Re: [PATCH 11/12] hw/xen: automatically assign device index to block devices

[Kevin Wolf <kwolf@xxxxxxxxxx>]

Am 18.10.2023 um 12:52 hat David Woodhouse geschrieben:
> > Actually, how does this play together with xen_config_dev_blk()? This
> > looks like it tried to implement a very similar thing (which is IF_XEN
> > even already existed).
> 
> Ah yes, thanks for spotting that! I hadn't been looking at the xenfv
> 
> > Are we now trying to attach each if=xen disk twice in the 'xenpv'
> > machine? Or if something prevents this, is it dead code.
> 
> I suspect we end up creating them twice (and probably thus failing to
> lock the backing file).
>
> [...]
>
> ... but this just reinforces what I said there about "if
> qmp_device_add() can find the damn bus and do this right, why do we
> have to litter it through platform code?"

Indeed, if you can do -device, it's always the best option. For block
devices not the least because it gives you a way to use -blockdev with
it. I'm happy whenever I see a drive_get() call go away.

Kevin