* [PATCH v2] tty: serial: fsl_lpuart: fix del_timer_sync() vs timer routine deadlock
@ 2016-12-04 15:49 Nikita Yushchenko
2016-12-07 1:44 ` Stefan Agner
0 siblings, 1 reply; 2+ messages in thread
From: Nikita Yushchenko @ 2016-12-04 15:49 UTC (permalink / raw)
To: Greg Kroah-Hartman, Jiri Slaby, Bhuvanchandra DV, Stefan Agner,
Wei Yongjun, Aaron Brice, Nicolae Rosia, linux-serial
Cc: Chris Healy, linux-kernel, Nikita Yushchenko
Problem found via lockdep:
- lpuart_set_termios() calls del_timer_sync(&sport->lpuart_timer) while
holding sport->port.lock
- sport->lpuart_timer routine is lpuart_timer_func() that calls
lpuart_copy_rx_to_tty() that acquires same lock.
To fix, move Rx DMA stopping out of lock, as it already is in other places
in the same file.
While at it, also make Rx DMA start/stop code to look the same is in
other places in the same file.
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
---
Changes from v1:
- keep reallocation of Rx DMA buffer conditional on 'old'
- do not set sport->lpuart_dma_rx_use to true when it is already true
drivers/tty/serial/fsl_lpuart.c | 35 +++++++++++++++++------------------
1 file changed, 17 insertions(+), 18 deletions(-)
diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c
index a1c6519837a4..f02934ffb329 100644
--- a/drivers/tty/serial/fsl_lpuart.c
+++ b/drivers/tty/serial/fsl_lpuart.c
@@ -1407,6 +1407,18 @@ lpuart_set_termios(struct uart_port *port, struct ktermios *termios,
/* ask the core to calculate the divisor */
baud = uart_get_baud_rate(port, termios, old, 50, port->uartclk / 16);
+ /*
+ * Need to update the Ring buffer length according to the selected
+ * baud rate and restart Rx DMA path.
+ *
+ * Since timer function acqures sport->port.lock, need to stop before
+ * acquring same lock because otherwise del_timer_sync() can deadlock.
+ */
+ if (old && sport->lpuart_dma_rx_use) {
+ del_timer_sync(&sport->lpuart_timer);
+ lpuart_dma_rx_free(&sport->port);
+ }
+
spin_lock_irqsave(&sport->port.lock, flags);
sport->port.read_status_mask = 0;
@@ -1456,22 +1468,11 @@ lpuart_set_termios(struct uart_port *port, struct ktermios *termios,
/* restore control register */
writeb(old_cr2, sport->port.membase + UARTCR2);
- /*
- * If new baud rate is set, we will also need to update the Ring buffer
- * length according to the selected baud rate and restart Rx DMA path.
- */
- if (old) {
- if (sport->lpuart_dma_rx_use) {
- del_timer_sync(&sport->lpuart_timer);
- lpuart_dma_rx_free(&sport->port);
- }
-
- if (sport->dma_rx_chan && !lpuart_start_rx_dma(sport)) {
- sport->lpuart_dma_rx_use = true;
+ if (old && sport->lpuart_dma_rx_use) {
+ if (!lpuart_start_rx_dma(sport))
rx_dma_timer_init(sport);
- } else {
+ else
sport->lpuart_dma_rx_use = false;
- }
}
spin_unlock_irqrestore(&sport->port.lock, flags);
@@ -2131,12 +2132,10 @@ static int lpuart_resume(struct device *dev)
if (sport->lpuart_dma_rx_use) {
if (sport->port.irq_wake) {
- if (!lpuart_start_rx_dma(sport)) {
- sport->lpuart_dma_rx_use = true;
+ if (!lpuart_start_rx_dma(sport))
rx_dma_timer_init(sport);
- } else {
+ else
sport->lpuart_dma_rx_use = false;
- }
}
}
--
2.1.4
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v2] tty: serial: fsl_lpuart: fix del_timer_sync() vs timer routine deadlock
2016-12-04 15:49 [PATCH v2] tty: serial: fsl_lpuart: fix del_timer_sync() vs timer routine deadlock Nikita Yushchenko
@ 2016-12-07 1:44 ` Stefan Agner
0 siblings, 0 replies; 2+ messages in thread
From: Stefan Agner @ 2016-12-07 1:44 UTC (permalink / raw)
To: Nikita Yushchenko
Cc: Greg Kroah-Hartman, Jiri Slaby, Bhuvanchandra DV, Wei Yongjun,
Aaron Brice, Nicolae Rosia, linux-serial, Chris Healy,
linux-kernel
On 2016-12-04 07:49, Nikita Yushchenko wrote:
> Problem found via lockdep:
>
> - lpuart_set_termios() calls del_timer_sync(&sport->lpuart_timer) while
> holding sport->port.lock
>
> - sport->lpuart_timer routine is lpuart_timer_func() that calls
> lpuart_copy_rx_to_tty() that acquires same lock.
>
> To fix, move Rx DMA stopping out of lock, as it already is in other places
> in the same file.
>
> While at it, also make Rx DMA start/stop code to look the same is in
> other places in the same file.
Looks good here:
Tested-by: Stefan Agner <stefan@agner.ch>
>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> ---
> Changes from v1:
> - keep reallocation of Rx DMA buffer conditional on 'old'
> - do not set sport->lpuart_dma_rx_use to true when it is already true
>
> drivers/tty/serial/fsl_lpuart.c | 35 +++++++++++++++++------------------
> 1 file changed, 17 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c
> index a1c6519837a4..f02934ffb329 100644
> --- a/drivers/tty/serial/fsl_lpuart.c
> +++ b/drivers/tty/serial/fsl_lpuart.c
> @@ -1407,6 +1407,18 @@ lpuart_set_termios(struct uart_port *port,
> struct ktermios *termios,
> /* ask the core to calculate the divisor */
> baud = uart_get_baud_rate(port, termios, old, 50, port->uartclk / 16);
>
> + /*
> + * Need to update the Ring buffer length according to the selected
> + * baud rate and restart Rx DMA path.
> + *
> + * Since timer function acqures sport->port.lock, need to stop before
> + * acquring same lock because otherwise del_timer_sync() can deadlock.
> + */
> + if (old && sport->lpuart_dma_rx_use) {
> + del_timer_sync(&sport->lpuart_timer);
> + lpuart_dma_rx_free(&sport->port);
> + }
> +
> spin_lock_irqsave(&sport->port.lock, flags);
>
> sport->port.read_status_mask = 0;
> @@ -1456,22 +1468,11 @@ lpuart_set_termios(struct uart_port *port,
> struct ktermios *termios,
> /* restore control register */
> writeb(old_cr2, sport->port.membase + UARTCR2);
>
> - /*
> - * If new baud rate is set, we will also need to update the Ring buffer
> - * length according to the selected baud rate and restart Rx DMA path.
> - */
> - if (old) {
> - if (sport->lpuart_dma_rx_use) {
> - del_timer_sync(&sport->lpuart_timer);
> - lpuart_dma_rx_free(&sport->port);
> - }
> -
> - if (sport->dma_rx_chan && !lpuart_start_rx_dma(sport)) {
> - sport->lpuart_dma_rx_use = true;
> + if (old && sport->lpuart_dma_rx_use) {
> + if (!lpuart_start_rx_dma(sport))
> rx_dma_timer_init(sport);
> - } else {
> + else
> sport->lpuart_dma_rx_use = false;
> - }
> }
>
> spin_unlock_irqrestore(&sport->port.lock, flags);
> @@ -2131,12 +2132,10 @@ static int lpuart_resume(struct device *dev)
>
> if (sport->lpuart_dma_rx_use) {
> if (sport->port.irq_wake) {
> - if (!lpuart_start_rx_dma(sport)) {
> - sport->lpuart_dma_rx_use = true;
> + if (!lpuart_start_rx_dma(sport))
> rx_dma_timer_init(sport);
> - } else {
> + else
> sport->lpuart_dma_rx_use = false;
> - }
> }
> }
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2016-12-07 1:51 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-12-04 15:49 [PATCH v2] tty: serial: fsl_lpuart: fix del_timer_sync() vs timer routine deadlock Nikita Yushchenko
2016-12-07 1:44 ` Stefan Agner
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).