Dev Builds » 20260525-0919

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 15. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo
ncm-dbt-01 06:42:19 582549 4000 1817 341 1842 +134.55 ± 4.43 0 19 526 1415 40 +313.91 ± 14.81
ncm-dbt-02 06:42:29 587187 4000 1775 319 1906 +132.54 ± 4.4 0 20 537 1410 33 +309.29 ± 14.66
ncm-dbt-03 06:43:10 585045 4000 1813 300 1887 +138.28 ± 4.38 0 14 504 1437 45 +325.67 ± 15.13
ncm-dbt-04 06:42:28 567626 4000 1792 303 1905 +135.86 ± 4.32 0 14 518 1433 35 +320.45 ± 14.91
ncm-dbt-05 06:44:24 583522 4000 1788 317 1895 +134.05 ± 4.39 0 17 532 1414 37 +313.19 ± 14.72
20000 8985 1580 9435 +135.05 ± 1.96 0 84 2617 7109 190 +316.43 ± 6.63

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
470574 ncm-dbt-05 580987 500 221 43 236 +129.35 ± 13.36 0 4 71 168 7 +290.66 ± 40.76
470573 ncm-dbt-04 568236 500 227 37 236 +138.99 ± 12.5 0 1 65 177 7 +324.17 ± 42.47
470572 ncm-dbt-02 588132 500 222 47 231 +126.97 ± 12.47 0 4 69 175 2 +295.94 ± 41.39
470571 ncm-dbt-01 583028 500 227 30 243 +144.71 ± 12.88 0 3 55 184 8 +342.85 ± 46.64
470570 ncm-dbt-03 582611 500 232 38 230 +142.25 ± 11.95 0 1 59 185 5 +342.85 ± 44.76
470569 ncm-dbt-05 584034 500 224 42 234 +132.54 ± 12.71 0 1 73 169 7 +301.33 ± 39.86
470568 ncm-dbt-02 586774 500 228 41 231 +136.56 ± 12.39 0 1 67 176 6 +318.25 ± 41.77
470567 ncm-dbt-04 566770 500 214 34 252 +130.94 ± 12.17 0 1 72 173 4 +304.07 ± 40.16
470566 ncm-dbt-01 581444 500 232 38 230 +142.25 ± 12.16 0 2 57 186 5 +342.85 ± 45.71
470565 ncm-dbt-03 585421 500 225 39 236 +135.76 ± 12.62 0 1 69 173 7 +312.48 ± 41.11
470564 ncm-dbt-05 583405 500 219 36 245 +133.34 ± 11.46 0 0 69 179 2 +318.25 ± 40.94
470563 ncm-dbt-01 580199 500 216 46 238 +123.02 ± 12.56 0 3 77 167 3 +280.42 ± 38.93
470562 ncm-dbt-04 567522 500 228 44 228 +134.15 ± 12.67 0 2 68 174 6 +309.64 ± 41.56
470561 ncm-dbt-02 587834 500 225 30 245 +143.07 ± 11.47 0 0 59 187 4 +349.43 ± 44.59
470560 ncm-dbt-03 587409 500 227 38 235 +138.18 ± 12.53 0 4 57 185 4 +330.23 ± 45.79
470559 ncm-dbt-05 583992 500 225 40 235 +134.95 ± 12.64 0 2 67 175 6 +312.48 ± 41.9
470558 ncm-dbt-01 577766 500 227 45 228 +132.54 ± 12.52 0 3 66 177 4 +309.64 ± 42.33
470557 ncm-dbt-04 566730 500 225 45 230 +130.94 ± 12.17 0 1 72 173 4 +304.07 ± 40.16
470556 ncm-dbt-03 586604 500 221 43 236 +129.35 ± 12.8 0 2 74 168 6 +293.29 ± 39.69
470555 ncm-dbt-02 587325 500 221 31 248 +138.99 ± 12.3 0 2 61 182 5 +330.23 ± 44.08
470554 ncm-dbt-01 587367 500 230 43 227 +136.56 ± 11.97 0 2 62 183 3 +327.18 ± 43.69
470553 ncm-dbt-04 567244 500 231 42 227 +138.18 ± 12.12 0 2 61 183 4 +330.23 ± 44.08
470552 ncm-dbt-02 586816 500 219 41 240 +129.35 ± 12.99 0 4 69 172 5 +295.94 ± 41.39
470551 ncm-dbt-05 583866 500 229 35 236 +142.25 ± 12.37 0 3 55 187 5 +342.85 ± 46.64
470550 ncm-dbt-03 587452 500 226 38 236 +137.37 ± 11.94 0 2 61 184 3 +330.23 ± 44.08
470549 ncm-dbt-04 568037 500 220 35 245 +134.95 ± 11.83 0 2 63 183 2 +324.17 ± 43.32
470548 ncm-dbt-03 585000 500 231 37 232 +142.25 ± 11.51 0 0 60 186 4 +346.11 ± 44.19
470547 ncm-dbt-02 586774 500 215 49 236 +119.89 ± 12.61 0 2 84 160 4 +268.17 ± 36.99
470546 ncm-dbt-01 584453 500 220 41 239 +130.14 ± 12.19 0 1 73 172 4 +301.33 ± 39.86
470545 ncm-dbt-05 584243 500 229 46 225 +133.34 ± 12.49 0 1 71 172 6 +306.84 ± 40.47
470544 ncm-dbt-04 569310 500 223 31 246 +140.62 ± 11.59 0 1 59 187 3 +342.85 ± 44.76
470543 ncm-dbt-01 582945 500 230 48 222 +132.54 ± 13.1 0 3 69 171 7 +301.33 ± 41.33
470542 ncm-dbt-05 587579 500 223 39 238 +134.15 ± 11.86 0 2 64 182 2 +321.19 ± 42.95
470541 ncm-dbt-02 587325 500 223 35 242 +137.37 ± 12.76 0 4 59 182 5 +324.17 ± 44.97
470540 ncm-dbt-03 580821 500 225 37 238 +137.37 ± 13.34 0 2 68 170 10 +309.64 ± 41.56
470539 ncm-dbt-04 567165 500 224 35 241 +138.18 ± 12.73 0 4 58 183 5 +327.18 ± 45.37
470538 ncm-dbt-01 583196 500 235 50 215 +134.95 ± 12.64 0 2 67 175 6 +312.48 ± 41.9
470537 ncm-dbt-03 585042 500 226 30 244 +143.89 ± 12.3 0 2 56 186 6 +346.12 ± 46.15
470536 ncm-dbt-05 580075 500 218 36 246 +132.54 ± 12.32 0 4 62 182 2 +315.35 ± 43.81
470535 ncm-dbt-02 586520 500 222 45 233 +128.55 ± 12.24 0 3 69 176 2 +301.33 ± 41.33

Commit

Commit ID 05f48577e15beb15b72363a36ecb4ca6ec40a6f9
Author anematode
Date 2026-05-25 09:19:40 UTC
Fix partial MADV_POPULATE_WRITE Potentially fixes https://github.com/official-stockfish/Stockfish/issues/6846 I'm not a Linux guru but here's my understanding of what's going on, from looking at strace logs. If `/dev/shm` is undersized or too full, `MADV_POPULATE_WRITE` can partially fault in the pages and then return an error. The crash then happens in the second process, on the first read to `header_ptr_->initialized`. There's a TOCTOU between when the region is attachable (the second process can see it in `/dev/shm`) and when we can actually safely read from it. The SIGBUS happens because `/dev/shm` is so full from partially faulted pages that we can't even fault in a single additional page. This didn't happen before because a failed `posix_fallocate` *rolls back any faulted pages*. Then the second process faults in one page (which succeeds), reads a `0` from `header_ptr_->initialized`, declares the region invalid, and continues to try (and fail) to set up its own region. There's still a problem here, methinks, if the process gets killed/interrupted between the failed `madvise` and the `ftruncate`. I think a more robust approach is to create the file anonymously first (to prevent other processes from attaching to it), then `linkat` to publish it atomically, but that would be more implementation work. (Could create a follow-up issue for that as an nice first issue for new contribs?) closes https://github.com/official-stockfish/Stockfish/pull/6852 No functional change
Copyright 2011–2026 Next Chess Move LLC