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
[WIP] Add Tpetra tests. #16611
base: master
Are you sure you want to change the base?
[WIP] Add Tpetra tests. #16611
Conversation
Very nice! A few questions for the direct solver tests:
|
Yes, with have a number of tests that have outputs like |
With #16613, we are now at 55/284 tests succeeding. |
And with #16616, we're now at 71/284 tests succeeding. There's progress :-) |
cc7ed16
to
5cfe554
Compare
Okay good to know! |
With the Update: On it and almost completed. Currently recompiling as |
We now have #16625 |
On 2/10/24 00:07, Jan Philipp Thiele wrote:
Then the question would be how the current setups on the tester look like
and whether we want to have a 'feature complete' Amesos2 test setup.
I don't actually know. The various testers may also have different setups.
@tamiko -- do you know?
|
f67eb95
to
0c8ad8e
Compare
0c8ad8e
to
72db4d0
Compare
I like to keep using this PR to keep track of the work we have done so far.
Implemented by: #16613
Implemented by: #16616
Implemented by: #16664 The current master passes 93 tests out of 284. |
72db4d0
to
9aac8c4
Compare
Now at 114 out of 288 tests succeeding :-) |
I have to admit that I don't understand the deadlock (nor perhaps what the test does). I think what the test does is create a Tpetra vector locally on a process and copy it to a deal.II vector:
I believe that the |
The test was introduced in #5794. Before that patch we were doing
when copying to a |
We're now at 141/288 succeeding tests. |
This is definitely a WIP patch: 284 tests for the Tpetra wrappers, or which currently 41 succeed. I created these tests by cloning the existing
tests/trilinos/
directory and replacingTrilinosWrappers::
byLinearAlgebra::TpetraWrappers::
plus some minor adjustments such as adding template arguments. I'm sure many tests need additional adjustments.Many (=the vast majority) of these tests clearly do not currently work. But the patch may help us isolate what needs to be fixed to make them work (such as the issue with
VectorReference::operator Number()
mentioned in #16610) and I will rebase this branch frequently to see how things proceed.I'll also keep a list of things that need to be fixed for a substantial number of tests, and that I'll keep updated as I see things. Here's a start:
VectorReference::operator Number()
SparseMatrix::operator() (int,int)
does not existSparseMatrix::el(int,int)
does not existVector::operator==
is missing.@kinnewig @jpthiele FYI