[VectorCombine] Use hasOneUser in shuffle-to-identity fold (#152675)

We need to check that the node is part of the graph being converted, so
will not contain external uses when transformed.

Author: davemgreen

llvm/lib/Transforms/Vectorize/VectorCombine.cpp

@@ -2910,7 +2910,7 @@ bool VectorCombine::foldShuffleToIdentity(Instruction &I) {
       if (!IL.first)
         return true;
       Value *V = IL.first->get();
-      if (auto *I = dyn_cast<Instruction>(V); I && !I->hasOneUse())
+      if (auto *I = dyn_cast<Instruction>(V); I && !I->hasOneUser())
         return false;
       if (V->getValueID() != FrontV->getValueID())
         return false;

llvm/test/Transforms/PhaseOrdering/AArch64/interleave_vec.ll

@@ -320,30 +320,17 @@ define void @same_op3_splat(ptr noalias noundef %a, ptr noundef %b, ptr noundef
 ; CHECK-NEXT:  [[ENTRY:.*]]:
 ; CHECK-NEXT:    [[TMP0:%.*]] = load float, ptr [[C]], align 4
 ; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[TMP0]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x float> [[BROADCAST_SPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <4 x float> [[BROADCAST_SPLATINSERT]], <4 x float> poison, <12 x i32> zeroinitializer
 ; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
 ; CHECK:       [[VECTOR_BODY]]:
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = mul i64 [[INDEX]], 3
 ; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds nuw float, ptr [[B]], i64 [[OFFSET_IDX]]
 ; CHECK-NEXT:    [[WIDE_VEC:%.*]] = load <12 x float>, ptr [[TMP1]], align 4
-; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = shufflevector <12 x float> [[WIDE_VEC]], <12 x float> poison, <4 x i32> <i32 0, i32 3, i32 6, i32 9>
-; CHECK-NEXT:    [[STRIDED_VEC12:%.*]] = shufflevector <12 x float> [[WIDE_VEC]], <12 x float> poison, <4 x i32> <i32 1, i32 4, i32 7, i32 10>
-; CHECK-NEXT:    [[STRIDED_VEC13:%.*]] = shufflevector <12 x float> [[WIDE_VEC]], <12 x float> poison, <4 x i32> <i32 2, i32 5, i32 8, i32 11>
-; CHECK-NEXT:    [[TMP2:%.*]] = fmul fast <4 x float> [[STRIDED_VEC]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP3:%.*]] = getelementptr inbounds nuw float, ptr [[A]], i64 [[OFFSET_IDX]]
 ; CHECK-NEXT:    [[WIDE_VEC14:%.*]] = load <12 x float>, ptr [[TMP3]], align 4
-; CHECK-NEXT:    [[STRIDED_VEC15:%.*]] = shufflevector <12 x float> [[WIDE_VEC14]], <12 x float> poison, <4 x i32> <i32 0, i32 3, i32 6, i32 9>
-; CHECK-NEXT:    [[STRIDED_VEC16:%.*]] = shufflevector <12 x float> [[WIDE_VEC14]], <12 x float> poison, <4 x i32> <i32 1, i32 4, i32 7, i32 10>
-; CHECK-NEXT:    [[TMP4:%.*]] = fadd fast <4 x float> [[STRIDED_VEC15]], [[TMP2]]
-; CHECK-NEXT:    [[TMP5:%.*]] = fmul fast <4 x float> [[STRIDED_VEC12]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT:    [[TMP6:%.*]] = fadd fast <4 x float> [[STRIDED_VEC16]], [[TMP5]]
-; CHECK-NEXT:    [[TMP7:%.*]] = fmul fast <4 x float> [[STRIDED_VEC13]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT:    [[TMP8:%.*]] = shufflevector <4 x float> [[TMP4]], <4 x float> [[TMP6]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
-; CHECK-NEXT:    [[TMP9:%.*]] = shufflevector <12 x float> [[WIDE_VEC14]], <12 x float> poison, <8 x i32> <i32 2, i32 5, i32 8, i32 11, i32 poison, i32 poison, i32 poison, i32 poison>
-; CHECK-NEXT:    [[TMP10:%.*]] = shufflevector <4 x float> [[TMP7]], <4 x float> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison>
-; CHECK-NEXT:    [[TMP11:%.*]] = fadd fast <8 x float> [[TMP9]], [[TMP10]]
-; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = shufflevector <8 x float> [[TMP8]], <8 x float> [[TMP11]], <12 x i32> <i32 0, i32 4, i32 8, i32 1, i32 5, i32 9, i32 2, i32 6, i32 10, i32 3, i32 7, i32 11>
+; CHECK-NEXT:    [[TMP4:%.*]] = fmul fast <12 x float> [[WIDE_VEC]], [[TMP2]]
+; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = fadd fast <12 x float> [[WIDE_VEC14]], [[TMP4]]
 ; CHECK-NEXT:    store <12 x float> [[INTERLEAVED_VEC]], ptr [[TMP3]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], 384

llvm/test/Transforms/VectorCombine/AArch64/shuffletoidentity.ll

@@ -716,10 +716,9 @@ define <4 x i64> @zext_chain(<4 x i16> %x) {
 
 define <4 x i32> @add_chain(<4 x i32> %x) {
 ; CHECK-LABEL: @add_chain(
-; CHECK-NEXT:    [[SHUF:%.*]] = shufflevector <4 x i32> [[X:%.*]], <4 x i32> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
-; CHECK-NEXT:    [[ADD:%.*]] = add <4 x i32> [[SHUF]], [[SHUF]]
-; CHECK-NEXT:    [[ADD2:%.*]] = add <4 x i32> [[ADD]], [[ADD]]
-; CHECK-NEXT:    [[REVSHUF:%.*]] = shufflevector <4 x i32> [[ADD2]], <4 x i32> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
+; CHECK-NEXT:    [[ADD:%.*]] = add <4 x i32> [[SHUF:%.*]], [[SHUF]]
+; CHECK-NEXT:    [[TMP2:%.*]] = add <4 x i32> [[SHUF]], [[SHUF]]
+; CHECK-NEXT:    [[REVSHUF:%.*]] = add <4 x i32> [[ADD]], [[TMP2]]
 ; CHECK-NEXT:    ret <4 x i32> [[REVSHUF]]
 ;
   %shuf = shufflevector <4 x i32> %x, <4 x i32> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
@@ -731,10 +730,9 @@ define <4 x i32> @add_chain(<4 x i32> %x) {
 
 define <4 x i64> @zext_add_chain(<4 x i32> %x) {
 ; CHECK-LABEL: @zext_add_chain(
-; CHECK-NEXT:    [[SHUF:%.*]] = shufflevector <4 x i32> [[X:%.*]], <4 x i32> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
-; CHECK-NEXT:    [[ZEXT:%.*]] = zext <4 x i32> [[SHUF]] to <4 x i64>
-; CHECK-NEXT:    [[ADD:%.*]] = add <4 x i64> [[ZEXT]], [[ZEXT]]
-; CHECK-NEXT:    [[REVSHUF:%.*]] = shufflevector <4 x i64> [[ADD]], <4 x i64> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
+; CHECK-NEXT:    [[ZEXT:%.*]] = zext <4 x i32> [[SHUF:%.*]] to <4 x i64>
+; CHECK-NEXT:    [[TMP2:%.*]] = zext <4 x i32> [[SHUF]] to <4 x i64>
+; CHECK-NEXT:    [[REVSHUF:%.*]] = add <4 x i64> [[ZEXT]], [[TMP2]]
 ; CHECK-NEXT:    ret <4 x i64> [[REVSHUF]]
 ;
   %shuf = shufflevector <4 x i32> %x, <4 x i32> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>