[PATCH] D18340: AMDGPU: allow specifying a workgroup size that needs to fit in a compute unit

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Mon Mar 21 15:50:58 PDT 2016

arsenm added a comment.

I have another patch which already fixes a hardcoded maximum of 256: http://reviews.llvm.org/D16834

You can try merging this what that one.

New attributes should be lowercased,use  '-' separated words and start with an "amdgpu-" prefix

Comment at: lib/Target/AMDGPU/AMDGPU.h:131-132
@@ -130,3 +130,4 @@
     GEOMETRY = 2,
-    COMPUTE = 3
+    COMPUTE = 3,
+    GL_COMPUTE = 4
I think theses should be distinguished by the ABI rather than introducing a new shader type. Also semi-related, I would like to replace shader types with calling conventions


More information about the llvm-commits mailing list