iio: light: fix scale in veml3235 and add helpers to iio-gts
Subsystems (3)
Javier Carrasco
maintainerv2 cover iio: light: fix scale in veml3235 and add helpers to iio-gts
This series addresses an issue in the veml3235 that was inherited from an older driver (veml6030, not covered here but probably addressed after discussing this series), where the scale is does not follow ABI.
To simplify the gain/integration time handling, the iio-gts helpers have been used. And to further simplify the process, two new helpers have been proposed to address repetitive patterns that are found in all users of iio-gts.
The additions to iio-gts are wrappers around existing helpers, and I have tried to keep their names short, as adding more prefixes to the existing functions looked too cumbersome and inconvenient to follow the 80-char/line recommendation. I have not added any test for the new helpers because I would prefer to discuss them first.
This series has been tested with a veml3235sl under all supported gains and integration times as well as with a few unsupported values to make sure the operations fail in those cases.
---
Changes in v2:
- industrialio-gts-helper.c: explicitly return -EINVAL if iio_gts_find_new_gain_vy_gain_time_min() fails for the minimum gain fallback.
- industrialio-gts-helper.c: check the non-zero value of ret to continue
to the next iteration in iio_gts_find_gan_sel_in_times().
- veml3235.c: drop code style fixes and move them to a separate patch.
- veml3235.c: add patch to support regmap cache to simplify value
updates.
- veml3235.c: put 'val' back in the arguments of set_it() to drop the
diff, and check it internally as before.
- veml3235.c: fix typo in comment.
- Link to v1: https://lore.kernel.org/r/[email protected]
---
Javier Carrasco (4):
iio: gts-helper: add helpers to ease searches of gain_sel and new_gain
iio: light: veml3235: fix code style
iio: light: veml3235: extend regmap to add cache
iio: veml3235: fix scale to conform to ABI
drivers/iio/industrialio-gts-helper.c | 76 ++++++++++ drivers/iio/light/Kconfig | 1 + drivers/iio/light/veml3235.c | 274 ++++++++++++++++++++-------------- include/linux/iio/iio-gts-helper.h | 5 + 4 files changed, 245 insertions(+), 111 deletions(-)
---
base-commit: e25c8d66f6786300b680866c0e0139981273feba
change-id: 20241215-veml3235_scale-62de98c7b5fa
Best regards,
Signature · 1 line
Javier Carrasco <[email protected]>
Javier Carrasco
maintainerPATCH v2 1/4 iio: gts-helper: add helpers to ease searches of gain_sel and new_gain
This helper functions reduce the burden in the drivers that want to fetch a gain selector in all available times or a new optimal gain.
The former is currently achieved by calling iio_gts_find_gain_sel_for_scale_using_time() for the current time selector, and then iterating over the rest of time selectors if the gain selector was not found.
The latter requires a combination of multiple iio-gts helpers to find the new gain, look for an optimal gain if there was no exact match, and set a minimum gain if the optimal gain is not in the range of available gains.
Provide simpler workflows by means of functions that address common patterns in the users of the iio-gts helpers.
--- drivers/iio/industrialio-gts-helper.c | 76 +++++++++++++++++++++++++++++++++++ include/linux/iio/iio-gts-helper.h | 5 +++ 2 files changed, 81 insertions(+)
drivers/iio/industrialio-gts-helper.c +76 -0
@@ -915,6 +915,40 @@ int iio_gts_find_gain_sel_for_scale_using_time(struct iio_gts *gts, int time_sel }EXPORT_SYMBOL_NS_GPL (iio_gts_find_gain_sel_for_scale_using_time,"IIO_GTS_HELPER" ); +/** +* iio_gts_find_gain_sel_in_times - Fetch gain selector in the available times. +* @gts: Gain time scale descriptor +* @scale_int: Integral part of the scale (typically val1) +* @scale_nano: Fractional part of the scale (nano or ppb) +* @gain_sel: Pointer to value where gain selector is stored. +* @time_sel: Pointer to value where time selector is stored. +* +* Wrapper around iio_gts_find_gain_for_scale_using_time() to fetch the +* gain selector for all supported integration times. +* +* Return: 0 on success and -EINVAL on error. +*/ +int iio_gts_find_gain_sel_in_times (struct iio_gts *gts,int scale_int, +int scale_nano,int *gain_sel,int *time_sel) +{ +int i, ret; + +for (i =0 ; i < gts->num_itime; i++) { + *time_sel = gts->itime_table[i].sel; + ret =iio_gts_find_gain_sel_for_scale_using_time (gts, *time_sel, + scale_int, + scale_nano, + gain_sel); +if (ret) +continue ; + +return 0 ; + } + +return -EINVAL ; +} +EXPORT_SYMBOL_NS_GPL (iio_gts_find_gain_sel_in_times,"IIO_GTS_HELPER" ); +static int iio_gts_get_total_gain (struct iio_gts *gts,int gain,int time) {const struct iio_itime_sel_mul *itime;
@@ -1086,6 +1120,48 @@ int iio_gts_find_new_gain_by_old_gain_time(struct iio_gts *gts, int old_gain, }EXPORT_SYMBOL_NS_GPL (iio_gts_find_new_gain_by_old_gain_time,"IIO_GTS_HELPER" ); +/** +* iio_gts_find_new_gain_by_gain_time_min - compensate for time change +* @gts: Gain time scale descriptor +* @old_gain: Previously set gain +* @old_time: Selector corresponding previously set time +* @new_time: Selector corresponding new time to be set +* @new_gain: Pointer to value where new gain is to be written +* @in_range: Indicate if the @new_gain was in the range of +* supported gains. +* +* Wrapper around iio_gts_find_new_gain_by_old_gain_time() that tries to +* set an optimal value if no exact match was found, defaulting to the +* minimum gain to avoid saturations if the optimal value is not in the +* range of supported gains. +* +* Return: 0 on success and a negative value if no gain was found. +*/ +int iio_gts_find_new_gain_by_gain_time_min (struct iio_gts *gts,int old_gain, +int old_time,int new_time, +int *new_gain,bool *in_range) +{ +int ret; + + *in_range = true; + ret =iio_gts_find_new_gain_by_old_gain_time (gts, old_gain, old_time, + new_time, new_gain); +if (*new_gain <0 ) +return -EINVAL ; + +if (ret) { + *new_gain =iio_find_closest_gain_low (gts, *new_gain, in_range); +if (*new_gain <0 ) { + *new_gain =iio_gts_get_min_gain (gts); +if (*new_gain <0 ) +return -EINVAL ; + } + } + +return 0 ; +} +EXPORT_SYMBOL_NS_GPL (iio_gts_find_new_gain_by_gain_time_min,"IIO_GTS_HELPER" ); +MODULE_LICENSE ("GPL" );MODULE_AUTHOR ("Matti Vaittinen <[email protected]>" );MODULE_DESCRIPTION ("IIO light sensor gain-time-scale helpers" );
include/linux/iio/iio-gts-helper.h +5 -0
@@ -188,6 +188,8 @@ int iio_gts_total_gain_to_scale(struct iio_gts *gts, int total_gain,int iio_gts_find_gain_sel_for_scale_using_time (struct iio_gts *gts,int time_sel,int scale_int,int scale_nano,int *gain_sel); +int iio_gts_find_gain_sel_in_times (struct iio_gts *gts,int scale_int, +int scale_nano,int *gain_sel,int *time_sel);int iio_gts_get_scale (struct iio_gts *gts,int gain,int time,int *scale_int,int *scale_nano); int iio_gts_find_new_gain_sel_by_old_gain_time(struct iio_gts *gts,
@@ -196,6 +198,9 @@ int iio_gts_find_new_gain_sel_by_old_gain_time(struct iio_gts *gts,int iio_gts_find_new_gain_by_old_gain_time (struct iio_gts *gts,int old_gain,int old_time,int new_time,int *new_gain); +int iio_gts_find_new_gain_by_gain_time_min (struct iio_gts *gts,int old_gain, +int old_time,int new_time, +int *new_gain,bool *in_range);int iio_gts_avail_times (struct iio_gts *gts,const int **vals,int *type,int *length); int iio_gts_all_avail_scales(struct iio_gts *gts, const int **vals, int *type,
Signature · 1 line
2.43.0
Jonathan Cameron
maintainerRe: [PATCH v2 1/4] iio: gts-helper: add helpers to ease searches of gain_sel and new_gain
Quoted 13 lines · On Tue, 24 Dec 2024 11:59:00 +0100 Javier Carrasco…
This helper functions reduce the burden in the drivers that want to fetch a gain selector in all available times or a new optimal gain.
The former is currently achieved by calling
iio_gts_find_gain_sel_for_scale_using_time()for the current time selector, and then iterating over the rest of time selectors if the gain selector was not found.The latter requires a combination of multiple iio-gts helpers to find the new gain, look for an optimal gain if there was no exact match, and set a minimum gain if the optimal gain is not in the range of available gains.
Quoted 4 lines
Provide simpler workflows by means of functions that address common patterns in the users of the iio-gts helpers.
Signed-off-by Javier Carrasco <[email protected]>
Matti gave an Ack. If you intentionally dropped it due to significant
changes, you should say so...
Quoted 1 line
---
here.
Other than that, looks fine to me.
Jonathan
Javier Carrasco
maintainerRe: [PATCH v2 1/4] iio: gts-helper: add helpers to ease searches of gain_sel and new_gain
Quoted 23 lines · Jonathan Cameron
Quoted 13 lines · On Tue, 24 Dec 2024 11:59:00 +0100 Javier Carrasco…
This helper functions reduce the burden in the drivers that want to fetch a gain selector in all available times or a new optimal gain.
The former is currently achieved by calling
iio_gts_find_gain_sel_for_scale_using_time()for the current time selector, and then iterating over the rest of time selectors if the gain selector was not found.The latter requires a combination of multiple iio-gts helpers to find the new gain, look for an optimal gain if there was no exact match, and set a minimum gain if the optimal gain is not in the range of available gains.
Quoted 4 lines
Provide simpler workflows by means of functions that address common patterns in the users of the iio-gts helpers.
Signed-off-by Javier Carrasco <[email protected]>Matti gave an Ack. If you intentionally dropped it due to significant
changes, you should say so...
Quoted 7 lines
Quoted 1 line
---
here.
Other than that, looks fine to me.
Jonathan
Hi Jonathan, you are absolutely right.
I did not add the Ack on purpose because I thought that I had to modify the helpers functions, but in the end that was not the case. Matti's tag should have stayed.
Best regards,
Javier Carrasco
Javier Carrasco
maintainerPATCH v2 2/4 iio: light: veml3235: fix code style
In reply to Javier CarrascoTrivial fixes to drop double spacings.
--- drivers/iio/light/veml3235.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/iio/light/veml3235.c +3 -3
@@ -321,7 +321,7 @@ static void veml3235_read_id(struct veml3235_data *data) {int ret, reg; - ret =regmap_field_read (data->rf.id, ®); + ret =regmap_field_read (data->rf.id, ®);if (ret) {dev_info (data->dev,"failed to read ID\n" );return ;
@@ -389,8 +389,8 @@ static int veml3235_hw_init(struct iio_dev *indio_dev) }static const struct iio_info veml3235_info = { - .read_raw = veml3235_read_raw, - .read_avail = veml3235_read_avail, + .read_raw = veml3235_read_raw, + .read_avail = veml3235_read_avail, .write_raw = veml3235_write_raw, };
Signature · 1 line
2.43.0
Jonathan Cameron
maintainerRe: [PATCH v2 2/4] iio: light: veml3235: fix code style
Quoted 3 lines · On Tue, 24 Dec 2024 11:59:01 +0100 Javier Carrasco…
Trivial fixes to drop double spacings.
Signed-off-by Javier Carrasco <[email protected]>
This stands fine on it's own.
Applied to the togreg branch of iio.git and pushed out as testing for 0-day to take a look.
Thanks,
Jonathan
Quoted 29 lines
--- drivers/iio/light/veml3235.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)drivers/iio/light/veml3235.c +3
-3@@ -321,7 +321,7 @@ static void veml3235_read_id(struct veml3235_data *data) {int ret, reg; - ret =regmap_field_read (data->rf.id, ®); + ret =regmap_field_read (data->rf.id, ®);if (ret) {dev_info (data->dev,"failed to read ID\n" );return ;@@ -389,8 +389,8 @@ static int veml3235_hw_init(struct iio_dev *indio_dev) }static const struct iio_info veml3235_info = { - .read_raw = veml3235_read_raw, - .read_avail = veml3235_read_avail, + .read_raw = veml3235_read_raw, + .read_avail = veml3235_read_avail, .write_raw = veml3235_write_raw, };
Javier Carrasco
maintainerPATCH v2 3/4 iio: light: veml3235: extend regmap to add cache
In reply to Javier CarrascoThe configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
--- drivers/iio/light/veml3235.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)
drivers/iio/light/veml3235.c +31 -0
@@ -101,12 +101,43 @@ static const struct iio_chan_spec veml3235_channels[] = { }, }; +static const struct regmap_range veml3235_readable_ranges[] = { +regmap_reg_range (VEML3235_REG_CONF ,VEML3235_REG_ID ), +}; + +static const struct regmap_access_table veml3235_readable_table = { + .yes_ranges = veml3235_readable_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_readable_ranges), +}; + +static const struct regmap_range veml3235_writable_ranges[] = { +regmap_reg_range (VEML3235_REG_CONF ,VEML3235_REG_CONF ), +}; + +static const struct regmap_access_table veml3235_writable_table = { + .yes_ranges = veml3235_writable_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_writable_ranges), +}; + +static const struct regmap_range veml3235_volatile_ranges[] = { +regmap_reg_range (VEML3235_REG_WH_DATA ,VEML3235_REG_ALS_DATA ), +}; + +static const struct regmap_access_table veml3235_volatile_table = { + .yes_ranges = veml3235_volatile_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_volatile_ranges), +}; +static const struct regmap_config veml3235_regmap_config = { .name ="veml3235_regmap" , .reg_bits =8 , .val_bits =16 , .max_register =VEML3235_REG_ID , .val_format_endian =REGMAP_ENDIAN_LITTLE , + .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type =REGCACHE_RBTREE , };static int veml3235_get_it (struct veml3235_data *data,int *val,int *val2)
Signature · 1 line
2.43.0
Re: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 5 lines · On Tue, 24 Dec 2024 11:59:02 +0100 Javier Carrasco…
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
Signed-off-by Javier Carrasco <[email protected]>
This is also fine on it's own.
Applied.
Quoted 53 lines
--- drivers/iio/light/veml3235.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)drivers/iio/light/veml3235.c +31
-0@@ -101,12 +101,43 @@ static const struct iio_chan_spec veml3235_channels[] = { }, }; +static const struct regmap_range veml3235_readable_ranges[] = { +regmap_reg_range (VEML3235_REG_CONF ,VEML3235_REG_ID ), +}; + +static const struct regmap_access_table veml3235_readable_table = { + .yes_ranges = veml3235_readable_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_readable_ranges), +}; + +static const struct regmap_range veml3235_writable_ranges[] = { +regmap_reg_range (VEML3235_REG_CONF ,VEML3235_REG_CONF ), +}; + +static const struct regmap_access_table veml3235_writable_table = { + .yes_ranges = veml3235_writable_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_writable_ranges), +}; + +static const struct regmap_range veml3235_volatile_ranges[] = { +regmap_reg_range (VEML3235_REG_WH_DATA ,VEML3235_REG_ALS_DATA ), +}; + +static const struct regmap_access_table veml3235_volatile_table = { + .yes_ranges = veml3235_volatile_ranges, + .n_yes_ranges =ARRAY_SIZE (veml3235_volatile_ranges), +}; +static const struct regmap_config veml3235_regmap_config = { .name ="veml3235_regmap" , .reg_bits =8 , .val_bits =16 , .max_register =VEML3235_REG_ID , .val_format_endian =REGMAP_ENDIAN_LITTLE , + .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type =REGCACHE_RBTREE , };static int veml3235_get_it (struct veml3235_data *data,int *val,int *val2)
Re: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
In reply to Javier CarrascoTue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
drivers/iio/light/veml3235.c static const struct iio_chan_spec veml3235_channels[] = { · line 140
Quoted 10 lines
static const struct
regmap_configveml3235_regmap_config= {drivers/iio/light/veml3235.c +4
-0.name = "veml3235_regmap", .reg_bits = 8, .val_bits = 16, .max_register = VEML3235_REG_ID, .val_format_endian = REGMAP_ENDIAN_LITTLE, + .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type = REGCACHE_RBTREE,
Any specific reason why this is NOT a maple tree?
Quoted 1 line
};
Signature · 2 lines
With Best Regards, Andy Shevchenko
Javier Carrasco
maintainerRe: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 7 lines · Andy Shevchenko
Tue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
Quoted 13 lines
drivers/iio/light/veml3235.c static const struct iio_chan_spec veml3235_channels[] = { · line 140
Quoted 10 lines
static const struct
regmap_configveml3235_regmap_config= {drivers/iio/light/veml3235.c +4
-0.name = "veml3235_regmap", .reg_bits = 8, .val_bits = 16, .max_register = VEML3235_REG_ID, .val_format_endian = REGMAP_ENDIAN_LITTLE, + .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type = REGCACHE_RBTREE,Any specific reason why this is NOT a maple tree?
Hello Andy,
I followed the most common approach in IIO (52 RBTREE vs 2 MAPLE), assuming that the "low-end systems" comment for the different REGCACHE_* applies well to the typical systems that will make use of this driver, and many others under IIO. I considered that *possible* performance advantage for low-end systems above other considerations, like the general rule about using maple tree.
Thanks and best regards,
Javier Carrasco
Re: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 20 lines · Javier Carrasco
Quoted 18 lines · Andy Shevchenko
Tue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
Quoted 10 lines
static const struct regmap_config veml3235_regmap_config = { .name = "veml3235_regmap", .reg_bits = 8, .val_bits = 16, .max_register = VEML3235_REG_ID, .val_format_endian = REGMAP_ENDIAN_LITTLE,drivers/iio/light/veml3235.c +4
-0+ .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type = REGCACHE_RBTREE,
Quoted 5 lines
Quoted 1 line
Any specific reason why this is NOT a maple tree?
Hello Andy,
I followed the most common approach in IIO (52 RBTREE vs 2 MAPLE),
But it's historical and can't be taken as an argument.
Quoted 5 lines
assuming that the "low-end systems" comment for the different REGCACHE_* applies well to the typical systems that will make use of this driver, and many others under IIO. I considered that *possible* performance advantage for low-end systems above other considerations, like the general rule about using maple tree.
https://elixir.bootlin.com/linux/v6.13-rc3/source/include/linux/regmap.h#L58
"Any new caches
* should usually use the maple tree cache unless they specifically
* require that there are never any allocations at runtime and can't
* provide defaults in which case they should use the flat cache."
Can you reconsider?
Signature · 2 lines
With Best Regards, Andy Shevchenko
Javier Carrasco
maintainerRe: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 36 lines · Andy Shevchenko
Quoted 20 lines · Javier Carrasco
Quoted 18 lines · Andy Shevchenko
Tue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
Quoted 10 lines
static const struct regmap_config veml3235_regmap_config = { .name = "veml3235_regmap", .reg_bits = 8, .val_bits = 16, .max_register = VEML3235_REG_ID, .val_format_endian = REGMAP_ENDIAN_LITTLE,drivers/iio/light/veml3235.c +4
-0+ .rd_table = &veml3235_readable_table, + .wr_table = &veml3235_writable_table, + .volatile_table = &veml3235_volatile_table, + .cache_type = REGCACHE_RBTREE,Quoted 5 lines
Quoted 1 line
Any specific reason why this is NOT a maple tree?
Hello Andy,
I followed the most common approach in IIO (52 RBTREE vs 2 MAPLE),
But it's historical and can't be taken as an argument.
Quoted 5 lines
assuming that the "low-end systems" comment for the different REGCACHE_* applies well to the typical systems that will make use of this driver, and many others under IIO. I considered that *possible* performance advantage for low-end systems above other considerations, like the general rule about using maple tree.
Quoted 8 lines
https://elixir.bootlin.com/linux/v6.13-rc3/source/include/linux/regmap.h#L58
"Any new caches
* should usually use the maple tree cache unless they specifically
* require that there are never any allocations at runtime and can't
* provide defaults in which case they should use the flat cache."Can you reconsider?
That was exactly the comment I referenced, actually the part about low-end systems that appears right after what you highlighted.
I have nothing against switching to MAPLE, if that is preferred even if the main user of this driver will be a low-end system. I think that IIO is a typical subsystem that addresses needs for very low-end systems that are sometimes slightly more powerful than a microcontroller, but on the other hand I am by no means an expert, and if MAPLE is the way to go here as well, I will send a follow-up patch for it.
Thank you for your feedback!
Javier Carrasco
Re: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 8 lines · Javier Carrasco
Quoted 7 lines · Andy Shevchenko
Quoted 5 lines · Javier Carrasco
Quoted 4 lines · Andy Shevchenko
Tue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
Quoted 26 lines
Quoted 14 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line
+ .cache_type =
REGCACHE_RBTREE,Any specific reason why this is NOT a maple tree?
I followed the most common approach in IIO (52 RBTREE vs 2 MAPLE),
But it's historical and can't be taken as an argument.
Quoted 5 lines
assuming that the "low-end systems" comment for the different REGCACHE_* applies well to the typical systems that will make use of this driver, and many others under IIO. I considered that *possible* performance advantage for low-end systems above other considerations, like the general rule about using maple tree.
Quoted 8 lines
https://elixir.bootlin.com/linux/v6.13-rc3/source/include/linux/regmap.h#L58
"Any new caches
* should usually use the maple tree cache unless they specifically
* require that there are never any allocations at runtime and can't
* provide defaults in which case they should use the flat cache."Can you reconsider?
That was exactly the comment I referenced, actually the part about low-end systems that appears right after what you highlighted.
Quoted 6 lines
I have nothing against switching to MAPLE, if that is preferred even if the main user of this driver will be a low-end system. I think that IIO is a typical subsystem that addresses needs for very low-end systems that are sometimes slightly more powerful than a microcontroller, but on the other hand I am by no means an expert, and if MAPLE is the way to go here as well, I will send a follow-up patch for it.
Ah, I see now. Okay, I leave it then to Jonathan as I am okay with any choice as long as it's understood and justified.
Signature · 2 lines
With Best Regards, Andy Shevchenko
Re: [PATCH v2 3/4] iio: light: veml3235: extend regmap to add cache
Quoted 46 lines · On Sun, 12 Jan 2025 20:50:41 +0200 Andy Shevchenko <[email protected]>…
Quoted 8 lines · Javier Carrasco
Quoted 7 lines · Andy Shevchenko
Quoted 5 lines · Javier Carrasco
Quoted 4 lines · Andy Shevchenko
Tue, Dec 24, 2024 at 11:59:02AM +0100, Javier Carrasco kirjoitti:
Quoted 3 lines
The configuration and ID registers are not volatile and are not affected by read operations (i.e. not precious), making them suitable to be cached in order to reduce the number of accesses to the device.
...
Quoted 26 lines
Quoted 14 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line
+ .cache_type =
REGCACHE_RBTREE,Any specific reason why this is NOT a maple tree?
I followed the most common approach in IIO (52 RBTREE vs 2 MAPLE),
But it's historical and can't be taken as an argument.
Quoted 5 lines
assuming that the "low-end systems" comment for the different REGCACHE_* applies well to the typical systems that will make use of this driver, and many others under IIO. I considered that *possible* performance advantage for low-end systems above other considerations, like the general rule about using maple tree.
Quoted 8 lines
https://elixir.bootlin.com/linux/v6.13-rc3/source/include/linux/regmap.h#L58
"Any new caches
* should usually use the maple tree cache unless they specifically
* require that there are never any allocations at runtime and can't
* provide defaults in which case they should use the flat cache."Can you reconsider?
That was exactly the comment I referenced, actually the part about low-end systems that appears right after what you highlighted.
Quoted 6 lines
I have nothing against switching to MAPLE, if that is preferred even if the main user of this driver will be a low-end system. I think that IIO is a typical subsystem that addresses needs for very low-end systems that are sometimes slightly more powerful than a microcontroller, but on the other hand I am by no means an expert, and if MAPLE is the way to go here as well, I will send a follow-up patch for it.
Quoted 2 lines
Ah, I see now. Okay, I leave it then to Jonathan as I am okay with any choice as long as it's understood and justified.
For a fairly small regmap, I doubt there is a strong reason to care about the choice. So stick to what you have. For future drivers we can reassess as makes sense.
Jonathan
Quoted 2 lines
Javier Carrasco
maintainerPATCH v2 4/4 iio: veml3235: fix scale to conform to ABI
In reply to Javier CarrascoThe current scale is not ABI-compliant as it is just the sensor gain instead of the value that acts as a multiplier to be applied to the raw value (there is no offset).
Use the iio-gts helpers to obtain the proper scale values according to the gain and integration time to match the resolution tables from the datasheet. When at it, use 'scale' instead of 'gain' consistently for the get/set functions to avoid misunderstandings.
--- drivers/iio/light/Kconfig | 1 + drivers/iio/light/veml3235.c | 237 +++++++++++++++++++++++-------------------- 2 files changed, 130 insertions(+), 108 deletions(-)
drivers/iio/light/Kconfig +1 -0
@@ -656,6 +656,7 @@ config VCNL4035 config VEML3235 tristate "VEML3235 ambient light sensor" select REGMAP_I2C + select IIO_GTS_HELPER depends on I2C help Say Y here if you want to build a driver for the Vishay VEML3235
drivers/iio/light/veml3235.c +129 -108
333 lines not on this page · raw message
Signature · 1 line
2.43.0
Jonathan Cameron
maintainerRe: [PATCH v2 4/4] iio: veml3235: fix scale to conform to ABI
Quoted 11 lines · On Tue, 24 Dec 2024 11:59:03 +0100 Javier Carrasco…
The current scale is not ABI-compliant as it is just the sensor gain instead of the value that acts as a multiplier to be applied to the raw value (there is no offset).
Use the iio-gts helpers to obtain the proper scale values according to the gain and integration time to match the resolution tables from the datasheet. When at it, use 'scale' instead of 'gain' consistently for the get/set functions to avoid misunderstandings.
Fixes c5a23f80c164 ("iio: light: add support for veml3235")Signed-off-by Javier Carrasco <[email protected]>
Hi Javier,
To me this looks fine, but I'd like Matti to take another quick look at this and patch 1 if time allows given Matti is a lot more familiar with the GTS helpers than I am.
Jonathan
Re: [PATCH v2 4/4] iio: veml3235: fix scale to conform to ABI
In reply to Javier CarrascoQuoted 4 lines · Javier Carrasco
The current scale is not ABI-compliant as it is just the sensor gain instead of the value that acts as a multiplier to be applied to the raw value (there is no offset).
Quoted 8 lines
Use the iio-gts helpers to obtain the proper scale values according to the gain and integration time to match the resolution tables from the datasheet. When at it, use 'scale' instead of 'gain' consistently for the get/set functions to avoid misunderstandings.
Fixes c5a23f80c164 ("iio: light: add support for veml3235")Signed-off-by Javier Carrasco <[email protected]>---
This looks good to me, although I now think we made a mistake with the naming of the iio_gts_find_gain_sel_in_times().
The intended use is finding the gain and time (selector) for the new scale (while preferring keeping the time unchanged if possible), right?
So, in this regard it'd be better to use name which reflects the fact that the function finds gain and time for given scale.
I would now (after having to look the doc of this new function while reviewing the code 2 weeks after reviewing this new function :rolleyes:) name it something like:
iio_gts_find_gain_time_sel_for_scale()
Well, it's not really in the scope of the review anymore, but I'd love to see a renaming patch while we have only one user... :)
Anyways:
Yours,
-- Matti
Javier Carrasco
maintainerRe: [PATCH v2 4/4] iio: veml3235: fix scale to conform to ABI
Quoted 32 lines · Matti Vaittinen
Quoted 4 lines · Javier Carrasco
The current scale is not ABI-compliant as it is just the sensor gain instead of the value that acts as a multiplier to be applied to the raw value (there is no offset).
Quoted 8 lines
Use the iio-gts helpers to obtain the proper scale values according to the gain and integration time to match the resolution tables from the datasheet. When at it, use 'scale' instead of 'gain' consistently for the get/set functions to avoid misunderstandings.
Fixes c5a23f80c164 ("iio: light: add support for veml3235")Signed-off-by Javier Carrasco <[email protected]>---
This looks good to me, although I now think we made a mistake with the naming of the
iio_gts_find_gain_sel_in_times().The intended use is finding the gain and time (selector) for the new scale (while preferring keeping the time unchanged if possible), right?
So, in this regard it'd be better to use name which reflects the fact that the function finds gain and time for given scale.
I would now (after having to look the doc of this new function while reviewing the code 2 weeks after reviewing this new function :rolleyes:) name it something like:
iio_gts_find_gain_time_sel_for_scale()Well, it's not really in the scope of the review anymore, but I'd love to see a renaming patch while we have only one user... :)
Quoted 6 lines
Anyways:
Reviewed-by Matti Vaittinen <[email protected]>Yours,
-- Matti
Hi Matti,
Thank you for your suggestion, I will add it to v3 as this patch and the one that introduced the helper functions have not been applied yet, so we don't need an extra patch to rename the function. I will add your tag too because I will only change what you suggested.
Best regards,
Javier Carrasco
Matti Vaittinen
maintainerRe: [PATCH v2 4/4] iio: veml3235: fix scale to conform to ABI
Quoted 35 lines · Javier Carrasco
Quoted 34 lines · Matti Vaittinen
Quoted 4 lines · Javier Carrasco
The current scale is not ABI-compliant as it is just the sensor gain instead of the value that acts as a multiplier to be applied to the raw value (there is no offset).
Quoted 8 lines
Use the iio-gts helpers to obtain the proper scale values according to the gain and integration time to match the resolution tables from the datasheet. When at it, use 'scale' instead of 'gain' consistently for the get/set functions to avoid misunderstandings.
Fixes c5a23f80c164 ("iio: light: add support for veml3235")Signed-off-by Javier Carrasco <[email protected]>---
This looks good to me, although I now think we made a mistake with the naming of the
iio_gts_find_gain_sel_in_times().The intended use is finding the gain and time (selector) for the new scale (while preferring keeping the time unchanged if possible), right?
So, in this regard it'd be better to use name which reflects the fact that the function finds gain and time for given scale.
I would now (after having to look the doc of this new function while reviewing the code 2 weeks after reviewing this new function :rolleyes:) name it something like:
iio_gts_find_gain_time_sel_for_scale()Well, it's not really in the scope of the review anymore, but I'd love to see a renaming patch while we have only one user... :)
Anyways:
Quoted 7 lines
Quoted 2 lines
Reviewed-by Matti Vaittinen <[email protected]>Thank you for your suggestion, I will add it to v3 as this patch and the one that introduced the helper functions have not been applied yet, so we don't need an extra patch to rename the function.
Great, Thanks!
Quoted 2 lines
I will add your
tag too because I will only change what you suggested.
Sure!
Yours,
-- Matti