feat[dace][next]: Adjusted the handling of non standard GPU Memlets. #1913
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
Once [PR#1976](spcl/dace#1976) is merged in DaCe the code generator is able to handle more Memlets directly as Cuda `memcpy()` calls. This PR modifies the GPU transformation of GT4Py in such a way that these Memlets are no longer transformed into Maps. However, it can only be merged if the DaCe dependency was bumped to a version that includes PR#1976!
edopao
reviewed
Mar 18, 2025
src/gt4py/next/program_processors/runners/dace/transformations/gpu_utils.py
Show resolved
Hide resolved
edopao
approved these changes
Apr 15, 2025
Contributor
edopao
left a comment
edopao
left a comment
There was a problem hiding this comment.
The GPU tests pass, although we are on a dace baseline that does not include the PR mentioned in the PR description. I suspect this is because gt4py (and icon4py as well) use symbolic field strides, by default. If this is the case, is it safe to merge this PR?
philip-paul-mueller
added a commit
to philip-paul-mueller/gt4py
that referenced
this pull request
Apr 28, 2025
We have to expand some Memlets into copy Maps, because there is no suitable call to `cudaMemcpy*`, because we have to set their order right. But this is a gross hack, it should be solved once [PR#1913](GridTools#1913) has been merged, but for that DaCe [PR#1976](spcl/dace#1976) has to be merged. The error was that there was the assumption that at least one of these newly created Maps survived. However, this does not seem to be the case, depending in which order MapFusion is applied. Thus the assert was removed. We now assume that if we do not find a newly created MapEntry that is was integrated into a Map that was already present, we further assume that the Map parameter of that Map were used and that they are correct.
philip-paul-mueller
added a commit
that referenced
this pull request
Apr 29, 2025
We have to expand some Memlets into copy Maps, because there is no suitable call to `cudaMemcpy*`, because we have to set their order right. But this is a gross hack, it should be solved once [PR#1913](#1913) has been merged, but for that DaCe [PR#1976](spcl/dace#1976) has to be merged. The error was that there was the assumption that at least one of these newly created Maps survived. However, this does not seem to be the case, depending in which order MapFusion is applied. Thus the assert was removed. We now assume that if we do not find a newly created MapEntry that is was integrated into a Map that was already present, we further assume that the Map parameter of that Map were used and that they are correct.
Contributor
|
This PR can merged, since the dace source now points to the GridTools fork repo. |
stubbiali
pushed a commit
to stubbiali/gt4py
that referenced
this pull request
Aug 19, 2025
…2004) We have to expand some Memlets into copy Maps, because there is no suitable call to `cudaMemcpy*`, because we have to set their order right. But this is a gross hack, it should be solved once [PR#1913](GridTools#1913) has been merged, but for that DaCe [PR#1976](spcl/dace#1976) has to be merged. The error was that there was the assumption that at least one of these newly created Maps survived. However, this does not seem to be the case, depending in which order MapFusion is applied. Thus the assert was removed. We now assume that if we do not find a newly created MapEntry that is was integrated into a Map that was already present, we further assume that the Map parameter of that Map were used and that they are correct.
stubbiali
pushed a commit
to stubbiali/gt4py
that referenced
this pull request
Aug 19, 2025
…ridTools#1913) Due to the update to DaCe in [PR#1976](spcl/dace#1976) the code generator is now able to handle more Memlets directly as Cuda `memcpy()` calls. Thus we have to change this function to reflect this case. It even stands to reason that these functions can be removed altogether. Note that the merge has become possible because of our switch to our DaCe fork in [PR#2012](GridTools#2012).
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
Due to the update to DaCe in PR#1976 the code generator is now able to handle more Memlets directly as Cuda
memcpy()calls. Thus we have to change this function to reflect this case.It even stands to reason that these functions can be removed altogether.
Note that the merge has become possible because of our switch to our DaCe fork in PR#2012.