Fixed regression in iteration space simplifier, fixed arguments of numpy to usm_ndarray copying function #1202
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
For nd==1, and when simplifying iteration over multiple arrays, strides should only change sign if all strides are negative.
When figuring out the bytes-boundary of numpy array to form the sycl::buffer we use simplified strides and shapes, and hence must start with src_offset, not with zero. Now, both of these copy calls correctly populate the destination usm_ndarray. ``` In [2]: n, m = 10**2, 10**3 In [3]: a = np.hstack((np.arange(n, dtype=np.int32),) * m) In [4]: b = np.flipud(a) In [5]: In [5]: c = dpt.asarray(b, device='cpu') In [6]: c Out[6]: usm_ndarray([99, 98, 97, ..., 2, 1, 0], dtype=int32) In [7]: import dpctl.tensor._tensor_impl as ti In [8]: c = dpt.empty(b.shape, dtype=b.dtype)[::-1] In [9]: ti._copy_numpy_ndarray_into_usm_ndarray(src=b, dst=c, sycl_queue=c.sycl_queue) In [10]: c Out[10]: usm_ndarray([99, 98, 97, ..., 2, 1, 0], dtype=int32) In [11]: c = dpt.empty(b.shape, dtype=b.dtype) In [12]: ti._copy_numpy_ndarray_into_usm_ndarray(src=b, dst=c, sycl_queue=c.sycl_queue) In [13]: c Out[13]: usm_ndarray([99, 98, 97, ..., 2, 1, 0], dtype=int32) ```
|
View rendered docs @ https://intelpython.github.io/dpctl/pulls/1202/index.html |
|
Array API standard conformance tests for dpctl=0.14.3dev1=py310h76be34b_48 ran successfully. |
1 similar comment
|
Array API standard conformance tests for dpctl=0.14.3dev1=py310h76be34b_48 ran successfully. |
antonwolfy
reviewed
May 8, 2023
|
Deleted rendered PR docs from intelpython.github.com/dpctl, latest should be updated shortly. 🤞 |
|
Array API standard conformance tests for dpctl=0.14.3dev1=py310h76be34b_48 ran successfully. |
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 fixes issues in
simplify_iteration_space*functions introduced in gh-1198.It also corrects calculation of numpy array byte-boundaries to form
sycl::bufferaccessingcontent of NumPy's array from a SYCL kernel.
Closes gh-1201