[PATCH] D122703: [AArch64][InstCombine] Fold MLOAD and zero extensions into MLOAD

[Allen zhong via Phabricator via llvm-commits]

Allen updated this revision to Diff 419693.
Allen added a comment.
Herald added a subscriber: ecnelises.

rewrite to support splat_vec in DAGCombiner::visitAND


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D122703/new/

https://reviews.llvm.org/D122703

Files:
  llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
  llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
  llvm/test/CodeGen/AArch64/sve-masked-ldst-zext.ll


Index: llvm/test/CodeGen/AArch64/sve-masked-ldst-zext.ll
===================================================================
--- llvm/test/CodeGen/AArch64/sve-masked-ldst-zext.ll
+++ llvm/test/CodeGen/AArch64/sve-masked-ldst-zext.ll
@@ -97,7 +97,7 @@
 ; CHECK-LABEL: masked_zload_2i16_2f64:
 ; CHECK:       ld1h { z0.d }, p0/z, [x0]
 ; CHECK-NEXT:  ptrue p0.d
-; CHECK-NEXT:  ucvtf z0.d, p0/m, z0.s
+; CHECK-NEXT:  ucvtf z0.d, p0/m, z0.d
 ; CHECK-NEXT:  ret
   %wide.load = call <vscale x 2 x i16> @llvm.masked.load.nxv2i16(<vscale x 2 x i16>* %in, i32 2, <vscale x 2 x i1> %mask, <vscale x 2 x i16> undef)
   %zext = zext <vscale x 2 x i16> %wide.load to <vscale x 2 x i32>
Index: llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
===================================================================
--- llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -1247,7 +1247,6 @@
       setLoadExtAction(Op, MVT::nxv2i64, MVT::nxv2i16, Legal);
       setLoadExtAction(Op, MVT::nxv2i64, MVT::nxv2i32, Legal);
       setLoadExtAction(Op, MVT::nxv4i32, MVT::nxv4i8, Legal);
-      setLoadExtAction(Op, MVT::nxv2i32, MVT::nxv2i16, Legal);
       setLoadExtAction(Op, MVT::nxv4i32, MVT::nxv4i16, Legal);
       setLoadExtAction(Op, MVT::nxv8i16, MVT::nxv8i8, Legal);
     }
Index: llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -6054,9 +6054,11 @@
       return N0;
 
     // fold (and (masked_load) (build_vec (x, ...))) to zext_masked_load
+    // fold (and (masked_load) (splat_vec (x, ...))) to zext_masked_load
     auto *MLoad = dyn_cast<MaskedLoadSDNode>(N0);
-    auto *BVec = dyn_cast<BuildVectorSDNode>(N1);
-    if (MLoad && BVec && MLoad->getExtensionType() == ISD::EXTLOAD &&
+    if (MLoad && MLoad->getExtensionType() == ISD::EXTLOAD &&
+        (dyn_cast<BuildVectorSDNode>(N1) ||
+         N1->getOpcode() == ISD::SPLAT_VECTOR) &&
         N0.hasOneUse() && N1.hasOneUse()) {
       EVT LoadVT = MLoad->getMemoryVT();
       EVT ExtVT = VT;
@@ -6064,7 +6066,7 @@
         // For this AND to be a zero extension of the masked load the elements
         // of the BuildVec must mask the bottom bits of the extended element
         // type
-        if (ConstantSDNode *Splat = BVec->getConstantSplatNode()) {
+        if (ConstantSDNode *Splat = isConstOrConstSplat(N1, true, true)) {
           uint64_t ElementSize =
               LoadVT.getVectorElementType().getScalarSizeInBits();
           if (Splat->getAPIntValue().isMask(ElementSize)) {


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D122703.419693.patch
Type: text/x-patch
Size: 2672 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220401/c7ca2dfb/attachment.bin>