[flang-commits] [clang] [flang] [llvm] [mlir] [IR] Generalize !amdgpu.ignore.denormal.mode into !atomic.ignore.denormal.mode (PR #217585)

Christian Sigg via flang-commits flang-commits at lists.llvm.org
Mon Sep 14 04:49:23 PDT 2026


chsigg wrote:

Friendly ping @arsenm @jayfoad:

Artem approved this PR to generalize `!amdgpu.ignore.denormal.mode` into the target-independent `!atomic.ignore.denormal.mode`, but asked for an OK from the AMD side since it touches AMDGPU atomic lowering and tests.

Summary of the change for AMDGPU:

- Mechanical generalization of `!amdgpu.ignore.denormal.mode` → `!atomic.ignore.denormal.mode` across AMDGPU ISel, OpenMP, and Clang.
- AutoUpgrade preserves backwards compatibility when materializing older bitcode containing `!amdgpu.ignore.denormal.mode`.
- All AMDGPU and SPIR-V lit tests have been updated and all CI checks are passing green.

Could you take a quick look and let me know if this looks good to land from the AMD perspective? Thanks!

https://github.com/llvm/llvm-project/pull/217585


More information about the flang-commits mailing list