SYCL : support non-contiguous tensors in binary ops (add, sub, etc) #12399
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
github-actions
bot
added
ggml
aubreyli
reviewed
Mar 15, 2025
| @@ -625,6 +638,26 @@ struct bin_bcast_sycl { | |||
| size_t s12 = nb12 / sizeof(src1_t); | |||
| size_t s13 = nb13 / sizeof(src1_t); | |||
|
|
|||
| size_t s00 = nb00 / sizeof(src0_t); | |||
There was a problem hiding this comment.
I saw a bunch of warning of s00, do you want to fix it?
/home/aubrey/work/llama-gpu/ggml/src/ggml-sycl/common.hpp:641:20: warning: unused variable 's00' [-Wunused-variable]
641 | size_t s00 = nb00 / sizeof(src0_t);
There was a problem hiding this comment.
Thx, they got lost in a sea of deprecation warnings.
|
This PR looks good to me now, thanks @fairydreaming for your excellent work! |
|
I test the CI of this PR! all passed! Thank you! |
arthw
approved these changes
Mar 15, 2025
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Mar 19, 2025
…gml-org#12399) * sycl : support non-contiguous tensors in binary ops * sycl : silence unused variable warning --------- Co-authored-by: Stanisław Szymczyk <[email protected]>
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Outdated suggestions cannot be applied.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 adds support for non-contiguous tensors in SYCL backend binary OPs like ADD, SUB etc.
Note that I have little experience writing GPU kernels, this is basically a copy from ggml/src/ggml-cuda/binbcast.cu. I checked that test-backend-ops passes and confirmed that with this patch DeepSeek V2 Lite works correctly.
But someone more knowledgeable shall take a look and test this on Intel hardware.
Fixes #12390.