Re: sysfs: wait for GPIO interrupt

[Javier Martinez Canillas <javier@xxxxxxxxxxxx>]

Hello Yegor,

On Thu, May 8, 2014 at 10:19 AM, Yegor Yefremov
<yegorslists@xxxxxxxxxxxxxx> wrote:
> I've stumbled upon following issue. When waiting for interrupt with
> poll() like described in this example [1], the program returns from
> poll() immediately. This thread [2] describes a workaround for this
> issue via making read() before invoking poll().
>
> My system is am335x, kernel 3.15-rc4
>
> I experience this behavior by both SoC's own GPIO and i2c expander tca6416.
>

I tested with an OMAP3 board which uses the same GPIO driver than the
am335x and I was able to reproduce your issue.

> Is this the proper way (invoke read() before poll()) to work with
> GPIOs from user space, so that this approach should be incorporated
> into such libraries as libsoc [3]
>

I don't think that this should be an expected behavior, it seems that
this is either a bug on gpiolib sysfs interface or gpio-omap.c and
gpio-pca953x.c drivers (most probably the later than the former).

The question is what side effect has to do a read(2) syscall before a
poll(2) to make it succeed. I looked at
drivers/gpio/gpiolib.c:gpio_value_show() but didn't find anything
relevant.

I didn't have time to dig futher on this but I'll try to take a look
this weekend when I've more time. But for now I've added to Linus and
Alexandre in cc in case they have an idea of what's going on.

> Yegor
>
> [1] https://developer.ridgerun.com/wiki/index.php/Gpio-int-test.c
> [2] http://comments.gmane.org/gmane.linux.distributions.gumstix.general/61286
> [3] https://github.com/jackmitch/libsoc

Best regards,
Javier
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html