[Bug tree-optimization/125551] New: [17 Regression] 5% slowdown of parest_r with PGO on Zen2 since r17-141-g76b8869f08a6ca

[jmelcr at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125551

            Bug ID: 125551
           Summary: [17 Regression] 5% slowdown of parest_r with PGO on
                    Zen2 since r17-141-g76b8869f08a6ca
           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=286.457.0

there was a 5% exec time slowdown of the parest_r SPEC 2017
benchmark when compiled with  -Ofast -march=native -g -flto=128 with PGO on a
Zen2 machine. I bisected it to r17-141-g76b8869f08a6ca.

76b8869f08a6ca99c72596e481dd91b2d545643a is the first bad commit
commit 76b8869f08a6ca99c72596e481dd91b2d545643a
Author: Richard Biener <rguent...@suse.de>
Date:   Wed Feb 25 09:08:07 2026 +0100

    tree-optimization/124222 - rewrite BB SLP costing scalar coverage

This is a regression against previous GCC versions. See the comparison
here:

https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=326.457.0&plot.1=313.457.0&plot.2=350.457.0&plot.3=279.457.0&plot.4=402.457.0&plot.5=415.457.0&plot.6=775.457.0&plot.7=929.457.0&plot.8=1077.457.0&plot.9=1207.457.0&plot.10=1520.457.0&plot.11=286.457.0&;


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)