https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125174
Bug ID: 125174
Summary: [17 Regression] 5% slowdown of tonto on Zen3 since
r17-223-ga22b31304e0a1a
Product: gcc
Version: 17.0
Status: UNCONFIRMED
Keywords: missed-optimization
Severity: normal
Priority: P3
Component: target
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
As seen here
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=288.70.0
there was a 5% exec time slowdown of the tonto SPEC 2006
benchmark when compiled with -Ofast -march=native -g -flto=128 on a Zen3
machine. I bisected it to r17-223-ga22b31304e0a1a.
a22b31304e0a1ad21751f882c02c32c167c78793 is the first bad commit
commit a22b31304e0a1ad21751f882c02c32c167c78793
Author: Richard Biener <[email protected]>
Date: Fri Apr 24 14:35:49 2026 +0200
flip --param ix86-vect-compare-costs default
This is a regression against GCC{13,14,15}. See the comparison here:
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=542.230.0&plot.1=529.230.0&plot.2=558.230.0&plot.3=482.230.0&plot.4=511.230.0&plot.5=494.230.0&plot.6=738.230.0&plot.7=889.230.0&plot.8=1029.230.0&plot.9=1140.230.0&plot.10=1471.230.0&plot.11=476.230.0&;
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)
