| Author |
Message |
thunderrd
|
|
Post subject: m-a a-i error after d-u, nvidia upgrade
 Posted: 01.02.2013, 17:07
|
|
Joined: 2010-09-12
Posts: 8
Status: Offline
|
|
My latest d-u updated the kernel and nvidia drivers, but upon reboot dmakms fails to build the nvidia stuff. The machine successfully boots to the new kernel, but the nvidia driver isn't used, although it is installed.
Following the manual, 'm-a a-i nvidia-kernel-source' gives me the following error:
Code:
'Bad luck, the kernel headers for the target kernel version could not be found and you did not specify other valid kernel headers to use.'
The /usr/src symlink points to the new kernel:
Code:
thunderrd@Q6600:~$ ls -l /usr/src
total 13268
lrwxrwxrwx 1 root src 39 Feb 1 23:10 linux -> linux-headers-3.7-5.slh.1-aptosid-amd64
drwxr-xr-x 3 root root 4096 Sep 27 2011 linux-headers-3.0-0.slh.3-aptosid-amd64
drwxr-xr-x 20 root root 4096 Sep 12 22:46 linux-headers-3.5-3.slh.3-aptosid-amd64
drwxr-xr-x 20 root root 4096 Feb 1 23:18 linux-headers-3.7-5.slh.1-aptosid-amd64
drwxr-xr-x 2 root src 4096 Sep 12 23:11 modules
drwxr-xr-x 2 root root 4096 Feb 1 22:27 nvidia-304.64
-rw-r--r-- 1 root src 4720834 Sep 12 23:11 nvidia-kernel-3.5-3.slh.3-aptosid-amd64_304.43-1+3.5-13_amd64.deb
-rw-r--r-- 1 root root 8842154 Jan 8 21:50 nvidia-kernel.tar.bz2
Apt-get reports that the newest image and headers, as well as the nvidia packages, are installed:
Code:
root@Q6600:/home/thunderrd# apt-get install nvidia-kernel-source nvidia-kernel-common nvidia-glx dmakms linux-image-aptosid-amd64 linux-headers-aptosid-amd64
Reading package lists... Done
Building dependency tree
Reading state information... Done
dmakms is already the newest version.
linux-image-aptosid-amd64 is already the newest version.
linux-headers-aptosid-amd64 is already the newest version.
nvidia-kernel-common is already the newest version.
nvidia-glx is already the newest version.
nvidia-kernel-source is already the newest version.
0 upgraded, 0 newly installed, 0 to remove and 3 not upgraded.
I will admit it has been several months since my last d-u, but I have not seen this problem in 3 years of running aptosid, so I've done something wrong. I wish I knew what it was  |
|
|
| |
|
|
|
 |
|
thunderrd
|
|
|
Post subject: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 04.02.2013, 03:25
|
|
Joined: 2010-09-12
Posts: 8
Status: Offline
|
|
Sorry to bump, but can anyone shed any light on this error?
I'm confused as to why the kernel headers cannot be found when the symlink clearly points to them.
Thanks |
|
|
| |
|
|
|
|
|
piper
|
|
|
Post subject: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 05.02.2013, 00:07
|
|
Moderator
Joined: 2010-09-11
Posts: 467
Location: cheektowaga, ny
Status: Offline
|
|
|
|
|
|
|
thunderrd
|
|
|
Post subject: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 05.02.2013, 13:05
|
|
Joined: 2010-09-12
Posts: 8
Status: Offline
|
|
Thanks Piper, and yes, that is exactly what I followed.
This system has been running for so long (several years)with the proprietary driver that I am surprised that I have the problem now. After the 3rd step of the instructions, I get the error in the OP. It seems to indicate some problem finding the kernel sources or headers, although the symlink is correct, as you can see. For some reason, the blob installation can't do its magic with 3.7-5, and so the .deb file is never created. I have probably updated dozens of kernels/nvidia versions without this issue in the past.
The blob is the only option for me, as it is needed for gpu folding@home. |
|
|
| |
|
|
|
|
|
towo
|
|
|
Post subject: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 05.02.2013, 14:01
|
|
Joined: 2010-09-13
Posts: 488
Location: Pößneck / Thüringen
Status: Offline
|
|
Use dkms instead of m-a.
module-assistant is a dead horse. |
|
|
| |
|
|
|
|
|
thunderrd
|
|
|
Post subject: Re: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 05.02.2013, 17:13
|
|
Joined: 2010-09-12
Posts: 8
Status: Offline
|
|
|
towo wrote:
Use dkms instead of m-a.
module-assistant is a dead horse.
Thank you, towo,
I'm not too clear on how that is done. Does that take the place of dmakms? If so, do I delete dmakms and install dkms in its place? Is there anything special to be done after that? |
|
|
| |
|
|
|
|
|
towo
|
|
|
Post subject: RE: Re: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 05.02.2013, 17:18
|
|
Joined: 2010-09-13
Posts: 488
Location: Pößneck / Thüringen
Status: Offline
|
|
| apt-get purge dmakms nvidia-kernel-source && apt-get install nvidia-kernel-dkms |
|
|
| |
|
|
|
|
|
thunderrd
|
|
|
Post subject: RE: Re: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 06.02.2013, 14:33
|
|
Joined: 2010-09-12
Posts: 8
Status: Offline
|
|
Yep, thank you towo, that did the trick.
Nvidia-kernel-dkms was already installed; simply purging the other packages removed the error. I see now that nvidia-glx depends on nvidia-kernel-dkms.
Thanks again |
|
|
| |
|
|
|
|
|
mz
|
|
|
Post subject: RE: Re: RE: m-a a-i error after d-u, nvidia upgrade
Posted: 10.02.2013, 16:45
|
|
Joined: 2010-09-11
Posts: 53
Status: Offline
|
|
| So it appears that the nvidia problems with the 3.7 kernel are corrected? I have been holding back at 3.6-6. |
|
|
| |
|
|
|
|
|
drb
|
|
|
Post subject:
Posted: 11.02.2013, 07:50
|
|
Joined: 2010-09-11
Posts: 94
Status: Offline
|
|
I think the above is a different issue.
Code:
Host/Kernel/OS "hypnos32" running Linux 3.7-6.slh.2-aptosid-686 i686 [ sidux 2010-01 Ύπνος - kde-full - (201006131622) ]
CPU Info 2x AMD Athlon 64 X2 Dual Core 5600+ 1024 KB cache flags( sse3 ht nx lm svm ) clocked at [ 1000.000 MHz ]
Videocard NVIDIA GT218 [GeForce 210] X.Org 1.12.4 [ 1680x1050@60.0hz ]
Network cards NVIDIA MCP55
Processes 170 | Uptime 6min | Memory 646.1/3041.9MB | HDD WDC WD3200AAKS-0,WDC WD5000AAKS-0,WDC WD6400AAKS-0,WDC WD15EADS-00P Size 2960GB (34%used) | GLX Renderer GeForce 210/PCIe/SSE2/3DNOW! | GLX Version 3.3.0 NVIDIA 313.18 | Client Shell | Infobash v3.46
|
|
|
| |
|
|
|
|
|
k1tt3h
|
|
|
Post subject:
Posted: 10.05.2013, 10:21
|
|
Joined: 2012-11-18
Posts: 15
Location: lap, window sill, or top of refrigerator
Status: Offline
|
|
| I am also experiencing issues with module-assistant almost exactly as thunderrd. Same ``bad luck'' message when I know
Code:
m-a prepare
tells me truth by stating the headers are there.
I worked around this by going to tty1, changing to init 3, and running
Code:
m-a -k /usr/src/linux-headers-3.9-1.slh.1-aptosid-amd64 build nvidia-kernel-source
m-a -k /usr/src/linux-headers-3.9-1.slh.1-aptosid-amd64 install nvidia-kernel-source
I removed the nvidia-kernel-source entry in /etc/default/dmakms, since to me it seemed the automation at shutdown was not working.
It is really simple for me to do the above when I am already running the d-u, but for the sake of testing and participating in Aptosid's growth, should I follow towo's recommendation to use dkms instead or is module-assistant sticking around  It seems like thunderdd and I both are in need of some definitive answer on this. I appreciate all the effort put forth by the Aptosid team, and I truly enjoy using it. I don't like it when things break, but I sure do love to try and ``fix'' them. Let me know or not, I am still satisfied. :3 I thought to at least share my experience with m-a and how I got it to work. |
|
|
| |
|
|
|
|
|
DonKult
|
|
|
Post subject:
Posted: 10.05.2013, 12:23
|
|
Team Member
Joined: 2010-09-02
Posts: 420
Status: Offline
|
|
dmakms itself is nice and curtsy, problem is module-assistant which seems to be in maintenance limbo currently as a patch fixing this exists for a while but isn't picked up so far ( http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=697269 ) and we really don't like adopting patches which aren't accepted by upstream.
The right thing to do™ is using m-a, but the rest of the world likes dkms better currently for unknown reasons. I am playing with dkms currently, but I am not really a big fan of building and installing stuff outside the knowledge of APT (surprise surprise).
So really no definite answer for now, I am afraid. |
_________________
MfG. DonKult
"I never make stupid mistakes. Only very, very clever ones." ~ The Doctor
|
| |
|
|
|
|
|
towo
|
|
|
Post subject:
Posted: 10.05.2013, 15:52
|
|
Joined: 2010-09-13
Posts: 488
Location: Pößneck / Thüringen
Status: Offline
|
|
|
Quote:
The right thing to do™ is using m-a
Nice joke, debian itself is focusing on dkms.
No one realy cares about module-assistant nowdays, since all maintainers have switched their sources to dkms.
Maybe that's the point, why m-a is broken since more than one year. |
|
|
| |
|
|
|
|
|
hubi
|
|
|
Post subject:
Posted: 10.05.2013, 18:42
|
|
Moderator
Joined: 2010-09-11
Posts: 220
Location: Vienna (AT)
Status: Offline
|
|
|
DonKult wrote:
The right thing to do™ is using m-a, but the rest of the world likes dkms better currently for unknown reasons. I am playing with dkms currently, but I am not really a big fan of building and installing stuff outside the knowledge of APT (surprise surprise).
I have one laptop with an nvidia card, and I switched to dkms approx. a year ago, no reason to complain. My experience is, the dkms versions of the nvidia blob in Debian are very stable (m-a was in a very bad state when I switched).
hubi |
_________________
Tilos Rádió Budapest
|
| |
|
|
|
|
|
k1tt3h
|
|
|
Post subject:
Posted: 11.05.2013, 04:13
|
|
Joined: 2012-11-18
Posts: 15
Location: lap, window sill, or top of refrigerator
Status: Offline
|
|
|
DonKult wrote:
The right thing to do™ is using m-a
:3, for once I am doing it right. 
Quote:
I am not really a big fan of building and installing stuff outside the knowledge of APT.
Neither am I.
Quote:
So really no definite answer for now, I am afraid.
Roger dodger. My fingers are crossed that the patches are pushed upstream and I can go back to a by-the-book install.
Thanks for the reply. |
|
|
| |
|
|
|
|
|
|
