[PATCH] D52081: [InstCombine] do not expand 8 byte memcpy if optimising for minsize

[Sanjay Patel via Phabricator via llvm-commits]

spatel added a comment.

I agree with all of the earlier review comments:

1. The way the code is written currently is not good.
2. If we're going to expand at all, we should be using the datalayout to decide when that is appropriate (no magic numbers).
3. If we want to distinguish optimizing for size from optimizing for speed, that belongs in the backend - and it's already there. For example, see SelectionDAG::getMemcpy() and MemCpyOptimizer.cpp.


https://reviews.llvm.org/D52081