Fix non-contiguous tensor problem in keyed_jagged_index_select_dim1 #2061
Conversation
This file contains 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
✅ Deploy Preview for pytorch-fbgemm-docs canceled.
|
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
sryap
force-pushed
the
export-D49939713
branch
from
e0a88d2 to
31126e7
Compare
sryap
pushed a commit
to sryap/FBGEMM
that referenced
this pull request
Oct 9, 2023
…ytorch#2061) Summary: Before this diff, `keyed_jagged_index_select_dim1` kernels take raw pointers as arguments. This requires the input tensors to be contiguous. However, the `keyed_jagged_index_select_dim1` operator did not make sure that the tensors are contiguous before extracting and passing the raw pointers to the kernels causing the correctness issue. This diff replaces the raw pointer arguments with PyTorch's `PackedTensorAccessor` which handles non-contiguous tensor accesses automatically. For some tensors that their raw pointers are still being used, the operator makes sure that the tensors are contiguous before using them. Reviewed By: jasonjk-park Differential Revision: D49939713
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
sryap
pushed a commit
to sryap/FBGEMM
that referenced
this pull request
Oct 9, 2023
…ytorch#2061) Summary: Before this diff, `keyed_jagged_index_select_dim1` kernels take raw pointers as arguments. This requires the input tensors to be contiguous. However, the `keyed_jagged_index_select_dim1` operator did not make sure that the tensors are contiguous before extracting and passing the raw pointers to the kernels causing the correctness issue. This diff replaces the raw pointer arguments with PyTorch's `PackedTensorAccessor` which handles non-contiguous tensor accesses automatically. For some tensors that their raw pointers are still being used, the operator makes sure that the tensors are contiguous before using them. Reviewed By: jasonjk-park Differential Revision: D49939713
sryap
force-pushed
the
export-D49939713
branch
from
31126e7 to
e3d38ba
Compare
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
…ytorch#2061) Summary: Before this diff, `keyed_jagged_index_select_dim1` kernels take raw pointers as arguments. This requires the input tensors to be contiguous. However, the `keyed_jagged_index_select_dim1` operator did not make sure that the tensors are contiguous before extracting and passing the raw pointers to the kernels causing the correctness issue. This diff replaces the raw pointer arguments with PyTorch's `PackedTensorAccessor` which handles non-contiguous tensor accesses automatically. For some tensors that their raw pointers are still being used, the operator makes sure that the tensors are contiguous before using them. Reviewed By: jasonjk-park, venkatrsrinivas Differential Revision: D49939713
sryap
force-pushed
the
export-D49939713
branch
from
e3d38ba to
3111b77
Compare
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
1 similar comment
|
This pull request was exported from Phabricator. Differential Revision: D49939713 |
|
This pull request has been merged in 1e194b7. |
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.
Summary:
Before this diff,
keyed_jagged_index_select_dim1kernels take rawpointers as arguments. This requires the input tensors to be
contiguous. However, the
keyed_jagged_index_select_dim1operatordid not make sure that the tensors are contiguous before extracting
and passing the raw pointers to the kernels causing the correctness
issue. This diff replaces the raw pointer arguments with PyTorch's
PackedTensorAccessorwhich handles non-contiguous tensor accessesautomatically. For some tensors that their raw pointers are still
being used, the operator makes sure that the tensors are contiguous
before using them.
Differential Revision: D49939713