Hi! When working on the last patch, I've noticed that the parity(bswap(x)) optimization only optimizes the 16/32/64/128-bit bswaps, but not generic _BitInt bswap, and doesn't optimize any of the bitreverses. Both all bswap and all bitreverse builtins/ifns preserve values of all the bits, just permute them, so parity (and popcount too) can be optimized.
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? 2026-07-29 Jakub Jelinek <[email protected]> * match.pd (parity(bswap(x)) is parity(x)): Use BSWAP BITREVERSE instead of BUILT_IN_BSWAP16 BUILT_IN_BSWAP32 BUILT_IN_BSWAP64 BUILT_IN_BSWAP128. * gcc.dg/bitint-140.c: New test. --- gcc/match.pd.jj 2026-07-29 16:13:46.327108421 +0200 +++ gcc/match.pd 2026-07-29 16:24:13.355136668 +0200 @@ -10786,8 +10786,7 @@ (define_operator_list SYNC_FETCH_AND_AND /* parity(bswap(x)) is parity(x). */ (for parity (PARITY) - (for bswap (BUILT_IN_BSWAP16 BUILT_IN_BSWAP32 - BUILT_IN_BSWAP64 BUILT_IN_BSWAP128) + (for bswap (BSWAP BITREVERSE) (simplify (parity (convert?@0 (bswap:s@1 @2))) (if (INTEGRAL_TYPE_P (TREE_TYPE (@0)) --- gcc/testsuite/gcc.dg/bitint-140.c.jj 2026-07-29 16:30:57.379000085 +0200 +++ gcc/testsuite/gcc.dg/bitint-140.c 2026-07-29 16:30:41.617200472 +0200 @@ -0,0 +1,17 @@ +/* { dg-do compile { target bitint575 } } */ +/* { dg-options "-O2 -fdump-tree-cplxlower1" } */ +/* { dg-final { scan-tree-dump-not "\.BSWAP \\\(" "cplxlower1" } } */ +/* { dg-final { scan-tree-dump-not "\.BITREVERSE \\\(" "cplxlower1" } } */ +/* { dg-final { scan-tree-dump-times "\.PARITY \\\(x_" 2 "cplxlower1" } } */ + +int +foo (unsigned _BitInt(512) x) +{ + return __builtin_parityg (__builtin_bswapg (x)); +} + +int +bar (unsigned _BitInt(575) x) +{ + return __builtin_parityg (__builtin_bitreverseg (x)); +} Jakub
