Open
Conversation
Passed simplification STC: LLR: 2.96 (-2.94,2.94) <-1.75,0.25> Total: 217184 W: 56218 L: 56195 D: 104771 Ptnml(0-2): 627, 25673, 55985, 25664, 643 https://tests.stockfishchess.org/tests/view/68a21e0eb6fb3300203bbcf3 Passed simplification LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 30474 W: 7923 L: 7713 D: 14838 Ptnml(0-2): 20, 3205, 8569, 3431, 12 https://tests.stockfishchess.org/tests/view/68a2adacb6fb3300203bc3f9 closes #6246 bench 2626263
Passed simplification STC LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 99840 W: 25965 L: 25815 D: 48060 Ptnml(0-2): 326, 11708, 25688, 11886, 312 https://tests.stockfishchess.org/tests/view/68a24504b6fb3300203bbdae Passed simplification LTC LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 248388 W: 63544 L: 63556 D: 121288 Ptnml(0-2): 125, 27132, 69709, 27086, 142 https://tests.stockfishchess.org/tests/view/68a24f3bb6fb3300203bbdee closes #6250 Bench: 2432765
Passed Non-regression STC: LLR: 2.96 (-2.94,2.94) <-1.75,0.25> Total: 107392 W: 27959 L: 27818 D: 51615 Ptnml(0-2): 317, 12094, 28735, 12231, 319 https://tests.stockfishchess.org/tests/view/68a65d8ab6fb3300203bc825 Passed Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 94014 W: 24129 L: 23986 D: 45899 Ptnml(0-2): 47, 9917, 26934, 10064, 45 https://tests.stockfishchess.org/tests/view/68a8f45cb6fb3300203bcb5e Passed Stalemate 10k: Elo: 2.22 ± 1.3 (95%) LOS: 100.0% Total: 10000 W: 4626 L: 4562 D: 812 Ptnml(0-2): 1, 137, 4659, 203, 0 nElo: 12.00 ± 6.8 (95%) PairsRatio: 1.47 https://tests.stockfishchess.org/tests/view/68a65d8ab6fb3300203bc825 Supersedes #6114 Closes #6232 closes #6255 Bench: 2473929
Passed Non-regression STC: LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 46272 W: 12076 L: 11866 D: 22330 Ptnml(0-2): 140, 5407, 11860, 5561, 168 https://tests.stockfishchess.org/tests/view/68a9c26575da51a345a5a6a2 Passed Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 140442 W: 35984 L: 35886 D: 68572 Ptnml(0-2): 72, 15333, 39305, 15447, 64 https://tests.stockfishchess.org/tests/view/68aa245e75da51a345a5a80d closes #6258 Bench: 2931171
Passed Non-regression STC: LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 16352 W: 4336 L: 4090 D: 7926 Ptnml(0-2): 54, 1832, 4157, 2080, 53 https://tests.stockfishchess.org/tests/view/68a808f0b6fb3300203bca08 Passed Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 94014 W: 24129 L: 23986 D: 45899 Ptnml(0-2): 47, 9917, 26934, 10064, 45 https://tests.stockfishchess.org/tests/view/68a8f45cb6fb3300203bcb5e closes #6256 Bench: 2483704
Remove no longer required condition for tt moves. The idea is that if tt the value is greater than probCut beta try tt move that is a capture anyway - even if it doesn't pass SEE check. This idea in various implementations passed some STCs and was looking decent at LTCs as a gainer. Passed STC: https://tests.stockfishchess.org/tests/view/68aa1ee075da51a345a5a805 LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 136160 W: 35189 L: 35079 D: 65892 Ptnml(0-2): 436, 16083, 34891, 16275, 395 Passed LTC: https://tests.stockfishchess.org/tests/view/68aa91d375da51a345a5a884 LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 80022 W: 20652 L: 20492 D: 38878 Ptnml(0-2): 33, 8717, 22357, 8865, 39 closes #6259 Bench: 2307940
Passed STC: LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 127264 W: 33159 L: 33042 D: 61063 Ptnml(0-2): 336, 14509, 33866, 14544, 377 https://tests.stockfishchess.org/tests/view/68a63fccb6fb3300203bc818 Passed LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 469512 W: 120094 L: 120331 D: 229087 Ptnml(0-2): 238, 51317, 131885, 51076, 240 https://tests.stockfishchess.org/tests/view/68a8d867b6fb3300203bcab5 closes #6260 Bench: 2433974
Give "(thread id mod 8) * 64" less reductions (up to nearly a half ply). Passed SMP STC: LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 413176 W: 106496 L: 105666 D: 201014 Ptnml(0-2): 504, 46732, 111266, 47602, 484 https://tests.stockfishchess.org/tests/view/68a24aeeb6fb3300203bbdc4 Passed SMP LTC: LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 60420 W: 15622 L: 15268 D: 29530 Ptnml(0-2): 11, 6106, 17632, 6440, 21 https://tests.stockfishchess.org/tests/view/68a2c2ffb6fb3300203bc516 closes #6249 No functional change
Passed Non-regression STC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 236896 W: 61683 L: 61683 D: 113530 Ptnml(0-2): 730, 28057, 60901, 28003, 757 https://tests.stockfishchess.org/tests/view/68a9c20475da51a345a5a6a0 Passed Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 73644 W: 18936 L: 18770 D: 35938 Ptnml(0-2): 33, 7951, 20685, 8123, 30 https://tests.stockfishchess.org/tests/view/68aa95c575da51a345a5a88a closes #6265 Bench: 2693376
closes #6276 No functional change
If the depth is low enough, don't TT cut if the rule50 count is high and the TT move is zeroing. Passed STC: https://tests.stockfishchess.org/tests/view/68a8fdd8b6fb3300203bcb92 LLR: 3.05 (-2.94,2.94) <0.00,2.00> Total: 110304 W: 28805 L: 28402 D: 53097 Ptnml(0-2): 275, 11174, 31875, 11529, 299 Passed LTC: https://tests.stockfishchess.org/tests/view/68aa200f75da51a345a5a809 LLR: 3.00 (-2.94,2.94) <0.50,2.50> Total: 187956 W: 48489 L: 47928 D: 91539 Ptnml(0-2): 59, 16118, 61075, 16655, 71 closes #6271 bench: 2641840
The recent commit af181d9 was merged as a simplification, but unfortunately hurts mate finding efficiency. https://github.com/vondele/matetrack/blob/69f5c5e8627c163a6a8480b869ee09bc44dc44d4/matetrack1000000.csv#L4075-L4076 So this PR proposes to revert it, while adding a comment in the code for future reference. closes #6269 Bench: 2566711
Passed Non-regression STC: LLR: 2.97 (-2.94,2.94) <-1.75,0.25> Total: 187904 W: 48639 L: 48583 D: 90682 Ptnml(0-2): 560, 22150, 48502, 22154, 586 https://tests.stockfishchess.org/tests/view/68aad56075da51a345a5a9e3 Passed Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 94302 W: 24246 L: 24101 D: 45955 Ptnml(0-2): 44, 10274, 26371, 10417, 45 https://tests.stockfishchess.org/tests/view/68ab541975da51a345a5aacf closes #6266 Bench: 2376717
closes #6267 No functional change Co-authored-by: Daniel Monroe <github.com@danielmonroe.net>
Passed Non-regression STC: LLR: 2.98 (-2.94,2.94) <-1.75,0.25> Total: 191520 W: 49633 L: 49582 D: 92305 Ptnml(0-2): 557, 20676, 53260, 20693, 574 https://tests.stockfishchess.org/tests/view/68ab570075da51a345a5abe1 Passed Non-regression LTC: LLR: 2.96 (-2.94,2.94) <-1.75,0.25> Total: 212934 W: 54643 L: 54618 D: 103673 Ptnml(0-2): 117, 22417, 61364, 22462, 107 https://tests.stockfishchess.org/tests/view/68ab84e975da51a345a5ac6b 10k Stalemate: Elo: 0.35 ± 1.2 (95%) LOS: 71.6% Total: 10000 W: 4602 L: 4592 D: 806 Ptnml(0-2): 0, 148, 4694, 158, 0 nElo: 1.99 ± 6.8 (95%) PairsRatio: 1.07 https://tests.stockfishchess.org/tests/view/68abeb8175da51a345a5af82 closes #6268 Bench: 2420973
The function does not fulfill its purpose and is not used anywhere. See https://discord.com/channels/435943710472011776/1101022188313772083/1409801409855094874 closes #6275 no functional change
closes #6277 No functional change
Remove the AfterMove template from the adjust_key50 function, which is only ever called with false. closes #6278 No functional change
fixes #6280 prevents probcut from extending depth as analyzed here: #6254 (comment) passed STC: LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 86688 W: 22591 L: 22426 D: 41671 Ptnml(0-2): 305, 10125, 22311, 10306, 297 https://tests.stockfishchess.org/tests/view/68b418ab467ff96994ae4cd5 passed LTC: LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 82914 W: 21287 L: 21130 D: 40497 Ptnml(0-2): 39, 8959, 23305, 9114, 40 https://tests.stockfishchess.org/tests/view/68b47ffa78ed7a752a9e8f36 closes #6286 Bench: 2787731 Co-Authored-By: xu-shawn <50402888+xu-shawn@users.noreply.github.com>
Simplify sign term in quiet histories Passed simplification STC LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 497824 W: 129130 L: 129418 D: 239276 Ptnml(0-2): 1546, 59040, 128008, 58792, 1526 https://tests.stockfishchess.org/tests/view/68a2a9c1b6fb3300203bc3ed Passed simplification LTC LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 70830 W: 18185 L: 18016 D: 34629 Ptnml(0-2): 36, 7658, 19861, 7821, 39 https://tests.stockfishchess.org/tests/view/68af36c96217b8721dca98d9 closes #6282 Bench: 2393762
Non-regression STC: LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 155200 W: 40650 L: 40562 D: 73988 Ptnml(0-2): 533, 17588, 41258, 17700, 521 https://tests.stockfishchess.org/tests/view/68abe11375da51a345a5adec Non-regression LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 82824 W: 21442 L: 21287 D: 40095 Ptnml(0-2): 51, 8738, 23675, 8901, 47 https://tests.stockfishchess.org/tests/view/68b205606217b8721dca9c8e 10k Stalemate: Elo: 1.46 ± 1.2 (95%) LOS: 99.0% Total: 10000 W: 4640 L: 4598 D: 762 Ptnml(0-2): 0, 140, 4678, 182, 0 https://tests.stockfishchess.org/tests/view/68b2059b6217b8721dca9c90 closes #6283 Bench: 2431727
Passed Non-regression SMP STC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 232784 W: 59845 L: 59841 D: 113098 Ptnml(0-2): 289, 26459, 62934, 26379, 331 https://tests.stockfishchess.org/tests/view/68ab96bf75da51a345a5acd6 Passed Non-regression SMP LTC: LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 374270 W: 95978 L: 96113 D: 182179 Ptnml(0-2): 118, 38575, 109888, 38432, 122 https://tests.stockfishchess.org/tests/view/68abcf1c75da51a345a5adb6 closes #6285 Bench: 2667107
It has been repeatedly shown that such tunes are suspectible to become anti-scaling. Below are some recent examples: 2e91a86 d11f49b Passed STC: LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 32448 W: 8651 L: 8342 D: 15455 Ptnml(0-2): 81, 3695, 8408, 3914, 126 https://tests.stockfishchess.org/tests/view/6899489b0049e8ccef9d64ad Passed LTC: LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 73854 W: 19042 L: 18649 D: 36163 Ptnml(0-2): 37, 7908, 20659, 8271, 52 https://tests.stockfishchess.org/tests/view/689abbe7fd8719b088c8d514 Revert VVLTC with STC bound: LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 53802 W: 14030 L: 13740 D: 26032 Ptnml(0-2): 5, 4924, 16754, 5212, 6 https://tests.stockfishchess.org/tests/view/68a9a9f575da51a345a5a675 Revert VVLTC with LTC bound: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 53658 W: 14022 L: 13699 D: 25937 Ptnml(0-2): 3, 4894, 16712, 5217, 3 https://tests.stockfishchess.org/tests/view/68a8d2b2b6fb3300203bca77 https://tests.stockfishchess.org/tests/view/688cf38bf17748b4d23c8057 https://tests.stockfishchess.org/tests/view/6890bc7792fcad741b804a19 Passed LTC: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 74928 W: 19466 L: 19071 D: 36391 Ptnml(0-2): 37, 8048, 20901, 8439, 39 Failed Non-regression VLTC: LLR: -2.94 (-2.94,2.94) <-1.75,0.25> Total: 57704 W: 14643 L: 14928 D: 28133 Ptnml(0-2): 5, 5925, 17280, 5634, 8 https://tests.stockfishchess.org/tests/view/6890bc7792fcad741b804a19 (Note that an STC-tuned version passed non-regression, but was shortly simplified) #6040 Passed STC: LLR: 2.93 (-2.94,2.94) <0.00,2.00> Total: 27776 W: 7352 L: 7054 D: 13370 Ptnml(0-2): 68, 3126, 7221, 3386, 87 https://tests.stockfishchess.org/tests/view/680ec0f83629b02d74b1605b Passed LTC: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 161304 W: 41432 L: 40864 D: 79008 Ptnml(0-2): 61, 17305, 45357, 17863, 66 https://tests.stockfishchess.org/tests/view/680ec7f93629b02d74b16084 Failed Non-regression VVLTC: LLR: -2.94 (-2.94,2.94) <-1.75,0.25> Total: 313466 W: 80573 L: 81089 D: 151804 Ptnml(0-2): 38, 29689, 97782, 29199, 25 https://tests.stockfishchess.org/tests/view/6810d0533629b02d74b16756 #5907 #5887 Passed LTC with STC bounds: https://tests.stockfishchess.org/tests/view/67b115dd6c6b9e172ad1592f LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 75756 W: 19393 L: 19044 D: 37319 Ptnml(0-2): 60, 8251, 20913, 8588, 66 Passed LTC with LTC bounds: https://tests.stockfishchess.org/tests/view/67af5f5d6c6b9e172ad15765 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 108126 W: 27880 L: 27412 D: 52834 Ptnml(0-2): 85, 11786, 29866, 12228, 98 Revert VVLTC w/ STC bounds: LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 56342 W: 14536 L: 14246 D: 27560 Ptnml(0-2): 7, 5061, 17741, 5359, 3 https://tests.stockfishchess.org/tests/view/67be4f8ad8d5c2c657c52d10 Revert VVLTC w/ LTC bounds: LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 66562 W: 17364 L: 17016 D: 32182 Ptnml(0-2): 3, 6145, 20637, 6493, 3 https://tests.stockfishchess.org/tests/view/67bcd25ff6b602bd7222ea40 closes #6284 no functional change
Reintroduces af181d9, which no longer regresses on matetrack after #6286 Using ./stockfish on matetrack.epd with --nodes 1000000 Engine ID: Stockfish dev-20250902-adfddd2c Total FENs: 6554 Found mates: 3490 Best mates: 2429 Passed Non-regression LTC: LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 37608 W: 9726 L: 9523 D: 18359 Ptnml(0-2): 16, 4001, 10578, 4182, 27 https://tests.stockfishchess.org/tests/view/68b886778f94a4e5a7fe77d9 closes #6289 Bench: 2493363
For the new bound in the opposite direction of the fail, use a weighted average of alpha, beta and best value +- delta. In the case of a fail high, different average weights are used depending on whether or not there was a best move change during the last search. The weights are determined from the following two consecutive LTC tunings. First tuning: https://tests.stockfishchess.org/tests/view/68ab727975da51a345a5ac2e Second tuning: https://tests.stockfishchess.org/tests/view/68aba3fe75da51a345a5ad52 Passed LTC: LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 39504 W: 10243 L: 9947 D: 19314 Ptnml(0-2): 25, 4182, 11041, 4480, 24 https://tests.stockfishchess.org/tests/view/68acbb6d6217b8721dca95f8 Passed VLTC: LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 91196 W: 23574 L: 23167 D: 44455 Ptnml(0-2): 13, 8943, 27276, 9356, 10 https://tests.stockfishchess.org/tests/view/68af64786217b8721dca993d closes #6292 Bench: 2785713
Previously the upload step used the os of the artifact to create the upload and used an extra msys2 step. This is in fact not needed and we can do all required changes on ubuntu instead. closes #6294 No functional change
The macOS 13 runner image will be retired by December 4th, 2025. closes #6330 No functional change
closes #6685 No functional change
Passed simp STC LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 29632 W: 7772 L: 7552 D: 14308 Ptnml(0-2): 76, 3155, 8147, 3349, Id expect there to be a small speedup (bc the compiler can skip some stores) but i haven’t tried to measure it closes #6688 No functional change
Passed non-regression STC at tests.stockfishchess.org/tests/view/69c9b8c3a54395a761335044 (set to gainer bounds, *mea culpa*) There's a major performance penalty when compiling the fish with GCC 16. @Torom 's measurements (similar to mine, but more precise): ``` GCC 15 (master) vs. 16 (master) sf_base = 2549174 +/- 2984 (95%) sf_test = 2435456 +/- 2239 (95%) diff = -113718 +/- 3599 (95%) speedup = -4.46100% +/- 0.141% (95%) GCC 15 (master) vs. 16 (patch) sf_base = 2544107 +/- 3091 (95%) sf_test = 2568837 +/- 3164 (95%) diff = 24729 +/- 4359 (95%) speedup = 0.97204% +/- 0.171% (95%) ``` Based on looking at the disassembly + profiling data of the builds, the root cause appears to be overly aggressive code duplication from constant propagation of various functions especially in `search.cpp`. The worst offender is `Worker::search`, which I think is being cloned and optimized based on the value of `cutNode`. `perf stat` indicates that the instruction cache miss rate goes up by ~50% on the GCC 16 build. closes #6691 No functional change
Reduce # instructions using widening add/sub calls on AArch64. Add shims for 32-bit ARM. Passed non-regression STC: https://tests.stockfishchess.org/tests/view/69c093d3f690a4b7f5fb0d6e LLR: 5.15 (-2.94,2.94) <-1.75,0.25> Total: 338624 W: 86931 L: 86853 D: 164840 Ptnml(0-2): 667, 35888, 96079, 36056, 622 closes #6692 No functional change
Makes eval calls more cache-friendly on a hot path. Hints of a speedup found locally: ``` Result of 30 runs ================== base (...ockfish-base) = 1754124 +/- 2887 test (...h-eval-order) = 1757848 +/- 2412 diff = +3724 +/- 1368 speedup = +0.0021 P(speedup > 0) = 1.0000 CPU: 16 x arm Hyperthreading: off ``` Passed STC: https://tests.stockfishchess.org/tests/view/69c84f3cf40d95b975597c21 LLR: 3.36 (-2.94,2.94) <0.00,2.00> Total: 335840 W: 87022 L: 86257 D: 162561 Ptnml(0-2): 932, 36760, 91816, 37435, 977 closes #6693 No functional change
[STC](https://tests.stockfishchess.org/tests/live_elo/69cb991bc025c305b7daa219) LLR: 2.97 (-2.94,2.94) <0.00,2.00> Total: 47520 W: 12558 L: 12219 D: 22743 Ptnml(0-2): 190, 5474, 12102, 5795, 199 [LTC](https://tests.stockfishchess.org/tests/live_elo/69cc6b42731dfb72d2e92128) LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 123756 W: 31996 L: 31497 D: 60263 Ptnml(0-2): 81, 13512, 34221, 13955, 109 Adds features for pawns pushing against other pawns. For indexing, we use the existing threats feature set by adding the forward square to the squares threatened by a pawn. This threat is only enabled when another pawn occupies that square. We then add appropriate logic changes in `update_piece_threats` and `features/full_threats.cpp`. The wonderful Ciekce (of Stormphrax) and peregrine (of Reckless) also had similar ideas with adding forward pawn movements to TI, and I appreciate their giving me motivation to figure out `nnue-pytorch` and test it. closes #6694 Bench: 2926703
closes #6525 No functional change
Passed Stc: LLR: 3.05 (-2.94,2.94) <-1.75,0.25> Total: 289088 W: 74727 L: 74773 D: 139588 Ptnml(0-2): 903, 30974, 80841, 30918, 908 https://tests.stockfishchess.org/tests/view/699d9093eaae015cd278eda3 Passed Ltc: LLR: 3.02 (-2.94,2.94) <-1.75,0.25> Total: 144438 W: 36859 L: 36766 D: 70813 Ptnml(0-2): 62, 13391, 45235, 13454, 77 https://tests.stockfishchess.org/tests/view/699f02913ece4c464328a264 Passed Sudden death: LLR: 2.99 (-2.94,2.94) <-1.75,0.25> Total: 53406 W: 13960 L: 13753 D: 25693 Ptnml(0-2): 265, 5848, 14253, 6089, 248 https://tests.stockfishchess.org/tests/view/69a47548a70d5f3736be6482 closes #6644 No functional change
https://tests.stockfishchess.org/tests/view/69bb1fafd7d60419badf31a3 LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 116544 W: 30024 L: 29895 D: 56625 Ptnml(0-2): 317, 12875, 31729, 13064, 287 closes #6677 No functional change
Passed simplification STC https://tests.stockfishchess.org/tests/live_elo/69c9c050a54395a76133504c LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 304544 W: 78502 L: 78574 D: 147468 Ptnml(0-2): 906, 33590, 83358, 33506, 912 closes #6697 No functional change
``` clang++: error: unknown argument: '-fno-ipa-cp-clone' clang++: error: unknown argument: '-fno-ipa-cp-clone' make[1]: *** [bitboard.o] Error 1 ``` The flag is unsupported by Clang, which pretends to be GCC on macOS, so only passing -fno-ipa-cp-clone to real GCC fixes macOS builds. closes #6698 No functional change
Avoids compilation errors on macOS by ensuring flags specific to g++ are applied to actual g++ closes #6709 No functional change
Using the already existing pawn_single_push_bb helper which encapsulates the exact same logic. closes #6700 No functional change
[Passed STC](https://tests.stockfishchess.org/tests/live_elo/69cf6a5d1668971c9da23ae1) LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 51104 W: 13376 L: 13180 D: 24548 Ptnml(0-2): 134, 5555, 13963, 5781, 119 [Passed LTC](https://tests.stockfishchess.org/tests/live_elo/69d00a7ee2b443cb2670b5c6) LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 101256 W: 25867 L: 25732 D: 49657 Ptnml(0-2): 59, 10423, 29520, 10576, 50 When processing a promotion we add/remove nonsensical threats, namely, piece threats of pawns on the 1st or 8th rank. This is not only inefficient but also seems to lead to an obscure bug during a promotion–capture + `double_inc_update` that causes a lingering invalid feature. Currently – as an artifact of the training process – these parts of the net are filled with random values in [-1,1], which is why there is a bench change. @ces42 did a helpful analysis: > checking bench with depth 17 actually shows that the bug in has been there all the time since 8e5392d (when threat inputs was merged into master) > i.e. double_inc_update for ThreatFeatureSet has always been slightly bugged After this patch is applied we'll be able to better compress the net by filling the invalid features with 0, w/o changing bench. I expect a 15%-ish size savings. closes #6701 Bench: 2700393
Instead of checking whether a threat index is valid and then writing to a vector if it isn't, we can instead always write to the buffer of the vector, and only increase the size of the vector if the index is valid. This saves some branch mispredictions. passed STC: https://tests.stockfishchess.org/tests/view/69ceb1689f7a7e3fdfc9a44b LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 205760 W: 53109 L: 52561 D: 100090 Ptnml(0-2): 603, 22552, 56076, 22992, 657 local speedtest shows ``` Result of 100 runs (cycles) =========================== base (...ockfish.orig) = 296886 +/- 1645 test (...kfish.df0822) = 298862 +/- 1662 speedup % = +0.67 +/- 0.19 [95% CI; t-statistic] p(speedup > 0) = 1.0000 p(speedup > .5%) = 0.9562 CPU: 16 x Intel(R) Core(TM) Ultra 9 185H Hyperthreading: on ``` closes #6703 No functional change
The uci pv suppression and PV roll-back logic in master is a bit convoluted, which makes it hard to reason about the code. In fact, subtle bugs that led to wrong mated-in scores in game play or a mismatch between bestmove and first PV move were only recently fixed. Moreover, in master the uci pv output through `pv()` is called in four different places. This PR proposes to simplify this logic. In this patch, the PV is sent to the GUI from within `iterative_deepening()` only (a) for fail highs/lows or (b) when an iteration for a root move is completed. All other PV outputs are now handled by `start_searching()`, just before the bestmove extraction. This easily ensures that bestmove (and ponder move) will always be in sync with the final PV output. The only noticeable change to master is for multi-threaded searches that do not involve `limits.depth`. Here master would show both the PV from the aborted search in main thread, as well as the new PV from the selected best thread. This patch would only show the latter. While at it, we also remove the requirement to finish at least a depth 1 search and simplify the logic around stalemates/checkmates at root and in particular avoid any PVs that start with `Move::none()`. Passed STC non-reg: LLR: 2.96 (-2.94,2.94) <-1.75,0.25> Total: 72416 W: 18784 L: 18609 D: 35023 Ptnml(0-2): 200, 7762, 20127, 7901, 218 https://tests.stockfishchess.org/tests/view/69d3ef0c33584dad27b3c85a closes #6704 No functional change
Passed STC non-reg: LLR: 2.93 (-2.94,2.94) <-1.75,0.25> Total: 102272 W: 26553 L: 26411 D: 49308 Ptnml(0-2): 269, 11212, 28036, 11346, 273 https://tests.stockfishchess.org/tests/view/69d0f1c5e2b443cb2670b710 closes #6707 No functional change
For some time now we can trust the threads' exact win/loss scores. This allows for some simplification in the best thread selection code. closes #6708 No functional change
Passed non-regression STC: https://tests.stockfishchess.org/tests/view/69d2d3de61a12cebe17ede38 LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 68384 W: 17646 L: 17459 D: 33279 Ptnml(0-2): 200, 8058, 17496, 8231, 207 Passed non-regression LTC: https://tests.stockfishchess.org/tests/view/69d4293433584dad27b3c8cb LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 37290 W: 9582 L: 9380 D: 18328 Ptnml(0-2): 18, 3915, 10579, 4113, 20 closes #6711 Bench: 2923401
the assert checks for `bestThreadDecisive && newThreadDecisive`. However `newThreadDecisive` is not guaranteed at this point, leading to assertion failures. This PR rearranges the conditions to ensure correctness. closes #6716 No functional change
A bit tricky to identify, but in `apple-silicon` builds, specifying a value for `SyzygyPath` could still lead to crashes at certain depths from a SIGBUS/SIGSEGV. This PR replaces the `std::count_if()` adjustment in `tbprobe.cpp`'s `do_probe_table()` with an annotated handwritten loop, avoiding the clang/LTO miscompilation behind the crash. closes #6721 No functional change.
STC: https://tests.stockfishchess.org/tests/view/69d586a64088e069540a22aa LLR: 2.93 (-2.94,2.94) <0.00,2.00> Total: 63520 W: 16579 L: 16233 D: 30708 Ptnml(0-2): 138, 7179, 16825, 7435, 183 LTC: https://tests.stockfishchess.org/tests/view/69d6ddd79ffee997bab13a98 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 227100 W: 58279 L: 57591 D: 111230 Ptnml(0-2): 95, 23435, 65828, 24071, 121 Transition highBestMoveEffort smoothly instead of abruptly. May also be more amenable to future tuning. closes #6729 Bench: 2923401
Trained with the following yaml definition https://github.com/Disservin/nettest/blob/8da0a21503adb341bc89ca20b4e2f166c7c9720d/threats.yaml and the following trainer change official-stockfish/nnue-pytorch@d87c248 Initially I saw a lot of weird 0 scores in the binpack, since then @linrock has explained to me/us their meaning and mentioned that they are actually mostly skipped.. except a very small percentage, however it seems this change was still beneficial. Passed STC: https://tests.stockfishchess.org/tests/view/69db95f43ca80bdf151d4913 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 106272 W: 27916 L: 27489 D: 50867 Ptnml(0-2): 420, 12533, 26841, 12884, 458 Passed LTC: https://tests.stockfishchess.org/tests/view/69ddd9003ca80bdf151d4c4d LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 29748 W: 7775 L: 7470 D: 14503 Ptnml(0-2): 24, 3136, 8247, 3445, 22 closes #6730 Bench: 3108996
STC: https://tests.stockfishchess.org/tests/view/69cf29ed1668971c9da23a5a LLR: 3.47 (-2.94,2.94) <-1.75,0.25> Total: 193760 W: 50142 L: 50051 D: 93567 Ptnml(0-2): 534, 21375, 52980, 21448, 543 * Simplify retval and arithmetic semantics * Remove mixing of int and uint8_t I find this overall much easier to reason about (Also rename a DEPTH constant per mstembera's suggestion #5766 (comment)) closes #6702 No functional change
Conditions for no pawn pushes, non-pawn material and last captured piece are already such restrictive that we don't need to artificially set checkers bitboard for legal move generation. Also moving the check for available pawn pushes at first place, as it should already filter about 90% of cases. Passed STC and LTC non-regression tests. LLR: 2.96 (-2.94,2.94) <-1.75,0.25> Total: 171232 W: 44202 L: 44131 D: 82899 Ptnml(0-2): 466, 18809, 47023, 18824, 494 https://tests.stockfishchess.org/tests/view/69d3623b61a12cebe17ededa LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 55662 W: 14424 L: 14247 D: 26991 Ptnml(0-2): 38, 5479, 16609, 5678, 27 https://tests.stockfishchess.org/tests/view/69d64cada07a818b0a25f95e And also neutral on 10k fixed games test with custom stalemate book. Elo: 0.24 ± 0.9 (95%) LOS: 69.9% Total: 10000 W: 4681 L: 4674 D: 645 Ptnml(0-2): 0, 87, 4819, 94, 0 nElo: 1.81 ± 6.8 (95%) PairsRatio: 1.08 https://tests.stockfishchess.org/tests/view/69d7e4183ca80bdf151d43b8 closes #6715 Bench: 2984258
In C++, all standard input streams (including std::istringstream) are initialized with the skipws format flag enabled by default. Because these stream objects are scoped locally and freshly constructed on each loop iteration, we are guaranteed that skipws is already active. closes #6719 No functional change
[Some build environments](https://tests.stockfishchess.org/actions?max_count=400&sort=time&order=desc&action=failed_task&user=gaster319&text=executing) do not have this intrinsic. closes #6722 No functional change.
Use a dedicated struct for keeping track of the pv moves. anematode pointed out that we can use the ValueList for this as well, left that open for a future patch. Passed Non Regression https://tests.stockfishchess.org/tests/view/69d399cfec2dc6794cf18376 LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 267264 W: 68895 L: 68927 D: 129442 Ptnml(0-2): 790, 29238, 73566, 29290, 748 closes #6726 No functional change
Since as of 9417552 the en passant square is only set if there is a legal en passant capture, the special handling is redundant Passed Non-Regression STC: https://tests.stockfishchess.org/tests/view/696ffb6512ee1f6231b96fe8 LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 79072 W: 20380 L: 20210 D: 38482. closes #6560 No functional change
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.