[SOLVED] MGA6Dev1 updated and now boots to black

[rolf]

I had installed from Mageia-6-dev1-x86_64-DVD.iso and had KDE working with nvidia driver, dkms, etc. then did not boot it for a while.
Today, I booted into this installation and `urpmi --auto-select` gave more than 800 updates, including the kernel, nvidia packages, and dkms was working in terminal at one point but I didn't watch closely.

When booting to the new kernel, I see the kernel messages as I have chosen 'text boot' and the screen goes black after a few seconds. The monitor shows no input. There are no virtual terminals at Ctrl-Alt-F? After waiting a few to many minutes with no change, I can reboot with Ctrl-Alt-Del.

I have appended 3 to the kernel command line and chosen Failsafe from the grub menu but there is no change with either try. Some entries from menu.lst:

Code: Select all

timeout 5
color black/cyan yellow/cyan
default 0

title linux
kernel (hd4,8)/boot/vmlinuz BOOT_IMAGE=linux root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot noiswmd resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788
root (hd4,8)
initrd /boot/initrd.img

title linux-nonfb
kernel (hd4,8)/boot/vmlinuz BOOT_IMAGE=linux-nonfb root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3
root (hd4,8)
initrd /boot/initrd.img

title failsafe
kernel (hd4,8)/boot/vmlinuz BOOT_IMAGE=failsafe root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot failsafe noiswmd
root (hd4,8)
initrd /boot/initrd.img


and what works in MGA5 is much the same:

Code: Select all

timeout 5
color black/cyan yellow/cyan
default 0

title linux
kernel (hd0,4)/boot/vmlinuz BOOT_IMAGE=linux root=UUID=cc0475e9-34dc-40ad-88e6-e656e6040a03  nokmsboot noiswmd resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788
root (hd0,4)
initrd /boot/initrd.img

title linux-nonfb
kernel (hd0,4)/boot/vmlinuz BOOT_IMAGE=linux-nonfb root=UUID=cc0475e9-34dc-40ad-88e6-e656e6040a03  nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3
root (hd0,4)
initrd /boot/initrd.img

title failsafe
kernel (hd0,4)/boot/vmlinuz BOOT_IMAGE=failsafe root=UUID=cc0475e9-34dc-40ad-88e6-e656e6040a03  nokmsboot failsafe noiswmd
root (hd0,4)
initrd /boot/initrd.img


[gohlip]

Try booting with different (hdx,8) in the root line (change it manually and boot). [x= 0,1,2,3]
instead of "root (hd4,8)"

Even if it boots, be aware that 'x' may change in future as bios will change the device mapping at different boots.
Better you remove unnecessary external drives at boot time to have better chance of hitting random correct device.
grub 2 will have no such issue, but your choice.

Oh, one more thing, in your menu.lst, you use vmlinuz and initrd.img
It should be vmlinuz-desktop and initrd-desktop.img otherwise you may be booting obsolete kernel instead.

Also, there is no 'ro' in the kernel line? Is that correct?

ps: I don't use grub-legacy.

[rolf]

Try booting with different (hdx,8) in the root line (change it manually and boot). [x= 0,1,2,3]
instead of "root (hd4,8)"
Even if it boots, be aware that 'x' may change in future as bios will change the device mapping at different boots.
Better you remove unnecessary external drives at boot time to have better chance of hitting random correct device.
grub 2 will have no such issue, but your choice.

When booting to the new kernel, I see the kernel messages

The kernel is booting, the device is correct.

Oh, one more thing, in your menu.lst, you use vmlinuz and initrd.img
It should be vmlinuz-desktop and initrd-desktop.img otherwise you may be booting obsolete kernel instead.

Kernel install script makes these links and they point to the same kernel:

Code: Select all

# ll /mnt/disk/boot
...
lrwxrwxrwx 1 root root       34 May  6 16:15 vmlinuz -> vmlinuz-4.6.0-desktop-0.rc6.1.mga6
-rw-r--r-- 1 root root  4457824 Mar 10 08:22 vmlinuz-4.4.5-desktop-1.mga6
-rw-r--r-- 1 root root  4459136 Mar 16 22:38 vmlinuz-4.4.6-desktop-1.mga6
-rw-r--r-- 1 root root  4597296 May  3 16:31 vmlinuz-4.6.0-desktop-0.rc6.1.mga6
lrwxrwxrwx 1 root root       34 May  6 16:15 vmlinuz-desktop -> vmlinuz-4.6.0-desktop-0.rc6.1.mga6

Also, there is no 'ro' in the kernel line? Is that correct?

It's how the Mageia scripts have set it, how it is in MGA5, which is working.

ps: I don't use grub-legacy.

I think it's not a matter of the bootloader, see my next post.

[rolf]

I don't think I said that, when I boot the original kernel, the behavior is the same, some kernel messages then no video. It's puzzling why init 3 and failsafe have no video, either.

However, it seems clear the breakdown comes with installation of the nividia driver:

Code: Select all

# cat /mnt/disk/var/log/Xorg.0.log
[    87.716] (--) Log file renamed from "/var/log/Xorg.pid-3184.log" to "/var/log/Xorg.0.log"
[    87.716]
X.Org X Server 1.18.3
Release Date: 2016-04-04
[    87.716] X Protocol Version 11, Revision 0
[    87.716] Build Operating System: rabbit 4.4.6-server-1.mga5
[    87.716] Current Operating System: Linux p8z68 4.4.5-desktop-1.mga6 #1 SMP Thu Mar 10 16:19:23 UTC 2016 x86_64
[    87.716] Kernel command line: BOOT_IMAGE=desktop_4.4.5-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788
[    87.716] Build Date: 20 April 2016  10:30:50AM
[    87.716] Build ID: x11-server 1.18.3-7.mga6
[    87.716] Current version of pixman: 0.34.0
[    87.716]    Before reporting problems, check http://bugs.mageia.org
        to make sure that you have the latest version.
[    87.716] Markers: (--) probed, (**) from config file, (==) default setting,
        (++) from command line, (!!) notice, (II) informational,
        (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[    87.716] (==) Log file: "/var/log/Xorg.0.log", Time: Sat May  7 06:49:23 2016
[    87.716] (==) Using config file: "/etc/X11/xorg.conf"
[    87.716] (==) Using config directory: "/etc/X11/xorg.conf.d"
[    87.716] (==) Using system config directory "/usr/share/X11/xorg.conf.d"
[    87.716] (==) ServerLayout "Layout0"
[    87.716] (**) |-->Screen "Screen0" (0)
[    87.716] (**) |   |-->Monitor "Monitor0"
[    87.716] (**) |   |-->Device "Device0"
[    87.716] (**) |-->Input Device "Keyboard0"
[    87.716] (**) |-->Input Device "Mouse0"
[    87.716] (**) Option "Xinerama" "0"
[    87.716] (==) Automatically adding devices
[    87.716] (==) Automatically enabling devices
[    87.716] (==) Automatically adding GPU devices
[    87.716] (==) Max clients allowed: 256, resource mask: 0x1fffff
[    87.716] (**) FontPath set to:
        /usr/share/fonts/default/Type1,
        catalogue:/etc/X11/fontpath.d,
        built-ins
[    87.716] (==) ModulePath set to "/usr/lib64/xorg/modules"
[    87.716] (WW) Hotplugging is on, devices using drivers 'kbd', 'mouse' or 'vmmouse' will be disabled.
[    87.716] (WW) Disabling Keyboard0
[    87.716] (WW) Disabling Mouse0
[    87.716] (II) Loader magic: 0x80cd00
[    87.716] (II) Module ABI versions:
[    87.716]    X.Org ANSI C Emulation: 0.4
[    87.716]    X.Org Video Driver: 20.0
[    87.716]    X.Org XInput driver : 22.1
[    87.716]    X.Org Server Extension : 9.0
[    87.716] (++) using VT number 1

[    87.716] (II) systemd-logind: logind integration requires -keeptty and -keeptty was not provided, disabling logind integration
[    87.717] (II) xfree86: Adding drm device (/dev/dri/card1)
[    87.717] (II) xfree86: Adding drm device (/dev/dri/card0)
[    87.718] (--) PCI: (0:0:2:0) 8086:0122:1043:844d rev 9, Mem @ 0xf7400000/4194304, 0xd0000000/268435456, I/O @ 0x0000f000/64
[    87.718] (--) PCI:*(0:2:0:0) 10de:1401:19da:2375 rev 161, Mem @ 0xf6000000/16777216, 0xe0000000/268435456, 0xf0000000/33554432, I/O @ 0x0000e000/128, BIOS @ 0x????????/524288
[    87.718] (II) Open ACPI successful (/var/run/acpid.socket)
[    87.718] (II) LoadModule: "glx"
[    87.719] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[    87.724] (II) Module glx: vendor="X.Org Foundation"
[    87.724]    compiled for 1.18.3, module version = 1.0.0
[    87.724]    ABI class: X.Org Server Extension, version 9.0
[    87.724] (==) AIGLX enabled
[    87.724] (II) LoadModule: "nvidia"
[    87.725] (WW) Warning, couldn't open module nvidia
[    87.725] (II) UnloadModule: "nvidia"
[    87.725] (II) Unloading nvidia
[    87.725] (EE) Failed to load module "nvidia" (module does not exist, 0)
[    87.725] (EE) No drivers available.
[    87.725] (EE)
Fatal server error:
[    87.725] (EE) no screens found(EE)
[    87.725] (EE)
Please consult the Mageia support
         at http://bugs.mageia.org
 for help.
[    87.725] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    87.725] (EE)

[doktor5000]

Code: Select all

urpme -a nvidia

and setting up nvidia driver via drakx11 should help - you're currently missing the nvidia kernel module, so nvidia driver will not work. Maybe update stalled in between, as mirror was out of sync or so ?
Or you could also try to use another driver via xdriver=nouveau or xdriver=vesa as kernel option.

[gohlip]

Okay, understand why you think strongly it's the graphics driver, not the kernel or device.
But still, like you, if that is the case, don't understand why failsafe or '3' won't boot to prompt (so you can fix the graphics).

So, one more try and ....to make sure it's the kernel, we'll boot directly to kernel and not the sym-link


At grub prompt,

Code: Select all

grub> find /boot/grub/stage1     <just to be sure>
grub> root (hdx,y)                    <# if it is indeed (hd4,8) given above>
grub> kernel (hdx,y)/boot/vmlinuz-4.6.0-desktop-0.rc6.1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot ro   <# if it is hd4,8 or use root=/dev/sdxy if easier and just put in 'ro' as well>
grub> initrd (hdx,y)/boot/initrd-4.6.0-desktop-0.rc6.1.mga6.img              <# is this there at all? please recheck.>
grub> boot


@doktor5000, how is he going to do that if he cannot even boot to prompt?
Unless you suggest he chroot to the partition? Then he should also do "dracut -f"

[rolf]

Maybe update stalled in between, as mirror was out of sync or so ?
Or you could also try to use another driver via xdriver=nouveau or xdriver=vesa as kernel option.

I noticed dkms working on the module during the update but did not pay close attention. I tried xdriver=vesa with the default entry but nvidia is still tried, see below for files with the time of this attempt, and the end result is the same, no video. The keyboard works. I can Ctrl-Alt-Del or magic keys rseiub, so kernel is loading.

Code: Select all

[root@p8z68 rolf]# cat /mnt/disk/var/log/dmesg | grep xdriver
[    0.000000] Command line: BOOT_IMAGE=linux root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot noiswmd resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 xdriver=vesa
[    0.000000] Kernel command line: BOOT_IMAGE=linux root=UUID=131bba44-fab2-42ff-93f7-007934a41181 nokmsboot noiswmd resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 xdriver=vesa

[root@p8z68 rolf]# ll /mnt/disk/var/log/Xorg.0.log
-rw-r--r-- 1 root root 3991 May  7 08:58 /mnt/disk/var/log/Xorg.0.log

[root@p8z68 rolf]# tail -n 20 /mnt/disk/var/log/Xorg.0.log
[    81.993] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[    81.999] (II) Module glx: vendor="X.Org Foundation"
[    81.999]    compiled for 1.18.3, module version = 1.0.0
[    81.999]    ABI class: X.Org Server Extension, version 9.0
[    81.999] (==) AIGLX enabled
[    81.999] (II) LoadModule: "nvidia"
[    82.000] (WW) Warning, couldn't open module nvidia
[    82.000] (II) UnloadModule: "nvidia"
[    82.000] (II) Unloading nvidia
[    82.000] (EE) Failed to load module "nvidia" (module does not exist, 0)
[    82.000] (EE) No drivers available.
[    82.000] (EE)
Fatal server error:
[    82.000] (EE) no screens found(EE)
[    82.000] (EE)
Please consult the Mageia support
         at http://bugs.mageia.org
 for help.
[    82.000] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    82.000] (EE)
[root@p8z68 rolf]#

Trying the kernel that was running when I did the update and this, also, boots to no video:

Code: Select all

[root@p8z68 rolf]# cat /mnt/disk/var/log/dmesg | grep xdriver
[    0.000000] Command line: BOOT_IMAGE=desktop_4.4.6-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 xdriver=vesa
[    0.000000] Kernel command line: BOOT_IMAGE=desktop_4.4.6-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 xdriver=vesa
[root@p8z68 rolf]# tail -n 20 /mnt/disk/var/log/Xorg.0.log
[    83.706] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[    83.712] (II) Module glx: vendor="X.Org Foundation"
[    83.712]    compiled for 1.18.3, module version = 1.0.0
[    83.712]    ABI class: X.Org Server Extension, version 9.0
[    83.712] (==) AIGLX enabled
[    83.712] (II) LoadModule: "nvidia"
[    83.712] (WW) Warning, couldn't open module nvidia
[    83.712] (II) UnloadModule: "nvidia"
[    83.712] (II) Unloading nvidia
[    83.712] (EE) Failed to load module "nvidia" (module does not exist, 0)
[    83.712] (EE) No drivers available.
[    83.712] (EE)
Fatal server error:
[    83.712] (EE) no screens found(EE)
[    83.712] (EE)
Please consult the Mageia support
         at http://bugs.mageia.org
 for help.
[    83.712] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    83.712] (EE)
[root@p8z68 rolf]# ll /mnt/disk/var/log/dmesg
-rw-r--r-- 1 root root 89879 May  7 09:17 /mnt/disk/var/log/dmesg
[root@p8z68 rolf]# ll /mnt/disk/var/log/Xorg.0.log
-rw-r--r-- 1 root root 4001 May  7 09:18 /mnt/disk/var/log/Xorg.0.log

One try with xdriver=nouveau and 3, which does not start X but also no video (would this be called framebuffer, if working?):

Code: Select all

[root@p8z68 rolf]# cat /mnt/disk/var/log/dmesg | grep xdriver
[    0.000000] Command line: BOOT_IMAGE=desktop_4.4.6-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot xdriver=nouveau resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 3
[    0.000000] Kernel command line: BOOT_IMAGE=desktop_4.4.6-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot xdriver=nouveau resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788 3
[root@p8z68 rolf]# ll /mnt/disk/var/log/dmesg
-rw-r--r-- 1 root root 87082 May  7 09:45 /mnt/disk/var/log/dmesg
[root@p8z68 rolf]# ll /mnt/disk/var/log/Xorg.0.log
-rw-r--r-- 1 root root 4001 May  7 09:18 /mnt/disk/var/log/Xorg.0.log
[root@p8z68 rolf]#

do "dracut -f"

Code: Select all

[root@p8z68 rolf]# ll /mnt/disk/boot/initrd*
-rw------- 1 root root 21555155 Mar 27 11:27 /mnt/disk/boot/initrd-4.4.5-desktop-1.mga6.img
-rw------- 1 root root 21553334 Mar 27 11:47 /mnt/disk/boot/initrd-4.4.6-desktop-1.mga6.img
-rw------- 1 root root 11001337 May  6 16:15 /mnt/disk/boot/initrd-4.6.0-desktop-0.rc6.1.mga6.img

The new initrd is half the size of the previous two, which haven't been touched, in case dracut is needed whenever the nvidia driver is updated.

I could do something in chroot but, in case there is some bug in Dev1, it seems a good idea to find out what is wrong before erasing what was done. Maybe there is some information from dkms status in chroot:

Code: Select all

[root@p8z68 /]# dkms status
nvidia-current, 361.42-3.mga6.nonfree, 4.4.5-desktop-1.mga6, x86_64: installed
nvidia-current, 361.42-3.mga6.nonfree, 4.4.6-desktop-1.mga6, x86_64: installed
nvidia-current, 352.79-2.mga6.nonfree, 4.4.5-desktop-1.mga6, x86_64: installed-binary from 4.4.5-desktop-1.mga6
nvidia-current, 352.79-2.mga6.nonfree, 4.4.6-desktop-1.mga6, x86_64: installed-binary from 4.4.6-desktop-1.mga6
nvidia-current, 361.42-3.mga6.nonfree, 4.6.0-desktop-0.rc6.1.mga6, x86_64: installed-binary from 4.6.0-desktop-0.rc6.1.mga6
[root@p8z68 /]# rpm -qa | grep nvidia
nvidia-current-kernel-4.6.0-desktop-0.rc6.1.mga6-361.42-6.mga6.nonfree
nvidia-current-kernel-desktop-latest-361.42-6.mga6.nonfree
nvidia-current-doc-html-361.42-3.mga6.nonfree
nvidia-current-kernel-4.4.5-desktop-1.mga6-352.79-11.mga6.nonfree
x11-driver-video-nvidia-current-361.42-3.mga6.nonfree
nvidia-current-kernel-4.4.6-desktop-1.mga6-352.79-12.mga6.nonfree
dkms-nvidia-current-361.42-3.mga6.nonfree

[gohlip]

One thing we haven't tried.
How about booting into 4.4.5 instead (boot directly - not sym-link).
Looks like, from your output, the graphics for this seems okay.

If that works, we can remove (completely) 4.6.0 and then reinstall that again.

ps: I have M6, updated all without problem, but on grub2 and uefi.
Just that after installation, it won't boot without modification to its grub.cfg.
After updates, grub.cfg boots well on its own.

[rolf]

One thing we haven't tried.
How about booting into 4.4.5 instead

I mentioned doing that

when I boot the original kernel, the behavior is the same

Looks like, from your output, the graphics for this seems okay.

Where do you see that? Double-checking:

Code: Select all

[root@p8z68 rolf]# cat /mnt/disk/var/log/dmesg | grep Command
[    0.000000] Command line: BOOT_IMAGE=desktop_4.4.5-1.mga6 root=UUID=131bba44-fab2-42ff-93f7-007934a41181 noiswmd nokmsboot resume=UUID=70a32201-ef9f-47ec-a84c-38aba7d010f3 vga=788
[root@p8z68 rolf]# tail -n 20 /mnt/disk/var/log/Xorg.0.log
[    82.240] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[    82.246] (II) Module glx: vendor="X.Org Foundation"
[    82.246]    compiled for 1.18.3, module version = 1.0.0
[    82.246]    ABI class: X.Org Server Extension, version 9.0
[    82.246] (==) AIGLX enabled
[    82.246] (II) LoadModule: "nvidia"
[    82.246] (WW) Warning, couldn't open module nvidia
[    82.246] (II) UnloadModule: "nvidia"
[    82.246] (II) Unloading nvidia
[    82.246] (EE) Failed to load module "nvidia" (module does not exist, 0)
[    82.246] (EE) No drivers available.
[    82.246] (EE)
Fatal server error:
[    82.246] (EE) no screens found(EE)
[    82.246] (EE)
Please consult the Mageia support
         at http://bugs.mageia.org
 for help.
[    82.246] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[    82.246] (EE)
[root@p8z68 rolf]#

I have M6, updated all without problem

Do you use dkms-nvidia?

[gohlip]

Do you use dkms-nvidia?

Yes. And it works.
At this stage, I would suggest you either chroot to fix everything or just reinstall.
Sorry, I think I've reached the limit of my erh... ability.
But good luck.

[doktor5000]

At this stage, I would suggest you either chroot to fix everything or just reinstall.

What do you mean by everything?

Even if dkms-nvidia work and the kernel module would be present and loadable, you still need the respective X driver - so please uninstall the nvidia packages, and let drakx11 reinstall them, that's your easiest bet. Otherwise it would be really helpful to see the output of

Code: Select all

rpm -qa|sort|grep nvidia

[rolf]

First of all, my aim was to find out what went wrong with this development version and posted to this Testing forum, as I implied, earlier.

I have done `urpmi -v --auto-update` and saw dkms-nvidia at work, again. I'll attach a text file of console output of that segment of the update. After the update, re-booting to the default kernel gives the same result: no video, log reports

Code: Select all

[   164.216] (II) LoadModule: "nvidia"
[   164.216] (WW) Warning, couldn't open module nvidia
[   164.216] (II) UnloadModule: "nvidia"
[   164.216] (II) Unloading nvidia
[   164.216] (EE) Failed to load module "nvidia" (module does not exist, 0)
[   164.216] (EE) No drivers available.

As requested,

Code: Select all

[root@p8z68 /]# rpm -qa|sort|grep nvidia
dkms-nvidia-current-361.42-10.mga6.nonfree
nvidia-current-doc-html-361.42-10.mga6.nonfree
nvidia-current-kernel-4.4.5-desktop-1.mga6-352.79-11.mga6.nonfree
nvidia-current-kernel-4.4.6-desktop-1.mga6-352.79-12.mga6.nonfree
nvidia-current-kernel-4.6.0-desktop-0.rc6.1.mga6-361.42-6.mga6.nonfree
x11-driver-video-nvidia-current-361.42-10.mga6.nonfree

and, since I think dkms can provide clues, sometimes,

Code: Select all

[root@p8z68 /]# dkms status|sort
nvidia-current, 352.79-2.mga6.nonfree, 4.4.5-desktop-1.mga6, x86_64: installed-binary from 4.4.5-desktop-1.mga6
nvidia-current, 352.79-2.mga6.nonfree, 4.4.6-desktop-1.mga6, x86_64: installed-binary from 4.4.6-desktop-1.mga6
nvidia-current, 361.42-10.mga6.nonfree, 4.6.0-desktop-0.rc6.2.mga6, x86_64: installed
nvidia-current, 361.42-3.mga6.nonfree, 4.6.0-desktop-0.rc6.1.mga6, x86_64: installed-binary from 4.6.0-desktop-0.rc6.1.mga6

plus bonus, no extra charge:

Code: Select all

[root@p8z68 /]# rpm -qa | grep kernel|sort
kernel-desktop-4.4.5-1.mga6-1-1.mga6
kernel-desktop-4.4.6-1.mga6-1-1.mga6
kernel-desktop-4.6.0-0.rc6.1.mga6-1-1.mga6
kernel-desktop-4.6.0-0.rc6.2.mga6-1-1.mga6
kernel-desktop-devel-4.4.5-1.mga6-1-1.mga6
kernel-desktop-devel-4.4.6-1.mga6-1-1.mga6
kernel-desktop-devel-4.6.0-0.rc6.1.mga6-1-1.mga6
kernel-desktop-devel-4.6.0-0.rc6.2.mga6-1-1.mga6
kernel-desktop-devel-latest-4.6.0-0.rc6.2.mga6
kernel-desktop-latest-4.6.0-0.rc6.2.mga6
kernel-firmware-20160409-1.mga6
kernel-source-4.4.5-1.mga6-1-1.mga6
kernel-source-4.4.6-1.mga6-1-1.mga6
kernel-source-4.6.0-0.rc6.1.mga6-1-1.mga6
kernel-source-4.6.0-0.rc6.2.mga6-1-1.mga6
kernel-source-latest-4.6.0-0.rc6.2.mga6
kernel-userspace-headers-4.6.0-3.mga6
nvidia-current-kernel-4.4.5-desktop-1.mga6-352.79-11.mga6.nonfree
nvidia-current-kernel-4.4.6-desktop-1.mga6-352.79-12.mga6.nonfree
nvidia-current-kernel-4.6.0-desktop-0.rc6.1.mga6-361.42-6.mga6.nonfree

[rolf]

I ran drakx11 in a chroot from MGA5 and now it has booted to X with

Code: Select all

bash-4.3$ /sbin/lsmod | grep nv
nvidia_modeset        741376  7
nvidia              10072064  292 nvidia_modeset
drm                   360448  5 i915,drm_kms_helper,nvidia
nvram                  16384  0
bash-4.3$ uname -r
4.6.0-desktop-0.rc6.2.mga6

and no complaints in Xorg.0.log.
Thanks.

[laidlaws]

Do you know what you fixed? Since recent updates, I am unable to start X, and get dropped to a "fixit" shell to run XFdrake. Messages in a console say that one of the xorg packages is unknown (sorry, didn't write down which) and dkms can't be run.

[rolf]

I tried to move slowly in order to detect what the problem was but, in the end, running drakx11 seemed to fix it and I don't know why.
There is a difference in that I had no shell, no video.
What the unknown package is would be interesting.
Cauldron, I think, can be in an unworkable state, from time to time, and waiting for the proper package(s) to be built and updated solves some problems.