From: Alex Williamson <alex.williamson@nvidia.com>
To: <alex@shazbot.org>
Cc: Alex Williamson <alex.williamson@nvidia.com>,
<yishaih@nvidia.com>, <kvm@vger.kernel.org>,
<guojinhui.liam@bytedance.com>, <linux-kernel@vger.kernel.org>,
<virtualization@lists.linux.dev>
Subject: [PATCH 0/4] vfio/virtio: Fix list_lock type and modernize locking
Date: Tue, 14 Apr 2026 14:06:18 -0600 [thread overview]
Message-ID: <20260414200625.3601509-1-alex.williamson@nvidia.com> (raw)
Jinhui Guo reported a mismatched spin_lock()/spin_unlock_irq() pair
in virtiovf_read_device_context_chunk() where spin_unlock_irq() would
unconditionally enable interrupts despite spin_lock() never having
disabled them. On closer inspection, the list_lock spinlock with IRQ
disabling was copied from the mlx5 variant driver where a hardirq
completion callback justifies it, but the virtio driver has no
interrupt context usage of list_lock. Patch 1 converts list_lock to
a mutex, fixing the mismatch and aligning with peer vfio-pci variant
drivers.
Patch 2 converts the list_lock acquisitions to guard()/scoped_guard()
where the lock scope aligns naturally with function or block scope.
Patches 3 and 4 extend the same guard() conversion to the remaining
two mutexes in the driver (migf->lock and bar_mutex). These are
relatively independent of the list_lock fix but complete the
conversion across the driver. Thanks,
Alex
Alex Williamson (4):
vfio/virtio: Convert list_lock from spinlock to mutex
vfio/virtio: Use guard() for list_lock where applicable
vfio/virtio: Use guard() for migf->lock where applicable
vfio/virtio: Use guard() for bar_mutex in legacy I/O
drivers/vfio/pci/virtio/common.h | 2 +-
drivers/vfio/pci/virtio/legacy_io.c | 17 +++---
drivers/vfio/pci/virtio/migrate.c | 90 ++++++++++++-----------------
3 files changed, 46 insertions(+), 63 deletions(-)
--
2.51.0
next reply other threads:[~2026-04-14 20:07 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-04-14 20:06 Alex Williamson [this message]
2026-04-14 20:06 ` [PATCH 1/4] vfio/virtio: Convert list_lock from spinlock to mutex Alex Williamson
2026-04-14 20:06 ` [PATCH 2/4] vfio/virtio: Use guard() for list_lock where applicable Alex Williamson
2026-04-14 20:06 ` [PATCH 3/4] vfio/virtio: Use guard() for migf->lock " Alex Williamson
2026-04-14 20:06 ` [PATCH 4/4] vfio/virtio: Use guard() for bar_mutex in legacy I/O Alex Williamson
2026-04-15 15:12 ` [PATCH 0/4] vfio/virtio: Fix list_lock type and modernize locking Yishai Hadas
2026-04-15 15:39 ` Alex Williamson
2026-04-15 17:23 ` Yishai Hadas
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=20260414200625.3601509-1-alex.williamson@nvidia.com \
--to=alex.williamson@nvidia.com \
--cc=alex@shazbot.org \
--cc=guojinhui.liam@bytedance.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=virtualization@lists.linux.dev \
--cc=yishaih@nvidia.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