-
Notifications
You must be signed in to change notification settings - Fork 161
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
slate: fix kernel argument ordering #4003
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Makes sense. Thanks for taking this on.
|
|
5aa2a35
to
6692db7
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks good to me.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If we leave this test here for now (and I do think we should until downstream testing gets a more formalised solution), then at least we can make it a bit smaller.
a) the solve doesn't need to be accurate, it just needs to run.
b) the second solve to reconstruct buoyancy doesn't test anything that isn't definitely covered elsewhere (it's a fairly simple form).
I'm happy to merge with these updates.
e77d9cb
to
9a182b9
Compare
9a182b9
to
33fc671
Compare
Co-authored-by: Josh Hope-Collins <[email protected]>
33fc671
to
2a6ab23
Compare
Approved pending CI passing.
Can I just try rerunning it or does something need fixing? |
I think this is a side effect of me messing about with CI for #4011. I believe I have fixed this and that it shouldn't recur. If you notice this again please let me know. |
Ok, I've set the failed job off to rerun. |
Fix #4001.
Consider a slate expression:
where:
Without aggressive zero simplification, we had:
which lead to T.coefficients() = (f, g, f, g)->(f, g) == B.coefficients().
Current implementation in Slate had happened to be working as B.coefficients() had been the same as T.coefficients().
With aggressive zero simplification (now merged in UFL), we have:
which lead to T.coefficients() = (g, f, g)->(g, f) != B.coefficients(), and the current Slate breaks.
This PR is to fix this issue.