| Message ID | 20220909175522.179175-13-sebastian.reichel@collabora.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Introduce RK806 Support | expand |
pe 9. syysk. 2022 klo 21.21 Sebastian Reichel (sebastian.reichel@collabora.com) kirjoitti: > > Expose and document the table lookup logic used by > regulator_set_ramp_delay_regmap, so that it can be > reused for devices that cannot be configured via > regulator_set_ramp_delay_regmap. > I am always in favor of adding helpers to be used for common tasks. I am not demanding this (so please ignore my comment if you feel so) but I guess finding the "closest bigger" from a table is not ramp-delay or even regulator specific. I believe something like this might live under lib - if exported. Anyways, and whatever it is worth: Acked-by: Matti Vaittinen <matti.vaittinen@mazziesaccount@gmail.com> for exporting such a helper. (Or please disregard the ack if acks are reserved only for maintainers - It's just a sign that I think this is a good idea [even if it could live in lib]) Yours -- Matti
Hi, On Sat, Sep 10, 2022 at 08:17:09PM +0300, Matti Vaittinen wrote: > pe 9. syysk. 2022 klo 21.21 Sebastian Reichel > (sebastian.reichel@collabora.com) kirjoitti: > > > > Expose and document the table lookup logic used by > > regulator_set_ramp_delay_regmap, so that it can be > > reused for devices that cannot be configured via > > regulator_set_ramp_delay_regmap. > > > > I am always in favor of adding helpers to be used for common tasks. I > am not demanding this (so please ignore my comment if you feel so) but > I guess finding the "closest bigger" from a table is not ramp-delay or > even regulator specific. I believe something like this might live > under lib - if exported. I'm bad with naming :) If you or Mark have a specific suggestion I can change it accordingly. > Anyways, and whatever it is worth: > Acked-by: Matti Vaittinen <matti.vaittinen@mazziesaccount@gmail.com> > for exporting such a helper. (Or please disregard the ack if acks are > reserved only for maintainers - It's just a sign that I think this is > a good idea [even if it could live in lib]) I usually use 'Reviewed-by: <...>' in that case. But your Acked line is broken. Looks like you have split personality problem between work and private mail account :) -- Sebastian
On Fri, Sep 09, 2022 at 07:55:20PM +0200, Sebastian Reichel wrote: > -static int find_closest_bigger(unsigned int target, const unsigned int *table, > +int regmap_find_closest_bigger(unsigned int target, const unsigned int *table, > unsigned int num_sel, unsigned int *sel) So this will create a function in the regmap namespace in the regulator code which doesn't seem to interact with regmaps in any way? I'm a bit confused.
diff --git a/drivers/regulator/helpers.c b/drivers/regulator/helpers.c index ad2237a95572..eb1f18917d05 100644 --- a/drivers/regulator/helpers.c +++ b/drivers/regulator/helpers.c @@ -902,7 +902,20 @@ bool regulator_is_equal(struct regulator *reg1, struct regulator *reg2) } EXPORT_SYMBOL_GPL(regulator_is_equal); -static int find_closest_bigger(unsigned int target, const unsigned int *table, +/** + * regmap_find_closest_bigger - helper to find offset in ramp delay table + * + * @target: targeted ramp_delay + * @table: table with supported ramp delays + * @num_sel: number of entries in the table + * @sel: Pointer to store table offset + * + * This is the internal helper used by regulator_set_ramp_delay_regmap to + * map ramp delay to register value. It should only be used directly if + * regulator_set_ramp_delay_regmap cannot handle a specific device setup + * (e.g. because the value is split over multiple registers). + */ +int regmap_find_closest_bigger(unsigned int target, const unsigned int *table, unsigned int num_sel, unsigned int *sel) { unsigned int s, tmp, max, maxsel = 0; @@ -933,6 +946,7 @@ static int find_closest_bigger(unsigned int target, const unsigned int *table, return 0; } +EXPORT_SYMBOL_GPL(regmap_find_closest_bigger); /** * regulator_set_ramp_delay_regmap - set_ramp_delay() helper @@ -951,8 +965,8 @@ int regulator_set_ramp_delay_regmap(struct regulator_dev *rdev, int ramp_delay) if (WARN_ON(!rdev->desc->n_ramp_values || !rdev->desc->ramp_delay_table)) return -EINVAL; - ret = find_closest_bigger(ramp_delay, rdev->desc->ramp_delay_table, - rdev->desc->n_ramp_values, &sel); + ret = regmap_find_closest_bigger(ramp_delay, rdev->desc->ramp_delay_table, + rdev->desc->n_ramp_values, &sel); if (ret) { dev_warn(rdev_get_dev(rdev), diff --git a/include/linux/regulator/driver.h b/include/linux/regulator/driver.h index f9a7461e72b8..9501dbec64da 100644 --- a/include/linux/regulator/driver.h +++ b/include/linux/regulator/driver.h @@ -757,6 +757,8 @@ int regulator_set_current_limit_regmap(struct regulator_dev *rdev, int min_uA, int max_uA); int regulator_get_current_limit_regmap(struct regulator_dev *rdev); void *regulator_get_init_drvdata(struct regulator_init_data *reg_init_data); +int regmap_find_closest_bigger(unsigned int target, const unsigned int *table, + unsigned int num_sel, unsigned int *sel); int regulator_set_ramp_delay_regmap(struct regulator_dev *rdev, int ramp_delay); int regulator_sync_voltage_rdev(struct regulator_dev *rdev);
Expose and document the table lookup logic used by regulator_set_ramp_delay_regmap, so that it can be reused for devices that cannot be configured via regulator_set_ramp_delay_regmap. Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com> --- drivers/regulator/helpers.c | 20 +++++++++++++++++--- include/linux/regulator/driver.h | 2 ++ 2 files changed, 19 insertions(+), 3 deletions(-)