Re: [Xen-devel] Is: SKB_MAX_LEN bites again. Was: Re: bug disabling guest interface

[James Harper <james.harper@xxxxxxxxxxxxxxxx>]

> >
> > On this, while we just hacked this in the kernel source to increase
> > MAX_SKB_FRAGS to 19, I believe Jacek Milewicz tested 18 to work fine on
> > linux guests, but it seems Windows guests would still fail unless it was
> > increased to 19.
> 
> No netback has TTBOMK ever support 19, so I don't know how those guests
> ever worked, but regardless doing coalescing in the backend would fix
> this at least for frames which are <64K in size.
> 

Well Windows could easily generate a packet with more than 19 buffers, and I 
have engineered situations where it could be much larger than that, but you 
aren't going to see such packets during regular usage.

Is there any consensus yet on what form max_buffers negotiation would take? Is 
it as simple as the frontend and backend each writing the maximum value they 
can support (without further work) to their respective xenstores?

What backends exist these days? Are BSD and Solaris backends still in 
existence? What are their respective limitations?

James


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel