Re: [PATCH net-next] xen-netback: improve guest-receive-side flow control

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

On Mon, Dec 02, 2013 at 11:55:50AM +0000, Paul Durrant wrote:
> > -----Original Message-----
> > From: David Vrabel
> > Sent: 02 December 2013 11:52
> > To: Paul Durrant
> > Cc: xen-devel@lists.xen.org; Wei Liu; Ian Campbell
> > Subject: Re: [PATCH net-next] xen-netback: improve guest-receive-side flow
> > control
> > 
> > On 28/11/13 13:11, Paul Durrant wrote:
> > > The flow control code relies on a double pass of tke skb, firstly to count
> > > the number of ring slots needed to supply sufficient grant references, and
> > > another to actually consume those references and build the grant copy
> > > operations. It transpires that, in some scenarios, the initial count and the
> > > number of references consumed differs and, after this happens a number
> > of
> > > times, flow control is completely broken.
> > 
> > Can you describe in more detail the cases where it gets it wrong?
> > 
> 
> I've already elaborated on this thread (at Annie Li's request). That explanation is probably a little long for inclusion in the comment though.


Please do add it. It never hurts to have more documentation.

> 
> > > -unsigned int xenvif_count_skb_slots(struct xenvif *vif, struct sk_buff
> > *skb)
> > 
> > Any patch that gets rid of this gets my vote (provided it works!).
> > 
> > Can you list all the guest frontends that you've tested this with?
> > 
> 
> So, far a couple of windows guests for dev test purposes, but it is going through XenRT too which will exercise linux guests too.
> 
>   Paul
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel