From c872e809d1ac4aa405ae510e271f93d7662e26dd Mon Sep 17 00:00:00 2001 From: Florian Hahn Date: Wed, 15 Jul 2020 10:30:48 +0100 Subject: [Matrix] Only pass vector arg as overloaded type in MatrixBuilder. In 2b3c505, the pointer arguments for the matrix load and store intrinsics was changed to always be the element type of the vector argument. This patch updates the MatrixBuilder to not add the pointer type to the overloaded types and adjusts the clang/mlir tests. This should fix a few build failures on GreenDragon, including http://green.lab.llvm.org/green/job/test-suite-verify-machineinstrs-x86_64-O0-g/7891/ --- clang/test/CodeGen/matrix-type-builtins.c | 36 ++++++++++++++------------ clang/test/CodeGenCXX/matrix-type-builtins.cpp | 22 ++++++++-------- clang/test/CodeGenObjC/matrix-type-builtins.m | 4 +-- llvm/include/llvm/IR/MatrixBuilder.h | 4 +-- mlir/test/Target/llvmir-intrinsics.mlir | 8 +++--- 5 files changed, 39 insertions(+), 35 deletions(-) diff --git a/clang/test/CodeGen/matrix-type-builtins.c b/clang/test/CodeGen/matrix-type-builtins.c index 58fde6f..f7e9587 100644 --- a/clang/test/CodeGen/matrix-type-builtins.c +++ b/clang/test/CodeGen/matrix-type-builtins.c @@ -1,5 +1,9 @@ // RUN: %clang_cc1 -fenable-matrix -triple x86_64-apple-darwin %s -emit-llvm -disable-llvm-passes -o - | FileCheck %s +// Also check we do not crash when running some middle-end passes. Most +// importantly this includes the IR verifier, to ensure we emit valid IR. +// RUN: %clang_cc1 -fenable-matrix -emit-llvm -triple x86_64-apple-darwin %s -o %t + // Tests for the matrix type builtins. typedef double dx5x5_t __attribute__((matrix_type(5, 5))); @@ -100,7 +104,7 @@ void transpose_global() { void column_major_load_with_const_stride_double(double *Ptr) { // CHECK-LABEL: define void @column_major_load_with_const_stride_double(double* %Ptr) // CHECK: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) dx5x5_t m_a1 = __builtin_matrix_column_major_load(Ptr, 5, 5, 5); } @@ -108,7 +112,7 @@ void column_major_load_with_const_stride_double(double *Ptr) { void column_major_load_with_const_stride2_double(double *Ptr) { // CHECK-LABEL: define void @column_major_load_with_const_stride2_double(double* %Ptr) // CHECK: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 8 [[PTR]], i64 15, i1 false, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 8 [[PTR]], i64 15, i1 false, i32 5, i32 5) dx5x5_t m_a2 = __builtin_matrix_column_major_load(Ptr, 5, 5, 2 * 3 + 9); } @@ -117,7 +121,7 @@ void column_major_load_with_variable_stride_ull_float(float *Ptr, unsigned long // CHECK-LABEL: define void @column_major_load_with_variable_stride_ull_float(float* %Ptr, i64 %S) // CHECK: [[S:%.*]] = load i64, i64* %S.addr, align 8 // CHECK-NEXT: [[PTR:%.*]] = load float*, float** %Ptr.addr, align 8 - // CHECK-NEXT: call <6 x float> @llvm.matrix.column.major.load.v6f32.p0f32(float* align 4 [[PTR]], i64 [[S]], i1 false, i32 2, i32 3) + // CHECK-NEXT: call <6 x float> @llvm.matrix.column.major.load.v6f32(float* align 4 [[PTR]], i64 [[S]], i1 false, i32 2, i32 3) fx2x3_t m_b = __builtin_matrix_column_major_load(Ptr, 2, 3, S); } @@ -128,7 +132,7 @@ void column_major_load_with_stride_math_int(int *Ptr, int S) { // CHECK-NEXT: [[STRIDE:%.*]] = add nsw i32 [[S]], 32 // CHECK-NEXT: [[STRIDE_EXT:%.*]] = sext i32 [[STRIDE]] to i64 // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <80 x i32> @llvm.matrix.column.major.load.v80i32.p0i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 4, i32 20) + // CHECK-NEXT: call <80 x i32> @llvm.matrix.column.major.load.v80i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 4, i32 20) ix4x20_t m_c = __builtin_matrix_column_major_load(Ptr, 4, 20, S + 32); } @@ -140,7 +144,7 @@ void column_major_load_with_stride_math_s_int(int *Ptr, short S) { // CHECK-NEXT: [[STRIDE:%.*]] = add nsw i32 [[S_EXT]], 32 // CHECK-NEXT: [[STRIDE_EXT:%.*]] = sext i32 [[STRIDE]] to i64 // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: %matrix = call <80 x i32> @llvm.matrix.column.major.load.v80i32.p0i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 4, i32 20) + // CHECK-NEXT: %matrix = call <80 x i32> @llvm.matrix.column.major.load.v80i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 4, i32 20) ix4x20_t m_c = __builtin_matrix_column_major_load(Ptr, 4, 20, S + 32); } @@ -148,7 +152,7 @@ void column_major_load_with_stride_math_s_int(int *Ptr, short S) { void column_major_load_array1(double Ptr[25]) { // CHECK-LABEL: define void @column_major_load_array1(double* %Ptr) // CHECK: [[ADDR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 8 [[ADDR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 8 [[ADDR]], i64 5, i1 false, i32 5, i32 5) dx5x5_t m = __builtin_matrix_column_major_load(Ptr, 5, 5, 5); } @@ -158,7 +162,7 @@ void column_major_load_array2() { // CHECK-NEXT: entry: // CHECK-NEXT: [[PTR:%.*]] = alloca [25 x double], align 16 // CHECK: [[ARRAY_DEC:%.*]] = getelementptr inbounds [25 x double], [25 x double]* [[PTR]], i64 0, i64 0 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 16 [[ARRAY_DEC]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 16 [[ARRAY_DEC]], i64 5, i1 false, i32 5, i32 5) double Ptr[25]; dx5x5_t m = __builtin_matrix_column_major_load(Ptr, 5, 5, 5); @@ -167,7 +171,7 @@ void column_major_load_array2() { void column_major_load_const(const double *Ptr) { // CHECK-LABEL: define void @column_major_load_const(double* %Ptr) // CHECK: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) dx5x5_t m_a1 = __builtin_matrix_column_major_load(Ptr, 5, 5, 5); } @@ -175,7 +179,7 @@ void column_major_load_const(const double *Ptr) { void column_major_load_volatile(volatile double *Ptr) { // CHECK-LABEL: define void @column_major_load_volatile(double* %Ptr) // CHECK: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64.p0f64(double* align 8 [[PTR]], i64 5, i1 true, i32 5, i32 5) + // CHECK-NEXT: call <25 x double> @llvm.matrix.column.major.load.v25f64(double* align 8 [[PTR]], i64 5, i1 true, i32 5, i32 5) dx5x5_t m_a1 = __builtin_matrix_column_major_load(Ptr, 5, 5, 5); } @@ -184,7 +188,7 @@ void column_major_store_with_const_stride_double(double *Ptr) { // CHECK-LABEL: define void @column_major_store_with_const_stride_double(double* %Ptr) // CHECK: [[M:%.*]] = load <25 x double>, <25 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64.p0f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) dx5x5_t m; __builtin_matrix_column_major_store(m, Ptr, 5); @@ -194,7 +198,7 @@ void column_major_store_with_const_stride2_double(double *Ptr) { // CHECK-LABEL: define void @column_major_store_with_const_stride2_double(double* %Ptr) // CHECK: [[M:%.*]] = load <25 x double>, <25 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64.p0f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 15, i1 false, i32 5, i32 5) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 15, i1 false, i32 5, i32 5) // dx5x5_t m; __builtin_matrix_column_major_store(m, Ptr, 2 * 3 + 9); @@ -207,7 +211,7 @@ void column_major_store_with_stride_math_int(int *Ptr, int S) { // CHECK-NEXT: [[S:%.*]] = load i32, i32* %S.addr, align 4 // CHECK-NEXT: [[ADD:%.*]] = add nsw i32 [[S]], 32 // CHECK-NEXT: [[IDX:%.*]] = sext i32 [[ADD]] to i64 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v80i32.p0i32(<80 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX]], i1 false, i32 4, i32 20) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v80i32(<80 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX]], i1 false, i32 4, i32 20) ix4x20_t m; __builtin_matrix_column_major_store(m, Ptr, S + 32); @@ -221,7 +225,7 @@ void column_major_store_with_stride_math_s_int(int *Ptr, short S) { // CHECK-NEXT: [[EXT:%.*]] = sext i16 [[S]] to i32 // CHECK-NEXT: [[ADD:%.*]] = add nsw i32 [[EXT]], 2 // CHECK-NEXT: [[IDX:%.*]] = sext i32 [[ADD]] to i64 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v80i32.p0i32(<80 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX]], i1 false, i32 4, i32 20) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v80i32(<80 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX]], i1 false, i32 4, i32 20) ix4x20_t m; __builtin_matrix_column_major_store(m, Ptr, S + 2); @@ -231,7 +235,7 @@ void column_major_store_array1(double Ptr[25]) { // CHECK-LABEL: define void @column_major_store_array1(double* %Ptr) // CHECK: [[M:%.*]] = load <25 x double>, <25 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64.p0f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 false, i32 5, i32 5) dx5x5_t m; __builtin_matrix_column_major_store(m, Ptr, 5); @@ -241,7 +245,7 @@ void column_major_store_array2() { // CHECK-LABEL: define void @column_major_store_array2() // CHECK: [[M:%.*]] = load <25 x double>, <25 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = getelementptr inbounds [25 x double], [25 x double]* %Ptr, i64 0, i64 0 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64.p0f64(<25 x double> [[M]], double* align 16 [[PTR]], i64 5, i1 false, i32 5, i32 5) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64(<25 x double> [[M]], double* align 16 [[PTR]], i64 5, i1 false, i32 5, i32 5) double Ptr[25]; dx5x5_t m; @@ -252,7 +256,7 @@ void column_major_store_volatile(volatile double *Ptr) { // CHECK-LABEL: define void @column_major_store_volatile(double* %Ptr) #0 { // CHECK: [[M:%.*]] = load <25 x double>, <25 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64.p0f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 true, i32 5, i32 5) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v25f64(<25 x double> [[M]], double* align 8 [[PTR]], i64 5, i1 true, i32 5, i32 5) dx5x5_t m; __builtin_matrix_column_major_store(m, Ptr, 5); diff --git a/clang/test/CodeGenCXX/matrix-type-builtins.cpp b/clang/test/CodeGenCXX/matrix-type-builtins.cpp index 3141687..dd341d2 100644 --- a/clang/test/CodeGenCXX/matrix-type-builtins.cpp +++ b/clang/test/CodeGenCXX/matrix-type-builtins.cpp @@ -94,7 +94,7 @@ void test_column_major_load_with_stride_template_double(double *Ptr) { // CHECK-LABEL: define linkonce_odr <40 x double> @_Z29column_major_load_with_strideIdLj10ELj4ELj15EEU11matrix_typeXT0_EXT1_ET_PS0_(double* %Ptr) // CHECK: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call <40 x double> @llvm.matrix.column.major.load.v40f64.p0f64(double* align 8 [[PTR]], i64 15, i1 false, i32 10, i32 4) + // CHECK-NEXT: call <40 x double> @llvm.matrix.column.major.load.v40f64(double* align 8 [[PTR]], i64 15, i1 false, i32 10, i32 4) matrix_t M1 = column_major_load_with_stride(Ptr); } @@ -106,7 +106,7 @@ void test_column_major_load_with_stride_template_int(int *Ptr) { // CHECK-LABEL: define linkonce_odr <6 x i32> @_Z29column_major_load_with_strideIiLj3ELj2ELj12EEU11matrix_typeXT0_EXT1_ET_PS0_(i32* %Ptr) // CHECK: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <6 x i32> @llvm.matrix.column.major.load.v6i32.p0i32(i32* align 4 [[PTR]], i64 12, i1 false, i32 3, i32 2) + // CHECK-NEXT: call <6 x i32> @llvm.matrix.column.major.load.v6i32(i32* align 4 [[PTR]], i64 12, i1 false, i32 3, i32 2) matrix_t M1 = column_major_load_with_stride(Ptr); } @@ -124,7 +124,7 @@ void test_column_major_load_stride_wrapper(int *Ptr, UnsignedWrapper &W) { // CHECK-NEXT: [[STRIDE:%.*]] = call i32 @_ZN15UnsignedWrappercvjEv(%struct.UnsignedWrapper* [[W]]) // CHECK-NEXT: [[STRIDE_EXT:%.*]] = zext i32 [[STRIDE]] to i64 // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <4 x i32> @llvm.matrix.column.major.load.v4i32.p0i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 2, i32 2) + // CHECK-NEXT: call <4 x i32> @llvm.matrix.column.major.load.v4i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 2, i32 2) matrix_t M1 = __builtin_matrix_column_major_load(Ptr, 2, 2, W); } @@ -133,7 +133,7 @@ constexpr int constexpr3() { return 3; } void test_column_major_load_constexpr_num_rows(int *Ptr) { // CHECK-LABEL: define void @_Z41test_column_major_load_constexpr_num_rowsPi(i32* %Ptr) // CHECK: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <6 x i32> @llvm.matrix.column.major.load.v6i32.p0i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 3, i32 2) + // CHECK-NEXT: call <6 x i32> @llvm.matrix.column.major.load.v6i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 3, i32 2) matrix_t M1 = __builtin_matrix_column_major_load(Ptr, constexpr3(), 2, 3); } @@ -143,7 +143,7 @@ constexpr int constexpr1() { return 1; } void test_column_major_load_constexpr_num_columns(int *Ptr) { // CHECK-LABEL: define void @_Z44test_column_major_load_constexpr_num_columnsPi(i32* %Ptr) // CHECK: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <2 x i32> @llvm.matrix.column.major.load.v2i32.p0i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 2, i32 1) + // CHECK-NEXT: call <2 x i32> @llvm.matrix.column.major.load.v2i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 2, i32 1) matrix_t M1 = __builtin_matrix_column_major_load(Ptr, 2, constexpr1(), 3); } @@ -153,7 +153,7 @@ constexpr int constexpr_plus1() { return N + 1; } void test_column_major_load_constexpr_num_columns_temp(int *Ptr) { // CHECK-LABEL: define void @_Z49test_column_major_load_constexpr_num_columns_tempPi(i32* %Ptr) // CHECK: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <10 x i32> @llvm.matrix.column.major.load.v10i32.p0i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 2, i32 5) + // CHECK-NEXT: call <10 x i32> @llvm.matrix.column.major.load.v10i32(i32* align 4 [[PTR]], i64 3, i1 false, i32 2, i32 5) matrix_t M1 = __builtin_matrix_column_major_load(Ptr, 2, constexpr_plus1<4>(), 3); } @@ -162,7 +162,7 @@ void test_column_major_load_constexpr_stride_constexpr(int *Ptr) { // CHECK: [[STRIDE:%.*]] = call i32 @_Z10constexpr3v() // CHECK-NEXT: [[STRIDE_EXT:%.*]] = sext i32 [[STRIDE]] to i64 // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call <4 x i32> @llvm.matrix.column.major.load.v4i32.p0i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 2, i32 2) + // CHECK-NEXT: call <4 x i32> @llvm.matrix.column.major.load.v4i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 2, i32 2) matrix_t M1 = __builtin_matrix_column_major_load(Ptr, 2, 2, constexpr3()); } @@ -200,7 +200,7 @@ void test_column_major_store_with_stride_template_double(double *Ptr) { // CHECK-LABEL: define linkonce_odr void @_Z30column_major_store_with_strideIdLj10ELj4ELj15EEvRU11matrix_typeXT0_EXT1_ET_PS0_([40 x double]* nonnull align 8 dereferenceable(320) %m, double* %Ptr) // CHECK: [[M:%.*]] = load <40 x double>, <40 x double>* {{.*}}, align 8 // CHECK-NEXT: [[PTR:%.*]] = load double*, double** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v40f64.p0f64(<40 x double> [[M]], double* align 8 [[PTR]], i64 15, i1 false, i32 10, i32 4) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v40f64(<40 x double> [[M]], double* align 8 [[PTR]], i64 15, i1 false, i32 10, i32 4) matrix_t M1; column_major_store_with_stride(M1, Ptr); @@ -214,7 +214,7 @@ void test_column_major_store_with_stride_template_int(int *Ptr) { // CHECK-LABEL: define linkonce_odr void @_Z30column_major_store_with_strideIiLj3ELj2ELj3EEvRU11matrix_typeXT0_EXT1_ET_PS0_([6 x i32]* nonnull align 4 dereferenceable(24) %m, i32* %Ptr) // CHECK: [[M:%.*]] = load <6 x i32>, <6 x i32>* {{.*}}, align 4 // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v6i32.p0i32(<6 x i32> [[M]], i32* align 4 [[PTR]], i64 3, i1 false, i32 3, i32 2) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v6i32(<6 x i32> [[M]], i32* align 4 [[PTR]], i64 3, i1 false, i32 3, i32 2) matrix_t M1; column_major_store_with_stride(M1, Ptr); @@ -227,7 +227,7 @@ void test_column_major_store_stride_wrapper(int *Ptr, UnsignedWrapper &W) { // CHECK-NEXT: [[W:%.*]] = load %struct.UnsignedWrapper*, %struct.UnsignedWrapper** %W.addr, align 8 // CHECK-NEXT: [[IDX:%.*]] = call i32 @_ZN15UnsignedWrappercvjEv(%struct.UnsignedWrapper* [[W]]) // CHECK-NEXT: [[IDX_EXT:%.*]] = zext i32 [[IDX]] to i64 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v4i32.p0i32(<4 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 2, i32 2) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v4i32(<4 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 2, i32 2) matrix_t M1; __builtin_matrix_column_major_store(M1, Ptr, W); @@ -239,7 +239,7 @@ void test_column_major_store_constexpr_stride_constexpr(int *Ptr) { // CHECK-NEXT: [[PTR:%.*]] = load i32*, i32** %Ptr.addr, align 8 // CHECK-NEXT: [[IDX:%.*]] = call i32 @_Z10constexpr3v() // CHECK-NEXT: [[IDX_EXT:%.*]] = sext i32 [[IDX]] to i64 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v4i32.p0i32(<4 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 2, i32 2) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v4i32(<4 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 2, i32 2) matrix_t M; __builtin_matrix_column_major_store(M, Ptr, constexpr3()); diff --git a/clang/test/CodeGenObjC/matrix-type-builtins.m b/clang/test/CodeGenObjC/matrix-type-builtins.m index 78dfad2..19bd957 100644 --- a/clang/test/CodeGenObjC/matrix-type-builtins.m +++ b/clang/test/CodeGenObjC/matrix-type-builtins.m @@ -56,7 +56,7 @@ void test_column_major_load(PtrValue *Ptr, IntValue *Stride) { // CHECK: [[STRIDE:%.*]] = call i32 bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to i32 (i8*, i8*)*) // CHECK-NEXT: [[STRIDE_EXT:%.*]] = sext i32 [[STRIDE]] to i64 // CHECK: [[PTR:%.*]] = call i32* bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to i32* (i8*, i8*)*) - // CHECK-NEXT: call <12 x i32> @llvm.matrix.column.major.load.v12i32.p0i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 3, i32 4) + // CHECK-NEXT: call <12 x i32> @llvm.matrix.column.major.load.v12i32(i32* align 4 [[PTR]], i64 [[STRIDE_EXT]], i1 false, i32 3, i32 4) u3x4 m = __builtin_matrix_column_major_load(Ptr.value, 3, 4, Stride.value); } @@ -67,7 +67,7 @@ void test_column_major_store(UnsignedMatrixValue *M, PtrValue *Ptr, IntValue *St // CHECK: [[PTR:%.*]] = call i32* bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to i32* (i8*, i8*)*) // CHECK: [[IDX:%.*]] = call i32 bitcast (i8* (i8*, i8*, ...)* @objc_msgSend to i32 (i8*, i8*)*) // CHECK-NEXT: [[IDX_EXT:%.*]] = sext i32 [[IDX]] to i64 - // CHECK-NEXT: call void @llvm.matrix.column.major.store.v12i32.p0i32(<12 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 3, i32 4) + // CHECK-NEXT: call void @llvm.matrix.column.major.store.v12i32(<12 x i32> [[M]], i32* align 4 [[PTR]], i64 [[IDX_EXT]], i1 false, i32 3, i32 4) __builtin_matrix_column_major_store(M.value, Ptr.value, Stride.value); } diff --git a/llvm/include/llvm/IR/MatrixBuilder.h b/llvm/include/llvm/IR/MatrixBuilder.h index 515b86b..5d04b35 100644 --- a/llvm/include/llvm/IR/MatrixBuilder.h +++ b/llvm/include/llvm/IR/MatrixBuilder.h @@ -69,7 +69,7 @@ public: Value *Ops[] = {DataPtr, Stride, B.getInt1(IsVolatile), B.getInt32(Rows), B.getInt32(Columns)}; - Type *OverloadedTypes[] = {RetType, PtrTy}; + Type *OverloadedTypes[] = {RetType}; Function *TheFn = Intrinsic::getDeclaration( getModule(), Intrinsic::matrix_column_major_load, OverloadedTypes); @@ -92,7 +92,7 @@ public: Value *Ops[] = {Matrix, Ptr, Stride, B.getInt1(IsVolatile), B.getInt32(Rows), B.getInt32(Columns)}; - Type *OverloadedTypes[] = {Matrix->getType(), Ptr->getType()}; + Type *OverloadedTypes[] = {Matrix->getType()}; Function *TheFn = Intrinsic::getDeclaration( getModule(), Intrinsic::matrix_column_major_store, OverloadedTypes); diff --git a/mlir/test/Target/llvmir-intrinsics.mlir b/mlir/test/Target/llvmir-intrinsics.mlir index ffbbf35..1595edf 100644 --- a/mlir/test/Target/llvmir-intrinsics.mlir +++ b/mlir/test/Target/llvmir-intrinsics.mlir @@ -181,11 +181,11 @@ llvm.func @matrix_intrinsics(%A: !llvm<"<64 x float>">, %B: !llvm<"<48 x float>" // CHECK: call <48 x float> @llvm.matrix.transpose.v48f32(<48 x float> %1, i32 3, i32 16) %D = llvm.intr.matrix.transpose %B { rows = 3: i32, columns = 16: i32} : !llvm<"<48 x float>"> into !llvm<"<48 x float>"> - // CHECK: call <48 x float> @llvm.matrix.column.major.load.v48f32.p0f32(float* align 4 %2, i64 %3, i1 false, i32 3, i32 16) + // CHECK: call <48 x float> @llvm.matrix.column.major.load.v48f32(float* align 4 %2, i64 %3, i1 false, i32 3, i32 16) %E = llvm.intr.matrix.column.major.load %ptr, { isVolatile = 0: i1, rows = 3: i32, columns = 16: i32} : !llvm<"<48 x float>"> from !llvm<"float*"> stride !llvm.i64 - // CHECK: call void @llvm.matrix.column.major.store.v48f32.p0f32(<48 x float> %7, float* align 4 %2, i64 %3, i1 false, i32 3, i32 16) + // CHECK: call void @llvm.matrix.column.major.store.v48f32(<48 x float> %7, float* align 4 %2, i64 %3, i1 false, i32 3, i32 16) llvm.intr.matrix.column.major.store %E, %ptr, { isVolatile = 0: i1, rows = 3: i32, columns = 16: i32} : !llvm<"<48 x float>"> to !llvm<"float*"> stride !llvm.i64 @@ -242,8 +242,8 @@ llvm.func @memcpy_test(%arg0: !llvm.i32, %arg1: !llvm.i1, %arg2: !llvm<"i8*">, % // CHECK-DAG: declare float @llvm.copysign.f32(float, float) // CHECK-DAG: declare <12 x float> @llvm.matrix.multiply.v12f32.v64f32.v48f32(<64 x float>, <48 x float>, i32 immarg, i32 immarg, i32 immarg) // CHECK-DAG: declare <48 x float> @llvm.matrix.transpose.v48f32(<48 x float>, i32 immarg, i32 immarg) -// CHECK-DAG: declare <48 x float> @llvm.matrix.column.major.load.v48f32.p0f32(float* nocapture, i64, i1 immarg, i32 immarg, i32 immarg) -// CHECK-DAG: declare void @llvm.matrix.column.major.store.v48f32.p0f32(<48 x float>, float* nocapture writeonly, i64, i1 immarg, i32 immarg, i32 immarg) +// CHECK-DAG: declare <48 x float> @llvm.matrix.column.major.load.v48f32(float* nocapture, i64, i1 immarg, i32 immarg, i32 immarg) +// CHECK-DAG: declare void @llvm.matrix.column.major.store.v48f32(<48 x float>, float* nocapture writeonly, i64, i1 immarg, i32 immarg, i32 immarg) // CHECK-DAG: declare <7 x float> @llvm.masked.load.v7f32.p0v7f32(<7 x float>*, i32 immarg, <7 x i1>, <7 x float>) // CHECK-DAG: declare void @llvm.masked.store.v7f32.p0v7f32(<7 x float>, <7 x float>*, i32 immarg, <7 x i1>) // CHECK-DAG: declare void @llvm.memcpy.p0i8.p0i8.i32(i8* noalias nocapture writeonly, i8* noalias nocapture readonly, i32, i1 immarg) -- cgit v1.1