[Bug tree-optimization/36401] [4.4 Regression] 164.gzip miscompares

2008-06-02 Thread rguenth at gcc dot gnu dot org
--- Comment #3 from rguenth at gcc dot gnu dot org 2008-06-02 08:56 --- It's working again, and a check with/without the patch produces exactly the same code. So I guess the tester absorbed some HEP. -- rguenth at gcc dot gnu dot org changed: What|Removed

[Bug tree-optimization/36401] [4.4 Regression] 164.gzip miscompares

2008-06-01 Thread rguenth at gcc dot gnu dot org
--- Comment #2 from rguenth at gcc dot gnu dot org 2008-06-01 11:39 --- Reverting my patch on the tester fixes it. It's now back in and if it again miscompares tomorrow I'll investigate. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36401

[Bug tree-optimization/36401] [4.4 Regression] 164.gzip miscompares

2008-05-31 Thread rguenth at gcc dot gnu dot org
--- Comment #1 from rguenth at gcc dot gnu dot org 2008-05-31 13:56 --- Two likely changes: +2008-05-30 Richard Guenther <[EMAIL PROTECTED]> + + * builtins.c (build_string_literal): Avoid generating + a non-gimple_val result. +2008-05-30 Steven Bosscher <[EMAIL PROTECTE