Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.debian.user > #207713 > unrolled thread

testing: weird resume (black or frozen screen)

Started byAndrea Borgia <andrea@borgia.bo.it>
First post2019-04-20 17:50 +0200
Last post2019-05-05 12:00 +0200
Articles 7 — 2 participants

Back to article view | Back to linux.debian.user


Contents

  testing: weird resume (black or frozen screen) Andrea Borgia <andrea@borgia.bo.it> - 2019-04-20 17:50 +0200
    Re: testing: weird resume (black or frozen screen) "hdv@gmail" <hdv.jadev@gmail.com> - 2019-04-20 18:20 +0200
      Re: testing: weird resume (black or frozen screen) Andrea Borgia <andrea@borgia.bo.it> - 2019-04-21 10:50 +0200
        Re: testing: weird resume (black or frozen screen) "hdv@gmail" <hdv.jadev@gmail.com> - 2019-04-21 13:50 +0200
          Re: testing: weird resume (black or frozen screen) Andrea Borgia <andrea@borgia.bo.it> - 2019-04-21 20:00 +0200
    Re: testing: weird resume (black or frozen screen) Andrea Borgia <andrea@borgia.bo.it> - 2019-04-22 21:30 +0200
      Re: testing: weird resume (black or frozen screen) Andrea Borgia <andrea@borgia.bo.it> - 2019-05-05 12:00 +0200

#207713 — testing: weird resume (black or frozen screen)

FromAndrea Borgia <andrea@borgia.bo.it>
Date2019-04-20 17:50 +0200
Subjecttesting: weird resume (black or frozen screen)
Message-ID<xP0pA-6LE-3@gated-at.bofh.it>
Hi.

Running the testing branch and using xfce, the last couple of resumes 
from hibernation were problematic: the first one had the suspend dialog 
still visible with the "yes" button still pushed in, all applications 
were open, the mouse could move but not actually click anything. 
Keyboard worked enough to switch to console and reboot.

Second instance was slightly different: screen completely black except 
for the mouse pointer. I killed xfce from console and logged in again.

Were it not for the fact that my system is a desktop, this bugreport 
would appear to be a good match: 
https://github.com/the-cavalry/light-locker/issues/108

I found nothing interesting in the Debian bugtracker, perhaps the issue 
is too new, after all I've been using this system for months and this 
issue only came up a couple of days ago.

Any ideas? Anyone else seeing this?

Thanks,
Andrea.

[toc] | [next] | [standalone]


#207714

From"hdv@gmail" <hdv.jadev@gmail.com>
Date2019-04-20 18:20 +0200
Message-ID<xP0SB-7at-1@gated-at.bofh.it>
In reply to#207713
On 20/04/2019 17.43, Andrea Borgia wrote:
> Hi.
> 
> Running the testing branch and using xfce, the last couple of resumes from
> hibernation were problematic: the first one had the suspend dialog still visible
> with the "yes" button still pushed in, all applications were open, the mouse
> could move but not actually click anything. Keyboard worked enough to switch to
> console and reboot.
> 
> Second instance was slightly different: screen completely black except for the
> mouse pointer. I killed xfce from console and logged in again.
> 
> Were it not for the fact that my system is a desktop, this bugreport would
> appear to be a good match: https://github.com/the-cavalry/light-locker/issues/108
> 
> I found nothing interesting in the Debian bugtracker, perhaps the issue is too
> new, after all I've been using this system for months and this issue only came
> up a couple of days ago.
> 
> Any ideas? Anyone else seeing this?
> 
> Thanks,
> Andrea.
> 

Hi Andrea,

I see the same type of problems. I run an up-to-date testing as well, but use
KDE as my DE. Suspend works properly, as far as I can see. I have a Thinkpad P1
with a red led on the lid that blinks during suspend, and that is still as it
was. However, as soon as I try to resume the system it hangs hard. It is not
just a black screen or X that hangs. I cannot even access this system through
SSH anymore when this happens. I still haven't found the culprit causing this,
so I'll be watching this thread with interest.

Grx HdV

[toc] | [prev] | [next] | [standalone]


#207729

FromAndrea Borgia <andrea@borgia.bo.it>
Date2019-04-21 10:50 +0200
Message-ID<xPgkF-86O-1@gated-at.bofh.it>
In reply to#207714
Il 20/04/19 18:16, hdv@gmail ha scritto:


> I see the same type of problems.

Hmm, I'm not so sure about it: in my case the sytem works except (parts 
of) X and it is most likely not kernel-related. In your case it doesn't 
even resume properly, it seems.

You might want to have a look at bugzilla.kernel.org: I had issues with 
my laptop in the recent past and saw that Thinkpads had issues like yours.

https://bugzilla.kernel.org/buglist.cgi?quicksearch=thinkpad

HTH,
Andrea.

[toc] | [prev] | [next] | [standalone]


#207730

From"hdv@gmail" <hdv.jadev@gmail.com>
Date2019-04-21 13:50 +0200
Message-ID<xPj8R-1oc-1@gated-at.bofh.it>
In reply to#207729
On 21/04/2019 10.42, Andrea Borgia wrote:
> Il 20/04/19 18:16, hdv@gmail ha scritto:
> 
> 
>> I see the same type of problems.
> 
> Hmm, I'm not so sure about it: in my case the sytem works except (parts of) X
> and it is most likely not kernel-related. In your case it doesn't even resume
> properly, it seems.
> 
> You might want to have a look at bugzilla.kernel.org: I had issues with my
> laptop in the recent past and saw that Thinkpads had issues like yours.
> 
> https://bugzilla.kernel.org/buglist.cgi?quicksearch=thinkpad

Hi Andrea,

You were right. It was indeed the kernel. If I disable intel_iommu at boot, then
resume works fine again. I haven't noticed any negative impact from doing this
yet. Now I need to find out what changed recently with IOMMU. I went back a few
iterations (to 4.19.0-2) and the problem showed up in those versions as well. It
seem I hadn't rebooted in some while...

I hope you can get rid of your problem soon as well!

Grx HdV

[toc] | [prev] | [next] | [standalone]


#207734

FromAndrea Borgia <andrea@borgia.bo.it>
Date2019-04-21 20:00 +0200
Message-ID<xPoUV-4QL-1@gated-at.bofh.it>
In reply to#207730
Il 21/04/19 13:44, hdv@gmail ha scritto:


> I hope you can get rid of your problem soon as well!

Glad I could help, let's see if I get lucky.


Regards,
Andrea.

[toc] | [prev] | [next] | [standalone]


#207774

FromAndrea Borgia <andrea@borgia.bo.it>
Date2019-04-22 21:30 +0200
Message-ID<xPMNz-2IU-1@gated-at.bofh.it>
In reply to#207713

[Multipart message — attachments visible in raw view] — view raw

Still no solution but an interesting tidbit: I've tried running
pm-hibernate from the console with debugging.

This meant that X session was not locked and, now after resuming, the apps
are still there and the mouse moves. Thing is, nothing else works: can't
click anywhere, alt-tab does nothing, only switching back to console works.
Clock is stuck at 20 minutes ago, too.

Also, the mouse pointer occasionally changes shape when I hover above
certain spots and weirdly enough it uses the desktop icon.

Any ideas?

[toc] | [prev] | [next] | [standalone]


#208187

FromAndrea Borgia <andrea@borgia.bo.it>
Date2019-05-05 12:00 +0200
Message-ID<xUm65-7jV-9@gated-at.bofh.it>
In reply to#207774
Il 22/04/19 21:20, Andrea Borgia ha scritto:


> This meant that X session was not locked and, now after resuming, the 
> apps are still there and the mouse moves. Thing is, nothing else works: 
> can't click anywhere, alt-tab does nothing, only switching back to 
> console works. Clock is stuck at 20 minutes ago, too.

Replying to self for documentation, in case someone else has this 
problem: it is most likely to be kernel-related, just not the one from 
Debian.

I'm using a custom kernel for my Ryzen "G" (with embedded graphics):
https://github.com/M-Bab/linux-kernel-amdgpu-binaries

* 5.0.1 works (with either ucode 2.50 or 2.60)
* anything above exhibits this issue (and occasionally another one, no 
poweroff at the end of hibernation).

Here is the ticket:
https://github.com/M-Bab/linux-kernel-amdgpu-binaries/issues/81

Regards,
Andrea.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.debian.user


csiph-web