From: "David Matlack" <dmatlack@google.com>
To: <sashiko-reviews@lists.linux.dev>,
"David Matlack" <dmatlack@google.com>,
<kexec@lists.infradead.org>, <linux-mm@kvack.org>,
<linux-pci@vger.kernel.org>
Cc: "Adithya Jayachandran" <ajayachandra@nvidia.com>,
"Alexander Graf" <graf@amazon.com>,
"Alex Williamson" <alex@shazbot.org>,
"Alex Williamson" <alex.williamson@nvidia.com>,
"Bjorn Helgaas" <bhelgaas@google.com>,
"Chris Down" <chris@chrisdown.name>,
"Chris Li" <chrisl@kernel.org>,
"David Rientjes" <rientjes@google.com>,
"Jacob Pan" <jacob.pan@linux.microsoft.com>,
"Jason Gunthorpe" <jgg@nvidia.com>,
"Jonathan Corbet" <corbet@lwn.net>,
"Josh Hilke" <jrhilke@google.com>,
"Leon Romanovsky" <leonro@nvidia.com>,
"Lukas Wunner" <lukas@wunner.de>,
"Mike Rapoport" <rppt@kernel.org>,
"Parav Pandit" <parav@nvidia.com>,
"Pasha Tatashin" <pasha.tatashin@soleen.com>,
"Pranjal Shrivastava" <praan@google.com>,
"Pratyush Yadav" <pratyush@kernel.org>,
"Randy Dunlap" <rdunlap@infradead.org>,
"Saeed Mahameed" <saeedm@nvidia.com>,
"Samiullah Khawaja" <skhawaja@google.com>,
"Shuah Khan" <skhan@linuxfoundation.org>,
"Vipin Sharma" <vipinsh@google.com>,
"William Tu" <witu@nvidia.com>, "Yi Liu" <yi.l.liu@intel.com>
Subject: Re: [PATCH v11 11/13] PCI: liveupdate: Freeze preservation status during shutdown
Date: Fri, 20 Nov 2026 11:11:04 -0800 [thread overview]
Message-ID: <DN09Y7673R13.3GB4PTE9ZPXAE@google.com> (raw)
In-Reply-To: <sashiko-outbox-224146@kernel.org>
On Thu Nov 19, 2026 at 2:38 PM PST, sashiko-bot wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential
> issue to consider:
> - [High] pci: liveupdate: panic on session teardown after
> pci_liveupdate_freeze()
> --
>
> commit a3120ec70b92ea26f3e1d92d3bd0d1531b5a73b6
> Author: David Matlack <dmatlack@google.com>
>
> PCI: liveupdate: Freeze preservation status during shutdown
>
> This commit freezes a PCI device's outgoing live update preservation
> status in pci_device_shutdown() so that devices cannot be preserved or
> unpreserved once shutdown begins.
>
>> diff --git a/drivers/pci/liveupdate.c b/drivers/pci/liveupdate.c
>> index a45ad86370d5d..e26d2f5c2338c 100644
>> --- a/drivers/pci/liveupdate.c
>> +++ b/drivers/pci/liveupdate.c
>
> [ ... ]
>
>> @@ -520,6 +523,11 @@ static int pci_liveupdate_unpreserve_device(struct pci_flb_outgoing *outgoing,
>> {
>> struct pci_dev_ser *dev_ser = dev->liveupdate.outgoing;
>>
>> + if (dev->liveupdate.frozen) {
>> + pci_warn(dev, "Cannot unpreserve device after it is frozen!\n");
>> + return -EINVAL;
>> + }
>
> [Severity: High]
> Can rejecting pci_liveupdate_unpreserve_device() on a frozen device trigger a
> panic in pci_check_all_devices_released() when a Live Update session is
> closed during a non-kexec shutdown?
>
> Unlike kernel_kexec(), which calls liveupdate_reboot() to serialize sessions
> before device_shutdown(), non-kexec shutdown paths (kernel_restart(),
> kernel_halt(), and kernel_power_off()) call device_shutdown() ->
> pci_device_shutdown() -> pci_liveupdate_freeze() without serializing Live
> Update sessions.
>
> If a task closes an open session FD with a preserved PCI device during or
> after pci_device_shutdown(), luo_file_unpreserve_files() invokes the file
> handler's .unpreserve() callback and then unconditionally drops the outgoing
> FLB reference:
>
> kernel/liveupdate/luo_file.c:luo_file_unpreserve_files() {
> ...
> luo_file->fh->ops->unpreserve(&args);
> luo_flb_file_unpreserve(luo_file->fh);
> ...
> }
>
> Because dev->liveupdate.frozen is set, pci_liveupdate_unpreserve_device()
> returns -EINVAL without decrementing dev_ser->refcount or
> outgoing->ser->nr_devices. When luo_flb_file_unpreserve() drops the final
> FLB reference, pci_flb_unpreserve() calls pci_check_all_devices_released(),
> which panics because outgoing->ser->nr_devices is still non-zero.
Yes, this can happen, although only in a narrow window.
kernel_restart(), kernel_halt(), and kernel_power_off() do not freeze
userspace or serialize Live Update sessions, so a task that still holds
a session FD when reboot(2) is called (e.g. reboot -f, which skips
killing userspace first) can close it after its devices have been
frozen. In v10 this only resulted in a warning, but the panic added to
pci_flb_unpreserve() in v11 turns it into a panic during reboot.
The freeze only matters during kexec. pci_device_shutdown() only checks
pci_liveupdate_is_outgoing() when kexec_in_progress is set, and
kernel_kexec() serializes all sessions via liveupdate_reboot() before
device_shutdown(), so nothing can be unpreserved after devices are
frozen.
I will fix this in v12 by only freezing devices during kexec, so that a
non-kexec shutdown no longer refuses to unpreserve devices:
diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
index 08b768b9e905..2748ca22e9cf 100644
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -560,7 +560,8 @@ static void pci_device_shutdown(struct device *dev)
struct pci_dev *pci_dev = to_pci_dev(dev);
struct pci_driver *drv = pci_dev->driver;
- pci_liveupdate_freeze(pci_dev);
+ if (kexec_in_progress)
+ pci_liveupdate_freeze(pci_dev);
pm_runtime_resume(dev);
if (drv && drv->shutdown)
next prev parent reply other threads:[~2026-11-20 19:11 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-18 22:36 [PATCH v11 00/13] PCI: liveupdate: PCI core support for Live Update David Matlack
2026-11-18 22:36 ` [PATCH v11 01/13] PCI: liveupdate: Set up FLB handler for the PCI core David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 02/13] PCI: liveupdate: Track outgoing preserved PCI devices David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 03/13] PCI: liveupdate: Track incoming " David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 04/13] PCI: liveupdate: Document driver binding responsibilities David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 05/13] PCI: liveupdate: Auto-preserve upstream bridges across Live Update David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 06/13] PCI: liveupdate: Preserve bus numbers during " David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 07/13] PCI: Refactor matching logic for pci_dev_acs_ops David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:36 ` [PATCH v11 08/13] PCI: Save and restore the ACS Control register and Egress Control Vector David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-20 14:44 ` Pranjal Shrivastava
2026-11-18 22:37 ` [PATCH v11 09/13] PCI: liveupdate: Adopt ACS controls in incoming preserved devices David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-20 0:10 ` Alex Williamson
2026-11-20 15:17 ` Pranjal Shrivastava
2026-11-20 19:29 ` David Matlack
2026-11-18 22:37 ` [PATCH v11 10/13] PCI: liveupdate: Adopt ARI Forwarding Enable on preserved bridges David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-20 19:25 ` David Matlack
2026-11-18 22:37 ` [PATCH v11 11/13] PCI: liveupdate: Freeze preservation status during shutdown David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-20 19:11 ` David Matlack [this message]
2026-11-18 22:37 ` [PATCH v11 12/13] PCI: liveupdate: Do not disable bus mastering on preserved devices during kexec David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-18 22:37 ` [PATCH v11 13/13] Documentation: PCI: Add documentation for Live Update David Matlack
2026-11-19 22:38 ` sashiko-bot
2026-11-19 11:17 ` [PATCH v11 00/13] PCI: liveupdate: PCI core support " Mike Rapoport
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=DN09Y7673R13.3GB4PTE9ZPXAE@google.com \
--to=dmatlack@google.com \
--cc=ajayachandra@nvidia.com \
--cc=alex.williamson@nvidia.com \
--cc=alex@shazbot.org \
--cc=bhelgaas@google.com \
--cc=chris@chrisdown.name \
--cc=chrisl@kernel.org \
--cc=corbet@lwn.net \
--cc=graf@amazon.com \
--cc=jacob.pan@linux.microsoft.com \
--cc=jgg@nvidia.com \
--cc=jrhilke@google.com \
--cc=kexec@lists.infradead.org \
--cc=leonro@nvidia.com \
--cc=linux-mm@kvack.org \
--cc=linux-pci@vger.kernel.org \
--cc=lukas@wunner.de \
--cc=parav@nvidia.com \
--cc=pasha.tatashin@soleen.com \
--cc=praan@google.com \
--cc=pratyush@kernel.org \
--cc=rdunlap@infradead.org \
--cc=rientjes@google.com \
--cc=rppt@kernel.org \
--cc=saeedm@nvidia.com \
--cc=sashiko-reviews@lists.linux.dev \
--cc=skhan@linuxfoundation.org \
--cc=skhawaja@google.com \
--cc=vipinsh@google.com \
--cc=witu@nvidia.com \
--cc=yi.l.liu@intel.com \
/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