1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
|
/*
* Copyright (c) 2015-2025 The Khronos Group Inc.
* Copyright (c) 2015-2025 Valve Corporation
* Copyright (c) 2015-2025 LunarG, Inc.
* Copyright (c) 2015-2024 Google, Inc.
* Modifications Copyright (C) 2020 Advanced Micro Devices, Inc. All rights reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*/
#include "../framework/layer_validation_tests.h"
#include "../framework/pipeline_helper.h"
#include "../framework/shader_object_helper.h"
#include "cooperative_matrix_helper.h"
void CooperativeMatrixTest::InitCooperativeMatrixKHR() {
AddRequiredExtensions(VK_KHR_COOPERATIVE_MATRIX_EXTENSION_NAME);
// glslang will generate OpCapability VulkanMemoryModel and need entension enabled
AddRequiredExtensions(VK_KHR_VULKAN_MEMORY_MODEL_EXTENSION_NAME);
AddRequiredFeature(vkt::Feature::cooperativeMatrix);
AddRequiredFeature(vkt::Feature::vulkanMemoryModel);
RETURN_IF_SKIP(Init());
}
class PositiveShaderCooperativeMatrix : public CooperativeMatrixTest {};
TEST_F(PositiveShaderCooperativeMatrix, CooperativeMatrixKHR) {
TEST_DESCRIPTION("Test VK_KHR_cooperative_matrix.");
SetTargetApiVersion(VK_API_VERSION_1_3);
AddRequiredFeature(vkt::Feature::shaderFloat16);
AddRequiredFeature(vkt::Feature::storageBuffer16BitAccess);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
CooperativeMatrixHelper helper(*this);
VkCooperativeMatrixPropertiesKHR subgroup_prop = vku::InitStructHelper();
bool found_scope_subgroup = false;
for (const auto &prop : helper.coop_matrix_props) {
// We only have the 16-bit features enabled, but 32-bit also works
if (prop.scope == VK_SCOPE_SUBGROUP_KHR && !helper.Has8BitComponentType(prop) && !helper.Has64BitComponentType(prop)) {
found_scope_subgroup = true;
subgroup_prop = prop;
break;
}
}
if (!found_scope_subgroup) {
GTEST_SKIP() << "VK_SCOPE_SUBGROUP_KHR not Found";
}
const vkt::DescriptorSetLayout dsl(*m_device,
{
{0, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr},
{1, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr},
{2, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr},
{3, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr},
});
std::string css = R"glsl(
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_KHR_shader_subgroup_basic : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : enable
layout(local_size_x = 64) in;
layout(set=0, binding=0) coherent buffer InputA { %type_A% x[]; } inputA;
layout(set=0, binding=1) coherent buffer InputB { %type_B% x[]; } inputB;
layout(set=0, binding=2) coherent buffer InputC { %type_C% x[]; } inputC;
layout(set=0, binding=3) coherent buffer Output { %type_R% x[]; } outputO;
coopmat<%type_A%, gl_ScopeSubgroup, %M%, %K%, gl_MatrixUseA> matA;
coopmat<%type_B%, gl_ScopeSubgroup, %K%, %N%, gl_MatrixUseB> matB;
coopmat<%type_C%, gl_ScopeSubgroup, %M%, %N%, gl_MatrixUseAccumulator> matC;
coopmat<%type_R%, gl_ScopeSubgroup, %M%, %N%, gl_MatrixUseAccumulator> matO;
void main()
{
coopMatLoad(matA, inputA.x, 0, %M%, gl_CooperativeMatrixLayoutRowMajor);
coopMatLoad(matB, inputB.x, 0, %K%, gl_CooperativeMatrixLayoutRowMajor);
coopMatLoad(matC, inputC.x, 0, %M%, gl_CooperativeMatrixLayoutRowMajor);
matO = coopMatMulAdd(matA, matB, matC);
coopMatStore(matO, outputO.x, 0, %M%, gl_CooperativeMatrixLayoutRowMajor);
}
)glsl";
auto replace = [](std::string &str, const std::string &from, const std::string &to) {
size_t pos;
while ((pos = str.find(from)) != std::string::npos) str.replace(pos, from.length(), to);
};
replace(css, "%M%", std::to_string(subgroup_prop.MSize));
replace(css, "%N%", std::to_string(subgroup_prop.NSize));
replace(css, "%K%", std::to_string(subgroup_prop.KSize));
replace(css, "%type_A%", helper.VkComponentTypeToGLSL(subgroup_prop.AType));
replace(css, "%type_B%", helper.VkComponentTypeToGLSL(subgroup_prop.BType));
replace(css, "%type_C%", helper.VkComponentTypeToGLSL(subgroup_prop.CType));
replace(css, "%type_R%", helper.VkComponentTypeToGLSL(subgroup_prop.ResultType));
CreateComputePipelineHelper pipe(*this);
pipe.cs_ = VkShaderObj(*m_device, css.c_str(), VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_3);
pipe.pipeline_layout_ = vkt::PipelineLayout(*m_device, {&dsl});
pipe.CreateComputePipeline();
m_errorMonitor->VerifyFound();
}
TEST_F(PositiveShaderCooperativeMatrix, RequiredSubgroupSize) {
TEST_DESCRIPTION("https://github.com/KhronosGroup/Vulkan-ValidationLayers/issues/9843");
SetTargetApiVersion(VK_API_VERSION_1_3);
AddRequiredFeature(vkt::Feature::shaderFloat16);
AddRequiredFeature(vkt::Feature::storageBuffer16BitAccess);
AddRequiredFeature(vkt::Feature::subgroupSizeControl);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
if (!IsPlatformMockICD()) {
GTEST_SKIP() << "This makes assumption about possible coop matrix subgroup size and support.";
}
const vkt::DescriptorSetLayout dsl(*m_device, {0, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr});
const vkt::PipelineLayout pipeline_layout(*m_device, {&dsl});
const char *cs_source = R"glsl(
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_KHR_shader_subgroup_basic : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : enable
layout(local_size_x = 16) in;
layout(set=0, binding=0) coherent buffer InputA { uint32_t x[]; } inputA;
coopmat<uint32_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> matA;
void main() {
coopMatLoad(matA, inputA.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
}
)glsl";
VkShaderObj cs(*m_device, cs_source, VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_3);
VkPhysicalDeviceSubgroupSizeControlPropertiesEXT subgroup_properties = vku::InitStructHelper();
VkPhysicalDeviceVulkan11Properties props11 = vku::InitStructHelper(&subgroup_properties);
GetPhysicalDeviceProperties2(props11);
if ((subgroup_properties.requiredSubgroupSizeStages & VK_SHADER_STAGE_COMPUTE_BIT) == 0) {
GTEST_SKIP() << "Required shader stage not present in requiredSubgroupSizeStages";
}
if (subgroup_properties.minSubgroupSize != 16) {
GTEST_SKIP() << "Testing when we go under the limit";
}
VkPipelineShaderStageRequiredSubgroupSizeCreateInfo subgroup_size_control = vku::InitStructHelper();
subgroup_size_control.requiredSubgroupSize = subgroup_properties.minSubgroupSize;
CreateComputePipelineHelper pipe(*this);
pipe.cp_ci_.stage = cs.GetStageCreateInfo();
pipe.cp_ci_.stage.pNext = &subgroup_size_control;
pipe.cp_ci_.layout = pipeline_layout;
pipe.CreateComputePipeline(false);
}
TEST_F(PositiveShaderCooperativeMatrix, RequiredVulkanVersionPipeline) {
TEST_DESCRIPTION("https://gitlab.khronos.org/spirv/SPIR-V/-/issues/847");
SetTargetApiVersion(VK_API_VERSION_1_1);
AddRequiredExtensions(VK_KHR_SHADER_FLOAT16_INT8_EXTENSION_NAME);
AddRequiredExtensions(VK_EXT_SUBGROUP_SIZE_CONTROL_EXTENSION_NAME);
AddRequiredExtensions(VK_KHR_16BIT_STORAGE_EXTENSION_NAME);
AddRequiredFeature(vkt::Feature::shaderFloat16);
AddRequiredFeature(vkt::Feature::storageBuffer16BitAccess);
AddRequiredFeature(vkt::Feature::computeFullSubgroups);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
if (!IsPlatformMockICD()) {
GTEST_SKIP() << "This makes assumption about possible coop matrix subgroup size and support.";
}
const vkt::DescriptorSetLayout dsl(*m_device, {0, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr});
const vkt::PipelineLayout pipeline_layout(*m_device, {&dsl});
const char *cs_source = R"glsl(
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_KHR_shader_subgroup_basic : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : enable
layout(local_size_x = 32) in;
layout(set=0, binding=0) coherent buffer InputA { uint32_t x[]; } inputA;
coopmat<uint32_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> matA;
void main() {
coopMatLoad(matA, inputA.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
}
)glsl";
VkShaderObj cs(*m_device, cs_source, VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_1);
CreateComputePipelineHelper pipe(*this);
pipe.cp_ci_.stage = cs.GetStageCreateInfo();
pipe.cp_ci_.stage.flags = VK_PIPELINE_SHADER_STAGE_CREATE_REQUIRE_FULL_SUBGROUPS_BIT;
pipe.cp_ci_.layout = pipeline_layout;
pipe.CreateComputePipeline(false);
}
TEST_F(PositiveShaderCooperativeMatrix, RequiredVulkanVersionShaderObject) {
TEST_DESCRIPTION("https://gitlab.khronos.org/spirv/SPIR-V/-/issues/847");
SetTargetApiVersion(VK_API_VERSION_1_1);
AddRequiredExtensions(VK_KHR_SHADER_FLOAT16_INT8_EXTENSION_NAME);
AddRequiredExtensions(VK_EXT_SUBGROUP_SIZE_CONTROL_EXTENSION_NAME);
AddRequiredExtensions(VK_KHR_16BIT_STORAGE_EXTENSION_NAME);
AddRequiredExtensions(VK_EXT_SHADER_OBJECT_EXTENSION_NAME);
AddRequiredFeature(vkt::Feature::shaderObject);
AddRequiredFeature(vkt::Feature::shaderFloat16);
AddRequiredFeature(vkt::Feature::storageBuffer16BitAccess);
AddRequiredFeature(vkt::Feature::computeFullSubgroups);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
if (!IsPlatformMockICD()) {
GTEST_SKIP() << "This makes assumption about possible coop matrix subgroup size and support.";
}
const vkt::DescriptorSetLayout dsl(*m_device,
{0, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER_DYNAMIC, 1, VK_SHADER_STAGE_COMPUTE_BIT, nullptr});
const char *cs_source = R"glsl(
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_KHR_shader_subgroup_basic : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_shader_explicit_arithmetic_types_float16 : enable
layout(local_size_x = 32) in;
layout(set=0, binding=0) coherent buffer InputA { uint32_t x[]; } inputA;
coopmat<uint32_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> matA;
void main() {
coopMatLoad(matA, inputA.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
}
)glsl";
const auto spv = GLSLToSPV(VK_SHADER_STAGE_COMPUTE_BIT, cs_source, SPV_ENV_VULKAN_1_1);
auto shader_ci = ShaderCreateInfoNoNextStage(spv, VK_SHADER_STAGE_COMPUTE_BIT, 1, &dsl.handle());
shader_ci.flags = VK_SHADER_CREATE_REQUIRE_FULL_SUBGROUPS_BIT_EXT;
const vkt::Shader comp_shader(*m_device, shader_ci);
}
TEST_F(PositiveShaderCooperativeMatrix, BFloat16) {
SetTargetApiVersion(VK_API_VERSION_1_3);
AddRequiredExtensions(VK_KHR_SHADER_BFLOAT16_EXTENSION_NAME);
AddRequiredExtensions(VK_KHR_SHADER_FLOAT16_INT8_EXTENSION_NAME);
AddRequiredFeature(vkt::Feature::shaderFloat16);
AddRequiredFeature(vkt::Feature::shaderBFloat16Type);
AddRequiredFeature(vkt::Feature::shaderBFloat16CooperativeMatrix);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
const char *cs_source = R"glsl(
#version 450 core
#extension GL_EXT_bfloat16 : require
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
layout(local_size_x = 32) in;
void main() {
coopmat<bfloat16_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> cmA = coopmat<bfloat16_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA>(3.0);
}
)glsl";
CreateComputePipelineHelper pipe(*this);
pipe.cs_ = VkShaderObj(*m_device, cs_source, VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_1);
pipe.CreateComputePipeline();
}
TEST_F(PositiveShaderCooperativeMatrix, Float8) {
SetTargetApiVersion(VK_API_VERSION_1_3);
AddRequiredExtensions(VK_EXT_SHADER_FLOAT8_EXTENSION_NAME);
AddRequiredFeature(vkt::Feature::storageBuffer8BitAccess);
AddRequiredFeature(vkt::Feature::shaderFloat8);
AddRequiredFeature(vkt::Feature::shaderFloat8CooperativeMatrix);
AddRequiredFeature(vkt::Feature::shaderInt8);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
const char *cs_source = R"glsl(
#version 450 core
#extension GL_EXT_float_e4m3 : require
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
layout(local_size_x = 32) in;
void main() {
coopmat<floate4m3_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> cmA = coopmat<floate4m3_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA>(3.0);
}
)glsl";
CreateComputePipelineHelper pipe(*this);
pipe.cs_ = VkShaderObj(*m_device, cs_source, VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_1);
pipe.CreateComputePipeline();
}
TEST_F(PositiveShaderCooperativeMatrix, Int8) {
SetTargetApiVersion(VK_API_VERSION_1_3);
AddRequiredFeature(vkt::Feature::shaderInt8);
AddRequiredFeature(vkt::Feature::storageBuffer8BitAccess);
RETURN_IF_SKIP(InitCooperativeMatrixKHR());
CooperativeMatrixHelper helper(*this);
if (!helper.Has16x16UintProperty()) {
GTEST_SKIP() << "desired VkCooperativeMatrixPropertiesKHR not found";
}
OneOffDescriptorSet descriptor_set(m_device, {{0, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_ALL, nullptr},
{1, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_ALL, nullptr},
{2, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_ALL, nullptr},
{3, VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, 1, VK_SHADER_STAGE_ALL, nullptr}});
vkt::PipelineLayout pl(*m_device, {&descriptor_set.layout_});
std::string css = R"glsl(
#version 450 core
#pragma use_vulkan_memory_model
#extension GL_KHR_shader_subgroup_basic : enable
#extension GL_KHR_memory_scope_semantics : enable
#extension GL_KHR_cooperative_matrix : enable
#extension GL_EXT_shader_explicit_arithmetic_types : enable
#extension GL_EXT_shader_explicit_arithmetic_types_int8 : enable
layout(local_size_x = 64) in;
layout(set=0, binding=0) coherent buffer InputA { uint8_t x[]; } inputA;
layout(set=0, binding=1) coherent buffer InputB { uint8_t x[]; } inputB;
layout(set=0, binding=2) coherent buffer InputC { uint32_t x[]; } inputC;
layout(set=0, binding=3) coherent buffer Output { uint32_t x[]; } outputO;
coopmat<uint8_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseA> matA;
coopmat<uint8_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseB> matB;
coopmat<uint32_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseAccumulator> matC;
coopmat<uint32_t, gl_ScopeSubgroup, 16, 16, gl_MatrixUseAccumulator> matO;
void main() {
coopMatLoad(matA, inputA.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
coopMatLoad(matB, inputB.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
coopMatLoad(matC, inputC.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
matO = coopMatMulAdd(matA, matB, matC);
coopMatStore(matO, outputO.x, 0, 16, gl_CooperativeMatrixLayoutRowMajor);
}
)glsl";
CreateComputePipelineHelper pipe(*this);
pipe.cs_ = VkShaderObj(*m_device, css.c_str(), VK_SHADER_STAGE_COMPUTE_BIT, SPV_ENV_VULKAN_1_3);
pipe.cp_ci_.layout = pl;
pipe.CreateComputePipeline();
m_errorMonitor->VerifyFound();
}
|
