kernel/drivers/video/backlight/mp3309c.c, branch linux-rolling-stable

backlight: mp3309c: Initialize backlight properties without memset

2025-09-02T10:36:59Z

Assigning values to a struct using a compound literal (since C99) also guarantees that all unspecified struct members are empty-initialized, so it properly replaces the memset to zero. The code looks a bit nicer and more idiomatic (though that might be subjective?). The resulting binary is a bit smaller. On ARCH=arm with an allnoconfig + minimal changes to enable the mp3309c driver the difference is 12 bytes. Signed-off-by: Uwe Kleine-König Tested-by: Flavio Suligoi Reviewed-by: Daniel Thompson (RISCstar) Link: https://lore.kernel.org/r/14514a1b0d3df6438aa10bb74f1c4fc2367d9987.1751361465.git.u.kleine-koenig@baylibre.com Signed-off-by: Lee Jones

backlight: mp3309c: Drop pwm_apply_args()

2025-09-02T10:36:47Z

pwm_apply_args() sole purpose is to initialize all parameters specified in the device tree for consumers that rely on pwm_config() and pwm_enable(). The mp3309c backlight driver uses pwm_apply_might_sleep() which gets passed the full configuration and so doesn't rely on the default being explicitly applied. Signed-off-by: Uwe Kleine-König Tested-by: Flavio Suligoi Reviewed-by: Daniel Thompson (RISCstar) Link: https://lore.kernel.org/r/2d1075f5dd45c7c135e4326e279468de699f9d17.1751361465.git.u.kleine-koenig@baylibre.com Signed-off-by: Lee Jones

backlight: mp3309c: Use backlight power constants

2024-07-04T15:45:25Z

Replace FB_BLANK_ constants with their counterparts from the backlight subsystem. The values are identical, so there's no change in functionality. Signed-off-by: Thomas Zimmermann Reviewed-by: Daniel Thompson Link: https://lore.kernel.org/r/20240624152033.25016-13-tzimmermann@suse.de Signed-off-by: Lee Jones

backlight: Drop explicit initialization of struct i2c_device_id::driver_data to 0

2024-06-20T17:36:25Z

These drivers don't use the driver_data member of struct i2c_device_id, so don't explicitly initialize this member. This prepares putting driver_data in an anonymous union which requires either no initialization or named designators. But it's also a nice cleanup on its own. While add it, also remove commas after the sentinel entries. Signed-off-by: Uwe Kleine-König Link: https://lore.kernel.org/r/20240619193558.2543645-2-u.kleine-koenig@baylibre.com Signed-off-by: Lee Jones

backlight: mp3309c: Fix LEDs flickering in PWM mode

2024-05-10T14:44:29Z

The mp3309 has two configuration registers, named according to their address (0x00 and 0x01). In the second register (0x01), the bit DIMS (Dimming Mode Select) must be always 0 (zero), in both analog (via I2C commands) and PWM dimming mode. In the initial driver version, the DIMS bit was set in PWM mode and reset in analog mode. But if the DIMS bit is set in pwm dimming mode and other devices are connected on the same I2C bus, every I2C commands on the bus generates a flickering on the LEDs powered by the mp3309c. This change concerns the chip initialization and does not impact any existing device-tree configuration. Signed-off-by: Flavio Suligoi Reviewed-by: Daniel Thompson Link: https://lore.kernel.org/r/20240417153105.1794134-2-f.suligoi@asem.it Signed-off-by: Lee Jones

backlight: mp3309c: Fix signedness bug in mp3309c_parse_fwnode()

2024-05-10T14:44:28Z

The "num_levels" variable is used to store error codes from device_property_count_u32() so it needs to be signed. This doesn't cause an issue at runtime because devm_kcalloc() won't allocate negative sizes. However, it's still worth fixing. Fixes: b54c828bdba9 ("backlight: mp3309c: Make use of device properties") Signed-off-by: Dan Carpenter Reviewed-by: Daniel Thompson Tested-by: Flavio Suligoi Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/74347f67-360d-4513-8939-595e3c4764fa@moroto.mountain Signed-off-by: Lee Jones

backlight: Remove fb_blank from struct backlight_properties

2024-03-28T10:16:26Z

Remove the field fb_blank from struct backlight_properties and remove all code that still sets or reads it. Backlight blank status is now tracked exclusively in struct backlight_properties.state. The core backlight code keeps the fb_blank and state fields in sync, but doesn't do anything else with fb_blank. Several drivers initialize fb_blank to FB_BLANK_UNBLANK to enable the backlight. This is already the default for the state field. So we can delete the fb_blank code from core and drivers and rely on the state field. Signed-off-by: Thomas Zimmermann Cc: Flavio Suligoi Cc: Nicolas Ferre Cc: Alexandre Belloni Cc: Claudiu Beznea Tested-by: Flavio Suligoi Reviewed-by: Daniel Thompson Reviewed-by: Dan Carpenter Link: https://lore.kernel.org/r/20240319093915.31778-7-tzimmermann@suse.de Signed-off-by: Lee Jones

backlight: mp3309c: Fully initialize backlight_properties during probe

2024-03-07T09:03:29Z

props is stack allocated and, although this driver initializes all the fields that are not "owned" by the framework, we'd still like to ensure it is zeroed to avoid problems from this driver if the fields change. Signed-off-by: Daniel Thompson Link: https://lore.kernel.org/r/20240220153532.76613-5-daniel.thompson@linaro.org Signed-off-by: Lee Jones

backlight: mp3309c: Utilise temporary variable for struct device

2024-03-07T09:03:28Z

We have a temporary variable to keep pointer to struct device. Utilise it where it makes sense. Signed-off-by: Andy Shevchenko Reviewed-by: Daniel Thompson Tested-by: Flavio Suligoi Link: https://lore.kernel.org/r/20240208184313.2224579-4-andriy.shevchenko@linux.intel.com Signed-off-by: Lee Jones

backlight: mp3309c: Use dev_err_probe() instead of dev_err()

2024-03-07T09:03:28Z

Replace dev_err() with dev_err_probe(). This helps in simplifing code and standardizing the error output. Signed-off-by: Andy Shevchenko Tested-by: Flavio Suligoi Reviewed-by: Daniel Thompson Link: https://lore.kernel.org/r/20240208184313.2224579-3-andriy.shevchenko@linux.intel.com Signed-off-by: Lee Jones