linux/drivers/leds, branch v6.17

treewide: rename GPIO set callbacks back to their original names

2025-08-07T08:07:06Z

The conversion of all GPIO drivers to using the .set_rv() and .set_multiple_rv() callbacks from struct gpio_chip (which - unlike their predecessors - return an integer and allow the controller drivers to indicate failures to users) is now complete and the legacy ones have been removed. Rename the new callbacks back to their original names in one sweeping change. Signed-off-by: Bartosz Golaszewski

Merge tag 'leds-next-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/leds

2025-07-31T18:54:01Z

Pull LED updates from Lee Jones: "Improvements & Fixes: - A fix for QCOM Flash to prevent incorrect register access when the driver is re-bound. This is solved by duplicating a static register array during the probe function, which prevents register addresses from being miscalculated after multiple binds - The LP50xx driver now correctly handles the 'reg' property in device tree child nodes to ensure the multi_index is set correctly. This prevents issues where LEDs could be controlled incorrectly if the device tree nodes were processed in a different order. An error is returned if the reg property is missing or out of range - Add a Kconfig dependency on between TPS6131x and V4L2_FLASH_LED_CLASS to prevent a build failure when the driver is built-in and the V4L2 flash infrastructure is a loadable module - Fix a potential buffer overflow warning in PCA955x reported by older GCC versions by using a more precise format specifier when creating the default LED label. Cleanups & Refactoring: - Correct the MAINTAINERS file entry for the TPS6131X flash LED driver to point to the correct device tree binding file name - Fix a comment in the Flash Class for the flash_timeout setter to "flash timeout" from "flash duration" for accuracy - The of_led_get() function is no longer exported as it has no users outside of its own module. Removals: - Revert the commit to configure LED blink intervals for hardware offload in the Netdev Trigger. This change was found to break existing PHY drivers by putting their LEDs into a permanent, unconditional blinking state. Device Tree Bindings Updates: - Update the binding for LP50xx to document that the child reg property is the index within the LED bank. The example was also updated to use correct values - Update the JNCP5623 binding to add 0x39 as a valid I2C address, as it is used by the NCP5623C variant" * tag 'leds-next-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/lee/leds: dt-bindings: leds: ncp5623: Add 0x39 as a valid I2C address Revert "leds: trigger: netdev: Configure LED blink interval for HW offload" leds: pca955x: Avoid potential overflow when filling default_label (take 2) leds: Unexport of_led_get() leds: tps6131x: Add V4L2_FLASH_LED_CLASS dependency dt-bindings: leds: lp50xx: Document child reg, fix example leds: leds-lp50xx: Handle reg to get correct multi_index leds: led-class-flash:: Fix flash_timeout comment MAINTAINERS: Adjust file entry in TPS6131X FLASH LED DRIVER leds: flash: leds-qcom-flash: Fix registry access after re-bind

Merge tag 'locking-core-2025-07-29' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

2025-07-30T01:11:32Z

Pull locking updates from Ingo Molnar: "Locking primitives: - Mark devm_mutex_init() as __must_check and fix drivers that didn't check the return code (Thomas Weißschuh) - Reorganize to better expose the internal APIs to local variables (Sebastian Andrzej Siewior) - Remove OWNER_SPINNABLE in rwsem (Jinliang Zheng) - Remove redundant #ifdefs in the mutex code (Ran Xiaokai) Lockdep: - Avoid returning struct in lock_stats() (Arnd Bergmann) - Change `static const` into enum for LOCKF_*_IRQ_* (Arnd Bergmann) - Temporarily use synchronize_rcu_expedited() in lockdep_unregister_key() to speed things up. (Breno Leitao) Rust runtime: - Add #[must_use] to Lock::try_lock() (Jason Devers)" * tag 'locking-core-2025-07-29' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: lockdep: Speed up lockdep_unregister_key() with expedited RCU synchronization locking/mutex: Remove redundant #ifdefs locking/lockdep: Change 'static const' variables to enum values locking/lockdep: Avoid struct return in lock_stats() locking/rwsem: Use OWNER_NONSPINNABLE directly instead of OWNER_SPINNABLE rust: sync: Add #[must_use] to Lock::try_lock() locking/mutex: Mark devm_mutex_init() as __must_check leds: lp8860: Check return value of devm_mutex_init() spi: spi-nxp-fspi: Check return value of devm_mutex_init() local_lock: Move this_cpu_ptr() notation from internal to main header

Revert "leds: trigger: netdev: Configure LED blink interval for HW offload"

2025-07-18T13:26:32Z

This reverts commit c629c972b310af41e9e072febb6dae9a299edde6. While .led_blink_set() would previously put an LED into an unconditional permanently blinking state, the offending commit now uses same operation to (also?) set the blink timing of the netdev trigger when offloading. This breaks many if not all of the existing PHY drivers which offer offloading LED operations, as those drivers would just put the LED into blinking state after .led_blink_set() has been called. Unfortunately the change even made it into stable kernels for unknown reasons, so it should be reverted there as well. Fixes: c629c972b310a ("leds: trigger: netdev: Configure LED blink interval for HW offload") Link: https://lore.kernel.org/linux-leds/c6134e26-2e45-4121-aa15-58aaef327201@lunn.ch/T/#m9d6fe81bbcb273e59f12bbedbd633edd32118387 Suggested-by: Andrew Lunn Cc: stable@vger.kernel.org Signed-off-by: Daniel Golle Reviewed-by: Andrew Lunn Link: https://lore.kernel.org/r/6dcc77ee1c9676891d6250d8994850f521426a0f.1752334655.git.daniel@makrotopia.org Signed-off-by: Lee Jones

leds: lp8860: Check return value of devm_mutex_init()

2025-07-11T22:11:19Z

devm_mutex_init() can fail. With CONFIG_DEBUG_MUTEXES=y the mutex will be marked as unusable and trigger errors on usage. Add the missed check. Fixes: 87a59548af95 ("leds: lp8860: Use new mutex guards to cleanup function exits") Signed-off-by: Thomas Weißschuh Acked-by: Andrew Davis Acked-by: Lee Jones Signed-off-by: Boqun Feng Link: https://lore.kernel.org/r/20250617-must_check-devm_mutex_init-v7-2-d9e449f4d224@weissschuh.net

leds: pca955x: Avoid potential overflow when filling default_label (take 2)

2025-07-02T16:18:44Z

GCC compiler v8.5.0 is not happy about printing into a too short buffer (when build with `make W=1`): drivers/leds/leds-pca955x.c:696:5: note: 'snprintf' output between 2 and 11 bytes into a destination of size 8 Unfortunately this is a false positive from the old GCC versions, but we may still improve the code by using '%hhu' format specifier and reduce buffer size by 4 bytes. Fixes: bd3d14932923 ("leds: pca955x: Avoid potential overflow when filling default_label") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202506282159.TXfvorYl-lkp@intel.com/ Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20250630093906.1715800-1-andriy.shevchenko@linux.intel.com Signed-off-by: Lee Jones

leds: Unexport of_led_get()

2025-07-02T16:18:01Z

There are no users outside the module. Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20250630092639.1574860-1-andriy.shevchenko@linux.intel.com Signed-off-by: Lee Jones

leds: tps6131x: Add V4L2_FLASH_LED_CLASS dependency

2025-07-01T13:26:03Z

This driver can optionally use the v4l2_flash infrastructure, but fails to link built=in if that is in a loadable module: ld.lld-21: error: undefined symbol: v4l2_flash_release >>> referenced by leds-tps6131x.c:792 (drivers/leds/flash/leds-tps6131x.c:792) Add the usual Kconfig dependency for it, still allowing it to be built when CONFIG_V4L2_FLASH_LED_CLASS is disabled. Fixes: b338a2ae9b31 ("leds: tps6131x: Add support for Texas Instruments TPS6131X flash LED driver") Signed-off-by: Arnd Bergmann Reviewed-by: Randy Dunlap Tested-by: Randy Dunlap Link: https://lore.kernel.org/r/20250620114440.4080938-1-arnd@kernel.org Signed-off-by: Lee Jones

leds: leds-lp50xx: Handle reg to get correct multi_index

2025-06-27T12:58:31Z

mc_subled used for multi_index needs well defined array indexes, to guarantee the desired result, use reg for that. If devicetree child nodes is processed in random or reverse order you may end up with multi_index "blue green red" instead of the expected "red green blue". If user space apps uses multi_index to deduce how to control the leds they would most likely be broken without this patch if devicetree processing is reversed (which it appears to be). arch/arm/boot/dts/aspeed/aspeed-bmc-facebook-fuji.dts has reg set but I don't see how it can have worked without this change. If reg is not set, an error is returned, If reg is out of range, an error is returned. reg within led child nodes starts with 0, to map to the iout in each bank. Signed-off-by: Johan Adolfsson Reviewed-by: Jacek Anaszewski Link: https://lore.kernel.org/r/20250617-led-fix-v7-1-cdbe8efc88fa@axis.com Signed-off-by: Lee Jones

sysfs: treewide: switch back to attribute_group::bin_attrs

2025-06-17T08:44:15Z

The normal bin_attrs field can now handle const pointers. This makes the _new variant unnecessary. Switch all users back. Signed-off-by: Thomas Weißschuh Link: https://lore.kernel.org/r/20250530-sysfs-const-bin_attr-final-v3-4-724bfcf05b99@weissschuh.net Signed-off-by: Greg Kroah-Hartman