================ @@ -5,7 +5,8 @@ ; RUN: llc -mtriple=amdgpu11.00-amd-amdhsa -O0 -mattr=+cumode < %s | FileCheck --check-prefixes=GFX11-CU %s ; RUN: llc -mtriple=amdgpu12.00-amd-amdhsa -O0 < %s | FileCheck --check-prefixes=GFX12-WGP %s ; RUN: llc -mtriple=amdgpu12.00-amd-amdhsa -O0 -mattr=+cumode < %s | FileCheck --check-prefixes=GFX12-CU %s -; RUN: llc -mtriple=amdgpu12.50-amd-amdhsa -O0 < %s | FileCheck --check-prefixes=GFX1250 %s +; RUN: llc -mtriple=amdgpu12.50-amd-amdhsa -O0 < %s | FileCheck --check-prefixes=GFX1250,GFX1250_NOSTRICT %s ---------------- ritter-x2a wrote:
There are still quite a few cases where gfx1250 and gfx1250s produce the same code, without GFX1250 as a shared prefix those will also lead to new check lines: Then, this patch will have ~69k additions and 0 deletions vs. the current ~33k additions and ~16k deletions. Is that preferable? https://github.com/llvm/llvm-project/pull/236846 _______________________________________________ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
