[llvm] r339166 - [SLP] Fix insert point for reused extract instructions.
Alexey Bataev via llvm-commits
llvm-commits at lists.llvm.org
Tue Aug 7 12:21:05 PDT 2018
Author: abataev
Date: Tue Aug 7 12:21:05 2018
New Revision: 339166
URL: http://llvm.org/viewvc/llvm-project?rev=339166&view=rev
Log:
[SLP] Fix insert point for reused extract instructions.
Summary:
Reworked the previously committed patch to insert shuffles for reused
extract element instructions in the correct position. Previous logic was
incorrect, and might lead to the crash with PHIs and EH instructions.
Reviewers: efriedma, javed.absar
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D50143
Modified:
llvm/trunk/lib/Transforms/Vectorize/SLPVectorizer.cpp
llvm/trunk/test/Transforms/SLPVectorizer/AArch64/PR38339.ll
Modified: llvm/trunk/lib/Transforms/Vectorize/SLPVectorizer.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Vectorize/SLPVectorizer.cpp?rev=339166&r1=339165&r2=339166&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/Vectorize/SLPVectorizer.cpp (original)
+++ llvm/trunk/lib/Transforms/Vectorize/SLPVectorizer.cpp Tue Aug 7 12:21:05 2018
@@ -3109,14 +3109,8 @@ Value *BoUpSLP::vectorizeTree(TreeEntry
}
if (NeedToShuffleReuses) {
// TODO: Merge this shuffle with the ReorderShuffleMask.
- if (!E->ReorderIndices.empty())
+ if (E->ReorderIndices.empty())
Builder.SetInsertPoint(VL0);
- else if (auto *I = dyn_cast<Instruction>(V))
- Builder.SetInsertPoint(I->getParent(),
- std::next(I->getIterator()));
- else
- Builder.SetInsertPoint(&F->getEntryBlock(),
- F->getEntryBlock().getFirstInsertionPt());
V = Builder.CreateShuffleVector(V, UndefValue::get(VecTy),
E->ReuseShuffleIndices, "shuffle");
}
Modified: llvm/trunk/test/Transforms/SLPVectorizer/AArch64/PR38339.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/SLPVectorizer/AArch64/PR38339.ll?rev=339166&r1=339165&r2=339166&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/SLPVectorizer/AArch64/PR38339.ll (original)
+++ llvm/trunk/test/Transforms/SLPVectorizer/AArch64/PR38339.ll Tue Aug 7 12:21:05 2018
@@ -27,3 +27,98 @@ define void @f1(<2 x i16> %x, i16* %a) {
store i16 %t2, i16* %ptr3
ret void
}
+
+define void @f2(<2 x i16> %x, i16* %a) {
+; CHECK-LABEL: @f2(
+; CHECK-NEXT: entry:
+; CHECK-NEXT: br label [[CONT:%.*]]
+; CHECK: cont:
+; CHECK-NEXT: [[XX:%.*]] = phi <2 x i16> [ [[X:%.*]], [[ENTRY:%.*]] ], [ undef, [[CONT]] ]
+; CHECK-NEXT: [[AA:%.*]] = phi i16* [ [[A:%.*]], [[ENTRY]] ], [ undef, [[CONT]] ]
+; CHECK-NEXT: [[SHUFFLE:%.*]] = shufflevector <2 x i16> [[XX]], <2 x i16> undef, <4 x i32> <i32 0, i32 1, i32 1, i32 0>
+; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 0
+; CHECK-NEXT: [[PTR1:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 1
+; CHECK-NEXT: [[PTR2:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 2
+; CHECK-NEXT: [[PTR3:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 3
+; CHECK-NEXT: [[TMP0:%.*]] = extractelement <4 x i16> [[SHUFFLE]], i32 0
+; CHECK-NEXT: store i16 [[TMP0]], i16* [[A]]
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast i16* [[PTR0]] to <4 x i16>*
+; CHECK-NEXT: store <4 x i16> [[SHUFFLE]], <4 x i16>* [[TMP1]], align 2
+; CHECK-NEXT: [[A_VAL:%.*]] = load i16, i16* [[A]], align 2
+; CHECK-NEXT: [[CMP:%.*]] = icmp eq i16 [[A_VAL]], 0
+; CHECK-NEXT: br i1 [[CMP]], label [[CONT]], label [[EXIT:%.*]]
+; CHECK: exit:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %cont
+
+cont: ; preds = %entry, %cont
+ %xx = phi <2 x i16> [ %x, %entry ], [ undef, %cont ]
+ %aa = phi i16* [ %a, %entry ], [ undef, %cont ]
+ %t2 = extractelement <2 x i16> %xx, i32 0
+ %t3 = extractelement <2 x i16> %xx, i32 1
+ %ptr0 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 0
+ %ptr1 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 1
+ %ptr2 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 2
+ %ptr3 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 3
+ store i16 %t2, i16* %a
+ store i16 %t2, i16* %ptr0
+ store i16 %t3, i16* %ptr1
+ store i16 %t3, i16* %ptr2
+ store i16 %t2, i16* %ptr3
+ %a_val = load i16, i16* %a, align 2
+ %cmp = icmp eq i16 %a_val, 0
+ br i1 %cmp, label %cont, label %exit
+
+exit: ; preds = %cont
+ ret void
+}
+
+define void @f3(<2 x i16> %x, i16* %a) {
+; CHECK-LABEL: @f3(
+; CHECK-NEXT: entry:
+; CHECK-NEXT: br label [[CONT:%.*]]
+; CHECK: cont:
+; CHECK-NEXT: [[XX:%.*]] = phi <2 x i16> [ [[X:%.*]], [[ENTRY:%.*]] ], [ undef, [[CONT]] ]
+; CHECK-NEXT: [[AA:%.*]] = phi i16* [ [[A:%.*]], [[ENTRY]] ], [ undef, [[CONT]] ]
+; CHECK-NEXT: [[REORDER_SHUFFLE:%.*]] = shufflevector <2 x i16> [[XX]], <2 x i16> undef, <2 x i32> <i32 1, i32 0>
+; CHECK-NEXT: [[SHUFFLE:%.*]] = shufflevector <2 x i16> [[REORDER_SHUFFLE]], <2 x i16> undef, <4 x i32> <i32 0, i32 1, i32 1, i32 0>
+; CHECK-NEXT: [[PTR0:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 0
+; CHECK-NEXT: [[PTR1:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 1
+; CHECK-NEXT: [[PTR2:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 2
+; CHECK-NEXT: [[PTR3:%.*]] = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 3
+; CHECK-NEXT: [[TMP0:%.*]] = extractelement <4 x i16> [[SHUFFLE]], i32 0
+; CHECK-NEXT: store i16 [[TMP0]], i16* [[A]]
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast i16* [[PTR0]] to <4 x i16>*
+; CHECK-NEXT: store <4 x i16> [[SHUFFLE]], <4 x i16>* [[TMP1]], align 2
+; CHECK-NEXT: [[A_VAL:%.*]] = load i16, i16* [[A]], align 2
+; CHECK-NEXT: [[CMP:%.*]] = icmp eq i16 [[A_VAL]], 0
+; CHECK-NEXT: br i1 [[CMP]], label [[CONT]], label [[EXIT:%.*]]
+; CHECK: exit:
+; CHECK-NEXT: ret void
+;
+entry:
+ br label %cont
+
+cont: ; preds = %entry, %cont
+ %xx = phi <2 x i16> [ %x, %entry ], [ undef, %cont ]
+ %aa = phi i16* [ %a, %entry ], [ undef, %cont ]
+ %t2 = extractelement <2 x i16> %xx, i32 0
+ %t3 = extractelement <2 x i16> %xx, i32 1
+ %ptr0 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 0
+ %ptr1 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 1
+ %ptr2 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 2
+ %ptr3 = getelementptr inbounds [4 x i16], [4 x i16]* undef, i16 0, i16 3
+ store i16 %t3, i16* %a
+ store i16 %t3, i16* %ptr0
+ store i16 %t2, i16* %ptr1
+ store i16 %t2, i16* %ptr2
+ store i16 %t3, i16* %ptr3
+ %a_val = load i16, i16* %a, align 2
+ %cmp = icmp eq i16 %a_val, 0
+ br i1 %cmp, label %cont, label %exit
+
+exit: ; preds = %cont
+ ret void
+}
More information about the llvm-commits
mailing list