***SPAM*** Re: [ofa-general] Re: [PATCH] infiniband-diags/mcm_rereg_test.c: Add missing mad_rpc_close_port call

[Hal Rosenstock]

On Thu, Mar 12, 2009 at 10:54 AM, Sasha Khapyorsky <sashak at voltaire.com> wrote:
> On 08:53 Wed 11 Mar     , Hal Rosenstock wrote:
>>
>> Signed-off-by: Hal Rosenstock <hal.rosenstock at gmail.com>
>>
>> ---
>> diff --git a/infiniband-diags/src/mcm_rereg_test.c b/infiniband-diags/src/mcm_rereg_test.c
>> index de5ea90..d087e4a 100644
>> --- a/infiniband-diags/src/mcm_rereg_test.c
>> +++ b/infiniband-diags/src/mcm_rereg_test.c
>> @@ -492,9 +492,13 @@ int main(int argc, char **argv)
>>       mad = umad_get_mad(umad);
>>
>>       free(umad);
>> +#if 1
>> +     mad_rpc_close_port(srcport);
>> +#else
>>       umad_unregister(port, agent);
>>       umad_close_port(port);
>>       umad_done();
>> +#endif
>
> This is not so correct - SA class is registered directly by using
> umad_register() and not with rpc. So it rather should be:
>
>        umad_unregister(port, agent);
>        mad_rpc_close_port(srcport);
>
> OTOH libibmad's rpc is not used in this program (only mad formatting
> stuff). So I would recommend to remove mad_rpc_* things completely.

Sure; makes sense to do that if we're keeping this around.

-- Hal

>
> Sasha
> _______________________________________________
> general mailing list
> general at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
>
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
>