Recent regression. Bisection points to g:b6ff3ddecfa93d53867afaaa078f85f $ gcc-trunk -v gcc version 11.0.0 20200516 (experimental) [master revision 53b4d52f114:91c1d944166:f5b461d453043c6b6dda50db0439e4c78b241f03] (GCC) $ gcc-trunk -O1 abc.c ; ./a.out 1 $ gcc-trunk abc.c ; ./a.out 0 $ cat abc.c int a, d; int *b = &a; short c; int main() { for (; c <= 4; c--) { for (; d;) ; a = 1; *b = 0; } printf("%d\n", a); }
Confirmed.
Mine.
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>: https://gcc.gnu.org/g:52a0f83980082c9995f2d8ec9b88548520fb8a5f commit r11-450-g52a0f83980082c9995f2d8ec9b88548520fb8a5f Author: Richard Biener <rguenther@suse.de> Date: Mon May 18 09:17:24 2020 +0200 tree-optimization/95172 - avoid mixing conditionalized and ordered SM The following testcase shows a missed optimization that then leads to wrong-code when issueing SMed stores on exits. When we were able to compute an ordered sequence of stores for an exit we need to emit that in the correct order and we can emit it disregarding to any conditional for whether a store actually happened (we know it did). We can also improve detection as of whether we need conditional processing at all. Both parts fix the testcase. 2020-05-18 Richard Biener <rguenther@suse.de> PR tree-optimization/95172 * tree-ssa-loop-im.c (execute_sm): Get flag whether we eventually need the conditional processing. (execute_sm_exit): When processing an orderd sequence avoid doing any conditional processing. (hoist_memory_references): Pass down whether all edges have ordered processing for a ref to execute_sm. * gcc.dg/torture/pr95172.c: New testcase.
Fixed.