No screens found after update to 4.9.32

spira_mirabilis · 2017-07-03 21:17:10

Hello, after updating linux-lts to 4.9.32, my laptop boots to a black screen. (I reproduce this problem with any version 4.9.32+)
Booting my computer to command line and starting Xorg manually, the error "No screens found" appears.

lspci -v
nvidia-lts and bumblebee packages installed.

/usr/share/X11/xorg.conf.d
10-nvidia-drm-outputclass.conf
10-quirks.conf
40-libinput.conf
70-synaptics.conf

/etc/X11/xorg.conf.d
10-monitor.conf
20-intel.conf
20-nvidia.conf

working 4.9.31-1-lts
dmesg
Xorg.0.log
Xorg.8.log

not working 4.9.36-1-lts
dmesg
Xorg.0.log
Xorg.8.log

Is there anyone who can help me?
Thank you in advance.

Last edited by spira_mirabilis (2017-08-06 12:58:35)

loqs · 2017-07-03 21:39:40

Bumblebee is not my area of expertise but it will probably help anyone examining the issue if you provide both xorg.logs
(the one for the integrated card and the one for the discrete card ) and dmesg from both a working kernvel version and a none working kernel version.
Does the issue also occur on a kernel based on the 4.11 or 4.12 series?

spira_mirabilis · 2017-07-08 09:01:21

I edited the first post with dmesg and Xorg logs from both working and not working kernel versions.
With kernel 4.11 I have this problem https://bugs.archlinux.org/task/54234, this is why I'm on linux-lts.

spira_mirabilis · 2017-07-27 05:54:01

I edited the first post adding Xorg configuration files.

seth · 2017-07-27 07:36:04

In general, please don't bump (posting empty)

That said and wrt your Xorg logs: the "not working" server runs fine for ~23 seconds, so this has nothing to do with the X server but with the desktop session and/or login manager you attempt to run, why you need to provide details on this matter (as well as on your "Booting my computer to command line and starting Xorg manually" approach, including the command you run and a full output)

spira_mirabilis · 2017-08-06 13:03:29

seth wrote:

In general, please don't bump (posting empty)

Sorry about that, I bumped because no one was trying to help me. Now that I bumped, you are helping me, so it worked!

seth wrote:

That said and wrt your Xorg logs: the "not working" server runs fine for ~23 seconds, so this has nothing to do with the X server but with the desktop session and/or login manager you attempt to run, why you need to provide details on this matter (as well as on your "Booting my computer to command line and starting Xorg manually" approach, including the command you run and a full output)

I'm using LightDM with i3wm.
My "Booting my computer to command line and starting Xorg manually" approach is this: I have booted my computer adding the arguments "nomodeset" and "single" to linux line in the grub config; then i try to lounch the "startx" command, producing the log in my first post (Xorg.0.log).

seth · 2017-08-06 14:18:35

"nomodeset" kills the intel driver, so with bumblebee installed/configured, X will not start for sure. no surprise there.
drop that idea, boot into the multi-user.target, login and try to startx (ensure to have a sane ~/.xinitrc) to get lightdm out of the equation.

spira_mirabilis · 2017-08-06 15:40:46

Booting into the multi-user.target ends up to black screen, as well as without kernel parameters.

seth · 2017-08-06 16:59:23

Ok, then you're hit by a modesetting regression in the kernel. Enable sshd, reboot *with* modesetting (ie. into the black screen), ssh into the machine, dump a dmesg and hope for some error messages allowing us to track this down.

loqs · 2017-08-06 17:18:40

The journal may contain the dmesg even though the screen was blank. Check the journal for the boot with multi-user.target and post the journal for that boot if it is present.

spira_mirabilis · 2017-08-06 18:22:01

seth wrote:

Ok, then you're hit by a modesetting regression in the kernel. Enable sshd, reboot *with* modesetting (ie. into the black screen), ssh into the machine, dump a dmesg and hope for some error messages allowing us to track this down.

When the black screen appears I can't do nothing but restart the computer, so how can I ssh into the same machine?. Am I missing something?

Here the journal from the boot with multi-user.target https://pastebin.com/nPVjDm0X

loqs · 2017-08-06 19:00:33

https://bugs.archlinux.org/task/54471

seth · 2017-08-06 19:08:19

And ftr: https://wiki.archlinux.org/index.php/Secure_Shell

spira_mirabilis · 2017-08-06 19:45:54

loqs wrote:

https://bugs.archlinux.org/task/54471

It seems to be the same problem. I'll add my configurations and logs to the task.

seth wrote:

And ftr: https://wiki.archlinux.org/index.php/Secure_Shell

I'm aware of the ssh command, but there are two things that confuse me:
~ ssh into local machine from local machine. Why would I do that?
~ I can't do anything while my computer is stuck in the black screen of death, including ssh.

Thank you both for helping me so far!

seth · 2017-08-06 19:48:49

It's the same backtrace. Can you disable CSM in your BIOS (or attach an external display)?

The idea is of course to use a second machine - didn't occur to me that this might be the only syste you've access to.
Sorry, I'm surrounded by computers ;-)

loqs · 2017-08-06 19:56:17

From https://cdn.kernel.org/pub/linux/kernel … Log-4.9.32

commit 09fcb3561d9ede40610d8f7b7fd64d2651179403
Author: Jani Nikula <jani.nikula@intel.com>
Date: Fri Mar 10 15:27:58 2017 +0200
drm/i915/vbt: split out defaults that are set when there is no VBT

commit bb1d132935c2f87cd261eb559759fe49d5e5dc43 upstream.

The main thing are the DDI ports. If there's a VBT that says there are
no outputs, we should trust that, and not have semi-random
defaults. Unfortunately, the defaults have resulted in some Chromebooks
without VBT to rely on this behaviour, so we split out the defaults for
the missing VBT case.

Reviewed-by: Manasi Navare <manasi.d.navare@intel.com>
Cc: Manasi Navare <manasi.d.navare@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: http://patchwork.freedesktop.org/patch/ … @intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 555c443a1ab9f6f6935ad23fa24a795d64220730
Author: Jani Nikula <jani.nikula@intel.com>
Date: Fri Mar 10 15:27:57 2017 +0200
drm/i915/vbt: don't propagate errors from intel_bios_init()

commit 665788572c6410b7efadc2e3009c5d830b6d8ef9 upstream.

We don't use the error return for anything other than reporting and
logging that there is no VBT. We can pull the logging in the function,
and remove the error status return. Moreover, if we needed the
information for something later on, we'd probably be better off storing
the bit in dev_priv, and using it where it's needed, instead of using
the error return.

While at it, improve the comments.

Cc: Manasi Navare <manasi.d.navare@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: http://patchwork.freedesktop.org/patch/ … @intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Those are the only i915 changes you could bisect between 4.9.31 and 4.9.32 and find the bad commit most likely one of those two.
Then report it upstream if there is not already an upstream report about it.

Arch Linux

#1 2017-07-03 21:17:10

No screens found after update to 4.9.32

#2 2017-07-03 21:39:40

Re: No screens found after update to 4.9.32

#3 2017-07-08 09:01:21

Re: No screens found after update to 4.9.32

#4 2017-07-27 05:54:01

Re: No screens found after update to 4.9.32

#5 2017-07-27 07:36:04

Re: No screens found after update to 4.9.32

#6 2017-08-06 13:03:29

Re: No screens found after update to 4.9.32

#7 2017-08-06 14:18:35

Re: No screens found after update to 4.9.32

#8 2017-08-06 15:40:46

Re: No screens found after update to 4.9.32

#9 2017-08-06 16:59:23

Re: No screens found after update to 4.9.32

#10 2017-08-06 17:18:40

Re: No screens found after update to 4.9.32

#11 2017-08-06 18:22:01

Re: No screens found after update to 4.9.32

#12 2017-08-06 19:00:33

Re: No screens found after update to 4.9.32

#13 2017-08-06 19:08:19

Re: No screens found after update to 4.9.32

#14 2017-08-06 19:45:54

Re: No screens found after update to 4.9.32

#15 2017-08-06 19:48:49

Re: No screens found after update to 4.9.32

#16 2017-08-06 19:56:17

Re: No screens found after update to 4.9.32

Board footer