File: inspect_gpu.cc

package info (click to toggle)
pytorch 1.13.1%2Bdfsg-4
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 139,252 kB
  • sloc: cpp: 1,100,274; python: 706,454; ansic: 83,052; asm: 7,618; java: 3,273; sh: 2,841; javascript: 612; makefile: 323; xml: 269; ruby: 185; yacc: 144; objc: 68; lex: 44
file content (58 lines) | stat: -rw-r--r-- 1,664 bytes parent folder | download | duplicates (4)
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
/**
 * Copyright (c) 2016-present, Facebook, Inc.
 *
 * 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.
 */

#include <cuda_runtime.h>

#include <sstream>
#include <vector>

#include "c10/util/Flags.h"
#include "caffe2/core/common_gpu.h"
#include "caffe2/core/init.h"
#include "caffe2/core/logging.h"

using std::vector;

C10_DECLARE_int(caffe2_log_level);

int main(int argc, char** argv) {
  caffe2::GlobalInit(&argc, &argv);
  c10::SetUsageMessage(
      "Inspects the GPUs on the current machine and prints out their details "
      "provided by cuda.");

  int gpu_count;
  CUDA_ENFORCE(cudaGetDeviceCount(&gpu_count));
  for (int i = 0; i < gpu_count; ++i) {
    LOG(INFO) << "Querying device ID = " << i;
    caffe2::DeviceQuery(i);
  }

  vector<vector<bool> > access_pattern;
  CAFFE_ENFORCE(caffe2::GetCudaPeerAccessPattern(&access_pattern));

  std::stringstream sstream;
  // Find topology
  for (int i = 0; i < gpu_count; ++i) {
    for (int j = 0; j < gpu_count; ++j) {
      sstream << (access_pattern[i][j] ? "+" : "-") << " ";
    }
    sstream << std::endl;
  }
  LOG(INFO) << "Access pattern: " << std::endl << sstream.str();

  return 0;
}