Jacob King

Hi Sherry, I've tested the maint branch and it is better. Thank you. I've run into a new issue that I believe you can reproduce with a simple modification to...

I should also mention that I needed to make this minor patch: ``` Index: FORTRAN/superlu_c2f_dwrap.c =================================================================== --- FORTRAN/superlu_c2f_dwrap.c (revision 2100) +++ FORTRAN/superlu_c2f_dwrap.c (working copy) @@ -95,7 +95,7 @@ dLUstructFree(LUstruct_ptr); }...

Thanks! I'm now getting through the first factorization and solve with the 3d code. However, it seems that there is another problem when doing a subsequent factorization with the same...

Thanks, Sherry. Everything is now running through with 4 processors for the 3d layouts that I've tried (nzdep=1, 2 and 4). However, I'm running into trouble with 8 cores. It...

I've confirmed that this is a matrix layout issue. When I modify the ordering to be the same as the application I get the failures with identical layouts (It works...

I fixed a problem with my debugging code (I forgot the break statements). I tested with 4-core runs and ended up back on this issue. I've now moved to 8-core...

@drew-parsons @balay @xiaoyeli Hi, I have a related issue with this part of the build system when using the fortran interface. Right now there is a dependency chain for the...

Hi @drew-parsons, after I re-examine the dependencies you are correct: ftestmod -> superlu_dist_fortran -> superlu_dist. I'd like to use the interface for the set_superlu_options that is available in superlu_mod.F90 which...

Sorry for the delay. I just tested and the patch works with "-j 4" on Cori.