serac
serac copied to clipboard
FirstOrderEquilibriumLua test in serac_newmark_test.cpp occasionally fails
This test will sporadically fail if run on more than one processor after a different test is run. It is probably a memory issue within the first-order-system version of the second order ODE solver class. It is currently broken in develop.
What is the nature of the failing test? Is it segfaulting, exceeding some tolerance, or something else?
This may be due to the CG linear solver by default using iterative mode. This means initial guesses for the linear solver must be initialized.
I believe this has already been fixed in our linear solver refactoring update.