[openib-general] Shared Receive Queue

[Galen M. Shipman]

Tziporet,

This did appear to be a firmware issue. After the update I have a  
latency difference on the order of .2 uSec which seems reasonable,  
much better than a ~5. uSec penalty!

Thanks for the help,

Galen


On Jul 27, 2005, at 12:38 AM, Tziporet Koren wrote:

> Hi Galen,
> You are working with an old FW version that had issues with SRQ  
> performance.
> We did performance improvements and now the latency and BW with SRQ  
> are the same as with regular QP.
>
> The changes are available in the new FW release we did few weeks  
> ago: see http://www.mellanox.com/products/firmware.html
>
> Regarding your specific device: you are using MT25208 InfiniHost  
> III Ex, so you need to use FW 4.7.0
>
> In case you need help with FW burning please approach our FAE Todd  
> (cc here too).
>
> After you have new results please reply if the performance numbers  
> are OK now.
>
> Thanks,
> Tziporet
>
>  -----Original Message-----
> From: Galen M. Shipman [mailto:gshipman at lanl.gov]
> Sent: Tuesday, July 26, 2005 5:55 PM
> To: Tziporet Koren
> Subject: Re: [openib-general] Shared Receive Queue
>
> The following is from vstat..
>
>  hca_id=InfiniHost0
>         vendor_id=0x02C9
>         vendor_part_id=0x6278
>         hw_ver=0xA0
>         fw_ver=0x400050003
>         num_phys_ports=2
>                 port=1
>                 port_state=PORT_DOWN
>                 sm_lid=0x0000
>                 port_lid=0x0000
>                 port_lmc=0x00
>                 max_mtu=2048
>
>                 port=2
>                 port_state=PORT_ACTIVE
>                 sm_lid=0x0001
>                 port_lid=0x010a
>                 port_lmc=0x00
>                 max_mtu=2048
>
>
> On Jul 26, 2005, at 8:46 AM, Tziporet Koren wrote:
>
>> which HW & FW are you using?
>>
>> -----Original Message-----
>> From: Roland Dreier [mailto:rolandd at cisco.com]
>> Sent: Tuesday, July 26, 2005 5:38 PM
>> To: Galen M. Shipman
>> Cc: openib-general at openib.org
>> Subject: Re: [openib-general] Shared Receive Queue
>>
>>
>>     Galen> We have software which uses the mellanox shared receive
>>     Galen> queue and we are seeing a substantial performance penalty
>>     Galen> (on the order of 4-5 usec on 0 byte messages). I don't  
>> know
>>     Galen> all the technical details but if you could keep this in
>>     Galen> mind when working on the shared receive queue perhaps the
>>     Galen> performance penalty could be much less in gen2.
>>
>> Is this performance penalty in the software?  It seems much more
>> likely that the issue is in the hardware implementation of SRQ.
>>
>>  - R.
>> _______________________________________________
>> openib-general mailing list
>> openib-general at openib.org
>> http://openib.org/mailman/listinfo/openib-general
>>
>> To unsubscribe, please visit http://openib.org/mailman/listinfo/ 
>> openib-general
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20050727/676391ed/attachment.html>