[PATCH] D64212: [ARM] MVE Vector Shifts
Sjoerd Meijer via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue Jul 9 05:05:09 PDT 2019
SjoerdMeijer added a comment.
Just a general question first, I was wondering about this:
> NEON was previously lowering the shifts into intrinsics and selecting on the intrinsic. I've converted that to a node shared between NEON and MVE which shifts left whilst still storing sign (ARMISD::VSHLs and ARMISD:VSHLu)
Would it perhaps be useful to do this in 2 steps? First the refactoring, the NEON part and the introduction of the nodes, which should be a NFC, and then in a 2nd follow up patch the MVE changes?
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D64212/new/
https://reviews.llvm.org/D64212
More information about the llvm-commits
mailing list