-
I did a little further debugging (with the `mpi_pfasst` example on the `feature/fix-mpi` branch of PR #212) and it turned out, that PFASST is not triggering non-blocking sends on the finest level. Alt…
-
Aside from the Boris example, no other part of PFASST uses random numbers. Following YAGNI, let's move this out of PFASST proper and into the Boris example.
This will also allow us to simplify the `…
-
The **really** weird mismatch of the _blocking_ status communication is still there on the `feature/fix-mpi`-branch of PR #212.
I put in some debug logging of the following in `mpi_communicator_impl.…
-
As we've been doing a bunch of bug fixes, I think it's time for thinking about a release before v1.0.0 incorporating those things. I don't want to call it a patch (i.e. v0.4.1) mainly because we've al…
-
I think we're getting close to being able to release 0.4.0.
Blockers
- [x] #31 (see #150) - LU decomp
- [x] #127 - iteration control for Boris
- [x] #84 - split up decls/impls
- [x] #173 - getting it…
-
When running in parallel with MPI, the logging output should somehow be differently handled than in a serial environment. There come two options to my mind:
1. make all calls to easylogging++ a no-op …
-
Based on #43, iteration control, e.g. based on residual, should be implemented for PFASST as well. This will require communicating status between the processors.
-
As mentioned in #113, we should discuss, whether we want to put the examples into their own namespace for better separation from the library.
I suggest using
```
pfasst::examples
{
advection_diffu…
-
Initially this should be very simple, like a boolean function `converged()` whose default implementation just checks the current iteration number vs the max iteration count.
-
Should we rename the repository to `PFASST++`? It seems a little more appropriate given that it's going to be cited as `PFASST++` shortly and named as that in the two release publications.
The git-UR…