OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 12:20:52 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

Fix test failure in TestCudaMultiGPU.test_cuda_device_memory_allocated (#105501)

Browse Source 
The test

f508d3564c/test/test_cuda_multigpu.py (L1282-L1290)

Torch cuda caching allocator may cache the allocation and cause the "new_alloc" being the same as the "old_alloc".
```python
     self.assertGreater(memory_allocated(0), current_alloc[0])
```

I suggest that we use `assertGreaterEqual` instead of `assertGreater` in the test.

Individually running only this test does not make it fail but running it together with other tests from the same test module will make it fail.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/105501
Approved by: https://github.com/zou3519
This commit is contained in:
Xiao Wang 2023-07-20 19:59:07 +00:00 committed by PyTorch MergeBot
parent 6abb8c382c
commit e6fd8ca3ee
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
test/test_cuda_multigpu.py
View File

@ -1285,7 +1285,7 @@ t2.start()
device_count = torch.cuda.device_count() device_count = torch.cuda.device_count()
current_alloc = [memory_allocated(idx) for idx in range(device_count)] current_alloc = [memory_allocated(idx) for idx in range(device_count)]
x = torch.ones(10, device="cuda:0") x = torch.ones(10, device="cuda:0")
self.assertGreater(memory_allocated(0), current_alloc[0]) self.assertGreaterEqual(memory_allocated(0), current_alloc[0])
self.assertTrue(all(memory_allocated(torch.cuda.device(idx)) == current_alloc[idx] for idx in range(1, device_count))) self.assertTrue(all(memory_allocated(torch.cuda.device(idx)) == current_alloc[idx] for idx in range(1, device_count)))