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
Reduced minimum size for sparse linear and nonlinear solvers #7083
Conversation
… solvers to 10, see ticket OpenModelica#6342
…t-in dense Newton/homotopy solver is used in more cases during initialization, though with reasonable initialization time.
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.
LGTM
@casella Just in case you are not aware: You can fix the failing tests automatically by running e.g. Andreas@DESKTOP-OSJIVA0 MINGW64 /D/workspace/OpenModelica/testsuite/simulation/modelica/linear_system
# ../../../rtest -b NPendulum40.mos with MINGW, but check if it didn't change to much. You can find |
Thank you for the info. I'd like to make sure that nothing actually got broken. Where exactly do I find the diff between the previous and actual test results? The log file is huge and I have some difficulties orienting myself in it. |
Poke! All tests look fine, they just need to be updated because the prints changed. You can
|
@phannebohm, this PR was a drastic solution, basically turning off tearing completely for linear systems. The idea was to use sparse solvers instead. The problem is that if you don't enable them explicitly, they won't be used, so the performance will become a lot worse. So, this PR is not really the ideal solution. I think we should first check why strict tearing is not working as expected, see #6196 and fix that. |
See discussion in ticket #6342