[ofa-general] Re: [ewg] [PATCH] libibmad: Handle MAD redirection

[Jason Gunthorpe]

On Tue, Jun 30, 2009 at 02:04:03PM +0200, Joachim Fenkes wrote:
> On Tuesday 30 June 2009 00:01, Hal Rosenstock wrote:
> > On Mon, Jun 29, 2009 at 8:10 AM, Joachim Fenkes<fenkes at de.ibm.com> wrote:
> 
> > > Previously, libibmad reacted to GSI MAD responses with a "redirect" status
> > > by throwing an error. IBM eHCA adapters use redirection, so most
> > > infiniband_diags tools didn't work against eHCA.
> > 
> > Are there GS classes other than PerfMgt which would be redirected by eHCA ?
> 
> Not right now, no. If you're interested in the details of how and when the
> eHCA driver redirects, please have a look at drivers/infiniband/hw/ehca/ehca_sqp.c.

Hmm.. That definately doesn't look right overall. You are not forming
the redirect reply in a way that will work with all possible
fabrics. You can't just return a 0 SL and the default PKey and assume
things will work out.

It looks like all you want to do is redirect to a different QPN? If so
I recommend you copy all the values from the incoming MAD's LRH and,
if present, GRH into the ClassPortInfo reply. Copy the PKey too.

This way you have the best chance of sending back the right information.

If there is no GRH then you can use GID index 0 and a 0 TC and a 0
FL. According to the spec returning the port GID is NOT optional.

> > > +            /* update dport for next request and retry */
> > > +            dport->lid = mad_get_field(mad, 64, IB_CPI_REDIRECT_LID_F);
> > > +            dport->qp = mad_get_field(mad, 64, IB_CPI_REDIRECT_QP_F);
> > > +            dport->qkey = mad_get_field(mad, 64, IB_CPI_REDIRECT_QKEY_F);

This code sould also check for 0 LID and bail.

Jason