https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123603
Bug ID: 123603
Summary: [16 Regression] 13% slowdown of exchange2_r on Zen4
since r16-6767-g948d33f490a6b0
Product: gcc
Version: 16.0
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: pheeck at gcc dot gnu.org, 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=1108.407.0
there was a 13% exec time slowdown of the exchange2_r SPEC 2017
benchmark when compiled with -O2 -march=x86-64-v3 -g -flto=128 with PGO on a
Zen4 machine. I bisected it to r16-6767-g948d33f490a6b0.
948d33f490a6b0051376da6bdcf55223a552b30f is the first bad commit
commit 948d33f490a6b0051376da6bdcf55223a552b30f
Author: Richard Biener <[email protected]>
Date: Wed Jan 14 12:45:19 2026 +0100
tree-optimization/123190 - fix costing of permuted contiguous loads
This is a regression against GCC 15. See the comparison
here:
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=1113.407.0&plot.1=1132.407.0&plot.2=1125.407.0&plot.3=1139.407.0&plot.4=1169.407.0&plot.5=1108.407.0&;
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=26163
[Bug 26163] [meta-bug] missed optimization in SPEC (2k17, 2k and 2k6 and 95)
