[Qemu-devel] Re: [PATCH 0/9] Virtio cleanups

[Michael S. Tsirkin]

On Thu, Mar 18, 2010 at 03:21:59PM +0100, Juan Quintela wrote:
> > My patch removes lines of code. It is actually simpler than
> > what we had: no casts, no assumptions.
> 
> It is more complex.  You need to add a new offset field to be able to
> free it in common code.  To add insult to injury, you have to do there a
> cast (without containeroff) to be able to free it there.  And once that
> you add the offset, the number of lines argument is also lost.

You seem to be looking at version 1 of the patch.
Please look at PATCHv2 I posted. No cast, no offset.

> >> (Yes, that is paraphrasing yourself in a funny way, pardon for the
> >> license )
> >> 
> >> about vdev & pcidev.  I still think that a lot of things would be easier
> >> if a vdev device is a pci_device _always_, not just some times.  And
> >> VirtIOPCIProxy shows it.
> >> 
> >> Later, Juan.
> >
> > Yes. But we don't always have pci. So the world we model does
> > not match single inheritance: a cow is both a mammal and a quadruped,
> > even if java programmers prefer it to be a mammal first of all :)
> 
> Look at it this other way, If I assume that I always have pci, how much
> I can simplify?
> 
> Later, Juan.

Quite a lot. We started this way, BTW.

-- 
MST