https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125543
Bug ID: 125543
Summary: [17 Regression] 21% slowdown of cactusADM on Zen2 with
PGO since r17-111-g27b2a84e580ac1
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: ptomsich 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=289.100.0
there was a 21% exec time slowdown of the cactusADM SPEC 2006
benchmark when compiled with -Ofast -march=native -g -flto=128 with PGO on a
Zen2 machine. I bisected it to r17-111-g27b2a84e580ac1.
27b2a84e580ac14677821daf736b9190f7aa5fc0 is the first bad commit
commit 27b2a84e580ac14677821daf736b9190f7aa5fc0
Author: Philipp Tomsich <[email protected]>
Date: Thu Mar 12 21:58:09 2026 +0100
match.pd: Relax single_use for fold-to-zero comparisons
This is a regression against GCC{15,16}. See the comparison here:
lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=322.100.0&plot.1=307.100.0&plot.2=343.100.0&plot.3=266.100.0&plot.4=395.100.0&plot.5=412.100.0&plot.6=765.100.0&plot.7=927.100.0&plot.8=1036.100.0&plot.9=1194.100.0&plot.10=1489.100.0&plot.11=289.100.0&
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)
