[Darshan-users] Darshan w/ OpenMPI

[Phil Carns]

On 07/13/2010 06:23 PM, Rob Latham wrote:
> On Tue, Jul 13, 2010 at 03:11:51PM -0700, Noah Watkins wrote:
>    
>> Here is the best I have so far:
>>
>> Program received signal SIGSEGV, Segmentation fault.
>> [Switching to Thread 47478960279584 (LWP 9272)]
>> 0x0000000000000000 in ?? ()
>> (gdb) backtrace
>> #0  0x0000000000000000 in ?? ()
>> #1  0x0000000000409d6d in hash (h=Variable "h" is not available.
>> ) at ../../utilities/src/hashtable.c:61
>> #2  0x0000000000418cc7 in darshan_file_by_name (name=0x535df0 "/net/scratch1/noah/fs_test.bin") at lib/darshan-posix.c:1183
>>      
> one weird thing: that hashtable.c didn't come from darshan...  the
> hash() in 'darshan_file_by_name' is supposed to come from
> lib/lookup8.c
>
> ==rob
>    

Ah, good catch.  Noah, can you try out the latest trunk update and see 
if that fixes the problem?  I just renamed the hash() function to 
darshan_hash().  Here is the diff:

https://trac.mcs.anl.gov/projects/darshan/changeset/290?format=diff&new=290

thanks,
-Phil