r342592 | dim | 2018-12-29 16:13:49 +0100 (Sat, 29 Dec 2018) | 19 lines
Pull in r342397 from upstream llvm trunk (by Amara Emerson):
Revert "Revert r342183 "[DAGCombine] Fix crash when store merging
created an extract_subvector with invalid index.""
Fixed the assertion failure.
Differential Revision: https://reviews.llvm.org/D51831
This fixes 'Assertion failed: ((VT.getVectorNumElements() +
N2C->getZExtValue() <= N1.getValueType().getVectorNumElements()) &&
"Extract subvector overflow!"), function getNode' when building the
multimedia/aom port (with AVX2 enabled).
Reported by: jbeich
PR: 234480
Index: lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- lib/CodeGen/SelectionDAG/DAGCombiner.cpp (revision 342591)
+++ lib/CodeGen/SelectionDAG/DAGCombiner.cpp (revision 342592)
@@ -13727,10 +13727,19 @@ bool DAGCombiner::MergeStoresOfConstantsOrVecElts(
Val.getOpcode() == ISD::EXTRACT_SUBVECTOR)) {
SDValue Vec = Val.getOperand(0);
EVT MemVTScalarTy = MemVT.getScalarType();
+ SDValue Idx = Val.getOperand(1);
// We may need to add a bitcast here to get types to line up.
if (MemVTScalarTy != Vec.getValueType()) {
unsigned Elts = Vec.getValueType().getSizeInBits() /
MemVTScalarTy.getSizeInBits();
+ if (Val.getValueType().isVector() && MemVT.isVector()) {
+ unsigned IdxC = cast<ConstantSDNode>(Idx)->getZExtValue();
+ unsigned NewIdx =
+ ((uint64_t)IdxC * MemVT.getVectorNumElements()) / Elts;
+ Idx = DAG.getConstant(NewIdx, SDLoc(Val), Idx.getValueType());
+ }
+ if (!MemVT.isVector() && Val.getValueType().isVector())
+ dbgs() << "hit!\n";
EVT NewVecTy =
EVT::getVectorVT(*DAG.getContext(), MemVTScalarTy, Elts);
Vec = DAG.getBitcast(NewVecTy, Vec);
@@ -13737,7 +13746,7 @@ bool DAGCombiner::MergeStoresOfConstantsOrVecElts(
}
auto OpC = (MemVT.isVector()) ? ISD::EXTRACT_SUBVECTOR
: ISD::EXTRACT_VECTOR_ELT;
- Val = DAG.getNode(OpC, SDLoc(Val), MemVT, Vec, Val.getOperand(1));
+ Val = DAG.getNode(OpC, SDLoc(Val), MemVT, Vec, Idx);
}
Ops.push_back(Val);
}
Index: test/CodeGen/X86/merge-vector-stores-scale-idx-crash.ll
===================================================================
--- test/CodeGen/X86/merge-vector-stores-scale-idx-crash.ll (nonexistent)
+++ test/CodeGen/X86/merge-vector-stores-scale-idx-crash.ll (revision 342592)
@@ -0,0 +1,49 @@
+; RUN: llc < %s -mtriple=x86_64-apple-osx10.14 -mattr=+avx2 | FileCheck %s
+
+; Check that we don't crash due creating invalid extract_subvector indices in store merging.
+; CHECK-LABEL: testfn
+; CHECK: retq
+define void @testfn(i32* nocapture %p) {
+ %v0 = getelementptr i32, i32* %p, i64 12
+ %1 = bitcast i32* %v0 to <2 x i64>*
+ %2 = bitcast i32* %v0 to <4 x i32>*
+ %3 = getelementptr <2 x i64>, <2 x i64>* %1, i64 -3
+ store <2 x i64> undef, <2 x i64>* %3, align 16
+ %4 = shufflevector <4 x i64> zeroinitializer, <4 x i64> undef, <2 x i32> <i32 0, i32 1>
+ %5 = getelementptr <2 x i64>, <2 x i64>* %1, i64 -2
+ store <2 x i64> %4, <2 x i64>* %5, align 16
+ %6 = shufflevector <8 x i32> zeroinitializer, <8 x i32> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
+ %7 = getelementptr <4 x i32>, <4 x i32>* %2, i64 -1
+ store <4 x i32> %6, <4 x i32>* %7, align 16
+ ret void
+}
+
+%struct.o.1.5.7.9.13.15.17.19.57 = type { [0 x %struct.d.0.4.6.8.12.14.16.18.56] }
+%struct.d.0.4.6.8.12.14.16.18.56 = type { float, float }
+
+; CHECK-LABEL: testfn_scalar
+; CHECK: retq
+define void @testfn_scalar(%struct.o.1.5.7.9.13.15.17.19.57* nocapture %j) local_unnamed_addr #0 align 2 {
+entry:
+ %0 = bitcast i64 undef to <2 x float>
+ br i1 undef, label %if.end, label %if.then
+
+if.then: ; preds = %entry
+ unreachable
+
+if.end: ; preds = %entry
+ %call.i.i17 = tail call <4 x float> @_Z1bv()
+ %1 = bitcast <4 x float> %call.i.i17 to <2 x i64>
+ %extract.i.i15 = extractelement <2 x i64> %1, i64 0
+ %arrayidx6 = getelementptr inbounds %struct.o.1.5.7.9.13.15.17.19.57, %struct.o.1.5.7.9.13.15.17.19.57* %j, i64 0, i32 0, i64 1
+ %2 = bitcast %struct.d.0.4.6.8.12.14.16.18.56* %arrayidx6 to i64*
+ store i64 %extract.i.i15, i64* %2, align 4
+ %r.sroa.0.0..sroa_cast12 = bitcast %struct.o.1.5.7.9.13.15.17.19.57* %j to <2 x float>*
+ store <2 x float> %0, <2 x float>* %r.sroa.0.0..sroa_cast12, align 4
+ ret void
+}
+
+declare <4 x float> @_Z1bv() local_unnamed_addr
+
+attributes #0 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "less-precise-fpmad"="false" "min-legal-vector-width"="128" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
+