[PATCH] [X86] Split information collection from actual transformation in call frame optimization (mov -> push)
michael.m.kuperstein at intel.com
Mon Feb 9 04:43:58 PST 2015
This splits collecting information from actually performing the transformation, so that we can add a heuristic in between the two.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 12957 bytes
Desc: not available
More information about the llvm-commits