Re: [PATCH v2 3/3] kvm/arm: Align the VMID allocation with the arm64 ASID one

From: Marco Elver <elver@google.com>
To: Quentin Perret <qperret@google.com>
Cc: Shameerali Kolothum Thodi <shameerali.kolothum.thodi@huawei.com>,
	Will Deacon <will@kernel.org>,
	"linux-arm-kernel@lists.infradead.org" 
	<linux-arm-kernel@lists.infradead.org>,
	"kvmarm@lists.cs.columbia.edu" <kvmarm@lists.cs.columbia.edu>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"maz@kernel.org" <maz@kernel.org>,
	"catalin.marinas@arm.com" <catalin.marinas@arm.com>,
	"james.morse@arm.com" <james.morse@arm.com>,
	"julien.thierry.kdev@gmail.com" <julien.thierry.kdev@gmail.com>,
	"suzuki.poulose@arm.com" <suzuki.poulose@arm.com>,
	"jean-philippe@linaro.org" <jean-philippe@linaro.org>,
	"Alexandru.Elisei@arm.com" <Alexandru.Elisei@arm.com>,
	Linuxarm <linuxarm@huawei.com>,
	mark.rutland@arm.com
Subject: Re: [PATCH v2 3/3] kvm/arm: Align the VMID allocation with the arm64 ASID one
Date: Thu, 22 Jul 2021 21:33:52 +0200	[thread overview]
Message-ID: <YPnIIAARHNhx9npt@elver.google.com> (raw)
In-Reply-To: <YPk2XqrOeP6dEtPL@google.com>

On Thu, Jul 22, 2021 at 10:11AM +0100, Quentin Perret wrote:
> On Thursday 22 Jul 2021 at 06:45:14 (+0000), Shameerali Kolothum Thodi wrote:
> > > From: Will Deacon [mailto:will@kernel.org]
> > > > diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> > > b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> > > > index 4b60c0056c04..a02c4877a055 100644
> > > > --- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> > > > +++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
> > > > @@ -106,8 +106,7 @@ int kvm_host_prepare_stage2(void *mem_pgt_pool,
> > > void *dev_pgt_pool)
> > > >  	mmu->pgd_phys = __hyp_pa(host_kvm.pgt.pgd);
> > > >  	mmu->arch = &host_kvm.arch;
> > > >  	mmu->pgt = &host_kvm.pgt;
> > > > -	mmu->vmid.vmid_gen = 0;
> > > > -	mmu->vmid.vmid = 0;
> > > > +	atomic64_set(&mmu->vmid.id, 0);
> > > 
> > > I think this is the first atomic64 use in the EL2 object, which may pull in
> > > some fatal KCSAN instrumentation. Quentin, have you run into this before?
> > > 
> > > Might be simple just to zero-initialise mmu for now, if it isn't already.
> > 
> > I will check that.
> 
> Yes I think what saves us here is that, AFAICT. arm64 doesn't support
> KCSAN yet. But the day it does, this should fail to link (hopefully)
> because of out-of-line calls into e.g. __kasan_check_write().
> 
> So yes, a simple zeroing here is probably preferable.

Note: Do not worry about hypothetically breaking with sanitizers here --
whether it's KASAN or KCSAN, they both instrument atomics. In files that
enable instrumentation but the atomic instrumentation should not be
pulled in, use the arch_ variants, but this doesn't apply here because
instrumentation shouldn't even be on.

The indicator that when KCSAN is supported on arm64, the Makefile here
just needs KCSAN_SANITIZE := n, is that all other instrumentation is
also killed entirely:

  $ grep -E "(PROFILE|SANITIZE|INSTRUMENT)" arch/arm64/kvm/hyp/nvhe/Makefile
  GCOV_PROFILE	:= n
  KASAN_SANITIZE	:= n
  UBSAN_SANITIZE	:= n
  KCOV_INSTRUMENT	:= n

KCSAN isn't supported on arm64 yet, and when it does, I believe Mark's
arm64 KCSAN series should take care of things like this.

Thanks,
-- Marco