From: Oliver Upton <oliver.upton@linux.dev>
To: Sean Christopherson <seanjc@google.com>
Cc: kvm@vger.kernel.org, Marc Zyngier <maz@kernel.org>,
Will Deacon <will@kernel.org>,
kvmarm@lists.linux.dev, Ben Gardon <bgardon@google.com>,
David Matlack <dmatlack@google.com>,
kvmarm@lists.cs.columbia.edu,
linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v3 09/15] KVM: arm64: Free removed stage-2 tables in RCU callback
Date: Tue, 1 Nov 2022 20:46:14 +0000 [thread overview]
Message-ID: <Y2GFliAVxui9VyK2@google.com> (raw)
In-Reply-To: <Y2GBVML5MWXZE9Na@google.com>
On Tue, Nov 01, 2022 at 08:28:04PM +0000, Sean Christopherson wrote:
> On Thu, Oct 27, 2022, Oliver Upton wrote:
> > There is no real urgency to free a stage-2 subtree that was pruned.
> > Nonetheless, KVM does the tear down in the stage-2 fault path while
> > holding the MMU lock.
> >
[ copy ]
> This is _very_ misleading. The above paints RCU as an optimization of sorts to
> avoid doing work while holding mmu_lock. Freeing page tables in an RCU callback
> is _required_ for correctness when allowing parallel page faults to remove page
> tables, as holding mmu_lock for read in that case doesn't ensure no other CPU is
> accessing and/or holds a reference to the to-be-freed page table.
Agree, but it is still important to reason about what is changing here
too. Moving work out of the vCPU fault path _is_ valuable, though
ancillary to the correctness requirements.
> IMO, this patch should to be squashed with the previous patch, "Protect stage-2
> traversal with RCU". One doesn't make any sense without the other.
I had split these up back when this series was a lot more gnarly and
there was too much slop in a single diff. That isn't the case any more,
so yeah I'll squash them.
[ paste ]
> > Free removed stage-2 subtrees after an RCU grace period. To guarantee
> > all stage-2 table pages are freed before killing a VM, add an
> > rcu_barrier() to the flush path.
An aside, this is flat-out wrong now.
--
Thanks,
Oliver
_______________________________________________
kvmarm mailing list
kvmarm@lists.cs.columbia.edu
https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
WARNING: multiple messages have this Message-ID (diff)
From: Oliver Upton <oliver.upton@linux.dev>
To: Sean Christopherson <seanjc@google.com>
Cc: Marc Zyngier <maz@kernel.org>, James Morse <james.morse@arm.com>,
Alexandru Elisei <alexandru.elisei@arm.com>,
linux-arm-kernel@lists.infradead.org,
kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org,
Reiji Watanabe <reijiw@google.com>,
Ricardo Koller <ricarkol@google.com>,
David Matlack <dmatlack@google.com>,
Quentin Perret <qperret@google.com>,
Ben Gardon <bgardon@google.com>, Gavin Shan <gshan@redhat.com>,
Peter Xu <peterx@redhat.com>, Will Deacon <will@kernel.org>,
kvmarm@lists.linux.dev
Subject: Re: [PATCH v3 09/15] KVM: arm64: Free removed stage-2 tables in RCU callback
Date: Tue, 1 Nov 2022 20:46:14 +0000 [thread overview]
Message-ID: <Y2GFliAVxui9VyK2@google.com> (raw)
Message-ID: <20221101204614.kqimwRxDINXrD-NIeaUu5fFWmyX9Rf88HNJi7E4YSak@z> (raw)
In-Reply-To: <Y2GBVML5MWXZE9Na@google.com>
On Tue, Nov 01, 2022 at 08:28:04PM +0000, Sean Christopherson wrote:
> On Thu, Oct 27, 2022, Oliver Upton wrote:
> > There is no real urgency to free a stage-2 subtree that was pruned.
> > Nonetheless, KVM does the tear down in the stage-2 fault path while
> > holding the MMU lock.
> >
[ copy ]
> This is _very_ misleading. The above paints RCU as an optimization of sorts to
> avoid doing work while holding mmu_lock. Freeing page tables in an RCU callback
> is _required_ for correctness when allowing parallel page faults to remove page
> tables, as holding mmu_lock for read in that case doesn't ensure no other CPU is
> accessing and/or holds a reference to the to-be-freed page table.
Agree, but it is still important to reason about what is changing here
too. Moving work out of the vCPU fault path _is_ valuable, though
ancillary to the correctness requirements.
> IMO, this patch should to be squashed with the previous patch, "Protect stage-2
> traversal with RCU". One doesn't make any sense without the other.
I had split these up back when this series was a lot more gnarly and
there was too much slop in a single diff. That isn't the case any more,
so yeah I'll squash them.
[ paste ]
> > Free removed stage-2 subtrees after an RCU grace period. To guarantee
> > all stage-2 table pages are freed before killing a VM, add an
> > rcu_barrier() to the flush path.
An aside, this is flat-out wrong now.
--
Thanks,
Oliver
WARNING: multiple messages have this Message-ID (diff)
From: Oliver Upton <oliver.upton@linux.dev>
To: Sean Christopherson <seanjc@google.com>
Cc: Marc Zyngier <maz@kernel.org>, James Morse <james.morse@arm.com>,
Alexandru Elisei <alexandru.elisei@arm.com>,
linux-arm-kernel@lists.infradead.org,
kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org,
Reiji Watanabe <reijiw@google.com>,
Ricardo Koller <ricarkol@google.com>,
David Matlack <dmatlack@google.com>,
Quentin Perret <qperret@google.com>,
Ben Gardon <bgardon@google.com>, Gavin Shan <gshan@redhat.com>,
Peter Xu <peterx@redhat.com>, Will Deacon <will@kernel.org>,
kvmarm@lists.linux.dev
Subject: Re: [PATCH v3 09/15] KVM: arm64: Free removed stage-2 tables in RCU callback
Date: Tue, 1 Nov 2022 20:46:14 +0000 [thread overview]
Message-ID: <Y2GFliAVxui9VyK2@google.com> (raw)
In-Reply-To: <Y2GBVML5MWXZE9Na@google.com>
On Tue, Nov 01, 2022 at 08:28:04PM +0000, Sean Christopherson wrote:
> On Thu, Oct 27, 2022, Oliver Upton wrote:
> > There is no real urgency to free a stage-2 subtree that was pruned.
> > Nonetheless, KVM does the tear down in the stage-2 fault path while
> > holding the MMU lock.
> >
[ copy ]
> This is _very_ misleading. The above paints RCU as an optimization of sorts to
> avoid doing work while holding mmu_lock. Freeing page tables in an RCU callback
> is _required_ for correctness when allowing parallel page faults to remove page
> tables, as holding mmu_lock for read in that case doesn't ensure no other CPU is
> accessing and/or holds a reference to the to-be-freed page table.
Agree, but it is still important to reason about what is changing here
too. Moving work out of the vCPU fault path _is_ valuable, though
ancillary to the correctness requirements.
> IMO, this patch should to be squashed with the previous patch, "Protect stage-2
> traversal with RCU". One doesn't make any sense without the other.
I had split these up back when this series was a lot more gnarly and
there was too much slop in a single diff. That isn't the case any more,
so yeah I'll squash them.
[ paste ]
> > Free removed stage-2 subtrees after an RCU grace period. To guarantee
> > all stage-2 table pages are freed before killing a VM, add an
> > rcu_barrier() to the flush path.
An aside, this is flat-out wrong now.
--
Thanks,
Oliver
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2022-11-01 20:46 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-27 22:17 [PATCH v3 00/15] KVM: arm64: Parallel stage-2 fault handling Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 01/15] KVM: arm64: Combine visitor arguments into a context structure Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 02/15] KVM: arm64: Stash observed pte value in visitor context Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 03/15] KVM: arm64: Pass mm_ops through the " Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 04/15] KVM: arm64: Don't pass kvm_pgtable through kvm_pgtable_walk_data Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 05/15] KVM: arm64: Add a helper to tear down unlinked stage-2 subtrees Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 06/15] KVM: arm64: Tear down unlinked stage-2 subtree after break-before-make Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 07/15] KVM: arm64: Use an opaque type for pteps Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` [PATCH v3 08/15] KVM: arm64: Protect stage-2 traversal with RCU Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-28 7:36 ` kernel test robot
2022-10-28 7:36 ` kernel test robot
2022-10-28 9:19 ` Oliver Upton
2022-10-28 9:19 ` Oliver Upton
2022-10-28 9:19 ` Oliver Upton
2022-11-01 20:22 ` Sean Christopherson
2022-11-01 20:22 ` Sean Christopherson
2022-11-01 20:22 ` Sean Christopherson
2022-10-27 22:17 ` [PATCH v3 09/15] KVM: arm64: Free removed stage-2 tables in RCU callback Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-10-27 22:17 ` Oliver Upton
2022-11-01 20:28 ` Sean Christopherson
2022-11-01 20:28 ` Sean Christopherson
2022-11-01 20:28 ` Sean Christopherson
2022-11-01 20:46 ` Oliver Upton [this message]
2022-11-01 20:46 ` Oliver Upton
2022-11-01 20:46 ` Oliver Upton
2022-11-01 21:07 ` Sean Christopherson
2022-11-01 21:07 ` Sean Christopherson
2022-11-01 21:07 ` Sean Christopherson
2022-10-27 22:22 ` [PATCH v3 10/15] KVM: arm64: Atomically update stage 2 leaf attributes in parallel walks Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-10-28 10:18 ` kernel test robot
2022-10-28 10:18 ` kernel test robot
2022-10-27 22:22 ` [PATCH v3 11/15] KVM: arm64: Split init and set for table PTE Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-10-27 22:22 ` [PATCH v3 12/15] KVM: arm64: Make block->table PTE changes parallel-aware Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-10-27 22:22 ` Oliver Upton
2022-11-02 2:22 ` Ricardo Koller
2022-11-02 2:22 ` Ricardo Koller
2022-11-02 2:22 ` Ricardo Koller
2022-11-02 23:03 ` Oliver Upton
2022-11-02 23:03 ` Oliver Upton
2022-11-02 23:03 ` Oliver Upton
2022-10-27 22:23 ` [PATCH v3 13/15] KVM: arm64: Make leaf->leaf " Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 22:23 ` [PATCH v3 14/15] KVM: arm64: Make table->block " Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 22:23 ` [PATCH v3 15/15] KVM: arm64: Handle stage-2 faults in parallel Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 22:23 ` Oliver Upton
2022-10-27 23:39 ` [PATCH v3 00/15] KVM: arm64: Parallel stage-2 fault handling Oliver Upton
2022-10-27 23:39 ` Oliver Upton
2022-10-27 23:39 ` Oliver Upton
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=Y2GFliAVxui9VyK2@google.com \
--to=oliver.upton@linux.dev \
--cc=bgardon@google.com \
--cc=dmatlack@google.com \
--cc=kvm@vger.kernel.org \
--cc=kvmarm@lists.cs.columbia.edu \
--cc=kvmarm@lists.linux.dev \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=maz@kernel.org \
--cc=seanjc@google.com \
--cc=will@kernel.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.