[PATCH] D119510: AMDGPU: Clamp min value of effective waves-per-eu instead of discarding

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUSubtarget.cpp:562-563
   // minimum/maximum flat work group sizes.
   if (Requested.first < MinImpliedByFlatWorkGroupSize)
-    return Default;
+    Requested.first = MinImpliedByFlatWorkGroupSize;
 
----------------
Just use std::max?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D119510/new/

https://reviews.llvm.org/D119510