[PATCH] D65400: [AArch64] Skip isZIPMask check for masks with an odd number of elements.

[Florian Hahn via Phabricator via llvm-commits]

fhahn marked an inline comment as done.
fhahn added a comment.

In D65400#1604658 <https://reviews.llvm.org/D65400#1604658>, @dmgreen wrote:

> Do we want this in any of the other other functions? isUZP_v_undef_Mask looks suspicious.


I'm not entirely sure, but it currently probably only makes sense for vectors with even number of elements. But  I don't think we get out-of-bounds errors there. However I think it would make sense to look into isUZP_v_undef_Mask independently. Also, we probably could handle that slightly better than just bailing out in the long run.



================
Comment at: llvm/test/CodeGen/AArch64/arm64-neon-vector-shuffle-extract.ll:35
+
+define void @zip_mask_check(<3 x float>* %p1, <3 x float>* %p2, i32* %p3) {
+  %tmp3 = load <3 x float>, <3 x float>* %p1, align 16
----------------
dmgreen wrote:
> Update checks!
Don, thanks!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D65400/new/

https://reviews.llvm.org/D65400