[WA] Fix the lack of exceptions vector in getrf_batch #1916
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Plan to remove this WA after upgrading to oneMKL 2025.3. |
There was a problem hiding this comment.
Pull Request Overview
This PR fixes a segmentation fault issue in the MKL batch linear algebra error handling by providing a workaround for missing exception vectors in batch_error exceptions. The fix includes adding debug output when exceptions are missing and creating a CPU-based temporary tensor for info handling.
- Adds error handling for cases where batch_error.exceptions() returns an empty vector
- Modifies lu_factor_mkl to use a CPU-based temporary tensor for info data instead of zeroing the original tensor
- Includes debug output to log batch error details when exceptions are unavailable
CuiYifeng
force-pushed
the
yifeng/getrf_batch_wa
branch
from
15dac16 to
29605c0
Compare
chunhuanMeng
approved these changes
Aug 11, 2025
jianyizh
reviewed
Aug 11, 2025
| info.zero_(); | ||
| int32_t* infos_data = info.data_ptr<int32_t>(); | ||
| Tensor info_ = at::zeros_like(info, Device(at::kCPU)); | ||
| int32_t* infos_data = info_.data_ptr<int32_t>(); |
There was a problem hiding this comment.
can you provide more explanation here? why you create a zero tensor on cpu and pass it to apply_lu_xpu_？ and then you copy info_ to xpu? mkl will get seg fault when info is on xpu?
There was a problem hiding this comment.
On stock Pytorch side, original info Tensor is on device instead of host.
jianyizh
approved these changes
Aug 11, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR provides a work-around for the lack of exception vector within batch_error exception, otherwise segmentation fault would happen while accessing exceptions.