All of lore.kernel.org
 help / color / mirror / Atom feed
From: Christoffer Dall <christoffer.dall@linaro.org>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: Marc Zyngier <marc.zyngier@arm.com>,
	Claudio Fontana <claudio.fontana@huawei.com>,
	Jan Kiszka <jan.kiszka@web.de>,
	"kvmarm@lists.cs.columbia.edu" <kvmarm@lists.cs.columbia.edu>,
	arm-mail-list <linux-arm-kernel@lists.infradead.org>,
	kvm-devel <kvm@vger.kernel.org>,
	Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [RFC PATCH] KVM: arm/arm64: Don't let userspace update CNTVOFF once guest is running
Date: Thu, 9 Jul 2015 18:06:31 +0200	[thread overview]
Message-ID: <20150709160631.GA25650@cbox> (raw)
In-Reply-To: <CAFEAcA81LRcRu1r+H-6T_yT-AQ4XLAzX1osAWZYNPEDHSTiAQg@mail.gmail.com>

On Thu, Jul 09, 2015 at 03:26:20PM +0100, Peter Maydell wrote:
> On 9 July 2015 at 15:17, Christoffer Dall <christoffer.dall@linaro.org> wrote:
> > On Thu, Jul 09, 2015 at 02:24:06PM +0200, Christoffer Dall wrote:
> > So I ran this through GDB, and this happens when the guest probes the
> > virtio devices, specifically the backtrace tells me that
> >
> > virtio_current_cpu_endian () at /root/src/qemu/hw/virtio/virtio.c:594
> > => cc->virtio_is_big_endian
> >   -> arm_cpu_is_big_endian
> >     -> cpu_synchronize_state
> >       -> kvm_cpu_synchronize_state
> >
> > which causes cpu->kvm_vcpu_dirty = true, which causes the run-loop to
> > write the CNTVOFF on a per-vcpu basis without stopping anything, as far
> > as I can tell.
> 
> Ah, I was wondering if it was going to turn out to be this,
> but I hadn't figured out why it was going to cause us to do
> a write-back of state rather than just a read.
> 
> > So yeah, I guess the only required fix here is to fix QEMU in some way
> > as to not fiddle with the timer registers in this way, and then I
> > honestly don't know if we should try to fix legacy userspace in the
> > kernel, but based on the feedback from Jan, I suppose not.
> 
> arm_cpu_is_big_endian() doesn't actually want to write
> back any state -- all it's interested in is reading.
> So we really ought not to need to write anything back there.
> kvm-all.c's sync functions don't seem to provide a "sync
> kernel state to userspace but I promise I'm not going to
> dirty it" function though. I guess we could add one.
> 
> (Overall it's kind of fragile even if we can avoid this
> specific case where we're writing back the counter state,
> though -- we should probably think about the sync level
> stuff as well.)

Sounds to me like we should add the sync level stuff first, as it should
catch all callers, and afterwards consider adding the sync-one-way
optimization.

-Christoffer

WARNING: multiple messages have this Message-ID (diff)
From: christoffer.dall@linaro.org (Christoffer Dall)
To: linux-arm-kernel@lists.infradead.org
Subject: [RFC PATCH] KVM: arm/arm64: Don't let userspace update CNTVOFF once guest is running
Date: Thu, 9 Jul 2015 18:06:31 +0200	[thread overview]
Message-ID: <20150709160631.GA25650@cbox> (raw)
In-Reply-To: <CAFEAcA81LRcRu1r+H-6T_yT-AQ4XLAzX1osAWZYNPEDHSTiAQg@mail.gmail.com>

On Thu, Jul 09, 2015 at 03:26:20PM +0100, Peter Maydell wrote:
> On 9 July 2015 at 15:17, Christoffer Dall <christoffer.dall@linaro.org> wrote:
> > On Thu, Jul 09, 2015 at 02:24:06PM +0200, Christoffer Dall wrote:
> > So I ran this through GDB, and this happens when the guest probes the
> > virtio devices, specifically the backtrace tells me that
> >
> > virtio_current_cpu_endian () at /root/src/qemu/hw/virtio/virtio.c:594
> > => cc->virtio_is_big_endian
> >   -> arm_cpu_is_big_endian
> >     -> cpu_synchronize_state
> >       -> kvm_cpu_synchronize_state
> >
> > which causes cpu->kvm_vcpu_dirty = true, which causes the run-loop to
> > write the CNTVOFF on a per-vcpu basis without stopping anything, as far
> > as I can tell.
> 
> Ah, I was wondering if it was going to turn out to be this,
> but I hadn't figured out why it was going to cause us to do
> a write-back of state rather than just a read.
> 
> > So yeah, I guess the only required fix here is to fix QEMU in some way
> > as to not fiddle with the timer registers in this way, and then I
> > honestly don't know if we should try to fix legacy userspace in the
> > kernel, but based on the feedback from Jan, I suppose not.
> 
> arm_cpu_is_big_endian() doesn't actually want to write
> back any state -- all it's interested in is reading.
> So we really ought not to need to write anything back there.
> kvm-all.c's sync functions don't seem to provide a "sync
> kernel state to userspace but I promise I'm not going to
> dirty it" function though. I guess we could add one.
> 
> (Overall it's kind of fragile even if we can avoid this
> specific case where we're writing back the counter state,
> though -- we should probably think about the sync level
> stuff as well.)

Sounds to me like we should add the sync level stuff first, as it should
catch all callers, and afterwards consider adding the sync-one-way
optimization.

-Christoffer

  reply	other threads:[~2015-07-09 16:06 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-24 14:54 [RFC PATCH] KVM: arm/arm64: Don't let userspace update CNTVOFF once guest is running Marc Zyngier
2015-06-24 14:54 ` Marc Zyngier
2015-06-25  8:04 ` Christoffer Dall
2015-06-25  8:04   ` Christoffer Dall
2015-06-25  8:48   ` Marc Zyngier
2015-06-25  8:48     ` Marc Zyngier
2015-06-25  8:59   ` Claudio Fontana
2015-06-25  8:59     ` Claudio Fontana
2015-06-25  9:10     ` Peter Maydell
2015-06-25  9:10       ` Peter Maydell
2015-06-25  9:25       ` Claudio Fontana
2015-06-25  9:25         ` Claudio Fontana
2015-06-26  4:49         ` Jan Kiszka
2015-06-26  4:49           ` Jan Kiszka
2015-06-29 17:20           ` Claudio Fontana
2015-06-29 17:20             ` Claudio Fontana
2015-06-29 17:37             ` Peter Maydell
2015-06-29 17:37               ` Peter Maydell
2015-07-08 15:56               ` Marc Zyngier
2015-07-08 15:56                 ` Marc Zyngier
2015-07-08 16:06                 ` Peter Maydell
2015-07-08 16:06                   ` Peter Maydell
2015-07-08 16:37                   ` Marc Zyngier
2015-07-08 16:37                     ` Marc Zyngier
2015-07-08 19:13                     ` Peter Maydell
2015-07-08 19:13                       ` Peter Maydell
2015-07-09 10:22                       ` Christoffer Dall
2015-07-09 10:22                         ` Christoffer Dall
2015-07-09 10:38                         ` Peter Maydell
2015-07-09 10:38                           ` Peter Maydell
2015-07-09 12:05                           ` Christoffer Dall
2015-07-09 12:05                             ` Christoffer Dall
2015-07-09 12:07                             ` Peter Maydell
2015-07-09 12:07                               ` Peter Maydell
2015-07-09 12:24                               ` Christoffer Dall
2015-07-09 12:24                                 ` Christoffer Dall
2015-07-09 14:17                                 ` Christoffer Dall
2015-07-09 14:17                                   ` Christoffer Dall
2015-07-09 14:26                                   ` Peter Maydell
2015-07-09 14:26                                     ` Peter Maydell
2015-07-09 16:06                                     ` Christoffer Dall [this message]
2015-07-09 16:06                                       ` Christoffer Dall
2015-07-09 10:40                         ` Jan Kiszka
2015-07-09 10:40                           ` Jan Kiszka
2015-07-09 12:08                           ` Christoffer Dall
2015-07-09 12:08                             ` Christoffer Dall

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=20150709160631.GA25650@cbox \
    --to=christoffer.dall@linaro.org \
    --cc=claudio.fontana@huawei.com \
    --cc=jan.kiszka@web.de \
    --cc=kvm@vger.kernel.org \
    --cc=kvmarm@lists.cs.columbia.edu \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=marc.zyngier@arm.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.