Re: [6.12.107] MSI G63 9SEK - Synaptics touchpad falls back to PS/2, multitouch lost at boot, fixed by delayed protocol write

On Sun, Sep 13, 2026 at 05:23:43AM +0000, Cameron Spencer wrote:

I experienced an issue where my Synaptics touchpad is registered as a 
basic PS/2 device at boot, losing all multitouch functionality 
(two-finger scroll, gestures). Writing "synaptics" to the sysfs protocol 
file a few seconds after boot restores the SynPS/2 extended protocol and 
full functionality.

System:

- Kernel: 6.12.107+deb13-amd64

Hi,

Thanks for the report, could you see if the issue is reproducible in mainline
first?

     $ dmesg | grep -iE 'psmouse|synaptics|serio|i8042'
     [    1.006462] i8042: PNP: PS/2 Controller 
[PNP0303:PS2K,PNP0f13:PS2M] at 0x60,0x64 irq 1,12
     [    1.010384] serio: i8042 KBD port at 0x60,0x64 irq 1
     [    1.010392] serio: i8042 AUX port at 0x60,0x64 irq 12
     [    1.018706] input: AT Translated Set 2 keyboard as 
/devices/platform/i8042/serio0/input/input0
     # psmouse probes at boot, fails the Synaptics handshake, registers 
as plain "PS/2 Synaptics TouchPad" (no multitouch)
     [    2.384949] input: PS/2 Synaptics TouchPad as 
/devices/platform/i8042/serio1/input/input6
     # psmouse-reload service runs after systemd-udev-settle and writes 
synaptics to /sys/bus/serio/devices/serio1/protocol, then re-probes with 
the Synaptics extended protocol, queries coordinates, and registers as 
"SynPS/2 Synaptics TouchPad" (multitouch restored)
     [    8.654815] psmouse serio1: synaptics: queried max coordinates: 
x [..5656], y [..4728]
     [    8.683422] psmouse serio1: synaptics: queried min coordinates: 
x [1284..], y [1128..]
     [    8.739382] psmouse serio1: synaptics: Touchpad model: 1, fw: 
8.2, id: 0x1e2b1, caps: 0xf00323/0x840300/0x26800/0x0, board id: 3163, 
fw id: 2286728
     [    8.775128] input: SynPS/2 Synaptics TouchPad as 
/devices/platform/i8042/serio1/input/input21

Seems like firmware is unable to answer extended queries for some time
after initialization and just falls back to PS/2,

perhaps we could just add a retry loop around
synaptics_query_hardware() to give firmware a little extra time.

Thanks,
Lovekesh