From: Will Deacon <will@kernel.org>
To: Marc Zyngier <maz@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org,
kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, qperret@google.com,
dbrazdil@google.com, Srivatsa Vaddagiri <vatsa@codeaurora.org>,
Shanker R Donthineni <sdonthineni@nvidia.com>,
James Morse <james.morse@arm.com>,
Suzuki K Poulose <suzuki.poulose@arm.com>,
Alexandru Elisei <alexandru.elisei@arm.com>,
kernel-team@android.com
Subject: Re: [PATCH 06/16] KVM: arm64: Force a full unmap on vpcu reinit
Date: Fri, 30 Jul 2021 13:50:09 +0100 [thread overview]
Message-ID: <20210730125008.GB23756@willie-the-truck> (raw)
In-Reply-To: <87wnpad8pg.wl-maz@kernel.org>
On Wed, Jul 28, 2021 at 11:38:35AM +0100, Marc Zyngier wrote:
> On Tue, 27 Jul 2021 19:11:33 +0100,
> Will Deacon <will@kernel.org> wrote:
> >
> > On Thu, Jul 15, 2021 at 05:31:49PM +0100, Marc Zyngier wrote:
> > > As we now keep information in the S2PT, we must be careful not
> > > to keep it across a VM reboot, which could otherwise lead to
> > > interesting problems.
> > >
> > > Make sure that the S2 is completely discarded on reset of
> > > a vcpu, and remove the flag that enforces the MMIO check.
> > >
> > > Signed-off-by: Marc Zyngier <maz@kernel.org>
> > > ---
> > > arch/arm64/kvm/arm.c | 8 +++++++-
> > > 1 file changed, 7 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
> > > index 97ab1512c44f..b0d2225190d2 100644
> > > --- a/arch/arm64/kvm/arm.c
> > > +++ b/arch/arm64/kvm/arm.c
> > > @@ -1096,12 +1096,18 @@ static int kvm_arch_vcpu_ioctl_vcpu_init(struct kvm_vcpu *vcpu,
> > > * ensuring that the data side is always coherent. We still
> > > * need to invalidate the I-cache though, as FWB does *not*
> > > * imply CTR_EL0.DIC.
> > > + *
> > > + * If the MMIO guard was enabled, we pay the price of a full
> > > + * unmap to get back to a sane state (and clear the flag).
> > > */
> > > if (vcpu->arch.has_run_once) {
> > > - if (!cpus_have_final_cap(ARM64_HAS_STAGE2_FWB))
> > > + if (!cpus_have_final_cap(ARM64_HAS_STAGE2_FWB) ||
> > > + test_bit(KVM_ARCH_FLAG_MMIO_GUARD, &vcpu->kvm->arch.flags))
> > > stage2_unmap_vm(vcpu->kvm);
> > > else
> > > icache_inval_all_pou();
> > > +
> > > + clear_bit(KVM_ARCH_FLAG_MMIO_GUARD, &vcpu->kvm->arch.flags);
> >
> > What prevents this racing with another vCPU trying to set the bit?
>
> Not much. We could take the kvm lock on both ends to serialize it, but
> that's pretty ugly. And should we care? What is the semantic of
> resetting a vcpu while another is still running?
It's definitely weird, but given that this is an attack vector I don't think
we can rule out attackers trying whacky stuff like this (although maybe
we end up forbidding vcpu reset in pKVM -- dunno).
> If we want to support this sort of behaviour, then our tracking is
> totally bogus, because it is VM-wide. And you don't even have to play
> with that bit from another vcpu: all the information is lost at the
> point where we unmap the S2 PTs.
>
> Maybe an alternative is to move this to the halt/reboot PSCI handlers,
> making it clearer what we expect?
I think that's probably worth looking at. The race is quite hard to reason
about otherwise, so if clearing the bit can be done on the teardown path
in single-threaded context then I think that's better. It looks like
kvm_prepare_system_event() has all the synchronisation we need there.
Will
next prev parent reply other threads:[~2021-07-30 12:50 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-15 16:31 [PATCH 00/16] KVM: arm64: MMIO guard PV services Marc Zyngier
2021-07-15 16:31 ` [PATCH 01/16] KVM: arm64: Generalise VM features into a set of flags Marc Zyngier
2021-07-27 18:10 ` Will Deacon
2021-07-28 9:41 ` Marc Zyngier
2021-07-28 14:51 ` Steven Price
2021-07-15 16:31 ` [PATCH 02/16] KVM: arm64: Don't issue CMOs when the physical address is invalid Marc Zyngier
2021-07-19 17:18 ` Quentin Perret
2021-07-20 8:04 ` Marc Zyngier
2021-07-27 18:10 ` Will Deacon
2021-07-28 9:45 ` Marc Zyngier
2021-07-15 16:31 ` [PATCH 03/16] KVM: arm64: Turn kvm_pgtable_stage2_set_owner into kvm_pgtable_stage2_annotate Marc Zyngier
2021-07-20 10:09 ` Quentin Perret
2021-07-20 10:21 ` Marc Zyngier
2021-07-20 10:38 ` Quentin Perret
2021-07-20 11:20 ` Marc Zyngier
2021-07-20 11:36 ` Quentin Perret
2021-07-20 13:13 ` Marc Zyngier
2021-07-15 16:31 ` [PATCH 04/16] KVM: arm64: Add MMIO checking infrastructure Marc Zyngier
2021-07-20 11:13 ` Quentin Perret
2021-07-20 13:15 ` Marc Zyngier
2021-07-20 15:49 ` Quentin Perret
2021-07-22 18:04 ` Marc Zyngier
2021-07-23 10:16 ` Quentin Perret
2021-07-27 18:11 ` Will Deacon
2021-07-28 9:57 ` Marc Zyngier
2021-07-30 12:26 ` Will Deacon
2021-07-30 13:04 ` Marc Zyngier
2021-07-30 12:58 ` Quentin Perret
2021-07-15 16:31 ` [PATCH 05/16] KVM: arm64: Plumb MMIO checking into the fault handling Marc Zyngier
2021-07-27 18:11 ` Will Deacon
2021-07-28 10:21 ` Marc Zyngier
2021-07-30 12:38 ` Will Deacon
2021-07-15 16:31 ` [PATCH 06/16] KVM: arm64: Force a full unmap on vpcu reinit Marc Zyngier
2021-07-27 18:11 ` Will Deacon
2021-07-28 10:38 ` Marc Zyngier
2021-07-30 12:50 ` Will Deacon [this message]
2021-07-15 16:31 ` [PATCH 07/16] KVM: arm64: Wire MMIO guard hypercalls Marc Zyngier
2021-07-27 18:11 ` Will Deacon
2021-07-28 10:47 ` Marc Zyngier
2021-07-30 13:11 ` Will Deacon
2021-08-01 11:20 ` Marc Zyngier
2021-07-15 16:31 ` [PATCH 08/16] KVM: arm64: Add tracepoint for failed MMIO guard check Marc Zyngier
2021-07-15 16:31 ` [PATCH 09/16] KVM: arm64: Advertise a capability for MMIO guard Marc Zyngier
2021-07-15 16:31 ` [PATCH 10/16] KVM: arm64: Add some documentation for the MMIO guard feature Marc Zyngier
2021-07-21 21:17 ` Andrew Jones
2021-07-23 13:30 ` Marc Zyngier
2021-07-23 13:38 ` Andrew Jones
2021-07-23 13:52 ` Marc Zyngier
2021-07-15 16:31 ` [PATCH 11/16] firmware/smccc: Call arch-specific hook on discovering KVM services Marc Zyngier
2021-07-15 16:31 ` [PATCH 12/16] mm/ioremap: Add arch-specific callbacks on ioremap/iounmap calls Marc Zyngier
2021-07-27 18:12 ` Will Deacon
2021-07-28 11:01 ` Marc Zyngier
2021-07-30 14:07 ` Will Deacon
2021-07-15 16:31 ` [PATCH 13/16] arm64: Implement ioremap/iounmap hooks calling into KVM's MMIO guard Marc Zyngier
2021-07-15 16:31 ` [PATCH 14/16] arm64: Enroll into KVM's MMIO guard if required Marc Zyngier
2021-07-15 16:31 ` [PATCH 15/16] arm64: Add a helper to retrieve the PTE of a fixmap Marc Zyngier
2021-07-20 11:16 ` Quentin Perret
2021-07-15 16:31 ` [PATCH 16/16] arm64: Register earlycon fixmap with the MMIO guard Marc Zyngier
2021-07-21 21:42 ` [PATCH 00/16] KVM: arm64: MMIO guard PV services Andrew Jones
2021-07-22 10:00 ` Marc Zyngier
2021-07-22 13:25 ` Andrew Jones
2021-07-22 15:30 ` Marc Zyngier
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210730125008.GB23756@willie-the-truck \
--to=will@kernel.org \
--cc=alexandru.elisei@arm.com \
--cc=dbrazdil@google.com \
--cc=james.morse@arm.com \
--cc=kernel-team@android.com \
--cc=kvm@vger.kernel.org \
--cc=kvmarm@lists.cs.columbia.edu \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maz@kernel.org \
--cc=qperret@google.com \
--cc=sdonthineni@nvidia.com \
--cc=suzuki.poulose@arm.com \
--cc=vatsa@codeaurora.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).