[PATCH] D46596: [AMDGPU] Optimze old value of v_mov_b32_dpp

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec created this revision.
rampitec added reviewers: arsenm, kzhuravl, b-sumner, nhaehnle.
Herald added subscribers: t-tye, tpr, dstuttard, yaxunl, wdng.

We can eliminate old value if bound_ctrl = 1 and row_mask = bank_mask = 0xf.
This is alternative implementation working with the intrinsic in InstCombine.
Original review for past-ISel optimization: https://reviews.llvm.org/D46570.


https://reviews.llvm.org/D46596

Files:
  lib/Transforms/InstCombine/InstCombineCalls.cpp
  test/CodeGen/AMDGPU/llvm.amdgcn.update.dpp.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D46596.145754.patch
Type: text/x-patch
Size: 5367 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180508/f4ad8af9/attachment-0001.bin>