[Pvfs2-developers] Possible bug in pvfs2-cp

[Phil Carns]

Hi Nick,

I just committed a fix for this in trunk that you might want to merge to 
your branch if you can confirm it fixes your problem.

I think the client side function that waits for state machines to 
complete was grabbing the sm frame too soon.  In your case that caused 
it to get the error code from an intermediate msgpairarray nested 
machine rather than from the top level lookup machine that it was 
supposed to.  That made the lookup call return success when it should 
have failed.

-Phil

Nicholas Mills wrote:
> Hey all,
> 
> David and I were using pvfs2-cp and couldn't get it to work. We keep
> getting errors of the form "Target file XXX already exists." We tried
> switching to HEAD and recreating the storage space but the error is
> always the same. For example, the command "pvfs2-cp /usr/lib/libc.a
> /pvfs2" gives the error "Target file libc.a already exists."
> 
> After stepping through the code with the debugger I've narrowed the
> problem down to pvfs2-cp.c:554. On this line the code calls
> PVFS_sys_ref_lookup on the destination path. If the lookup succeeds
> (returns 0) then it assumes the file already exists and refuses to
> copy over it. Is this the correct behavior? In the example command
> above the lookup succeeds but the lookup response contains the null
> handle.
> 
> Not sure what happened here but somewhere along the line the semantics
> of PVFS_sys_ref_lookup must have changed. I had pvfs2-cp working just
> fine about a month ago.
> 
> --Nick
> _______________________________________________
> Pvfs2-developers mailing list
> Pvfs2-developers at beowulf-underground.org
> http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers