[LV] Limits the splat operations be hoisted must not be defined by a recipe.

Author: Mel-Chen

llvm/lib/Transforms/Vectorize/VPlan.cpp

@@ -267,7 +267,7 @@ Value *VPTransformState::get(VPValue *Def, bool NeedsScalar) {
     return Data.VPV2Vector[Def];
 
   auto GetBroadcastInstrs = [this, Def](Value *V) {
-    bool SafeToHoist = Def->isDefinedOutsideLoopRegions();
+    bool SafeToHoist = !Def->hasDefiningRecipe();
     if (VF.isScalar())
       return V;
     // Place the code for broadcasting invariant variables in the new preheader.

llvm/test/Transforms/LoopVectorize/AArch64/extractvalue-no-scalarization-required.ll

@@ -23,9 +23,9 @@
 ; Check that the extractvalue operands are actually free in vector code.
 
 ; FORCED:         [[E1:%.+]] = extractvalue { i64, i64 } %sv, 0
-; FORCED-NEXT:    [[E2:%.+]] = extractvalue { i64, i64 } %sv, 1
 ; FORCED-NEXT:    %broadcast.splatinsert = insertelement <2 x i64> poison, i64 [[E1]], i64 0
 ; FORCED-NEXT:    %broadcast.splat = shufflevector <2 x i64> %broadcast.splatinsert, <2 x i64> poison, <2 x i32> zeroinitializer
+; FORCED-NEXT:    [[E2:%.+]] = extractvalue { i64, i64 } %sv, 1
 ; FORCED-NEXT:    %broadcast.splatinsert1 = insertelement <2 x i64> poison, i64 [[E2]], i64 0
 ; FORCED-NEXT:    %broadcast.splat2 = shufflevector <2 x i64> %broadcast.splatinsert1, <2 x i64> poison, <2 x i32> zeroinitializer
 ; FORCED-NEXT:    [[ADD:%.+]] = add <2 x i64> %broadcast.splat, %broadcast.splat2
@@ -75,9 +75,9 @@ declare float @powf(float, float) readnone nounwind
 ; FORCED-LABEL: define void @test_getVectorCallCost
 
 ; FORCED:         [[E1:%.+]] = extractvalue { float, float } %sv, 0
-; FORCED-NEXT:    [[E2:%.+]] = extractvalue { float, float } %sv, 1
 ; FORCED-NEXT:    %broadcast.splatinsert = insertelement <2 x float> poison, float [[E1]], i64 0
 ; FORCED-NEXT:    %broadcast.splat = shufflevector <2 x float> %broadcast.splatinsert, <2 x float> poison, <2 x i32> zeroinitializer
+; FORCED-NEXT:    [[E2:%.+]] = extractvalue { float, float } %sv, 1
 ; FORCED-NEXT:    %broadcast.splatinsert1 = insertelement <2 x float> poison, float [[E2]], i64 0
 ; FORCED-NEXT:    %broadcast.splat2 = shufflevector <2 x float> %broadcast.splatinsert1, <2 x float> poison, <2 x i32> zeroinitializer
 

llvm/test/Transforms/LoopVectorize/AArch64/sve-widen-extractvalue.ll

@@ -5,9 +5,9 @@ target triple = "aarch64-unknown-linux-gnu"
 define void @widen_extractvalue(ptr %dst, {i64, i64} %sv) #0 {
 ; CHECK-LABEL: @widen_extractvalue(
 ; CHECK:        [[EXTRACT0:%.*]] = extractvalue { i64, i64 } [[SV:%.*]], 0
-; CHECK-NEXT:   [[EXTRACT1:%.*]] = extractvalue { i64, i64 } [[SV]], 1
 ; CHECK-NEXT:   [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[EXTRACT0]], i64 0
 ; CHECK-NEXT:   [[DOTSPLAT:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+; CHECK-NEXT:   [[EXTRACT1:%.*]] = extractvalue { i64, i64 } [[SV]], 1
 ; CHECK-NEXT:   [[DOTSPLATINSERT1:%.*]] = insertelement <vscale x 2 x i64> poison, i64 [[EXTRACT1]], i64 0
 ; CHECK-NEXT:   [[DOTSPLAT2:%.*]] = shufflevector <vscale x 2 x i64> [[DOTSPLATINSERT1]], <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
 ; CHECK:        [[ADD:%.*]] = add <vscale x 2 x i64> [[DOTSPLAT]], [[DOTSPLAT2]]

llvm/test/Transforms/LoopVectorize/X86/epilog-vectorization-inductions.ll

@@ -137,6 +137,8 @@ define void @test_induction_step_needs_expansion(ptr noalias %j, ptr %k, i64 %l,
 ; CHECK-LABEL: @test_induction_step_needs_expansion(
 ; CHECK-NEXT:  iter.check:
 ; CHECK-NEXT:    [[TMP0:%.*]] = sub i16 0, [[OFF:%.*]]
+; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <16 x i16> poison, i16 [[TMP0]], i64 0
+; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <16 x i16> [[BROADCAST_SPLATINSERT]], <16 x i16> poison, <16 x i32> zeroinitializer
 ; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[L:%.*]], 8
 ; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH:%.*]], label [[VECTOR_MAIN_LOOP_ITER_CHECK:%.*]]
 ; CHECK:       vector.main.loop.iter.check:
@@ -147,15 +149,13 @@ define void @test_induction_step_needs_expansion(ptr noalias %j, ptr %k, i64 %l,
 ; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[L]], [[N_MOD_VF]]
 ; CHECK-NEXT:    [[DOTCAST:%.*]] = trunc i64 [[N_VEC]] to i16
 ; CHECK-NEXT:    [[IND_END:%.*]] = mul i16 [[DOTCAST]], [[TMP0]]
-; CHECK-NEXT:    [[DOTSPLATINSERT2:%.*]] = insertelement <16 x i16> poison, i16 [[TMP0]], i64 0
-; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <16 x i16> [[DOTSPLATINSERT2]], <16 x i16> poison, <16 x i32> zeroinitializer
 ; CHECK-NEXT:    [[TMP1:%.*]] = mul <16 x i16> splat (i16 16), [[TMP2]]
 ; CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <16 x i16> poison, i16 [[TMP0]], i64 0
 ; CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <16 x i16> [[DOTSPLATINSERT]], <16 x i16> poison, <16 x i32> zeroinitializer
 ; CHECK-NEXT:    [[TMP11:%.*]] = mul <16 x i16> <i16 0, i16 1, i16 2, i16 3, i16 4, i16 5, i16 6, i16 7, i16 8, i16 9, i16 10, i16 11, i16 12, i16 13, i16 14, i16 15>, [[DOTSPLAT]]
 ; CHECK-NEXT:    [[INDUCTION:%.*]] = add <16 x i16> zeroinitializer, [[TMP11]]
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <16 x i16> poison, i16 [[OFF]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT3:%.*]] = shufflevector <16 x i16> [[BROADCAST_SPLATINSERT]], <16 x i16> poison, <16 x i32> zeroinitializer
+; CHECK-NEXT:    [[BROADCAST_SPLATINSERT2:%.*]] = insertelement <16 x i16> poison, i16 [[OFF]], i64 0
+; CHECK-NEXT:    [[BROADCAST_SPLAT3:%.*]] = shufflevector <16 x i16> [[BROADCAST_SPLATINSERT2]], <16 x i16> poison, <16 x i32> zeroinitializer
 ; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
 ; CHECK:       vector.body:
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -171,11 +171,11 @@ define void @test_induction_step_needs_expansion(ptr noalias %j, ptr %k, i64 %l,
 ; CHECK-NEXT:    [[TMP8:%.*]] = getelementptr inbounds i16, ptr [[K:%.*]], i64 [[TMP3]]
 ; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr inbounds i16, ptr [[TMP8]], i32 0
 ; CHECK-NEXT:    [[TMP10:%.*]] = getelementptr inbounds i16, ptr [[TMP8]], i32 16
-; CHECK-NEXT:    [[TMP11:%.*]] = getelementptr inbounds i16, ptr [[TMP8]], i32 32
+; CHECK-NEXT:    [[TMP21:%.*]] = getelementptr inbounds i16, ptr [[TMP8]], i32 32
 ; CHECK-NEXT:    [[TMP12:%.*]] = getelementptr inbounds i16, ptr [[TMP8]], i32 48
 ; CHECK-NEXT:    store <16 x i16> [[TMP4]], ptr [[TMP9]], align 2
 ; CHECK-NEXT:    store <16 x i16> [[TMP5]], ptr [[TMP10]], align 2
-; CHECK-NEXT:    store <16 x i16> [[TMP6]], ptr [[TMP11]], align 2
+; CHECK-NEXT:    store <16 x i16> [[TMP6]], ptr [[TMP21]], align 2
 ; CHECK-NEXT:    store <16 x i16> [[TMP7]], ptr [[TMP12]], align 2
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 64
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <16 x i16> [[STEP_ADD_3]], [[TMP1]]