[PATCH] D48826: [AMDGPU] Add support for TFE/LWE in image intrinsics

Tim Renouf via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed Aug 29 06:32:21 PDT 2018


tpr added a comment.
Herald added a subscriber: jvesely.

Please also fold in the fix to disable InstCombine for image ops with tfe/lfe (AMDVLK fixes c5675e2d, fea7c135, 7b7276ec).



================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:7798
+  unsigned TFEIdx = DmaskIdx + 5;
+  unsigned LWEIdx = DmaskIdx + 6;
+  unsigned UsesTFC = (Node->getConstantOperandVal(TFEIdx) ||
----------------
Please use operand names (per AMDVLK fix 458248e9).


Repository:
  rL LLVM

https://reviews.llvm.org/D48826





More information about the llvm-commits mailing list