[all-commits] [llvm/llvm-project] 1f21f1: [X86] - Prevent the wrong fold of x86_avx512_mask_...

[Rohit Aggarwal via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 1f21f15e6503053b3f5cfed3689829ba16a60ce0
      https://github.com/llvm/llvm-project/commit/1f21f15e6503053b3f5cfed3689829ba16a60ce0
  Author: Rohit Aggarwal <Rohit.Aggarwal at amd.com>
  Date:   2026-06-12 (Fri, 12 Jun 2026)

  Changed paths:
    M clang/test/CodeGen/X86/avx512f-builtins-constrained-cmp.c

  Log Message:
  -----------
  [X86] - Prevent the wrong fold of x86_avx512_mask_cmp_ss/sd to fcmp (#202321)

The issue is based upon the SemiAnalysisAI by @jlebar.
[058-mask-cmp-ss-imm-immediate-not-validated](https://github.com/SemiAnalysisAI/FuzzX/blob/master/x86/bugs/058-mask-cmp-ss-imm-immediate-not-validated/NOTES.md)

It is not a real bug, just a warning for the future fold implementation
of mask_cmp → fcmp.

There is non to fix as of now in the source code. Added a few comments
and test cases for the future implementation of the folds.

@topperc @phoebewang



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications