Sebastian Grimberg

Is there any chance that anyone can comment on the customization of the `IntegrationRule` order in the Tesla miniapp ([code](https://github.com/mfem/mfem/blob/master/miniapps/electromagnetics/tesla_solver.cpp#L77))? ``` int irOrder = H1FESpace_->GetElementTransformation(0)->OrderW() + 2 * order; int...

Update SuperLU_DIST solver interface

1

Updates and upgrades for SuperLU_DIST sparse direct solver interface: - HYPRE_BIGINT and HYPRE_MIXEDINT support. - `Operator::Mult(const Array &X, Array &Y)` overrides for SuperLU solver classes (depends on https://github.com/mfem/mfem/pull/3075). - Enable...

`Multigrid::Mult` with multiple RHS

10

Includes: - `Operator::Mult(const Array &X, Array &Y)` overload for application to multiple RHS. Base class implementation just loops over columns of the input and calls Mult on each Vector. -...

Update STRUMPACK solver interface

2

Updates and upgrades for STRUMPACK sparse direct solver interface: - HYPRE_BIGINT and HYPRE_MIXEDINT support. - `Operator::Mult(const Array &X, Array &Y)` overrides for STRUMPACK solver classes (depends on https://github.com/mfem/mfem/pull/3075). - Enable...

Update MUMPS solver interface

2

Updates and upgrades for MUMPS sparse direct solver interface: - HYPRE_BIGINT and HYPRE_MIXEDINT support. - `Operator::Mult(const Array &X, Array &Y)` overrides for MUMPS solver classes (depends on https://github.com/mfem/mfem/pull/3075). - Block...

[WIP] Add support for pyramid elements in Gmsh meshes

1

This PR uncomments the blocks from #1571 to allow for reading of pyramid elements in Gmsh meshes, in light of upgrades introduced in #2400. It also adds a correction to...

I came across these comments today: https://github.com/mfem/mfem/blob/master/fem/fespace.cpp#L1393 https://github.com/mfem/mfem/blob/master/fem/fespace.cpp#L1880 https://github.com/mfem/mfem/blob/master/fem/fespace.cpp#L2044 I have not dug into this too much, but should users be wary of using mesh refinement/derefinement when dealing with tetrahedral...

Fixes a bug for prism mesh VTU output with `high_order_output = false` and `ref > 1`. NOTE: This is still a WIP. There are still issues with high order output...

Consistent edge ordering for `MarkTetMeshForRefinement` for tetrahedral meshes

7

Even after the fix in #3843 I experienced correctness issues with parallel tetrahedral meshes. This PR adds a few improvements to `Mesh` and `ParMesh`, but in particular fixes issues I...

Fixes for `mfem::Device`

4

This PR adds two updates to `mfem::Device`: - Add `GetNumGPU` method (can rename as desired) in order to query the number of available devices before initializing a `Device`. This is...