[PATCH] D86894: [SVE] Disable INSERT_SUBVECTOR DAGCombine for scalable vectors

[Cameron McInally via Phabricator via llvm-commits]

cameron.mcinally created this revision.
cameron.mcinally added reviewers: ctetreau, sdesmalen, fpetrogalli, c-rhodes, david-arm, kmclaughlin, paulwalker-arm.
Herald added subscribers: llvm-commits, psnobl, hiraditya, tschuett.
Herald added a reviewer: efriedma.
Herald added a project: LLVM.
cameron.mcinally requested review of this revision.

Looking for some input on this combine wrt scalable vectors. Right now, it uses code like this:

        unsigned NumElts = VT.getVectorNumElements();
  ...
          if ((NumElts % Scale) == 0 && (InsIdx % Scale) == 0) {

So it's obviously not compatible with scalable vectors right now.

Has anyone thought through overloading % for ElementCount? At first glance, it seems like we can't since we don't know the vscale. Am I missing something there?


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D86894

Files:
  llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp


Index: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -20588,7 +20588,8 @@
     EVT N0SrcSVT = N0Src.getValueType().getScalarType();
     EVT N1SrcSVT = N1Src.getValueType().getScalarType();
     if ((N0.isUndef() || N0SrcSVT == N1SrcSVT) &&
-        N0Src.getValueType().isVector() && N1Src.getValueType().isVector()) {
+        N0Src.getValueType().isFixedLengthVector() &&
+        N1Src.getValueType().isFixedLengthVector()) {
       EVT NewVT;
       SDLoc DL(N);
       SDValue NewIdx;


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D86894.289016.patch
Type: text/x-patch
Size: 662 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200831/8a90a0c8/attachment.bin>