The lmr reduction init loop was using a misleading name, maxdepth, even when the field wasn't indexed with depth but total moves, at the same time if totalmoves ever reached a value > 127 (extremely unlikely given lmp and more but potentially possible, especially in an engine with less pruning) the table would index OOB, this removes the misleading name and caps the reduction index.
The lmr reduction init loop was using a misleading name,
maxdepth, even when the field wasn't indexed with depth but total moves, at the same time if totalmoves ever reached a value > 127 (extremely unlikely given lmp and more but potentially possible, especially in an engine with less pruning) the table would index OOB, this removes the misleading name and caps the reduction index.passed stc non reg Elo | 5.12 +- 6.19 (95%) SPRT | 8.0+0.08s Threads=1 Hash=16MB LLR | 2.95 (-2.94, 2.94) [-5.00, 0.00] Games | N: 5428 W: 1259 L: 1179 D: 2990 Penta | [11, 546, 1522, 622, 13] https://chess.swehosting.se/test/6562/