[PATCH] D30852: StructurizeCFG: Directly invert cmp instructions

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
Herald added subscribers: nhaehnle, wdng.

The most common case for a branch condition is
a single use compare. Directly invert the branch
predicate rather than adding a lot of xor i1 true
which the DAG will have to fold later.

      

This produces nicer to read structurizer output.

      

This produces some random changes in codegen
due to the DAG swapping branch conditions itself,
 and then does a poor job of dealing with those
inverts.


https://reviews.llvm.org/D30852

Files:
  lib/Transforms/Scalar/StructurizeCFG.cpp
  test/CodeGen/AMDGPU/loop_break.ll
  test/CodeGen/AMDGPU/multi-divergent-exit-region.ll
  test/CodeGen/AMDGPU/ret_jump.ll
  test/Transforms/StructurizeCFG/one-loop-multiple-backedges.ll
  test/Transforms/StructurizeCFG/post-order-traversal-bug.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D30852.91430.patch
Type: text/x-patch
Size: 17416 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170311/ac44c98a/attachment-0001.bin>