https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125566
Bug ID: 125566
Summary: [17 Regression] 10% slowdown of cactusADM on Zen2
since r17-302-g015813e3573be0
Product: gcc
Version: 17.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: jmelcr at gcc dot gnu.org
CC: rguenth at gcc dot gnu.org
Blocks: 26163
Target Milestone: ---
Host: x86_64-suse-linux
Target: x86_64-suse-linux
Build: x86_64-suse-linux
As seen here
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=292.100.0
there was a 10% exec time slowdown of the cactusADM SPEC 2006
benchmark when compiled with -O2 -g -flto=128 on a Zen2 machine. I bisected it
to r17-302-g015813e3573be0 (revert of r16-303-g1a13684dfc7286).
015813e3573be0d1aec92b2936d18ac972d8fd8a is the first bad commit
commit 015813e3573be0d1aec92b2936d18ac972d8fd8a
Author: Richard Biener <[email protected]>
Date: Mon May 4 13:11:27 2026 +0200
Revert "tree-optimization/120003 - missed jump threading"
This is a regression against GCC 15, on par with GCC 16. See the comparison
here:
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=320.100.0&plot.1=305.100.0&plot.2=341.100.0&plot.3=270.100.0&plot.4=393.100.0&plot.5=413.100.0&plot.6=752.100.0&plot.7=920.100.0&plot.8=1043.100.0&plot.9=1190.100.0&plot.10=1486.100.0&plot.11=292.100.0&;
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=26163
[Bug 26163] [meta-bug] missed optimization in SPEC (2026, 2k17, 2k and 2k6 and
95)
