[PATCH] Reimplement heuristic for estimating complete-unroll optimization effects.

[Michael Zolotukhin]

Thanks, Chandler!

I've committed the patch and will follow-up with the tests today.

> There are a bunch of things I think should be cleaned up here, but they're fairly minor and I'm happy to just fix those and for a few that have more impact, send you patches.


Sure, go ahead!

Michael


REPOSITORY
  rL LLVM

http://reviews.llvm.org/D8816

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/