* [PATCH] serial: 8250_aspeed_vuart: Fix potential stale port on probe failure
@ 2026-11-16 14:24 Guangshuo Li
2026-11-16 14:39 ` sashiko-bot
0 siblings, 1 reply; 2+ messages in thread
From: Guangshuo Li @ 2026-11-16 14:24 UTC (permalink / raw)
To: Greg Kroah-Hartman, Jiri Slaby, Joel Stanley, Andrew Jeffery,
Ryan Chen, Billy Tsai
Cc: linux-kernel, linux-serial, linux-arm-kernel, linux-aspeed,
stable
Invalid LPC configuration can fail probe after the UART port has been
registered. Removing only the sysfs group leaves a port whose private
data and mappings are released by device-managed cleanup.
Disable the VUART, shut down its timer and unregister the port before
returning from these failures.
This issue was found by our static analysis tool under development.
Fixes: ca03042f0f12 ("serial: 8250_aspeed_vuart: add aspeed, lpc-io-reg and aspeed, lpc-interrupts DT properties")
Cc: stable@vger.kernel.org
Signed-off-by: Guangshuo Li <liguangshuo@mail.sdu.edu.cn>
---
drivers/tty/serial/8250/8250_aspeed_vuart.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/drivers/tty/serial/8250/8250_aspeed_vuart.c b/drivers/tty/serial/8250/8250_aspeed_vuart.c
index c8dc87f6886c..2cd1c9146ec8 100644
--- a/drivers/tty/serial/8250/8250_aspeed_vuart.c
+++ b/drivers/tty/serial/8250/8250_aspeed_vuart.c
@@ -513,7 +513,7 @@ static int aspeed_vuart_probe(struct platform_device *pdev)
rc = aspeed_vuart_set_lpc_address(vuart, prop);
if (rc < 0) {
dev_err_probe(dev, rc, "invalid value in aspeed,lpc-io-reg property\n");
- goto err_sysfs_remove;
+ goto err_unregister_port;
}
rc = of_property_read_u32_array(np, "aspeed,lpc-interrupts", sirq, 2);
@@ -525,14 +525,14 @@ static int aspeed_vuart_probe(struct platform_device *pdev)
rc = aspeed_vuart_set_sirq(vuart, sirq[0]);
if (rc < 0) {
dev_err_probe(dev, rc, "invalid sirq number in aspeed,lpc-interrupts property\n");
- goto err_sysfs_remove;
+ goto err_unregister_port;
}
sirq_polarity = aspeed_vuart_map_irq_polarity(sirq[1]);
if (sirq_polarity < 0) {
rc = dev_err_probe(dev, sirq_polarity,
"invalid sirq polarity in aspeed,lpc-interrupts property\n");
- goto err_sysfs_remove;
+ goto err_unregister_port;
}
aspeed_vuart_set_sirq_polarity(vuart, sirq_polarity);
@@ -543,6 +543,10 @@ static int aspeed_vuart_probe(struct platform_device *pdev)
return 0;
+err_unregister_port:
+ aspeed_vuart_set_enabled(vuart, false);
+ timer_shutdown_sync(&vuart->unthrottle_timer);
+ serial8250_unregister_port(vuart->port);
err_sysfs_remove:
sysfs_remove_group(&vuart->dev->kobj, &aspeed_vuart_attr_group);
return rc;
base-commit: 6628e7d4c6b25c8d18f1345e3a20d21829aae7e4
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] serial: 8250_aspeed_vuart: Fix potential stale port on probe failure
2026-11-16 14:24 [PATCH] serial: 8250_aspeed_vuart: Fix potential stale port on probe failure Guangshuo Li
@ 2026-11-16 14:39 ` sashiko-bot
0 siblings, 0 replies; 2+ messages in thread
From: sashiko-bot @ 2026-11-16 14:39 UTC (permalink / raw)
To: Guangshuo Li; +Cc: linux-serial
Thank you for your contribution! Sashiko AI review found no new issues
in this patch, but noticed 1 related pre-existing issue it does not
introduce:
- [Critical] serial: 8250_aspeed_vuart: use-after-free in
aspeed_vuart_remove() [1]
--
commit 4eac67f7d5031d7988614ed386938a68b8050e5b
Author: Guangshuo Li <liguangshuo@mail.sdu.edu.cn>
serial: 8250_aspeed_vuart: Fix potential stale port on probe failure
This commit disables the VUART, shuts down its unthrottle timer, and
unregisters the 8250 port when LPC configuration fails during probe.
> diff --git a/drivers/tty/serial/8250/8250_aspeed_vuart.c b/drivers/tty/serial/8250/8250_aspeed_vuart.c
> index c8dc87f6886c1..2cd1c9146ec89 100644
> --- a/drivers/tty/serial/8250/8250_aspeed_vuart.c
> +++ b/drivers/tty/serial/8250/8250_aspeed_vuart.c
[ ... ]
> @@ -543,6 +543,10 @@ static int aspeed_vuart_probe(struct platform_device *pdev)
>
> return 0;
>
> +err_unregister_port:
> + aspeed_vuart_set_enabled(vuart, false);
> + timer_shutdown_sync(&vuart->unthrottle_timer);
> + serial8250_unregister_port(vuart->port);
[Severity: Critical]
This problem wasn't introduced by this patch, but can [1]
aspeed_vuart_remove() cause a use-after-free on vuart when
unthrottle_timer is re-armed during removal?
drivers/tty/serial/8250/8250_aspeed_vuart.c:aspeed_vuart_remove() {
struct aspeed_vuart *vuart = platform_get_drvdata(pdev);
timer_delete_sync(&vuart->unthrottle_timer);
aspeed_vuart_set_enabled(vuart, false);
serial8250_unregister_port(vuart->port);
sysfs_remove_group(&vuart->dev->kobj, &aspeed_vuart_attr_group);
}
Because aspeed_vuart_remove() calls timer_delete_sync() instead of
timer_shutdown_sync() before serial8250_unregister_port(), a UART
interrupt arriving while the TTY flip buffers are full can re-arm
vuart->unthrottle_timer before the port is unregistered:
CPU0 CPU1
aspeed_vuart_remove()
timer_delete_sync(&vuart->unthrottle_timer)
aspeed_vuart_handle_irq()
if (!space)
mod_timer(&vuart->unthrottle_timer, ...)
serial8250_unregister_port(...)
// vuart freed by devres on return
aspeed_vuart_unthrottle_exp()
<-- use-after-free
Unregistering the port before stopping the timer, or using
timer_shutdown_sync() as this patch does on the probe error path, would
prevent the timer from being re-armed.
[1] https://sashiko.dev/#/bug/linux-2b3eb48a-af6f-4ebd-a642-413d2498b067
--
Sashiko AI review · https://sashiko.dev/#/patchset/20261116-c29ba55625-serial-aspeed@mail.sdu.edu.cn?part=1
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-11-16 14:39 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-16 14:24 [PATCH] serial: 8250_aspeed_vuart: Fix potential stale port on probe failure Guangshuo Li
2026-11-16 14:39 ` sashiko-bot
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox