[PATCH] D31693: [AMDGPU] Unroll more to eliminate phis and conditions

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

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

Increase threshold to unroll a loop which contains an "if" statement
whose condition defined by a PHI belonging to the loop. This may help
to eliminate if region and potentially even PHI itself, saving on
both divergence and registers used for the PHI.

Add a small bonus for each of such "if" statements.


Repository:
  rL LLVM

https://reviews.llvm.org/D31693

Files:
  lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
  test/CodeGen/AMDGPU/unroll.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D31693.94150.patch
Type: text/x-patch
Size: 4608 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170405/c88fce79/attachment.bin>