[Bug target/121993] New: [16 Regression] 20-30% slowdown of 470.lbm on AMD Zen3 and 5-8% slowdown of 519.lbm_r on Zen2 since r16-3485-gae689f89fb4059

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

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121993

            Bug ID: 121993
           Summary: [16 Regression] 20-30% slowdown of 470.lbm on AMD Zen3
                    and 5-8% slowdown of 519.lbm_r on Zen2 since
                    r16-3485-gae689f89fb4059
           Product: gcc
           Version: 16.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: pheeck at gcc dot gnu.org
                CC: lili.cui at intel dot com
            Blocks: 26163
  Target Milestone: ---
              Host: x86_64-linux
            Target: x86_64-linux

With commit r16-3485-gae689f89fb4059

commit ae689f89fb4059ba1950117a5b8e457eb46d9747
Commit:     Cui, Lili <lili....@intel.com>
CommitDate: Mon Sep 1 14:35:59 2025 +0800

    ira: Remove soft conflict related code in improve_allocation. [PR117838]

SPEC 2006 and 2017 lbm benchmarks slowed down.

470.lbm slowed down by 20-30% on AMD Zen3 machines, compile flags -Ofast
-march=native -flto
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=476.240.0

519.lbm slowed down by 5-8% on AMD Zen2 machines, compile flags -O2 -flto
(generic march)
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=298.477.0


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)