fix[dace][next]: Propagates view upon changes in strides. #1784
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
…timization pipeline when confronted with scans.
However the actuall modifier function is not modified yet.
Which is funny then if you look at the last commit, the number of `not`s in this function was correct.
It seems that we alsohave to handle alias. It makes thing a bit handler, instead of only looking at the NestedSDFG, we now look at the `(NameOfDataDescriptorInside, NestedSDFG)` pair. However, it still has some errors.
…ts` flag was not implemented properly.
Before the function had a special mode in which it performed the renaming through the `symbol_mapping`. However, this made testing a bit harder and so I decided that there should be a flag to disable this.
There are some functioanlity missing, but it is looking good.
…trides. However, it is not yet fully tested, tehy are on their wa.
It also seems that it inferes with something.
Because a scalar has a shape of `(1,)` but a stride of `()`. Thus we have first to handle this case. However, now we are back at the index stuff, let's fix it.
However, it still seems to fail in some cases.
The type is now a bit better estimated.
The type are now extracted from the stuff we get from `free_symbols`.
I realized that allowing this is not very safe. I also added a test to show that.
…ces errors in certain cases.
philip-paul-mueller
added a commit
that referenced
this pull request
Dec 20, 2024
Added functionality to properly handle changes of strides. During the implementation of the scan we found that the strides were not handled properly. Most importantly a change on one level was not propagated into the next levels, i.e. they were still using the old strides. This PR Solves most of the problems, but there are still some issues that are unsolved: - Views are not adjusted yet (Fixed in [PR@1784](#1784)). - It is not properly checked if the symbols of the propagated strides are safe to introduce into the nested SDFG. The initial functionality of this PR was done by Edoardo Paone (@edopao). --------- Co-authored-by: edopao <[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.
Some aspects of the stride propagation were not addressed in the original PR.
This PR fixes the behaviour for Views.