Re: [OpenBIOS] [PATCH 0/2] Sparc64: virtio-block and interrupt support

[Mark Cave-Ayland]

On 12/05/12 15:14, Artyom Tarasenko wrote:
> On Sat, May 12, 2012 at 3:54 PM, Blue Swirl  wrote:
> > On Sat, May 12, 2012 at 1:34 PM, Artyom Tarasenko  wrote:
> > > On Sat, May 12, 2012 at 1:42 PM, Blue Swirl  wrote:
> > > > Based on Artyom's Sparc64 work (thanks a lot!), I added some minimal
> > > > support for virtio-block and interrupt nodes.
> > >
> > > Hey, that was fast! Does it mean your patches also allow booting from
> > > a virtio device?
> >
> > No, only a warning for the unknown device is avoided and Linux can use
> > the interrupts.
> >
> > Making a simple driver does not look very difficult, here are some examples:
> > http://fxr.watson.org/fxr/source/dev/virtio/block/virtio_blk.c
> > http://fxr.watson.org/fxr/source/drivers/block/virtio_blk.c?v=linux-2.6
> > http://git.qemu.org/?p=seabios.git;a=blob;f=src/virtio-blk.c;h=194deaf98b41a...
>
> Fair enough. On the other hand, if we implement it too soon, there
> will be less motivation in fixing the cmd646 emulation in qemu... ;-)

What's the actual bug with cmd646 in QEMU?

ATB,

Mark.

[PATCH 0/2] Sparc64: virtio-block and interrupt support

[Artyom Tarasenko]

On Sat, May 12, 2012 at 5:15 PM, Mark Cave-Ayland
 wrote:
> On 12/05/12 15:14, Artyom Tarasenko wrote:
> > On Sat, May 12, 2012 at 3:54 PM, Blue Swirl  wrote:
> > > On Sat, May 12, 2012 at 1:34 PM, Artyom Tarasenko
> > >  wrote:
> > > > On Sat, May 12, 2012 at 1:42 PM, Blue Swirl
> > > >  wrote:
> > > > > Based on Artyom's Sparc64 work (thanks a lot!), I added some minimal
> > > > > support for virtio-block and interrupt nodes.
> > > >
> > > > Hey, that was fast! Does it mean your patches also allow booting from
> > > > a virtio device?
> > >
> > > No, only a warning for the unknown device is avoided and Linux can use
> > > the interrupts.
> > >
> > > Making a simple driver does not look very difficult, here are some
> > > examples:
> > > http://fxr.watson.org/fxr/source/dev/virtio/block/virtio_blk.c
> > > http://fxr.watson.org/fxr/source/drivers/block/virtio_blk.c?v=linux-2.6
> > >
> > > http://git.qemu.org/?p=seabios.git;a=blob;f=src/virtio-blk.c;h=194deaf98b41a...
> >
> > Fair enough. On the other hand, if we implement it too soon, there
> > will be less motivation in fixing the cmd646 emulation in qemu... ;-)
>
> What's the actual bug with cmd646 in QEMU?

It doesn't work. :)
There seems to be quite  a few cmd646 modifications. Some were capable
of DMA, which Linux tries to use.
The qemu implementation doesn't provide dma: IOMMU/DVMA are not implemented.
But why it doesn't work in PIO mode - I don't know.