[Matrix] Use fmuladd for matrix.multiply if allowed.

If the matrix.multiply calls have the contract fast math flag, we can
use fmuladd. This als adds a command line option to force fmuladd
generation. We can retire this option once there is a clang-level
option.

Reviewers: anemet, Gerolf, hfinkel, andrew.w.kaylor

Reviewed By: anemet

Differential Revision: https://reviews.llvm.org/D70951

diff --git a/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp b/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
index b318800..d03b557 100644 (file)
--- a/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
+++ b/llvm/lib/Transforms/Scalar/LowerMatrixIntrinsics.cpp
@@ -43,6 +43,11 @@ using namespace PatternMatch;
 static cl::opt<bool> EnableShapePropagation("matrix-propagate-shape",
                                             cl::init(true));
 
+static cl::opt<bool> AllowContractEnabled(
+    "matrix-allow-contract", cl::init(false), cl::Hidden,
+    cl::desc("Allow the use of FMAs if available and profitable. This may "
+             "result in different results, due to less rounding error."));
+
 namespace {
 
 // Given an element poitner \p BasePtr to the start of a (sub) matrix, compute
@@ -536,12 +541,25 @@ public:
   }
 
   Value *createMulAdd(Value *Sum, Value *A, Value *B, bool UseFPOp,
-                      IRBuilder<> &Builder) {
-    Value *Mul = UseFPOp ? Builder.CreateFMul(A, B) : Builder.CreateMul(A, B);
+                      IRBuilder<> &Builder, bool AllowContraction) {
+
     if (!Sum)
-      return Mul;
+      return UseFPOp ? Builder.CreateFMul(A, B) : Builder.CreateMul(A, B);
+
+    if (UseFPOp) {
+      if (AllowContraction) {
+        // Use fmuladd for floating point operations and let the backend decide
+        // if that's profitable.
+        Value *FMulAdd = Intrinsic::getDeclaration(
+            Func.getParent(), Intrinsic::fmuladd, A->getType());
+        return Builder.CreateCall(FMulAdd, {A, B, Sum});
+      }
+      Value *Mul = Builder.CreateFMul(A, B);
+      return Builder.CreateFAdd(Sum, Mul);
+    }
 
-    return UseFPOp ? Builder.CreateFAdd(Sum, Mul) : Builder.CreateAdd(Sum, Mul);
+    Value *Mul = Builder.CreateMul(A, B);
+    return Builder.CreateAdd(Sum, Mul);
   }
 
   /// Cache \p Matrix as result of \p Inst and update the uses of \p Inst. For
@@ -591,6 +609,8 @@ public:
                                      EltType->getPrimitiveSizeInBits(),
                                  uint64_t(1));
 
+    bool AllowContract = AllowContractEnabled || (isa<FPMathOperator>(MatMul) &&
+                                                  MatMul->hasAllowContract());
     // Multiply columns from the first operand with scalars from the second
     // operand.  Then move along the K axes and accumulate the columns.  With
     // this the adds can be vectorized without reassociation.
@@ -607,7 +627,7 @@ public:
           Value *RH = Builder.CreateExtractElement(Rhs.getColumn(J), K);
           Value *Splat = Builder.CreateVectorSplat(BlockSize, RH, "splat");
           Sum = createMulAdd(Sum, L, Splat, EltType->isFloatingPointTy(),
-                             Builder);
+                             Builder, AllowContract);
         }
         Result.setColumn(J, insertVector(Result.getColumn(J), I, Sum, Builder));
       }

diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction-fmf.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction-fmf.ll
new file mode 100644 (file)
index 0000000..1b1e8dc
--- /dev/null
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction-fmf.ll
@@ -0,0 +1,69 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -lower-matrix-intrinsics -S < %s | FileCheck %s
+; RUN: opt -passes='lower-matrix-intrinsics' -S < %s | FileCheck %s
+
+
+define <4 x double> @multiply_2x2(<4 x double> %a, <4 x double> %b) {
+; CHECK-LABEL: @multiply_2x2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SPLIT:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT1:%.*]] = shufflevector <4 x double> [[A]], <4 x double> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[SPLIT2:%.*]] = shufflevector <4 x double> [[B:%.*]], <4 x double> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT3:%.*]] = shufflevector <4 x double> [[B]], <4 x double> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[BLOCK:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP0:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x double> undef, double [[TMP0]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP1:%.*]] = fmul <1 x double> [[BLOCK]], [[SPLAT_SPLAT]]
+; CHECK-NEXT:    [[BLOCK4:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT5:%.*]] = insertelement <1 x double> undef, double [[TMP2]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT6:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT5]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP3:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK4]], <1 x double> [[SPLAT_SPLAT6]], <1 x double> [[TMP1]])
+; CHECK-NEXT:    [[TMP4:%.*]] = shufflevector <1 x double> [[TMP3]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP5:%.*]] = shufflevector <2 x double> undef, <2 x double> [[TMP4]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK7:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT8:%.*]] = insertelement <1 x double> undef, double [[TMP6]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT9:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT8]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP7:%.*]] = fmul <1 x double> [[BLOCK7]], [[SPLAT_SPLAT9]]
+; CHECK-NEXT:    [[BLOCK10:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT11:%.*]] = insertelement <1 x double> undef, double [[TMP8]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT12:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT11]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP9:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK10]], <1 x double> [[SPLAT_SPLAT12]], <1 x double> [[TMP7]])
+; CHECK-NEXT:    [[TMP10:%.*]] = shufflevector <1 x double> [[TMP9]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP11:%.*]] = shufflevector <2 x double> [[TMP5]], <2 x double> [[TMP10]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[BLOCK13:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP12:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT14:%.*]] = insertelement <1 x double> undef, double [[TMP12]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT15:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT14]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP13:%.*]] = fmul <1 x double> [[BLOCK13]], [[SPLAT_SPLAT15]]
+; CHECK-NEXT:    [[BLOCK16:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP14:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT17:%.*]] = insertelement <1 x double> undef, double [[TMP14]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT18:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT17]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP15:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK16]], <1 x double> [[SPLAT_SPLAT18]], <1 x double> [[TMP13]])
+; CHECK-NEXT:    [[TMP16:%.*]] = shufflevector <1 x double> [[TMP15]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP17:%.*]] = shufflevector <2 x double> undef, <2 x double> [[TMP16]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK19:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT20:%.*]] = insertelement <1 x double> undef, double [[TMP18]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT21:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT20]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP19:%.*]] = fmul <1 x double> [[BLOCK19]], [[SPLAT_SPLAT21]]
+; CHECK-NEXT:    [[BLOCK22:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP20:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT23:%.*]] = insertelement <1 x double> undef, double [[TMP20]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT24:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT23]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP21:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK22]], <1 x double> [[SPLAT_SPLAT24]], <1 x double> [[TMP19]])
+; CHECK-NEXT:    [[TMP22:%.*]] = shufflevector <1 x double> [[TMP21]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP23:%.*]] = shufflevector <2 x double> [[TMP17]], <2 x double> [[TMP22]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[TMP24:%.*]] = shufflevector <2 x double> [[TMP11]], <2 x double> [[TMP23]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT:    ret <4 x double> [[TMP24]]
+;
+entry:
+  %c = call contract <4 x double> @llvm.matrix.multiply.v4f64.v4f64.v4f64(<4 x double> %a, <4 x double> %b, i32 2, i32 2, i32 2)
+  ret <4 x double> %c
+}
+
+declare <4 x double> @llvm.matrix.multiply.v4f64.v4f64.v4f64(<4 x double>, <4 x double>, i32, i32, i32)

diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction.ll
new file mode 100644 (file)
index 0000000..6245996
--- /dev/null
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-double-contraction.ll
@@ -0,0 +1,69 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -lower-matrix-intrinsics -matrix-allow-contract -S < %s | FileCheck %s
+; RUN: opt -passes='lower-matrix-intrinsics' -matrix-allow-contract -S < %s | FileCheck %s
+
+
+define <4 x double> @multiply_2x2(<4 x double> %a, <4 x double> %b) {
+; CHECK-LABEL: @multiply_2x2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SPLIT:%.*]] = shufflevector <4 x double> [[A:%.*]], <4 x double> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT1:%.*]] = shufflevector <4 x double> [[A]], <4 x double> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[SPLIT2:%.*]] = shufflevector <4 x double> [[B:%.*]], <4 x double> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT3:%.*]] = shufflevector <4 x double> [[B]], <4 x double> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[BLOCK:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP0:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x double> undef, double [[TMP0]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP1:%.*]] = fmul <1 x double> [[BLOCK]], [[SPLAT_SPLAT]]
+; CHECK-NEXT:    [[BLOCK4:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT5:%.*]] = insertelement <1 x double> undef, double [[TMP2]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT6:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT5]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP3:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK4]], <1 x double> [[SPLAT_SPLAT6]], <1 x double> [[TMP1]])
+; CHECK-NEXT:    [[TMP4:%.*]] = shufflevector <1 x double> [[TMP3]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP5:%.*]] = shufflevector <2 x double> undef, <2 x double> [[TMP4]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK7:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT8:%.*]] = insertelement <1 x double> undef, double [[TMP6]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT9:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT8]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP7:%.*]] = fmul <1 x double> [[BLOCK7]], [[SPLAT_SPLAT9]]
+; CHECK-NEXT:    [[BLOCK10:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <2 x double> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT11:%.*]] = insertelement <1 x double> undef, double [[TMP8]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT12:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT11]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP9:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK10]], <1 x double> [[SPLAT_SPLAT12]], <1 x double> [[TMP7]])
+; CHECK-NEXT:    [[TMP10:%.*]] = shufflevector <1 x double> [[TMP9]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP11:%.*]] = shufflevector <2 x double> [[TMP5]], <2 x double> [[TMP10]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[BLOCK13:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP12:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT14:%.*]] = insertelement <1 x double> undef, double [[TMP12]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT15:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT14]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP13:%.*]] = fmul <1 x double> [[BLOCK13]], [[SPLAT_SPLAT15]]
+; CHECK-NEXT:    [[BLOCK16:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP14:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT17:%.*]] = insertelement <1 x double> undef, double [[TMP14]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT18:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT17]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP15:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK16]], <1 x double> [[SPLAT_SPLAT18]], <1 x double> [[TMP13]])
+; CHECK-NEXT:    [[TMP16:%.*]] = shufflevector <1 x double> [[TMP15]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP17:%.*]] = shufflevector <2 x double> undef, <2 x double> [[TMP16]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK19:%.*]] = shufflevector <2 x double> [[SPLIT]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT20:%.*]] = insertelement <1 x double> undef, double [[TMP18]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT21:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT20]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP19:%.*]] = fmul <1 x double> [[BLOCK19]], [[SPLAT_SPLAT21]]
+; CHECK-NEXT:    [[BLOCK22:%.*]] = shufflevector <2 x double> [[SPLIT1]], <2 x double> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP20:%.*]] = extractelement <2 x double> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT23:%.*]] = insertelement <1 x double> undef, double [[TMP20]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT24:%.*]] = shufflevector <1 x double> [[SPLAT_SPLATINSERT23]], <1 x double> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP21:%.*]] = call <1 x double> @llvm.fmuladd.v1f64(<1 x double> [[BLOCK22]], <1 x double> [[SPLAT_SPLAT24]], <1 x double> [[TMP19]])
+; CHECK-NEXT:    [[TMP22:%.*]] = shufflevector <1 x double> [[TMP21]], <1 x double> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP23:%.*]] = shufflevector <2 x double> [[TMP17]], <2 x double> [[TMP22]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[TMP24:%.*]] = shufflevector <2 x double> [[TMP11]], <2 x double> [[TMP23]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT:    ret <4 x double> [[TMP24]]
+;
+entry:
+  %c = call <4 x double> @llvm.matrix.multiply.v4f64.v4f64.v4f64(<4 x double> %a, <4 x double> %b, i32 2, i32 2, i32 2)
+  ret <4 x double> %c
+}
+
+declare <4 x double> @llvm.matrix.multiply.v4f64.v4f64.v4f64(<4 x double>, <4 x double>, i32, i32, i32)

diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction-fmf.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction-fmf.ll
new file mode 100644 (file)
index 0000000..f9ac729
--- /dev/null
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction-fmf.ll
@@ -0,0 +1,69 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -lower-matrix-intrinsics -S < %s | FileCheck %s
+; RUN: opt -passes='lower-matrix-intrinsics' -S < %s | FileCheck %s
+
+
+define <4 x float> @multiply_2x2(<4 x float> %a, <4 x float> %b) {
+; CHECK-LABEL: @multiply_2x2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SPLIT:%.*]] = shufflevector <4 x float> [[A:%.*]], <4 x float> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT1:%.*]] = shufflevector <4 x float> [[A]], <4 x float> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[SPLIT2:%.*]] = shufflevector <4 x float> [[B:%.*]], <4 x float> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT3:%.*]] = shufflevector <4 x float> [[B]], <4 x float> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[BLOCK:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP0:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x float> undef, float [[TMP0]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP1:%.*]] = fmul <1 x float> [[BLOCK]], [[SPLAT_SPLAT]]
+; CHECK-NEXT:    [[BLOCK4:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT5:%.*]] = insertelement <1 x float> undef, float [[TMP2]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT6:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT5]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP3:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK4]], <1 x float> [[SPLAT_SPLAT6]], <1 x float> [[TMP1]])
+; CHECK-NEXT:    [[TMP4:%.*]] = shufflevector <1 x float> [[TMP3]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP5:%.*]] = shufflevector <2 x float> undef, <2 x float> [[TMP4]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK7:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT8:%.*]] = insertelement <1 x float> undef, float [[TMP6]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT9:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT8]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP7:%.*]] = fmul <1 x float> [[BLOCK7]], [[SPLAT_SPLAT9]]
+; CHECK-NEXT:    [[BLOCK10:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT11:%.*]] = insertelement <1 x float> undef, float [[TMP8]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT12:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT11]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP9:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK10]], <1 x float> [[SPLAT_SPLAT12]], <1 x float> [[TMP7]])
+; CHECK-NEXT:    [[TMP10:%.*]] = shufflevector <1 x float> [[TMP9]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP11:%.*]] = shufflevector <2 x float> [[TMP5]], <2 x float> [[TMP10]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[BLOCK13:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP12:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT14:%.*]] = insertelement <1 x float> undef, float [[TMP12]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT15:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT14]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP13:%.*]] = fmul <1 x float> [[BLOCK13]], [[SPLAT_SPLAT15]]
+; CHECK-NEXT:    [[BLOCK16:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP14:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT17:%.*]] = insertelement <1 x float> undef, float [[TMP14]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT18:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT17]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP15:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK16]], <1 x float> [[SPLAT_SPLAT18]], <1 x float> [[TMP13]])
+; CHECK-NEXT:    [[TMP16:%.*]] = shufflevector <1 x float> [[TMP15]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP17:%.*]] = shufflevector <2 x float> undef, <2 x float> [[TMP16]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK19:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT20:%.*]] = insertelement <1 x float> undef, float [[TMP18]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT21:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT20]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP19:%.*]] = fmul <1 x float> [[BLOCK19]], [[SPLAT_SPLAT21]]
+; CHECK-NEXT:    [[BLOCK22:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP20:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT23:%.*]] = insertelement <1 x float> undef, float [[TMP20]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT24:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT23]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP21:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK22]], <1 x float> [[SPLAT_SPLAT24]], <1 x float> [[TMP19]])
+; CHECK-NEXT:    [[TMP22:%.*]] = shufflevector <1 x float> [[TMP21]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP23:%.*]] = shufflevector <2 x float> [[TMP17]], <2 x float> [[TMP22]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[TMP24:%.*]] = shufflevector <2 x float> [[TMP11]], <2 x float> [[TMP23]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT:    ret <4 x float> [[TMP24]]
+;
+entry:
+  %c = call contract <4 x float> @llvm.matrix.multiply.v4f32.v4f32.v4f32(<4 x float> %a, <4 x float> %b, i32 2, i32 2, i32 2)
+  ret <4 x float> %c
+}
+
+declare <4 x float> @llvm.matrix.multiply.v4f32.v4f32.v4f32(<4 x float>, <4 x float>, i32, i32, i32)

diff --git a/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction.ll b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction.ll
new file mode 100644 (file)
index 0000000..2e2cecf
--- /dev/null
+++ b/llvm/test/Transforms/LowerMatrixIntrinsics/multiply-float-contraction.ll
@@ -0,0 +1,69 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -lower-matrix-intrinsics -matrix-allow-contract -S < %s | FileCheck %s
+; RUN: opt -passes='lower-matrix-intrinsics' -matrix-allow-contract -S < %s | FileCheck %s
+
+
+define <4 x float> @multiply_2x2(<4 x float> %a, <4 x float> %b) {
+; CHECK-LABEL: @multiply_2x2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SPLIT:%.*]] = shufflevector <4 x float> [[A:%.*]], <4 x float> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT1:%.*]] = shufflevector <4 x float> [[A]], <4 x float> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[SPLIT2:%.*]] = shufflevector <4 x float> [[B:%.*]], <4 x float> undef, <2 x i32> <i32 0, i32 1>
+; CHECK-NEXT:    [[SPLIT3:%.*]] = shufflevector <4 x float> [[B]], <4 x float> undef, <2 x i32> <i32 2, i32 3>
+; CHECK-NEXT:    [[BLOCK:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP0:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT:%.*]] = insertelement <1 x float> undef, float [[TMP0]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP1:%.*]] = fmul <1 x float> [[BLOCK]], [[SPLAT_SPLAT]]
+; CHECK-NEXT:    [[BLOCK4:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP2:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT5:%.*]] = insertelement <1 x float> undef, float [[TMP2]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT6:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT5]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP3:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK4]], <1 x float> [[SPLAT_SPLAT6]], <1 x float> [[TMP1]])
+; CHECK-NEXT:    [[TMP4:%.*]] = shufflevector <1 x float> [[TMP3]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP5:%.*]] = shufflevector <2 x float> undef, <2 x float> [[TMP4]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK7:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT8:%.*]] = insertelement <1 x float> undef, float [[TMP6]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT9:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT8]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP7:%.*]] = fmul <1 x float> [[BLOCK7]], [[SPLAT_SPLAT9]]
+; CHECK-NEXT:    [[BLOCK10:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP8:%.*]] = extractelement <2 x float> [[SPLIT2]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT11:%.*]] = insertelement <1 x float> undef, float [[TMP8]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT12:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT11]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP9:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK10]], <1 x float> [[SPLAT_SPLAT12]], <1 x float> [[TMP7]])
+; CHECK-NEXT:    [[TMP10:%.*]] = shufflevector <1 x float> [[TMP9]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP11:%.*]] = shufflevector <2 x float> [[TMP5]], <2 x float> [[TMP10]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[BLOCK13:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP12:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT14:%.*]] = insertelement <1 x float> undef, float [[TMP12]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT15:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT14]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP13:%.*]] = fmul <1 x float> [[BLOCK13]], [[SPLAT_SPLAT15]]
+; CHECK-NEXT:    [[BLOCK16:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP14:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT17:%.*]] = insertelement <1 x float> undef, float [[TMP14]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT18:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT17]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP15:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK16]], <1 x float> [[SPLAT_SPLAT18]], <1 x float> [[TMP13]])
+; CHECK-NEXT:    [[TMP16:%.*]] = shufflevector <1 x float> [[TMP15]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP17:%.*]] = shufflevector <2 x float> undef, <2 x float> [[TMP16]], <2 x i32> <i32 2, i32 1>
+; CHECK-NEXT:    [[BLOCK19:%.*]] = shufflevector <2 x float> [[SPLIT]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 0
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT20:%.*]] = insertelement <1 x float> undef, float [[TMP18]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT21:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT20]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP19:%.*]] = fmul <1 x float> [[BLOCK19]], [[SPLAT_SPLAT21]]
+; CHECK-NEXT:    [[BLOCK22:%.*]] = shufflevector <2 x float> [[SPLIT1]], <2 x float> undef, <1 x i32> <i32 1>
+; CHECK-NEXT:    [[TMP20:%.*]] = extractelement <2 x float> [[SPLIT3]], i64 1
+; CHECK-NEXT:    [[SPLAT_SPLATINSERT23:%.*]] = insertelement <1 x float> undef, float [[TMP20]], i32 0
+; CHECK-NEXT:    [[SPLAT_SPLAT24:%.*]] = shufflevector <1 x float> [[SPLAT_SPLATINSERT23]], <1 x float> undef, <1 x i32> zeroinitializer
+; CHECK-NEXT:    [[TMP21:%.*]] = call <1 x float> @llvm.fmuladd.v1f32(<1 x float> [[BLOCK22]], <1 x float> [[SPLAT_SPLAT24]], <1 x float> [[TMP19]])
+; CHECK-NEXT:    [[TMP22:%.*]] = shufflevector <1 x float> [[TMP21]], <1 x float> undef, <2 x i32> <i32 0, i32 undef>
+; CHECK-NEXT:    [[TMP23:%.*]] = shufflevector <2 x float> [[TMP17]], <2 x float> [[TMP22]], <2 x i32> <i32 0, i32 2>
+; CHECK-NEXT:    [[TMP24:%.*]] = shufflevector <2 x float> [[TMP11]], <2 x float> [[TMP23]], <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK-NEXT:    ret <4 x float> [[TMP24]]
+;
+entry:
+  %c = call contract <4 x float> @llvm.matrix.multiply.v4f32.v4f32.v4f32(<4 x float> %a, <4 x float> %b, i32 2, i32 2, i32 2)
+  ret <4 x float> %c
+}
+
+declare <4 x float> @llvm.matrix.multiply.v4f32.v4f32.v4f32(<4 x float>, <4 x float>, i32, i32, i32)