I have seen several posts on setting being changed or lost after a wakeup from hibernation.
This inspired me to look at the udev rules (as koki has been asleep for 48 hours, another wierd dups or no show problem like severs gui. Many things sort of out of sorts.
After a bit up reading (my right eye is actually injured from this alchemy but when one gets the wireless for a few days, one makes hay) I have a couple of theories.
1. The udev rules have different state, default, persistent, restore. (no, I don't know how or when but I think it must be systemd talking to udev after a wakeup event.)
So the alias sound, the Norwegian keyboard, and the tablet with the clicker turned oof (I just made up the last one) all get restored to default or some set of persistantcy.
2. Usb - the kernel log is full of usb bits being set and the bug was discovered that the time to set the bits was unknow before the event (i.e. sleep.) so it was allowed a time of 1ms.evidently it needs more.
Also some (echi?) controllers can not be slept during hibernate, so lan wakeup signals are lost, but they wouldn't work anyway.
Is it componded, sleeping to a usb drive with an older kernel?
why not? Some of the stuff is going to be pushed back to 2.6.32.
So the best test is to update to the lastest kernel 3.4.4-5.
Try the stuff above and file bugs if they are still failing.
So the question is, am I on base about udev rules or standing outside the park listing to crowd echos.
Any comments about udev rules.
SJ
Sleep and udev rules - Lang, keyborads, sound, mice.
Linear Mode
