Project-MONAI / Project-MONAI/MONAI
Random Segmentation fault in test_masked_autoencoder_vit
Open
Nobody has claimed this yet.
bug
- Dominant language
- Python
- Stars
- 8.7k
- Forks
- 1.6k
- Avg merge
- 5d 1h
- Merged PRs (30d)
- 20
Description
[2025-06-17T20:37:02.135Z] Starting test: test_access_attn_matrix (tests.test_masked_autoencoder_vit.TestMaskedAutoencoderViT)...
[2025-06-17T20:37:02.135Z] 2025-06-17 20:37:02,024 - py.warnings - WARNING - torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at /opt/pytorch/pytorch/aten/src/ATen/native/TensorShape.cpp:3593.)
[2025-06-17T20:37:02.135Z]
[2025-06-17T20:37:02.135Z] [monai-latest-image-24-08-a100-jenkins-monai-latest-image--qxv1d:8707 :0:8707] Caught signal 11 (Segmentation fault: address not mapped to object at address (nil))
[2025-06-17T20:37:02.135Z] [1750192622.043102] [monai-latest-image-24-08-a100-jenkins-monai-latest-image--qxv1d:8707 :1] debug.c:1297 UCX WARN ucs_debug_disable_signal: signal 8 was not set in ucs
[2025-06-17T20:37:02.135Z] [monai-latest-image-24-08-a100-jenkins-monai-latest-image--qxv1d:8707 :1:8731] Caught signal 11 (Segmentation fault: address not mapped to object at address (nil))
[2025-06-17T20:37:02.135Z] ==== backtrace (tid: 8707) ====
[2025-06-17T20:37:02.135Z] 0 0x0000000000042520 __sigaction() ???:0
[2025-06-17T20:37:02.135Z] =================================
[2025-06-17T20:37:02.698Z] ./runtests.sh: line 724: 8707 Segmentation fault
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with tests/test_masked_autoencoder_vit.py and run test_access_attn_matrix to determine whether the segmentation fault is reproducible. Review runtests.sh around the reported failure and the test's execution environment, including the PyTorch warning and UCX output. Done means identifying a reproducible cause or clearly documenting the conditions that trigger the fault.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, testing-qa
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100