-
Notifications
You must be signed in to change notification settings - Fork 219
/
test_cuda_shared_memory.py
168 lines (132 loc) · 6.69 KB
/
test_cuda_shared_memory.py
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
# Copyright 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
import unittest
import numpy
# Torch support read / write DLPack object on GPU
import torch
import tritonclient.utils as utils
import tritonclient.utils.cuda_shared_memory as cudashm
class DLPackTest(unittest.TestCase):
"""
Testing DLPack implementation in CUDA shared memory utilities
"""
def test_from_gpu(self):
# Create GPU tensor via PyTorch and CUDA shared memory region with
# enough space
gpu_tensor = torch.ones(4, 4).cuda(0)
byte_size = 64
shm_handle = cudashm.create_shared_memory_region("cudashm_data", byte_size, 0)
# Set data from DLPack specification of PyTorch tensor
cudashm.set_shared_memory_region_from_dlpack(shm_handle, [gpu_tensor])
# Make sure the DLPack specification of the shared memory region can
# be consumed by PyTorch
smt = cudashm.as_shared_memory_tensor(shm_handle, "FP32", [4, 4])
generated_torch_tensor = torch.from_dlpack(smt)
self.assertTrue(torch.allclose(gpu_tensor, generated_torch_tensor))
cudashm.destroy_shared_memory_region(shm_handle)
def test_from_cpu(self):
# Create CPU tensor via numpy and CUDA shared memory region with
# enough space
cpu_tensor = numpy.ones([4, 4], dtype=numpy.float32)
byte_size = 64
shm_handle = cudashm.create_shared_memory_region("cudashm_data", byte_size, 0)
# Set data from DLPack specification of Numpy array
cudashm.set_shared_memory_region_from_dlpack(shm_handle, [cpu_tensor])
# Make sure the DLPack specification of the shared memory region can
# be consumed by PyTorch.
# Need to pass to PyTorch first as numpy doesn't consume GPU DLPack
smt = cudashm.as_shared_memory_tensor(shm_handle, "FP32", [4, 4])
generated_torch_tensor = torch.from_dlpack(smt)
self.assertTrue(
numpy.allclose(cpu_tensor, numpy.from_dlpack(generated_torch_tensor.cpu()))
)
cudashm.destroy_shared_memory_region(shm_handle)
class NumpyTest(unittest.TestCase):
"""
Testing Numpy implementation in CUDA shared memory utilities
"""
def test_from_numpy(self):
# Retrieve data from shared memory region via DLPack as it has been
# verified in 'DLPackTest'
# Create CPU tensor via numpy and CUDA shared memory region with
# enough space
cpu_tensor = numpy.ones([4, 4], dtype=numpy.float32)
byte_size = 64
shm_handle = cudashm.create_shared_memory_region("cudashm_data", byte_size, 0)
# Set data from Numpy array
cudashm.set_shared_memory_region(shm_handle, [cpu_tensor])
# Make sure the DLPack specification of the shared memory region can
# be consumed by PyTorch.
# Need to pass to PyTorch first as numpy doesn't consume GPU DLPack
smt = cudashm.as_shared_memory_tensor(shm_handle, "FP32", [4, 4])
generated_torch_tensor = torch.from_dlpack(smt)
self.assertTrue(
numpy.allclose(cpu_tensor, numpy.from_dlpack(generated_torch_tensor.cpu()))
)
cudashm.destroy_shared_memory_region(shm_handle)
def test_to_numpy(self):
# Retrieve data from shared memory region via DLPack as it has been
# verified in 'DLPackTest'
# Create CPU tensor via numpy and CUDA shared memory region with
# enough space
cpu_tensor = numpy.ones([4, 4], dtype=numpy.float32)
byte_size = 64
shm_handle = cudashm.create_shared_memory_region("cudashm_data", byte_size, 0)
# Set data from Numpy array
cudashm.set_shared_memory_region(shm_handle, [cpu_tensor])
# Make sure the DLPack specification of the shared memory region can
# be consumed by PyTorch.
# Need to pass to PyTorch first as numpy doesn't consume GPU DLPack
generated_tensor = cudashm.get_contents_as_numpy(
shm_handle, numpy.float32, [4, 4]
)
self.assertTrue(numpy.allclose(cpu_tensor, generated_tensor))
cudashm.destroy_shared_memory_region(shm_handle)
def test_numpy_bytes(self):
int_tensor = numpy.arange(start=0, stop=16, dtype=numpy.int32)
bytes_tensor = numpy.array(
[str(x).encode("utf-8") for x in int_tensor.flatten()], dtype=object
)
bytes_tensor = bytes_tensor.reshape(int_tensor.shape)
bytes_tensor_serialized = utils.serialize_byte_tensor(bytes_tensor)
byte_size = utils.serialized_byte_size(bytes_tensor_serialized)
shm_handle = cudashm.create_shared_memory_region("cudashm_data", byte_size, 0)
# Set data from Numpy array
cudashm.set_shared_memory_region(shm_handle, [bytes_tensor_serialized])
# Make sure the DLPack specification of the shared memory region can
# be consumed by PyTorch.
# Need to pass to PyTorch first as numpy doesn't consume GPU DLPack
generated_tensor = cudashm.get_contents_as_numpy(
shm_handle,
numpy.object_,
[
16,
],
)
self.assertTrue(numpy.array_equal(bytes_tensor, generated_tensor))
cudashm.destroy_shared_memory_region(shm_handle)
if __name__ == "__main__":
unittest.main()