[openib-general] Re: [kDAPL][PATCH] fix panic in server side

[Tom Duffy]

On Fri, 2005-05-06 at 13:30 -0400, James Lentini wrote:
> Tom,
> 
> There were two changes in this patch:
> 
> - check for a NULL comm_id pointer
> - add error returns
> 
> My understanding of the subsequent discussion is that:
> 
> - we should not have been using the comm_id->state value. We should
>    have been using the event->event value instead
> 
> - if we return a non 0 error code, our CM ID will be destroyed
> 
> Given those pieces of information:
> 
> Do we still want to check for a NULL comm_id pointer?
> Should we always return 0 from this function?
> 
> I think the answer is yes to both of the above, but I want to make 
> sure there is consensus.

My patch was wrong as pointed out.

The panic I saw did have a NULL comm_id.  I am not sure how this
happened.  Apparently, this function should never be called with a null
pointer.

-tduffy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20050506/31f4e423/attachment.sig>