Opened 3 years ago
Closed 2 years ago
#17806 closed bug (fixed)
keyboard is reset during ps2 active multiplexing setup
Reported by: | pulkomandy | Owned by: | |
---|---|---|---|
Priority: | normal | Milestone: | R1/beta4 |
Component: | Drivers/Input/PS2 | Version: | R1/beta3 |
Keywords: | Cc: | ||
Blocked By: | Blocking: | ||
Platform: | All |
Description
On my machine, there is a PS2 controller with nothing wired to the the AUX port (the touchpad and trackpoint are both on I2C).
The PS2 controller appears to support active multiplexing. So we send command trying to configure it and publish the corresponding devices (even if nothing is connected to the ports).
The commands sent to set the active multiplexing seem to be disturbing the keyboard, which sends several FE responses (asking to re-send the previous command) and eventually resets (sending FA AA).
On Linux, they detect that the AUX port is not used from DMI data, as far as can see. The DMI data gives the interrupt that should be used, and if that field is not present, they disable the mouse port and never try to enable the active multiplexing. Which avoids the issue, of course.
I think this keyboard reset causes the typematic rate setting to be reinitialized. As a result, I have to go to input preferences and slightly move the slider to get my key repeat settings applied after every boot.
I don't know if a lot of people use non-default repeat settings, so possibly there are more machines affected, but no one noticed until now. So I would like confirmation: does the setting work at boot for other people? When/if it doesn't, does the PS2 driver log anything interesting in the syslog?
Change History (2)
comment:1 by , 3 years ago
comment:2 by , 2 years ago
Milestone: | Unscheduled → R1/beta4 |
---|---|
Resolution: | → fixed |
Status: | new → closed |
Fix merged in hrev56268.
It could be that the touchpad can be switched to PS2 in the BIOS. I have one laptop with a hidden setting for this in the BIOS.