USB Mouse Turns Off and Won't Turn Back On?

TracyTiger · 05-06-2014, 12:45 PM

DELETED (Changed my mind)

K-Wizzz · 05-06-2014, 02:28 PM

Quote:

Originally Posted by gengisdave

I looked at laptop's /var/log/messages and disconnect are still there; every time it increments input number, after 8 uptime hours mouse is connected as input28, but always as /dev/input/mouse1;

it should be an x.org issue, the mouse reconnects successfly, but the grahpic driver won't use it; i have xf86-input-mouse-19.0 build 3; in that case, mouse won't work in x, but if you switch on a terminal (ctrl+alt+f1) you get the pointer (alt+f7 to return in x)

Hm, don't know if this helps, but I had a similar issue with my laptop and its touchpad that did not work properly after suspend (C-M-F1, followed by C-M-F7 restored regular mouse behaviour).

The information from http://www.linuxquestions.org/questi...on-4175494532/ solved my issue.

tomtomjkw · 05-07-2014, 09:29 AM

Glad to see I'm not alone
Have the same problem with my XBMC Slackware-current based machine and some cheap Logitech wireless keyboard/mouse combo. Some weird power management event disconnects it to the point that even unplug/plug sequence doesn't help. What helps is moving a mouse once in a while, which apparently stops the inactivity counter. I've never put any effort to the problem, but might try. The problem occured also on the older machine, which was based on XBMCBuntu distro, so we can safely rule out kernel or hardware issue. Therefore I agree with gengisdave's assumption - it's an x.org wrongdoing.

gengisdave · 08-11-2014, 07:21 AM

I know this is an old thread, but I haven't used any more my laptop, I studied the problem a bit more, because it came back

https://www.kernel.org/doc/Documenta...management.txt

I think the problem is related to the powersaving policies in the kernel: the usb device goes into idle mode, then the device (or the kernel module) chooses to disconnect himself rather than do a wakeup.

My trick is to disable the autosuspend just for the mouse through an udev rule.

Code:

#cat /etc/udev/rules.d/99-local.rules
SUBSYSTEM=="usb", ACTION=="add", ATTRS{idVendor}=="0461", ATTR{idProduct}=="4d03", TEST=="power/control", ATTR{power/control}="on"

(replace idVendor and idProduct with the correct ones, lsusb will show you which ones)

This will not prevent the device to disconnect, but in 90% of cases, the mouse is able to reconnect without problems. I will debug of USB to understand why disconnects so much (but I'm a bit lazy)