Xorg fail with startx

[kgdrenefort]

Hello,

Profile: 23
GPU: NVIDIA Corporation GF106GLM [Quadro 2000M]
Driver: nouveau
Kernel: Linux satan 6.6.21-gentoo-dist #1 SMP PREEMPT_DYNAMIC Mon Apr 8 22:13:15 CEST 2024 x86_64 Intel(R) Core(TM) i7-2820QM CPU @ 2.30GHz GenuineIntel GNU/Linux

It's also a fresh install on a HP Elitebook 8560w, replacing my first ever-installed Gentoo because… Because.

The problem:

I want to use the startx command (and not a DM as lightdm) to start Awesome.

But each time I try, I get an ouput saying it closed successfully (glad to know !), full log here: https://bpa.st/LAUQ

The only errors I see, at least warning, are about some modules not loaded and some missing directories:

[bstaletic]

Last time I have seen that error was when I was running xorg-server with USE=-udev.
Is that the case for you? Or do you have an /etc/X11/xorg.conf?

What is in emerge --info xorg-server?

[pingtoo]

kgdrenefort,

I see no critical error in your log and

[flexibeast]

Further to pingtoo's comment, assuming that you have an ~/.xinitrc with nothing more than the line:

[kgdrenefort]

Thanks for all replies.

[flexibeast]

[kgdrenefort]

[kgdrenefort]

And it failed: https://bpa.st/MU6A

It fails on this line into /var/tmp/portage/x11-wm/awesome-4.3-r103/work/awesome-4.3/tests/examples/runner.sh (line 22):

[pingtoo]

kgdrenefort,

Reading the awesome source code made my headache I don't understand why is modern development went so far to create a complex solution for some simple problem

Anyway, it looks to me is binary that installed on your client (as in client to the binhost) may have problem. So you are correct from your early post, the "lua" binary most likely is build for wrong CPU.

So I would suggest you resolve this issue first, further testing X server and or awesomeWM will not get you anywhere.

I am going to assume "binhost" is also your "build server" for your client machine. (in my mind "binhost" is where binary package is stored and "binhost" will serve the binary package use defined protocol to requester. "binhost" is not necessary the machine that actually create the binary package)

You probably need to review how you create binary packages for client ARCH or the other possibility is that your client setting is wrong, it pick up wrong ARCH from binhost (That is assume your binhost serve multiple ARCH)

You may (or may not) want to create a different thread for the binhost ARCH issue. since it is not relate to X (or this thread) someone else may pickup and help with the binhost issue.

[NeddySeagoon]

kgdrenefort,

Lets try giving Xorg it's default programs to run to try to split the problem into Xorg/Window manager parts.

[kgdrenefort]

[kgdrenefort]

So, not better.

I installed from binaries of the binhost Xfce and LightDM as well.

I enabled LightDM at boot, got a tty.

Trying to restart it give me an error and asking to run this command below:

[NeddySeagoon]

kgdrenefort,

That Xorg failed with its default applications puts the problem (or at least, the first part of it) with Xorg, or the configuration..
Pastebin the entire Xorg.0.log from that startup.

Pastebin the kernel .config file too please. That may save some time. The video configuration in the kernel matters to Xorg. Some things must be off, others must be on.
Exactly what depends on the video card and driver.

Until Xorg can start with twm, it can't start with anything else either. That's why I picked twm as that's Xorgs hard wired default.

Keep the problem space small, we know its Xorg related. We have to fix that first.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

[kgdrenefort]

[NeddySeagoon]

kgdrenefort,

[pingtoo]

kgdrenefort,

Did you rebuild "lua" on your client?

I understand you believe current condition is due to X, but we need to prove it before we dig deep in to X server problem, In this matter I hold a different opinion (I believe X server is functional).

I hope you will continue with us on same train of thoughts. first understand why X server terminate successfully but not doing as expected. May be you can let us know your plan before try something new,

if you did rebuild "lua" on your client then rebuild "awesome" and still not working please share what was the condition of not work. In the past post your local rebuild "awesome" failed due to the build system perform test fail, so what you did to correct this problem and made local build "awesome" successful?

Neddy's suggestion of making use of "twm", "xterm" is to verify X server functionality, you said it did not work, can you elaborate what not work?

[kgdrenefort]

[NeddySeagoon]

kgdrenefort,

Step 1 is to boot to a console with no automatic attempts to start Xorg or a display manager.

The display-manager service must not be any any runlevel.
If it is, it will start Xorg, regardless of anything else installed.

Make a note of the time, then reboot.

Your /var/log/Xorg.0.log, if it exists, must be older that the time you rebooted. If its newer, something is starting Xorg as root.
Likely a display manager. FInd it and stop it.

Check your user Xorg.0.log. As you save not run startx yet, that must be older than the reboot time too.

Once all of your Xorg.0.log files are older than the reboot time, run startx.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

[pingtoo]

kgdrenefort,

Please feel free to experiment, I am not suggest you stop trying anything, I just suggest you may be share your idea so we can understand the progress. Part of this suggestion is that I have seem many time that the experiment introduce a new issue that may or may not related to the original problem and it gets everybody chasing a different rabbit. Something this could give a workaround, but most of time it will turn out to be waste of time.

Because we don't know anything about the context at the time when this issue rise so our step one is to try to get in sync with you so we understand the state. Because there could be many factors influence what you see therefor we need to clear out those known interfering factor (or at lease we know their existent) so we can come up right suggestion.

For example Neddy's suggestion to make sure there is nothing else is running X (or try to start X) so we know we are looking at right log.

Another example I am not sure if you are aware that Xorg.0.log could be at two locations depend on if Xserver was started with root priveleges, so it could be at /var/log/Xorg.0.log or ~<user>/.local/share/xorg/Xorg.0.log.

anyway, let go back to current problem at hand, I don't understand what happen in your post the log for running startx for default (xclock, xterm and twm) the context is kind of muddy. Is the X log messages was display on screen? it did not go into a log file? (I call it muddy is because your shell prompt does not include your normal "kgdrenefort@satan ~ $ ", so I am not sure if you used another user ID to test run, or you just cut off that part)

The message

[kgdrenefort]

[NeddySeagoon]

kgdrenefort,

[kgdrenefort]

[NeddySeagoon]

kgdrenefort,

Good. That proves that Xorg works. It follows that the problem is with awesome or its configuration.

Save that ~/.xinitrc and create a new one with one line.

[kgdrenefort]

[NeddySeagoon]

kgdrenefort,