[PATCH] D21136: AMDGPU/SI: Set INDEX_STRIDE for scratch coalescing

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Thu Jun 9 11:43:32 PDT 2016


arsenm accepted this revision.
arsenm added a comment.
This revision is now accepted and ready to land.

LGTM


================
Comment at: lib/Target/AMDGPU/SIInstrInfo.cpp:3098
@@ -3098,1 +3097,3 @@
+  Rsrc23 |= (EltSizeValue << AMDGPU::RSRC_ELEMENT_SIZE_SHIFT) |
+            (3LL << AMDGPU::RSRC_INDEX_STRIDE_SHIFT); // IndexStride = 64
 
----------------
Ditto

================
Comment at: lib/Target/AMDGPU/SIInstrInfo.h:553
@@ -553,1 +552,3 @@
+  const uint64_t RSRC_INDEX_STRIDE_SHIFT = (32 + 21);
+  const uint64_t RSRC_TID_ENABLE = 1LL << (32 + 23);
 } // End namespace AMDGPU
----------------
UINT64_C() instead of LL


http://reviews.llvm.org/D21136





More information about the llvm-commits mailing list