|
From: | Anthony Liguori |
Subject: | Re: [Qemu-devel] [PATCH 28/35] kvm: x86: Introduce kvmclock device to save/restore its state |
Date: | Tue, 18 Jan 2011 11:31:19 -0600 |
User-agent: | Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.15) Gecko/20101027 Lightning/1.0b1 Thunderbird/3.0.10 |
On 01/18/2011 11:20 AM, Jan Kiszka wrote:
Which only works as along as we expose a single bus. You don't need to be an oracle to predict that this is not a stable interface.
Today we only have a very low level factory interface--device creation and deletion.
I think we should move to higher level bus factory interfaces. An interface to create a PCI device and to delete PCI devices. This is the only sane way to do hot plug.
This also makes supporting multiple busses a lot more reasonable since this factory interface could be a method of the controller.
It's automatically created as part of the CPUs or as part of the chipset. How to enable/disable kvm assistance is a property of the CPU and/or chipset.If we exclude creation via command line / config files, we could also pass the kvm_state directly from the machine or chipset setup code and save us at least the kvm system buses.
Which is fine in the short term. This is exactly why we don't want the device model to be an ABI. It gives us the ability to make changes as they make sense instead of trying to be perfect from the start (which we never will be).
Regards, Anthony Liguori
Jan
[Prev in Thread] | Current Thread | [Next in Thread] |