[Mlir-commits] [mlir] [mlir][AMDGPU] Enable emulating vector buffer_atomic_fadd for bf16 on gfx942 (PR #129029)

[Jakub Kuderski]

================
@@ -189,6 +189,14 @@ void mlir::amdgpu::populateAmdgpuEmulateAtomicsPatterns(
     } else {
       target.addIllegalOp<RawBufferAtomicFmaxOp>();
     }
+    // gfx950 has bf16 atomics
----------------
kuhar wrote:

Could you open an issue to refactor how we handle target features in this pass and add a TODO linking to it here? These can be passed in as a struct, so that downstream compilers like IREE or Triton can provide target features.

https://github.com/llvm/llvm-project/pull/129029