[llvm] r214634 - [AArch64] Teach DAGCombiner that converting two consecutive loads into a vector load is not a good transform when paired loads are available.

Tim Northover t.p.northover at gmail.com
Sat Aug 2 09:49:52 PDT 2014


> Wouldn't it be better to use AArch64 triples? Or is that not
> possible/desirable with Apple triples?

It's mostly irrelevant either way. Both are tested (probably rightly);
any eventual change will need a massive "sed s//" anyway; and in the
meantime they should behave the same.

Cheers.

Tim.



More information about the llvm-commits mailing list