maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[mlir][spirv] Remove layout decoration on unneeded storage classes 

Per SPIR-V validation rules, explict layout decorations are only
needed for StorageBuffer, PhysicalStorageBuffer, Uniform, and
PushConstant storage classes. (And even that is for Shader
capabilities). So we don't need such decorations on the rest.

Reviewed By: hanchung

Differential Revision: https://reviews.llvm.org/D124543
This commit is contained in:
Lei Zhang 2022-04-28 08:18:08 -04:00
parent 8854b73606
commit bbffece383
7 changed files with 79 additions and 70 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
mlir/lib/Dialect/SPIRV/Transforms/SPIRVConversion.cpp
View File

@ -370,7 +370,7 @@ static Type convertTensorType(const spirv::TargetEnv &targetEnv,
return nullptr;
}
return spirv::ArrayType::get(arrayElemType, arrayElemCount, *arrayElemSize);
return spirv::ArrayType::get(arrayElemType, arrayElemCount);
}
static Type convertBoolMemrefType(const spirv::TargetEnv &targetEnv,
@ -407,15 +407,15 @@ static Type convertBoolMemrefType(const spirv::TargetEnv &targetEnv,
}
if (!type.hasStaticShape()) {
auto arrayType =
spirv::RuntimeArrayType::get(arrayElemType, *arrayElemSize);
int64_t stride = needsExplicitLayout(*storageClass) ? *arrayElemSize : 0;
auto arrayType = spirv::RuntimeArrayType::get(arrayElemType, stride);
return wrapInStructAndGetPointer(arrayType, *storageClass);
}
int64_t memrefSize = (type.getNumElements() * numBoolBits + 7) / 8;
auto arrayElemCount = (memrefSize + *arrayElemSize - 1) / *arrayElemSize;
auto arrayType =
spirv::ArrayType::get(arrayElemType, arrayElemCount, *arrayElemSize);
int64_t stride = needsExplicitLayout(*storageClass) ? *arrayElemSize : 0;
auto arrayType = spirv::ArrayType::get(arrayElemType, arrayElemCount, stride);
return wrapInStructAndGetPointer(arrayType, *storageClass);
}
@ -470,8 +470,8 @@ static Type convertMemrefType(const spirv::TargetEnv &targetEnv,
}
if (!type.hasStaticShape()) {
auto arrayType =
spirv::RuntimeArrayType::get(arrayElemType, *arrayElemSize);
int64_t stride = needsExplicitLayout(*storageClass) ? *arrayElemSize : 0;
auto arrayType = spirv::RuntimeArrayType::get(arrayElemType, stride);
return wrapInStructAndGetPointer(arrayType, *storageClass);
}
@ -483,10 +483,8 @@ static Type convertMemrefType(const spirv::TargetEnv &targetEnv,
}
auto arrayElemCount = *memrefSize / *elementSize;
auto arrayType =
spirv::ArrayType::get(arrayElemType, arrayElemCount, *arrayElemSize);
int64_t stride = needsExplicitLayout(*storageClass) ? *arrayElemSize : 0;
auto arrayType = spirv::ArrayType::get(arrayElemType, arrayElemCount, stride);
return wrapInStructAndGetPointer(arrayType, *storageClass);
}

48
mlir/test/Conversion/ArithmeticToSPIRV/arithmetic-to-spirv.mlir
View File

@ -435,17 +435,17 @@ func.func @constant() {
%3 = arith.constant dense<[2, 3]> : vector<2xi32>
// CHECK: spv.Constant 1 : i32
%4 = arith.constant 1 : index
// CHECK: spv.Constant dense<1> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<1> : tensor<6xi32> : !spv.array<6 x i32>
%5 = arith.constant dense<1> : tensor<2x3xi32>
// CHECK: spv.Constant dense<1.000000e+00> : tensor<6xf32> : !spv.array<6 x f32, stride=4>
// CHECK: spv.Constant dense<1.000000e+00> : tensor<6xf32> : !spv.array<6 x f32>
%6 = arith.constant dense<1.0> : tensor<2x3xf32>
// CHECK: spv.Constant dense<{{\[}}1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00, 5.000000e+00, 6.000000e+00]> : tensor<6xf32> : !spv.array<6 x f32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00, 5.000000e+00, 6.000000e+00]> : tensor<6xf32> : !spv.array<6 x f32>
%7 = arith.constant dense<[[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]]> : tensor<2x3xf32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%8 = arith.constant dense<[[1, 2, 3], [4, 5, 6]]> : tensor<2x3xi32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%9 = arith.constant dense<[[1, 2], [3, 4], [5, 6]]> : tensor<3x2xi32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%10 = arith.constant dense<[1, 2, 3, 4, 5, 6]> : tensor<6xi32>
return
}
@ -458,7 +458,7 @@ func.func @constant_16bit() {
%1 = arith.constant 5.0 : f16
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi16>
%2 = arith.constant dense<[2, 3]> : vector<2xi16>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf16> : !spv.array<5 x f16, stride=2>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf16> : !spv.array<5 x f16>
%3 = arith.constant dense<4.0> : tensor<5xf16>
return
}
@ -471,7 +471,7 @@ func.func @constant_64bit() {
%1 = arith.constant 5.0 : f64
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi64>
%2 = arith.constant dense<[2, 3]> : vector<2xi64>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf64> : !spv.array<5 x f64, stride=8>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf64> : !spv.array<5 x f64>
%3 = arith.constant dense<4.0> : tensor<5xf64>
return
}
@ -504,9 +504,9 @@ func.func @constant_16bit() {
%1 = arith.constant 5.0 : f16
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi32>
%2 = arith.constant dense<[2, 3]> : vector<2xi16>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32, stride=4>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32>
%3 = arith.constant dense<4.0> : tensor<5xf16>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32, stride=4>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32>
%4 = arith.constant dense<[[1.0, 2.0], [3.0, 4.0]]> : tensor<2x2xf16>
return
}
@ -519,9 +519,9 @@ func.func @constant_64bit() {
%1 = arith.constant 5.0 : f64
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi32>
%2 = arith.constant dense<[2, 3]> : vector<2xi64>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32, stride=4>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32>
%3 = arith.constant dense<4.0> : tensor<5xf64>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32, stride=4>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32>
%4 = arith.constant dense<[[1.0, 2.0], [3.0, 4.0]]> : tensor<2x2xf16>
return
}
@ -1360,17 +1360,17 @@ func.func @constant() {
%3 = arith.constant dense<[2, 3]> : vector<2xi32>
// CHECK: spv.Constant 1 : i32
%4 = arith.constant 1 : index
// CHECK: spv.Constant dense<1> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<1> : tensor<6xi32> : !spv.array<6 x i32>
%5 = arith.constant dense<1> : tensor<2x3xi32>
// CHECK: spv.Constant dense<1.000000e+00> : tensor<6xf32> : !spv.array<6 x f32, stride=4>
// CHECK: spv.Constant dense<1.000000e+00> : tensor<6xf32> : !spv.array<6 x f32>
%6 = arith.constant dense<1.0> : tensor<2x3xf32>
// CHECK: spv.Constant dense<{{\[}}1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00, 5.000000e+00, 6.000000e+00]> : tensor<6xf32> : !spv.array<6 x f32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00, 5.000000e+00, 6.000000e+00]> : tensor<6xf32> : !spv.array<6 x f32>
%7 = arith.constant dense<[[1.0, 2.0, 3.0], [4.0, 5.0, 6.0]]> : tensor<2x3xf32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%8 = arith.constant dense<[[1, 2, 3], [4, 5, 6]]> : tensor<2x3xi32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%9 = arith.constant dense<[[1, 2], [3, 4], [5, 6]]> : tensor<3x2xi32>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32, stride=4>
// CHECK: spv.Constant dense<{{\[}}1, 2, 3, 4, 5, 6]> : tensor<6xi32> : !spv.array<6 x i32>
%10 = arith.constant dense<[1, 2, 3, 4, 5, 6]> : tensor<6xi32>
return
}
@ -1383,7 +1383,7 @@ func.func @constant_16bit() {
%1 = arith.constant 5.0 : f16
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi16>
%2 = arith.constant dense<[2, 3]> : vector<2xi16>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf16> : !spv.array<5 x f16, stride=2>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf16> : !spv.array<5 x f16>
%3 = arith.constant dense<4.0> : tensor<5xf16>
return
}
@ -1396,7 +1396,7 @@ func.func @constant_64bit() {
%1 = arith.constant 5.0 : f64
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi64>
%2 = arith.constant dense<[2, 3]> : vector<2xi64>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf64> : !spv.array<5 x f64, stride=8>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf64> : !spv.array<5 x f64>
%3 = arith.constant dense<4.0> : tensor<5xf64>
return
}
@ -1418,9 +1418,9 @@ func.func @constant_16bit() {
%1 = arith.constant 5.0 : f16
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi32>
%2 = arith.constant dense<[2, 3]> : vector<2xi16>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32, stride=4>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32>
%3 = arith.constant dense<4.0> : tensor<5xf16>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32, stride=4>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32>
%4 = arith.constant dense<[[1.0, 2.0], [3.0, 4.0]]> : tensor<2x2xf16>
return
}
@ -1433,9 +1433,9 @@ func.func @constant_64bit() {
%1 = arith.constant 5.0 : f64
// CHECK: spv.Constant dense<[2, 3]> : vector<2xi32>
%2 = arith.constant dense<[2, 3]> : vector<2xi64>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32, stride=4>
// CHECK: spv.Constant dense<4.000000e+00> : tensor<5xf32> : !spv.array<5 x f32>
%3 = arith.constant dense<4.0> : tensor<5xf64>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32, stride=4>
// CHECK: spv.Constant dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32> : !spv.array<4 x f32>
%4 = arith.constant dense<[[1.0, 2.0], [3.0, 4.0]]> : tensor<2x2xf16>
return
}

55
mlir/test/Conversion/FuncToSPIRV/types-to-spirv.mlir
View File

@ -282,6 +282,17 @@ func.func @memref_mem_space(
%arg5: memref<4xf32, 6>
) { return }
// CHECK-LABEL: func @memref_1bit_type
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<8 x i32, stride=4> [0])>, StorageBuffer>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<8 x i32>)>, Function>
// NOEMU-LABEL: func @memref_1bit_type
// NOEMU-SAME: memref<4x8xi1>
// NOEMU-SAME: memref<4x8xi1, 6>
func.func @memref_1bit_type(
%arg0: memref<4x8xi1, 0>,
%arg1: memref<4x8xi1, 6>
) { return }
} // end module
// -----
@ -337,13 +348,13 @@ func.func @memref_16bit_Uniform(%arg0: memref<16xsi16, 4>) { return }
func.func @memref_16bit_PushConstant(%arg0: memref<16xui16, 7>) { return }
// CHECK-LABEL: spv.func @memref_16bit_Input
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f32, stride=4>)>, Input>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f32>)>, Input>
// NOEMU-LABEL: func @memref_16bit_Input
// NOEMU-SAME: memref<16xf16, 9>
func.func @memref_16bit_Input(%arg3: memref<16xf16, 9>) { return }
// CHECK-LABEL: spv.func @memref_16bit_Output
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f32, stride=4>)>, Output>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f32>)>, Output>
// NOEMU-LABEL: func @memref_16bit_Output
// NOEMU-SAME: memref<16xf16, 10>
func.func @memref_16bit_Output(%arg4: memref<16xf16, 10>) { return }
@ -451,15 +462,15 @@ module attributes {
} {
// CHECK-LABEL: spv.func @memref_16bit_Input
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f16, stride=2>)>, Input>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f16>)>, Input>
// NOEMU-LABEL: spv.func @memref_16bit_Input
// NOEMU-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f16, stride=2>)>, Input>
// NOEMU-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x f16>)>, Input>
func.func @memref_16bit_Input(%arg3: memref<16xf16, 9>) { return }
// CHECK-LABEL: spv.func @memref_16bit_Output
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x i16, stride=2>)>, Output>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x i16>)>, Output>
// NOEMU-LABEL: spv.func @memref_16bit_Output
// NOEMU-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x i16, stride=2>)>, Output>
// NOEMU-SAME: !spv.ptr<!spv.struct<(!spv.array<16 x i16>)>, Output>
func.func @memref_16bit_Output(%arg4: memref<16xi16, 10>) { return }
} // end module
@ -564,13 +575,13 @@ func.func @memref_16bit_Uniform(%arg0: memref<?xsi16, 4>) { return }
func.func @memref_16bit_PushConstant(%arg0: memref<?xui16, 7>) { return }
// CHECK-LABEL: spv.func @memref_16bit_Input
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.rtarray<f32, stride=4>)>, Input>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.rtarray<f32>)>, Input>
// NOEMU-LABEL: func @memref_16bit_Input
// NOEMU-SAME: memref<?xf16, 9>
func.func @memref_16bit_Input(%arg3: memref<?xf16, 9>) { return }
// CHECK-LABEL: spv.func @memref_16bit_Output
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.rtarray<f32, stride=4>)>, Output>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.rtarray<f32>)>, Output>
// NOEMU-LABEL: func @memref_16bit_Output
// NOEMU-SAME: memref<?xf16, 10>
func.func @memref_16bit_Output(%arg4: memref<?xf16, 10>) { return }
@ -629,10 +640,10 @@ module attributes {
} {
// CHECK-LABEL: spv.func @int_tensor_types
// CHECK-SAME: !spv.array<32 x i64, stride=8>
// CHECK-SAME: !spv.array<32 x i32, stride=4>
// CHECK-SAME: !spv.array<32 x i16, stride=2>
// CHECK-SAME: !spv.array<32 x i8, stride=1>
// CHECK-SAME: !spv.array<32 x i64>
// CHECK-SAME: !spv.array<32 x i32>
// CHECK-SAME: !spv.array<32 x i16>
// CHECK-SAME: !spv.array<32 x i8>
func.func @int_tensor_types(
%arg0: tensor<8x4xi64>,
%arg1: tensor<8x4xi32>,
@ -641,9 +652,9 @@ func.func @int_tensor_types(
) { return }
// CHECK-LABEL: spv.func @float_tensor_types
// CHECK-SAME: !spv.array<32 x f64, stride=8>
// CHECK-SAME: !spv.array<32 x f32, stride=4>
// CHECK-SAME: !spv.array<32 x f16, stride=2>
// CHECK-SAME: !spv.array<32 x f64>
// CHECK-SAME: !spv.array<32 x f32>
// CHECK-SAME: !spv.array<32 x f16>
func.func @float_tensor_types(
%arg0: tensor<8x4xf64>,
%arg1: tensor<8x4xf32>,
@ -660,10 +671,10 @@ module attributes {
} {
// CHECK-LABEL: spv.func @int_tensor_types
// CHECK-SAME: !spv.array<32 x i32, stride=4>
// CHECK-SAME: !spv.array<32 x i32, stride=4>
// CHECK-SAME: !spv.array<32 x i32, stride=4>
// CHECK-SAME: !spv.array<32 x i32, stride=4>
// CHECK-SAME: !spv.array<32 x i32>
// CHECK-SAME: !spv.array<32 x i32>
// CHECK-SAME: !spv.array<32 x i32>
// CHECK-SAME: !spv.array<32 x i32>
func.func @int_tensor_types(
%arg0: tensor<8x4xi64>,
%arg1: tensor<8x4xi32>,
@ -672,9 +683,9 @@ func.func @int_tensor_types(
) { return }
// CHECK-LABEL: spv.func @float_tensor_types
// CHECK-SAME: !spv.array<32 x f32, stride=4>
// CHECK-SAME: !spv.array<32 x f32, stride=4>
// CHECK-SAME: !spv.array<32 x f32, stride=4>
// CHECK-SAME: !spv.array<32 x f32>
// CHECK-SAME: !spv.array<32 x f32>
// CHECK-SAME: !spv.array<32 x f32>
func.func @float_tensor_types(
%arg0: tensor<8x4xf64>,
%arg1: tensor<8x4xf32>,

2
mlir/test/Conversion/GPUToSPIRV/module-structure-opencl.mlir
View File

@ -9,7 +9,7 @@ module attributes {
// CHECK: spv.func
// CHECK-SAME: {{%.*}}: f32
// CHECK-NOT: spv.interface_var_abi
// CHECK-SAME: {{%.*}}: !spv.ptr<!spv.struct<(!spv.array<12 x f32, stride=4>)>, CrossWorkgroup>
// CHECK-SAME: {{%.*}}: !spv.ptr<!spv.struct<(!spv.array<12 x f32>)>, CrossWorkgroup>
// CHECK-NOT: spv.interface_var_abi
// CHECK-SAME: spv.entry_point_abi = {local_size = dense<[32, 4, 1]> : vector<3xi32>}
gpu.func @basic_module_structure(%arg0 : f32, %arg1 : memref<12xf32, 11>) kernel

12
mlir/test/Conversion/MemRefToSPIRV/alloc.mlir
View File

@ -13,7 +13,7 @@ module attributes {
return
}
}
// CHECK: spv.GlobalVariable @[[VAR:.+]] : !spv.ptr<!spv.struct<(!spv.array<20 x f32, stride=4>)>, Workgroup>
// CHECK: spv.GlobalVariable @[[VAR:.+]] : !spv.ptr<!spv.struct<(!spv.array<20 x f32>)>, Workgroup>
// CHECK: func @alloc_dealloc_workgroup_mem
// CHECK-NOT: memref.alloc
// CHECK: %[[PTR:.+]] = spv.mlir.addressof @[[VAR]]
@ -40,7 +40,7 @@ module attributes {
}
// CHECK: spv.GlobalVariable @__workgroup_mem__{{[0-9]+}}
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<20 x i32, stride=4>)>, Workgroup>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<20 x i32>)>, Workgroup>
// CHECK_LABEL: spv.func @alloc_dealloc_workgroup_mem
// CHECK: %[[VAR:.+]] = spv.mlir.addressof @__workgroup_mem__0
// CHECK: %[[LOC:.+]] = spv.SDiv
@ -67,9 +67,9 @@ module attributes {
}
// CHECK-DAG: spv.GlobalVariable @__workgroup_mem__{{[0-9]+}}
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<6 x i32, stride=4>)>, Workgroup>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<6 x i32>)>, Workgroup>
// CHECK-DAG: spv.GlobalVariable @__workgroup_mem__{{[0-9]+}}
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<20 x f32, stride=4>)>, Workgroup>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<20 x f32>)>, Workgroup>
// CHECK: func @two_allocs()
// -----
@ -87,9 +87,9 @@ module attributes {
}
// CHECK-DAG: spv.GlobalVariable @__workgroup_mem__{{[0-9]+}}
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<2 x vector<2xi32>, stride=8>)>, Workgroup>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<2 x vector<2xi32>>)>, Workgroup>
// CHECK-DAG: spv.GlobalVariable @__workgroup_mem__{{[0-9]+}}
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<4 x vector<4xf32>, stride=16>)>, Workgroup>
// CHECK-SAME: !spv.ptr<!spv.struct<(!spv.array<4 x vector<4xf32>>)>, Workgroup>
// CHECK: func @two_allocs_vector()

10
mlir/test/Conversion/MemRefToSPIRV/alloca.mlir
View File

@ -10,7 +10,7 @@ module attributes {spv.target_env = #spv.target_env<#spv.vce<v1.3, [Shader], []>
}
// CHECK-LABEL: func @alloc_function_variable
// CHECK: %[[VAR:.+]] = spv.Variable : !spv.ptr<!spv.struct<(!spv.array<20 x f32, stride=4>)>, Function>
// CHECK: %[[VAR:.+]] = spv.Variable : !spv.ptr<!spv.struct<(!spv.array<20 x f32>)>, Function>
// CHECK: %[[LOADPTR:.+]] = spv.AccessChain %[[VAR]]
// CHECK: %[[VAL:.+]] = spv.Load "Function" %[[LOADPTR]] : f32
// CHECK: %[[STOREPTR:.+]] = spv.AccessChain %[[VAR]]
@ -28,8 +28,8 @@ module attributes {spv.target_env = #spv.target_env<#spv.vce<v1.3, [Shader], []>
}
// CHECK-LABEL: func @two_allocs
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<6 x i32, stride=4>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<20 x f32, stride=4>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<6 x i32>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<20 x f32>)>, Function>
// -----
@ -42,8 +42,8 @@ module attributes {spv.target_env = #spv.target_env<#spv.vce<v1.3, [Shader], []>
}
// CHECK-LABEL: func @two_allocs_vector
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<2 x vector<2xi32>, stride=8>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<4 x vector<4xf32>, stride=16>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<2 x vector<2xi32>>)>, Function>
// CHECK-DAG: spv.Variable : !spv.ptr<!spv.struct<(!spv.array<4 x vector<4xf32>>)>, Function>
// -----

2
mlir/test/Conversion/TensorToSPIRV/tensor-ops-to-spirv.mlir
View File

@ -9,7 +9,7 @@
func.func @tensor_extract_constant(%a : index, %b: index, %c: index) -> i32 {
// CHECK: %[[CST:.+]] = spv.Constant dense<[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]>
%cst = arith.constant dense<[[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]]> : tensor<2x2x3xi32>
// CHECK: %[[VAR:.+]] = spv.Variable init(%[[CST]]) : !spv.ptr<!spv.array<12 x i32, stride=4>, Function>
// CHECK: %[[VAR:.+]] = spv.Variable init(%[[CST]]) : !spv.ptr<!spv.array<12 x i32>, Function>
// CHECK: %[[C0:.+]] = spv.Constant 0 : i32
// CHECK: %[[C6:.+]] = spv.Constant 6 : i32
// CHECK: %[[MUL0:.+]] = spv.IMul %[[C6]], %[[A]] : i32