This was creating natural aligned loads and stores, which may not be
the case. The target could request a wider type load with less
alignment.
Details
Details
Diff Detail
Diff Detail
Event Timeline
Comment Actions
I think this makes sense. LGTM, one nit.
llvm/lib/Transforms/Utils/LowerMemIntrinsics.cpp | ||
---|---|---|
64 | Isn't PartSize == LoopOpSize ? |
Isn't PartSize == LoopOpSize ?