forked from tensorflow/tfjs
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathbackend_webgpu_test.ts
261 lines (207 loc) · 10.2 KB
/
backend_webgpu_test.ts
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
/**
* @license
* Copyright 2019 Google LLC. 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
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* =============================================================================
*/
import * as tf from '@tensorflow/tfjs-core';
import glslangInit from '@webgpu/glslang/dist/web-devel/glslang.onefile';
import {WebGPUBackend, WebGPUMemoryInfo} from './backend_webgpu';
import {describeWebGPU} from './test_util';
describeWebGPU('backend webgpu cpu forwarding turned on', () => {
let cpuForwardFlagSaved: boolean;
beforeAll(() => {
cpuForwardFlagSaved = tf.env().getBool('WEBGPU_CPU_FORWARD');
tf.env().set('WEBGPU_CPU_FORWARD', true);
});
afterAll(() => {
tf.env().set('WEBGPU_CPU_FORWARD', cpuForwardFlagSaved);
});
it('should not allocate GPU memory when CPU forwarding', async () => {
const savedFlag = tf.env().get('WEBGPU_IMMEDIATE_EXECUTION_ENABLED');
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', true);
const a = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c = tf.mul(a, b);
const startNumBytes = tf.memory().numBytes;
const startNumTensors = tf.memory().numTensors;
const startNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
expect(startNumBytes).toEqual(48);
expect(startNumTensors).toEqual(3);
expect(startNumBytesInGPU).toEqual(0);
const f = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
const d = tf.matMul(c, f);
const dData = await d.data();
const endNumBytes = tf.memory().numBytes;
const endNumTensors = tf.memory().numTensors;
const endNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
expect(endNumBytes - startNumBytes).toEqual(48);
expect(endNumTensors - startNumTensors).toEqual(2);
expect(endNumBytesInGPU - startNumBytesInGPU).toEqual(40);
tf.test_util.expectArraysClose(
dData, new Float32Array([9, 12, 15, 19, 26, 33]));
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', savedFlag);
});
});
describeWebGPU('backend webgpu', () => {
it('should not leak memory in delayed mode', async () => {
const savedFlag = tf.env().get('WEBGPU_IMMEDIATE_EXECUTION_ENABLED');
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', false);
const a = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c = tf.mul(a, b);
const startNumBytes = tf.memory().numBytes;
const startNumTensors = tf.memory().numTensors;
const startNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
const f = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
const d = tf.matMul(c, f);
const dData = await d.data();
const endNumBytes = tf.memory().numBytes;
const endNumTensors = tf.memory().numTensors;
const endNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
expect(endNumBytes - startNumBytes).toEqual(48);
expect(endNumTensors - startNumTensors).toEqual(2);
expect(endNumBytesInGPU - startNumBytesInGPU).toEqual(0);
tf.test_util.expectArraysClose(
dData, new Float32Array([9, 12, 15, 19, 26, 33]));
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', savedFlag);
});
it('should not leak memory in immediate mode', async () => {
const savedFlag = tf.env().get('WEBGPU_IMMEDIATE_EXECUTION_ENABLED');
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', true);
const a = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c = tf.mul(a, b);
const startNumBytes = tf.memory().numBytes;
const startNumTensors = tf.memory().numTensors;
const startNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
const f = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
const d = tf.matMul(c, f);
const dData = await d.data();
const endNumBytes = tf.memory().numBytes;
const endNumTensors = tf.memory().numTensors;
const endNumBytesInGPU = (tf.memory() as WebGPUMemoryInfo).numBytesInGPU;
expect(endNumBytes - startNumBytes).toEqual(48);
expect(endNumTensors - startNumTensors).toEqual(2);
expect(endNumBytesInGPU - startNumBytesInGPU).toEqual(24);
tf.test_util.expectArraysClose(
dData, new Float32Array([9, 12, 15, 19, 26, 33]));
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', savedFlag);
});
it('should recycle buffers in immediate mode', () => {
const savedFlag = tf.env().get('WEBGPU_IMMEDIATE_EXECUTION_ENABLED');
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', true);
const backend = tf.backend() as WebGPUBackend;
const bufferManager = backend.getBufferManager();
bufferManager.reset();
const a = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c = tf.mul(a, b);
const freeBuffersAfterFirstMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterFirstMul = bufferManager.getNumUsedBuffers();
const f = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
tf.matMul(c, f);
const freeBuffersAfterFirstMatMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterFirstMatMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterFirstMatMul - freeBuffersAfterFirstMul)
.toEqual(1); // from released uniform
expect(usedBuffersAfterFirstMatMul - usedBuffersAfterFirstMul).toEqual(2);
const a2 = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b2 = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c2 = tf.mul(a2, b2);
const freeBuffersAfterSecondMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterSecondMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterSecondMul - freeBuffersAfterFirstMatMul)
.toEqual(0); // released a uniform buffer and reused a buffer
expect(usedBuffersAfterSecondMul - usedBuffersAfterFirstMatMul).toEqual(3);
const f2 = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
tf.matMul(c2, f2);
const freeBuffersAfterSecondMatMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterSecondMatMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterSecondMatMul - freeBuffersAfterSecondMul).toEqual(0);
expect(usedBuffersAfterSecondMatMul - usedBuffersAfterSecondMul).toEqual(2);
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', savedFlag);
});
it('should not recycle buffers in delayed mode', async () => {
const savedFlag = tf.env().get('WEBGPU_IMMEDIATE_EXECUTION_ENABLED');
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', false);
const backend = tf.backend() as WebGPUBackend;
const bufferManager = backend.getBufferManager();
bufferManager.reset();
const a = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c = tf.mul(a, b);
const freeBuffersAfterFirstMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterFirstMul = bufferManager.getNumUsedBuffers();
const f = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
tf.matMul(c, f);
const freeBuffersAfterFirstMatMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterFirstMatMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterFirstMatMul - freeBuffersAfterFirstMul).toEqual(0);
expect(usedBuffersAfterFirstMatMul - usedBuffersAfterFirstMul).toEqual(3);
const a2 = tf.tensor2d([2, 4, 6, 8], [2, 2]);
const b2 = tf.tensor2d([0.5, 0.5, 0.5, 0.5], [2, 2]);
const c2 = tf.mul(a2, b2);
const freeBuffersAfterSecondMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterSecondMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterSecondMul - freeBuffersAfterFirstMatMul).toEqual(0);
expect(usedBuffersAfterSecondMul - usedBuffersAfterFirstMatMul).toEqual(4);
const f2 = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
const c3 = tf.matMul(c2, f2);
const freeBuffersAfterSecondMatMul = bufferManager.getNumFreeBuffers();
const usedBuffersAfterSecondMatMul = bufferManager.getNumUsedBuffers();
expect(freeBuffersAfterSecondMatMul - freeBuffersAfterSecondMul).toEqual(0);
expect(usedBuffersAfterSecondMatMul - usedBuffersAfterSecondMul).toEqual(3);
// Tests happen within a tidy so we need to read a tensor at the end of a
// test in delayed mode in order to force flush the disposal queue.
await c3.data();
tf.env().set('WEBGPU_IMMEDIATE_EXECUTION_ENABLED', savedFlag);
});
it('readSync should throw if tensors are on the GPU', async () => {
const a = tf.tensor2d([1, 2, 3, 4], [2, 2]);
const b = tf.tensor2d([1, 2, 3, 4, 5, 6], [2, 3]);
const c = tf.matMul(a, b);
expect(() => c.dataSync())
.toThrowError(
'WebGPU readSync is only available for CPU-resident tensors.');
await c.data();
// Now that data has been downloaded to the CPU, dataSync should work.
expect(() => c.dataSync()).not.toThrow();
});
it('lazily upload', async () => {
const glslang = await glslangInit();
const adapter = await navigator.gpu.requestAdapter({});
const device = await adapter.requestDevice({});
const backend = new WebGPUBackend(device, glslang);
tf.registerBackend('test-storage', () => backend);
tf.setBackend('test-storage');
const bufferManager = backend.getBufferManager();
const t = tf.tensor1d([1, 2, 3], 'float32');
expect(bufferManager.getNumUsedBuffers()).toBe(0);
backend.getBuffer(t.dataId);
expect(bufferManager.getNumUsedBuffers()).toBe(1);
});
it('should be possible to move data from webgl to webgpu', async () => {
tf.setBackend('webgl');
const a = tf.randomNormal([1, 65, 65, 256]);
const b = tf.randomNormal([1, 65, 65, 256]);
const c = tf.add(a, b);
await c.data();
const f = async () => {
tf.setBackend('webgpu');
const d = tf.add(a, b);
await d.data();
};
expect(f).not.toThrow();
});
});