linux-toradex.git/drivers/hid/intel-thc-hid/intel-quicki2c, branch master Linux kernel for Apalis and Colibri modules HID: Intel-thc-hid: Intel-quicki2c: Add NVL Device IDs 2026-03-17T22:59:50+00:00 Even Xu even.xu@intel.com 2026-03-17T05:56:28+00:00 d56c9cca2e7f8f0ad91f37abd4e6357c2318b0db Add Nova Lake THC QuickI2C device IDs to support list. Signed-off-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Add Nova Lake THC QuickI2C device IDs to support list.

Signed-off-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: intel-thc-hid: Set HID_PHYS with PCI BDF 2026-03-16T10:25:07+00:00 Daniel Schaefer git@danielschaefer.me 2026-03-13T13:39:25+00:00 4bc7bc457922742d38915458e630195e761c1efd Currently HID_PHYS is empty, which means userspace tools (e.g. fwupd) that depend on it for distinguishing the devices, are unable to do so. Other drivers like i2c-hid, usbhid, surface-hid, all populate it. With this change it's set to, for example: HID_PHYS=0000:00:10.0 Each function has just a single HID device, as far as I can tell, so there is no need to add a suffix. Tested with fwupd 2.1.1, can avoid https://github.com/fwupd/fwupd/pull/9995 Cc: Even Xu <even.xu@intel.com> Cc: Xinpeng Sun <xinpeng.sun@intel.com> Cc: Jiri Kosina <jikos@kernel.org> Cc: Benjamin Tissoires <bentiss@kernel.org> Cc: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Daniel Schaefer <git@danielschaefer.me> Reviewed-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Currently HID_PHYS is empty, which means userspace tools (e.g. fwupd)
that depend on it for distinguishing the devices, are unable to do so.
Other drivers like i2c-hid, usbhid, surface-hid, all populate it.

With this change it's set to, for example: HID_PHYS=0000:00:10.0

Each function has just a single HID device, as far as I can tell, so
there is no need to add a suffix.

Tested with fwupd 2.1.1, can avoid https://github.com/fwupd/fwupd/pull/9995

Cc: Even Xu <even.xu@intel.com>
Cc: Xinpeng Sun <xinpeng.sun@intel.com>
Cc: Jiri Kosina <jikos@kernel.org>
Cc: Benjamin Tissoires <bentiss@kernel.org>
Cc: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Daniel Schaefer <git@danielschaefer.me>
Reviewed-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: Intel-thc-hid: Intel-quicki2c: Add power management for touch device 2026-01-07T21:04:37+00:00 Even Xu even.xu@intel.com 2025-12-22T01:41:34+00:00 6761f9ccd6247682430413274e7af955d39cfee4 Send POWER_SLEEP command to touch device when system enters into sleep mode or hibernate mode to save more power; and send POWER_ON command to take device back when system exits sleep mode. Signed-off-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Send POWER_SLEEP command to touch device when system enters into sleep
mode or hibernate mode to save more power; and send POWER_ON command to
take device back when system exits sleep mode.

Signed-off-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: Intel-thc-hid: Intel-quicki2c: Add output report support 2026-01-07T21:03:47+00:00 Even Xu even.xu@intel.com 2025-12-09T07:52:15+00:00 a788b2057029d831ee610c064bb607984a00367d Add support for HID output reports in the intel-quicki2c driver by implementing the output_report callback in the HID low-level driver interface. This enables proper communication with HID devices that require output report functionality, such as setting device configuration or updating device firmware. Tested-by: Rui Zhang <rui1.zhang@intel.com> Signed-off-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Add support for HID output reports in the intel-quicki2c driver by
implementing the output_report callback in the HID low-level driver
interface.

This enables proper communication with HID devices that require
output report functionality, such as setting device configuration or
updating device firmware.

Tested-by: Rui Zhang <rui1.zhang@intel.com>
Signed-off-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: Intel-thc-hid: Intel-quicki2c: Support writing output report format 2026-01-07T21:03:47+00:00 Even Xu even.xu@intel.com 2025-12-09T07:52:14+00:00 e149af9ce108d7b9e39d731b93c9fb55fd9e7842 There are two output formats requested in the HID-over-I2C specification: - Command format (set feature/set report): encoded command written to command register, followed by data written to data register - Output report format: all data written directly to output register Current quicki2c_init_write_buf() implementation only supports the command format. Extend quicki2c_init_write_buf() to automatically detect the output format based on the presence of command parameters and prepare the appropriate output buffer accordingly. Tested-by: Rui Zhang <rui1.zhang@intel.com> Signed-off-by: Even Xu <even.xu@intel.com> Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
There are two output formats requested in the HID-over-I2C specification:
- Command format (set feature/set report): encoded command written to
  command register, followed by data written to data register
- Output report format: all data written directly to output register

Current quicki2c_init_write_buf() implementation only supports the
command format.

Extend quicki2c_init_write_buf() to automatically detect the output
format based on the presence of command parameters and prepare the
appropriate output buffer accordingly.

Tested-by: Rui Zhang <rui1.zhang@intel.com>
Signed-off-by: Even Xu <even.xu@intel.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: Intel-thc-hid: Intel-quicki2c: Use put_unaligned_le16 for __le16 writes 2026-01-07T21:03:47+00:00 Even Xu even.xu@intel.com 2025-12-09T07:52:13+00:00 678cbaa24f09115352be669884f7775ccd28f757 Replace memcpy operations with put_unaligned_le16() when writing 16-bit little-endian values to the write buffer. This change improves code clarity and ensures proper handling of unaligned memory access. Signed-off-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Replace memcpy operations with put_unaligned_le16() when writing 16-bit
little-endian values to the write buffer.

This change improves code clarity and ensures proper handling of unaligned
memory access.

Signed-off-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: Intel-thc-hid: Intel-quicki2c: Use size_t for all length variables 2026-01-07T21:03:47+00:00 Even Xu even.xu@intel.com 2025-12-09T07:52:12+00:00 dabc153c79890911b23c93c1dd65e15b53e87740 Convert all length-related variables in the QuickI2C protocol layer to use size_t type to follow kernel coding conventions. This includes: - All buffer length parameters and variables - Return values of quicki2c_encode_cmd() function which represents encoded command buffer size. - Return values of quicki2c_init_write_buf() function which represents process result: either prepared output buffer size or error code. This change improves type consistency and aligns with standard kernel practices for memory size representation, reducing potential issues with size calculations and comparisons. Signed-off-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
Convert all length-related variables in the QuickI2C protocol layer to use
size_t type to follow kernel coding conventions.

This includes:
- All buffer length parameters and variables
- Return values of quicki2c_encode_cmd() function which represents
  encoded command buffer size.
- Return values of quicki2c_init_write_buf() function which represents
  process result: either prepared output buffer size or error code.

This change improves type consistency and aligns with standard kernel
practices for memory size representation, reducing potential issues
with size calculations and comparisons.

Signed-off-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: intel-thc-hid: Remove redundant pm_runtime_mark_last_busy() calls 2025-10-30T10:27:17+00:00 Sakari Ailus sakari.ailus@linux.intel.com 2025-10-27T12:01:23+00:00 c430f56be32ec098999158717d04e64da5edce06 pm_runtime_put_autosuspend(), pm_runtime_put_sync_autosuspend(), pm_runtime_autosuspend() and pm_request_autosuspend() now include a call to pm_runtime_mark_last_busy(). Remove the now-reduntant explicit call to pm_runtime_mark_last_busy(). Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com> Reviewed-by: Even Xu <even.xu@intel.com> Signed-off-by: Jiri Kosina <jkosina@suse.com> 
pm_runtime_put_autosuspend(), pm_runtime_put_sync_autosuspend(),
pm_runtime_autosuspend() and pm_request_autosuspend() now include a call
to pm_runtime_mark_last_busy(). Remove the now-reduntant explicit call to
pm_runtime_mark_last_busy().

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Even Xu <even.xu@intel.com>
Signed-off-by: Jiri Kosina <jkosina@suse.com>
HID: intel-thc-hid: intel-quicki2c: Fix wrong type casting 2025-10-14T09:27:01+00:00 Xinpeng Sun xinpeng.sun@intel.com 2025-10-09T03:31:08+00:00 6c26c055523d915afb8d18e7277848eff66a3085 The type definition of qcdev->i2c_max_frame_size is already u32, so remove the unnecessary type casting le16_to_cpu. Signed-off-by: Xinpeng Sun <xinpeng.sun@intel.com> Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202509280841.pxmgBzKW-lkp@intel.com/ Signed-off-by: Jiri Kosina <jkosina@suse.com> 
The type definition of qcdev->i2c_max_frame_size is already
u32, so remove the unnecessary type casting le16_to_cpu.

Signed-off-by: Xinpeng Sun <xinpeng.sun@intel.com>
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202509280841.pxmgBzKW-lkp@intel.com/
Signed-off-by: Jiri Kosina <jkosina@suse.com>
Merge branch 'for-6.18/intel-thc-hid' into for-linus 2025-09-30T14:58:07+00:00 Benjamin Tissoires bentiss@kernel.org 2025-09-30T14:58:07+00:00 cd3770675874ae72e4d26dd34f4e02dfc7ed8067 - quicki2c: support ACPI config for advanced features: max input size and interrupt delay (Xinpeng Sun) - some more str_true_false() conversions 
- quicki2c: support ACPI config for advanced features: max input size
  and interrupt delay (Xinpeng Sun)
- some more str_true_false() conversions