refactor[next-dace]: Add library node for reduce with skip values#2603
Open
edopao wants to merge 15 commits into
Open
refactor[next-dace]: Add library node for reduce with skip values#2603edopao wants to merge 15 commits into
edopao wants to merge 15 commits into
Conversation
4e1c0e5 to
308cfc7
Compare
Contributor
Author
|
cscs-ci run dace |
Contributor
Author
|
cscs-ci run default |
5fcb578 to
bf658c3
Compare
## Description Do not apply `open_mp` flags on device compilers - we don't do multithreading host code for GPU backends (and device is plain wrong). This flag usage was discovered in the case of an `icx/nvcc` where the `-qopenmp` intel flag bled into the command line of `nvcc`. ## Requirements - [ ] All fixes and/or new features come with corresponding tests. - [ ] Important design decisions have been documented in the appropriate ADR inside the [docs/development/ADRs/](docs/development/ADRs/README.md) folder.
- Use same `dace` version for GT4Py cartesian and next. - Make `dace` a GT4Py dependency, and remove the development groups `dace-cartesian` and `dace-next`. - Instead of using the `gridtools` package index, pull the dace pre-release package from PyPI index. Note: if in the future a custom (non-pypi) version is needed, cartesian and next need to create a version from a common branch (and publish to the gridtools index).
b8f0627 to
14443e2
Compare
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
DaCe has a library node in its standard library for reduction. However, GT4Py and ICON4Py also need to lower reduction expressions with skip values. The lowering was already done, on main, inside a nested SDFG. All this PR does is to move the refactor the code that creates the nested SDFG and make an application library node in GT4Py. In this way, the SDFG nodes for regular reduction and reduction with skip values are represented in the same way in the SDFG.