Re: [Qemu-devel] Re: [PATCH, RFC] virtio_blk: add cache control support

From: Anthony Liguori
Date: Wed Mar 23 2011 - 23:11:44 EST

On 03/17/2011 09:21 AM, Christoph Hellwig wrote:
On Thu, Mar 17, 2011 at 03:36:08PM +1030, Rusty Russell wrote:
I'm happ to switch strcmp.
Of course, that's assuming buf is nul terminated.
It's the string the user writes into it, which normally should be

No, it's intentional. config space writes can't return errors, so we need
to check that the value has really changed. I'll add a comment explaining it.
OK, under what circumstances could it fail?

If you're using this mechanism to indicate that the host doesn't support
the feature, that's making an assumption about the nature of config
space writes which isn't true for non-PCI virtio.

ie. lguest and S/390 don't trap writes to config space.

Or perhaps they should? But we should be explicit about needing it...
We have the features flag to indicate if updating the caching mode is
supported, but we we could still fail it for other reasons - e.g. we could fail
to reopen the file with/without O_SYNC. But if lguest or S/390 don't support
trapping config space write this feature won't work for them at all. As do
other features that make use of config space write, e.g. updating the MAC
address for virtio-net.

QEMU does not rely on config space writes for supporting mac address updates.

Whenever the config space is updated, we update the mac address in the virtio-net structure but since PCI only supports 4 byte accesses, it will only be partially updated at certain points in time.

This is okay though because as long as a guest updates it before we send a network packet out, when we refer to the mac address, it will be correct. We could just as well have the config space be in shared memory and only refer to the mac address in that shared memory area when transmitting a packet.

So the fact that we respond to config space writes doesn't mean that writes have a side effect other than updating the config space, which is really what I think the important point is here.


Anthony Liguori

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at