[PATCH] D60789: [x86] try to widen 'shl' as part of LEA formation

[Sanjay Patel via Phabricator via llvm-commits]

spatel updated this revision to Diff 195481.
spatel added a comment.

Patch updated:

1. Try to make code comment clearer.
2. Set scale/index parts of the AddressMode directly rather than recursing.
3. Remove dead node and explicitly RAUW.
4. Added some shift constant variation to the tests and a negative test.

#3 is unclear to me...some cases within here appear to do part of this node updating explicitly while others seem to leave it to the callers (or other passes?). My understanding of what's expected at this level isn't good, so what I have there right now might be the worst of all worlds. :)


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D60789/new/

https://reviews.llvm.org/D60789

Files:
  llvm/lib/Target/X86/X86ISelDAGToDAG.cpp
  llvm/test/CodeGen/X86/lea-dagdag.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D60789.195481.patch
Type: text/x-patch
Size: 3946 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190416/0ecbef6e/attachment.bin>