lkml.org 
[lkml]   [2016]   [Dec]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRe: [PATCH] reset: Make optional functions really optional.
From
Date
Hi Ramiro,

Am Donnerstag, den 15.12.2016, 18:05 +0000 schrieb Ramiro Oliveira:
> Up until now optional functions in the reset API were similar to the non
> optional.
>
> This patch corrects that, while maintaining compatibility with existing
> drivers.
>
> As suggested here: https://lkml.org/lkml/2016/12/14/502
>
> Signed-off-by: Ramiro Oliveira <Ramiro.Oliveira@synopsys.com>
> ---
> drivers/reset/core.c | 21 +++++++++++++++++++--
> include/linux/reset.h | 46 ++++++++++++++++++++++++++++++++++++++++++----
> 2 files changed, 61 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/reset/core.c b/drivers/reset/core.c
> index 395dc9c..6150e7c 100644
> --- a/drivers/reset/core.c
> +++ b/drivers/reset/core.c
> @@ -135,9 +135,14 @@ EXPORT_SYMBOL_GPL(devm_reset_controller_register);
> * @rstc: reset controller
> *
> * Calling this on a shared reset controller is an error.
> + *
> + * If it's an optional reset it will return 0.

I'd prefer this to explicitly mention that rstc==NULL means this is an
optional reset:

"If rstc is NULL it is an optional reset and the function will just
return 0."

> */
> int reset_control_reset(struct reset_control *rstc)
> {
> + if (!rstc)
> + return 0;
> +
> if (WARN_ON(rstc->shared))
> return -EINVAL;
>
> @@ -158,9 +163,14 @@ EXPORT_SYMBOL_GPL(reset_control_reset);
> *
> * For shared reset controls a driver cannot expect the hw's registers and
> * internal state to be reset, but must be prepared for this to happen.
> + *
> + * If it's an optional reset it will return 0.

Same as above.

> */
> int reset_control_assert(struct reset_control *rstc)
> {
> + if (!rstc)
> + return 0;
> +
> if (!rstc->rcdev->ops->assert)
> return -ENOTSUPP;
>
> @@ -180,10 +190,14 @@ EXPORT_SYMBOL_GPL(reset_control_assert);
> * reset_control_deassert - deasserts the reset line
> * @rstc: reset controller
> *
> - * After calling this function, the reset is guaranteed to be deasserted.
> + * After calling this function, the reset is guaranteed to be deasserted, if
> + * it's not optional.

Same as above.

> */
> int reset_control_deassert(struct reset_control *rstc)
> {
> + if (!rstc)
> + return 0;
> +
> if (!rstc->rcdev->ops->deassert)
> return -ENOTSUPP;
>
> @@ -199,11 +213,14 @@ EXPORT_SYMBOL_GPL(reset_control_deassert);
> /**
> * reset_control_status - returns a negative errno if not supported, a
> * positive value if the reset line is asserted, or zero if the reset
> - * line is not asserted.
> + * line is not asserted or if the desc is NULL (optional reset).
> * @rstc: reset controller
> */
> int reset_control_status(struct reset_control *rstc)
> {
> + if (!rstc)
> + return 0;
> +
> if (rstc->rcdev->ops->status)
> return rstc->rcdev->ops->status(rstc->rcdev, rstc->id);
>
> diff --git a/include/linux/reset.h b/include/linux/reset.h
> index 5daff15..1af1e62 100644
> --- a/include/linux/reset.h
> +++ b/include/linux/reset.h
> @@ -138,13 +138,33 @@ static inline struct reset_control *reset_control_get_shared(
> static inline struct reset_control *reset_control_get_optional_exclusive(
> struct device *dev, const char *id)
> {
> - return __of_reset_control_get(dev ? dev->of_node : NULL, id, 0, 0);
> + struct reset_control *desc;
> +
> + desc = __of_reset_control_get(dev ? dev->of_node : NULL, id, 0, 0);

Note that the __of_reset_control_get stub returns -ENOTSUPP if
CONFIG_RESET_CONTROLLER is disabled.

> + if (IS_ERR(desc)) {
> + if (PTR_ERR(desc) == -ENOENT)
> + return NULL;
> + }
> +
> + return desc;
> +
> }
>
> static inline struct reset_control *reset_control_get_optional_shared(
> struct device *dev, const char *id)
> {
> - return __of_reset_control_get(dev ? dev->of_node : NULL, id, 0, 1);
> +
> + struct reset_control *desc;
> +
> + desc = __of_reset_control_get(dev ? dev->of_node : NULL, id, 0, 1);
> +
> + if (IS_ERR(desc)) {
> + if (PTR_ERR(desc) == -ENOENT)
> + return NULL;
> + }

With this duplication, I think it might be better to add an int optional
parameter to __of_reset_control_get and let that return NULL if optional
is set and either of_property_match_string or of_parse_phandle_with_args
would cause an -ENOENT return.

The stub could then
return optional ? NULL : ERR_PTR(-EONOENT);

> + return desc;
> }
>
> /**
> @@ -273,13 +293,31 @@ static inline struct reset_control *devm_reset_control_get_shared(
> static inline struct reset_control *devm_reset_control_get_optional_exclusive(
> struct device *dev, const char *id)
> {
> - return __devm_reset_control_get(dev, id, 0, 0);
> + struct reset_control *desc;
> +
> + desc = __devm_reset_control_get(dev, id, 0, 0);
> +
> + if (IS_ERR(desc)) {
> + if (PTR_ERR(desc) == -ENOENT)
> + return NULL;
> + }

Same as for __of_reset_control_get above.

> + return desc;
> }
>
> static inline struct reset_control *devm_reset_control_get_optional_shared(
> struct device *dev, const char *id)
> {
> - return __devm_reset_control_get(dev, id, 0, 1);
> + struct reset_control *desc;
> +
> + desc = __devm_reset_control_get(dev, id, 0, 1);
> +
> + if (IS_ERR(desc)) {
> + if (PTR_ERR(desc) == -ENOENT)
> + return NULL;
> + }
> +
> + return desc;
> }
>
> /**

regards
Philipp

\
 
 \ /
  Last update: 2016-12-23 11:59    [W:0.073 / U:0.040 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site