Thanks for reporting.

Could you try setting udev_log="err" in /etc/udev/udev.conf and try again? Also change rc.sysinit to remove the "--quiet" after "udevadm settle".

Also could you look in /dev/input and tell me what is missing with 168-2.

Please attach the output of "udevadm info --query=all --name=/dev/input/mouse0" with udev 168-1 (assuming mouse0 is the one missing from 168-2).

Hi Tom, some questions and data :
quote "Could you try setting udev_log="err" in /etc/udev/udev.conf and try again? Also change rc.sysinit to remove the "--quiet" after "udevadm settle"."
I did that now, question , what output shall I attach now , dmesg ?
dmesg_udev_168_2_with_all_debug : http://pastie.org/private/55vjywfhukyikk7yypyu8a
_ls_of_dev_input_168_2_without_replug : http://pastie.org/private/jolmhbxebaarkss0batza
_udevadm_of_168_2_without_replug : http://pastie.org/private/qcratm8mnwk8q1cqjc8gw ( NULL)
_cat_of_var_log_errors_boot_with_Debug_168_2 : http://pastie.org/private/1vhtitxjwys0fotkzwnkg

-with 168-1

_udevadm_of_168_1: http://pastie.org/private/7ie2tb5wmnucyaxwv7qg (udevadm info --query=all --name=/dev/input/mouse0)
ls /dev/input shows mouse0 , but will check again now ..

What else shall I provide ?
I'm using custom kernel 2.6.38.6 just as a reminder, would have to move around nvidia module to test other orig kernel
Thx for sticking with me hope nothing machine specific
One more thing might it be librazer ? This is a "Razer Deathadder "mouse , which has it's own daemon "razerd" in this package : AUR/razercfg-git, however no updates for last 6 weeks in that git ...

Steffen,

I'm at a bit of a loss to be honest. I cannot imagine why this change would cause your mouse not to be detected at all...

What happens if you do "udevadm trigger --action=add" after booting with udev-168-2? Does the mouse appear? What if you follow it with "udevadm settle"?

Thanks or your reporting so far.

I had a look at the AUR package. Is razerd in your MODULES array? Could you give me the output of lsmod in both the working and non-working state. Also, when you replug your mouse, does that make the output of lsmod the same as in the working case?

I still have to test udevadm trigger --action=add , but I know for the first time experienced the same missing device with 168-1 to which i downgraded. But I never had that before, so I will try to run this ( udev 167-2) some boots and see whether it appears again. The razerd is a daemon not a module and just runs in DAEMONS=() , it enables me to talk to the mouse, turn off/on all the leds etc.

Btw, I just uploaded a new udev to testing (169), so please compare 168-1 and 169-1, if you are going to test more :-)

I just experienced this fault with udev 167-2 , it can't be a udev issue , it must be something kernel related, still investigating , one guy here , https://bbs.archlinux.org/viewtopic.php?pid=936487#p936487, ubuntu : https://bugs.launchpad.net/ubuntu/+source/linux/+bug/775498
still investigating , nothing much in the web ...

@tuxflusion: I'm assuming you meant "udev-167-1"? Can you upgrade to the latest udev (170), and try to show that the problem is not present on older kernels, just to make sure it is not udev related? I have anyway assigned this bug to the kernel (but I'll follow it too in case it turns out to be udev after all).

With udve-170 you should get output in case of errors, this might help. If it does not, you could try setting udev_log=warn or even info, to find out more.

Have you remembered to rebuild your aur package? Could you try without the aur package at all, would the mouse still work? I would like to exclude that package from the equation as I don't have a clue about what it does (and don't want to wade through code unless I absolutely must ;-) ).

@tomegun, I've disabled the razerd daemon just to be sure. I really don't know what might have been the problem but I can't recreate it know ( I created a complete new system user and migrated my whole data to fix multiple kde issues ) . It must have been something kernel related I think i disabled kernel debug , maybe some exception didn't show up. I got way less device disconnect/reconnect msg in dmesg now, I also disabled OHCI. Sorry for all the noise and thanks for sticking with me.

Good to hear that it seems to have worked out :-)

I'm closing, please reopen if it resurfaces.