Fix missing strides validation in dpnp.tensor.usm_ndarray#2927
Open
vlad-perevezentsev wants to merge 5 commits into
Open
Fix missing strides validation in dpnp.tensor.usm_ndarray#2927vlad-perevezentsev wants to merge 5 commits into
dpnp.tensor.usm_ndarray#2927vlad-perevezentsev wants to merge 5 commits into
Conversation
vlad-perevezentsev
requested review from
antonwolfy and
ndgrigorian
as code owners
May 20, 2026 12:19
Contributor
|
View rendered docs @ https://intelpython.github.io/dpnp/pull/2927/index.html |
Contributor
|
Array API standard conformance tests for dpnp=0.21.0dev3=py314h509198e_16 ran successfully. |
vlad-perevezentsev
marked this pull request as draft
May 20, 2026 14:54
ndgrigorian
reviewed
May 20, 2026
ndgrigorian
reviewed
May 20, 2026
7 tasks
Contributor
Author
|
#2939 to fix fft falling test |
vlad-perevezentsev
added a commit
that referenced
this pull request
Jun 26, 2026
…y allocation (#2939) This PR fixes an issue discovered while implementing #2927 where several FFT tests started failing after adding validation for stride configurations with oversized memory footprints. The problem was that FFT could allocate output arrays using the same strided layout as a non-contiguous input. For inputs such as `a[::2`, this resulted in oversized allocations followed by an additional copy to a contiguous array. This fix checks whether the memory footprint implied by the input strides exceeds the number of elements in the array. If so, the input is copied to a contiguous layout before configuring the FFT descriptor allowing oneMKL FFT to produce a contiguous output directly By avoiding oversized allocations and the extra copy this significantly improves the performance of all FFT operations in dpnp with complex strided inputs Additionally, the cuFFT C-contiguous requirement is now enforced for all code paths. Previously it was only applied within `batch_fft` branch which was a latent bug for 1D non-batch inputs on CUDA.
github-actions Bot
added a commit
that referenced
this pull request
Jun 26, 2026
…y allocation (#2939) This PR fixes an issue discovered while implementing #2927 where several FFT tests started failing after adding validation for stride configurations with oversized memory footprints. The problem was that FFT could allocate output arrays using the same strided layout as a non-contiguous input. For inputs such as `a[::2`, this resulted in oversized allocations followed by an additional copy to a contiguous array. This fix checks whether the memory footprint implied by the input strides exceeds the number of elements in the array. If so, the input is copied to a contiguous layout before configuring the FFT descriptor allowing oneMKL FFT to produce a contiguous output directly By avoiding oversized allocations and the extra copy this significantly improves the performance of all FFT operations in dpnp with complex strided inputs Additionally, the cuFFT C-contiguous requirement is now enforced for all code paths. Previously it was only applied within `batch_fft` branch which was a latent bug for 1D non-batch inputs on CUDA. 77e000b
vlad-perevezentsev
marked this pull request as ready for review
July 23, 2026 14:22
Collaborator
|
coverage: 78.375% (-0.01%) from 78.386% — extend_stride_validation into master |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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 proposes to fix #2881 issue.
The problem was that
usm_ndarrayconstructor did not validate user-provided strides when allocating new memory.This made it possible to create arrays with invalid stride patterns producing negative memory displacement or oversized memory footprints.
This PR adds validation for invalid stride configurations when allocating new memory and raises
ValueErrorfor unsupported strides.It also adds a new
test_ctor_invalid_stridesintest_usm_ndarray_ctor.pyto cover these cases