Greetings,
Vital details: kernel 2.6.20, pvfs-2.6.2,
mpich2-1.0.5
I have a seemingly working pvfs2 filesystem up and running
on three nodes. That is, I can perform file operations
using the pvfs2-* utilities and mpi-io-test appears to
work fine.
I can run my single process MPI-IO program just fine
on a unix fs, but I get segfaults when I run it on the
pvfs2 fs. I'm not getting any useful debugging output
when I run it under mpiexec. Basically the app
opens a file, reads it, writes out the data as a new
file. It repeats this for 64 files. At some point in
the process, seemingly random,
the process crashes on file open, after a significant number of
reads and writes, such as 33, for example. I poked
around with gdb and things look kosher. Any clues about how
I would go about debugging this would be very
appreciated.
Here's the gdb bt:
#0 0x00002b84c3ace1d5 in free () from /lib/libc.so.6
#1 0x0000000000550b6e in PINT_free_object_attr ()
#2 0x00000000005206d0 in finalize_context ()
#3 0x0000000000521cf8 in lookup_cleanup ()
#4 0x000000000051a025 in PINT_state_machine_next ()
#5 0x000000000051a80e in PINT_client_state_machine_test ()
#6 0x000000000051ad30 in PINT_client_wait_internal ()
#7 0x000000000051af30 in PVFS_sys_wait ()
#8 0x0000000000520ac4 in PVFS_sys_ref_lookup ()
#9 0x0000000000520c09 in PVFS_sys_lookup ()
#10 0x00000000004fb8b6 in fake_an_open ()
#11 0x00000000004fbcb3 in ADIOI_PVFS2_Open ()
#12 0x00000000004ae97d in ADIO_Open ()
#13 0x00000000004a6185 in PMPI_File_open ()
#14 0x0000000000410ebd in parcfdio::xyzq::Meta::read_indices (this=0x76dfa0,
fname=0x21 <Address 0x21 out of bounds>, [EMAIL PROTECTED],
indices=0x7fffe7a88de0,
[EMAIL PROTECTED], [EMAIL PROTECTED]) at
/usr/local/mpi/include/mpicxx.h:2094
#15 0x000000000041139a in parcfdio::xyzq::get_topology (
fname=0x76e0e8 "pvfs2:/mnt/pvfs/plot3d/grid64/q64_8_2_2.dat", sid=<value
optimized out>,
[EMAIL PROTECTED], indices=0x7fffe7a88de0, [EMAIL PROTECTED],
[EMAIL PROTECTED]) at xyzqMeta.cpp:19
#16 0x00000000004149ff in parcfdio::plot3d::get_topology (
fname=0x2d7580002d758 <Address 0x2d7580002d758 out of bounds>, sid=33,
ngrid=0x2d7580002d758,
indices=0x0, my_firstzone=0x0, my_nzones=0x3a6b) at
../parcfdio/plot3dIO.h:24
#17 0x0000000000414a25 in plot3d_get_topology (
fname=0x2d7580002d758 <Address 0x2d7580002d758 out of bounds>, sid=33,
ngrid=0x2d7580002d758,
indices=0x0, firstzone=0x0, my_nzones=0x3a6b) at parcfdio.cpp:18
#18 0x000000000040b0e5 in tester<double>::open_session (this=0x7fffe7a88db0,
argc=<value optimized out>,
argv=<value optimized out>, idir=<value optimized out>, ifile=<value
optimized out>)
at driver.cpp:209
#19 0x00000000004094c4 in main () at driver.cpp:309
(gdb)
--
Thanks,
Russell
_______________________________________________
Pvfs2-users mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
