[openib-general] Re: [PATCH] iWARP Include File Changes

[Tom Tucker]

Sean:

I don't think mtcha needs to set these unless it supports FMR or SRQ.
When the device does, it should set them. The bits aren't currently
usedby any of the core code or ULP. However...

This change is intended to allow an application (and the core for that
matter) to determine if a device has an optional capability. This will
be important going forward when attempting to optimize a ULP and making
sure that optional features are not relied upon (e.g. SRQ), but are
available if they can be used to good effect. 

I think the mtcha device would set FMR, and ultimately SRQ when that is
ready. 

On Wed, 2006-01-18 at 16:38 -0800, Sean Hefty wrote:
> Tom Tucker wrote:
> >  enum ib_device_cap_flags {
> > @@ -86,6 +87,14 @@
> >  	IB_DEVICE_RC_RNR_NAK_GEN	= (1<<12),
> >  	IB_DEVICE_SRQ_RESIZE		= (1<<13),
> >  	IB_DEVICE_N_NOTIFY_CQ		= (1<<14),
> > +	IB_DEVICE_IN_ORD_PLCMNT         = (1<<15),
> > +	IB_DEVICE_ZERO_STAG             = (1<<16),
> > +	IB_DEVICE_SEND_W_INV            = (1<<17),
> > +	IB_DEVICE_MW                    = (1<<18),
> > +	IB_DEVICE_FMR                   = (1<<19),
> > +	IB_DEVICE_SRQ                   = (1<<20),
> > +	IB_DEVICE_ARP                   = (1<<21),
> > +	IB_DEVICE_LLP                   = (1<<22),
> >  };
> 
> Does this change imply that devices need to set these capabilities?  I.e. should 
> mthca and the other drivers be updated to set the device capabilities correctly?
> 
> - Sean