Merge tag 'hid-for-linus-2026081901' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid

Pull HID updates from Jiri Kosina:
"Core:
- fix long-standing force-feedback initialization race acr

Merge tag 'hid-for-linus-2026081901' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid

Pull HID updates from Jiri Kosina:
"Core:
- fix long-standing force-feedback initialization race across the
subsystem (Dmitry Torokhov)
- switch to system_dfl_wq (Marco Crivellari)

AMD-SFH:
- support for tablet-mode switch for AMD SFH-based systems (Basavaraj
Natikar)

HyperX:
- support for HyperX QuadCast 2 (Benjamin Blume)

I2C-HID:
- support for devices that provide HID descriptor solely through
the ACPI _DSM method (XIE Zhibang)

Intel-THC-HID:
- support for full I2C bus config parameters (Even Xu)

Logitech:
- HID++ 2.0 repogrammable button support (Elliot Douglas)
- Bolt receiver support for HID++ devices (Erik Håkansson)

MSI:
- support for MSI Claw (Derek J. Clark)

Steam:
- initial support for 2026 Steam Controller (Vicki Pfau)
- support for sensor events on the 2025 Steam Controller (Vicki Pfau)

And many, many other fixes for various long standing issues that were
found by new modern tools, and quite a few device ID additions"

* tag 'hid-for-linus-2026081901' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid: (146 commits)
HID: tmff: Use 64-bit arithmetic for force feedback scaling
HID: multitouch: reclassify HTIX5288 to WIN_8_FORCE_MULTI_INPUT_NSMU
HID: sensor: custom: Fix field sysfs group cleanup on failure
HID: sensor: custom: Fix use-after-free in enable_sensor
HID: intel-thc-hid: intel-quickspi: bound GET_REPORT response to the caller buffer
HID: haptic: don't write an uninitialized value to unhandled usages
HID: intel-thc-hid: intel-quickspi: fix autosuspend cleanup during teardown
HID: intel-thc-hid: intel-quicki2c: fix autosuspend cleanup during teardown
HID: steam: Zero out inputs when disabling gamepad mode
HID: steam: Clean up locking
HID: steam: Don't set feature reports when disconnecting
HID: steam: Fix wording of connect/disconnect logs
HID: steam: Initial 2026 Steam Controller support
HID: steam: Refactor registration
HID: logitech: add Bolt receiver support for Logitech HID++ devices
HID: sensor-hub: Fix out-of-bounds write in sensor_hub_get_feature
HID: universal-pidff: stop the device when force-feedback init fails
HID: haptic: move FF initialization into .input_configured()
HID: logitech-hidpp: move FF initialization to .input_configured()
HID: megaworld: move FF initialization to .input_configured()
...

show more ...

Merge branch 'for-7.3/hyperx' into for-linus

- support for HyperX QuadCast 2 (Benjamin Blume)

HID: hyperx: add driver for the HyperX QuadCast 2 mute button

The tap-to-mute button of the HyperX QuadCast 2 (03f0:07b4) is handled
entirely in the device firmware. The firmware gates the audio in

HID: hyperx: add driver for the HyperX QuadCast 2 mute button

The tap-to-mute button of the HyperX QuadCast 2 (03f0:07b4) is handled
entirely in the device firmware. The firmware gates the audio internally
but never sends the Telephony "Phone Mute" usage (0x2f) that its own report
descriptor advertises, and it does not touch the UAC feature unit either.
Consequently neither an evdev key event nor an ALSA mixer change is ever
generated, and userspace has no way to learn that the microphone was muted:
desktops keep showing the microphone as live, and conferencing applications
keep displaying an unmuted microphone while transmitting silence.

The mute state is reported through a vendor-defined collection instead:

06 c0 ff Usage Page (Vendor-Defined 0xFFC0)
a1 01 Collection (Application)
06 c1 ff Usage Page (Vendor-Defined 0xFFC1)
85 77 Report ID (0x77)
09 f0 Usage (0xF0)
75 08 95 3f Report Size (8), Report Count (63)
81 02 Input (Data,Var,Abs)

Pressing the button emits a 64-byte report on that collection:

77 06 00 00 ... microphone unmuted
77 06 01 00 ... microphone muted

where byte 1 identifies the mute event and byte 2 carries the resulting
state. As the payload is an opaque vendor blob carrying no HID usages,
hid-input cannot map it and a hwdb entry cannot express it either.

Add a driver that decodes the report and emits KEY_MICMUTE, which makes the
button behave like any other microphone mute key.

Note that the device reports the resulting absolute state, whereas
KEY_MICMUTE is a momentary key that userspace acts on as a toggle, so the
driver emits one keypress per state change.

Tested on a HyperX QuadCast 2 (03f0:07b4).

Signed-off-by: Benjamin Blume <benjaminblume@posteo.de>
Signed-off-by: Jiri Kosina <jkosina@suse.com>

show more ...