F20 gnome inactivity causes mouse/kbd unresponsiveness, must restart gnome-shell

[Pekka Savola]

As an update:

Removing the additional input source and downgrading from 
updates-testing to updates did not help. Cron is also not a culprit.

On Wed, 12 Feb 2014, Pekka Savola wrote:

> Hello,
>
> Thanks for ideas; I'll need to check at least reverting to updates and 
> checking the input source and report back tomorrow.
>
> Combining replies:
>
> On Wed, 12 Feb 2014, g wrote:
>> >   After inactivity timeout, you need to swipe the screen or use keyboard
>> >   to reactivate it (I've disabled screen locking). The problem that
>> >   often occurs is that mouse movement works, but clicking doesn't do
>> >   anything. Similarly keyboard is unresponsibe. The clock has been
>> >   frozen to the inactivity time.
>>
>>  "problem that often occurs"
>
> It depends. On my desktop, I get it 1-2 times a week, so relatively rarely. 
> On my laptop, I get it every day. After rebooting (or logging in, I don't 
> remember which), for a few hours (5-10 inactivity switches) it works OK every 
> time. After that point, every inactivity seems to trigger the problem.
>
>>  are there other problems?
>
> Not ones I think are related.
>
> On laptop, the screen gets garbled sometimes e.g. when restarting gnome-shell 
> and switching between terminals. The same restart or suspend/resume often 
> fixes it.
>
> On desktop, firefox consumes 99% of CPU if I have facebook maximized (this 
> has occurred for at least a month now) and some other times it runs havoc in 
> any case.
>
>>  does this problem not always happen?
>
> See above. On laptop after a while, it appears to always happen after it has 
> been triggered the first time. Because it's more easily reproducible there, 
> I've focused on that (it isn't with me at the moment).
>
>>  is there anything special that you have on both systems that
>>  relates to keyboard and/or mouse?
>
> Not really. Internal devices on laptop, USB on desktop.
>
>>  ie, being that i see you are from "fi" land, i wonder if by chance
>>  you are using 'multi language' on your systems.
>>
>>  if so, can/have you create another user as "en" only to see if
>>  problem occurs?
>
> I suppose you mean multiple Input Sources in Settings - Region/Language. On 
> desktop, just one. On laptop, I might have had two (at least initially after 
> install I did). I think I removed it. I'll need to check tomorrow.
>
>>  have you tired/considered setting up one of the systems without
>>  using "updates-testing"?
>
> It has been a habit to use updates-testing (I'm a RHL user for 15 years now, 
> so I've usually lived a bit on the edge). I could use just updates. I thought 
> "downgrading" would be impossible, but after googling, it seems distro-sync 
> back should do it. I'll try that.
>
> On Wed, 12 Feb 2014, Tom Horsley wrote:
>>  Are the periods of inactivity long enough for cron to have run
>>  while inactive? I'm so leery of this bug now, that I think
>>  almost anything strange could be another symptom:
>>
>>  https://bugzilla.redhat.com/show_bug.cgi?id=1043212
>>
>>  (The title of that bug is kind of misleading, it refers
>>  to one of about a zillion symptoms of the same problem).
>
> Thanks. I don't see any nologin issues at least, though sometimes I've seen 
> notifications of failed logins which I have found suspect but haven't 
> investigated.
>
> The inactivity times (esp. with laptop, this is where I've mostly seen this) 
> have typically been quite short. I think even less than a minute. So I 
> suppose it's something else.
>
>

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings