Tpetra: Make sparse matrix-matrix multiply thread-parallel

[mhoemmen]

@trilinos/tpetra
This is related to #148 and #430. It may not require completion of either of those.

[mhoemmen]

@csiefer2 has been working on this, so I'm assigning him just so he gets credit for finishing it :-)

[mhoemmen]

Note that @csiefer2 finished OpenMP integration in FY17 Q1. Unclear whether CUDA is part of the requirement for FY17. (CUDA must run correctly, but unclear whether we need to plug in sparse matrix-matrix multiply for CUDA; is it acceptable for now to run on host?)

[csiefer2]

To build w/ the OpenMP kernels, you need these configure options

-D Trilinos_ENABLE_OpenMP:BOOL=ON 
-D TpetraKernels_ENABLE_Experimental:BOOL=ON

[mhoemmen]

Next step is to enable this capability for OpenMP by default. We'll need to make sure that tests pass. Since CUDA is not an immediate priority, we should for now only enable this by default for non-CUDA builds.

[github-actions]

This issue has had no activity for 365 days and is marked for closure. It will be closed after an additional 30 days of inactivity.
If you would like to keep this issue open please add a comment and remove the MARKED_FOR_CLOSURE label.
If this issue should be kept open even with no activity beyond the time limits you can add the label DO_NOT_AUTOCLOSE.

[github-actions]

This issue was closed due to inactivity for 395 days.