[Libmesh-devel] Hermites now Thread Safe!

2010-02-10 Thread Derek Gaston
So... I got tired of not being able to use threading with Hermites... so I finally went in there and fixed it. I fixed it by turning off the caching that Roy was doing when libMesh is compiled with thread support. Note that this might slow things down if you happen to have compiled your librar

Re: [Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread Jed Brown
On Wed, 10 Feb 2010 15:25:41 -0700, Derek Gaston wrote: > Nope... not important... I just want to make sure we don't get errors > when using Hypre and calling set_preconditioner multiple times (using > the same settings). As long as that is satisfied I'm good. That should be fine. Speak up if y

Re: [Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread Derek Gaston
On Feb 10, 2010, at 3:25 PM, Jed Brown wrote: > Redirecting through something trivially cheap like Jacobi should work, > but it's a hack. Is runtime switching between Hypre solvers something > that is important to you? There is nothing fundamental in the design of > PC_Hypre preventing such dyn

Re: [Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread John Peterson
On Wed, Feb 10, 2010 at 4:16 PM, Derek Gaston wrote: > On Feb 10, 2010, at 3:12 PM, John Peterson wrote: > >> Right now the generic LinearSolver::set_preconditioner_type() >> interface just sets its own _preconditioner_type variable and returns. >> This makes it hard to actually change preconditio

Re: [Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread Jed Brown
On Wed, 10 Feb 2010 15:16:34 -0700, Derek Gaston wrote: > I'm all for this by the way... as long as it works. Like I mentioned > yesterday I seem to remember some issue with calling set > preconditioner multiple times with Petsc... especially involving > Hypre. This is supposed to work, but swit

Re: [Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread Derek Gaston
On Feb 10, 2010, at 3:12 PM, John Peterson wrote: > Right now the generic LinearSolver::set_preconditioner_type() > interface just sets its own _preconditioner_type variable and returns. > This makes it hard to actually change preconditioners through the > generic interface if you've already initi

[Libmesh-devel] Making LinearSolver::set_preconditioner_type() virtual

2010-02-10 Thread John Peterson
Right now the generic LinearSolver::set_preconditioner_type() interface just sets its own _preconditioner_type variable and returns. This makes it hard to actually change preconditioners through the generic interface if you've already initialized your LinearSolver object. For example, in the Pets