Re: [Qemu-devel] [PATCH v2] block: Introduce "null" driver

[Fam Zheng <famz@redhat.com>]

On Thu, 08/28 11:52, Markus Armbruster wrote:
> Fam Zheng <famz@redhat.com> writes:
> 
> > On Thu, 08/28 10:22, Markus Armbruster wrote:
> >> Fam Zheng <famz@redhat.com> writes:
> >> 
> >> > This is an analogue to Linux null_blk. It can be used for testing block
> >> > device emulation and general block layer functionalities such as
> >> > coroutines and throttling, where disk IO is not necessary or wanted.
> >> >
> >> > Use null:// for AIO version, and null-co:// for coroutine version.
> >> >
> >> > Signed-off-by: Fam Zheng <famz@redhat.com>
> >> >
> >> > ---
> >> > V2: Don't #ifdef code, add two drivers. (Benoit)
> >> >     Add to QAPI BlockdevOptions. (Eric)
> >> >     Add "file.size" option to override backend size. (What is a better
> >> >     way to associate /dev/vd{a,b,c} with command line devices, if sizes
> >> >     are the same?)
> >> 
> >> Is this a request for advice?  If yes, please explain the problem in
> >> more detail, because I'm too dense to get it :)
> >> 
> >
> > Yes :)
> >
> > I have more than one virtio-blk:
> >
> > -drive file=/dev/ram0,id=d0,if=none -device
> > virtio-blk-pci,drive=d0,x-datap-plane=on \
> > -drive file=/dev/ram1,id=d0,if=none -device virtio-blk-pci,drive=d1
> >
> > Then I get /dev/vda and /dev/vdb in guest. They are in the same size.
> >
> > The question is how do I tell which is /dev/ram0 and which is /dev/ram1,
> > without bothering peaking PCI addr, etc.?
> 
> Set virtio-blk-pci's serial property to a suitable value.  The serial
> number is visible in a Linux guest as
> /sys/bus/virtio/devices/virtio0/block/vda/serial or similar.
> 

Great, then I get a readable symlink in /dev/disk/by-id/.

Thanks,
Fam