[PATCH] D44735: [LV] Add TTI::shouldMaximizeVectorBandwidth to allow enabling it per target

[Hideki Saito via Phabricator via llvm-commits]

hsaito added a comment.

I think this is much cleaner than the other way. Your target can return extremely high cost for super-inefficient "less than full vector" loads/stores, right?

Added the author of https://reviews.llvm.org/D33341.


Repository:
  rL LLVM

https://reviews.llvm.org/D44735