This is a note to let you know that I've just added the patch titled
vfio/pci: Disable auto-enable of exclusive INTx IRQ
to the 5.4-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
vfio-pci-disable-auto-enable-of-exclusive-intx-irq.patch
and it can be found in the queue-5.4 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From stable+bounces-35122-greg=kroah.com@xxxxxxxxxxxxxxx Mon Apr 1 18:53:59 2024
From: Alex Williamson <alex.williamson@xxxxxxxxxx>
Date: Mon, 1 Apr 2024 10:52:55 -0600
Subject: vfio/pci: Disable auto-enable of exclusive INTx IRQ
To: stable@xxxxxxxxxxxxxxx
Cc: Alex Williamson <alex.williamson@xxxxxxxxxx>, sashal@xxxxxxxxxx, gregkh@xxxxxxxxxxxxxxxxxxx, eric.auger@xxxxxxxxxx, Kevin Tian <kevin.tian@xxxxxxxxx>
Message-ID: <20240401165302.3699643-2-alex.williamson@xxxxxxxxxx>
From: Alex Williamson <alex.williamson@xxxxxxxxxx>
[ Upstream commit fe9a7082684eb059b925c535682e68c34d487d43 ]
Currently for devices requiring masking at the irqchip for INTx, ie.
devices without DisINTx support, the IRQ is enabled in request_irq()
and subsequently disabled as necessary to align with the masked status
flag. This presents a window where the interrupt could fire between
these events, resulting in the IRQ incrementing the disable depth twice.
This would be unrecoverable for a user since the masked flag prevents
nested enables through vfio.
Instead, invert the logic using IRQF_NO_AUTOEN such that exclusive INTx
is never auto-enabled, then unmask as required.
Cc: <stable@xxxxxxxxxxxxxxx>
Fixes: 89e1f7d4c66d ("vfio: Add PCI device driver")
Reviewed-by: Kevin Tian <kevin.tian@xxxxxxxxx>
Reviewed-by: Eric Auger <eric.auger@xxxxxxxxxx>
Link: https://lore.kernel.org/r/20240308230557.805580-2-alex.williamson@xxxxxxxxxx
Signed-off-by: Alex Williamson <alex.williamson@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/vfio/pci/vfio_pci_intrs.c | 17 ++++++++++-------
1 file changed, 10 insertions(+), 7 deletions(-)
--- a/drivers/vfio/pci/vfio_pci_intrs.c
+++ b/drivers/vfio/pci/vfio_pci_intrs.c
@@ -199,8 +199,15 @@ static int vfio_intx_set_signal(struct v
vdev->ctx[0].trigger = trigger;
+ /*
+ * Devices without DisINTx support require an exclusive interrupt,
+ * IRQ masking is performed at the IRQ chip. The masked status is
+ * protected by vdev->irqlock. Setup the IRQ without auto-enable and
+ * unmask as necessary below under lock. DisINTx is unmodified by
+ * the IRQ configuration and may therefore use auto-enable.
+ */
if (!vdev->pci_2_3)
- irqflags = 0;
+ irqflags = IRQF_NO_AUTOEN;
ret = request_irq(pdev->irq, vfio_intx_handler,
irqflags, vdev->ctx[0].name, vdev);
@@ -211,13 +218,9 @@ static int vfio_intx_set_signal(struct v
return ret;
}
- /*
- * INTx disable will stick across the new irq setup,
- * disable_irq won't.
- */
spin_lock_irqsave(&vdev->irqlock, flags);
- if (!vdev->pci_2_3 && vdev->ctx[0].masked)
- disable_irq_nosync(pdev->irq);
+ if (!vdev->pci_2_3 && !vdev->ctx[0].masked)
+ enable_irq(pdev->irq);
spin_unlock_irqrestore(&vdev->irqlock, flags);
return 0;
Patches currently in stable-queue which might be from kroah.com@xxxxxxxxxxxxxxx are
queue-5.4/vfio-pci-create-persistent-intx-handler.patch
queue-5.4/loop-factor-out-configuring-loop-from-status.patch
queue-5.4/loop-call-loop_config_discard-only-after-new-config-is-applied.patch
queue-5.4/loop-refactor-loop_set_status-size-calculation.patch
queue-5.4/loop-factor-out-setting-loop-device-size.patch
queue-5.4/vfio-pci-disable-auto-enable-of-exclusive-intx-irq.patch
queue-5.4/loop-check-for-overflow-while-configuring-loop.patch
queue-5.4/vfio-pci-lock-external-intx-masking-ops.patch
queue-5.4/loop-remove-sector_t-truncation-checks.patch
queue-5.4/vfio-introduce-interface-to-flush-virqfd-inject-workqueue.patch
queue-5.4/revert-loop-check-for-overflow-while-configuring-loop.patch
queue-5.4/loop-loop_set_status_from_info-check-before-assignment.patch
queue-5.4/mm-migrate-set-swap-entry-values-of-thp-tail-pages-properly.patch
queue-5.4/vfio-platform-create-persistent-irq-handlers.patch
