[ofa-general] vendor_id in struct ib_device_attr

[Jeff Squyres]

On Jun 23, 2008, at 4:02 PM, Jason Gunthorpe wrote:

> So unless people are abusing this, a PCI ID is tied to a piece of
> sillicon, while OUI+Subsystem ID+GUID/MAC is tied to the board
> vendor.. So, yes, there tend to be many PCI IDs for sillicon that is
> very similar, but the OUI tells you nothing about the underlying
> sillicon.
>
> Using an OUI to trigger sillicon specific tuning only works while we
> have a small set of suppliers.

The same could be said for having a small set of PCI ID's, no?  There  
are both a small number of suppliers and PCI ID's today, so I guess it  
hasn't mattered to this point.

> I'm just commenting on Jeff's use of the vendor_id in OMPI, I have no
> real opinion on what should go in the structure member.

We use the (vendor_id, vendor_part_id) tuple to identify an HCA and  
the options that we should use with it because these were the two  
values available to us via the verbs API.  Is there something better  
that we should do?

I'm fairly ambivalent -- on what the actual value should be in vendor  
ID, as long as it's a consistent value/namespace.

-- 
Jeff Squyres
Cisco Systems