[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [RFC 0/4] virtio-mmio transport
From: |
Paul Brook |
Subject: |
Re: [Qemu-devel] [RFC 0/4] virtio-mmio transport |
Date: |
Mon, 12 Dec 2011 02:52:34 +0000 |
User-agent: |
KMail/1.13.7 (Linux/3.1.0-1-amd64; KDE/4.6.5; x86_64; ; ) |
> >> I'm not sure what guest software uses the syborg virtio transport.
> >
> > It is/was a virtual reference platform for SymbianOS. However since then
> > the Symbian Foundation got shot in the back of the head and the rest of
> > Nokia jumped ship to Windows, so I'd be surprised if anyone really cares
> > about it.
>
> Can we remove syborg virtio then?
If we do, then I recommend removing the whole of the rest of the syborg
machine at the same time.
AFAIK the only use of this machine was the Symbian Virtual Platform reference
environment provided by the Symbian Foundation. This organisation no longer
exists, and the code (both qemu and guest OS) they hosted is no longer
publicly available. The qemu.org code is also a stripped-down version of the
full SVP.
In short I do not have any good reasons to keep this code.
I've taken a look at the virtion-mmio spec, and it looks fairly reasonable.
The only thing I'd change is the GuestPageSize/QueuePFN mess. Seems like just
using straight 64-bit addresses would be a better solution. Maybe split into
a high/low pair to keep all registers as 32-bit registers.
Assuming a 4k page size, the pfn approach makes it painful to exceed a 44-bit
physical address space. Modern x86 cores have already supprt larger physical
address spaces, and architectural limits are even higher. Remember that
physical address space need not be densely populated.
Paul