[all-commits] [llvm/llvm-project] 6fec10: [AMDGPU] Enable typechecks for __builtin_amdgcn_ra...

[Juan Manuel Martinez Caamaño via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 6fec104b45734034f3772747e5adb47b1f7ee658
      https://github.com/llvm/llvm-project/commit/6fec104b45734034f3772747e5adb47b1f7ee658
  Author: Juan Manuel Martinez Caamaño <jmartinezcaamao at gmail.com>
  Date:   2025-11-05 (Wed, 05 Nov 2025)

  Changed paths:
    M clang/include/clang/Basic/BuiltinsAMDGPU.def
    M clang/test/SemaHIP/builtins-amdgcn-raw-buffer-atomic-add.hip

  Log Message:
  -----------
  [AMDGPU] Enable typechecks for __builtin_amdgcn_raw_ptr_buffer_atomic_fadd_v2f16 (#166547)

We didn't remove the `t` for this builtin in the past due to not being
sure if we should use `float16/half`.

This patch doesn't fix the _Float16/half question, I'll address that in
a separate patch later (after discussing the options on our weekly
meeting). At the moment we maintain the `h` for this builtin (which is
likely not what we want for HIP).



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