Thanks for the explanation. I also thought of same purpose for this patch but I 
got confused by function names and data structure names in the patch. They 
still have lguest in it.

Thanks
-Bharat

> -----Original Message-----
> From: Magnus Damm [mailto:[email protected]]
> Sent: Friday, June 24, 2011 6:44 AM
> To: Bhushan Bharat-R65777
> Cc: [email protected]; [email protected]; [email protected];
> [email protected]; [email protected]
> Subject: Re: [PATCH 02/02] virtio: Add virtio platform driver
> 
> Hi Bhushan Bharat,
> 
> On Thu, Jun 23, 2011 at 5:54 PM, Bhushan Bharat-R65777
> <[email protected]> wrote:
> > What purpose this virtio platform driver serves? Are you trying to make
> driver layers here and each virtio device driver can hook up into this
> platform driver. Is that correct ?
> 
> The idea behind the virto platform driver is to have a reusable layer
> that handles common virtio and vring functionality. Regular virtio
> drivers for console, network and disk and interfacing to the virtio
> platform driver, so these paravirtualized drivers can with this virtio
> platform driver easily be reused by embedded platforms.
> 
> Vendor specific code is abstracted away from the virtio platform driver
> and interfaced to using regular interrupts and SoC-specific callbacks. So
> with this patch virtio can be interfaced as any regular platform driver.
> This version of the driver is simply reusing the configuration method
> from lguest.
> 
> If you want to see all code involved (including extremely experimental
> backing drivers) then have a look at:
> http://www.spinics.net/lists/linux-sh/msg07188.html
> 
> Cheers,
> 
> / magnus


_______________________________________________
Virtualization mailing list
[email protected]
https://lists.linux-foundation.org/mailman/listinfo/virtualization

Reply via email to