kernel/drivers/input/keyboard/mtk-pmic-keys.c, branch linux-6.18.y

Input: mtk-pmic-keys - MT6359 has a specific release irq

2025-09-06T16:24:14Z

Support for MT6359 PMIC keys has been added recently. However, the key release event is not properly handled: only key press events are generated, leaving key states stuck in "pressed". This patch ensures that both key press and key release events are properly emitted by handling the release logic correctly. Introduce a 'key_release_irq' member to the 'mtk_pmic_regs' to identify the devices that have a separate irq for the release event. Fixes: bc25e6bf032e ("Input: mtk-pmic-keys - add support for MT6359 PMIC keys") Signed-off-by: Julien Massot Link: https://lore.kernel.org/r/20250905-radxa-nio-12-l-gpio-v3-1-40f11377fb55@collabora.com Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - add support for MT6359 PMIC keys

2025-07-03T23:34:10Z

Add PMIC key support on MT6359 SoC. Signed-off-by: Louis-Alexis Eyraud Link: https://lore.kernel.org/r/20250703-add-mt6359-pmic-keys-support-v1-1-21a4d2774e34@collabora.com Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - fix possible null pointer dereference

2025-04-29T16:27:49Z

In mtk_pmic_keys_probe, the regs parameter is only set if the button is parsed in the device tree. However, on hardware where the button is left floating, that node will most likely be removed not to enable that input. In that case the code will try to dereference a null pointer. Let's use the regs struct instead as it is defined for all supported platforms. Note that it is ok setting the key reg even if that latter is disabled as the interrupt won't be enabled anyway. Fixes: b581acb49aec ("Input: mtk-pmic-keys - transfer per-key bit in mtk_pmic_keys_regs") Signed-off-by: Gary Bisson Cc: stable@vger.kernel.org Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - switch to for_each_child_of_node_scoped

2024-10-20T04:21:37Z

Use the scoped variant of the macro to simplify the code and error handling. This makes the error handling more robust by ensuring that the child node is always freed. Signed-off-by: Javier Carrasco Link: https://lore.kernel.org/r/20241010-input_automate_of_node_put-v1-2-ebc62138fbf8@gmail.com Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys: add MT6357 support

2023-01-19T17:25:44Z

Add PMIC Keys support on MT6357 SoC. Signed-off-by: Fabien Parent Reviewed-by: AngeloGioacchino Del Regno Reviewed-by: Mattijs Korpershoek Acked-by: Dmitry Torokhov Signed-off-by: Alexandre Mergnat Link: https://lore.kernel.org/r/20221005-mt6357-support-v8-6-560caaafee53@baylibre.com Signed-off-by: Matthias Brugger

Input: mtk-pmic-keys - switch to DEFINE_SIMPLE_DEV_PM_OPS() and pm_sleep_ptr()

2022-12-07T21:27:10Z

SIMPLE_DEV_PM_OPS() is deprecated as it requires explicit protection against unused function warnings. The new combination of pm_sleep_ptr() and DEFINE_SIMPLE_DEV_PM_OPS() allows the compiler to see the functions, thus suppressing the warning, but still allowing the unused code to be removed. Thus also drop the __maybe_unused markings. Signed-off-by: Jonathan Cameron Reviewed-by: Mattijs Korpershoek Link: https://lore.kernel.org/r/20221204180841.2211588-10-jic23@kernel.org Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - add support for MT6331 PMIC keys

2022-09-13T16:52:14Z

Add support for PMIC Keys of the MT6331 PMIC. Signed-off-by: AngeloGioacchino Del Regno Reviewed-by: Mattijs Korpershoek Link: https://lore.kernel.org/r/20220913123941.385349-1-angelogioacchino.delregno@collabora.com Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - move long press debounce mask to mtk_pmic_regs

2022-06-01T04:03:45Z

As the second and last step of preparation to add support for more PMICs in this driver, move the long press debounce mask to struct mtk_pmic_regs and use that in mtk_pmic_keys_lp_reset_setup() instead of directly using the definition. While at it, remove the definition for MTK_PMIC_RST_DU_SHIFT as we are able to calculate it dynamically and spares us some unnecessary new definitions around for future per-PMIC variations of RST_DU_MASK. Lastly, it was necessary to change the function signature of mtk_pmic_keys_lp_reset_setup() to now pass a pointer to the main mtk_pmic_regs structure, since that's where the reset debounce mask now resides. This commit brings no functional changes. Signed-off-by: AngeloGioacchino Del Regno Reviewed-by: Mattijs Korpershoek Link: https://lore.kernel.org/r/20220524093505.85438-3-angelogioacchino.delregno@collabora.com Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - transfer per-key bit in mtk_pmic_keys_regs

2022-06-01T04:03:44Z

Place the key bit in struct mtk_pmic_keys_regs to enhance this driver's flexibility, in preparation for adding support for more PMICs. While at it, remove the definition of MTK_PMIC_RST_KEY_MASK as we are now dynamically setting the keymask relatively to the keys that are defined in the newly added rst_en_mask variable, on a per-key basis. This commit brings no functional changes. Signed-off-by: AngeloGioacchino Del Regno Reviewed-by: Mattijs Korpershoek Link: https://lore.kernel.org/r/20220524093505.85438-2-angelogioacchino.delregno@collabora.com Signed-off-by: Dmitry Torokhov

Input: mtk-pmic-keys - use single update when configuring long press behavior

2022-06-01T04:03:42Z

Instead of doing 3 read-modify-write operations when configuring behavior of long-press, consolidate everything into one. Reviewed-by: Mattijs Korpershoek Tested-by: Mattijs Korpershoek # on mt8183-pumpkin Link: https://lore.kernel.org/r/YowH0Q5HAGU7Grx2@google.com Signed-off-by: Dmitry Torokhov