[PATCH] D61075: [CodeGenPrepare] limit overflow intrinsic matching to a single basic block

[Sanjay Patel via Phabricator via llvm-commits]

spatel updated this revision to Diff 196682.
spatel retitled this revision from "[CodeGenPrepare] delay instruction deletion for efficiency" to "[CodeGenPrepare] limit overflow intrinsic matching to a single basic block".
spatel edited the summary of this revision.
spatel added a comment.

Patch updated:
Don't try to match this pattern across blocks. That means we are not using a dominator tree at all, so if this doesn't restore compile-time perf, I'm not sure what would. :)

Limiting this transform to a single block means it really shouldn't be in CGP anymore. It could be implemented in SDAG. But I'd prefer to leave that to a follow-up patch once we confirm that this (1) restores compile-time and (2) does not cause perf regressions.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61075/new/

https://reviews.llvm.org/D61075

Files:
  llvm/lib/CodeGen/CodeGenPrepare.cpp
  llvm/test/CodeGen/X86/cgp-usubo.ll
  llvm/test/Transforms/CodeGenPrepare/X86/optimizeSelect-DT.ll
  llvm/test/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D61075.196682.patch
Type: text/x-patch
Size: 10920 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190425/9b3921b8/attachment.bin>