Hi! powerpc64-linux apparently (not very surprisingly) behaves the same way as powerpc64le-linux and has 4 sunk statements rather than 5, so we should xfail it on powerpc64*-*-* rather than just powerpc64le-*-*. powerpc-linux has 3 sunk statements, but the scan pattern is done for lp64 only as the comment explains.
Tested in a cross to powerpc64-linux with make check-gcc RUNTESTFLAGS='--target_board=unix\{-m32,-m64\} tree-ssa.exp=ssa-sink-18.c' and committed to trunk as obvious. 2024-02-28 Jakub Jelinek <ja...@redhat.com> PR testsuite/111462 * gcc.dg/tree-ssa/ssa-sink-18.c: XFAIL also on powerpc64. --- gcc/testsuite/gcc.dg/tree-ssa/ssa-sink-18.c.jj 2024-01-09 09:22:57.685124089 +0100 +++ gcc/testsuite/gcc.dg/tree-ssa/ssa-sink-18.c 2024-02-28 12:05:38.040579565 +0100 @@ -213,6 +213,6 @@ compute_on_bytes (uint8_t *in_data, int expected, so this case is restricted to lp64 only so far. This different ivopts choice affects riscv64 as well, probably because it also lacks base+index addressing modes, so the ip[len] address computation can't be - made from the IV computation above. powerpc64le similarly is affected. */ + made from the IV computation above. powerpc64{,le} similarly is affected. */ - /* { dg-final { scan-tree-dump-times "Sunk statements: 5" 1 "sink2" { target lp64 xfail { riscv64-*-* powerpc64le-*-* hppa*64*-*-* } } } } */ + /* { dg-final { scan-tree-dump-times "Sunk statements: 5" 1 "sink2" { target lp64 xfail { riscv64-*-* powerpc64*-*-* hppa*64*-*-* } } } } */ Jakub