On 7/6/06, Michael S. Tsirkin <[EMAIL PROTECTED]> wrote:
> Quoting r. Sean Hefty <[EMAIL PROTECTED]>:
> > Subject: Re: ucma into kernel.org
> >
> > Michael S. Tsirkin wrote:
> > > TCP sockets just expose this to application through the
> > > TCP_SYNCNT option.
> > >
> > > Which leads again to my suggestion: since both TCP and IB CM have this, 
> > > let us
> > > change max_cm_retries to max_request_retries, and add this in rdma_cm as a
> > > generic option.
> >
> > I'm not against adding in generic option for this, but the retry values have
> > different limits.  IB is limited to a maximum of 15, while TCP has a 
> > maximum of 255.
>
> What limits IB retry cound to 15? I would expect it to be arbitrary.  Note I 
> am
> talking about CM REQ retries that we do in software and that are not passed in
> any message.

The CM REQ includes a "Max CM Retries" field, 4-bits, that specifies
the maximum number of retries that can be done for REQ, REP, and DREQ
before a REJ is sent to indicate that the connection is aborted.

- Fab

_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general

To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

Reply via email to