[PATCH] Reassociate x + -0.1234 * y into x - 0.1234 * y (part 2)

[Chad Rosier]

Hi resistor, chandlerc, erikjv, Jiangning, jmolloy,

This is a follow up patch to http://reviews.llvm.org/D4904.

This gives CSE/GVN more options to eliminate duplicate expressions.  However, this implementation is significantly less rigid and adds support for other operations such as interger add/sub/urem/srem/sdiv/etc.  I need to rerun the performance numbers, but I thought I would begin the review process in the meantime.

Please have a look,
  Chad

http://reviews.llvm.org/D5363

Files:
  lib/Transforms/Scalar/Reassociate.cpp
  test/Transforms/Reassociate/basictest.ll
  test/Transforms/Reassociate/canonicalize-neg-const.ll
  test/Transforms/Reassociate/fast-basictest.ll
  test/Transforms/Reassociate/liftsign.ll
  test/Transforms/Reassociate/shift-factor.ll
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D5363.13733.patch
Type: text/x-patch
Size: 17395 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20140915/1436df29/attachment.bin>