[openib-general] [PATCH for-2.6.18] IB/cma: option to limitMTU to 1K

[Hal Rosenstock]

On Wed, 2006-09-13 at 15:13, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <halr at voltaire.com>:
> > Subject: Re: [PATCH for-2.6.18] IB/cma: option to limitMTU to 1K
> > 
> > On Wed, 2006-09-13 at 12:22, Michael S. Tsirkin wrote:
> > > Quoting r. Hal Rosenstock <halr at voltaire.com>:
> > > > Subject: Re: [openib-general] [PATCH for-2.6.18] IB/cma: option to limitMTU to 1K
> > > > 
> > > > On Wed, 2006-09-13 at 11:57, Michael S. Tsirkin wrote:
> > > > > Tavor systems get better performance with 1K MTU. Since there does
> > > > > not seem to be any way to find out whether the remote system uses Tavor,
> > > > > add an option to limit the MTU globally.
> > > > 
> > > > Can't Tavor be determined locally ?
> > > 
> > > It can, but we need this for remote tavor as well, anyway.
> > > 
> > > > And couldn't the remote end negotiate the MTU down (if Tavor) as well ?
> > > 
> > > The way to do this is would be for SA to select 1K MTU if it detects Tavor on one side
> > > and if this does not conflict with MTU selector.
> > 
> > But it only needs the MTU on each local side (once for the REQ and on
> > the remote side for the REP). It would mean that if the local side were
> > capable of larger MTU and the remote side were Tavor, that the REQ would
> > be REJ with MTU too large and need to be retried at a smaller MTU.
> 
> This has 3 implications that make it impractical:
> . connection rate will suffer greatly
> . this will need ot be done in each ulp, and it's a lot of code
> . protocols such as sdp explicitly say what to do on rej
>   and do not seem to speak about retries

OK.

> > > However
> > > 1. Even opensm does not implement this optimization yet
> > 
> > What optimization ? I don't understand what you are saying OpenSM
> > doesn't support.
> > 
> > > 2. We need to work with existing SMs too
> > 
> > Not sure what the SA issue is here.
> 
> If path MTU selector in path query allows MTU 1K (e.g. "best MTU")
> and one of the sides is Tavor, select the best MTU that is 1K
> and not the largest possible.

How would it be identified if the SA supports this ?

> If path MTU selector requires 2K MTU, return path with 2K MTU.

Also, I'm not sure that this is the required difference in the SA
requests :-(

-- Hal