[llvm] [AMDGPU] Move `AMDGPUAttributorPass` to full LTO post link stage (PR #102086)

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Fri Aug 9 00:32:52 PDT 2024


================
@@ -1,6 +1,6 @@
-; RUN: opt -mtriple=amdgcn--amdhsa -S -O3 -enable-unsafe-fp-math %s  | FileCheck -check-prefix=GCN -check-prefix=UNSAFE %s
-; RUN: opt -mtriple=amdgcn--amdhsa -S -O3 -enable-no-nans-fp-math %s | FileCheck -check-prefix=GCN -check-prefix=NONANS %s
-; RUN: opt -mtriple=amdgcn--amdhsa -S -O3 -enable-no-infs-fp-math %s | FileCheck -check-prefix=GCN -check-prefix=NOINFS %s
----------------
arsenm wrote:

I'm surprised the attributor induced attributes are here in the first place. I would just leave this running non-LTO (also this test doesn't really belong in codegen. I'm also not sure why its running -O3 instead of just the inliner)

https://github.com/llvm/llvm-project/pull/102086


More information about the llvm-commits mailing list