Unwrap AutoSpecialize through SciMLBase - #1157
Draft
ChrisRackauckas-Claude wants to merge 1 commit into
Draft
Conversation
Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What changed and why
Teach the public
SciMLBase.unwrapped_finterface how to unwrapNonlinearSolveBase.AutoSpecializeCallable, and add a regression assertion alongside the wrapper construction tests.The wrapper fallback removal in d3e1692 exposed a downstream failure in the archived SciMLNLSolve package: its nested ForwardDiff tag does not match the wrappers precompiled by NonlinearSolveBase. NonlinearSolveBase 2.19.0 passes that downstream test, while 2.20.0 and later fail. SciMLNLSolve can now unwrap through a documented public SciMLBase API instead of reaching into NonlinearSolveBase internals.
This PR should be ignored until reviewed by @ChrisRackauckas.
Verification
Failing before the fix on clean current master (
f16755710), with the regression assertion applied:Passing with this fix:
I also applied the method to a checkout of NonlinearSolveBase 2.30.3 and developed that checkout into SciMLNLSolve. Its full test suite then passed:
Other local checks:
Not verified
I did not run the monorepo root/full
Everythingsuite, GPU jobs, or downstream packages other than SciMLNLSolve. The docs build emitted its existing missing-docstring and output-size warnings but completed successfully.The fix still needs a 2.30.x backport/release to affect SciMLNLSolve's currently resolvable dependency set. SciMLNLSolve itself is archived, so its public-API downstream adjustment cannot be submitted until the repository is unarchived.