[Solved] - Tower's 1st [no, 3rd] Hard-Reset since clean-reinstall.

[kdemeoz]

<<Still posting from Lappy>>

Thanks. I didn't see your reply til after i completed my "PS".

I rebooted Tower, got to hanging 2nd splash again, tried shift+alt+f12 to deactivate the effects (compositor) -- nothing at all happened / changed.

I toggled to tty, ran your command, & saw:

Info: the nvidia profile is already in use

Sad, tired. 

[kdemeoz]

Holy cow. Given it said "Info: the nvidia profile is already in use", ie, for whatever reason my previous...

Code:

sudo prime-select intel

... did not work, i decided to see what would happen if now, in tty, i repeated the attempted intel switchover. So i did it, rebooted, & this time it proceeded to the login screen... but don't cheer yet.

It is now stuck in 640x480 resolution, & to my shock, in Settings - Display & Monitor - Displays, that is the ONLY res offered to me. Oh noooooooooooooooo. This is unusable; where are all the other res selections?

[rocky7x]

So to sum it up: second splash screen means AFTER filling in the login credentials? If yes, then just Plasma is having difficulties to load. What you can do to test it is this:

Code:

when you get to a stuck splash screen, switch to a tty a issue this command
sudo su (enter sudo pwd)
adduser test (follow prompts to create the user, leave rooms and phones empty)

Now you've created a clean vanilla user. Go to the login screen, select that new user and try to login - if it works, means that your Plasma settings are wrong for the intel driver and we need to fix them.

Deleting the created user can be done with command:

Code:

deluser test (as root, after sudo su)

If the answer to the above question was NO and you don't even get to the login screen, then the X server has troubles starting and that is because of the Nvidia driver. When you installed the Nvidia driver, have you also installed the package nvidia-prime? When you issue command

Code:

sudo prime-select intel

what is the response?

[kdemeoz]

"nomodeset" is still in grub. Could that be forcing the intel gpu driver into low-res? Should i try removing it again [at risk of booting back into black screen all over again]?

[rocky7x]

Holy cow. Given it said "Info: the nvidia profile is already in use", ie, for whatever reason my previous...

Code:

sudo prime-select intel

... did not work, i decided to see what would happen if now, in tty, i repeated the attempted intel switchover. So i did it, rebooted, & this time it proceeded to the login screen... but don't cheer yet.

It is now stuck in 640x480 resolution, & to my shock, in Settings - Display & Monitor - Displays, that is the ONLY res offered to me. Oh noooooooooooooooo. This is unusable; where are all the other res selections?

Ah OK, I see you're faster than me in posting Anyway, open the konsole, go to /etc/X11/ and see if there is a file called xorg.conf - if yes, post here what it contains. Then check /etc/X11/xorg.conf.d for any files and post their contents. To summarize, there should be NO xorg.conf and in /etc/X11/xorg.conf.d there should be NO files.

[rocky7x]

"nomodeset" is still in grub. Could that be forcing the intel gpu driver into low-res? Should i try removing it again [at risk of booting back into black screen all over again]?

And yes, DEFINITELY REMOVE nomodeset for intel driver nomodeset is only for nvidia driver.

[kdemeoz]

Ah OK, I see you're faster than me in posting Anyway, open the konsole, go to /etc/X11/ and see if there is a file called xorg.conf - if yes, post here what it contains. Then check /etc/X11/xorg.conf.d for any files and post their contents. To summarize, there should be NO xorg.conf and in /etc/X11/xorg.conf.d there should be NO files.

1. /etc/X11/xorg.conf --> file does NOT exist.
2. /etc/X11/xorg.conf.d/ --> directory does NOT exist.
3. nomodeset is now removed again from grub, duly updated, & i'm about to reboot Tower again [listen for explosion]...

[kdemeoz]

Tower is now back in business

It was my unintended usage of nomodeset in grub that limited the screen resolution. Now it's all good again.

If my stuff-up of the attempted kernel 4.9.9 removal did not cause the last several hours of terror, then i have to assume it was the apparently unsuccessful initial attempt to change from the nvidia to intel gpu driver that did cause it. As i wrote a couple of hours ago, after making that initial attempt, it was not at all clear to me that it had succeeded... & IMO, if it had succeeded, i would not have got that boot black screen, which was why i had to then use nomodeset. Later, indeed only within the past 45', when my 2nd attempt to invoke the intel driver did work*, i'd temporarily forgotten i had to remove nomodeset again [& now i've learned how horrible the res becomes as a result]. 

*I assume those hanging splash-screens were coz at that stage neither nvidia or intel gpu drivers were properly activated. 

Does that post-mortem seem reasonable?

leszek & rocky -- i'm sooooooo grateful to you both, again, for rescuing me. I wish i could buy you a beer / scotch / milk / mineral water. 

This is my last post in this thread to be made from Lappy, now that Tower is running again. Next post will be my last for the night [tis now 3.15AM here, yawn], will be back on Tower, & will show you some pics for your thoughts / confirmation of what gpu driver is actually active now. 

[kdemeoz]

<<Posted from back on Tower; phew!>>

I'm a bit confused/worried about the "FAILED" in the Graphics section. What is your interpretation of what gpu & driver is active pls?:

Code:

Z97-HD3:/$ inxi -Fxz
System:    Host: Z97-HD3 Kernel: 4.4.0-63-generic x86_64 (64 bit gcc: 5.4.0)
          Desktop: KDE Plasma 5.8.4 (Qt 5.7.0) Distro: Ubuntu 2.1 xenial
Machine:   Mobo: Gigabyte model: Z97-HD3 v: x.x Bios: American Megatrends v: F7 date: 12/08/2014
CPU:       Quad core Intel Core i7-4790 (-HT-MCP-) cache: 8192 KB
          flags: (lm nx sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx) bmips: 28801
          clock speeds: max: 4000 MHz 1: 3891 MHz 2: 3963 MHz 3: 3823 MHz 4: 3866 MHz 5: 3956 MHz 6: 3931 MHz
          7: 3820 MHz 8: 3976 MHz
Graphics:  Card-1: Intel Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller bus-ID: 00:02.0
          Card-2: NVIDIA GF119 [GeForce GT 610] bus-ID: 01:00.0
          Display Server: X.Org 1.18.4 drivers: nouveau (unloaded: fbdev,vesa) FAILED: intel
          Resolution: 1920x1080@60.00hz
          GLX Renderer: Gallium 0.4 on NVD9 GLX Version: 3.0 Mesa 12.0.6 Direct Rendering: Yes

This implies Nouveau driver [good, that's what i want], but what actual gpu?
   

Ditto...?
   

This indicates successful deployment & activation of rocky's suggested intel-microcode? [annoying/confusing that Driver Manager wrongly(?) implies i'm using Nvidia rather than Nouveau driver].
   

Bed now. Once sun's up again, i'll try to get new kernel 4.8 HWE by running:

Code:

sudo apt-get install --install-recommends xserver-xorg-hwe-16.04

[rocky7x]

Seems you still use nouveau driver. In order to prevent this you have to blacklist it as described in article http://askubuntu.com/questions/481414/in...ad-nouveau

Seems you haven't installed the Nvidia driver correctly, because it should do the blacklisting itself. So either try to reinstall it or do the blacklisting yourself.