + 168;
+        futilityBase = bestValue + 190;
     }
 
     const PieceToHistory* contHist[] = { (ss-1)->continuationHistory, (ss-2)->continuationHistory,
@@ -1572,7 +1572,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, Value(-110)))
+      if (!pos.see_ge(move, Value(-94)))
           continue;
     }
 
@@ -1705,7 +1705,7 @@ moves_loop: // When in check, search starts here
 
     if (!pos.capture_stage(bestMove))
     {
-        int bonus2 = bestValue > beta + 153 ? bonus1               // larger bonus
+        int bonus2 = bestValue > beta + 143 ? bonus1               // larger bonus
                                             : stat_bonus(depth);   // smaller bonus
 
         // Increase stats for the best move in case it was a quiet move

From 4f24ee086828e28df7d0b2dce5c13732139e7c19 Mon Sep 17 00:00:00 2001
From: Stefan Geschwentner <stgeschwentner@gmail.com>
Date: Fri, 19 May 2023 23:16:48 +0200
Subject: [PATCH 149/192] Small simplification in history pruning.

Remove the constant term of the history threshold which lowers the chance that pruning occurs.
As compensation allow pruning at a slightly higher depth.

Passed STC:
https://tests.stockfishchess.org/tests/view/64634c9a87f6567dd4df4901
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 101536 W: 27156 L: 27012 D: 47368
Ptnml(0-2): 266, 11165, 27772, 11289, 276

Passed LTC:
https://tests.stockfishchess.org/tests/view/6463d68b17982fde89d2bc2b
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 32154 W: 8741 L: 8543 D: 14870
Ptnml(0-2): 8, 3093, 9687, 3271, 18

Passed LTC: retest on top of VLTC tuning PR 4571 because this changes the history depth factor (use this new factor here)
https://tests.stockfishchess.org/tests/view/6467300e165c4b29ec0afd3f
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 99270 W: 26840 L: 26707 D: 45723
Ptnml(0-2): 36, 9753, 29928, 9878, 40

closes https://github.com/official-stockfish/Stockfish/pull/4578

Bench: 2984341
---
 src/search.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 429db9a5..10844746 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1021,8 +1021,8 @@ moves_loop: // When in check, search starts here
                             + (*contHist[3])[movedPiece][to_sq(move)];
 
               // Continuation history based pruning (~2 Elo)
-              if (   lmrDepth < 5
-                  && history < -3792 * (depth - 1))
+              if (   lmrDepth < 6
+                  && history < -3792 * depth)
                   continue;
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];

From 4b085c4777c36939bd0a598f4bc3e0c04606e31b Mon Sep 17 00:00:00 2001
From: xoto10 <23479932+xoto10@users.noreply.github.com>
Date: Fri, 19 May 2023 19:58:18 +0100
Subject: [PATCH 150/192] Simplify optimism calculation

This change removes one of the constants in the calculation of optimism. It also changes the 2 constants used with the scale value so that they are independent, instead of applying a constant to the scale and then adjusting it again when it is applied to the optimism. This might make the tuning of these constants cleaner and more reliable in the future.

STC 10+0.1 (accidentally run as an Elo gainer:
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 154080 W: 41119 L: 40651 D: 72310
Ptnml(0-2): 375, 16840, 42190, 17212, 423
https://tests.stockfishchess.org/tests/live_elo/64653eabf3b1a4e86c317f77

LTC 60+0.6:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 217434 W: 58382 L: 58363 D: 100689
Ptnml(0-2): 66, 21075, 66419, 21088, 69
https://tests.stockfishchess.org/tests/live_elo/6465d077f3b1a4e86c318d6c

closes https://github.com/official-stockfish/Stockfish/pull/4576

bench: 3190961
---
 src/evaluate.cpp | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index d8f4e2e1..cc789e35 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1063,7 +1063,7 @@ Value Eval::evaluate(const Position& pos) {
   else
   {
       int nnueComplexity;
-      int scale = 967 + pos.non_pawn_material() / 64;
+      int npm = pos.non_pawn_material() / 64;
 
       Color stm = pos.side_to_move();
       Value optimism = pos.this_thread()->optimism[stm];
@@ -1071,12 +1071,12 @@ Value Eval::evaluate(const Position& pos) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  402 * nnueComplexity
-                        + (454 + optimism) * abs(psq - nnue)
+      nnueComplexity = (  397 * nnueComplexity
+                        + (477 + optimism) * abs(psq - nnue)
                         ) / 1024;
 
-      optimism = optimism * (274 + nnueComplexity) / 256;
-      v = (nnue * scale + optimism * (scale - 791)) / 1024;
+      optimism += optimism * nnueComplexity / 256;
+      v = (nnue * (945 + npm) + optimism * (174 + npm)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling

From 7cd650f435715f73550d1f8031315e65b701d631 Mon Sep 17 00:00:00 2001
From: pb00067 <pb00067@phxl0401.wp.lan>
Date: Wed, 17 May 2023 09:22:02 +0200
Subject: [PATCH 151/192] Simplify SEE verfication logic

Passed STC
https://tests.stockfishchess.org/tests/view/6461d51887f6567dd4df27d0
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 177056 W: 47181 L: 47118 D: 82757
Ptnml(0-2): 456, 19381, 48792, 19442, 457

Passed LTC
https://tests.stockfishchess.org/tests/view/64631a9287f6567dd4df4502
2.94 (-2.94,2.94) <-1.75,0.25>
Total: 104346 W: 28062 L: 27935 D: 48349
Ptnml(0-2): 25, 10190, 31631, 10287, 40

closes https://github.com/official-stockfish/Stockfish/pull/4578

bench: 2903251
---
 src/search.cpp | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 10844746..653cbf33 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -996,9 +996,7 @@ moves_loop: // When in check, search starts here
               // SEE based pruning (~11 Elo)
               if (!pos.see_ge(move, occupied, Value(-205) * depth))
               {
-                  if (depth < 2 - capture)
-                      continue;
-                  // Don't prune the move if opp. King/Queen/Rook gets a discovered attack during or after the exchanges
+                  // Don't prune the move if opponent King/Queen/Rook gets a discovered attack during or after the exchanges
                   Bitboard leftEnemies = pos.pieces(~us, KING, QUEEN, ROOK);
                   Bitboard attacks = 0;
                   occupied |= to_sq(move);
@@ -1006,7 +1004,7 @@ moves_loop: // When in check, search starts here
                   {
                       Square sq = pop_lsb(leftEnemies);
                       attacks = pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
-                      // Exclude Queen/Rook(s) which were already threatened before SEE (opp King can't be in check when it's our turn)
+                      // Exclude Queen/Rook(s) which were already threatened before SEE (opponent King can't be in check when it's our turn)
                       if (attacks && sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us)))
                           attacks = 0;
                   }

From df0fb8471e5015bb4ba0b398c203b7faad45840e Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Sun, 14 May 2023 00:30:35 +0300
Subject: [PATCH 152/192] Small simplification in low depth pruning

Uncap low depth pruning lmr depth.
It's anyway capped for most cases apart from futility pruning for captures - removes one std::min call.

Passed STC:
https://tests.stockfishchess.org/tests/view/645e8fa6d55cccb2e64225a1
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 184064 W: 49039 L: 48982 D: 86043
Ptnml(0-2): 462, 20353, 50349, 20402, 466

Passed LTC:
https://tests.stockfishchess.org/tests/view/645f4d48d55cccb2e6423335
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 83886 W: 22613 L: 22465 D: 38808
Ptnml(0-2): 31, 8090, 25546, 8252, 24

closes https://github.com/official-stockfish/Stockfish/pull/4566

bench 3201883
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 653cbf33..f58def60 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -979,7 +979,7 @@ moves_loop: // When in check, search starts here
           moveCountPruning = moveCount >= futility_move_count(improving, depth);
 
           // Reduced depth of the next LMR search
-          int lmrDepth = std::max(newDepth - r, 0);
+          int lmrDepth = newDepth - r;
 
           if (   capture
               || givesCheck)

From d7e72d801fd68f2ee3c7d6b814bbc82916c30041 Mon Sep 17 00:00:00 2001
From: candirufish <38038147+candirufish@users.noreply.github.com>
Date: Mon, 22 May 2023 00:14:55 +0200
Subject: [PATCH 153/192] More Depth Reduction

Reduce more for depth > 3 and depth < 12

LTC: https://tests.stockfishchess.org/tests/view/646c5abbd1f14fd69a6f2fab
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 197280 W: 53405 L: 52797 D: 91078
Ptnml(0-2): 62, 19025, 59886, 19577, 90

STC: https://tests.stockfishchess.org/tests/view/646bee71d1f14fd69a6f259d
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 100832 W: 26861 L: 26466 D: 47505
Ptnml(0-2): 240, 10985, 27622, 11278, 291

https://github.com/official-stockfish/Stockfish/pull/4585

bench: 2276617
---
 src/search.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index f58def60..130855c1 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1317,10 +1317,11 @@ moves_loop: // When in check, search starts here
               else
               {
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
+                  // Reduce more for depth > 3 and depth < 12 (~1 Elo)
                   if (   depth > 1
                       && beta  <  14001
                       && value > -12754)
-                      depth -= 1;
+                      depth -= depth > 3 && depth < 12 ? 2 : 1;
 
                   assert(depth > 0);
                   alpha = value; // Update alpha! Always alpha < beta

From b64c97825eb473d9b5cbdb67afe65a8ac0d5ec9f Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Sat, 20 May 2023 21:28:54 +0800
Subject: [PATCH 154/192] Simplify delta calculation in aspiration window

Simplification STC: https://tests.stockfishchess.org/tests/view/6468cb200db5177f2b76ecbb
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 220416 W: 58503 L: 58487 D: 103426
Ptnml(0-2): 596, 24384, 60188, 24488, 552

Simplification LTC: https://tests.stockfishchess.org/tests/view/646a15840db5177f2b770704
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 177756 W: 47882 L: 47825 D: 82049
Ptnml(0-2): 55, 17430, 53858, 17473, 62

closes https://github.com/official-stockfish/Stockfish/pull/4581

Bench: 2304063
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 130855c1..3632a469 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -409,7 +409,7 @@ void Thread::search() {
               else
                   break;
 
-              delta += delta / 4 + 2;
+              delta += delta / 3;
 
               assert(alpha >= -VALUE_INFINITE && beta <= VALUE_INFINITE);
           }

From a989aa1825503ab39e6b2cf77bba2f1f022f367c Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Sun, 21 May 2023 16:22:28 +0300
Subject: [PATCH 155/192] Simplify Prune moves with negative SEE

Passed STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 57760 W: 15472 L: 15286 D: 27002
Ptnml(0-2): 123, 6025, 16430, 6147, 155
https://tests.stockfishchess.org/tests/view/6468eb6b0db5177f2b76ef62

Passed LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 93966 W: 25274 L: 25141 D: 43551
Ptnml(0-2): 33, 8498, 29792, 8623, 37
https://tests.stockfishchess.org/tests/view/6469570b0db5177f2b76f81b

closes: https://github.com/official-stockfish/Stockfish/pull/4579

Bench: 2304063
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 3632a469..270c5e7c 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1037,7 +1037,7 @@ moves_loop: // When in check, search starts here
               lmrDepth = std::max(lmrDepth, 0);
 
               // Prune moves with negative SEE (~4 Elo)
-              if (!pos.see_ge(move, Value(-27 * lmrDepth * lmrDepth - 33 * lmrDepth / 2)))
+              if (!pos.see_ge(move, Value(-27 * lmrDepth * lmrDepth - 16 * lmrDepth)))
                   continue;
           }
       }

From cedd73f4aa9f83c2891105695ac11e743e6ceab7 Mon Sep 17 00:00:00 2001
From: FauziAkram <fauzi.dabat@hotmail.com>
Date: Tue, 23 May 2023 20:36:47 +0300
Subject: [PATCH 156/192] Simplify Futility pruning for captures

Passed STC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 412928 W: 109433 L: 109620 D: 193875
Ptnml(0-2): 1071, 45929, 112650, 45744, 1070
https://tests.stockfishchess.org/tests/view/6468eac40db5177f2b76ef4d

Passed LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 190200 W: 51465 L: 51420 D: 87315
Ptnml(0-2): 58, 18585, 57788, 18592, 77
https://tests.stockfishchess.org/tests/view/646b66520db5177f2b772a84

closes https://github.com/official-stockfish/Stockfish/pull/4583

bench: 2486604
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 270c5e7c..32eeedab 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -989,7 +989,7 @@ moves_loop: // When in check, search starts here
                   && lmrDepth < 7
                   && !ss->inCheck
                   && ss->staticEval + 207 + 223 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
-                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] * 1078 / 7000 < alpha)
+                   + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
               Bitboard occupied;

From c701745cf243f4816754167e85bf5fabf5b34e47 Mon Sep 17 00:00:00 2001
From: cj5716 <125858804+cj5716@users.noreply.github.com>
Date: Thu, 25 May 2023 14:34:52 +0800
Subject: [PATCH 157/192] Remove ss->ttHit condition where ttValue !=
 VALUE_NONE

Simplification is done at 3 separate places in the code.
Thanks to peregrineshahin for helping me find 2 of such places.
(See original PR #4584)

Passed non-regression test
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 120256 W: 32204 L: 32085 D: 55967
Ptnml(0-2): 292, 12473, 34483, 12584, 296
https://tests.stockfishchess.org/tests/view/646f045968661bfd984325e3

closes https://github.com/official-stockfish/Stockfish/pull/4587

No functional change
---
 AUTHORS        | 1 +
 src/search.cpp | 9 +++------
 2 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index b6723246..884bffab 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -76,6 +76,7 @@ George Sobala (gsobala)
 gguliash
 Giacomo Lorenzetti (G-Lorenz)
 Gian-Carlo Pascutto (gcp)
+Goh CJ (cj5716)
 Gontran Lemaire (gonlem)
 Goodkov Vasiliy Aleksandrovich (goodkov)
 Gregor Cramer
diff --git a/src/search.cpp b/src/search.cpp
index 32eeedab..93212e23 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -615,10 +615,9 @@ namespace {
 
     // At non-PV nodes we check for an early TT cutoff
     if (  !PvNode
-        && ss->ttHit
         && !excludedMove
         && tte->depth() > depth - (tte->bound() == BOUND_EXACT)
-        && ttValue != VALUE_NONE // Possible in case of TT access race
+        && ttValue != VALUE_NONE // Possible in case of TT access race or if !ttHit
         && (tte->bound() & (ttValue >= beta ? BOUND_LOWER : BOUND_UPPER)))
     {
         // If ttMove is quiet, update move sorting heuristics on TT hit (~2 Elo)
@@ -835,8 +834,7 @@ namespace {
         // there and in further interactions with transposition table cutoff depth is set to depth - 3
         // because probCut search has depth set to depth - 4 but we also do a move before it
         // so effective depth is equal to depth - 3
-        && !(   ss->ttHit
-             && tte->depth() >= depth - 3
+        && !(   tte->depth() >= depth - 3
              && ttValue != VALUE_NONE
              && ttValue < probCutBeta))
     {
@@ -1453,9 +1451,8 @@ moves_loop: // When in check, search starts here
 
     // At non-PV nodes we check for an early TT cutoff
     if (  !PvNode
-        && ss->ttHit
         && tte->depth() >= ttDepth
-        && ttValue != VALUE_NONE // Only in case of TT access race
+        && ttValue != VALUE_NONE // Only in case of TT access race or if !ttHit
         && (tte->bound() & (ttValue >= beta ? BOUND_LOWER : BOUND_UPPER)))
         return ttValue;
 

From 7f0b19dedf7bff7dbe2dd42e73788826486b36b6 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sat, 27 May 2023 18:01:08 +0200
Subject: [PATCH 158/192] Update CPU contributors list

update CPU contributors list, the previous update was a couple of months ago,
and unfortunately, was not quite accurate for the number of games played.

This version is based clean calculation from the DB and
an updated script that tracks things (see https://github.com/glinscott/fishtest/pull/1702).

closes https://github.com/official-stockfish/Stockfish/pull/4589

No functional change
---
 Top CPU Contributors.txt | 196 ++++++++++++++++++++-------------------
 1 file changed, 101 insertions(+), 95 deletions(-)

diff --git a/Top CPU Contributors.txt b/Top CPU Contributors.txt
index 30c963d7..7b279590 100644
--- a/Top CPU Contributors.txt	
+++ b/Top CPU Contributors.txt	
@@ -1,212 +1,218 @@
-Contributors to Fishtest with >10,000 CPU hours, as of 2022-11-19.
+Contributors to Fishtest with >10,000 CPU hours, as of 2023-05-27.
 Thank you!
 
 Username                                CPU Hours     Games played
 ------------------------------------------------------------------
-noobpwnftw                               36475307       2748033975
-technologov                              14570711        760073590
+noobpwnftw                               37304027       2833556221
+technologov                              13508659        714674674
+linrock                                   4121386        280027751
 mlang                                     3026000        200065824
-dew                                       1689222        100034318
-grandphish2                               1442171         86798057
-okrout                                    1439985        133471766
-pemo                                      1405374         44189811
-linrock                                   1299003         28382783
-TueRens                                   1163420         71159522
-JojoM                                      897158         55177114
+dew                                       1689162        100033738
+okrout                                    1541122        145085726
+pemo                                      1481818         47546583
+grandphish2                               1459364         91364265
+TueRens                                   1178700         69951886
+JojoM                                      937875         60821044
 tvijlbrief                                 796125         51897690
 mibere                                     703840         46867607
-gvreuls                                    635982         40652394
-oz                                         590763         41201352
-sebastronomy                               581517         23307132
-cw                                         517915         34865769
-fastgm                                     504266         30264740
-CSU_Dynasty                                479901         31846710
-ctoks                                      433503         28180725
+sebastronomy                               687502         35585318
+gvreuls                                    645570         42437926
+oz                                         541224         39133532
+cw                                         517856         34869499
+fastgm                                     503862         30260818
+CSU_Dynasty                                464691         31166478
+leszek                                     460426         32840277
+ctoks                                      434323         28497451
 crunchy                                    427035         27344275
-leszek                                     416883         27493447
-bcross                                     409982         28062127
-velislav                                   345954         22232274
+maximmasiutin                              424154         26534660
+bcross                                     415722         29060963
+rpngn                                      344368         24218047
+velislav                                   342559         22138408
 Fisherman                                  327231         21829379
+mgrabiak                                   297057         20260882
 Dantist                                    296386         18031762
-mgrabiak                                   288928         18869896
-rpngn                                      259965         16281463
-robal                                      237653         15148350
-ncfish1                                    231764         15275003
-nordlandia                                 226923         14624832
+nordlandia                                 242642         15922516
+robal                                      240199         15544104
+marrco                                     234581         17714473
+ncfish1                                    227517         15233777
 glinscott                                  208125         13277240
 drabel                                     204167         13930674
 mhoram                                     202894         12601997
 bking_US                                   198894         11876016
-thirdlife                                  198844          5453268
+olafm                                      192342         14968698
 Thanar                                     179852         12365359
 vdv                                        175544          9904472
-armo9494                                   168201         11136452
 spams                                      157128         10319326
-marrco                                     151599          9551115
 sqrt2                                      147963          9724586
-vdbergh                                    137690          8971569
+DesolatedDodo                              144759          9408038
+Calis007                                   143165          9478764
+vdbergh                                    138436          9042073
 CoffeeOne                                  137100          5024116
 malala                                     136182          8002293
-DesolatedDodo                              135276          8657464
+armo9494                                   136010          9447548
 xoto                                       133759          9159372
 davar                                      129023          8376525
+DMBK                                       122960          8980062
 dsmith                                     122059          7570238
 amicic                                     119661          7938029
 Data                                       113305          8220352
 BrunoBanani                                112960          7436849
 CypressChess                               108331          7759788
-skiminki                                   106518          7062598
+skiminki                                   107583          7218170
 MaZePallas                                 102823          6633619
 sterni1971                                 100532          5880772
+jcAEie                                     100392          7788270
 sunu                                       100167          7040199
 zeryl                                       99331          6221261
+thirdlife                                   99124          2242380
 ElbertoOne                                  99028          7023771
-DMBK                                        97572          6950312
-Calis007                                    96779          5611552
-cuistot                                     93111          5536500
+cuistot                                     98360          6017102
+bigpen0r                                    94809          6529203
 brabos                                      92118          6186135
-Wolfgang                                    91769          5720158
+Wolfgang                                    90855          5998076
 psk                                         89957          5984901
 racerschmacer                               85805          6122790
-jcAEie                                      85527          5630616
+Dubslow                                     84986          6042456
 Vizvezdenec                                 83761          5344740
-sschnee                                     83557          4853690
+sschnee                                     83564          4853834
 0x3C33                                      82614          5271253
 BRAVONE                                     81239          5054681
-Dubslow                                     78461          5042980
+Fifis                                       77355          5158211
 nssy                                        76497          5259388
 jromang                                     76106          5236025
 teddybaer                                   75125          5407666
-yurikvelo                                   73933          5031096
-tolkki963                                   73885          4721430
+Wencey                                      74181          4711488
+megaman7de                                  73866          4894960
 Pking_cda                                   73776          5293873
-Bobo1239                                    71675          4860987
+tolkki963                                   73531          5020500
+yurikvelo                                   72847          4972808
+Bobo1239                                    70579          4794999
 solarlight                                  70517          5028306
 dv8silencer                                 70287          3883992
-Gelma                                       69304          3980932
+markkulix                                   70278          5068326
 manap                                       66273          4121774
-megaman7de                                  65419          4120200
-markkulix                                   65331          4114860
-bigpen0r                                    64932          4683883
 tinker                                      64333          4268790
 qurashee                                    61208          3429862
-AGI                                         58325          4258646
+Mineta                                      58759          4399960
+AGI                                         58147          4325994
+Spprtr                                      58106          3858759
 robnjr                                      57262          4053117
 Freja                                       56938          3733019
 MaxKlaxxMiner                               56879          3423958
+MarcusTullius                               56746          3762951
 ttruscott                                   56010          3680085
 rkl                                         55132          4164467
 renouve                                     53811          3501516
-Spprtr                                      52736          3410019
 finfish                                     51360          3370515
 eva42                                       51272          3599691
 eastorwest                                  51117          3454811
 rap                                         49985          3219146
-unixwizard                                  49734          2536230
-pb00067                                     49727          3298270
+pb00067                                     49733          3298934
+javran                                      49178          4190632
+OuaisBla                                    48606          3442958
 ronaldjerum                                 47654          3240695
 biffhero                                    46564          3111352
-GPUex                                       45861          2926502
-Fifis                                       45843          3088497
-oryx                                        45578          3493978
 VoyagerOne                                  45476          3452465
-Wencey                                      44943          2654490
+oryx                                        44532          3450170
+jmdana                                      43849          2955821
 speedycpu                                   43842          3003273
 jbwiebe                                     43305          2805433
 Antihistamine                               41788          2761312
 mhunt                                       41735          2691355
-olafm                                       41277          3284344
+maposora                                    41534          3733078
+GPUex                                       41061          2998356
 homyur                                      39893          2850481
 gri                                         39871          2515779
-MarcusTullius                               38303          2251097
 Garf                                        37741          2999686
-kdave                                       37424          2557406
 SC                                          37299          2731694
 csnodgrass                                  36207          2688994
-jmdana                                      36157          2210661
 strelock                                    34716          2074055
 EthanOConnor                                33370          2090311
 slakovv                                     32915          2021889
-gopeto                                      31669          2060958
+Gelma                                       31771          1551204
+gopeto                                      31671          2060990
+szupaw                                      31248          2594920
+kdave                                       31157          2198362
 manapbk                                     30987          1810399
 Prcuvu                                      30377          2170122
 anst                                        30301          2190091
 jkiiski                                     30136          1904470
-spcc                                        30135          1903728
+spcc                                        29925          1901692
 hyperbolic.tom                              29840          2017394
-xwziegtm                                    29763          2347412
 chuckstablers                               29659          2093438
 Pyafue                                      29650          1902349
 belzedar94                                  28846          1811530
-OuaisBla                                    27636          1578800
 chriswk                                     26902          1868317
+xwziegtm                                    26897          2124586
 achambord                                   26582          1767323
 Patrick_G                                   26276          1801617
 yorkman                                     26193          1992080
-Ulysses                                     25289          1674274
+Ulysses                                     25285          1689346
 SFTUser                                     25182          1675689
 nabildanial                                 24942          1519409
 Sharaf_DG                                   24765          1786697
 rodneyc                                     24376          1416402
 agg177                                      23890          1395014
-Ente                                        23747          1674582
-Karpovbot                                   23629          1313186
+Ente                                        23639          1671638
 JanErik                                     23408          1703875
 Isidor                                      23388          1680691
 Norabor                                     23371          1603244
-cisco2015                                   22934          1763773
+Goatminola                                  23338          1910634
+cisco2015                                   22920          1763301
+Jopo12321                                   22890          1424926
 Zirie                                       22542          1472937
 team-oh                                     22272          1636708
 Roady                                       22220          1465606
 MazeOfGalious                               21978          1629593
 sg4032                                      21947          1643353
+jsys14                                      21935          1499128
 ianh2105                                    21725          1632562
 xor12                                       21628          1680365
 dex                                         21612          1467203
 nesoneg                                     21494          1463031
 user213718                                  21454          1404128
-AndreasKrug                                 21227          1577833
 sphinx                                      21211          1384728
 jjoshua2                                    21001          1423089
+Zake9298                                    20938          1565848
+AndreasKrug                                 20911          1615673
 horst.prack                                 20878          1465656
-jsys14                                      20729          1221010
 0xB00B1ES                                   20590          1208666
 j3corre                                     20405           941444
 Adrian.Schmidt123                           20316          1281436
-bonsi                                       20022          1300682
 wei                                         19973          1745989
-dapper                                      19754          1167758
-Zake9298                                    19745          1458416
+Serpensin                                   19840          1697528
 fishtester                                  19617          1257388
 rstoesser                                   19569          1293588
 eudhan                                      19274          1283717
+Gaster319                                   18934          1596772
 vulcan                                      18871          1729392
-Jopo12321                                   18803          1036284
+Karpovbot                                   18766          1053178
 jundery                                     18445          1115855
+votoanthuan                                 18012          1508836
 ville                                       17883          1384026
-5t0ckf15hTr4in3r                            17809          1105858
 chris                                       17698          1487385
-dju                                         17697           994333
 purplefishies                               17595          1092533
+qoo_charly_cai                              17494          1182667
+dju                                         17414           981289
 iisiraider                                  17275          1049015
 DragonLord                                  17014          1162790
-Karby                                       16457          1010138
-Goatminola                                  16278          1145026
+redstone59                                  16842          1461780
+Alb11747                                    16787          1213926
 IgorLeMasson                                16064          1147232
-Gaster319                                   16056          1109070
-redstone59                                  15953          1161664
+Karby                                       15982           979610
+notchris                                    15818          1426762
 scuzzi                                      15757           968735
 ako027ako                                   15671          1173203
 Nikolay.IT                                  15154          1068349
 Andrew Grant                                15114           895539
 Naven94                                     15054           834762
 OssumOpossum                                14857          1007129
-qoo_charly_cai                              14490           847865
+ZacHFX                                      14783          1021842
 enedene                                     14476           905279
-szupaw                                      14252           929130
 bpfliegel                                   14233           882523
 mpx86                                       14019           759568
 jpulman                                     13982           870599
+Skiff84                                     13826           721996
 crocogoat                                   13803          1117422
 Nesa92                                      13786          1114691
 joster                                      13710           946160
@@ -214,47 +220,47 @@ mbeier                                      13650          1044928
 Hjax                                        13535           915487
 Dark_wizzie                                 13422          1007152
 Rudolphous                                  13244           883140
+pirt                                        13100          1009897
 Machariel                                   13010           863104
 infinigon                                   12991           943216
-pirt                                        12925           985437
-Skiff84                                     12923           649994
+Maxim                                       12963           985594
 mabichito                                   12903           749391
 thijsk                                      12886           722107
 AdrianSA                                    12860           804972
 Flopzee                                     12698           894821
+korposzczur                                 12606           838168
+Nullvalue                                   12583          1048502
 fatmurphy                                   12547           853210
-woutboat                                    12419           836696
 SapphireBrand                               12416           969604
-Oakwen                                      12406           840961
+Oakwen                                      12399           844109
 deflectooor                                 12386           579392
 modolief                                    12386           896470
 Farseer                                     12249           694108
+Jackfish                                    12180           801372
 pgontarz                                    12151           848794
+dbernier                                    12103           860824
+getraideBFF                                 12072          1024966
 stocky                                      11954           699440
 mschmidt                                    11941           803401
-MooTheCow                                   11871           773654
-Jackfish                                    11867           773550
-dbernier                                    11705           821780
+MooTheCow                                   11870           773598
+FormazChar                                  11689           877727
 whelanh                                     11557           245188
-Maxim                                       11543           836024
-Nullvalue                                   11534           731410
-icewulf                                     11528           650470
-FormazChar                                  11523           861599
 infinity                                    11470           727027
 aga                                         11412           695127
 torbjo                                      11395           729145
 Thomas A. Anderson                          11372           732094
 savage84                                    11358           670860
-ali-al-zhrani                               11272           781310
 d64                                         11263           789184
-Bourbaki                                    11108           709144
+ali-al-zhrani                               11245           779246
 snicolet                                    11106           869170
-Alb11747                                    10855           696920
+dapper                                      11032           771402
+Karmatron                                   10828           677458
 basepi                                      10637           744851
 Cubox                                       10621           826448
-Karmatron                                   10616           674818
 michaelrpg                                  10509           739239
 OIVAS7572                                   10420           995586
-Garruk                                      10348           704905
+jojo2357                                    10419           929708
+WoodMan777                                  10380           873720
+Garruk                                      10365           706465
 dzjp                                        10343           732529
 ols                                         10259           570669

From 7e9b131efb832339ee6cd9e22b7c837c3e69a1b5 Mon Sep 17 00:00:00 2001
From: windfishballad <windfishballadtcec@gmail.com>
Date: Mon, 22 May 2023 20:13:44 -0400
Subject: [PATCH 159/192] Removed quadratic term in optimism

Remove term which is quadratic in optimism in the eval.
Simplifies and should also remove the bias towards side to move making the eval better for analysis.

STC: https://tests.stockfishchess.org/tests/view/6470a9d8c29e0d4352b0bca5
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 154432 W: 41127 L: 41040 D: 72265
Ptnml(0-2): 380, 17094, 42190, 17163, 389

LTC: https://tests.stockfishchess.org/tests/view/6471e9b3e549d9cf2fb219ef
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 127926 W: 34474 L: 34369 D: 59083
Ptnml(0-2): 43, 12505, 38776, 12582, 57

closes https://github.com/official-stockfish/Stockfish/pull/4590

Bench: 2541211
---
 AUTHORS          | 1 +
 src/evaluate.cpp | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/AUTHORS b/AUTHORS
index 884bffab..d01d23cd 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -213,6 +213,7 @@ tttak
 Unai Corzo (unaiic)
 Uri Blass (uriblass)
 Vince Negri (cuddlestmonkey)
+windfishballad
 xefoci7612
 zz4032
 
diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index cc789e35..7239fd1e 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1072,7 +1072,7 @@ Value Eval::evaluate(const Position& pos) {
 
       // Blend nnue complexity with (semi)classical complexity
       nnueComplexity = (  397 * nnueComplexity
-                        + (477 + optimism) * abs(psq - nnue)
+                        + 477 * abs(psq - nnue)
                         ) / 1024;
 
       optimism += optimism * nnueComplexity / 256;

From c1fff71650e2f8bf5a2d63bdc043161cdfe8e460 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Fri, 12 May 2023 18:07:20 -0400
Subject: [PATCH 160/192] Update NNUE architecture to SFNNv6 with larger L1
 size of 1536

Created by training a new net from scratch with L1 size increased from 1024 to 1536.
Thanks to Vizvezdenec for the idea of exploring larger net sizes after recent
training data improvements.

A new net was first trained with lambda 1.0 and constant LR 8.75e-4. Then a strong net
from a later epoch in the training run was chosen for retraining with start-lambda 1.0
and initial LR 4.375e-4 decaying with gamma 0.995. Retraining was performed a total of
3 times, for this 4-step process:

1. 400 epochs, lambda 1.0 on filtered T77+T79 v6 deduplicated data
2. 800 epochs, end-lambda 0.75 on T60T70wIsRightFarseerT60T74T75T76.binpack
3. 800 epochs, end-lambda 0.75 and early-fen-skipping 28 on the master dataset
4. 800 epochs, end-lambda 0.7 and early-fen-skipping 28 on the master dataset

In the training sequence that reached the new nn-8d69132723e2.nnue net,
the epochs used for the 3x retraining runs were:

1. epoch 379 trained on T77T79-filter-v6-dd.min.binpack
2. epoch 679 trained on T60T70wIsRightFarseerT60T74T75T76.binpack
3. epoch 799 trained on the master dataset

For training from scratch:

python3 easy_train.py \
  --experiment-name new-L1-1536-T77T79-filter-v6dd \
  --training-dataset /data/T77T79-filter-v6-dd.min.binpack \
  --max_epoch 400 \
  --lambda 1.0 \
  --start-from-engine-test-net False \
  --engine-test-branch linrock/Stockfish/L1-1536 \
  --nnue-pytorch-branch linrock/Stockfish/misc-fixes-L1-1536 \
  --tui False \
  --gpus "0," \
  --seed $RANDOM

Retraining commands were similar to each other. For the 3rd retraining run:

python3 easy_train.py \
  --experiment-name L1-1536-T77T79-v6dd-Re1-LeelaFarseer-Re2-masterDataset-Re3-sameData \
  --training-dataset /data/leela96-dfrc99-v2-T60novdecT80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd.binpack \
  --early-fen-skipping 28 \
  --max_epoch 800 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --start-from-engine-test-net False \
  --start-from-model /data/L1-1536-T77T79-v6dd-Re1-LeelaFarseer-Re2-masterDataset-nn-epoch799.nnue \
  --engine-test-branch linrock/Stockfish/L1-1536 \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes-L1-1536 \
  --tui False \
  --gpus "0," \
  --seed $RANDOM

The T77+T79 data used is a subset of the master dataset available at:
https://robotmoon.com/nnue-training-data/

T60T70wIsRightFarseerT60T74T75T76.binpack is available at:
https://drive.google.com/drive/folders/1S9-ZiQa_3ApmjBtl2e8SyHxj4zG4V8gG

Local elo at 25k nodes per move vs. nn-e1fb1ade4432.nnue (L1 size 1024):
nn-epoch759.nnue : 26.9 +/- 1.6

Failed STC
https://tests.stockfishchess.org/tests/view/64742485d29264e4cfa75f97
LLR: -2.94 (-2.94,2.94) <0.00,2.00>
Total: 13728 W: 3588 L: 3829 D: 6311
Ptnml(0-2): 71, 1661, 3610, 1482, 40

Failing LTC
https://tests.stockfishchess.org/tests/view/64752d7c4a36543c4c9f3618
LLR: -1.91 (-2.94,2.94) <0.50,2.50>
Total: 35424 W: 9522 L: 9603 D: 16299
Ptnml(0-2): 24, 3579, 10585, 3502, 22

Passed VLTC 180+1.8
https://tests.stockfishchess.org/tests/view/64752df04a36543c4c9f3638
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 47616 W: 13174 L: 12863 D: 21579
Ptnml(0-2): 13, 4261, 14952, 4566, 16

Passed VLTC SMP 60+0.6 th 8
https://tests.stockfishchess.org/tests/view/647446ced29264e4cfa761e5
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 19942 W: 5694 L: 5451 D: 8797
Ptnml(0-2): 6, 1504, 6707, 1749, 5

closes https://github.com/official-stockfish/Stockfish/pull/4593

bench 2222567
---
 src/evaluate.h               | 2 +-
 src/nnue/nnue_architecture.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index f5db1c1e..0990111c 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-e1fb1ade4432.nnue"
+  #define EvalFileDefaultName   "nn-8d69132723e2.nnue"
 
   namespace NNUE {
 
diff --git a/src/nnue/nnue_architecture.h b/src/nnue/nnue_architecture.h
index 508f3aae..d10434f3 100644
--- a/src/nnue/nnue_architecture.h
+++ b/src/nnue/nnue_architecture.h
@@ -39,7 +39,7 @@ namespace Stockfish::Eval::NNUE {
 using FeatureSet = Features::HalfKAv2_hm;
 
 // Number of input feature dimensions after conversion
-constexpr IndexType TransformedFeatureDimensions = 1024;
+constexpr IndexType TransformedFeatureDimensions = 1536;
 constexpr IndexType PSQTBuckets = 8;
 constexpr IndexType LayerStacks = 8;
 

From 07bd8adcbce41f076c36f4b65c7f9a786de0b02d Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Sun, 28 May 2023 14:45:24 -0400
Subject: [PATCH 161/192] Simplify nnue eval complexity calculation

Remove a multiplier when blending nnue complexity with semi-classical complexity.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/6473a71dd29264e4cfa75839
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 124768 W: 33180 L: 33060 D: 58528
Ptnml(0-2): 314, 13797, 34030, 13941, 302

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/6474af3dd29264e4cfa768f4
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 108180 W: 29008 L: 28884 D: 50288
Ptnml(0-2): 29, 10420, 33075, 10530, 36

closes https://github.com/official-stockfish/Stockfish/pull/4592

bench 2316827
---
 src/evaluate.cpp | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 7239fd1e..40c43d23 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1071,9 +1071,7 @@ Value Eval::evaluate(const Position& pos) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = (  397 * nnueComplexity
-                        + 477 * abs(psq - nnue)
-                        ) / 1024;
+      nnueComplexity = 25 * (nnueComplexity + abs(psq - nnue)) / 64;
 
       optimism += optimism * nnueComplexity / 256;
       v = (nnue * (945 + npm) + optimism * (174 + npm)) / 1024;

From d99942f25449789de78c9d36e3dcb67d4eb04e98 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Wed, 31 May 2023 09:22:26 +0300
Subject: [PATCH 162/192] Small simplification for probcut in check

Remove depth condition from there as not longer needed.

Passed STC:
https://tests.stockfishchess.org/tests/view/647367cad29264e4cfa753e6
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 254336 W: 67830 L: 67847 D: 118659
Ptnml(0-2): 580, 28181, 69697, 28096, 614

Passed LTC:
https://tests.stockfishchess.org/tests/view/647576184a36543c4c9f3af7
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 80706 W: 22048 L: 21898 D: 36760
Ptnml(0-2): 28, 7721, 24712, 7857, 35

closes https://github.com/official-stockfish/Stockfish/pull/4594

bench 2381945
---
 src/search.cpp | 1 -
 1 file changed, 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 93212e23..41116eb2 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -895,7 +895,6 @@ moves_loop: // When in check, search starts here
     probCutBeta = beta + 430;
     if (   ss->inCheck
         && !PvNode
-        && depth >= 2
         && ttCapture
         && (tte->bound() & BOUND_LOWER)
         && tte->depth() >= depth - 4

From 06186b786e4a73a29d6f0eef80fa7e20084a1e85 Mon Sep 17 00:00:00 2001
From: Muzhen Gaming <61100393+XInTheDark@users.noreply.github.com>
Date: Fri, 2 Jun 2023 19:55:25 +0800
Subject: [PATCH 163/192] Search tuning at very long time control with new net
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The most significant change would be the singularBeta formula.
It was first tested by cj5716 (see https://tests.stockfishchess.org/tests/view/647317c9d29264e4cfa74ec7),
and I took much inspiration from that idea.

LTC (fixed games): https://tests.stockfishchess.org/tests/view/6479d8da54dd118e1d990b12
Elo: 0.61 ± 1.2 (95%) LOS: 83.4%
Total: 60000 W: 16278 L: 16172 D: 27550
Ptnml(0-2): 16, 5845, 18179, 5937, 23
nElo: 1.38 ± 2.8 (95%) PairsRatio: 1.02

VLTC 180+1.8: https://tests.stockfishchess.org/tests/view/6479da1454dd118e1d990b2b
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 33224 W: 9261 L: 8984 D: 14979
Ptnml(0-2): 5, 2809, 10710, 3080, 8

SMP VLTC 8-thread: https://tests.stockfishchess.org/tests/view/647b0fe354dd118e1d992425
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 61398 W: 17386 L: 17081 D: 26931
Ptnml(0-2): 7, 4571, 21232, 4888, 1

closes https://github.com/official-stockfish/Stockfish/pull/4603

Bench: 2805878
---
 src/search.cpp | 72 ++++++++++++++++++++++++--------------------------
 1 file changed, 35 insertions(+), 37 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 41116eb2..16122315 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -64,7 +64,7 @@ namespace {
 
   // Futility margin
   Value futility_margin(Depth d, bool improving) {
-    return Value(148 * (d - improving));
+    return Value(140 * (d - improving));
   }
 
   // Reductions lookup table, initialized at startup
@@ -72,7 +72,7 @@ namespace {
 
   Depth reduction(bool i, Depth d, int mn, Value delta, Value rootDelta) {
     int r = Reductions[d] * Reductions[mn];
-    return (r + 1356 - int(delta) * 983 / int(rootDelta)) / 1024 + (!i && r > 901);
+    return (r + 1372 - int(delta) * 1073 / int(rootDelta)) / 1024 + (!i && r > 936);
   }
 
   constexpr int futility_move_count(bool improving, Depth depth) {
@@ -82,7 +82,7 @@ namespace {
 
   // History and stats update bonus, based on depth
   int stat_bonus(Depth d) {
-    return std::min(337 * d - 497, 1632);
+    return std::min(336 * d - 547, 1561);
   }
 
   // Add a small random component to draw evaluations to avoid 3-fold blindness
@@ -162,7 +162,7 @@ namespace {
 void Search::init() {
 
   for (int i = 1; i < MAX_MOVES; ++i)
-      Reductions[i] = int((20.89 + std::log(Threads.size()) / 2) * std::log(i));
+      Reductions[i] = int((20.57 + std::log(Threads.size()) / 2) * std::log(i));
 }
 
 
@@ -348,12 +348,12 @@ void Thread::search() {
 
           // Reset aspiration window starting size
           Value prev = rootMoves[pvIdx].averageScore;
-          delta = Value(11) + int(prev) * prev / 15368;
+          delta = Value(10) + int(prev) * prev / 15799;
           alpha = std::max(prev - delta,-VALUE_INFINITE);
           beta  = std::min(prev + delta, VALUE_INFINITE);
 
           // Adjust optimism based on root move's previousScore
-          int opt = 116 * prev / (std::abs(prev) + 143);
+          int opt = 109 * prev / (std::abs(prev) + 141);
           optimism[ us] = Value(opt);
           optimism[~us] = -optimism[us];
 
@@ -741,7 +741,7 @@ namespace {
     // Use static evaluation difference to improve quiet move ordering (~4 Elo)
     if (is_ok((ss-1)->currentMove) && !(ss-1)->inCheck && !priorCapture)
     {
-        int bonus = std::clamp(-19 * int((ss-1)->staticEval + ss->staticEval), -1717, 1717);
+        int bonus = std::clamp(-18 * int((ss-1)->staticEval + ss->staticEval), -1817, 1817);
         thisThread->mainHistory[~us][from_to((ss-1)->currentMove)] << bonus;
     }
 
@@ -751,13 +751,13 @@ namespace {
     // margin and the improving flag are used in various pruning heuristics.
     improvement =   (ss-2)->staticEval != VALUE_NONE ? ss->staticEval - (ss-2)->staticEval
                   : (ss-4)->staticEval != VALUE_NONE ? ss->staticEval - (ss-4)->staticEval
-                  :                                    163;
+                  :                                    173;
     improving = improvement > 0;
 
     // Step 7. Razoring (~1 Elo).
     // If eval is really low check with qsearch if it can exceed alpha, if it can't,
     // return a fail low.
-    if (eval < alpha - 467 - 266 * depth * depth)
+    if (eval < alpha - 456 - 252 * depth * depth)
     {
         value = qsearch<NonPV>(pos, ss, alpha - 1, alpha);
         if (value < alpha)
@@ -770,16 +770,16 @@ namespace {
         &&  depth < 9
         &&  eval - futility_margin(depth, improving) - (ss-1)->statScore / 306 >= beta
         &&  eval >= beta
-        &&  eval < 22761) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
+        &&  eval < 24923) // larger than VALUE_KNOWN_WIN, but smaller than TB wins
         return eval;
 
     // Step 9. Null move search with verification search (~35 Elo)
     if (   !PvNode
         && (ss-1)->currentMove != MOVE_NULL
-        && (ss-1)->statScore < 18404
+        && (ss-1)->statScore < 17329
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 19 * depth - improvement / 13 + 257
+        &&  ss->staticEval >= beta - 21 * depth - improvement * 99 / 1300 + 258
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly))
@@ -787,7 +787,7 @@ namespace {
         assert(eval - beta >= 0);
 
         // Null move dynamic reduction based on depth and eval
-        Depth R = std::min(int(eval - beta) / 172, 6) + depth / 3 + 4;
+        Depth R = std::min(int(eval - beta) / 173, 6) + depth / 3 + 4;
 
         ss->currentMove = MOVE_NULL;
         ss->continuationHistory = &thisThread->continuationHistory[0][0][NO_PIECE][0];
@@ -822,7 +822,7 @@ namespace {
         }
     }
 
-    probCutBeta = beta + 174 - 60 * improving;
+    probCutBeta = beta + 168 - 61 * improving;
 
     // Step 10. ProbCut (~10 Elo)
     // If we have a good enough capture (or queen promotion) and a reduced search returns a value
@@ -892,7 +892,7 @@ namespace {
 moves_loop: // When in check, search starts here
 
     // Step 12. A small Probcut idea, when we are in check (~4 Elo)
-    probCutBeta = beta + 430;
+    probCutBeta = beta + 413;
     if (   ss->inCheck
         && !PvNode
         && ttCapture
@@ -985,13 +985,13 @@ moves_loop: // When in check, search starts here
               if (   !givesCheck
                   && lmrDepth < 7
                   && !ss->inCheck
-                  && ss->staticEval + 207 + 223 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
+                  && ss->staticEval + 197 + 248 * lmrDepth + PieceValue[EG][pos.piece_on(to_sq(move))]
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
               Bitboard occupied;
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, occupied, Value(-205) * depth))
+              if (!pos.see_ge(move, occupied, Value(-212) * depth))
               {
                   // Don't prune the move if opponent King/Queen/Rook gets a discovered attack during or after the exchanges
                   Bitboard leftEnemies = pos.pieces(~us, KING, QUEEN, ROOK);
@@ -1017,18 +1017,18 @@ moves_loop: // When in check, search starts here
 
               // Continuation history based pruning (~2 Elo)
               if (   lmrDepth < 6
-                  && history < -3792 * depth)
+                  && history < -3832 * depth)
                   continue;
 
               history += 2 * thisThread->mainHistory[us][from_to(move)];
 
-              lmrDepth += history / 7019;
+              lmrDepth += history / 7011;
               lmrDepth = std::max(lmrDepth, -2);
 
               // Futility pruning: parent node (~13 Elo)
               if (   !ss->inCheck
                   && lmrDepth < 12
-                  && ss->staticEval + 111 + 136 * lmrDepth <= alpha)
+                  && ss->staticEval + 112 + 138 * lmrDepth <= alpha)
                   continue;
 
               lmrDepth = std::max(lmrDepth, 0);
@@ -1057,7 +1057,7 @@ moves_loop: // When in check, search starts here
               && (tte->bound() & BOUND_LOWER)
               &&  tte->depth() >= depth - 3)
           {
-              Value singularBeta = ttValue - (99 + 65 * (ss->ttPv && !PvNode)) * depth / 64;
+              Value singularBeta = ttValue - (82 + 65 * (ss->ttPv && !PvNode)) * depth / 64;
               Depth singularDepth = (depth - 1) / 2;
 
               ss->excludedMove = move;
@@ -1071,7 +1071,7 @@ moves_loop: // When in check, search starts here
 
                   // Avoid search explosion by limiting the number of double extensions
                   if (  !PvNode
-                      && value < singularBeta - 22
+                      && value < singularBeta - 21
                       && ss->doubleExtensions <= 11)
                   {
                       extension = 2;
@@ -1101,16 +1101,14 @@ moves_loop: // When in check, search starts here
           }
 
           // Check extensions (~1 Elo)
-          else if (   givesCheck
-                   && depth > 9
-                   && abs(ss->staticEval) > 87)
+          else if (   givesCheck && depth > 8)
               extension = 1;
 
           // Quiet ttMove extensions (~1 Elo)
           else if (   PvNode
                    && move == ttMove
                    && move == ss->killers[0]
-                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5480)
+                   && (*contHist[0])[movedPiece][to_sq(move)] >= 5168)
               extension = 1;
       }
 
@@ -1152,7 +1150,7 @@ moves_loop: // When in check, search starts here
 
       // Decrease reduction for PvNodes based on depth (~2 Elo)
       if (PvNode)
-          r -= 1 + 11 / (3 + depth);
+          r -= 1 + 12 / (3 + depth);
 
       // Decrease reduction if ttMove has been singularly extended (~1 Elo)
       if (singularQuietLMR)
@@ -1169,10 +1167,10 @@ moves_loop: // When in check, search starts here
                      + (*contHist[0])[movedPiece][to_sq(move)]
                      + (*contHist[1])[movedPiece][to_sq(move)]
                      + (*contHist[3])[movedPiece][to_sq(move)]
-                     - 3755;
+                     - 4006;
 
       // Decrease/increase reduction for moves with a good/bad history (~25 Elo)
-      r -= ss->statScore / (10445 + 4762 * (depth > 6 && depth < 21));
+      r -= ss->statScore / (11124 + 4740 * (depth > 5 && depth < 22));
 
       // Step 17. Late moves reduction / extension (LMR, ~117 Elo)
       // We use various heuristics for the sons of a node after the first son has
@@ -1196,8 +1194,8 @@ moves_loop: // When in check, search starts here
           {
               // Adjust full depth search based on LMR results - if result
               // was good enough search deeper, if it was bad enough search shallower
-              const bool doDeeperSearch = value > (bestValue + 63 + 11 * (newDepth - d));
-              const bool doEvenDeeperSearch = value > alpha + 662 && ss->doubleExtensions <= 6;
+              const bool doDeeperSearch = value > (bestValue + 64 + 11 * (newDepth - d));
+              const bool doEvenDeeperSearch = value > alpha + 711 && ss->doubleExtensions <= 6;
               const bool doShallowerSearch = value < bestValue + newDepth;
 
               ss->doubleExtensions = ss->doubleExtensions + doEvenDeeperSearch;
@@ -1316,8 +1314,8 @@ moves_loop: // When in check, search starts here
                   // Reduce other moves if we have found at least one score improvement (~1 Elo)
                   // Reduce more for depth > 3 and depth < 12 (~1 Elo)
                   if (   depth > 1
-                      && beta  <  14001
-                      && value > -12754)
+                      && beta  <  14362
+                      && value > -12393)
                       depth -= depth > 3 && depth < 12 ? 2 : 1;
 
                   assert(depth > 0);
@@ -1366,7 +1364,7 @@ moves_loop: // When in check, search starts here
     // Bonus for prior countermove that caused the fail low
     else if (!priorCapture && prevSq != SQ_NONE)
     {
-        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 100 * depth) + ((ss-1)->moveCount > 11);
+        int bonus = (depth > 5) + (PvNode || cutNode) + (bestValue < alpha - 113 * depth) + ((ss-1)->moveCount > 12);
         update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, stat_bonus(depth) * bonus);
     }
 
@@ -1494,7 +1492,7 @@ moves_loop: // When in check, search starts here
         if (PvNode && bestValue > alpha)
             alpha = bestValue;
 
-        futilityBase = bestValue + 190;
+        futilityBase = bestValue + 200;
     }
 
     const PieceToHistory* contHist[] = { (ss-1)->continuationHistory, (ss-2)->continuationHistory,
@@ -1567,7 +1565,7 @@ moves_loop: // When in check, search starts here
           continue;
 
       // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, Value(-94)))
+      if (!pos.see_ge(move, Value(-95)))
           continue;
     }
 
@@ -1700,7 +1698,7 @@ moves_loop: // When in check, search starts here
 
     if (!pos.capture_stage(bestMove))
     {
-        int bonus2 = bestValue > beta + 143 ? bonus1               // larger bonus
+        int bonus2 = bestValue > beta + 145 ? bonus1               // larger bonus
                                             : stat_bonus(depth);   // smaller bonus
 
         // Increase stats for the best move in case it was a quiet move

From 6cf8d938c5950ddedb8a92cdea4712f7d507c614 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Thu, 1 Jun 2023 09:50:19 -0400
Subject: [PATCH 164/192] Simplify blending nnue complexity with optimism

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/6478a26d54dd118e1d98f21c
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 241248 W: 64058 L: 64063 D: 113127
Ptnml(0-2): 644, 26679, 65960, 26720, 621

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/647b464854dd118e1d9928b2
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 24336 W: 6658 L: 6451 D: 11227
Ptnml(0-2): 8, 2316, 7312, 2525, 7

closes https://github.com/official-stockfish/Stockfish/pull/4602

bench 2425813
---
 src/evaluate.cpp | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index 40c43d23..bf6dd69a 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1070,10 +1070,8 @@ Value Eval::evaluate(const Position& pos) {
 
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
-      // Blend nnue complexity with (semi)classical complexity
-      nnueComplexity = 25 * (nnueComplexity + abs(psq - nnue)) / 64;
-
-      optimism += optimism * nnueComplexity / 256;
+      // Blend optimism with nnue complexity and (semi)classical complexity
+      optimism += 25 * optimism * (nnueComplexity + abs(psq - nnue)) / 16384;
       v = (nnue * (945 + npm) + optimism * (174 + npm)) / 1024;
   }
 

From 5930c0defbe01576315d7d081447f94a01daf337 Mon Sep 17 00:00:00 2001
From: Guenther Demetz <guenther.demetz@wuerth-phoenix.com>
Date: Wed, 31 May 2023 11:48:18 +0200
Subject: [PATCH 165/192] Simplify away SEE verification

After 4 simplificatons over PR#4453 the idea does not yield significant
improvement anymore. Maybe also
https://tests.stockfishchess.org/tests/view/640c88092644b62c3394c1c5 was
a fluke.

Passed non-regression bounds:

STC:
https://tests.stockfishchess.org/tests/view/64705389c079b6583146d873
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 131936 W: 35040 L: 34930 D: 61966
Ptnml(0-2): 336, 14559, 36035, 14735, 303

LTC:
https://tests.stockfishchess.org/tests/view/6471a2ade549d9cf2fb213cd
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 407700 W: 109999 L: 110164 D: 187537
Ptnml(0-2): 279, 39913, 123689, 39632, 337

closes https://github.com/official-stockfish/Stockfish/pull/4595

bench: 2675974
---
 src/position.cpp | 19 +++++++------------
 src/position.h   |  1 -
 src/search.cpp   | 18 +-----------------
 3 files changed, 8 insertions(+), 30 deletions(-)

diff --git a/src/position.cpp b/src/position.cpp
index 2a9d798f..af274d3f 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -1061,7 +1061,7 @@ Key Position::key_after(Move m) const {
 /// SEE value of move is greater or equal to the given threshold. We'll use an
 /// algorithm similar to alpha-beta pruning with a null window.
 
-bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
+bool Position::see_ge(Move m, Value threshold) const {
 
   assert(is_ok(m));
 
@@ -1080,7 +1080,7 @@ bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
       return true;
 
   assert(color_of(piece_on(from)) == sideToMove);
-  occupied = pieces() ^ from ^ to; // xoring to is important for pinned piece logic
+  Bitboard occupied = pieces() ^ from ^ to; // xoring to is important for pinned piece logic
   Color stm = sideToMove;
   Bitboard attackers = attackers_to(to, occupied);
   Bitboard stmAttackers, bb;
@@ -1111,43 +1111,43 @@ bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
       // the bitboard 'attackers' any X-ray attackers behind it.
       if ((bb = stmAttackers & pieces(PAWN)))
       {
-          occupied ^= least_significant_square_bb(bb);
           if ((swap = PawnValueMg - swap) < res)
               break;
+          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(KNIGHT)))
       {
-          occupied ^= least_significant_square_bb(bb);
           if ((swap = KnightValueMg - swap) < res)
               break;
+          occupied ^= least_significant_square_bb(bb);
       }
 
       else if ((bb = stmAttackers & pieces(BISHOP)))
       {
-          occupied ^= least_significant_square_bb(bb);
           if ((swap = BishopValueMg - swap) < res)
               break;
+          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(ROOK)))
       {
-          occupied ^= least_significant_square_bb(bb);
           if ((swap = RookValueMg - swap) < res)
               break;
+          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<ROOK>(to, occupied) & pieces(ROOK, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(QUEEN)))
       {
-          occupied ^= least_significant_square_bb(bb);
           if ((swap = QueenValueMg - swap) < res)
               break;
+          occupied ^= least_significant_square_bb(bb);
 
           attackers |=  (attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN))
                       | (attacks_bb<ROOK  >(to, occupied) & pieces(ROOK  , QUEEN));
@@ -1162,11 +1162,6 @@ bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
   return bool(res);
 }
 
-bool Position::see_ge(Move m, Value threshold) const {
-    Bitboard occupied;
-    return see_ge(m, occupied, threshold);
-}
-
 
 /// Position::is_draw() tests whether the position is drawn by 50-move rule
 /// or by repetition. It does not detect stalemates.
diff --git a/src/position.h b/src/position.h
index a736f3e6..780d463c 100644
--- a/src/position.h
+++ b/src/position.h
@@ -143,7 +143,6 @@ public:
   void undo_null_move();
 
   // Static Exchange Evaluation
-  bool see_ge(Move m, Bitboard& occupied, Value threshold = VALUE_ZERO) const;
   bool see_ge(Move m, Value threshold = VALUE_ZERO) const;
 
   // Accessing hash keys
diff --git a/src/search.cpp b/src/search.cpp
index 16122315..1e82203a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -989,25 +989,9 @@ moves_loop: // When in check, search starts here
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
-              Bitboard occupied;
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, occupied, Value(-212) * depth))
-              {
-                  // Don't prune the move if opponent King/Queen/Rook gets a discovered attack during or after the exchanges
-                  Bitboard leftEnemies = pos.pieces(~us, KING, QUEEN, ROOK);
-                  Bitboard attacks = 0;
-                  occupied |= to_sq(move);
-                  while (leftEnemies && !attacks)
-                  {
-                      Square sq = pop_lsb(leftEnemies);
-                      attacks = pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
-                      // Exclude Queen/Rook(s) which were already threatened before SEE (opponent King can't be in check when it's our turn)
-                      if (attacks && sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us)))
-                          attacks = 0;
-                  }
-                  if (!attacks)
+              if (!pos.see_ge(move, Value(-205) * depth))
                       continue;
-              }
           }
           else
           {

From ced0311890add58ab516b9c19608cbd1e1f295ed Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Tue, 30 May 2023 18:24:54 -0400
Subject: [PATCH 166/192] Remove static eval threshold for extensions when
 giving check

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/647685d54a36543c4c9f4f2a
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 114688 W: 30701 L: 30571 D: 53416
Ptnml(0-2): 336, 12708, 31136, 12818, 346

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/64774b02b81f005b572de770
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 107310 W: 28920 L: 28796 D: 49594
Ptnml(0-2): 33, 10427, 32621, 10531, 43

closes https://github.com/official-stockfish/Stockfish/pull/4599

bench 2597974
---
 src/search.cpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 1e82203a..4365b215 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1085,7 +1085,8 @@ moves_loop: // When in check, search starts here
           }
 
           // Check extensions (~1 Elo)
-          else if (   givesCheck && depth > 8)
+          else if (   givesCheck
+                   && depth > 9)
               extension = 1;
 
           // Quiet ttMove extensions (~1 Elo)

From 8dea070538dcad790de3c5b9720bdbb836a32440 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Thu, 1 Jun 2023 15:24:13 +0300
Subject: [PATCH 167/192] Move internal iterative reduction before probcut

This patch moves IIR before probcut which allows probcut
to be produced at lower depths. Comments in IIR are also slightly updated.

Passed STC:
https://tests.stockfishchess.org/tests/view/6472d604d29264e4cfa749fd
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 387616 W: 103295 L: 102498 D: 181823
Ptnml(0-2): 976, 42322, 106381, 43187, 942

Passed LTC:
https://tests.stockfishchess.org/tests/view/6475eb8c4a36543c4c9f42e8
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 202836 W: 54901 L: 54281 D: 93654
Ptnml(0-2): 85, 19609, 61422, 20205, 97

closes https://github.com/official-stockfish/Stockfish/pull/4597

bench 2551691
---
 src/search.cpp | 31 ++++++++++++++++---------------
 1 file changed, 16 insertions(+), 15 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 4365b215..0e82f04e 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -822,9 +822,24 @@ namespace {
         }
     }
 
+    // Step 10. If the position doesn't a have ttMove, decrease depth by 2
+    // (or by 4 if the TT entry for the current position was hit and the stored depth is greater than or equal to the current depth).
+    // Use qsearch if depth is equal or below zero (~9 Elo)
+    if (    PvNode
+        && !ttMove)
+        depth -= 2 + 2 * (ss->ttHit && tte->depth() >= depth);
+
+    if (depth <= 0)
+        return qsearch<PV>(pos, ss, alpha, beta);
+
+    if (    cutNode
+        &&  depth >= 8
+        && !ttMove)
+        depth -= 2;
+
     probCutBeta = beta + 168 - 61 * improving;
 
-    // Step 10. ProbCut (~10 Elo)
+    // Step 11. ProbCut (~10 Elo)
     // If we have a good enough capture (or queen promotion) and a reduced search returns a value
     // much above beta, we can (almost) safely prune the previous move.
     if (   !PvNode
@@ -875,20 +890,6 @@ namespace {
         Eval::NNUE::hint_common_parent_position(pos);
     }
 
-    // Step 11. If the position is not in TT, decrease depth by 2 (or by 4 if the TT entry for the current position was hit and the stored depth is greater than or equal to the current depth).
-    // Use qsearch if depth is equal or below zero (~9 Elo)
-    if (    PvNode
-        && !ttMove)
-        depth -= 2 + 2 * (ss->ttHit && tte->depth() >= depth);
-
-    if (depth <= 0)
-        return qsearch<PV>(pos, ss, alpha, beta);
-
-    if (    cutNode
-        &&  depth >= 8
-        && !ttMove)
-        depth -= 2;
-
 moves_loop: // When in check, search starts here
 
     // Step 12. A small Probcut idea, when we are in check (~4 Elo)

From b60738e01b5393e6f85bc10d2f257dd0cd26a2f9 Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Mon, 5 Jun 2023 00:59:31 +0300
Subject: [PATCH 168/192] Fix no previous moves on root.

guards against no previous move existing if qSearch is called on the root node (i.e. when razoring).

Passed Non-regression STC:
https://tests.stockfishchess.org/tests/view/647d242d726f6b400e408143
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 53120 W: 14167 L: 13976 D: 24977
Ptnml(0-2): 109, 5597, 14981, 5740, 133

closes https://github.com/official-stockfish/Stockfish/pull/4604

Bench: 2551691
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index 0e82f04e..c7b00766 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1489,7 +1489,7 @@ moves_loop: // When in check, search starts here
     // to search the moves. Because the depth is <= 0 here, only captures,
     // queen promotions, and other checks (only if depth >= DEPTH_QS_CHECKS)
     // will be generated.
-    Square prevSq = (ss-1)->currentMove != MOVE_NULL ? to_sq((ss-1)->currentMove) : SQ_NONE;
+    Square prevSq = is_ok((ss-1)->currentMove) ? to_sq((ss-1)->currentMove) : SQ_NONE;
     MovePicker mp(pos, ttMove, depth, &thisThread->mainHistory,
                                       &thisThread->captureHistory,
                                       contHist,

From 295f57829ea189248c8b4afa8d11222d170f78da Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Mon, 17 Apr 2023 22:16:22 +0200
Subject: [PATCH 169/192] Add binaries to releases with github actions

when a release is made with a tag matching sf_* the binaries will also be uploaded to the release as assets.

closes https://github.com/official-stockfish/Stockfish/pull/4596

No functional change.
---
 .github/workflows/stockfish.yml              |  6 +++--
 .github/workflows/stockfish_arm_binaries.yml |  8 ++++++
 .github/workflows/stockfish_binaries.yml     | 27 +++++++++++++-------
 3 files changed, 30 insertions(+), 11 deletions(-)

diff --git a/.github/workflows/stockfish.yml b/.github/workflows/stockfish.yml
index 6345b27c..082c65de 100644
--- a/.github/workflows/stockfish.yml
+++ b/.github/workflows/stockfish.yml
@@ -1,6 +1,8 @@
 name: Stockfish
 on:
   push:
+    tags:        
+      - '*' 
     branches:
       - master
       - tools
@@ -17,8 +19,8 @@ jobs:
   Compiles:
     uses: ./.github/workflows/stockfish_compile_test.yml
   Binaries:
-    if: github.ref == 'refs/heads/master'
+    if: github.ref == 'refs/heads/master' || (startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag')
     uses: ./.github/workflows/stockfish_binaries.yml
   ARM_Binaries:
-    if: github.ref == 'refs/heads/master'
+    if: github.ref == 'refs/heads/master' || (startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag')
     uses: ./.github/workflows/stockfish_arm_binaries.yml
diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
index e088c441..9a4734ee 100644
--- a/.github/workflows/stockfish_arm_binaries.yml
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -115,6 +115,8 @@ jobs:
           cp "Top CPU Contributors.txt" stockfish/
           cp Copying.txt stockfish/
           cp AUTHORS stockfish/
+          cp CITATION.cff stockfish/
+          cp README.md stockfish/
           tar -cvf stockfish-android-$BINARY.tar stockfish
 
       - name: Upload binaries
@@ -122,3 +124,9 @@ jobs:
         with:
           name: stockfish-android-${{ matrix.binaries }}
           path: stockfish-android-${{ matrix.binaries }}.tar
+
+      - name: Release
+        if: startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag'
+        uses: softprops/action-gh-release@v1
+        with:
+          files: stockfish-android-${{ matrix.binaries }}.tar
\ No newline at end of file
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index b22897cf..86449b97 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -9,23 +9,26 @@ jobs:
       COMPILER: ${{ matrix.config.compiler }}
       COMP: ${{ matrix.config.comp }}
       EXT: ${{ matrix.config.ext }}
-      OS: ${{ matrix.config.os }}
+      NAME: ${{ matrix.config.simple_name }}
       BINARY: ${{ matrix.binaries }}
     strategy:
       matrix:
         config:
           - name: Ubuntu 20.04 GCC
             os: ubuntu-20.04
+            simple_name: ubuntu
             compiler: g++
             comp: gcc
             shell: bash {0}
           - name: MacOS 12 Apple Clang
             os: macos-12
+            simple_name: macos
             compiler: clang++
             comp: clang
             shell: bash {0}
           - name: Windows 2022 Mingw-w64 GCC x86_64
             os: windows-2022
+            simple_name: windows
             compiler: g++
             comp: mingw
             msys_sys: mingw64
@@ -75,19 +78,17 @@ jobs:
 
       - name: Compile ${{ matrix.binaries }} build
         run: |
-          make clean
           make -j2 profile-build ARCH=$BINARY COMP=$COMP
           make strip ARCH=$BINARY COMP=$COMP
-          mv ./stockfish$EXT ../stockfish-$OS-$BINARY$EXT
+          mv ./stockfish$EXT ../stockfish-$NAME-$BINARY$EXT
 
       - name: Remove non src files
-        run: rm -f *.o .depend *.nnue
+        run: git clean -fx
 
       - name: Download wiki
         run: |
           git clone https://github.com/official-stockfish/Stockfish.wiki.git ../wiki
-          cd ../wiki
-          rm -rf .git
+          rm -rf ../wiki/.git
 
       - name: Create tar archive.
         run: |
@@ -95,14 +96,22 @@ jobs:
           mkdir stockfish
           cp -r wiki stockfish/
           cp -r src stockfish/
-          cp stockfish-$OS-$BINARY$EXT stockfish/
+          cp stockfish-$NAME-$BINARY$EXT stockfish/
           cp "Top CPU Contributors.txt" stockfish/
           cp Copying.txt stockfish/
           cp AUTHORS stockfish/
-          tar -cvf stockfish-$OS-$BINARY.tar stockfish
+          cp CITATION.cff stockfish/
+          cp README.md stockfish/
+          tar -cvf stockfish-$NAME-$BINARY.tar stockfish
 
       - name: Upload binaries
         uses: actions/upload-artifact@v3
         with:
           name: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}
-          path: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}.tar
+          path: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.tar
+
+      - name: Release
+        if: startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag'
+        uses: softprops/action-gh-release@v1
+        with:
+          files: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.tar

From 373359b44d0947cce2628a9a8c9b432a458615a8 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Sun, 4 Jun 2023 01:56:11 -0400
Subject: [PATCH 170/192] Update default net to nn-0dd1cebea573.nnue

Created by retraining an earlier epoch of the experiment leading to the first SFNNv6 net
on a more-randomized version of the nn-e1fb1ade4432.nnue dataset mixed with unfiltered
T80 apr2023 data. Trained using early-fen-skipping 28 and max-epoch 960.

The trainer settings and epochs used in the 5-step training sequence leading here were:
1. train from scratch for 400 epochs, lambda 1.0, constant LR 9.75e-4, T79T77-filter-v6-dd.min.binpack
2. retrain ep379, max-epoch 800, end-lambda 0.75, T60T70wIsRightFarseerT60T74T75T76.binpack
3. retrain ep679, max-epoch 800, end-lambda 0.75, skip 28, nn-e1fb1ade4432 dataset
4. retrain ep799, max-epoch 800, end-lambda 0.7, skip 28, nn-e1fb1ade4432 dataset
5. retrain ep439, max-epoch 960, end-lambda 0.7, skip 28, shuffled nn-e1fb1ade4432 + T80 apr2023

This net was epoch 559 of the final (step 5) retraining:

```bash
python3 easy_train.py \
  --experiment-name L1-1536-Re4-leela96-dfrc99-T60novdec-v2-T80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd-T80apr-shuffled-sk28 \
  --training-dataset /data/leela96-dfrc99-T60novdec-v2-T80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd-T80apr.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes-L1-1536 \
  --early-fen-skipping 28 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --max_epoch 960 \
  --start-from-engine-test-net False \
  --start-from-model /data/L1-1536-Re3-nn-epoch439.nnue \
  --engine-test-branch linrock/Stockfish/L1-1536 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --tui False \
  --seed $RANDOM \
  --gpus "0,"
```

During data preparation, most binpacks were unminimized by removing positions with
score 32002 (`VALUE_NONE`). This makes the tradeoff of increasing dataset filesize
on disk to increase the randomness of positions in interleaved datasets.
The code used for unminimizing is at:
https://github.com/linrock/Stockfish/tree/tools-unminify

For preparing the dataset used in this experiment:

```bash
python3 interleave_binpacks.py \
  leela96-filt-v2.binpack \
  dfrc99-16tb7p-eval-filt-v2.binpack \
  filt-v6-dd-min/test60-novdec2021-12tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd-min/test80-aug2022-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd-min/test80-sep2022-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd-min/test80-jun2022-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd/test80-jul2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-oct2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test80-nov2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd-min/test80-jan2023-3of3-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd-min/test80-feb2023-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd/test79-apr2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test79-may2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd-min/test78-jantomay2022-16tb7p-filter-v6-dd.min-mar2023.unmin.binpack \
  filt-v6-dd/test78-juntosep2022-16tb7p-filter-v6-dd.binpack \
  filt-v6-dd/test77-dec2021-16tb7p-filter-v6-dd.binpack \
  test80-apr2023-2tb7p.binpack \
  /data/leela96-dfrc99-T60novdec-v2-T80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd-T80apr.binpack
```

T80 apr2023 data was converted using lc0-rescorer with ~2tb of tablebases and can be found at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move vs. nn-e1fb1ade4432.nnue (L1 size 1024):
nn-epoch559.nnue : 25.7 +/- 1.6

Passed STC:
https://tests.stockfishchess.org/tests/view/647cd3b87cf638f0f53f9cbb
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 59200 W: 16000 L: 15660 D: 27540
Ptnml(0-2): 159, 6488, 15996, 6768, 189

Passed LTC:
https://tests.stockfishchess.org/tests/view/647d58de726f6b400e4085d8
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 58800 W: 16002 L: 15657 D: 27141
Ptnml(0-2): 44, 5607, 17748, 5962, 39

closes https://github.com/official-stockfish/Stockfish/pull/4606

bench 2141197
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 0990111c..dcbe6b3c 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-8d69132723e2.nnue"
+  #define EvalFileDefaultName   "nn-0dd1cebea573.nnue"
 
   namespace NNUE {
 

From 54ad986768eec524aeab721713ea2009931b51b3 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Mon, 5 Jun 2023 01:08:38 -0400
Subject: [PATCH 171/192] Remove optimism multiplier in nnue eval calculation

The same formula had passed SPRT against an earlier version of master.

Passed non-regression STC vs. d99942f:
https://tests.stockfishchess.org/tests/view/6478e76654dd118e1d98f72e
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 118720 W: 31402 L: 31277 D: 56041
Ptnml(0-2): 301, 13148, 32344, 13259, 308

Passed non-regression LTC vs. d99942f:
https://tests.stockfishchess.org/tests/view/647a22c154dd118e1d991146
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 74286 W: 20019 L: 19863 D: 34404
Ptnml(0-2): 31, 7189, 22540, 7359, 24

The earlier patch had conflicted with a faster SPRT passer, so this
was tested again after rebasing on latest master.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/647d6e46726f6b400e408790
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 166176 W: 44309 L: 44234 D: 77633
Ptnml(0-2): 461, 18252, 45557, 18387, 431

Passed non-regression LTC:
https://tests.stockfishchess.org/tests/view/647eb00ba268d1bc11255e7b
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 28170 W: 7713 L: 7513 D: 12944
Ptnml(0-2): 14, 2609, 8635, 2817, 10

closes https://github.com/official-stockfish/Stockfish/pull/4607

bench 2503095
---
 src/evaluate.cpp | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/evaluate.cpp b/src/evaluate.cpp
index bf6dd69a..35d05427 100644
--- a/src/evaluate.cpp
+++ b/src/evaluate.cpp
@@ -1071,8 +1071,8 @@ Value Eval::evaluate(const Position& pos) {
       Value nnue = NNUE::evaluate(pos, true, &nnueComplexity);
 
       // Blend optimism with nnue complexity and (semi)classical complexity
-      optimism += 25 * optimism * (nnueComplexity + abs(psq - nnue)) / 16384;
-      v = (nnue * (945 + npm) + optimism * (174 + npm)) / 1024;
+      optimism += optimism * (nnueComplexity + abs(psq - nnue)) / 512;
+      v = (nnue * (945 + npm) + optimism * (150 + npm)) / 1024;
   }
 
   // Damp down the evaluation linearly when shuffling

From a9a6915e0839d3f3f54659c86f15868a7db0e386 Mon Sep 17 00:00:00 2001
From: cj5716 <125858804+cj5716@users.noreply.github.com>
Date: Mon, 5 Jun 2023 07:59:56 +0800
Subject: [PATCH 172/192] Simplify multiplier for improvement

This simplifies a `* 99 / 1300` term into just `/ 13`.

Passed non-regression STC:
LLR: 2.92 (-2.94,2.94) <-1.75,0.25>
Total: 58816 W: 15727 L: 15540 D: 27549
Ptnml(0-2): 149, 6370, 16203, 6517, 169
https://tests.stockfishchess.org/tests/view/647d25e4726f6b400e408165

Passed non-regression LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 154386 W: 41749 L: 41669 D: 70968
Ptnml(0-2): 94, 14992, 46956, 15042, 109
https://tests.stockfishchess.org/tests/view/647d9b3c726f6b400e408b2a

closes https://github.com/official-stockfish/Stockfish/pull/4608

Bench:  2511327
---
 src/search.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index c7b00766..593fdc72 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -779,7 +779,7 @@ namespace {
         && (ss-1)->statScore < 17329
         &&  eval >= beta
         &&  eval >= ss->staticEval
-        &&  ss->staticEval >= beta - 21 * depth - improvement * 99 / 1300 + 258
+        &&  ss->staticEval >= beta - 21 * depth - improvement / 13 + 258
         && !excludedMove
         &&  pos.non_pawn_material(us)
         && (ss->ply >= thisThread->nmpMinPly))

From e1dd005583bd6c2aaf58468efc5de86a3936380a Mon Sep 17 00:00:00 2001
From: Guenther Demetz <guenther.demetz@wuerth-phoenix.com>
Date: Wed, 7 Jun 2023 09:01:05 +0200
Subject: [PATCH 173/192] Reintroduce SEE verification against discovered
 attacks

Reintroduces https://github.com/official-stockfish/Stockfish/pull/4453
along with https://github.com/official-stockfish/Stockfish/pull/4469

Leaving out
https://github.com/official-stockfish/Stockfish/pull/4533
https://github.com/official-stockfish/Stockfish/pull/4572

Passed STC:
https://tests.stockfishchess.org/tests/view/647d8c37726f6b400e408a0a
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 143168 W: 38346 L: 37892 D: 66930
Ptnml(0-2): 352, 15672, 39164, 15962, 434

Passed LTC:
https://tests.stockfishchess.org/tests/view/647ee8c528c4431bcb58e432
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 71538 W: 19560 L: 19190 D: 32788
Ptnml(0-2): 49, 6905, 21499, 7259, 57

closes https://github.com/official-stockfish/Stockfish/pull/4609

bench: 2595430
---
 src/position.cpp | 19 ++++++++++++-------
 src/position.h   |  1 +
 src/search.cpp   | 23 +++++++++++++++++++++--
 3 files changed, 34 insertions(+), 9 deletions(-)

diff --git a/src/position.cpp b/src/position.cpp
index af274d3f..2a9d798f 100644
--- a/src/position.cpp
+++ b/src/position.cpp
@@ -1061,7 +1061,7 @@ Key Position::key_after(Move m) const {
 /// SEE value of move is greater or equal to the given threshold. We'll use an
 /// algorithm similar to alpha-beta pruning with a null window.
 
-bool Position::see_ge(Move m, Value threshold) const {
+bool Position::see_ge(Move m, Bitboard& occupied, Value threshold) const {
 
   assert(is_ok(m));
 
@@ -1080,7 +1080,7 @@ bool Position::see_ge(Move m, Value threshold) const {
       return true;
 
   assert(color_of(piece_on(from)) == sideToMove);
-  Bitboard occupied = pieces() ^ from ^ to; // xoring to is important for pinned piece logic
+  occupied = pieces() ^ from ^ to; // xoring to is important for pinned piece logic
   Color stm = sideToMove;
   Bitboard attackers = attackers_to(to, occupied);
   Bitboard stmAttackers, bb;
@@ -1111,43 +1111,43 @@ bool Position::see_ge(Move m, Value threshold) const {
       // the bitboard 'attackers' any X-ray attackers behind it.
       if ((bb = stmAttackers & pieces(PAWN)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = PawnValueMg - swap) < res)
               break;
-          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(KNIGHT)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = KnightValueMg - swap) < res)
               break;
-          occupied ^= least_significant_square_bb(bb);
       }
 
       else if ((bb = stmAttackers & pieces(BISHOP)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = BishopValueMg - swap) < res)
               break;
-          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(ROOK)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = RookValueMg - swap) < res)
               break;
-          occupied ^= least_significant_square_bb(bb);
 
           attackers |= attacks_bb<ROOK>(to, occupied) & pieces(ROOK, QUEEN);
       }
 
       else if ((bb = stmAttackers & pieces(QUEEN)))
       {
+          occupied ^= least_significant_square_bb(bb);
           if ((swap = QueenValueMg - swap) < res)
               break;
-          occupied ^= least_significant_square_bb(bb);
 
           attackers |=  (attacks_bb<BISHOP>(to, occupied) & pieces(BISHOP, QUEEN))
                       | (attacks_bb<ROOK  >(to, occupied) & pieces(ROOK  , QUEEN));
@@ -1162,6 +1162,11 @@ bool Position::see_ge(Move m, Value threshold) const {
   return bool(res);
 }
 
+bool Position::see_ge(Move m, Value threshold) const {
+    Bitboard occupied;
+    return see_ge(m, occupied, threshold);
+}
+
 
 /// Position::is_draw() tests whether the position is drawn by 50-move rule
 /// or by repetition. It does not detect stalemates.
diff --git a/src/position.h b/src/position.h
index 780d463c..2e6014db 100644
--- a/src/position.h
+++ b/src/position.h
@@ -144,6 +144,7 @@ public:
 
   // Static Exchange Evaluation
   bool see_ge(Move m, Value threshold = VALUE_ZERO) const;
+  bool see_ge(Move m, Bitboard& occupied, Value threshold = VALUE_ZERO) const;
 
   // Accessing hash keys
   Key key() const;
diff --git a/src/search.cpp b/src/search.cpp
index 593fdc72..b2c2344a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -990,9 +990,28 @@ moves_loop: // When in check, search starts here
                    + captureHistory[movedPiece][to_sq(move)][type_of(pos.piece_on(to_sq(move)))] / 7 < alpha)
                   continue;
 
+              Bitboard occupied;
               // SEE based pruning (~11 Elo)
-              if (!pos.see_ge(move, Value(-205) * depth))
-                      continue;
+              if (!pos.see_ge(move, occupied, Value(-205) * depth))
+              {
+                 if (depth < 2 - capture)
+                    continue;
+                 // Don't prune the move if opponent Queen/Rook is under discovered attack after the exchanges
+                 // Don't prune the move if opponent King is under discovered attack after or during the exchanges
+                 Bitboard leftEnemies = (pos.pieces(~us, KING, QUEEN, ROOK)) & occupied;
+                 Bitboard attacks = 0;
+                 occupied |= to_sq(move);
+                 while (leftEnemies && !attacks)
+                 {
+                      Square sq = pop_lsb(leftEnemies);
+                      attacks |= pos.attackers_to(sq, occupied) & pos.pieces(us) & occupied;
+                      // don't consider pieces which were already threatened/hanging before SEE exchanges
+                      if (attacks && (sq != pos.square<KING>(~us) && (pos.attackers_to(sq, pos.pieces()) & pos.pieces(us))))
+                         attacks = 0;
+                 }
+                 if (!attacks)
+                    continue;
+              }
           }
           else
           {

From 932f5a2d657c846c282adcf2051faef7ca17ae15 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Wed, 7 Jun 2023 14:57:17 -0400
Subject: [PATCH 174/192] Update default net to nn-ea57bea57e32.nnue

Created by retraining an earlier epoch (ep659) of the experiment that led to the first SFNNv6 net:
- First retrained on the nn-0dd1cebea573 dataset
- Then retrained with skip 20 on a smaller dataset containing unfiltered Leela data
- And then retrained again with skip 27 on the nn-0dd1cebea573 dataset

The equivalent 7-step training sequence from scratch that led here was:

1. max-epoch 400, lambda 1.0, constant LR 9.75e-4, T79T77-filter-v6-dd.min.binpack
   ep379 chosen for retraining in step2

2. max-epoch 800, end-lambda 0.75, T60T70wIsRightFarseerT60T74T75T76.binpack
   ep679 chosen for retraining in step3

3. max-epoch 800, end-lambda 0.75, skip 28, nn-e1fb1ade4432 dataset
   ep799 chosen for retraining in step4

4. max-epoch 800, end-lambda 0.7, skip 28, nn-e1fb1ade4432 dataset
   ep759 became nn-8d69132723e2.nnue (first SFNNv6 net)
   ep659 chosen for retraining in step5

5. max-epoch 800, end-lambda 0.7, skip 28, nn-0dd1cebea573 dataset
   ep759 chosen for retraining in step6

6. max-epoch 800, end-lambda 0.7, skip 20, leela-dfrc-v2-T77decT78janfebT79aprT80apr.binpack
   ep639 chosen for retraining in step7

7. max-epoch 800, end-lambda 0.7, skip 27, nn-0dd1cebea573 dataset
   ep619 became nn-ea57bea57e32.nnue

For the last retraining (step7):

python3 easy_train.py
  --experiment-name L1-1536-Re6-masterShuffled-ep639-sk27-Re5-leela-dfrc-v2-T77toT80small-Re4-masterShuffled-ep659-Re3-sameAs-Re2-leela96-dfrc99-16t-v2-T60novdecT80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd-Re1-LeelaFarseer-new-T77T79 \
  --training-dataset /data/leela96-dfrc99-T60novdec-v2-T80juntonovjanfebT79aprmayT78jantosepT77dec-v6dd-T80apr.binpack \
  --nnue-pytorch-branch linrock/nnue-pytorch/misc-fixes-L1-1536 \
  --early-fen-skipping 27 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --max_epoch 800 \
  --start-from-engine-test-net False \
  --start-from-model /data/L1-1536-Re5-leela-dfrc-v2-T77toT80small-epoch639.nnue \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --tui False \
  --seed $RANDOM \
  --gpus "0,"

For preparing the step6 leela-dfrc-v2-T77decT78janfebT79aprT80apr.binpack dataset:

python3 interleave_binpacks.py \
  leela96-filt-v2.binpack \
  dfrc99-16tb7p-eval-filt-v2.binpack \
  test77-dec2021-16tb7p.no-db.min-mar2023.binpack \
  test78-janfeb2022-16tb7p.no-db.min-mar2023.binpack \
  test79-apr2022-16tb7p-filter-v6-dd.binpack \
  test80-apr2022-16tb7p.no-db.min-mar2023.binpack \
  /data/leela-dfrc-v2-T77decT78janfebT79aprT80apr.binpack

The unfiltered Leela data used for the step6 dataset can be found at:
https://robotmoon.com/nnue-training-data

Local elo at 25k nodes per move:
nn-epoch619.nnue : 2.3 +/- 1.9

Passed STC:
https://tests.stockfishchess.org/tests/view/6480d43c6e6ce8d9fc6d7cc8
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 40992 W: 11017 L: 10706 D: 19269
Ptnml(0-2): 113, 4400, 11170, 4689, 124

Passed LTC:
https://tests.stockfishchess.org/tests/view/648119ac6e6ce8d9fc6d8208
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 129174 W: 35059 L: 34579 D: 59536
Ptnml(0-2): 66, 12548, 38868, 13050, 55

closes https://github.com/official-stockfish/Stockfish/pull/4611

bench: 2370027
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index dcbe6b3c..fc852c8d 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-0dd1cebea573.nnue"
+  #define EvalFileDefaultName   "nn-ea57bea57e32.nnue"
 
   namespace NNUE {
 

From b7ee7290b552b21352491ec7f390565ff4748647 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Mon, 12 Jun 2023 09:51:28 +0200
Subject: [PATCH 175/192] Add network export to CI

verify the network written by export_net matches the original

closes https://github.com/official-stockfish/Stockfish/pull/4613

No functional change
---
 tests/instrumented.sh | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/tests/instrumented.sh b/tests/instrumented.sh
index e9455eab..1b37c7a8 100755
--- a/tests/instrumented.sh
+++ b/tests/instrumented.sh
@@ -70,7 +70,8 @@ for args in "eval" \
             "go depth 10" \
             "go movetime 1000" \
             "go wtime 8000 btime 8000 winc 500 binc 500" \
-            "bench 128 $threads 8 default depth"
+            "bench 128 $threads 8 default depth" \
+            "export_net verify.nnue"
 do
 
    echo "$prefix $exeprefix ./stockfish $args $postfix"
@@ -78,6 +79,11 @@ do
 
 done
 
+# verify the generated net equals the base net
+network=`./stockfish uci | grep 'option name EvalFile type string default' | awk '{print $NF}'`
+echo "Comparing $network to the written verify.nnue"
+diff $network verify.nnue
+
 # more general testing, following an uci protocol exchange
 cat << EOF > game.exp
  set timeout 240

From 38e61663d836e062af0bc002814ad5149c4b7729 Mon Sep 17 00:00:00 2001
From: AndrovT <31534597+AndrovT@users.noreply.github.com>
Date: Sun, 11 Jun 2023 03:24:04 +0200
Subject: [PATCH 176/192] Use block sparse input for the first layer.

Use block sparse input for the first fully connected layer on architectures with at least SSSE3.

Depending on the CPU architecture, this yields a speedup of up to 10%, e.g.

```
Result of 100 runs of 'bench 16 1 13 default depth NNUE'

base (...ockfish-base) =     959345  +/- 7477
test (...ckfish-patch) =    1054340  +/- 9640
diff                   =     +94995  +/- 3999

speedup        = +0.0990
P(speedup > 0) =  1.0000

CPU: 8 x AMD Ryzen 7 5700U with Radeon Graphics
Hyperthreading: on
```

Passed STC:
https://tests.stockfishchess.org/tests/view/6485aa0965ffe077ca12409c
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 8864 W: 2479 L: 2223 D: 4162
Ptnml(0-2): 13, 829, 2504, 1061, 25

This commit includes a net with reordered weights, to increase the likelihood of block sparse inputs,
but otherwise equivalent to the previous master net (nn-ea57bea57e32.nnue).

Activation data collected with https://github.com/AndrovT/Stockfish/tree/log-activations, running bench 16 1 13 varied_1000.epd depth NNUE on this data. Net parameters permuted with https://gist.github.com/AndrovT/9e3fbaebb7082734dc84d27e02094cb3.

closes https://github.com/official-stockfish/Stockfish/pull/4612

No functional change
---
 AUTHORS                                       |   1 +
 src/evaluate.h                                |   2 +-
 .../layers/affine_transform_sparse_input.h    | 286 ++++++++++++++++++
 src/nnue/nnue_architecture.h                  |   3 +-
 4 files changed, 290 insertions(+), 2 deletions(-)
 create mode 100644 src/nnue/layers/affine_transform_sparse_input.h

diff --git a/AUTHORS b/AUTHORS
index d01d23cd..63b862ce 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -159,6 +159,7 @@ Norman Schmidt (FireFather)
 notruck
 Ofek Shochat (OfekShochat, ghostway)
 Ondrej Mosnáček (WOnder93)
+Ondřej Mišina (AndrovT)
 Oskar Werkelin Ahlin
 Pablo Vazquez
 Panthee
diff --git a/src/evaluate.h b/src/evaluate.h
index fc852c8d..94cd42cc 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-ea57bea57e32.nnue"
+  #define EvalFileDefaultName   "nn-fdc1d0fe6455.nnue"
 
   namespace NNUE {
 
diff --git a/src/nnue/layers/affine_transform_sparse_input.h b/src/nnue/layers/affine_transform_sparse_input.h
new file mode 100644
index 00000000..00b17c19
--- /dev/null
+++ b/src/nnue/layers/affine_transform_sparse_input.h
@@ -0,0 +1,286 @@
+/*
+  Stockfish, a UCI chess playing engine derived from Glaurung 2.1
+  Copyright (C) 2004-2023 The Stockfish developers (see AUTHORS file)
+
+  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/>.
+*/
+
+// Definition of layer AffineTransformSparseInput of NNUE evaluation function
+
+#ifndef NNUE_LAYERS_AFFINE_TRANSFORM_SPARSE_INPUT_H_INCLUDED
+#define NNUE_LAYERS_AFFINE_TRANSFORM_SPARSE_INPUT_H_INCLUDED
+
+#include <iostream>
+#include <algorithm>
+#include <array>
+#include <type_traits>
+#include "../nnue_common.h"
+#include "affine_transform.h"
+#include "simd.h"
+
+/*
+  This file contains the definition for a fully connected layer (aka affine transform) with block sparse input.
+*/
+
+namespace Stockfish::Eval::NNUE::Layers {
+#if defined(__GNUC__)  // GCC, Clang, ICC
+
+  static inline IndexType lsb_(std::uint32_t b) {
+    assert(b);
+    return IndexType(__builtin_ctzl(b));
+  }
+
+#elif defined(_MSC_VER)  // MSVC
+
+  static inline IndexType lsb_(std::uint32_t b) {
+    assert(b);
+    unsigned long idx;
+    _BitScanForward(&idx, b);
+    return (IndexType) idx;
+  }
+
+#else  // Compiler is neither GCC nor MSVC compatible
+
+#error "Compiler not supported."
+
+#endif
+
+
+#if defined(USE_SSSE3)
+  alignas(CacheLineSize) static inline const std::array<std::array<std::uint16_t, 8>, 256> lookup_indices = [](){
+    std::array<std::array<std::uint16_t, 8>, 256> v{};
+    for (int i = 0; i < 256; ++i)
+    {
+      int j = i;
+      int k = 0;
+      while(j)
+      {
+        const IndexType lsbIndex = lsb_(std::uint32_t(j));
+        j &= j - 1;
+        v[i][k] = lsbIndex;
+        ++k;
+      }
+    }
+    return v;
+  }();
+  alignas(CacheLineSize) static inline const std::array<unsigned, 256> lookup_count = [](){
+    std::array<unsigned, 256> v;
+    for (int i = 0; i < 256; ++i)
+    {
+      int j = i;
+      int k = 0;
+      while(j)
+      {
+        j &= j - 1;
+        ++k;
+      }
+      v[i] = k;
+    }
+    return v;
+  }();
+
+  // Find indices of nonzero numbers in an int32_t array
+  template<const IndexType InputDimensions>
+  void find_nnz(const std::int32_t* input, std::uint16_t* out, IndexType& count_out) {
+#if defined (USE_AVX512)
+    using vec_t = __m512i;
+    #define vec_nnz(a) _mm512_cmpgt_epi32_mask(a, _mm512_setzero_si512())
+#elif defined (USE_AVX2)
+    using vec_t = __m256i;
+    #define vec_nnz(a) _mm256_movemask_ps((__m256)_mm256_cmpgt_epi32(a, _mm256_setzero_si256()))
+#elif defined (USE_SSSE3)
+    using vec_t = __m128i;
+    #define vec_nnz(a) _mm_movemask_ps((__m128)_mm_cmpgt_epi32(a, _mm_setzero_si128()))
+#endif
+    constexpr IndexType InputSimdWidth = sizeof(vec_t) / sizeof(std::int32_t);
+    // Inputs are processed InputSimdWidth at a time and outputs are processed 8 at a time so we process in chunks of max(InputSimdWidth, 8)
+    constexpr IndexType ChunkSize = std::max<IndexType>(InputSimdWidth, 8);
+    constexpr IndexType NumChunks = InputDimensions / ChunkSize;
+    constexpr IndexType InputsPerChunk = ChunkSize / InputSimdWidth;
+    constexpr IndexType OutputsPerChunk = ChunkSize / 8;
+
+    const auto inputVector = reinterpret_cast<const vec_t*>(input);
+    IndexType count = 0;
+    __m128i base = _mm_set1_epi16(0);
+    __m128i increment = _mm_set1_epi16(8);
+    for (IndexType i = 0; i < NumChunks; ++i)
+    {
+      // bitmask of nonzero values in this chunk
+      unsigned nnz = 0;
+      for (IndexType j = 0; j < InputsPerChunk; ++j)
+      {
+        const vec_t inputChunk = inputVector[i * InputsPerChunk + j];
+        nnz |= (unsigned)vec_nnz(inputChunk) << (j * InputSimdWidth);
+      }
+      for (IndexType j = 0; j < OutputsPerChunk; ++j)
+      {
+        const auto lookup = (nnz >> (j * 8)) & 0xFF;
+        const auto offsets = _mm_loadu_si128(reinterpret_cast<const __m128i*>(&lookup_indices[lookup]));
+        _mm_storeu_si128(reinterpret_cast<__m128i*>(out + count), _mm_add_epi16(base, offsets));
+        count += lookup_count[lookup];
+        base = _mm_add_epi16(base, increment);
+      }
+    }
+    count_out = count;
+  }
+# undef vec_nnz
+#endif
+
+  // Sparse input implementation
+  template <IndexType InDims, IndexType OutDims>
+  class AffineTransformSparseInput {
+   public:
+    // Input/output type
+    // Input/output type
+    using InputType = std::uint8_t;
+    using OutputType = std::int32_t;
+
+    // Number of input/output dimensions
+    static constexpr IndexType InputDimensions = InDims;
+    static constexpr IndexType OutputDimensions = OutDims;
+
+    static_assert(OutputDimensions % 16 == 0, "Only implemented for OutputDimensions divisible by 16.");
+
+    static constexpr IndexType PaddedInputDimensions =
+      ceil_to_multiple<IndexType>(InputDimensions, MaxSimdWidth);
+    static constexpr IndexType PaddedOutputDimensions =
+      ceil_to_multiple<IndexType>(OutputDimensions, MaxSimdWidth);
+
+#if defined (USE_SSSE3)
+    static constexpr IndexType ChunkSize = 4;
+#else
+    static constexpr IndexType ChunkSize = 1;
+#endif
+
+    using OutputBuffer = OutputType[PaddedOutputDimensions];
+
+    // Hash value embedded in the evaluation file
+    static constexpr std::uint32_t get_hash_value(std::uint32_t prevHash) {
+      std::uint32_t hashValue = 0xCC03DAE4u;
+      hashValue += OutputDimensions;
+      hashValue ^= prevHash >> 1;
+      hashValue ^= prevHash << 31;
+      return hashValue;
+    }
+
+    static IndexType get_weight_index_scrambled(IndexType i)
+    {
+      return
+        (i / ChunkSize) % (PaddedInputDimensions / ChunkSize) * OutputDimensions * ChunkSize +
+        i / PaddedInputDimensions * ChunkSize +
+        i % ChunkSize;
+    }
+
+    static IndexType get_weight_index(IndexType i)
+    {
+#if defined (USE_SSSE3)
+      return get_weight_index_scrambled(i);
+#else
+      return i;
+#endif
+    }
+
+    // Read network parameters
+    bool read_parameters(std::istream& stream) {
+      read_little_endian<BiasType>(stream, biases, OutputDimensions);
+      for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
+        weights[get_weight_index(i)] = read_little_endian<WeightType>(stream);
+
+      return !stream.fail();
+    }
+
+    // Write network parameters
+    bool write_parameters(std::ostream& stream) const {
+      write_little_endian<BiasType>(stream, biases, OutputDimensions);
+
+      for (IndexType i = 0; i < OutputDimensions * PaddedInputDimensions; ++i)
+        write_little_endian<WeightType>(stream, weights[get_weight_index(i)]);
+
+      return !stream.fail();
+    }
+    // Forward propagation
+    const OutputType* propagate(
+        const InputType* input, OutputType* output) const {
+
+#if defined (USE_SSSE3)
+#if defined (USE_AVX512)
+      using vec_t = __m512i;
+      #define vec_setzero _mm512_setzero_si512
+      #define vec_set_32 _mm512_set1_epi32
+      #define vec_add_dpbusd_32 Simd::m512_add_dpbusd_epi32
+#elif defined (USE_AVX2)
+      using vec_t = __m256i;
+      #define vec_setzero _mm256_setzero_si256
+      #define vec_set_32 _mm256_set1_epi32
+      #define vec_add_dpbusd_32 Simd::m256_add_dpbusd_epi32
+#elif defined (USE_SSSE3)
+      using vec_t = __m128i;
+      #define vec_setzero _mm_setzero_si128
+      #define vec_set_32 _mm_set1_epi32
+      #define vec_add_dpbusd_32 Simd::m128_add_dpbusd_epi32
+#endif
+      static constexpr IndexType OutputSimdWidth = sizeof(vec_t) / sizeof(OutputType);
+
+      constexpr IndexType NumChunks = ceil_to_multiple<IndexType>(InputDimensions, 8) / ChunkSize;
+      constexpr IndexType NumRegs = OutputDimensions / OutputSimdWidth;
+      std::uint16_t nnz[NumChunks];
+      IndexType count;
+
+      const auto input32 = reinterpret_cast<const std::int32_t*>(input);
+
+      // Find indices of nonzero 32bit blocks
+      find_nnz<NumChunks>(input32, nnz, count);
+
+      const vec_t* biasvec = reinterpret_cast<const vec_t*>(biases);
+      vec_t acc[NumRegs];
+      for (IndexType k = 0; k < NumRegs; ++k)
+        acc[k] = biasvec[k];
+
+      for (IndexType j = 0; j < count; ++j)
+      {
+        const auto i = nnz[j];
+        const vec_t in = vec_set_32(input32[i]);
+        const auto col = reinterpret_cast<const vec_t*>(&weights[i * OutputDimensions * ChunkSize]);
+        for (IndexType k = 0; k < NumRegs; ++k)
+          vec_add_dpbusd_32(acc[k], in, col[k]);
+      }
+
+      vec_t* outptr = reinterpret_cast<vec_t*>(output);
+      for (IndexType k = 0; k < NumRegs; ++k)
+        outptr[k] = acc[k];
+# undef vec_setzero
+# undef vec_set_32
+# undef vec_add_dpbusd_32
+#else
+      // Use dense implementation for the other architectures.
+      affine_transform_non_ssse3<
+        InputDimensions,
+        PaddedInputDimensions,
+        OutputDimensions>(output, weights, biases, input);
+#endif
+
+      return output;
+    }
+
+   private:
+    using BiasType = OutputType;
+    using WeightType = std::int8_t;
+
+    alignas(CacheLineSize) BiasType biases[OutputDimensions];
+    alignas(CacheLineSize) WeightType weights[OutputDimensions * PaddedInputDimensions];
+  };
+
+}  // namespace Stockfish::Eval::NNUE::Layers
+
+#endif // #ifndef NNUE_LAYERS_AFFINE_TRANSFORM_SPARSE_INPUT_H_INCLUDED
diff --git a/src/nnue/nnue_architecture.h b/src/nnue/nnue_architecture.h
index d10434f3..413dbb3d 100644
--- a/src/nnue/nnue_architecture.h
+++ b/src/nnue/nnue_architecture.h
@@ -27,6 +27,7 @@
 
 #include "features/half_ka_v2_hm.h"
 
+#include "layers/affine_transform_sparse_input.h"
 #include "layers/affine_transform.h"
 #include "layers/clipped_relu.h"
 #include "layers/sqr_clipped_relu.h"
@@ -48,7 +49,7 @@ struct Network
   static constexpr int FC_0_OUTPUTS = 15;
   static constexpr int FC_1_OUTPUTS = 32;
 
-  Layers::AffineTransform<TransformedFeatureDimensions, FC_0_OUTPUTS + 1> fc_0;
+  Layers::AffineTransformSparseInput<TransformedFeatureDimensions, FC_0_OUTPUTS + 1> fc_0;
   Layers::SqrClippedReLU<FC_0_OUTPUTS + 1> ac_sqr_0;
   Layers::ClippedReLU<FC_0_OUTPUTS + 1> ac_0;
   Layers::AffineTransform<FC_0_OUTPUTS * 2, FC_1_OUTPUTS> fc_1;

From ece90bca9c513fe7b252da1521fc5ff701396f61 Mon Sep 17 00:00:00 2001
From: Michael Chaly <Vizvezdenec@gmail.com>
Date: Mon, 12 Jun 2023 22:13:08 +0300
Subject: [PATCH 177/192] Improve comments

Fix comments for IIR, also document non-linear scaling in extensions.
Also add explicitly the bench, to fix an issue after the last commit.

closes https://github.com/official-stockfish/Stockfish/pull/4614

bench 2370027
---
 src/search.cpp | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/search.cpp b/src/search.cpp
index b2c2344a..7ee6d439 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -822,7 +822,7 @@ namespace {
         }
     }
 
-    // Step 10. If the position doesn't a have ttMove, decrease depth by 2
+    // Step 10. If the position doesn't have a ttMove, decrease depth by 2
     // (or by 4 if the TT entry for the current position was hit and the stored depth is greater than or equal to the current depth).
     // Use qsearch if depth is equal or below zero (~9 Elo)
     if (    PvNode
@@ -1052,6 +1052,9 @@ moves_loop: // When in check, search starts here
           // then that move is singular and should be extended. To verify this we do
           // a reduced search on all the other moves but the ttMove and if the
           // result is lower than ttValue minus a margin, then we will extend the ttMove.
+          // Depth margin and singularBeta margin are known for having non-linear scaling.
+          // Their values are optimized to time controls of 180+1.8 and longer
+          // so changing them requires tests at this type of time controls.
           if (   !rootNode
               &&  depth >= 4 - (thisThread->completedDepth > 22) + 2 * (PvNode && tte->is_pv())
               &&  move == ttMove

From 92c949e12e028cb4556de2786a77f2aec178d59f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?St=C3=A9phane=20Nicolet?= <Stephane.Nicolet@u-paris2.fr>
Date: Mon, 12 Jun 2023 23:26:42 +0200
Subject: [PATCH 178/192] Clean-up code indentation in qsearch

closes https://github.com/official-stockfish/Stockfish/pull/4615

No functional change
---
 src/search.cpp | 152 +++++++++++++++++++++++++------------------------
 1 file changed, 77 insertions(+), 75 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 7ee6d439..d3b5642a 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1481,10 +1481,11 @@ moves_loop: // When in check, search starts here
                 bestValue = ttValue;
         }
         else
+        {
             // In case of null move search use previous static eval with a different sign
-            ss->staticEval = bestValue =
-            (ss-1)->currentMove != MOVE_NULL ? evaluate(pos)
-                                             : -(ss-1)->staticEval;
+            ss->staticEval = bestValue = (ss-1)->currentMove != MOVE_NULL ? evaluate(pos)
+                                                                          : -(ss-1)->staticEval;
+        }
 
         // Stand pat. Return immediately if static value is at least beta
         if (bestValue >= beta)
@@ -1523,97 +1524,98 @@ moves_loop: // When in check, search starts here
     // or a beta cutoff occurs.
     while ((move = mp.next_move()) != MOVE_NONE)
     {
-      assert(is_ok(move));
+        assert(is_ok(move));
 
-      // Check for legality
-      if (!pos.legal(move))
-          continue;
+        // Check for legality
+        if (!pos.legal(move))
+            continue;
 
-      givesCheck = pos.gives_check(move);
-      capture = pos.capture_stage(move);
+        givesCheck = pos.gives_check(move);
+        capture = pos.capture_stage(move);
 
-      moveCount++;
+        moveCount++;
 
-    // Step 6. Pruning.
-    if (bestValue > VALUE_TB_LOSS_IN_MAX_PLY)
-    {
-      // Futility pruning and moveCount pruning (~10 Elo)
-      if (   !givesCheck
-          &&  to_sq(move) != prevSq
-          &&  futilityBase > -VALUE_KNOWN_WIN
-          &&  type_of(move) != PROMOTION)
-      {
-          if (moveCount > 2)
-              continue;
+        // Step 6. Pruning.
+        if (bestValue > VALUE_TB_LOSS_IN_MAX_PLY)
+        {
+            // Futility pruning and moveCount pruning (~10 Elo)
+            if (   !givesCheck
+                &&  to_sq(move) != prevSq
+                &&  futilityBase > -VALUE_KNOWN_WIN
+                &&  type_of(move) != PROMOTION)
+            {
+                if (moveCount > 2)
+                    continue;
 
-          futilityValue = futilityBase + PieceValue[EG][pos.piece_on(to_sq(move))];
+                futilityValue = futilityBase + PieceValue[EG][pos.piece_on(to_sq(move))];
 
-          if (futilityValue <= alpha)
-          {
-              bestValue = std::max(bestValue, futilityValue);
-              continue;
-          }
+                if (futilityValue <= alpha)
+                {
+                    bestValue = std::max(bestValue, futilityValue);
+                    continue;
+                }
 
-          if (futilityBase <= alpha && !pos.see_ge(move, VALUE_ZERO + 1))
-          {
-              bestValue = std::max(bestValue, futilityBase);
-              continue;
-          }
-      }
+                if (futilityBase <= alpha && !pos.see_ge(move, VALUE_ZERO + 1))
+                {
+                    bestValue = std::max(bestValue, futilityBase);
+                    continue;
+                }
+        }
 
-      // We prune after 2nd quiet check evasion where being 'in check' is implicitly checked through the counter
-      // and being a 'quiet' apart from being a tt move is assumed after an increment because captures are pushed ahead.
-      if (quietCheckEvasions > 1)
-          break;
+        // We prune after the second quiet check evasion move, where being 'in check' is
+        // implicitly checked through the counter, and being a 'quiet move' apart from
+        // being a tt move is assumed after an increment because captures are pushed ahead.
+        if (quietCheckEvasions > 1)
+            break;
 
-      // Continuation history based pruning (~3 Elo)
-      if (   !capture
-          && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
-          && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
-          continue;
+        // Continuation history based pruning (~3 Elo)
+        if (   !capture
+            && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
+            && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
+            continue;
 
-      // Do not search moves with bad enough SEE values (~5 Elo)
-      if (!pos.see_ge(move, Value(-95)))
-          continue;
-    }
+        // Do not search moves with bad enough SEE values (~5 Elo)
+        if (!pos.see_ge(move, Value(-95)))
+            continue;
+        }
 
-      // Speculative prefetch as early as possible
-      prefetch(TT.first_entry(pos.key_after(move)));
+        // Speculative prefetch as early as possible
+        prefetch(TT.first_entry(pos.key_after(move)));
 
-      // Update the current move
-      ss->currentMove = move;
-      ss->continuationHistory = &thisThread->continuationHistory[ss->inCheck]
-                                                                [capture]
-                                                                [pos.moved_piece(move)]
-                                                                [to_sq(move)];
+        // Update the current move
+        ss->currentMove = move;
+        ss->continuationHistory = &thisThread->continuationHistory[ss->inCheck]
+                                                                  [capture]
+                                                                  [pos.moved_piece(move)]
+                                                                  [to_sq(move)];
 
-      quietCheckEvasions += !capture && ss->inCheck;
+        quietCheckEvasions += !capture && ss->inCheck;
 
-      // Step 7. Make and search the move
-      pos.do_move(move, st, givesCheck);
-      value = -qsearch<nodeType>(pos, ss+1, -beta, -alpha, depth - 1);
-      pos.undo_move(move);
+        // Step 7. Make and search the move
+        pos.do_move(move, st, givesCheck);
+        value = -qsearch<nodeType>(pos, ss+1, -beta, -alpha, depth - 1);
+        pos.undo_move(move);
 
-      assert(value > -VALUE_INFINITE && value < VALUE_INFINITE);
+        assert(value > -VALUE_INFINITE && value < VALUE_INFINITE);
 
-      // Step 8. Check for a new best move
-      if (value > bestValue)
-      {
-          bestValue = value;
+        // Step 8. Check for a new best move
+        if (value > bestValue)
+        {
+            bestValue = value;
 
-          if (value > alpha)
-          {
-              bestMove = move;
+            if (value > alpha)
+            {
+                bestMove = move;
 
-              if (PvNode) // Update pv even in fail-high case
-                  update_pv(ss->pv, move, (ss+1)->pv);
+                if (PvNode) // Update pv even in fail-high case
+                    update_pv(ss->pv, move, (ss+1)->pv);
 
-              if (PvNode && value < beta) // Update alpha here!
-                  alpha = value;
-              else
-                  break; // Fail high
-          }
-       }
+                if (PvNode && value < beta) // Update alpha here!
+                    alpha = value;
+                else
+                    break; // Fail high
+            }
+        }
     }
 
     // Step 9. Check for mate

From 7922e07af83dd472da6e5b38fb84214cfe46a630 Mon Sep 17 00:00:00 2001
From: Andreas Matthies <a.matthies@online.de>
Date: Tue, 13 Jun 2023 06:24:04 +0200
Subject: [PATCH 179/192] Fix for MSVC compilation.

MSVC needs two more explicit casts to compile new affine_transform_sparse_input.
See https://www.intel.com/content/www/us/en/docs/intrinsics-guide/index.html#text=_mm256_castsi256_ps
and https://www.intel.com/content/www/us/en/docs/intrinsics-guide/index.html#text=_mm_castsi128_ps

closes https://github.com/official-stockfish/Stockfish/pull/4616

No functional change
---
 AUTHORS                                         | 1 +
 src/nnue/layers/affine_transform_sparse_input.h | 4 ++--
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/AUTHORS b/AUTHORS
index 63b862ce..a89dc130 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -19,6 +19,7 @@ Alexander Kure
 Alexander Pagel (Lolligerhans)
 Alfredo Menezes (lonfom169)
 Ali AlZhrani (Cooffe)
+Andreas Matthies (Matthies)
 Andrei Vetrov (proukornew)
 Andrew Grant (AndyGrant)
 Andrey Neporada (nepal)
diff --git a/src/nnue/layers/affine_transform_sparse_input.h b/src/nnue/layers/affine_transform_sparse_input.h
index 00b17c19..e0c3a8a0 100644
--- a/src/nnue/layers/affine_transform_sparse_input.h
+++ b/src/nnue/layers/affine_transform_sparse_input.h
@@ -98,10 +98,10 @@ namespace Stockfish::Eval::NNUE::Layers {
     #define vec_nnz(a) _mm512_cmpgt_epi32_mask(a, _mm512_setzero_si512())
 #elif defined (USE_AVX2)
     using vec_t = __m256i;
-    #define vec_nnz(a) _mm256_movemask_ps((__m256)_mm256_cmpgt_epi32(a, _mm256_setzero_si256()))
+    #define vec_nnz(a) _mm256_movemask_ps(_mm256_castsi256_ps(_mm256_cmpgt_epi32(a, _mm256_setzero_si256())))
 #elif defined (USE_SSSE3)
     using vec_t = __m128i;
-    #define vec_nnz(a) _mm_movemask_ps((__m128)_mm_cmpgt_epi32(a, _mm_setzero_si128()))
+    #define vec_nnz(a) _mm_movemask_ps(_mm_castsi128_ps(_mm_cmpgt_epi32(a, _mm_setzero_si128())))
 #endif
     constexpr IndexType InputSimdWidth = sizeof(vec_t) / sizeof(std::int32_t);
     // Inputs are processed InputSimdWidth at a time and outputs are processed 8 at a time so we process in chunks of max(InputSimdWidth, 8)

From 887bbd8b3df8a01307a38bfe529a49842f810a9c Mon Sep 17 00:00:00 2001
From: Viren6 <94880762+Viren6@users.noreply.github.com>
Date: Tue, 13 Jun 2023 22:26:20 +0100
Subject: [PATCH 180/192] Remove setting of static to none if in check in
 qsearch

Small simplification

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/6487924d713491385c8034ae
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 59616 W: 15885 L: 15703 D: 28028
Ptnml(0-2): 144, 6130, 17086, 6296, 152

closes https://github.com/official-stockfish/Stockfish/pull/4618

No functional change.
---
 AUTHORS        | 1 +
 src/search.cpp | 1 -
 2 files changed, 1 insertion(+), 1 deletion(-)

diff --git a/AUTHORS b/AUTHORS
index a89dc130..ff224954 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -215,6 +215,7 @@ tttak
 Unai Corzo (unaiic)
 Uri Blass (uriblass)
 Vince Negri (cuddlestmonkey)
+Viren
 windfishballad
 xefoci7612
 zz4032
diff --git a/src/search.cpp b/src/search.cpp
index d3b5642a..5de950eb 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1464,7 +1464,6 @@ moves_loop: // When in check, search starts here
     // Step 4. Static evaluation of the position
     if (ss->inCheck)
     {
-        ss->staticEval = VALUE_NONE;
         bestValue = futilityBase = -VALUE_INFINITE;
     }
     else

From 14bfec2a981e906d1bfc08331a2e15bddd07ffe4 Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Wed, 14 Jun 2023 15:57:36 +0300
Subject: [PATCH 181/192] Consistent bench extraction with fishtest.

Consistent with recent fishtest commit https://github.com/glinscott/fishtest/commit/c0d174396f7fb1c0b3243aaa6cc73769079f3ff9

closes https://github.com/official-stockfish/Stockfish/pull/4619

No functional change
---
 .github/workflows/stockfish_test.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/stockfish_test.yml b/.github/workflows/stockfish_test.yml
index 8c383fe7..28218402 100644
--- a/.github/workflows/stockfish_test.yml
+++ b/.github/workflows/stockfish_test.yml
@@ -115,7 +115,7 @@ jobs:
 
       - name: Extract the bench number from the commit history
         run: |
-          git log HEAD | grep "\b[Bb]ench[ :]\+[0-9]\{7\}" | head -n 1 | sed "s/[^0-9]*\([0-9]*\).*/\1/g" > git_sig
+          git log HEAD | grep -o "\b[Bb]ench[ :]\+[1-9][0-9]\{5,9\}\b" | head -n 1 | sed "s/[^0-9]//g" > git_sig
           [ -s git_sig ] && echo "benchref=$(cat git_sig)" >> $GITHUB_ENV && echo "Reference bench:" $(cat git_sig) || echo "No bench found"
 
       - name: Check compiler

From 32d3284df5b2fd395504efa5319d64856902fef1 Mon Sep 17 00:00:00 2001
From: AndrovT <31534597+AndrovT@users.noreply.github.com>
Date: Tue, 13 Jun 2023 03:00:04 +0200
Subject: [PATCH 182/192] Permute master net weights to increase sparsity

Activation data collection using https://github.com/AndrovT/Stockfish/commit/ac468039ab544b03ad9a22c859a4217729c10a77 run as

bench 16 1 13 varied_1000.epd depth NNUE log.bin

on FENs from https://gist.github.com/AndrovT/7eae6918eb50764227e2bafe7938953c.

Permutation found using https://gist.github.com/AndrovT/359c831b7223c637e9156b01eb96949e.
Uses a greedy algorithm that goes sequentially through the output positions and
chooses a neuron for that position such that the number of nonzero quartets is the smallest.

Net weights permuted using https://gist.github.com/AndrovT/9e3fbaebb7082734dc84d27e02094cb3.

Benchmark:

Result of 100 runs of 'bench 16 1 13 default depth NNUE'
========================================================
base (...kfish-master) =     885869  +/- 7395
test (./stockfish    ) =     895885  +/- 7368
diff                   =     +10016  +/- 2984

speedup        = +0.0113
P(speedup > 0) =  1.0000

Passed STC:
https://tests.stockfishchess.org/tests/view/648866c4713491385c804728
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 126784 W: 34003 L: 33586 D: 59195
Ptnml(0-2): 283, 13001, 36437, 13358, 313

closes https://github.com/official-stockfish/Stockfish/pull/4620

No functional change.
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 94cd42cc..c35b2f07 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-fdc1d0fe6455.nnue"
+  #define EvalFileDefaultName   "nn-cd2ff4716c34.nnue"
 
   namespace NNUE {
 

From 0187546275cb015d42a7d99789f2f6a650b03651 Mon Sep 17 00:00:00 2001
From: cj5716 <125858804+cj5716@users.noreply.github.com>
Date: Thu, 15 Jun 2023 21:05:01 +0800
Subject: [PATCH 183/192] Small cleanup

This non-functional change keeps formatting consistent.

closes https://github.com/official-stockfish/Stockfish/pull/4623

Bench 2370027
---
 src/search.cpp | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 5de950eb..8ace674d 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1463,9 +1463,7 @@ moves_loop: // When in check, search starts here
 
     // Step 4. Static evaluation of the position
     if (ss->inCheck)
-    {
         bestValue = futilityBase = -VALUE_INFINITE;
-    }
     else
     {
         if (ss->ttHit)
@@ -1480,11 +1478,9 @@ moves_loop: // When in check, search starts here
                 bestValue = ttValue;
         }
         else
-        {
             // In case of null move search use previous static eval with a different sign
             ss->staticEval = bestValue = (ss-1)->currentMove != MOVE_NULL ? evaluate(pos)
                                                                           : -(ss-1)->staticEval;
-        }
 
         // Stand pat. Return immediately if static value is at least beta
         if (bestValue >= beta)

From a46087ee30ace00e1edd8963ee33b0a1b87031b6 Mon Sep 17 00:00:00 2001
From: maxim <mmo@2021.ai>
Date: Tue, 13 Jun 2023 23:09:05 +0300
Subject: [PATCH 184/192] Compressed network parameters

Implemented LEB128 (de)compression for the feature transformer.
Reduces embedded network size from 70 MiB to 39 Mib.

The new nn-78bacfcee510.nnue corresponds to the master net compressed.

closes https://github.com/official-stockfish/Stockfish/pull/4617

No functional change
---
 src/evaluate.h                      |  2 +-
 src/nnue/nnue_common.h              | 77 +++++++++++++++++++++++++++++
 src/nnue/nnue_feature_transformer.h | 12 ++---
 3 files changed, 84 insertions(+), 7 deletions(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index c35b2f07..33effb1c 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-cd2ff4716c34.nnue"
+  #define EvalFileDefaultName   "nn-78bacfcee510.nnue"
 
   namespace NNUE {
 
diff --git a/src/nnue/nnue_common.h b/src/nnue/nnue_common.h
index 12309d26..d338527d 100644
--- a/src/nnue/nnue_common.h
+++ b/src/nnue/nnue_common.h
@@ -57,6 +57,9 @@ namespace Stockfish::Eval::NNUE {
   // Size of cache line (in bytes)
   constexpr std::size_t CacheLineSize = 64;
 
+  constexpr const char Leb128MagicString[] = "COMPRESSED_LEB128";
+  constexpr const std::size_t Leb128MagicStringSize = sizeof(Leb128MagicString) - 1;
+
   // SIMD width (in bytes)
   #if defined(USE_AVX2)
   constexpr std::size_t SimdWidth = 32;
@@ -159,6 +162,80 @@ namespace Stockfish::Eval::NNUE {
               write_little_endian<IntType>(stream, values[i]);
   }
 
+  template <typename IntType>
+  inline void read_leb_128(std::istream& stream, IntType* out, std::size_t count) {
+      static_assert(std::is_signed_v<IntType>, "Not implemented for unsigned types");
+      char leb128MagicString[Leb128MagicStringSize];
+      stream.read(leb128MagicString, Leb128MagicStringSize);
+      assert(strncmp(Leb128MagicString, leb128MagicString, Leb128MagicStringSize) == 0);
+      const std::uint32_t BUF_SIZE = 4096;
+      std::uint8_t buf[BUF_SIZE];
+      auto bytes_left = read_little_endian<std::uint32_t>(stream);
+      std::uint32_t buf_pos = BUF_SIZE;
+      for (std::size_t i = 0; i < count; ++i) {
+          IntType result = 0;
+          size_t shift = 0;
+          do {
+              if (buf_pos == BUF_SIZE) {
+                  stream.read(reinterpret_cast<char*>(buf), std::min(bytes_left, BUF_SIZE));
+                  buf_pos = 0;
+              }
+              std::uint8_t byte = buf[buf_pos++];
+              --bytes_left;
+              result |= (byte & 0x7f) << shift;
+              shift += 7;
+              if ((byte & 0x80) == 0) {
+                  out[i] = sizeof(IntType) * 8 <= shift || (byte & 0x40) == 0 ? result : result | ~((1 << shift) - 1);
+                  break;
+              }
+          } while (shift < sizeof(IntType) * 8);
+      }
+      assert(bytes_left == 0);
+  }
+
+  template <typename IntType>
+  inline void write_leb_128(std::ostream& stream, const IntType* values, std::size_t count) {
+      static_assert(std::is_signed_v<IntType>, "Not implemented for unsigned types");
+      stream.write(Leb128MagicString, Leb128MagicStringSize);
+      std::uint32_t byte_count = 0;
+      for (std::size_t i = 0; i < count; ++i) {
+          IntType value = values[i];
+          std::uint8_t byte;
+          do {
+              byte = value & 0x7f;
+              value >>= 7;
+              ++byte_count;
+          } while ((byte & 0x40) == 0 ? value != 0 : value != -1);
+      }
+      write_little_endian(stream, byte_count);
+      const std::uint32_t BUF_SIZE = 4096;
+      std::uint8_t buf[BUF_SIZE];
+      std::uint32_t buf_pos = 0;
+      auto flush = [&]() {
+          if (buf_pos > 0) {
+              stream.write(reinterpret_cast<char*>(buf), buf_pos);
+              buf_pos = 0;
+          }
+      };
+      auto write = [&](std::uint8_t byte) {
+          buf[buf_pos++] = byte;
+          if (buf_pos == BUF_SIZE) flush();
+      };
+      for (std::size_t i = 0; i < count; ++i) {
+          IntType value = values[i];
+          while (true) {
+              std::uint8_t byte = value & 0x7f;
+              value >>= 7;
+              if ((byte & 0x40) == 0 ? value == 0 : value == -1) {
+                  write(byte);
+                  break;
+              }
+              write(byte | 0x80);
+          }
+      }
+      flush();
+  }
+
 }  // namespace Stockfish::Eval::NNUE
 
 #endif // #ifndef NNUE_COMMON_H_INCLUDED
diff --git a/src/nnue/nnue_feature_transformer.h b/src/nnue/nnue_feature_transformer.h
index a1888c7a..7571f398 100644
--- a/src/nnue/nnue_feature_transformer.h
+++ b/src/nnue/nnue_feature_transformer.h
@@ -253,9 +253,9 @@ namespace Stockfish::Eval::NNUE {
     // Read network parameters
     bool read_parameters(std::istream& stream) {
 
-      read_little_endian<BiasType      >(stream, biases     , HalfDimensions                  );
-      read_little_endian<WeightType    >(stream, weights    , HalfDimensions * InputDimensions);
-      read_little_endian<PSQTWeightType>(stream, psqtWeights, PSQTBuckets    * InputDimensions);
+      read_leb_128<BiasType      >(stream, biases     , HalfDimensions                  );
+      read_leb_128<WeightType    >(stream, weights    , HalfDimensions * InputDimensions);
+      read_leb_128<PSQTWeightType>(stream, psqtWeights, PSQTBuckets    * InputDimensions);
 
       return !stream.fail();
     }
@@ -263,9 +263,9 @@ namespace Stockfish::Eval::NNUE {
     // Write network parameters
     bool write_parameters(std::ostream& stream) const {
 
-      write_little_endian<BiasType      >(stream, biases     , HalfDimensions                  );
-      write_little_endian<WeightType    >(stream, weights    , HalfDimensions * InputDimensions);
-      write_little_endian<PSQTWeightType>(stream, psqtWeights, PSQTBuckets    * InputDimensions);
+      write_leb_128<BiasType      >(stream, biases     , HalfDimensions                  );
+      write_leb_128<WeightType    >(stream, weights    , HalfDimensions * InputDimensions);
+      write_leb_128<PSQTWeightType>(stream, psqtWeights, PSQTBuckets    * InputDimensions);
 
       return !stream.fail();
     }

From a68a1c11543eef6808181c92e0e7e5fb3f826f21 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Tue, 13 Jun 2023 19:30:01 +0200
Subject: [PATCH 185/192] create prereleases upon push to master

using github actions, create a prerelease for the latest commit to master.
As such a development version will be available on github, in addition to the latest release.

closes https://github.com/official-stockfish/Stockfish/pull/4622

No functional change
---
 .github/workflows/stockfish.yml              | 24 +++++++++
 .github/workflows/stockfish_arm_binaries.yml | 26 +++++++++
 .github/workflows/stockfish_binaries.yml     | 57 ++++++++++++++++++--
 3 files changed, 104 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/stockfish.yml b/.github/workflows/stockfish.yml
index 082c65de..ca52ffe0 100644
--- a/.github/workflows/stockfish.yml
+++ b/.github/workflows/stockfish.yml
@@ -12,6 +12,30 @@ on:
       - master
       - tools
 jobs:
+  Prerelease:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 0
+
+      # returns null if no pre-release exists
+      - name: Get Commit SHA of Latest Pre-release
+        run: |
+          # Install required packages
+          sudo apt-get update
+          sudo apt-get install -y curl jq
+
+          echo "COMMIT_SHA=$(jq -r 'map(select(.prerelease)) | first | .tag_name' <<< $(curl -s https://api.github.com/repos/${{ github.repository_owner }}/Stockfish/releases))" >> $GITHUB_ENV
+
+      # delete old previous pre-release and tag
+      - uses: dev-drprasad/delete-tag-and-release@v0.2.1
+        if: env.COMMIT_SHA != 'null'
+        with:
+          tag_name: ${{ env.COMMIT_SHA }}
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
   Sanitizers:
     uses: ./.github/workflows/stockfish_sanitizers.yml
   Tests:
diff --git a/.github/workflows/stockfish_arm_binaries.yml b/.github/workflows/stockfish_arm_binaries.yml
index 9a4734ee..52105eb6 100644
--- a/.github/workflows/stockfish_arm_binaries.yml
+++ b/.github/workflows/stockfish_arm_binaries.yml
@@ -129,4 +129,30 @@ jobs:
         if: startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag'
         uses: softprops/action-gh-release@v1
         with:
+          files: stockfish-android-${{ matrix.binaries }}.tar
+
+      - name: Get last commit sha
+        id: last_commit
+        run: echo "COMMIT_SHA=$(git rev-parse HEAD | cut -c 1-8)" >> $GITHUB_ENV
+
+      - name: Get commit date
+        id: commit_date
+        run: echo "COMMIT_DATE=$(git show -s --date=format:'%Y%m%d' --format=%cd HEAD)" >> $GITHUB_ENV
+
+      # Make sure that an old ci which still runs on master doesn't recreate a prerelease
+      - name: Check Pullable Commits
+        id: check_commits
+        run: |
+          git fetch
+          CHANGES=$(git rev-list HEAD..origin/master --count)
+          echo "CHANGES=$CHANGES" >> $GITHUB_ENV
+
+      - name: Prerelease
+        if: github.ref_name == 'master' && env.CHANGES == '0'
+        continue-on-error: true
+        uses: softprops/action-gh-release@v1
+        with:
+          name: Stockfish dev-${{ env.COMMIT_DATE }}-${{ env.COMMIT_SHA }}
+          tag_name: stockfish-dev-${{ env.COMMIT_DATE }}-${{ env.COMMIT_SHA }}
+          prerelease: true
           files: stockfish-android-${{ matrix.binaries }}.tar
\ No newline at end of file
diff --git a/.github/workflows/stockfish_binaries.yml b/.github/workflows/stockfish_binaries.yml
index 86449b97..0a53cb03 100644
--- a/.github/workflows/stockfish_binaries.yml
+++ b/.github/workflows/stockfish_binaries.yml
@@ -20,12 +20,14 @@ jobs:
             compiler: g++
             comp: gcc
             shell: bash {0}
+            archive_ext: tar
           - name: MacOS 12 Apple Clang
             os: macos-12
             simple_name: macos
             compiler: clang++
             comp: clang
             shell: bash {0}
+            archive_ext: tar
           - name: Windows 2022 Mingw-w64 GCC x86_64
             os: windows-2022
             simple_name: windows
@@ -35,6 +37,7 @@ jobs:
             msys_env: x86_64-gcc
             shell: msys2 {0}
             ext: .exe
+            archive_ext: zip
         binaries:
           - x86-64
           - x86-64-modern
@@ -60,7 +63,7 @@ jobs:
         uses: msys2/setup-msys2@v2
         with:
           msystem: ${{ matrix.config.msys_sys }}
-          install: mingw-w64-${{ matrix.config.msys_env }} make git
+          install: mingw-w64-${{ matrix.config.msys_env }} make git zip
 
       - name: Download the used network from the fishtest framework
         run: make net
@@ -90,7 +93,7 @@ jobs:
           git clone https://github.com/official-stockfish/Stockfish.wiki.git ../wiki
           rm -rf ../wiki/.git
 
-      - name: Create tar archive.
+      - name: Create directory.
         run: |
           cd ..
           mkdir stockfish
@@ -102,16 +105,64 @@ jobs:
           cp AUTHORS stockfish/
           cp CITATION.cff stockfish/
           cp README.md stockfish/
+
+      - name: Create tar
+        if: runner.os != 'Windows'
+        run: |
+          cd ..
           tar -cvf stockfish-$NAME-$BINARY.tar stockfish
 
+      - name: Create zip
+        if: runner.os == 'Windows'
+        run: |
+          cd ..
+          zip -r stockfish-$NAME-$BINARY.zip stockfish
+
       - name: Upload binaries
+        if: runner.os != 'Windows'
         uses: actions/upload-artifact@v3
         with:
           name: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}
           path: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.tar
 
+      # Artifacts automatically get zipped
+      # to avoid double zipping, we use the unzipped directory
+      - name: Upload binaries
+        if: runner.os == 'Windows'
+        uses: actions/upload-artifact@v3
+        with:
+          name: stockfish-${{ matrix.config.os }}-${{ matrix.binaries }}
+          path: stockfish
+
       - name: Release
         if: startsWith(github.ref_name, 'sf_') && github.ref_type == 'tag'
         uses: softprops/action-gh-release@v1
         with:
-          files: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.tar
+          files: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.${{ matrix.config.archive_ext }}
+
+      - name: Get last commit sha
+        id: last_commit
+        run: echo "COMMIT_SHA=$(git rev-parse HEAD | cut -c 1-8)" >> $GITHUB_ENV
+
+      - name: Get commit date
+        id: commit_date
+        run: echo "COMMIT_DATE=$(git show -s --date=format:'%Y%m%d' --format=%cd HEAD)" >> $GITHUB_ENV
+
+      # Make sure that an old ci which still runs on master doesn't recreate a prerelease
+      - name: Check Pullable Commits
+        id: check_commits
+        run: |
+          git fetch
+          CHANGES=$(git rev-list HEAD..origin/master --count)
+          echo "CHANGES=$CHANGES" >> $GITHUB_ENV
+
+      - name: Prerelease
+        if: github.ref_name == 'master' && env.CHANGES == '0'
+        continue-on-error: true
+        uses: softprops/action-gh-release@v1
+        with:
+          name: Stockfish dev-${{ env.COMMIT_DATE }}-${{ env.COMMIT_SHA }}
+          tag_name: stockfish-dev-${{ env.COMMIT_DATE }}-${{ env.COMMIT_SHA }}
+          prerelease: true
+          files: stockfish-${{ matrix.config.simple_name }}-${{ matrix.binaries }}.${{ matrix.config.archive_ext }}
+

From 6eaa1c3ecd297404b28f9e80cddf81c4c6926a51 Mon Sep 17 00:00:00 2001
From: Joerg Oster <osterj165@googlemail.com>
Date: Tue, 20 Jun 2023 10:40:31 +0200
Subject: [PATCH 186/192] Fix indentation in qsearch.

https://github.com/official-stockfish/Stockfish/pull/4630

No functional change.
---
 src/search.cpp | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 8ace674d..9b686a52 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -1555,23 +1555,23 @@ moves_loop: // When in check, search starts here
                     bestValue = std::max(bestValue, futilityBase);
                     continue;
                 }
-        }
+            }
 
-        // We prune after the second quiet check evasion move, where being 'in check' is
-        // implicitly checked through the counter, and being a 'quiet move' apart from
-        // being a tt move is assumed after an increment because captures are pushed ahead.
-        if (quietCheckEvasions > 1)
-            break;
+            // We prune after the second quiet check evasion move, where being 'in check' is
+            // implicitly checked through the counter, and being a 'quiet move' apart from
+            // being a tt move is assumed after an increment because captures are pushed ahead.
+            if (quietCheckEvasions > 1)
+                break;
 
-        // Continuation history based pruning (~3 Elo)
-        if (   !capture
-            && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
-            && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
-            continue;
+            // Continuation history based pruning (~3 Elo)
+            if (   !capture
+                && (*contHist[0])[pos.moved_piece(move)][to_sq(move)] < 0
+                && (*contHist[1])[pos.moved_piece(move)][to_sq(move)] < 0)
+                continue;
 
-        // Do not search moves with bad enough SEE values (~5 Elo)
-        if (!pos.see_ge(move, Value(-95)))
-            continue;
+            // Do not search moves with bad enough SEE values (~5 Elo)
+            if (!pos.see_ge(move, Value(-95)))
+                continue;
         }
 
         // Speculative prefetch as early as possible

From aec8fb3fdec8309a0cc78222a4f674ea6fea9411 Mon Sep 17 00:00:00 2001
From: disservin <disservin.social@gmail.com>
Date: Tue, 20 Jun 2023 18:27:20 +0200
Subject: [PATCH 187/192] Fix failing CI of pull requests

adds a guard to prevent pull requests from trying to delete the previous pre-release

closing https://github.com/official-stockfish/Stockfish/pull/4631

No functional change.
---
 .github/workflows/stockfish.yml | 1 +
 1 file changed, 1 insertion(+)

diff --git a/.github/workflows/stockfish.yml b/.github/workflows/stockfish.yml
index ca52ffe0..99c4259a 100644
--- a/.github/workflows/stockfish.yml
+++ b/.github/workflows/stockfish.yml
@@ -13,6 +13,7 @@ on:
       - tools
 jobs:
   Prerelease:
+    if: github.ref == 'refs/heads/master'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v3

From 02728736edd5915fc0abc8698635e633a3cba201 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Tue, 20 Jun 2023 09:46:02 +0200
Subject: [PATCH 188/192] Update top CPU contributors

closes https://github.com/official-stockfish/Stockfish/pull/4629

No functional change
---
 Top CPU Contributors.txt | 117 ++++++++++++++++++++-------------------
 1 file changed, 59 insertions(+), 58 deletions(-)

diff --git a/Top CPU Contributors.txt b/Top CPU Contributors.txt
index 7b279590..74c471b7 100644
--- a/Top CPU Contributors.txt	
+++ b/Top CPU Contributors.txt	
@@ -1,55 +1,55 @@
-Contributors to Fishtest with >10,000 CPU hours, as of 2023-05-27.
+Contributors to Fishtest with >10,000 CPU hours, as of 2023-06-20.
 Thank you!
 
 Username                                CPU Hours     Games played
 ------------------------------------------------------------------
-noobpwnftw                               37304027       2833556221
-technologov                              13508659        714674674
-linrock                                   4121386        280027751
+noobpwnftw                               37457426       2850540907
+technologov                              14135647        742892808
+linrock                                   4423514        303254809
 mlang                                     3026000        200065824
 dew                                       1689162        100033738
-okrout                                    1541122        145085726
-pemo                                      1481818         47546583
-grandphish2                               1459364         91364265
-TueRens                                   1178700         69951886
-JojoM                                      937875         60821044
+okrout                                    1578136        148855886
+pemo                                      1508508         48814305
+grandphish2                               1461406         91540343
+TueRens                                   1194790         70400852
+JojoM                                      947612         61773190
 tvijlbrief                                 796125         51897690
+sebastronomy                               742434         38218524
 mibere                                     703840         46867607
-sebastronomy                               687502         35585318
-gvreuls                                    645570         42437926
-oz                                         541224         39133532
-cw                                         517856         34869499
+gvreuls                                    651026         42988582
+oz                                         543438         39314736
+cw                                         517858         34869755
 fastgm                                     503862         30260818
-CSU_Dynasty                                464691         31166478
-leszek                                     460426         32840277
-ctoks                                      434323         28497451
+leszek                                     467278         33514883
+CSU_Dynasty                                464940         31177118
+ctoks                                      434416         28506889
 crunchy                                    427035         27344275
-maximmasiutin                              424154         26534660
+maximmasiutin                              424795         26577722
 bcross                                     415722         29060963
-rpngn                                      344368         24218047
-velislav                                   342559         22138408
+olafm                                      395922         32268020
+rpngn                                      348378         24560289
+velislav                                   342567         22138992
 Fisherman                                  327231         21829379
-mgrabiak                                   297057         20260882
+mgrabiak                                   300612         20608380
 Dantist                                    296386         18031762
-nordlandia                                 242642         15922516
-robal                                      240199         15544104
+nordlandia                                 246201         16189678
+robal                                      241300         15656382
 marrco                                     234581         17714473
 ncfish1                                    227517         15233777
 glinscott                                  208125         13277240
 drabel                                     204167         13930674
 mhoram                                     202894         12601997
 bking_US                                   198894         11876016
-olafm                                      192342         14968698
 Thanar                                     179852         12365359
 vdv                                        175544          9904472
 spams                                      157128         10319326
 sqrt2                                      147963          9724586
-DesolatedDodo                              144759          9408038
+DesolatedDodo                              146350          9536172
 Calis007                                   143165          9478764
-vdbergh                                    138436          9042073
+vdbergh                                    138650          9064413
 CoffeeOne                                  137100          5024116
+armo9494                                   136191          9460264
 malala                                     136182          8002293
-armo9494                                   136010          9447548
 xoto                                       133759          9159372
 davar                                      129023          8376525
 DMBK                                       122960          8980062
@@ -59,43 +59,43 @@ Data                                       113305          8220352
 BrunoBanani                                112960          7436849
 CypressChess                               108331          7759788
 skiminki                                   107583          7218170
+jcAEie                                     105675          8238962
 MaZePallas                                 102823          6633619
 sterni1971                                 100532          5880772
-jcAEie                                     100392          7788270
 sunu                                       100167          7040199
 zeryl                                       99331          6221261
 thirdlife                                   99124          2242380
 ElbertoOne                                  99028          7023771
-cuistot                                     98360          6017102
+cuistot                                     98853          6069816
 bigpen0r                                    94809          6529203
 brabos                                      92118          6186135
-Wolfgang                                    90855          5998076
+Wolfgang                                    91939          6105872
 psk                                         89957          5984901
+sschnee                                     88235          5268000
 racerschmacer                               85805          6122790
+Fifis                                       85722          5709729
 Dubslow                                     84986          6042456
 Vizvezdenec                                 83761          5344740
-sschnee                                     83564          4853834
 0x3C33                                      82614          5271253
 BRAVONE                                     81239          5054681
-Fifis                                       77355          5158211
 nssy                                        76497          5259388
 jromang                                     76106          5236025
 teddybaer                                   75125          5407666
+tolkki963                                   74762          5149662
+megaman7de                                  74351          4940352
 Wencey                                      74181          4711488
-megaman7de                                  73866          4894960
 Pking_cda                                   73776          5293873
-tolkki963                                   73531          5020500
-yurikvelo                                   72847          4972808
+yurikvelo                                   73150          5004382
+markkulix                                   72607          5304642
 Bobo1239                                    70579          4794999
 solarlight                                  70517          5028306
 dv8silencer                                 70287          3883992
-markkulix                                   70278          5068326
 manap                                       66273          4121774
 tinker                                      64333          4268790
 qurashee                                    61208          3429862
-Mineta                                      58759          4399960
+Mineta                                      59357          4418202
+Spprtr                                      58723          3911011
 AGI                                         58147          4325994
-Spprtr                                      58106          3858759
 robnjr                                      57262          4053117
 Freja                                       56938          3733019
 MaxKlaxxMiner                               56879          3423958
@@ -103,35 +103,35 @@ MarcusTullius                               56746          3762951
 ttruscott                                   56010          3680085
 rkl                                         55132          4164467
 renouve                                     53811          3501516
+javran                                      53785          4627608
 finfish                                     51360          3370515
 eva42                                       51272          3599691
 eastorwest                                  51117          3454811
 rap                                         49985          3219146
 pb00067                                     49733          3298934
-javran                                      49178          4190632
-OuaisBla                                    48606          3442958
+OuaisBla                                    48626          3445134
 ronaldjerum                                 47654          3240695
 biffhero                                    46564          3111352
 VoyagerOne                                  45476          3452465
-oryx                                        44532          3450170
-jmdana                                      43849          2955821
+jmdana                                      44893          3065205
+maposora                                    44597          4039578
+oryx                                        44570          3454238
 speedycpu                                   43842          3003273
 jbwiebe                                     43305          2805433
+GPUex                                       42378          3133332
 Antihistamine                               41788          2761312
 mhunt                                       41735          2691355
-maposora                                    41534          3733078
-GPUex                                       41061          2998356
 homyur                                      39893          2850481
 gri                                         39871          2515779
 Garf                                        37741          2999686
 SC                                          37299          2731694
 csnodgrass                                  36207          2688994
 strelock                                    34716          2074055
+szupaw                                      34102          2880346
 EthanOConnor                                33370          2090311
 slakovv                                     32915          2021889
 Gelma                                       31771          1551204
 gopeto                                      31671          2060990
-szupaw                                      31248          2594920
 kdave                                       31157          2198362
 manapbk                                     30987          1810399
 Prcuvu                                      30377          2170122
@@ -147,52 +147,54 @@ xwziegtm                                    26897          2124586
 achambord                                   26582          1767323
 Patrick_G                                   26276          1801617
 yorkman                                     26193          1992080
-Ulysses                                     25285          1689346
+Ulysses                                     25288          1689730
 SFTUser                                     25182          1675689
 nabildanial                                 24942          1519409
 Sharaf_DG                                   24765          1786697
+Maxim                                       24705          1502062
 rodneyc                                     24376          1416402
 agg177                                      23890          1395014
+Goatminola                                  23763          1956036
 Ente                                        23639          1671638
+Jopo12321                                   23467          1483172
 JanErik                                     23408          1703875
 Isidor                                      23388          1680691
 Norabor                                     23371          1603244
-Goatminola                                  23338          1910634
 cisco2015                                   22920          1763301
-Jopo12321                                   22890          1424926
+jsys14                                      22824          1591906
 Zirie                                       22542          1472937
 team-oh                                     22272          1636708
 Roady                                       22220          1465606
 MazeOfGalious                               21978          1629593
 sg4032                                      21947          1643353
-jsys14                                      21935          1499128
 ianh2105                                    21725          1632562
 xor12                                       21628          1680365
 dex                                         21612          1467203
 nesoneg                                     21494          1463031
 user213718                                  21454          1404128
 sphinx                                      21211          1384728
+AndreasKrug                                 21097          1634811
 jjoshua2                                    21001          1423089
 Zake9298                                    20938          1565848
-AndreasKrug                                 20911          1615673
 horst.prack                                 20878          1465656
 0xB00B1ES                                   20590          1208666
 j3corre                                     20405           941444
 Adrian.Schmidt123                           20316          1281436
 wei                                         19973          1745989
+notchris                                    19958          1800128
 Serpensin                                   19840          1697528
+Gaster319                                   19712          1677310
 fishtester                                  19617          1257388
 rstoesser                                   19569          1293588
 eudhan                                      19274          1283717
-Gaster319                                   18934          1596772
+votoanthuan                                 19108          1609992
 vulcan                                      18871          1729392
 Karpovbot                                   18766          1053178
+qoo_charly_cai                              18543          1284937
 jundery                                     18445          1115855
-votoanthuan                                 18012          1508836
 ville                                       17883          1384026
 chris                                       17698          1487385
 purplefishies                               17595          1092533
-qoo_charly_cai                              17494          1182667
 dju                                         17414           981289
 iisiraider                                  17275          1049015
 DragonLord                                  17014          1162790
@@ -200,7 +202,6 @@ redstone59                                  16842          1461780
 Alb11747                                    16787          1213926
 IgorLeMasson                                16064          1147232
 Karby                                       15982           979610
-notchris                                    15818          1426762
 scuzzi                                      15757           968735
 ako027ako                                   15671          1173203
 Nikolay.IT                                  15154          1068349
@@ -218,33 +219,33 @@ Nesa92                                      13786          1114691
 joster                                      13710           946160
 mbeier                                      13650          1044928
 Hjax                                        13535           915487
+Nullvalue                                   13468          1140498
 Dark_wizzie                                 13422          1007152
 Rudolphous                                  13244           883140
 pirt                                        13100          1009897
 Machariel                                   13010           863104
 infinigon                                   12991           943216
-Maxim                                       12963           985594
 mabichito                                   12903           749391
 thijsk                                      12886           722107
 AdrianSA                                    12860           804972
 Flopzee                                     12698           894821
 korposzczur                                 12606           838168
-Nullvalue                                   12583          1048502
 fatmurphy                                   12547           853210
 SapphireBrand                               12416           969604
 Oakwen                                      12399           844109
 deflectooor                                 12386           579392
 modolief                                    12386           896470
 Farseer                                     12249           694108
-Jackfish                                    12180           801372
+Jackfish                                    12213           805008
 pgontarz                                    12151           848794
 dbernier                                    12103           860824
 getraideBFF                                 12072          1024966
 stocky                                      11954           699440
 mschmidt                                    11941           803401
 MooTheCow                                   11870           773598
-FormazChar                                  11689           877727
+FormazChar                                  11766           885707
 whelanh                                     11557           245188
+3cho                                        11494          1031076
 infinity                                    11470           727027
 aga                                         11412           695127
 torbjo                                      11395           729145
@@ -254,6 +255,7 @@ d64                                         11263           789184
 ali-al-zhrani                               11245           779246
 snicolet                                    11106           869170
 dapper                                      11032           771402
+ols                                         10947           624903
 Karmatron                                   10828           677458
 basepi                                      10637           744851
 Cubox                                       10621           826448
@@ -263,4 +265,3 @@ jojo2357                                    10419           929708
 WoodMan777                                  10380           873720
 Garruk                                      10365           706465
 dzjp                                        10343           732529
-ols                                         10259           570669

From 52e84e4b4675aae52a619c309479684dc5478bf5 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Thu, 22 Jun 2023 09:59:03 +0200
Subject: [PATCH 189/192] Update winrate model with June data

Retained 748191776 scored positions for analysis

const int NormalizeToPawnValue = 328;
Corresponding spread = 60;
Corresponding normalized spread = 0.18337766691628035;
Draw rate at 0.0 eval at move 32 = 0.9914715947898592;

closes https://github.com/official-stockfish/Stockfish/pull/4636

No functional change
---
 src/uci.cpp | 4 ++--
 src/uci.h   | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/uci.cpp b/src/uci.cpp
index 523d551e..ed16f24c 100644
--- a/src/uci.cpp
+++ b/src/uci.cpp
@@ -207,8 +207,8 @@ namespace {
      // The coefficients of a third-order polynomial fit is based on the fishtest data
      // for two parameters that need to transform eval to the argument of a logistic
      // function.
-     constexpr double as[] = {   1.07390458,   -6.94334517,   31.95090161,  317.75424048};
-     constexpr double bs[] = {  -2.82843814,   16.64518180,  -19.74439200,   68.39499088 };
+     constexpr double as[] = {   0.38036525,   -2.82015070,   23.17882135,  307.36768407};
+     constexpr double bs[] = {  -2.29434733,   13.27689788,  -14.26828904,   63.45318330 };
 
      // Enforce that NormalizeToPawnValue corresponds to a 50% win rate at ply 64
      static_assert(UCI::NormalizeToPawnValue == int(as[0] + as[1] + as[2] + as[3]));
diff --git a/src/uci.h b/src/uci.h
index 680d2d2c..8f1be00c 100644
--- a/src/uci.h
+++ b/src/uci.h
@@ -35,7 +35,7 @@ namespace UCI {
 // the win_rate_model() such that Stockfish outputs an advantage of
 // "100 centipawns" for a position if the engine has a 50% probability to win
 // from this position in selfplay at fishtest LTC time control.
-const int NormalizeToPawnValue = 343;
+const int NormalizeToPawnValue = 328;
 
 class Option;
 

From 48f7c74f15f4cae4b77596cd468802054314d701 Mon Sep 17 00:00:00 2001
From: peregrineshahin <peregrineshahin@gmail.com>
Date: Mon, 12 Jun 2023 12:57:41 +0300
Subject: [PATCH 190/192] Fix Potential in TB cutoffs for NMP.

Removes the second dependency on beta and caps the return value to VALUE_TB_WIN_IN_MAX_PLY - 1

Earlier tests:

STC:
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 193632 W: 51372 L: 51326 D: 90934
Ptnml(0-2): 447, 20111, 55687, 20091, 480
https://tests.stockfishchess.org/tests/view/6486ee4465ffe077ca125bc1

LTC:
LLR: 2.97 (-2.94,2.94) <-1.75,0.25>
Total: 331758 W: 89538 L: 89624 D: 152596
Ptnml(0-2): 114, 30121, 105516, 29993, 135
https://tests.stockfishchess.org/tests/view/6489401af42a44347ed7be42

updated constant:
LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 100260 W: 27143 L: 27017 D: 46100
Ptnml(0-2): 34, 8842, 32248, 8976, 30
https://tests.stockfishchess.org/tests/view/6492fcafdc7002ce609c818c

closes: https://github.com/official-stockfish/Stockfish/pull/4632
fixes: https://github.com/official-stockfish/Stockfish/issues/4598

bench: 2370027
---
 src/search.cpp | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/src/search.cpp b/src/search.cpp
index 9b686a52..740ad71e 100644
--- a/src/search.cpp
+++ b/src/search.cpp
@@ -801,10 +801,9 @@ namespace {
         if (nullValue >= beta)
         {
             // Do not return unproven mate or TB scores
-            if (nullValue >= VALUE_TB_WIN_IN_MAX_PLY)
-                nullValue = beta;
+            nullValue = std::min(nullValue, VALUE_TB_WIN_IN_MAX_PLY-1);
 
-            if (thisThread->nmpMinPly || (abs(beta) < VALUE_KNOWN_WIN && depth < 14))
+            if (thisThread->nmpMinPly || depth < 14)
                 return nullValue;
 
             assert(!thisThread->nmpMinPly); // Recursive verification is not allowed

From a49b3ba7ed5d9be9151c8ceb5eed40efe3387c75 Mon Sep 17 00:00:00 2001
From: Linmiao Xu <linmiao.xu@gmail.com>
Date: Wed, 21 Jun 2023 00:43:46 -0400
Subject: [PATCH 191/192] Update default net to nn-5af11540bbfe.nnue

Created by retraining the sparsified master net (nn-cd2ff4716c34.nnue) on
a 100% minified dataset including Leela transformers data from T80 may2023.

Weights permuted with the exact methods and code in:
https://github.com/official-stockfish/Stockfish/pull/4620

LEB128 compression done with the new serialize.py param in:
https://github.com/glinscott/nnue-pytorch/pull/251

Initially trained with max epoch 800. Around epoch 780, training was paused
and max epoch raised to 960.

python3 easy_train.py \
  --experiment-name L1-1536-sparse-master-retrain \
  --training-dataset /data/leela96-dfrc99-v2-T60novdecT77decT78jantosepT79aprmayT80juntonovjan-v6dd-T80febtomay2023.min.binpack \
  --early-fen-skipping 27 \
  --start-from-engine-test-net True \
  --max_epoch 960 \
  --lr 4.375e-4 \
  --gamma 0.995 \
  --start-lambda 1.0 \
  --end-lambda 0.7 \
  --tui False \
  --seed $RANDOM \
  --gpus 0

For preparing the training dataset (interleaved size 328G):

python3 interleave_binpacks.py \
  leela96-filt-v2.min.binpack \
  dfrc99-16tb7p-eval-filt-v2.min.binpack \
  filt-v6-dd-min/test60-novdec2021-12tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test77-dec2021-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test78-jantomay2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test78-juntosep2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test79-apr2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test79-may2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-jun2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-jul2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-aug2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-sep2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-oct2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-nov2022-16tb7p-filter-v6-dd.min.binpack \
  filt-v6-dd-min/test80-jan2023-16tb7p-filter-v6-dd.min.binpack \
  test80-2023/test80-feb2023-16tb7p-no-db.min.binpack \
  test80-2023/test80-mar2023-2tb7p-no-db.min.binpack \
  test80-2023/test80-apr2023-2tb7p-no-db.min.binpack \
  test80-2023/test80-may2023-2tb7p-no-db.min.binpack \
  /data/leela96-dfrc99-v2-T60novdecT77decT78jantosepT79aprmayT80juntonovjan-v6dd-T80febtomay2023.min.binpack

Minified binpacks and Leela T80 training data from 2023 available at:
https://robotmoon.com/nnue-training-data/

Local elo at 25k nodes per move:
nn-epoch879.nnue : 3.9 +/- 5.7

Passed STC:
https://tests.stockfishchess.org/tests/view/64928c1bdc7002ce609c7690
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 72000 W: 19242 L: 18889 D: 33869
Ptnml(0-2): 182, 7787, 19716, 8126, 189

Passed LTC:
https://tests.stockfishchess.org/tests/view/64930a37dc7002ce609c82e3
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 54552 W: 14978 L: 14647 D: 24927
Ptnml(0-2): 23, 5123, 16650, 5460, 20

closes https://github.com/official-stockfish/Stockfish/pull/4635

bench 2593605
---
 src/evaluate.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/evaluate.h b/src/evaluate.h
index 33effb1c..b9d7231d 100644
--- a/src/evaluate.h
+++ b/src/evaluate.h
@@ -39,7 +39,7 @@ namespace Eval {
   // The default net name MUST follow the format nn-[SHA256 first 12 digits].nnue
   // for the build process (profile-build and fishtest) to work. Do not change the
   // name of the macro, as it is used in the Makefile.
-  #define EvalFileDefaultName   "nn-78bacfcee510.nnue"
+  #define EvalFileDefaultName   "nn-5af11540bbfe.nnue"
 
   namespace NNUE {
 

From 68e1e9b3811e16cad014b590d7443b9063b3eb52 Mon Sep 17 00:00:00 2001
From: Joost VandeVondele <Joost.VandeVondele@gmail.com>
Date: Sat, 24 Jun 2023 08:58:17 +0200
Subject: [PATCH 192/192] Stockfish 16
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Official release version of Stockfish 16

Bench: 2593605

---

Stockfish 16

A new major release of Stockfish is now available at

https://stockfishchess.org/download/

*Quality of chess play*

Stockfish continues to demonstrate its ability to discover superior moves
with remarkable speed. In self-play against Stockfish 15, this new
release gains up to 50 Elo[1] and wins up to 12 times more game pairs[2]
than it loses. In major chess engine tournaments, Stockfish reliably tops
the rankings[3] winning the TCEC season 24 Superfinal, Swiss, Fischer
Random, and Double Random Chess tournaments and the CCC 19 Bullet,
20 Blitz, and 20 Rapid competitions. Leela Chess Zero[4] was the
challenger in most finals, putting top-engine chess now firmly in the
hands of teams embracing free and open-source software.

*Progress made*

This updated version of Stockfish introduces several enhancements,
including an upgraded neural net architecture (SFNNv6)[5], improved
implementation, and refined parameterization. The ongoing utilization
of Leela’s data combined with a novel inference approach exploiting
sparsity[6], and network compression[7] ensure a speedy evaluation and
modest binary sizes while allowing for more weights and higher accuracy.
The search has undergone more optimization, leading to improved
performance, particularly in longer analyses[8]. Additionally,
the Fishtest framework has been improved and is now able to run the
tests needed to validate new ideas with 10000s of CPU cores.

*Usability improvements*

Stockfish now comes with documentation, found in the wiki folder when
downloading it or on GitHub[9]. Additionally, Stockfish now includes
a clear and consistent forced tablebase win score, displaying a value
of 200 minus the number of plies required to reach a tablebase win[10].
Furthermore, the UCI_Elo option, to reduce its strength, has been
calibrated[11]. It is worth noting that the evaluation system remains
consistent with Stockfish 15.1[12], maintaining the choice that 100cp
means a 50% chance of winning the game against an equal opponent[13].
Finally, binaries of our latest development version are now provided
continuously as pre-releases on GitHub making it easier for
enthusiasts to download the latest and strongest version of
the program[14], we thank Roman Korba for having provided a similar
service for a long time.

*Thank you*

The success of the Stockfish project relies on the vibrant community
of passionate enthusiasts (we appreciate each and every one of you!)
who generously contribute their knowledge, time, and resources.
Together, this dedicated community works towards the common goal of
developing a powerful, freely accessible, and open-source chess engine.
We invite all chess enthusiasts to join the Fishtest testing framework
and contribute to the project[15].

The Stockfish team

[1] https://tests.stockfishchess.org/tests/view/649409f0dc7002ce609c99cc
[2] https://tests.stockfishchess.org/tests/view/649409d7dc7002ce609c99c6
[3] https://en.wikipedia.org/wiki/Stockfish_(chess)#Competition_results
[4] https://lczero.org/
[5] https://github.com/official-stockfish/Stockfish/commit/c1fff71
[6] https://github.com/official-stockfish/Stockfish/commit/38e6166
[7] https://github.com/official-stockfish/Stockfish/commit/a46087e
[8] https://github.com/official-stockfish/Stockfish/commit/472e726
[9] https://github.com/official-stockfish/Stockfish/wiki/
[10] https://github.com/official-stockfish/Stockfish/commit/def2966
[11] https://github.com/official-stockfish/Stockfish/commit/a08b8d4
[12] https://github.com/official-stockfish/Stockfish/commit/52e84e4
[13] https://github.com/official-stockfish/Stockfish/wiki/Stockfish-FAQ#interpretation-of-the-stockfish-evaluation
[14] https://github.com/official-stockfish/Stockfish/releases?q=prerelease%3Atrue
[15] https://stockfishchess.org/get-involved/
---
 src/misc.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/misc.cpp b/src/misc.cpp
index f1554060..bbfa4061 100644
--- a/src/misc.cpp
+++ b/src/misc.cpp
@@ -73,7 +73,7 @@ namespace Stockfish {
 namespace {
 
 /// Version number or dev.
-constexpr string_view version = "dev";
+constexpr string_view version = "16";
 
 /// 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