Hi Phil,

Thanks for the reply.  It ended up being a messup on my end.  I was
mallocing 1 less byte than actually needed (for a null termination
character) for a string I was using and also forgot to free something else
along the way.  Testing happily now, thanks.

Adam

On Mon, Oct 10, 2011 at 8:00 AM, Phil Carns <[email protected]> wrote:

>  Hi Adam,
>
> I don't have any guesses as to what's going wrong, but you might want to
> try running your application through valgrind to see if it gives you any
> further information.  That segfault that you listed below might not be the
> original source of the problem.
>
> -Phil
>
>
> On 10/04/2011 07:25 AM, Adam Yee wrote:
>
>  Dear OrangeFS developers,
>
>  I'm using the native API to write an interface to the file system.  I've
> been experiencing problems with PVFS_sys_lookup.  I've pasted a couple of
> call stacks below.  As I'm "stress-testing" functionality, for example the
> equivalent of pvfs2-rm or pvfs2-ls, I run into these runtime memory errors.
>  What I mean by stress-testing is running the program repeatedly.  Errors
> occur after a few consecutive tests.  Really the only difference in my code
> is I call PVFS_util_init_defaults once at the beginning and
> PVFS_sys_finalize once at the end, while making multiple calls to the file
> system in between (as opposed to each admintool utility calling both
> functions upon each execution).  I tested initializing and finalizing each
> time I access the file system, but that produced errors immediately.  Last
> piece of info you need to know is that I'm calling my API functions via TCP
> sockets.  But since I'm just passing ascii characters over the network and
> parsing them for commands (which function to run, ie rm, ls, mkdir), I don't
> see how that could be a source of error.
>
>  Anyone have an idea for a solution?  Ideas where the problem is coming
> from?
>
>  Thanks,
> Adam
>
>  orangefs API Server [C/C++ Application]
>  /home/adam/[...] [23913] [cores: 1]
>  Thread [1] 23913 [core: 1] (Suspended : Signal : SIGSEGV:Segmentation
> fault)
>  0xb7ec57ae
>  0xb7ec6ced
>  free() at 0xb7ec9e5d
>  BMI_tcp_memfree() at bmi-tcp.c:682 0x80b0a71
>  PINT_serv_free_msgpair_resources() at msgpairarray.sm:815 0x808d998
>  msgpairarray_completion_fn() at msgpairarray.sm:606 0x808dcc2
>  PINT_state_machine_invoke() at state-machine-fns.c:140 0x808fc19
>  PINT_state_machine_next() at state-machine-fns.c:317 0x809018d
>  PINT_state_machine_continue() at state-machine-fns.c:335 0x8090331
>  PINT_client_state_machine_test() at client-state-machine.c:735 0x8057a52
>  PINT_client_wait_internal() at client-state-machine.c:871 0x8057c56
>  PVFS_sys_wait() at client-state-machine.c:1,036 0x8057d3e
>  PVFS_sys_ref_lookup() at sys-lookup.sm:541 0x8062315
>  PVFS_sys_lookup() at sys-lookup.sm:597 0x8062492
>  api_delete() at api-delete.h:101 0x805361b
>  main() at api-server.c:600 0x8053fde
>  gdb
>
>
>  *** glibc detected *** /home/adam/[...] : malloc(): memory corruption:
> 0x080e8210 ***
>
>  orangefs API Server [C/C++ Application]
>  /home/adam/[...] [24167] [cores: 0]
>  Thread [1] 24167 [core: 0] (Suspended : Signal : SIGABRT:Aborted)
>  __kernel_vsyscall() at 0xb7fe1424
>  raise() at 0xb7e83941
>  abort() at 0xb7e86e42
>  0xb7ebb305
>  0xb7ec5501
>  0xb7ec82fc
>  malloc() at 0xb7ec9f33
>  PINT_smcb_alloc() at state-machine-fns.c:523 0x808f897
>  PVFS_isys_ref_lookup() at sys-lookup.sm:472 0x8061fde
>  PVFS_sys_ref_lookup() at sys-lookup.sm:528 0x80622e3
>  PVFS_sys_lookup() at sys-lookup.sm:597 0x8062492
>  api_lookup() at api-lookup.h:29 0x8052d3e
>  main() at api-server.c:506 0x8054337
>  gdb
>
>
>
>
> _______________________________________________
> Pvfs2-developers mailing 
> [email protected]http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
>
>
>
> _______________________________________________
> Pvfs2-developers mailing list
> [email protected]
> http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
>
>
_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers

Reply via email to