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

Sean Hefty mshefty at ichips.intel.com
Thu May 5 10:20:17 PDT 2005


Tom Duffy wrote:
> On Thu, 2005-05-05 at 08:44 +0300, Or Gerlitz wrote:
> 
>> Tom,
>>
>>
>>>This occurred when I did a connect without first setting up a listen
>>
>>using kdapltest. 
>>
>>>RIP: 0010:[<ffffffff882b2c61>]
>>
>><ffffffff882b2c61>{:ib_dat_provider:dapl_cm_passive_cb_handler+27}
>>
>>>Call Trace:<ffffffff882aae32>{:ib_cm:cm_process_work+50}
>>
>><ffffffff882abd1e>{:ib_cm:cm_work_handler+1438}
>>
>>If you you haven't set up a (kdapltest server) listener - why did the CM
>>invoked a (kdapl provider) consumer callback for this SID?
> 
> 
> I have no idea.  I loaded the kdapltest.ko kernel modules, but had not
> started the kdapltest userland program when this occured.
> 
> I am pretty sure I had rebooted right before trying this, but it could
> have been that I had just unloaded and reloaded all the (dapl) modules
> (so maybe something was left from a previous incarnation).
> 
> I will reboot both machines and try to reproduce.

If you had just unloaded/reloaded the modules, we should probably look at 
kdapl to see that it destroys all cm_id's.

- Sean




More information about the general mailing list