On Thu, Nov 14, 2013 at 4:01 PM, Geoffrey Irving <[email protected]> wrote:
> On Thu, Nov 14, 2013 at 1:03 PM, Matthew Knepley <[email protected]> > wrote: > > On Thu, Nov 14, 2013 at 2:01 PM, Geoffrey Irving <[email protected]> wrote: > >> > >> The SetupElement, SetupMaterialElement, and SetupBdElement functions > >> are almost entirely duplicates of each other: > >> > >> ./src/snes/examples/tutorials/ex12.c > >> > >> This makes it a bit difficult to understand what's going on: I'm new > >> the code, so I had to run a diff to confirm that the boundary spaces > >> are allocated identically to the interior spaces (I naively expected a > >> difference, but it's apparently elsewhere in the code). > >> > >> Are there any objections to me refactoring the code a bit to avoid the > >> duplication? > > > > > > No, however it would be nice to verify that the current tests run: > > > > ./config/builder2.py check src/snes/examples/tutorials/ex12.c > > They don't, but I haven't started changing anything yet. mpiexec > fails with a exceedingly unhelpful usage message (below). The test > seems to run fine if I use mpirun directory. I'm on openmpi 1.4.5. > The problem appears to be the "-hosts" option: it also runs fine with > mpiexec if I remove "-hosts localhost". > > Geoffrey > > --------------------------------- > > cayley:petsc% mpiexec -hosts localhost -n 1 -- > /home/irving/petsc/debug/lib/ex12-obj/ex12 -run_type test > -refinement_limit 0.0 -bc_type dirichlet -interpolate 0 > -petscspace_order 1 -show_initial -dm_plex_print_fem 1 > mpiexec (OpenRTE) 1.4.5 > > Usage: mpiexec [OPTION]... [PROGRAM]... > ... > Yes, I hate mpiexec since its non-standard. I put in -hosts so that it will run on my laptop when disconnected from Wifi. You can change builder.py:1701 to get it running. I need to put in a stupid configure tests for this. Matt -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener
