[FLASH-USERS] Trouble getting started with Flash

[David Williamson]

Hi,

I've been trying to get Flash working on our cluster. It looks like it's got hdf5 and mpi all linked, and I'm trying the Sedov test but still finding issues. It crashes right at the beginning with:

Warning: no access to tty (Bad file descriptor).
Thus no job control in this shell.
 RuntimeParameters_read:  ignoring unknown parameter "iProcs"...
 RuntimeParameters_read:  ignoring unknown parameter "jProcs"...
 RuntimeParameters_read:  ignoring unknown parameter "kProcs"...
[compute-2-0:26014] *** Process received signal ***
[compute-2-0:26014] Signal: Segmentation fault (11)
[compute-2-0:26014] Signal code: Address not mapped (1)
[compute-2-0:26014] Failing at address: 0x10
[compute-2-0:26015] *** Process received signal ***
[compute-2-0:26015] Signal: Segmentation fault (11)
[compute-2-0:26015] Signal code: Address not mapped (1)
[compute-2-0:26015] Failing at address: 0x10
[compute-2-0:26014] [ 0] /lib64/libpthread.so.0 [0x338e00de70]
[compute-2-0:26014] [ 1] /usr/mpi/gcc/openmpi-1.2.5/lib64/libopen-pal.so.0(_int_malloc+0x2a5) [0x2aaaab8a9145]
[compute-2-0:26014] [ 2] /usr/mpi/gcc/openmpi-1.2.5/lib64/libopen-pal.so.0(malloc+0x93) [0x2aaaab8aae83]
[compute-2-0:26014] [ 3] /usr/lib64/libgfortran.so.1 [0x2aaaabaef34d]
[compute-2-0:26014] [ 4] ./flash3(__namevaluell_data__namevaluell_addreal+0x60) [0x551c6a]
[compute-2-0:26014] [ 5] ./flash3(namevaluell_bcast_+0x423) [0x550593]
[compute-2-0:26014] [ 6] ./flash3(runtimeparameters_bcast_+0x1a) [0x446ff2]
[compute-2-0:26014] [ 7] ./flash3(runtimeparameters_init_+0x22b) [0x44767f]
[compute-2-0:26014] [ 8] ./flash3(driver_initflash_+0x48) [0x41d374]
[compute-2-0:26014] [ 9] ./flash3(MAIN__+0x1d) [0x421625]
[compute-2-0:26014] [10] ./flash3(main+0xe) [0x57179e]
[compute-2-0:26014] [11] /lib64/libc.so.6(__libc_start_main+0xf4) [0x338d81d8a4]
[compute-2-0:26014] [12] ./flash3 [0x4198f9]
[compute-2-0:26014] *** End of error message ***

(repeated for each thread)

Is this something that anyone has hit before? Is it likely an mpi issue, or something simpler?

Cheers,
-Dave