Register Single Precision Linear Solvers #13087
Conversation
|
Looks like every compiler is happy now. @loumalouomega can you please take a look again? I'd also appreciate if someone from the @KratosMultiphysics/technical-committee too could take a quick peek. |
| // using LinearSolverType = LinearSolver<SpaceType, LocalSpaceType>; | ||
| // using IterativeSolverType = IterativeSolver<SpaceType, LocalSpaceType>; | ||
| using CGSolverType = CGSolver<SpaceType, LocalSpaceType>; | ||
| using SpaceType = TUblasSparseSpace<TSparseDataType>; |
There was a problem hiding this comment.
BTW, we should eventually replace the linear solver factory with the register @KratosMultiphysics/technical-committee I don't know if a good moment to do that
There was a problem hiding this comment.
I was thinking about that too, but other parts of Kratos still expect linear solvers to be in KratosComponents and I didn't want to change too many unrelated things at once.
There was a problem hiding this comment.
A priori we can define in both database and add a warning in the factory, opinion @KratosMultiphysics/technical-committee ?
| @@ -541,6 +541,7 @@ class KRATOS_API(KRATOS_CORE) Serializer | |||
| KRATOS_SERIALIZATION_DIRECT_LOAD(std::size_t) | |||
| #endif | |||
| KRATOS_SERIALIZATION_DIRECT_LOAD(std::complex<double>) | |||
There was a problem hiding this comment.
We should remove those evil tabs
This PR is a follow-up on #13079.
Changelog
doubleas the value type in some functions.double.Serializerto supportstd::complex<float>.