Re: [virtio-dev] A simple example of virtio blk driver

[Srinivas Nayak <sinu.nayak2001@gmail.com>]

Dear Friends,

Many thanks for your links.
I want a driver, not device.

Long back in 2013 I wrote a driver,
following Rusty's paper, "Towards a de facto standard..."
which used to fill only one descriptor chain
(the header-buffer-status chain for blk device) into available ring.
And after kick, when I get an IRQ,
I collect the data and then proceed to fill another such a request.
It felt sequential.

I read that, even before a single kick,
many such chains can be fit into available ring.
And thus getting the benefit of parallel execution of all requests.
Which is parallel.
But I couldn't do it.

And meanwhile that standard 0.95 was obsolete.
Now we have 1.0. Not sure how big the difference it brings.
So wanted to see a driver which takes benefit of parallelism.
And works with IRQ mechanism.

Say, I fill up 4 such requests and fired a single kick.
Will device interrupt me 4 times with IRQ?
I hope so...then only the benefit of parallelism comes...right?
But if so, when driver accesses used ring,
how can device use it to place the next served index?
Can there be a chance of IRQ miss?

Wanted to know about all this...


With thanks and best regards,

Yours sincerely,
Srinivas Nayak

Home: http://www.mathmeth.com/sn/
Blog: http://srinivas-nayak.blogspot.in/

On 06/08/2016 04:40 PM, Ian Campbell wrote:
> On Wed, 2016-06-08 at 07:06 -0400, Paolo Bonzini wrote:
> > > > > I am studying the virtio spec 1.0 and trying to write a virtio
> > > > > blk
> > > > > driver.
> > > > >
> > > > > Is there anywhere a simple example virtio blk driver can be
> > > > > found,
> > > > > which may be written with conformance to virtio spec 1.0?
> > > > > I would like to understand a simple virtio blk driver, how it
> > > > > is
> > > > > written.
> > > > >
> > > > > Better if there is any clear document available regarding a
> > > > > virtio
> > > > > blk driver implementation.
> > > > If you want to see a minimal driver, look at the SeaBIOS
> > > > src/hw/virtio-blk.c driver.  It only uses a subset of virtio-blk
> > > > functionality.
> > > >
> > > > https://github.com/qemu/seabios/blob/master/src/hw/virtio-blk.c
> > > There's also a reasonably small/simple implementation in the
> > > FreeBSD
> > > bhyve stuff, e.g.
> > >
> > > https://github.com/freebsd/freebsd/blob/master/usr.sbin/bhyve/pci_v
> > > irtio_block.c
> > This is an implementation of the device though, not the driver. :)
>
> Oh yes, ENOTENOUGHCOFFEE.
>
> > It's not 100% clear from Srinivas message whether he wanted a
> > device or a driver.
>
> Reading back I think he probably did mean a driver. My point about the
> low level "smarts" being in common code still stands ;-)
>
> Ian.