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

[Michael S. Tsirkin]

Quoting r. Sean Hefty <mshefty at ichips.intel.com>:
> Subject: Re: [PATCH for-2.6.18] IB/cma: option to limitMTU to 1K
> 
> Hal Rosenstock wrote:
> > 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.
> 
> I agree with this approach.  The user should determine the proper MTU based on 
> local information, and either set it to 1k if sending a REQ, or REJ the REQ if 
> the MTU is too large.  I'm not sure that this policy should be in the CMA, 
> versus the consumer, but I can go with the CMA.
> 
> I do think that the MTU could be negotiated down as part of the private data in 
> the REP, but this would need to be done outside of the CMA.
> 
> - Sean

Putting knowledge about hw quirks in all protocols is really horrible.

MTU should be decided by SA as part of path information.
If ULPs have spicific limitations wrt MTU they should use mtu selector
in path record query.

-- 
MST