[PATCH] D13269: Improved X86-FMA3 mem-folding & coalescing

[Quentin Colombet via llvm-commits]

qcolombet accepted this revision.
qcolombet added a comment.

Hi Slava,

LGTM.

Regarding fixing the commute bug, forget about the sub register thing I was talking about. There is no precedence and on a second thought I am not sure it is the right to do. In other words, stick to the *_Int approach when you’ll get to it.

Thanks,
-Quentin


http://reviews.llvm.org/D13269