From 749dcd9ba8466fec5b51dd564cd63424c44f808b Mon Sep 17 00:00:00 2001 From: Richard Sandiford Date: Mon, 20 Jan 2025 20:05:05 +0000 Subject: vect: Preserve OMP info for conditional stores [PR118348] OMP reductions are lowered into the form: idx = .OMP_SIMD_LANE (simuid, 0); ... oldval = D.anon[idx]; newval = oldval op ...; D.anon[idx] = newval; So if the scalar loop has a {0, +, 1} iv i, idx = i % vf. Despite this wraparound, the vectoriser pretends that the D.anon accesses are linear. It records the .OMP_SIMD_LANE's second argument (val) in the data_reference aux field (-1 - val) and then copies this to the stmt_vec_info simd_lane_access_p field (val + 1). vectorizable_load and vectorizable_store use simd_lane_access_p to detect accesses of this form and suppress the vector pointer increments that would be used for genuine linear accesses. The difference in this PR is that the reduction is conditional, and so the store back to D.anon is recognised as a conditional store pattern. simd_lane_access_p was not being copied across from the original stmt_vec_info to the pattern stmt_vec_info, meaning that it was vectorised as a normal linear store. gcc/ PR tree-optimization/118348 * tree-vectorizer.cc (vec_info::move_dr): Copy STMT_VINFO_SIMD_LANE_ACCESS_P. gcc/testsuite/ PR tree-optimization/118348 * gcc.target/aarch64/pr118348_1.c: New test. * gcc.target/aarch64/pr118348_2.c: Likewise. --- gcc/tree-vectorizer.cc | 2 ++ 1 file changed, 2 insertions(+) (limited to 'gcc/tree-vectorizer.cc') diff --git a/gcc/tree-vectorizer.cc b/gcc/tree-vectorizer.cc index 01c6470..f38c8d2 100644 --- a/gcc/tree-vectorizer.cc +++ b/gcc/tree-vectorizer.cc @@ -633,6 +633,8 @@ vec_info::move_dr (stmt_vec_info new_stmt_info, stmt_vec_info old_stmt_info) = STMT_VINFO_GATHER_SCATTER_P (old_stmt_info); STMT_VINFO_STRIDED_P (new_stmt_info) = STMT_VINFO_STRIDED_P (old_stmt_info); + STMT_VINFO_SIMD_LANE_ACCESS_P (new_stmt_info) + = STMT_VINFO_SIMD_LANE_ACCESS_P (old_stmt_info); } /* Permanently remove the statement described by STMT_INFO from the -- cgit v1.1