Hi Russel,
From: Russell King [mailto:linux@armlinux.org.uk] Sent: Tuesday, November 1, 2022 11:24 AM To: Salil Mehta salil.mehta@huawei.com Cc: James Morse james.morse@arm.com; joyce.qi@linaro.org; linaro-open-discussions@op-lists.linaro.org; Jonathan Cameron jonathan.cameron@huawei.com; lorenzo.pieralisi@linaro.org; ilkka@os.amperecomputing.com; Jean-Philippe Brucker jean-philippe.brucker@arm.com; salil.mehta@opnsrc.net; mehta.salil.lnk@gmail.com Subject: Re: [Linaro-open-discussions] Re: Invitation: Linaro Open Discussions monthly meeting @ Wed 5 Oct 2022 19:00 - 20:00 (HKT) (linaro-open-discussions@op-lists.linaro.org)
On Tue, Nov 01, 2022 at 11:09:51AM +0000, Salil Mehta wrote:
PSCI has to be able to return an error if you try to online a disabled CPU,
(as without
enforcement the feature is pointless). So there is no need to track which
CPUs are
disabled in the MADT - just online the lot, and handle the error gracefully.
To reconfirm this point...
"without enforcement the feature is pointless" I completely agree - without that, having vCPU hotplug is an utterly pointless feature and a waste of time.
Sure, I agreed as well. There is no contention in that.
One of the whole points for vCPU hotplug is for the environment outside of the guest to control how many CPUs guests can use, and if the guest can online CPUs that the external environment has thought it has taken offline, then really this is a waste of time and is not what people want when they talk about vCPU hotplug.
Yes, of course. that's the key feature of vCPU Hotplug.
There has to be enforcement - and that enforcement can not be done solely by the guest kernel, it has to be done by the environment outside of the guest for the feature to have any meaningful application beyond an academic exercise.
Sure, agreed. We cannot trust guest as it can be compromised so there has to be a check in the VMM/firmware as well.
Thanks Salil