Debian 10, compatibility problem

[ispgardner]

I would like to report that the new release of debian buster is not compatible with my computer; the debian installer crashes during X installation. Clean netinst installation works fine, but X installation crashes. I suspect that Wayland is not compatible with my hardware. On the other note; NONE of my application uses wayland, ALL my user software use Xorg. I have reverted to stretch, which is working just fine.

Is it possible to install buster w/o wayland drivers?

Here is my computer;
It is Acer cb3-532 with the coreboot. CPU: Intel(R) Celeron(R) CPU N3060 @ 1.60GHz;
VGA compatible controller: Intel Corporation Atom/Celeron/Pentium Processor x5-E8000/J3xxx/N3xxx Integrated Graphics Controller (rev 35).

I hope this incompatibility will be fixed in the future updates and I will wait for the time that I could use this version of debian.

[ruwolf]

Yes, in the buster, the classic X11 is still present as you can see e.g. here: packages.debian.org: buster: x11: xserver-xorg-video-all

[Head_on_a_Stick]

Clean netinst installation works fine, but X installation crashes.

The netinstall image allows you to select a graphical desktop, did that also fail? Did you check TTY4 for any error messages?

Did you check the validity of the image and are you sure the USB stick is not bad? How did you transfer the ISO image to the installation medium?

You could install a console only system and then use tasksel to pull in the graphical desktop.

Is it possible to install buster w/o wayland drivers?

There are no Wayland drivers.

And there are no Debian developers here, if you want to report a problem then use the proper channels.

[ispgardner]

Thank you for picking up the discussion. I have not reported this “feature” as a bug, because I do not know it is a bug. However, I felt obligated to report this problem somewhere; why my post is here. Here are some facts:

- Debian 10 crashes during Xwindow installation on my computer, while Debian 9 works close to perfection (pls. Notice that debian 10 is LTS, while 9 is not).
- According to changelog; debian 10 uses wayland as the default X driver (xorg is used but not default).
- xserver-xorg-core depends on libwayland-client0 & libwayland-server0 and (for some reason) these wayland drivers are downloaded & installed during default xorg installation.
- Other distributions (RH, Ubuntu) have switched to wayland and it was frequently reported that wayland is not fully mature having hardware/software compatibility problems for some computers.

My knowledge of debian is limited to be certain if wayland on 10 is a real problem. I think my post (and reports of other problems) is a contribution to the debian project.

[Head_on_a_Stick]

- Debian 10 crashes during Xwindow installation on my computer

What does "Xwindow installation" mean, exactly? Please explain more fully what the problem is and how to reproduce it.

Notice that debian 10 is LTS, while 9 is not

That is incorrect, neither Debian 10 nor Debian 9 are currently covered by the LTS team. Debian 9 falls under the support of the LTS team next year and will continue to be covered by them until June 2022: https://wiki.debian.org/LTS

debian 10 uses wayland as the default

Only for the GNOME desktop and it is possible to select an X-based session in the login screen.

- xserver-xorg-core depends on libwayland-client0 & libwayland-server0 and (for some reason) these wayland drivers are downloaded & installed during default xorg installation.

They are not "drivers", those packages are libraries.

[ispgardner]

Sorry for the delay, but I needed time to make sure how buster and stretch differes.
First how I have installed (for the purpose of this testing) both debian releases on converted Googlebook; Acer c3-532:

• - Since wifi required non-free driver I have used “debian-10.0.0-amd64-netinst.iso” image to install bare debian w/o the network.
  - After debian installation, using dpkg, I have installed relevant wifi driver.
  - Using binary from the above iso I have installed, wpa-supplicant and I have obtained network connection.
  - After modifying sources.list I have run apt-get update.
  - Next I have run apt-get install xinit followed by apt-get install jwm (window manager used for this test).

Up to this point I have not noticed any single error.
BTW it looks complicated, bat I have used this method for years & I have written a simple script to do all this; It takes less than 3 minutes to complete all these tasks.

As a regular user:

• - I have modified .xsession in user directory.
  - Next I was trying to rum startx; at this moment computer crashed and only power button worked. I was able to recover Xorg.0.log and .xsession-errors in the next session, but these logs didn't bring much information.

I have repeated the above procedure using debian streatch and this time everything worked as expected !!!

I hesitate to report this as a bug; I'm using converted Goglebook that uses Seabios (https://en.wikipedia.org/wiki/SeaBIOS), which may not be standard bios. And some software (otherwise good) may not be compatible with this (why my LTS question for stretch).

[Head_on_a_Stick]

• - Since wifi required non-free driver I have used “debian-10.0.0-amd64-netinst.iso” image to install bare debian w/o the network.
  - After debian installation, using dpkg, I have installed relevant wifi driver.
  - Using binary from the above iso I have installed, wpa-supplicant and I have obtained network connection.

Why not just use the unofficial non-free image instead? That should include the firmware (not driver) required by your wireless card.

I was able to recover Xorg.0.log and .xsession-errors in the next session, but these logs didn't bring much information.

You should post the logs here (using code tags) so we can check if you've missed anything.

[ispgardner]

Today I've done one more thing:
I've used <firmware-10.0.0-amd64-netinst.iso> to install buster with graphics; during the installation the LXDE desktop was chosen. The installation went w/o errors, but after reboot system crashed. Installation took quite a long time & I didn't have time to recover the logs.

Here is <.xsession-errors> log from the attempt described above:

Code: Select all

Xsession: X session started for mypass at Tue 06 Aug 2019 12:45:10 AM UTC 
XIO:  fatal IO error 104 (Connection reset by peer) on X server ":0"^M
      after 2260 requests (2032 known processed) with 0 events remaining.^M

Also <Xorg.0.log>, (file is quite long, so I'm showing fragment from the first (EE)):

Code: Select all

 ...
[  1441.013] (II) This device may have been added with another device file.
[  1441.014] (II) config/udev: Adding input device AT Translated Set 2 keyboard (/dev/input/event0)
[  1441.014] (**) AT Translated Set 2 keyboard: Applying InputClass "libinput keyboard catchall"
[  1441.014] (II) Using input driver 'libinput' for 'AT Translated Set 2 keyboard'
[  1441.020] (II) systemd-logind: got fd for /dev/input/event0 13:64 fd 26 paused 0
[  1441.020] (**) AT Translated Set 2 keyboard: always reports core events
[  1441.020] (**) Option "Device" "/dev/input/event0"
[  1441.020] (**) Option "_source" "server/udev"
[  1441.022] (II) event0  - AT Translated Set 2 keyboard: is tagged by udev as: Keyboard
[  1441.022] (II) event0  - AT Translated Set 2 keyboard: device is a keyboard
[  1441.023] (II) event0  - AT Translated Set 2 keyboard: device removed
[  1441.023] (**) Option "config_info" "udev:/sys/devices/platform/i8042/serio0/input/input0/event0"
[  1441.024] (II) XINPUT: Adding extended input device "AT Translated Set 2 keyboard" (type: KEYBOARD, id 10)
[  1441.026] (II) event0  - AT Translated Set 2 keyboard: is tagged by udev as: Keyboard
[  1441.026] (II) event0  - AT Translated Set 2 keyboard: device is a keyboard
[  1441.028] (II) config/udev: Adding input device PC Speaker (/dev/input/event5)
[  1441.028] (II) No input driver specified, ignoring this device.
[  1441.028] (II) This device may have been added with another device file.
[  1441.040] (EE) modeset(0): failed to set mode: Invalid argument
[  1516.984] (**) Option "fd" "20"
[  1516.985] (II) event3  - Power Button: device removed
[  1516.986] (**) Option "fd" "23"
[  1516.986] (II) event2  - Power Button: device removed
[  1516.986] (**) Option "fd" "24"
[  1516.987] (II) event6  - HD WebCam: HD WebCam: device removed
[  1516.987] (**) Option "fd" "25"
[  1516.987] (II) event4  - Elan Touchpad: device removed
[  1516.988] (**) Option "fd" "26"
[  1516.988] (II) event0  - AT Translated Set 2 keyboard: device removed
[  1516.988] (II) AIGLX: Suspending AIGLX clients for VT switch
[  1517.044] (II) systemd-logind: got pause for 13:68
[  1517.044] (II) systemd-logind: got pause for 226:0
[  1517.044] (II) systemd-logind: got pause for 13:66
[  1517.044] (II) systemd-logind: got pause for 13:70
[  1517.044] (II) systemd-logind: got pause for 13:64
[  1517.044] (II) systemd-logind: got pause for 13:67
[  1528.211] (II) systemd-logind: got resume for 13:68
[  1528.211] (II) systemd-logind: got resume for 226:0
[  1528.211] (II) AIGLX: Resuming AIGLX clients after VT switch
[  1528.212] (EE) modeset(0): failed to set mode: Invalid argument
[  1528.212] (EE) 
Fatal server error:
[  1528.212] (EE) EnterVT failed for screen 0
[  1528.213] (EE) 
[  1528.213] (EE) 
Please consult the The X.Org Foundation support 
	 at http://wiki.x.org
 for help. 
[  1528.213] (EE) Please also check the log file at "/home/mypass/.local/share/xorg/Xorg.0.log" for additional information.
[  1528.213] (EE) 
[  1528.214] (II) AIGLX: Suspending AIGLX clients for VT switch
[  1528.298] (EE) Server terminated with error (1). Closing log file.

Comment; pls. check line "[1441.040] (EE) modeset(0): failed to set mode: Invalid argument"

There is some suggestion of some incompatibility of this machine/CPU (or failed discovery?).

[vbrummond]

X is just crashing. What is your graphics hardware, and did you try the non free image?

[Head_on_a_Stick]

Comment; pls. check line "[1441.040] (EE) modeset(0): failed to set mode: Invalid argument"

OK, looks like you have encountered the problem discussed in this thread: http://forums.debian.net/viewtopic.php?f=17&t=143041

To fix things switch to the Intel DDX driver:

Code: Select all

# apt install xserver-xorg-video-intel
# mkdir -p /etc/X11/xorg.conf.d
# tee /etc/X11/xorg.conf.d/20-intel.conf <<!
Section "Device"
   Identifier "Intel Graphics"
   Driver "Intel"
EndSection
!

Then log out & back in again.

If the configuration file (/etc/X11/xorg.conf.d/20-intel.conf, the contents are listed between the exclamation marks in my code block) doesn't work then we may have to add a bus address, the output of lspci will tell us that.

[ispgardner]

Success!! I've expected longer battle, but when I have followed “what doctor ordered” X statred working like a charm.

To (maybe) help in some further development I would like to point to some differences between before and after modification:

1. The firmware <xserver-xorg-video-intel> was already installed, so I've skipped the apt-get part; this “firmware” is a part of xinit standard installation. Here is the output of dpkg -l | grep “xorg-video”:

Code: Select all

ii  xserver-xorg-video-all        1:7.7+19                    amd64        X.Org X server -- output driver metapackage
ii  xserver-xorg-video-amdgpu     18.1.99+git20190207-1       amd64        X.Org X server -- AMDGPU display driver
ii  xserver-xorg-video-ati        1:19.0.1-1                  amd64        X.Org X server -- AMD/ATI display driver wrapper
ii  xserver-xorg-video-fbdev      1:0.5.0-1                   amd64        X.Org X server -- fbdev display driver
ii  xserver-xorg-video-intel      2:2.99.917+git20180925-2    amd64        X.Org X server -- Intel i8xx, i9xx display driver
ii  xserver-xorg-video-nouveau    1:1.0.16-1                  amd64        X.Org X server -- Nouveau display driver
ii  xserver-xorg-video-qxl        0.1.5-2+b1                  amd64        X.Org X server -- QXL display driver
ii  xserver-xorg-video-radeon     1:19.0.1-1                  amd64        X.Org X server -- AMD/ATI Radeon display driver
ii  xserver-xorg-video-vesa       1:2.4.0-1                   amd64        X.Org X server -- VESA display driver
ii  xserver-xorg-video-vmware     1:13.3.0-2                  amd64        X.Org X server -- VMware display driver

2. As expected the output of lspci didn't changed.
3. I have run lsmod before & after modification: first the video module is the same (i915), but after modification three hid modules are NOT loaded, e.i. output of lsmod after modification is shorter. Also, the order how modules are loaded changed significantly (I've saved these outputs & I can provide them if they are needed).
4. I have tried VESA firmware that helped me as a fallout firmware in the past; it didn't work.

Thank you very much for help; I will try to load buster into my other machines ( hope w/o problems)
Also, I'll thy to mark this thread (if I could) as [SOLVED]