[PATCH v3 0/4] Fix soc-button-array debounce

[Mario Limonciello <superm1@kernel.org>]

From: Mario Limonciello <mario.limonciello@amd.com>

I have some hardware in front of me that uses the soc-button-array
driver but the power button doesn't work.

Digging into it, it's because the ASL prescribes a debounce of 0 for
the power button, but the soc-button-array driver hardcodes 50ms.

Hardcoding it to what the ASL expects the power button works.

I looked at the callpath into the GPIO core and I believe it's
because the debounce value from _CRS is never programmed to the
hardware the way that the GPIO gets setup.

This series add that programming path and then sets the hardcoded
value on on some quirked systems.  Hopefully Hans can confirm this
continues to work on the hardware that he originally developed the
hardcoding for.

---
v3:
 * Use Hans suggestion to force software debounce instead of a quirk
 * Fix a resume issue identified from testing this series

Mario Limonciello (4):
  gpiolib: acpi: Add a helper for programming debounce
  gpiolib: acpi: Program debounce when finding GPIO
  Input: Don't program hw debounce for soc_button_array devices
  Input: Don't send fake button presses to wake system

 drivers/gpio/gpiolib-acpi-core.c      | 25 ++++++++++++++-----------
 drivers/input/keyboard/gpio_keys.c    | 14 ++++++--------
 drivers/input/misc/soc_button_array.c |  1 +
 include/linux/gpio_keys.h             |  2 ++
 4 files changed, 23 insertions(+), 19 deletions(-)

-- 
2.43.0