[PATCH v2 3/5] gpiolib: gpio_request(): add error message
Jean-Christophe PLAGNIOL-VILLARD
plagnioj at jcrosoft.com
Fri Feb 1 13:24:21 EST 2013
On 22:09 Fri 01 Feb , Antony Pavlov wrote:
> On 1 February 2013 12:13, Jean-Christophe PLAGNIOL-VILLARD
> <plagnioj at jcrosoft.com> wrote:
> > Nack the is need to handle by the caller not here
>
> You want to get duplicated code? Ok, you will get in the "patch v3" series! :)))
this is not duplicated code the caller known the dev or the part of the code
that fault
and we do auto request which is silent or other par of the code report the
erroor at higher level to yet I do not want the generic API to print something
Best Regards,
J.
>
> >
> > Best Regards,
> > J.
> >>
> >> Signed-off-by: Antony Pavlov <antonynpavlov at gmail.com>
> >> ---
> >> drivers/gpio/gpio.c | 29 ++++++++++++++++++++++-------
> >> 1 file changed, 22 insertions(+), 7 deletions(-)
> >>
> >> diff --git a/drivers/gpio/gpio.c b/drivers/gpio/gpio.c
> >> index 5876454..179fc7b 100644
> >> --- a/drivers/gpio/gpio.c
> >> +++ b/drivers/gpio/gpio.c
> >> @@ -29,22 +29,37 @@ int gpio_request(unsigned gpio, const char *label)
> >> struct gpio_chip *chip = gi->chip;
> >> int ret;
> >>
> >> - if (!gpio_is_valid(gpio))
> >> - return -EINVAL;
> >> - if (!chip)
> >> - return -EINVAL;
> >> - if (gi->requested)
> >> - return -EBUSY;
> >> + if (!gpio_is_valid(gpio)) {
> >> + ret = -EINVAL;
> >> + goto err;
> >> + }
> >> +
> >> + if (!chip) {
> >> + ret = -EINVAL;
> >> + goto err;
> >> + }
> >> +
> >> + if (gi->requested) {
> >> + ret = -EBUSY;
> >> + goto err;
> >> + }
> >> +
> >> if (chip->ops->request) {
> >> ret = chip->ops->request(chip, gpio - chip->base);
> >> if (ret)
> >> - return ret;
> >> + goto err;
> >> }
> >>
> >> gi->requested = true;
> >> gi->label = xstrdup(label);
> >>
> >> return 0;
> >> +
> >> +err:
> >> + pr_err("can't request gpio %d as %s (%d)\n",
> >> + gpio, label, ret);
> >> +
> >> + return ret;
> >> }
> >>
> >> int gpio_request_input(unsigned gpio, const char *label)
> >> --
> >> 1.7.10.4
> >>
>
>
>
> --
> Best regards,
> Antony Pavlov
More information about the barebox
mailing list