[llvm] AMDGPU: Add codegen for atomicrmw operations usub_cond and usub_sat (PR #141068)

[via llvm-commits]

================
@@ -0,0 +1,487 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1030 < %s | FileCheck %s -check-prefix=GFX10-GISEL
+; RUN: llc -global-isel -mtriple=amdgcn-amd-amdhsa -mcpu=gfx1031 < %s | FileCheck %s -check-prefix=GFX10-GISEL
----------------
anjenner wrote:

Ah, turns out the expansion was not working. I fixed it, and fixed the tests, with https://github.com/llvm/llvm-project/pull/141068/commits/28491ab8521f03db48091564b7e38aea61e17e3f .

https://github.com/llvm/llvm-project/pull/141068