[PATCH] D118979: [AArch64] Set maximum VF with shouldMaximizeVectorBandwidth

JinGu Kang via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed Feb 9 02:35:44 PST 2022


jaykang10 added inline comments.


================
Comment at: llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h:139
+  bool shouldMaximizeVectorBandwidth() const {
+    if (ST->hasSVE())
+      return false;
----------------
paulwalker-arm wrote:
> jaykang10 wrote:
> > dmgreen wrote:
> > > It's generally best if fixed length vectorization doesn't start behaving differently just because SVE is available (unless it can be better, of course). If we expect MaximizeVectorBandwidth to be better, but doesn't work for scalable vectors well, can we just try to disable the scalable VFs from being widened?
> > I think it could be good to have some comments from SVE people... If possible, can you add them as reviewer please? I do not know well who work on it...
> I agree with Dave, this decision is distinct from whether SVE is available or not. As well as this affecting the NEON side of things there are circumstances where SVE is also used for fixed length vectorisation. Perhaps this function should be changed to take a `TargetTransformInfo::RegisterKind` much like getRegisterBitWidth?
Thanks for comment @paulwalker-arm!
Let me update the function with the `TargetTransformInfo::RegisterKind`.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D118979/new/

https://reviews.llvm.org/D118979



More information about the llvm-commits mailing list