[RFC PATCH 0/3] kvm/arm: New VMID allocator based on asid(2nd approach)

[Shameerali Kolothum Thodi]

Hi Marc,

> -----Original Message-----
> From: Marc Zyngier [mailto:maz at kernel.org]
> Sent: 04 June 2021 14:55
> To: Shameerali Kolothum Thodi <shameerali.kolothum.thodi at huawei.com>
> Cc: linux-arm-kernel at lists.infradead.org; kvmarm at lists.cs.columbia.edu;
> linux-kernel at vger.kernel.org; will at kernel.org; catalin.marinas at arm.com;
> james.morse at arm.com; julien.thierry.kdev at gmail.com;
> suzuki.poulose at arm.com; jean-philippe at linaro.org; Alexandru Elisei
> <Alexandru.Elisei at arm.com>; Linuxarm <linuxarm at huawei.com>
> Subject: Re: [RFC PATCH 0/3] kvm/arm: New VMID allocator based on asid(2nd
> approach)
> 
> On Fri, 04 Jun 2021 09:13:10 +0100,
> Shameerali Kolothum Thodi <shameerali.kolothum.thodi at huawei.com>
> wrote:
> >
> > Hi,
> >
> > > -----Original Message-----
> > > From: Shameerali Kolothum Thodi
> > > Sent: 06 May 2021 17:52
> > > To: linux-arm-kernel at lists.infradead.org; kvmarm at lists.cs.columbia.edu;
> > > linux-kernel at vger.kernel.org
> > > Cc: maz at kernel.org; will at kernel.org; catalin.marinas at arm.com;
> > > james.morse at arm.com; julien.thierry.kdev at gmail.com;
> > > suzuki.poulose at arm.com; jean-philippe at linaro.org; Linuxarm
> > > <linuxarm at huawei.com>
> > > Subject: [RFC PATCH 0/3] kvm/arm: New VMID allocator based on asid(2nd
> > > approach)
> > >
> > > This is based on a suggestion from Will [0] to try out the asid
> > > based kvm vmid solution as a separate VMID allocator instead of
> > > the shared lib approach attempted in v4[1].
> > >
> > > The idea is to compare both the approaches and see whether the
> > > shared lib solution with callbacks make sense or not.
> >
> > A gentle ping on this. Please take a look and let me know.
> 
> I had a look and I don't overly dislike it. I'd like to see the code
> without the pinned stuff though, at least to ease the reviewing. I
> haven't tested it in anger, but I have pushed the rebased code at [1]
> as it really didn't apply to 5.13-rc4.

Thanks for taking a look and the rebase. I will remove the pinned stuff
in the next revision as that was added just to compare against the previous
version.

> 
> One thing I'm a bit worried about is that we so far relied on VMID 0
> never being allocated to a guest, which is now crucial for protected
> KVM. I can't really convince myself that this can never happen with
> this.

Hmm..not sure I quite follow that. As per the current logic vmid 0 is
reserved and is not allocated to Guest. 

> Plus, I've found this nugget:
> 
> <quote
> 	max_pinned_vmids = NUM_USER_VMIDS - num_possible_cpus() - 2;
> </quote>
> 
> What is this "- 2"? My hunch is that it should really be "- 1" as VMID
> 0 is reserved, and we have no equivalent of KPTI for S2.

I think this is more related to the "pinned vmid" stuff and was borrowed from
the asid_update_limit() fn in arch/arm64/mm/context.c. But I missed the
comment that explains the reason behind it. It says,

---x---
	/*
	 * There must always be an ASID available after rollover. Ensure that,
	 * even if all CPUs have a reserved ASID and the maximum number of ASIDs
	 * are pinned, there still is at least one empty slot in the ASID map.
	 */
	max_pinned_asids = num_available_asids - num_possible_cpus() - 2;
---x---

So this is to make sure we will have at least one VMID available after rollover
in case we have pinned the max number of VMIDs. I will include that comment
to make it clear.

Thanks,
Shameer

> 
> 	M.
> 
> [1]
> https://git.kernel.org/pub/scm/linux/kernel/git/maz/arm-platforms.git/log/?h
> =kvm-arm64/mmu/vmid
> 
> --
> Without deviation from the norm, progress is not possible.