[llvm] 7adf664 - [Matrix] Propagate and use shape information for loads.
Florian Hahn via llvm-commits
llvm-commits at lists.llvm.org
Thu Jan 9 02:21:50 PST 2020
Author: Florian Hahn
Date: 2020-01-09T10:21:20Z
New Revision: 7adf6644f50a60a25f81c976b0f3a08320a63665
URL: https://github.com/llvm/llvm-project/commit/7adf6644f50a60a25f81c976b0f3a08320a63665
DIFF: https://github.com/llvm/llvm-project/commit/7adf6644f50a60a25f81c976b0f3a08320a63665.diff
LOG: [Matrix] Propagate and use shape information for loads.
This patch extends to shape propagation to also include load
instructions and implements shape aware lowering for vector loads.
Reviewers: anemet, Gerolf, reames, hfinkel, andrew.w.kaylor
Reviewed By: anemet
Differential Revision: https://reviews.llvm.org/D70900
Added:
Modified:
llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
llvm/test/Transforms/LowerMatrixIntrinsics/bigger-expressions-double.ll
llvm/test/Transforms/LowerMatrixIntrinsics/propagate-backward.ll
Removed:
################################################################################
diff --git a/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp b/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
index c39fdac93b98..afe1b4e7cc78 100644
--- a/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
+++ b/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
@@ -95,20 +95,20 @@ Value *computeColumnAddr(Value *BasePtr, Value *Col, Value *Stride,
unsigned AS = cast<PointerType>(BasePtr->getType())->getAddressSpace();
// Compute the start of the column with index Col as Col * Stride.
- Value *ColumnStart = Builder.CreateMul(Col, Stride);
+ Value *ColumnStart = Builder.CreateMul(Col, Stride, "col.start");
// Get pointer to the start of the selected column. Skip GEP creation,
// if we select column 0.
if (isa<ConstantInt>(ColumnStart) && cast<ConstantInt>(ColumnStart)->isZero())
ColumnStart = BasePtr;
else
- ColumnStart = Builder.CreateGEP(EltType, BasePtr, ColumnStart);
+ ColumnStart = Builder.CreateGEP(EltType, BasePtr, ColumnStart, "col.gep");
// Cast elementwise column start pointer to a pointer to a column
// (EltType x NumRows)*.
Type *ColumnType = VectorType::get(EltType, NumRows);
Type *ColumnPtrType = PointerType::get(ColumnType, AS);
- return Builder.CreatePointerCast(ColumnStart, ColumnPtrType);
+ return Builder.CreatePointerCast(ColumnStart, ColumnPtrType, "col.cast");
}
/// LowerMatrixIntrinsics contains the methods used to lower matrix intrinsics.
@@ -317,7 +317,7 @@ class LowerMatrixIntrinsics {
default:
return false;
}
- return isUniformShape(V) || isa<StoreInst>(V);
+ return isUniformShape(V) || isa<StoreInst>(V) || isa<LoadInst>(V);
}
/// Propagate the shape information of instructions to their users.
@@ -481,6 +481,8 @@ class LowerMatrixIntrinsics {
Value *Op2;
if (auto *BinOp = dyn_cast<BinaryOperator>(&Inst))
Changed |= VisitBinaryOperator(BinOp);
+ if (match(&Inst, m_Load(m_Value(Op1))))
+ Changed |= VisitLoad(&Inst, Op1, Builder);
else if (match(&Inst, m_Store(m_Value(Op1), m_Value(Op2))))
Changed |= VisitStore(&Inst, Op1, Op2, Builder);
}
@@ -495,7 +497,7 @@ class LowerMatrixIntrinsics {
LoadInst *createColumnLoad(Value *ColumnPtr, Type *EltType,
IRBuilder<> Builder) {
unsigned Align = DL.getABITypeAlignment(EltType);
- return Builder.CreateAlignedLoad(ColumnPtr, Align);
+ return Builder.CreateAlignedLoad(ColumnPtr, Align, "col.load");
}
StoreInst *createColumnStore(Value *ColumnValue, Value *ColumnPtr,
@@ -536,17 +538,11 @@ class LowerMatrixIntrinsics {
return true;
}
- /// Lowers llvm.matrix.columnwise.load.
- ///
- /// The intrinsic loads a matrix from memory using a stride between columns.
- void LowerColumnwiseLoad(CallInst *Inst) {
+ void LowerLoad(Instruction *Inst, Value *Ptr, Value *Stride,
+ ShapeInfo Shape) {
IRBuilder<> Builder(Inst);
- Value *Ptr = Inst->getArgOperand(0);
- Value *Stride = Inst->getArgOperand(1);
auto VType = cast<VectorType>(Inst->getType());
Value *EltPtr = createElementPtr(Ptr, VType->getElementType(), Builder);
- ShapeInfo Shape(Inst->getArgOperand(2), Inst->getArgOperand(3));
-
ColumnMatrixTy Result;
// Distance between start of one column and the start of the next
for (unsigned C = 0, E = Shape.NumColumns; C < E; ++C) {
@@ -560,6 +556,16 @@ class LowerMatrixIntrinsics {
finalizeLowering(Inst, Result, Builder);
}
+ /// Lowers llvm.matrix.columnwise.load.
+ ///
+ /// The intrinsic loads a matrix from memory using a stride between columns.
+ void LowerColumnwiseLoad(CallInst *Inst) {
+ Value *Ptr = Inst->getArgOperand(0);
+ Value *Stride = Inst->getArgOperand(1);
+ LowerLoad(Inst, Ptr, Stride,
+ {Inst->getArgOperand(2), Inst->getArgOperand(3)});
+ }
+
void LowerStore(Instruction *Inst, Value *Matrix, Value *Ptr, Value *Stride,
ShapeInfo Shape) {
IRBuilder<> Builder(Inst);
@@ -755,6 +761,16 @@ class LowerMatrixIntrinsics {
finalizeLowering(Inst, Result, Builder);
}
+ /// Lower load instructions, if shape information is available.
+ bool VisitLoad(Instruction *Inst, Value *Ptr, IRBuilder<> &Builder) {
+ auto I = ShapeMap.find(Inst);
+ if (I == ShapeMap.end())
+ return false;
+
+ LowerLoad(Inst, Ptr, Builder.getInt32(I->second.NumRows), I->second);
+ return true;
+ }
+
bool VisitStore(Instruction *Inst, Value *StoredVal, Value *Ptr,
IRBuilder<> &Builder) {
auto I = ShapeMap.find(StoredVal);
diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/bigger-expressions-double.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/bigger-expressions-double.ll
index 246f4e42c2fa..b72ecf5c0e4c 100644
--- a/llvm/test/Transforms/LowerMatrixIntrinsics/bigger-expressions-double.ll
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/bigger-expressions-double.ll
@@ -6,59 +6,63 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-LABEL: @transpose_multiply(
; CHECK-NEXT: entry:
-; Load input matrixes %A and %B.
-
-; CHECK-NEXT: [[A:%.*]] = load <9 x double>, <9 x double>* [[A_PTR:%.*]]
-; CHECK-NEXT: [[B:%.*]] = load <9 x double>, <9 x double>* [[B_PTR:%.*]]
-
-; Extract columns from loaded value %A.
-
-; CHECK-NEXT: [[SPLIT:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
-; CHECK-NEXT: [[SPLIT1:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 3, i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT2:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 6, i32 7, i32 8>
+; Load columns of input matrixes %A and %B.
+
+; CHECK-NEXT: [[TMP0:%.*]] = bitcast <9 x double>* [[A_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST:%.*]] = bitcast double* [[TMP0]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST]], align 8
+; CHECK-NEXT: [[COL_GEP:%.*]] = getelementptr double, double* [[TMP0]], i32 3
+; CHECK-NEXT: [[COL_CAST1:%.*]] = bitcast double* [[COL_GEP]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD2:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST1]], align 8
+; CHECK-NEXT: [[COL_GEP3:%.*]] = getelementptr double, double* [[TMP0]], i32 6
+; CHECK-NEXT: [[COL_CAST4:%.*]] = bitcast double* [[COL_GEP3]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD5:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST4]], align 8
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <9 x double>* [[B_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST6:%.*]] = bitcast double* [[TMP1]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD7:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST6]], align 8
+; CHECK-NEXT: [[COL_GEP8:%.*]] = getelementptr double, double* [[TMP1]], i32 3
+; CHECK-NEXT: [[COL_CAST9:%.*]] = bitcast double* [[COL_GEP8]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD10:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST9]], align 8
+; CHECK-NEXT: [[COL_GEP11:%.*]] = getelementptr double, double* [[TMP1]], i32 6
+; CHECK-NEXT: [[COL_CAST12:%.*]] = bitcast double* [[COL_GEP11]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD13:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST12]], align 8
; Transpose %A.
-; CHECK-NEXT: [[TMP0:%.*]] = extractelement <3 x double> [[SPLIT]], i64 0
+; CHECK-NEXT: [[TMP0:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 0
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <3 x double> undef, double [[TMP0]], i64 0
-; CHECK-NEXT: [[TMP2:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 0
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 0
; CHECK-NEXT: [[TMP3:%.*]] = insertelement <3 x double> [[TMP1]], double [[TMP2]], i64 1
-; CHECK-NEXT: [[TMP4:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 0
; CHECK-NEXT: [[TMP5:%.*]] = insertelement <3 x double> [[TMP3]], double [[TMP4]], i64 2
-; CHECK-NEXT: [[TMP6:%.*]] = extractelement <3 x double> [[SPLIT]], i64 1
+; CHECK-NEXT: [[TMP6:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 1
; CHECK-NEXT: [[TMP7:%.*]] = insertelement <3 x double> undef, double [[TMP6]], i64 0
-; CHECK-NEXT: [[TMP8:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 1
+; CHECK-NEXT: [[TMP8:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 1
; CHECK-NEXT: [[TMP9:%.*]] = insertelement <3 x double> [[TMP7]], double [[TMP8]], i64 1
-; CHECK-NEXT: [[TMP10:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 1
; CHECK-NEXT: [[TMP11:%.*]] = insertelement <3 x double> [[TMP9]], double [[TMP10]], i64 2
-; CHECK-NEXT: [[TMP12:%.*]] = extractelement <3 x double> [[SPLIT]], i64 2
+; CHECK-NEXT: [[TMP12:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 2
; CHECK-NEXT: [[TMP13:%.*]] = insertelement <3 x double> undef, double [[TMP12]], i64 0
-; CHECK-NEXT: [[TMP14:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 2
+; CHECK-NEXT: [[TMP14:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 2
; CHECK-NEXT: [[TMP15:%.*]] = insertelement <3 x double> [[TMP13]], double [[TMP14]], i64 1
-; CHECK-NEXT: [[TMP16:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 2
+; CHECK-NEXT: [[TMP16:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 2
; CHECK-NEXT: [[TMP17:%.*]] = insertelement <3 x double> [[TMP15]], double [[TMP16]], i64 2
-; Extract columns from %B.
-
-; CHECK-NEXT: [[SPLIT3:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
-; CHECK-NEXT: [[SPLIT4:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 3, i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT5:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 6, i32 7, i32 8>
-
; Lower multiply(transpose(%A), %B)
; CHECK-NEXT: [[BLOCK:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP18:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP18:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x double> undef, double [[TMP18]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP19:%.*]] = fmul <1 x double> [[BLOCK]], [[SPLAT_SPLAT]]
; CHECK-NEXT: [[BLOCK6:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP20:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP20:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT7:%.*]] = insertelement <1 x double> undef, double [[TMP20]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT8:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT7]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP21:%.*]] = fmul <1 x double> [[BLOCK6]], [[SPLAT_SPLAT8]]
; CHECK-NEXT: [[TMP22:%.*]] = fadd <1 x double> [[TMP19]], [[TMP21]]
; CHECK-NEXT: [[BLOCK9:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP23:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP23:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT10:%.*]] = insertelement <1 x double> undef, double [[TMP23]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT11:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT10]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP24:%.*]] = fmul <1 x double> [[BLOCK9]], [[SPLAT_SPLAT11]]
@@ -66,18 +70,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP26:%.*]] = shufflevector <1 x double> [[TMP25]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP27:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP26]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK12:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP28:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP28:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT13:%.*]] = insertelement <1 x double> undef, double [[TMP28]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT14:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT13]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP29:%.*]] = fmul <1 x double> [[BLOCK12]], [[SPLAT_SPLAT14]]
; CHECK-NEXT: [[BLOCK15:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP30:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP30:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT16:%.*]] = insertelement <1 x double> undef, double [[TMP30]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT17:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT16]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP31:%.*]] = fmul <1 x double> [[BLOCK15]], [[SPLAT_SPLAT17]]
; CHECK-NEXT: [[TMP32:%.*]] = fadd <1 x double> [[TMP29]], [[TMP31]]
; CHECK-NEXT: [[BLOCK18:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP33:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP33:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT19:%.*]] = insertelement <1 x double> undef, double [[TMP33]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT20:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT19]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP34:%.*]] = fmul <1 x double> [[BLOCK18]], [[SPLAT_SPLAT20]]
@@ -85,18 +89,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP36:%.*]] = shufflevector <1 x double> [[TMP35]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP37:%.*]] = shufflevector <3 x double> [[TMP27]], <3 x double> [[TMP36]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK21:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP38:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP38:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT22:%.*]] = insertelement <1 x double> undef, double [[TMP38]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT23:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT22]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP39:%.*]] = fmul <1 x double> [[BLOCK21]], [[SPLAT_SPLAT23]]
; CHECK-NEXT: [[BLOCK24:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP40:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP40:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT25:%.*]] = insertelement <1 x double> undef, double [[TMP40]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT26:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT25]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP41:%.*]] = fmul <1 x double> [[BLOCK24]], [[SPLAT_SPLAT26]]
; CHECK-NEXT: [[TMP42:%.*]] = fadd <1 x double> [[TMP39]], [[TMP41]]
; CHECK-NEXT: [[BLOCK27:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP43:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP43:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT28:%.*]] = insertelement <1 x double> undef, double [[TMP43]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT29:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT28]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP44:%.*]] = fmul <1 x double> [[BLOCK27]], [[SPLAT_SPLAT29]]
@@ -104,18 +108,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP46:%.*]] = shufflevector <1 x double> [[TMP45]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP47:%.*]] = shufflevector <3 x double> [[TMP37]], <3 x double> [[TMP46]], <3 x i32> <i32 0, i32 1, i32 3>
; CHECK-NEXT: [[BLOCK30:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP48:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP48:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT31:%.*]] = insertelement <1 x double> undef, double [[TMP48]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT32:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT31]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP49:%.*]] = fmul <1 x double> [[BLOCK30]], [[SPLAT_SPLAT32]]
; CHECK-NEXT: [[BLOCK33:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP50:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP50:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT34:%.*]] = insertelement <1 x double> undef, double [[TMP50]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT35:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT34]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP51:%.*]] = fmul <1 x double> [[BLOCK33]], [[SPLAT_SPLAT35]]
; CHECK-NEXT: [[TMP52:%.*]] = fadd <1 x double> [[TMP49]], [[TMP51]]
; CHECK-NEXT: [[BLOCK36:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP53:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP53:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT37:%.*]] = insertelement <1 x double> undef, double [[TMP53]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT38:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT37]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP54:%.*]] = fmul <1 x double> [[BLOCK36]], [[SPLAT_SPLAT38]]
@@ -123,18 +127,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP56:%.*]] = shufflevector <1 x double> [[TMP55]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP57:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP56]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK39:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP58:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP58:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT40:%.*]] = insertelement <1 x double> undef, double [[TMP58]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT41:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT40]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP59:%.*]] = fmul <1 x double> [[BLOCK39]], [[SPLAT_SPLAT41]]
; CHECK-NEXT: [[BLOCK42:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP60:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP60:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT43:%.*]] = insertelement <1 x double> undef, double [[TMP60]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT44:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT43]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP61:%.*]] = fmul <1 x double> [[BLOCK42]], [[SPLAT_SPLAT44]]
; CHECK-NEXT: [[TMP62:%.*]] = fadd <1 x double> [[TMP59]], [[TMP61]]
; CHECK-NEXT: [[BLOCK45:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP63:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP63:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT46:%.*]] = insertelement <1 x double> undef, double [[TMP63]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT47:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT46]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP64:%.*]] = fmul <1 x double> [[BLOCK45]], [[SPLAT_SPLAT47]]
@@ -142,18 +146,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP66:%.*]] = shufflevector <1 x double> [[TMP65]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP67:%.*]] = shufflevector <3 x double> [[TMP57]], <3 x double> [[TMP66]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK48:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP68:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP68:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT49:%.*]] = insertelement <1 x double> undef, double [[TMP68]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT50:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT49]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP69:%.*]] = fmul <1 x double> [[BLOCK48]], [[SPLAT_SPLAT50]]
; CHECK-NEXT: [[BLOCK51:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP70:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP70:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT52:%.*]] = insertelement <1 x double> undef, double [[TMP70]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT53:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT52]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP71:%.*]] = fmul <1 x double> [[BLOCK51]], [[SPLAT_SPLAT53]]
; CHECK-NEXT: [[TMP72:%.*]] = fadd <1 x double> [[TMP69]], [[TMP71]]
; CHECK-NEXT: [[BLOCK54:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP73:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP73:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT55:%.*]] = insertelement <1 x double> undef, double [[TMP73]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT56:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT55]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP74:%.*]] = fmul <1 x double> [[BLOCK54]], [[SPLAT_SPLAT56]]
@@ -161,18 +165,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP76:%.*]] = shufflevector <1 x double> [[TMP75]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP77:%.*]] = shufflevector <3 x double> [[TMP67]], <3 x double> [[TMP76]], <3 x i32> <i32 0, i32 1, i32 3>
; CHECK-NEXT: [[BLOCK57:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP78:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP78:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT58:%.*]] = insertelement <1 x double> undef, double [[TMP78]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT59:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT58]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP79:%.*]] = fmul <1 x double> [[BLOCK57]], [[SPLAT_SPLAT59]]
; CHECK-NEXT: [[BLOCK60:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP80:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP80:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT61:%.*]] = insertelement <1 x double> undef, double [[TMP80]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT62:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT61]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP81:%.*]] = fmul <1 x double> [[BLOCK60]], [[SPLAT_SPLAT62]]
; CHECK-NEXT: [[TMP82:%.*]] = fadd <1 x double> [[TMP79]], [[TMP81]]
; CHECK-NEXT: [[BLOCK63:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP83:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP83:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT64:%.*]] = insertelement <1 x double> undef, double [[TMP83]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT65:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT64]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP84:%.*]] = fmul <1 x double> [[BLOCK63]], [[SPLAT_SPLAT65]]
@@ -180,18 +184,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP86:%.*]] = shufflevector <1 x double> [[TMP85]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP87:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP86]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK66:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP88:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP88:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT67:%.*]] = insertelement <1 x double> undef, double [[TMP88]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT68:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT67]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP89:%.*]] = fmul <1 x double> [[BLOCK66]], [[SPLAT_SPLAT68]]
; CHECK-NEXT: [[BLOCK69:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP90:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP90:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT70:%.*]] = insertelement <1 x double> undef, double [[TMP90]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT71:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT70]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP91:%.*]] = fmul <1 x double> [[BLOCK69]], [[SPLAT_SPLAT71]]
; CHECK-NEXT: [[TMP92:%.*]] = fadd <1 x double> [[TMP89]], [[TMP91]]
; CHECK-NEXT: [[BLOCK72:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP93:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP93:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT73:%.*]] = insertelement <1 x double> undef, double [[TMP93]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT74:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT73]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP94:%.*]] = fmul <1 x double> [[BLOCK72]], [[SPLAT_SPLAT74]]
@@ -199,18 +203,18 @@ define void @transpose_multiply(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr, <9 x
; CHECK-NEXT: [[TMP96:%.*]] = shufflevector <1 x double> [[TMP95]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP97:%.*]] = shufflevector <3 x double> [[TMP87]], <3 x double> [[TMP96]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK75:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP98:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP98:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT76:%.*]] = insertelement <1 x double> undef, double [[TMP98]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT77:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT76]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP99:%.*]] = fmul <1 x double> [[BLOCK75]], [[SPLAT_SPLAT77]]
; CHECK-NEXT: [[BLOCK78:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP100:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP100:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT79:%.*]] = insertelement <1 x double> undef, double [[TMP100]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT80:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT79]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP101:%.*]] = fmul <1 x double> [[BLOCK78]], [[SPLAT_SPLAT80]]
; CHECK-NEXT: [[TMP102:%.*]] = fadd <1 x double> [[TMP99]], [[TMP101]]
; CHECK-NEXT: [[BLOCK81:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP103:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP103:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT82:%.*]] = insertelement <1 x double> undef, double [[TMP103]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT83:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT82]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP104:%.*]] = fmul <1 x double> [[BLOCK81]], [[SPLAT_SPLAT83]]
@@ -248,59 +252,61 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-LABEL: @transpose_multiply_add(
; CHECK-NEXT: entry:
-; Load input matrixes %A and %B.
-
-; CHECK-NEXT: [[A:%.*]] = load <9 x double>, <9 x double>* [[A_PTR:%.*]]
-; CHECK-NEXT: [[B:%.*]] = load <9 x double>, <9 x double>* [[B_PTR:%.*]]
-
-; Extract columns from loaded value %A.
-
-; CHECK-NEXT: [[SPLIT:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
-; CHECK-NEXT: [[SPLIT1:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 3, i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT2:%.*]] = shufflevector <9 x double> [[A]], <9 x double> undef, <3 x i32> <i32 6, i32 7, i32 8>
+; CHECK-NEXT: [[TMP0:%.*]] = bitcast <9 x double>* [[A_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST:%.*]] = bitcast double* [[TMP0]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST]], align 8
+; CHECK-NEXT: [[COL_GEP:%.*]] = getelementptr double, double* [[TMP0]], i32 3
+; CHECK-NEXT: [[COL_CAST1:%.*]] = bitcast double* [[COL_GEP]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD2:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST1]], align 8
+; CHECK-NEXT: [[COL_GEP3:%.*]] = getelementptr double, double* [[TMP0]], i32 6
+; CHECK-NEXT: [[COL_CAST4:%.*]] = bitcast double* [[COL_GEP3]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD5:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST4]], align 8
+; CHECK-NEXT: [[TMP1:%.*]] = bitcast <9 x double>* [[B_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST6:%.*]] = bitcast double* [[TMP1]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD7:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST6]], align 8
+; CHECK-NEXT: [[COL_GEP8:%.*]] = getelementptr double, double* [[TMP1]], i32 3
+; CHECK-NEXT: [[COL_CAST9:%.*]] = bitcast double* [[COL_GEP8]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD10:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST9]], align 8
+; CHECK-NEXT: [[COL_GEP11:%.*]] = getelementptr double, double* [[TMP1]], i32 6
+; CHECK-NEXT: [[COL_CAST12:%.*]] = bitcast double* [[COL_GEP11]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD13:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST12]], align 8
; Transpose %A.
-; CHECK-NEXT: [[TMP0:%.*]] = extractelement <3 x double> [[SPLIT]], i64 0
+; CHECK-NEXT: [[TMP0:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 0
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <3 x double> undef, double [[TMP0]], i64 0
-; CHECK-NEXT: [[TMP2:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 0
+; CHECK-NEXT: [[TMP2:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 0
; CHECK-NEXT: [[TMP3:%.*]] = insertelement <3 x double> [[TMP1]], double [[TMP2]], i64 1
-; CHECK-NEXT: [[TMP4:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 0
; CHECK-NEXT: [[TMP5:%.*]] = insertelement <3 x double> [[TMP3]], double [[TMP4]], i64 2
-; CHECK-NEXT: [[TMP6:%.*]] = extractelement <3 x double> [[SPLIT]], i64 1
+; CHECK-NEXT: [[TMP6:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 1
; CHECK-NEXT: [[TMP7:%.*]] = insertelement <3 x double> undef, double [[TMP6]], i64 0
-; CHECK-NEXT: [[TMP8:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 1
+; CHECK-NEXT: [[TMP8:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 1
; CHECK-NEXT: [[TMP9:%.*]] = insertelement <3 x double> [[TMP7]], double [[TMP8]], i64 1
-; CHECK-NEXT: [[TMP10:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 1
; CHECK-NEXT: [[TMP11:%.*]] = insertelement <3 x double> [[TMP9]], double [[TMP10]], i64 2
-; CHECK-NEXT: [[TMP12:%.*]] = extractelement <3 x double> [[SPLIT]], i64 2
+; CHECK-NEXT: [[TMP12:%.*]] = extractelement <3 x double> [[COL_LOAD]], i64 2
; CHECK-NEXT: [[TMP13:%.*]] = insertelement <3 x double> undef, double [[TMP12]], i64 0
-; CHECK-NEXT: [[TMP14:%.*]] = extractelement <3 x double> [[SPLIT1]], i64 2
+; CHECK-NEXT: [[TMP14:%.*]] = extractelement <3 x double> [[COL_LOAD2]], i64 2
; CHECK-NEXT: [[TMP15:%.*]] = insertelement <3 x double> [[TMP13]], double [[TMP14]], i64 1
-; CHECK-NEXT: [[TMP16:%.*]] = extractelement <3 x double> [[SPLIT2]], i64 2
+; CHECK-NEXT: [[TMP16:%.*]] = extractelement <3 x double> [[COL_LOAD5]], i64 2
; CHECK-NEXT: [[TMP17:%.*]] = insertelement <3 x double> [[TMP15]], double [[TMP16]], i64 2
-; Extract columns from %B.
-
-; CHECK-NEXT: [[SPLIT3:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
-; CHECK-NEXT: [[SPLIT4:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 3, i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT5:%.*]] = shufflevector <9 x double> [[B]], <9 x double> undef, <3 x i32> <i32 6, i32 7, i32 8>
-
; Lower multiply(transpose(%A), %B)
; CHECK-NEXT: [[BLOCK:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP18:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP18:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x double> undef, double [[TMP18]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP19:%.*]] = fmul <1 x double> [[BLOCK]], [[SPLAT_SPLAT]]
; CHECK-NEXT: [[BLOCK6:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP20:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP20:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT7:%.*]] = insertelement <1 x double> undef, double [[TMP20]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT8:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT7]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP21:%.*]] = fmul <1 x double> [[BLOCK6]], [[SPLAT_SPLAT8]]
; CHECK-NEXT: [[TMP22:%.*]] = fadd <1 x double> [[TMP19]], [[TMP21]]
; CHECK-NEXT: [[BLOCK9:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP23:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP23:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT10:%.*]] = insertelement <1 x double> undef, double [[TMP23]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT11:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT10]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP24:%.*]] = fmul <1 x double> [[BLOCK9]], [[SPLAT_SPLAT11]]
@@ -308,18 +314,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP26:%.*]] = shufflevector <1 x double> [[TMP25]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP27:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP26]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK12:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP28:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP28:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT13:%.*]] = insertelement <1 x double> undef, double [[TMP28]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT14:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT13]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP29:%.*]] = fmul <1 x double> [[BLOCK12]], [[SPLAT_SPLAT14]]
; CHECK-NEXT: [[BLOCK15:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP30:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP30:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT16:%.*]] = insertelement <1 x double> undef, double [[TMP30]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT17:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT16]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP31:%.*]] = fmul <1 x double> [[BLOCK15]], [[SPLAT_SPLAT17]]
; CHECK-NEXT: [[TMP32:%.*]] = fadd <1 x double> [[TMP29]], [[TMP31]]
; CHECK-NEXT: [[BLOCK18:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP33:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP33:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT19:%.*]] = insertelement <1 x double> undef, double [[TMP33]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT20:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT19]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP34:%.*]] = fmul <1 x double> [[BLOCK18]], [[SPLAT_SPLAT20]]
@@ -327,18 +333,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP36:%.*]] = shufflevector <1 x double> [[TMP35]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP37:%.*]] = shufflevector <3 x double> [[TMP27]], <3 x double> [[TMP36]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK21:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP38:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT: [[TMP38:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT22:%.*]] = insertelement <1 x double> undef, double [[TMP38]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT23:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT22]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP39:%.*]] = fmul <1 x double> [[BLOCK21]], [[SPLAT_SPLAT23]]
; CHECK-NEXT: [[BLOCK24:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP40:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT: [[TMP40:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT25:%.*]] = insertelement <1 x double> undef, double [[TMP40]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT26:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT25]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP41:%.*]] = fmul <1 x double> [[BLOCK24]], [[SPLAT_SPLAT26]]
; CHECK-NEXT: [[TMP42:%.*]] = fadd <1 x double> [[TMP39]], [[TMP41]]
; CHECK-NEXT: [[BLOCK27:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP43:%.*]] = extractelement <3 x double> [[SPLIT3]], i64 2
+; CHECK-NEXT: [[TMP43:%.*]] = extractelement <3 x double> [[COL_LOAD7]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT28:%.*]] = insertelement <1 x double> undef, double [[TMP43]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT29:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT28]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP44:%.*]] = fmul <1 x double> [[BLOCK27]], [[SPLAT_SPLAT29]]
@@ -346,18 +352,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP46:%.*]] = shufflevector <1 x double> [[TMP45]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP47:%.*]] = shufflevector <3 x double> [[TMP37]], <3 x double> [[TMP46]], <3 x i32> <i32 0, i32 1, i32 3>
; CHECK-NEXT: [[BLOCK30:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP48:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP48:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT31:%.*]] = insertelement <1 x double> undef, double [[TMP48]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT32:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT31]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP49:%.*]] = fmul <1 x double> [[BLOCK30]], [[SPLAT_SPLAT32]]
; CHECK-NEXT: [[BLOCK33:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP50:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP50:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT34:%.*]] = insertelement <1 x double> undef, double [[TMP50]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT35:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT34]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP51:%.*]] = fmul <1 x double> [[BLOCK33]], [[SPLAT_SPLAT35]]
; CHECK-NEXT: [[TMP52:%.*]] = fadd <1 x double> [[TMP49]], [[TMP51]]
; CHECK-NEXT: [[BLOCK36:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP53:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP53:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT37:%.*]] = insertelement <1 x double> undef, double [[TMP53]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT38:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT37]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP54:%.*]] = fmul <1 x double> [[BLOCK36]], [[SPLAT_SPLAT38]]
@@ -365,18 +371,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP56:%.*]] = shufflevector <1 x double> [[TMP55]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP57:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP56]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK39:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP58:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP58:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT40:%.*]] = insertelement <1 x double> undef, double [[TMP58]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT41:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT40]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP59:%.*]] = fmul <1 x double> [[BLOCK39]], [[SPLAT_SPLAT41]]
; CHECK-NEXT: [[BLOCK42:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP60:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP60:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT43:%.*]] = insertelement <1 x double> undef, double [[TMP60]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT44:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT43]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP61:%.*]] = fmul <1 x double> [[BLOCK42]], [[SPLAT_SPLAT44]]
; CHECK-NEXT: [[TMP62:%.*]] = fadd <1 x double> [[TMP59]], [[TMP61]]
; CHECK-NEXT: [[BLOCK45:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP63:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP63:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT46:%.*]] = insertelement <1 x double> undef, double [[TMP63]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT47:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT46]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP64:%.*]] = fmul <1 x double> [[BLOCK45]], [[SPLAT_SPLAT47]]
@@ -384,18 +390,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP66:%.*]] = shufflevector <1 x double> [[TMP65]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP67:%.*]] = shufflevector <3 x double> [[TMP57]], <3 x double> [[TMP66]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK48:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP68:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 0
+; CHECK-NEXT: [[TMP68:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT49:%.*]] = insertelement <1 x double> undef, double [[TMP68]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT50:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT49]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP69:%.*]] = fmul <1 x double> [[BLOCK48]], [[SPLAT_SPLAT50]]
; CHECK-NEXT: [[BLOCK51:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP70:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 1
+; CHECK-NEXT: [[TMP70:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT52:%.*]] = insertelement <1 x double> undef, double [[TMP70]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT53:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT52]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP71:%.*]] = fmul <1 x double> [[BLOCK51]], [[SPLAT_SPLAT53]]
; CHECK-NEXT: [[TMP72:%.*]] = fadd <1 x double> [[TMP69]], [[TMP71]]
; CHECK-NEXT: [[BLOCK54:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP73:%.*]] = extractelement <3 x double> [[SPLIT4]], i64 2
+; CHECK-NEXT: [[TMP73:%.*]] = extractelement <3 x double> [[COL_LOAD10]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT55:%.*]] = insertelement <1 x double> undef, double [[TMP73]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT56:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT55]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP74:%.*]] = fmul <1 x double> [[BLOCK54]], [[SPLAT_SPLAT56]]
@@ -403,18 +409,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP76:%.*]] = shufflevector <1 x double> [[TMP75]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP77:%.*]] = shufflevector <3 x double> [[TMP67]], <3 x double> [[TMP76]], <3 x i32> <i32 0, i32 1, i32 3>
; CHECK-NEXT: [[BLOCK57:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP78:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP78:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT58:%.*]] = insertelement <1 x double> undef, double [[TMP78]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT59:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT58]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP79:%.*]] = fmul <1 x double> [[BLOCK57]], [[SPLAT_SPLAT59]]
; CHECK-NEXT: [[BLOCK60:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP80:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP80:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT61:%.*]] = insertelement <1 x double> undef, double [[TMP80]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT62:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT61]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP81:%.*]] = fmul <1 x double> [[BLOCK60]], [[SPLAT_SPLAT62]]
; CHECK-NEXT: [[TMP82:%.*]] = fadd <1 x double> [[TMP79]], [[TMP81]]
; CHECK-NEXT: [[BLOCK63:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> zeroinitializer
-; CHECK-NEXT: [[TMP83:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP83:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT64:%.*]] = insertelement <1 x double> undef, double [[TMP83]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT65:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT64]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP84:%.*]] = fmul <1 x double> [[BLOCK63]], [[SPLAT_SPLAT65]]
@@ -422,18 +428,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP86:%.*]] = shufflevector <1 x double> [[TMP85]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP87:%.*]] = shufflevector <3 x double> undef, <3 x double> [[TMP86]], <3 x i32> <i32 3, i32 1, i32 2>
; CHECK-NEXT: [[BLOCK66:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP88:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP88:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT67:%.*]] = insertelement <1 x double> undef, double [[TMP88]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT68:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT67]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP89:%.*]] = fmul <1 x double> [[BLOCK66]], [[SPLAT_SPLAT68]]
; CHECK-NEXT: [[BLOCK69:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP90:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP90:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT70:%.*]] = insertelement <1 x double> undef, double [[TMP90]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT71:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT70]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP91:%.*]] = fmul <1 x double> [[BLOCK69]], [[SPLAT_SPLAT71]]
; CHECK-NEXT: [[TMP92:%.*]] = fadd <1 x double> [[TMP89]], [[TMP91]]
; CHECK-NEXT: [[BLOCK72:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 1>
-; CHECK-NEXT: [[TMP93:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP93:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT73:%.*]] = insertelement <1 x double> undef, double [[TMP93]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT74:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT73]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP94:%.*]] = fmul <1 x double> [[BLOCK72]], [[SPLAT_SPLAT74]]
@@ -441,18 +447,18 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; CHECK-NEXT: [[TMP96:%.*]] = shufflevector <1 x double> [[TMP95]], <1 x double> undef, <3 x i32> <i32 0, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP97:%.*]] = shufflevector <3 x double> [[TMP87]], <3 x double> [[TMP96]], <3 x i32> <i32 0, i32 3, i32 2>
; CHECK-NEXT: [[BLOCK75:%.*]] = shufflevector <3 x double> [[TMP5]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP98:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 0
+; CHECK-NEXT: [[TMP98:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 0
; CHECK-NEXT: [[SPLAT_SPLATINSERT76:%.*]] = insertelement <1 x double> undef, double [[TMP98]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT77:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT76]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP99:%.*]] = fmul <1 x double> [[BLOCK75]], [[SPLAT_SPLAT77]]
; CHECK-NEXT: [[BLOCK78:%.*]] = shufflevector <3 x double> [[TMP11]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP100:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 1
+; CHECK-NEXT: [[TMP100:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 1
; CHECK-NEXT: [[SPLAT_SPLATINSERT79:%.*]] = insertelement <1 x double> undef, double [[TMP100]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT80:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT79]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP101:%.*]] = fmul <1 x double> [[BLOCK78]], [[SPLAT_SPLAT80]]
; CHECK-NEXT: [[TMP102:%.*]] = fadd <1 x double> [[TMP99]], [[TMP101]]
; CHECK-NEXT: [[BLOCK81:%.*]] = shufflevector <3 x double> [[TMP17]], <3 x double> undef, <1 x i32> <i32 2>
-; CHECK-NEXT: [[TMP103:%.*]] = extractelement <3 x double> [[SPLIT5]], i64 2
+; CHECK-NEXT: [[TMP103:%.*]] = extractelement <3 x double> [[COL_LOAD13]], i64 2
; CHECK-NEXT: [[SPLAT_SPLATINSERT82:%.*]] = insertelement <1 x double> undef, double [[TMP103]], i32 0
; CHECK-NEXT: [[SPLAT_SPLAT83:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT82]], <1 x double> undef, <1 x i32> zeroinitializer
; CHECK-NEXT: [[TMP104:%.*]] = fmul <1 x double> [[BLOCK81]], [[SPLAT_SPLAT83]]
@@ -465,19 +471,21 @@ define void @transpose_multiply_add(<9 x double>* %A.Ptr, <9 x double>* %B.Ptr,
; Load %C.
-; CHECK-NEXT: [[C:%.*]] = load <9 x double>, <9 x double>* [[C_PTR:%.*]]
-
-; Extract columns from %C.
-
-; CHECK-NEXT: [[SPLIT84:%.*]] = shufflevector <9 x double> [[C]], <9 x double> undef, <3 x i32> <i32 0, i32 1, i32 2>
-; CHECK-NEXT: [[SPLIT85:%.*]] = shufflevector <9 x double> [[C]], <9 x double> undef, <3 x i32> <i32 3, i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT86:%.*]] = shufflevector <9 x double> [[C]], <9 x double> undef, <3 x i32> <i32 6, i32 7, i32 8>
+; CHECK-NEXT: [[TMP110:%.*]] = bitcast <9 x double>* [[C_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST92:%.*]] = bitcast double* [[TMP110]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD93:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST92]], align 8
+; CHECK-NEXT: [[COL_GEP94:%.*]] = getelementptr double, double* [[TMP110]], i32 3
+; CHECK-NEXT: [[COL_CAST95:%.*]] = bitcast double* [[COL_GEP94]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD96:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST95]], align 8
+; CHECK-NEXT: [[COL_GEP97:%.*]] = getelementptr double, double* [[TMP110]], i32 6
+; CHECK-NEXT: [[COL_CAST98:%.*]] = bitcast double* [[COL_GEP97]] to <3 x double>*
+; CHECK-NEXT: [[COL_LOAD99:%.*]] = load <3 x double>, <3 x double>* [[COL_CAST98]], align 8
; Add column vectors.
-; CHECK-NEXT: [[TMP108:%.*]] = fadd <3 x double> [[SPLIT84]], [[TMP47]]
-; CHECK-NEXT: [[TMP109:%.*]] = fadd <3 x double> [[SPLIT85]], [[TMP77]]
-; CHECK-NEXT: [[TMP110:%.*]] = fadd <3 x double> [[SPLIT86]], [[TMP107]]
+; CHECK-NEXT: [[TMP108:%.*]] = fadd <3 x double> [[COL_LOAD93]], [[TMP47]]
+; CHECK-NEXT: [[TMP109:%.*]] = fadd <3 x double> [[COL_LOAD96]], [[TMP77]]
+; CHECK-NEXT: [[TMP110:%.*]] = fadd <3 x double> [[COL_LOAD99]], [[TMP107]]
; Store result columns.
diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/propagate-backward.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/propagate-backward.ll
index f4d7c0e5120e..89ca79649b87 100644
--- a/llvm/test/Transforms/LowerMatrixIntrinsics/propagate-backward.ll
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/propagate-backward.ll
@@ -45,19 +45,28 @@ entry:
define <8 x double> @load_fadd_transpose(<8 x double>* %A.Ptr, <8 x double> %b) {
; CHECK-LABEL: @load_fadd_transpose(
; CHECK-NEXT: entry:
-; CHECK-NEXT: [[A:%.*]] = load <8 x double>, <8 x double>* [[A_PTR:%.*]]
-; CHECK-NEXT: [[SPLIT:%.*]] = shufflevector <8 x double> [[A]], <8 x double> undef, <2 x i32> <i32 0, i32 1>
-; CHECK-NEXT: [[SPLIT1:%.*]] = shufflevector <8 x double> [[A]], <8 x double> undef, <2 x i32> <i32 2, i32 3>
-; CHECK-NEXT: [[SPLIT2:%.*]] = shufflevector <8 x double> [[A]], <8 x double> undef, <2 x i32> <i32 4, i32 5>
-; CHECK-NEXT: [[SPLIT3:%.*]] = shufflevector <8 x double> [[A]], <8 x double> undef, <2 x i32> <i32 6, i32 7>
+
+; CHECK-NEXT: [[TMP0:%.*]] = bitcast <8 x double>* [[A_PTR:%.*]] to double*
+; CHECK-NEXT: [[COL_CAST:%.*]] = bitcast double* [[TMP0]] to <2 x double>*
+; CHECK-NEXT: [[COL_LOAD:%.*]] = load <2 x double>, <2 x double>* [[COL_CAST]], align 8
+; CHECK-NEXT: [[COL_GEP:%.*]] = getelementptr double, double* [[TMP0]], i32 2
+; CHECK-NEXT: [[COL_CAST1:%.*]] = bitcast double* [[COL_GEP]] to <2 x double>*
+; CHECK-NEXT: [[COL_LOAD2:%.*]] = load <2 x double>, <2 x double>* [[COL_CAST1]], align 8
+; CHECK-NEXT: [[COL_GEP3:%.*]] = getelementptr double, double* [[TMP0]], i32 4
+; CHECK-NEXT: [[COL_CAST4:%.*]] = bitcast double* [[COL_GEP3]] to <2 x double>*
+; CHECK-NEXT: [[COL_LOAD5:%.*]] = load <2 x double>, <2 x double>* [[COL_CAST4]], align 8
+; CHECK-NEXT: [[COL_GEP6:%.*]] = getelementptr double, double* [[TMP0]], i32 6
+; CHECK-NEXT: [[COL_CAST7:%.*]] = bitcast double* [[COL_GEP6]] to <2 x double>*
+; CHECK-NEXT: [[COL_LOAD8:%.*]] = load <2 x double>, <2 x double>* [[COL_CAST7]], align 8
; CHECK-NEXT: [[SPLIT4:%.*]] = shufflevector <8 x double> [[B:%.*]], <8 x double> undef, <2 x i32> <i32 0, i32 1>
; CHECK-NEXT: [[SPLIT5:%.*]] = shufflevector <8 x double> [[B]], <8 x double> undef, <2 x i32> <i32 2, i32 3>
; CHECK-NEXT: [[SPLIT6:%.*]] = shufflevector <8 x double> [[B]], <8 x double> undef, <2 x i32> <i32 4, i32 5>
; CHECK-NEXT: [[SPLIT7:%.*]] = shufflevector <8 x double> [[B]], <8 x double> undef, <2 x i32> <i32 6, i32 7>
-; CHECK-NEXT: [[TMP0:%.*]] = fadd <2 x double> [[SPLIT]], [[SPLIT4]]
-; CHECK-NEXT: [[TMP1:%.*]] = fadd <2 x double> [[SPLIT1]], [[SPLIT5]]
-; CHECK-NEXT: [[TMP2:%.*]] = fadd <2 x double> [[SPLIT2]], [[SPLIT6]]
-; CHECK-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[SPLIT3]], [[SPLIT7]]
+
+; CHECK-NEXT: [[TMP0:%.*]] = fadd <2 x double> [[COL_LOAD]], [[SPLIT4]]
+; CHECK-NEXT: [[TMP1:%.*]] = fadd <2 x double> [[COL_LOAD2]], [[SPLIT5]]
+; CHECK-NEXT: [[TMP2:%.*]] = fadd <2 x double> [[COL_LOAD5]], [[SPLIT6]]
+; CHECK-NEXT: [[TMP3:%.*]] = fadd <2 x double> [[COL_LOAD8]], [[SPLIT7]]
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x double> [[TMP0]], i64 0
; CHECK-NEXT: [[TMP5:%.*]] = insertelement <4 x double> undef, double [[TMP4]], i64 0
; CHECK-NEXT: [[TMP6:%.*]] = extractelement <2 x double> [[TMP1]], i64 0
More information about the llvm-commits
mailing list