[llvm] [AMDGPU] Overload image atomic swap to allow float as well. (PR #107283)

via llvm-commits llvm-commits at lists.llvm.org
Mon Sep 9 08:54:24 PDT 2024


sstipanovic wrote:

> Why doesn't this need any codegen / pattern changes?

float version was already used out there, so codegen for float was already implemented.

https://github.com/llvm/llvm-project/pull/107283


More information about the llvm-commits mailing list