gpiolib: fix line event timestamps for nested irqs

Message ID 20190104103226.2878-1-brgl@bgdev.pl
State New
Headers show
Series
  • gpiolib: fix line event timestamps for nested irqs
Related show

Commit Message

Bartosz Golaszewski Jan. 4, 2019, 10:32 a.m.
From: Bartosz Golaszewski <bgolaszewski@baylibre.com>

Nested interrupts run inside the calling thread's context and the top
half handler is never called which means that we never read the
timestamp.

This issue came up when trying to read line events from a gpiochip
using regmap_irq_chip for interrupts.

Fix it by reading the timestamp from the irq thread function if it's
still 0 by the time the second handler is called.

Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
---
 drivers/gpio/gpiolib.c | 7 +++++++
 1 file changed, 7 insertions(+)

Comments

Linus Walleij Jan. 11, 2019, 12:19 p.m. | #1
On Fri, Jan 4, 2019 at 11:32 AM Bartosz Golaszewski <brgl@bgdev.pl> wrote:

> From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
>
> Nested interrupts run inside the calling thread's context and the top
> half handler is never called which means that we never read the
> timestamp.
>
> This issue came up when trying to read line events from a gpiochip
> using regmap_irq_chip for interrupts.
>
> Fix it by reading the timestamp from the irq thread function if it's
> still 0 by the time the second handler is called.
>
> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>

Oh that's right. Should this not have Fixes: and Cc: stable?

> +       /*
> +        * We may be running from a nested threaded interrupt in which case
> +        * we didn't get the timestamp from lineevent_irq_handler().
> +        */
> +       if (!le->timestamp)
> +               le->timestamp = ktime_get_real_ns();
> +
>         ge.timestamp = le->timestamp;

I would merge these:

if (!le->timestamp)
    ge.timestamp = ktime_get_real_ns();
else
    ge.timestamp = le->timestamp;

Yours,
Linus Walleij
Bartosz Golaszewski Jan. 11, 2019, 1:45 p.m. | #2
pt., 11 sty 2019 o 13:19 Linus Walleij <linus.walleij@linaro.org> napisał(a):
>
> On Fri, Jan 4, 2019 at 11:32 AM Bartosz Golaszewski <brgl@bgdev.pl> wrote:
>
> > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> >
> > Nested interrupts run inside the calling thread's context and the top
> > half handler is never called which means that we never read the
> > timestamp.
> >
> > This issue came up when trying to read line events from a gpiochip
> > using regmap_irq_chip for interrupts.
> >
> > Fix it by reading the timestamp from the irq thread function if it's
> > still 0 by the time the second handler is called.
> >
> > Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
>
> Oh that's right. Should this not have Fixes: and Cc: stable?
>
> > +       /*
> > +        * We may be running from a nested threaded interrupt in which case
> > +        * we didn't get the timestamp from lineevent_irq_handler().
> > +        */
> > +       if (!le->timestamp)
> > +               le->timestamp = ktime_get_real_ns();
> > +
> >         ge.timestamp = le->timestamp;
>
> I would merge these:
>
> if (!le->timestamp)
>     ge.timestamp = ktime_get_real_ns();
> else
>     ge.timestamp = le->timestamp;
>
> Yours,
> Linus Walleij

I'll fix both issues and resend.

Bart

Patch

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 1651d7f0a303..330c52e19b7e 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -828,6 +828,13 @@  static irqreturn_t lineevent_irq_thread(int irq, void *p)
 	/* Do not leak kernel stack to userspace */
 	memset(&ge, 0, sizeof(ge));
 
+	/*
+	 * We may be running from a nested threaded interrupt in which case
+	 * we didn't get the timestamp from lineevent_irq_handler().
+	 */
+	if (!le->timestamp)
+		le->timestamp = ktime_get_real_ns();
+
 	ge.timestamp = le->timestamp;
 
 	if (le->eflags & GPIOEVENT_REQUEST_RISING_EDGE