Re: [Xen-devel] [PATCH 0 of 7] libxl: refactor tap disk handling

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Thu, 2011-04-07 at 13:31 +0100, Christoph Egger wrote:
> On 04/07/11 11:52, Ian Campbell wrote:
> > I'm not personally convinced that support for blktap2 devices should
> > be conflated in libxl with the PV block backend support but given it's
> > there lets at least correct it.
> >
> > In a blktap2 system there is no "tap" PV backend type. blktap2 exposes
> > a standard block device and this is passed to a guest using the
> > standard blkback "vbd" (sometimes called "phy") backend. Drop the
> > "tap" backend type and associated (libxl internal ) DEVICE_TAP
> > enumeration value.
> 
> What happens on a no-blktap2 system?

Same as before, libxl__blktap_enabled still returns false and the same
things should happen as a result.

Ian.
> 
> >
> > Also try and clarify the paths which fallback from blktap2 to qdisk
> > when the former is not present. Falling through a switch statement is
> > a neat way of doing this in some cases, but I don't think this is one
> > of them.
> >
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@xxxxxxxxxxxxxxxxxxx
> > http://lists.xensource.com/xen-devel
> >
> 
> 



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel