[Bug target/128408] New: [17 Regression] 47% exec time slowdown of lbm_r on Zen5 since r17-5603-gcd7f479dfa3da1

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

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128408

            Bug ID: 128408
           Summary: [17 Regression] 47% exec time slowdown of lbm_r on
                    Zen5 since r17-5603-gcd7f479dfa3da1
           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: liuhongt at gcc dot gnu.org
            Blocks: 26163
  Target Milestone: ---
              Host: x86_64-suse-linux
            Target: x86_64-suse-linux
             Build: x86_64-suse-linux

I am seeing a 43% exec time slowdown of SPEC 2017 lbm_r on Zen5 when compiling
with -O2 -march=native (=znver5) -flto=128:

https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=1253.477.0

The slowdown bisects to commit r17-5603-gcd7f479dfa3da1:

commit cd7f479dfa3da1bde9bf331c647bc70a6185b4db
Author: liuhongt <hongtao....@intel.com>
Date:   Thu Oct 22 22:32:51 2026 -0700

    i386: Add REG_EQUAL note for x86_cse vector constant copies

This is a regression against GCC 16:

https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=1308.477.0&plot.1=1365.477.0&plot.2=1567.477.0&plot.3=1253.477.0&;


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)