xZero707/droidfish
1
0
Fork 0
mirror of https://github.com/peterosterlund2/droidfish.git synced 2024-11-23 11:31:33 +01:00
Code Issues Packages Projects Releases Wiki Activity

Update to current Stockfish development version

Browse Source 
Corresponds to commit 8b8a510fd6a1a17b39b2d4b166f60ac7be0dab23 in
Stockfish repository, from Wed Sep 16 17:39:11 2020 +0200.
This commit is contained in:
Peter Osterlund 2020-09-19 23:43:58 +02:00
parent ed5ef03dba
commit 1871f1d54a
13 changed files with 186 additions and 214 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
DroidFishApp/src/main/assets/nn-82215d0fd0df.nnue → DroidFishApp/src/main/assets/nn-03744f8d56d8.nnue
View File

Binary file not shown.

2
DroidFishApp/src/main/cpp/stockfish/benchmark.cpp
View File

@ -164,5 +164,7 @@ vector<string> setup_bench(const Position& current, istream& is) {
++posCounter;
}
list.emplace_back("setoption name Use NNUE value true");
return list;
}

11
DroidFishApp/src/main/cpp/stockfish/evaluate.cpp
View File

@ -1015,12 +1015,19 @@ make_v:
Value Eval::evaluate(const Position& pos) {
// Use classical eval if there is a large imbalance
// If there is a moderate imbalance, use classical eval with probability (1/8),
// as derived from the node counter.
bool useClassical = abs(eg_value(pos.psq_score())) * 16 > NNUEThreshold1 * (16 + pos.rule50_count());
bool classical = !Eval::useNNUE
|| abs(eg_value(pos.psq_score())) * 16 > NNUEThreshold1 * (16 + pos.rule50_count());
|| useClassical
|| (abs(eg_value(pos.psq_score())) > PawnValueMg / 4 && !(pos.this_thread()->nodes & 0xB));
Value v = classical ? Evaluation<NO_TRACE>(pos).value()
: NNUE::evaluate(pos) * 5 / 4 + Tempo;
if (classical && Eval::useNNUE && abs(v) * 16 < NNUEThreshold2 * (16 + pos.rule50_count()))
if ( useClassical
&& Eval::useNNUE
&& abs(v) * 16 < NNUEThreshold2 * (16 + pos.rule50_count()))
v = NNUE::evaluate(pos) * 5 / 4 + Tempo;
// Damp down the evaluation linearly when shuffling

2
DroidFishApp/src/main/cpp/stockfish/evaluate.h
View File

@ -38,7 +38,7 @@ 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-82215d0fd0df.nnue"
#define EvalFileDefaultName "nn-03744f8d56d8.nnue"
namespace NNUE {

2
DroidFishApp/src/main/cpp/stockfish/misc.cpp
View File

@ -65,7 +65,7 @@ namespace {
/// Version number. If Version is left empty, then compile date in the format
/// DD-MM-YY and show in engine_info.
const string Version = "12";
const string Version = "";
/// 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

38
DroidFishApp/src/main/cpp/stockfish/nnue/evaluate_nnue.cpp
View File

@ -115,31 +115,16 @@ namespace Eval::NNUE {
return stream && stream.peek() == std::ios::traits_type::eof();
}
// Proceed with the difference calculation if possible
static void UpdateAccumulatorIfPossible(const Position& pos) {
feature_transformer->UpdateAccumulatorIfPossible(pos);
}
// Calculate the evaluation value
static Value ComputeScore(const Position& pos, bool refresh) {
auto& accumulator = pos.state()->accumulator;
if (!refresh && accumulator.computed_score) {
return accumulator.score;
}
// Evaluation function. Perform differential calculation.
Value evaluate(const Position& pos) {
alignas(kCacheLineSize) TransformedFeatureType
transformed_features[FeatureTransformer::kBufferSize];
feature_transformer->Transform(pos, transformed_features, refresh);
feature_transformer->Transform(pos, transformed_features);
alignas(kCacheLineSize) char buffer[Network::kBufferSize];
const auto output = network->Propagate(transformed_features, buffer);
auto score = static_cast<Value>(output[0] / FV_SCALE);
accumulator.score = score;
accumulator.computed_score = true;
return accumulator.score;
return static_cast<Value>(output[0] / FV_SCALE);
}
// Load eval, from a file stream or a memory stream
@ -150,19 +135,4 @@ namespace Eval::NNUE {
return ReadParameters(stream);
}
// Evaluation function. Perform differential calculation.
Value evaluate(const Position& pos) {
return ComputeScore(pos, false);
}
// Evaluation function. Perform full calculation.
Value compute_eval(const Position& pos) {
return ComputeScore(pos, true);
}
// Proceed with the difference calculation if possible
void update_eval(const Position& pos) {
UpdateAccumulatorIfPossible(pos);
}
} // namespace Eval::NNUE

2
DroidFishApp/src/main/cpp/stockfish/nnue/nnue_accumulator.h
View File

@ -29,9 +29,7 @@ namespace Eval::NNUE {
struct alignas(kCacheLineSize) Accumulator {
std::int16_t
accumulation[2][kRefreshTriggers.size()][kTransformedFeatureDimensions];
Value score;
bool computed_accumulation;
bool computed_score;
};
} // namespace Eval::NNUE

269
DroidFishApp/src/main/cpp/stockfish/nnue/nnue_feature_transformer.h
View File

@ -29,6 +29,56 @@
namespace Eval::NNUE {
// If vector instructions are enabled, we update and refresh the
// accumulator tile by tile such that each tile fits in the CPU's
// vector registers.
#define TILING
#ifdef USE_AVX512
typedef __m512i vec_t;
#define vec_load(a) _mm512_loadA_si512(a)
#define vec_store(a,b) _mm512_storeA_si512(a,b)
#define vec_add_16(a,b) _mm512_add_epi16(a,b)
#define vec_sub_16(a,b) _mm512_sub_epi16(a,b)
static constexpr IndexType kNumRegs = 8; // only 8 are needed
#elif USE_AVX2
typedef __m256i vec_t;
#define vec_load(a) _mm256_loadA_si256(a)
#define vec_store(a,b) _mm256_storeA_si256(a,b)
#define vec_add_16(a,b) _mm256_add_epi16(a,b)
#define vec_sub_16(a,b) _mm256_sub_epi16(a,b)
static constexpr IndexType kNumRegs = 16;
#elif USE_SSE2
typedef __m128i vec_t;
#define vec_load(a) (*(a))
#define vec_store(a,b) *(a)=(b)
#define vec_add_16(a,b) _mm_add_epi16(a,b)
#define vec_sub_16(a,b) _mm_sub_epi16(a,b)
static constexpr IndexType kNumRegs = Is64Bit ? 16 : 8;
#elif USE_MMX
typedef __m64 vec_t;
#define vec_load(a) (*(a))
#define vec_store(a,b) *(a)=(b)
#define vec_add_16(a,b) _mm_add_pi16(a,b)
#define vec_sub_16(a,b) _mm_sub_pi16(a,b)
static constexpr IndexType kNumRegs = 8;
#elif USE_NEON
typedef int16x8_t vec_t;
#define vec_load(a) (*(a))
#define vec_store(a,b) *(a)=(b)
#define vec_add_16(a,b) vaddq_s16(a,b)
#define vec_sub_16(a,b) vsubq_s16(a,b)
static constexpr IndexType kNumRegs = 16;
#else
#undef TILING
#endif
// Input feature converter
class FeatureTransformer {
@ -36,6 +86,11 @@ namespace Eval::NNUE {
// Number of output dimensions for one side
static constexpr IndexType kHalfDimensions = kTransformedFeatureDimensions;
#ifdef TILING
static constexpr IndexType kTileHeight = kNumRegs * sizeof(vec_t) / 2;
static_assert(kHalfDimensions % kTileHeight == 0, "kTileHeight must divide kHalfDimensions");
#endif
public:
// Output type
using OutputType = TransformedFeatureType;
@ -50,11 +105,13 @@ namespace Eval::NNUE {
// Hash value embedded in the evaluation file
static constexpr std::uint32_t GetHashValue() {
return RawFeatures::kHashValue ^ kOutputDimensions;
}
// Read network parameters
bool ReadParameters(std::istream& stream) {
for (std::size_t i = 0; i < kHalfDimensions; ++i)
biases_[i] = read_little_endian<BiasType>(stream);
for (std::size_t i = 0; i < kHalfDimensions * kInputDimensions; ++i)
@ -64,23 +121,26 @@ namespace Eval::NNUE {
// Proceed with the difference calculation if possible
bool UpdateAccumulatorIfPossible(const Position& pos) const {
const auto now = pos.state();
if (now->accumulator.computed_accumulation) {
if (now->accumulator.computed_accumulation)
return true;
}
const auto prev = now->previous;
if (prev && prev->accumulator.computed_accumulation) {
UpdateAccumulator(pos);
return true;
}
return false;
}
// Convert input features
void Transform(const Position& pos, OutputType* output, bool refresh) const {
if (refresh || !UpdateAccumulatorIfPossible(pos)) {
void Transform(const Position& pos, OutputType* output) const {
if (!UpdateAccumulatorIfPossible(pos))
RefreshAccumulator(pos);
}
const auto& accumulation = pos.state()->accumulator.accumulation;
#if defined(USE_AVX2)
@ -177,74 +237,58 @@ namespace Eval::NNUE {
private:
// Calculate cumulative value without using difference calculation
void RefreshAccumulator(const Position& pos) const {
auto& accumulator = pos.state()->accumulator;
IndexType i = 0;
Features::IndexList active_indices[2];
RawFeatures::AppendActiveIndices(pos, kRefreshTriggers[i],
active_indices);
for (Color perspective : { WHITE, BLACK }) {
std::memcpy(accumulator.accumulation[perspective][i], biases_,
kHalfDimensions * sizeof(BiasType));
for (const auto index : active_indices[perspective]) {
const IndexType offset = kHalfDimensions * index;
#if defined(USE_AVX512)
auto accumulation = reinterpret_cast<__m512i*>(
&accumulator.accumulation[perspective][i][0]);
auto column = reinterpret_cast<const __m512i*>(&weights_[offset]);
constexpr IndexType kNumChunks = kHalfDimensions / kSimdWidth;
for (IndexType j = 0; j < kNumChunks; ++j)
_mm512_storeA_si512(&accumulation[j], _mm512_add_epi16(_mm512_loadA_si512(&accumulation[j]), column[j]));
#ifdef TILING
for (unsigned j = 0; j < kHalfDimensions / kTileHeight; ++j) {
auto biasesTile = reinterpret_cast<const vec_t*>(
&biases_[j * kTileHeight]);
auto accTile = reinterpret_cast<vec_t*>(
&accumulator.accumulation[perspective][i][j * kTileHeight]);
vec_t acc[kNumRegs];
#elif defined(USE_AVX2)
auto accumulation = reinterpret_cast<__m256i*>(
&accumulator.accumulation[perspective][i][0]);
auto column = reinterpret_cast<const __m256i*>(&weights_[offset]);
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
for (IndexType j = 0; j < kNumChunks; ++j)
_mm256_storeA_si256(&accumulation[j], _mm256_add_epi16(_mm256_loadA_si256(&accumulation[j]), column[j]));
for (unsigned k = 0; k < kNumRegs; ++k)
acc[k] = biasesTile[k];
#elif defined(USE_SSE2)
auto accumulation = reinterpret_cast<__m128i*>(
&accumulator.accumulation[perspective][i][0]);
auto column = reinterpret_cast<const __m128i*>(&weights_[offset]);
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
for (IndexType j = 0; j < kNumChunks; ++j)
accumulation[j] = _mm_add_epi16(accumulation[j], column[j]);
for (const auto index : active_indices[perspective]) {
const IndexType offset = kHalfDimensions * index + j * kTileHeight;
auto column = reinterpret_cast<const vec_t*>(&weights_[offset]);
#elif defined(USE_MMX)
auto accumulation = reinterpret_cast<__m64*>(
&accumulator.accumulation[perspective][i][0]);
auto column = reinterpret_cast<const __m64*>(&weights_[offset]);
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm_add_pi16(accumulation[j], column[j]);
for (unsigned k = 0; k < kNumRegs; ++k)
acc[k] = vec_add_16(acc[k], column[k]);
}
#elif defined(USE_NEON)
auto accumulation = reinterpret_cast<int16x8_t*>(
&accumulator.accumulation[perspective][i][0]);
auto column = reinterpret_cast<const int16x8_t*>(&weights_[offset]);
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
for (IndexType j = 0; j < kNumChunks; ++j)
accumulation[j] = vaddq_s16(accumulation[j], column[j]);
for (unsigned k = 0; k < kNumRegs; k++)
vec_store(&accTile[k], acc[k]);
}
#else
std::memcpy(accumulator.accumulation[perspective][i], biases_,
kHalfDimensions * sizeof(BiasType));
for (const auto index : active_indices[perspective]) {
const IndexType offset = kHalfDimensions * index;
for (IndexType j = 0; j < kHalfDimensions; ++j)
accumulator.accumulation[perspective][i][j] += weights_[offset + j];
#endif
}
#endif
}
#if defined(USE_MMX)
_mm_empty();
#endif
accumulator.computed_accumulation = true;
accumulator.computed_score = false;
}
// Calculate cumulative value using difference calculation
void UpdateAccumulator(const Position& pos) const {
const auto prev_accumulator = pos.state()->previous->accumulator;
auto& accumulator = pos.state()->accumulator;
IndexType i = 0;
@ -252,29 +296,55 @@ namespace Eval::NNUE {
bool reset[2];
RawFeatures::AppendChangedIndices(pos, kRefreshTriggers[i],
removed_indices, added_indices, reset);
for (Color perspective : { WHITE, BLACK }) {
#if defined(USE_AVX2)
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
auto accumulation = reinterpret_cast<__m256i*>(
&accumulator.accumulation[perspective][i][0]);
#ifdef TILING
for (IndexType j = 0; j < kHalfDimensions / kTileHeight; ++j) {
for (Color perspective : { WHITE, BLACK }) {
auto accTile = reinterpret_cast<vec_t*>(
&accumulator.accumulation[perspective][i][j * kTileHeight]);
vec_t acc[kNumRegs];
#elif defined(USE_SSE2)
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
auto accumulation = reinterpret_cast<__m128i*>(
&accumulator.accumulation[perspective][i][0]);
if (reset[perspective]) {
auto biasesTile = reinterpret_cast<const vec_t*>(
&biases_[j * kTileHeight]);
for (unsigned k = 0; k < kNumRegs; ++k)
acc[k] = biasesTile[k];
} else {
auto prevAccTile = reinterpret_cast<const vec_t*>(
&prev_accumulator.accumulation[perspective][i][j * kTileHeight]);
for (IndexType k = 0; k < kNumRegs; ++k)
acc[k] = vec_load(&prevAccTile[k]);
#elif defined(USE_MMX)
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
auto accumulation = reinterpret_cast<__m64*>(
&accumulator.accumulation[perspective][i][0]);
// Difference calculation for the deactivated features
for (const auto index : removed_indices[perspective]) {
const IndexType offset = kHalfDimensions * index + j * kTileHeight;
auto column = reinterpret_cast<const vec_t*>(&weights_[offset]);
#elif defined(USE_NEON)
constexpr IndexType kNumChunks = kHalfDimensions / (kSimdWidth / 2);
auto accumulation = reinterpret_cast<int16x8_t*>(
&accumulator.accumulation[perspective][i][0]);
for (IndexType k = 0; k < kNumRegs; ++k)
acc[k] = vec_sub_16(acc[k], column[k]);
}
}
{ // Difference calculation for the activated features
for (const auto index : added_indices[perspective]) {
const IndexType offset = kHalfDimensions * index + j * kTileHeight;
auto column = reinterpret_cast<const vec_t*>(&weights_[offset]);
for (IndexType k = 0; k < kNumRegs; ++k)
acc[k] = vec_add_16(acc[k], column[k]);
}
}
for (IndexType k = 0; k < kNumRegs; ++k)
vec_store(&accTile[k], acc[k]);
}
}
#if defined(USE_MMX)
_mm_empty();
#endif
#else
for (Color perspective : { WHITE, BLACK }) {
if (reset[perspective]) {
std::memcpy(accumulator.accumulation[perspective][i], biases_,
kHalfDimensions * sizeof(BiasType));
@ -286,83 +356,22 @@ namespace Eval::NNUE {
for (const auto index : removed_indices[perspective]) {
const IndexType offset = kHalfDimensions * index;
#if defined(USE_AVX2)
auto column = reinterpret_cast<const __m256i*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm256_sub_epi16(accumulation[j], column[j]);
}
#elif defined(USE_SSE2)
auto column = reinterpret_cast<const __m128i*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm_sub_epi16(accumulation[j], column[j]);
}
#elif defined(USE_MMX)
auto column = reinterpret_cast<const __m64*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm_sub_pi16(accumulation[j], column[j]);
}
#elif defined(USE_NEON)
auto column = reinterpret_cast<const int16x8_t*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = vsubq_s16(accumulation[j], column[j]);
}
#else
for (IndexType j = 0; j < kHalfDimensions; ++j) {
accumulator.accumulation[perspective][i][j] -=
weights_[offset + j];
}
#endif
for (IndexType j = 0; j < kHalfDimensions; ++j)
accumulator.accumulation[perspective][i][j] -= weights_[offset + j];
}
}
{ // Difference calculation for the activated features
for (const auto index : added_indices[perspective]) {
const IndexType offset = kHalfDimensions * index;
#if defined(USE_AVX2)
auto column = reinterpret_cast<const __m256i*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm256_add_epi16(accumulation[j], column[j]);
}
#elif defined(USE_SSE2)
auto column = reinterpret_cast<const __m128i*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm_add_epi16(accumulation[j], column[j]);
}
#elif defined(USE_MMX)
auto column = reinterpret_cast<const __m64*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = _mm_add_pi16(accumulation[j], column[j]);
}
#elif defined(USE_NEON)
auto column = reinterpret_cast<const int16x8_t*>(&weights_[offset]);
for (IndexType j = 0; j < kNumChunks; ++j) {
accumulation[j] = vaddq_s16(accumulation[j], column[j]);
}
#else
for (IndexType j = 0; j < kHalfDimensions; ++j) {
accumulator.accumulation[perspective][i][j] +=
weights_[offset + j];
}
#endif
for (IndexType j = 0; j < kHalfDimensions; ++j)
accumulator.accumulation[perspective][i][j] += weights_[offset + j];
}
}
}
#if defined(USE_MMX)
_mm_empty();
#endif
accumulator.computed_accumulation = true;
accumulator.computed_score = false;
}
using BiasType = std::int16_t;

2
DroidFishApp/src/main/cpp/stockfish/position.cpp
View File

@ -704,7 +704,6 @@ void Position::do_move(Move m, StateInfo& newSt, bool givesCheck) {
// Used by NNUE
st->accumulator.computed_accumulation = false;
st->accumulator.computed_score = false;
auto& dp = st->dirtyPiece;
dp.dirty_num = 1;
@ -1000,7 +999,6 @@ void Position::do_null_move(StateInfo& newSt) {
if (Eval::useNNUE)
{
std::memcpy(&newSt, st, sizeof(StateInfo));
st->accumulator.computed_score = false;
}
else
std::memcpy(&newSt, st, offsetof(StateInfo, accumulator));

67
DroidFishApp/src/main/cpp/stockfish/search.cpp
View File

@ -520,7 +520,7 @@ void Thread::search() {
totBestMoveChanges += th->bestMoveChanges;
th->bestMoveChanges = 0;
}
double bestMoveInstability = 1 + totBestMoveChanges / Threads.size();
double bestMoveInstability = 1 + 2 * totBestMoveChanges / Threads.size();
double totalTime = rootMoves.size() == 1 ? 0 :
Time.optimum() * fallingEval * reduction * bestMoveInstability;
@ -597,7 +597,7 @@ namespace {
Move ttMove, move, excludedMove, bestMove;
Depth extension, newDepth;
Value bestValue, value, ttValue, eval, maxValue, probCutBeta;
bool ttHit, formerPv, givesCheck, improving, didLMR, priorCapture;
bool formerPv, givesCheck, improving, didLMR, priorCapture;
bool captureOrPromotion, doFullDepthSearch, moveCountPruning,
ttCapture, singularQuietLMR;
Piece movedPiece;
@ -654,9 +654,7 @@ namespace {
// 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+4)->statScore = 0;
else
if (!rootNode)
(ss+2)->statScore = 0;
// Step 4. Transposition table lookup. We don't want the score of a partial
@ -664,12 +662,12 @@ namespace {
// position key in case of an excluded move.
excludedMove = ss->excludedMove;
posKey = excludedMove == MOVE_NONE ? pos.key() : pos.key() ^ make_key(excludedMove);
tte = TT.probe(posKey, ttHit);
ttValue = ttHit ? value_from_tt(tte->value(), ss->ply, pos.rule50_count()) : VALUE_NONE;
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]
: ttHit ? tte->move() : MOVE_NONE;
: ss->ttHit ? tte->move() : MOVE_NONE;
if (!excludedMove)
ss->ttPv = PvNode || (ttHit && tte->is_pv());
ss->ttPv = PvNode || (ss->ttHit && tte->is_pv());
formerPv = ss->ttPv && !PvNode;
if ( ss->ttPv
@ -681,11 +679,11 @@ namespace {
// thisThread->ttHitAverage can be used to approximate the running average of ttHit
thisThread->ttHitAverage = (TtHitAverageWindow - 1) * thisThread->ttHitAverage / TtHitAverageWindow
+ TtHitAverageResolution * ttHit;
+ TtHitAverageResolution * ss->ttHit;
// At non-PV nodes we check for an early TT cutoff
if ( !PvNode
&& ttHit
&& ss->ttHit
&& tte->depth() >= depth
&& ttValue != VALUE_NONE // Possible in case of TT access race
&& (ttValue >= beta ? (tte->bound() & BOUND_LOWER)
@ -778,7 +776,7 @@ namespace {
improving = false;
goto moves_loop;
}
else if (ttHit)
else if (ss->ttHit)
{
// Never assume anything about values stored in TT
ss->staticEval = eval = tte->eval();
@ -882,14 +880,14 @@ 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
&& !( ttHit
&& !( ss->ttHit
&& tte->depth() >= depth - 3
&& ttValue != VALUE_NONE
&& ttValue < probCutBeta))
{
// if ttMove is a capture and value from transposition table is good enough produce probCut
// cutoff without digging into actual probCut search
if ( ttHit
if ( ss->ttHit
&& tte->depth() >= depth - 3
&& ttValue != VALUE_NONE
&& ttValue >= probCutBeta
@ -933,7 +931,7 @@ namespace {
if (value >= probCutBeta)
{
// if transposition table doesn't have equal or more deep info write probCut data into it
if ( !(ttHit
if ( !(ss->ttHit
&& tte->depth() >= depth - 3
&& ttValue != VALUE_NONE))
tte->save(posKey, value_to_tt(value, ss->ply), ttPv,
@ -1058,7 +1056,6 @@ moves_loop: // When in check, search starts from here
if ( !givesCheck
&& lmrDepth < 6
&& !(PvNode && abs(bestValue) < 2)
&& PieceValue[MG][type_of(movedPiece)] >= PieceValue[MG][type_of(pos.piece_on(to_sq(move)))]
&& !ss->inCheck
&& ss->staticEval + 169 + 244 * lmrDepth
+ PieceValue[MG][type_of(pos.piece_on(to_sq(move)))] <= alpha)
@ -1129,11 +1126,6 @@ moves_loop: // When in check, search starts from here
&& pos.non_pawn_material() <= 2 * RookValueMg)
extension = 1;
// Castling extension
if ( type_of(move) == CASTLING
&& popcount(pos.pieces(us) & ~pos.pieces(PAWN) & (to_sq(move) & KingSide ? KingSide : QueenSide)) <= 2)
extension = 1;
// Late irreversible move extension
if ( move == ttMove
&& pos.rule50_count() > 80
@ -1168,13 +1160,6 @@ moves_loop: // When in check, search starts from here
{
Depth r = reduction(improving, depth, moveCount);
// Decrease reduction at non-check cut nodes for second move at low depths
if ( cutNode
&& depth <= 10
&& moveCount <= 2
&& !ss->inCheck)
r--;
// Decrease reduction if the ttHit running average is large
if (thisThread->ttHitAverage > 509 * TtHitAverageResolution * TtHitAverageWindow / 1024)
r--;
@ -1196,7 +1181,7 @@ moves_loop: // When in check, search starts from here
// Decrease reduction if ttMove has been singularly extended (~3 Elo)
if (singularQuietLMR)
r -= 1 + formerPv;
r--;
if (!captureOrPromotion)
{
@ -1430,7 +1415,7 @@ moves_loop: // When in check, search starts from here
Move ttMove, move, bestMove;
Depth ttDepth;
Value bestValue, value, ttValue, futilityValue, futilityBase, oldAlpha;
bool ttHit, pvHit, givesCheck, captureOrPromotion;
bool pvHit, givesCheck, captureOrPromotion;
int moveCount;
if (PvNode)
@ -1460,13 +1445,13 @@ moves_loop: // When in check, search starts from here
: DEPTH_QS_NO_CHECKS;
// Transposition table lookup
posKey = pos.key();
tte = TT.probe(posKey, ttHit);
ttValue = ttHit ? value_from_tt(tte->value(), ss->ply, pos.rule50_count()) : VALUE_NONE;
ttMove = ttHit ? tte->move() : MOVE_NONE;
pvHit = ttHit && tte->is_pv();
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();
if ( !PvNode
&& ttHit
&& ss->ttHit
&& tte->depth() >= ttDepth
&& ttValue != VALUE_NONE // Only in case of TT access race
&& (ttValue >= beta ? (tte->bound() & BOUND_LOWER)
@ -1481,7 +1466,7 @@ moves_loop: // When in check, search starts from here
}
else
{
if (ttHit)
if (ss->ttHit)
{
// Never assume anything about values stored in TT
if ((ss->staticEval = bestValue = tte->eval()) == VALUE_NONE)
@ -1500,7 +1485,7 @@ moves_loop: // When in check, search starts from here
// Stand pat. Return immediately if static value is at least beta
if (bestValue >= beta)
{
if (!ttHit)
if (!ss->ttHit)
tte->save(posKey, value_to_tt(bestValue, ss->ply), false, BOUND_LOWER,
DEPTH_NONE, MOVE_NONE, ss->staticEval);
@ -1564,7 +1549,9 @@ moves_loop: // When in check, search starts from here
}
// Do not search moves with negative SEE values
if (!ss->inCheck && !pos.see_ge(move))
if ( !ss->inCheck
&& !(givesCheck && pos.is_discovery_check_on_king(~pos.side_to_move(), move))
&& !pos.see_ge(move))
continue;
// Speculative prefetch as early as possible
@ -1716,8 +1703,8 @@ moves_loop: // When in check, search starts from here
else
captureHistory[moved_piece][to_sq(bestMove)][captured] << bonus1;
// Extra penalty for a quiet TT or main killer move in previous ply when it gets refuted
if ( ((ss-1)->moveCount == 1 || ((ss-1)->currentMove == (ss-1)->killers[0]))
// 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]))
&& !pos.captured_piece())
update_continuation_histories(ss-1, pos.piece_on(prevSq), prevSq, -bonus1);

1
DroidFishApp/src/main/cpp/stockfish/search.h
View File

@ -49,6 +49,7 @@ struct Stack {
int moveCount;
bool inCheck;
bool ttPv;
bool ttHit;
};

2
DroidFishApp/src/main/cpp/stockfish/uci.cpp
View File

@ -170,7 +170,7 @@ namespace {
if (token == "go" || token == "eval")
{
cerr << "\nPosition: " << cnt++ << '/' << num << endl;
cerr << "\nPosition: " << cnt++ << '/' << num << " (" << pos.fen() << ")" << endl;
if (token == "go")
{
go(pos, is, states);

2
DroidFishApp/src/main/java/org/petero/droidfish/engine/InternalStockFish.java
View File

@ -36,7 +36,7 @@ import org.petero.droidfish.EngineOptions;
/** Stockfish engine running as process, started from assets resource. */
public class InternalStockFish extends ExternalEngine {
private static final String defaultNet = "nn-82215d0fd0df.nnue";
private static final String defaultNet = "nn-03744f8d56d8.nnue";
private static final String netOption = "evalfile";
private File defaultNetFile; // To get the full path of the copied default network file