#574614 /usr/bin/module-assistant: fails building nvidia-kernel module and interface gets confused #574614
- Package:
- module-assistant
- Source:
- module-assistant
- Submitter:
- Nicola Manini
- Date:
- 2010-05-24 10:03:08 UTC
- Severity:
- normal
While building nvidia-kernel in squeeze, the kbuild fails as follows: make[2]: Entering directory `/usr/src/modules/nvidia-kernel' NVIDIA: calling KBUILD... make CC=gcc-4.3 -C /lib/modules/2.6.32-3-686/build SUBDIRS=/usr/src/modules/nvidia-kernel modules make[3]: Entering directory `/usr/src/linux-headers-2.6.32-3-686' Building modules, stage 2. MODPOST 0 modules make[3]: Leaving directory `/usr/src/linux-headers-2.6.32-3-686' NVIDIA: left KBUILD. nvidia.ko failed to build! The above might reflect a bug of linux-kbuild-2.6.32 , or of nvidia-kernel-source, rather than of m-a. After this failure, the m-a enters the menu which proposes to VIEW Examine the build log file CONTINUE Skip and continue with the next operation STOP Stop processing the build commands After VIEWing the log, the m-a remains in a confused state, by pressing down-arrow to reach CONTINUE os STOP, some undetectable lines scroll below the fake windows, and m-a exits. This is surely a bug of m-a. It may help to know that I was running it in a xterm. I provide a snapshot of the terminal after this failure, in attachment. I also attach the full m-a log. Thanx! NM
I attach the xterm window snaphot after m-a exited, and the m-a log. All the best, Nick
Hello, Nicola Manini <nicola.manini@mi.infm.it> (19/03/2010): looks like, yeah. Looks like, too. :) I'm currently busy with X11 stuff, I'll try to kill some m-a bugs next week. Feel free to ping back if you don't hear from me by then. Mraw, KiBi.
Hi. I hope not to spam your bug report: I just want you to know I was building nvidia-kernel in my squeeze 64bit system, getting from m-a your same error (Building modules, stage 2. MODPOST 0 modules), and I solved the problem by installing nvidia-kernel-common nvidia-kernel-source nvidia-glx from sid repositories. I hope I've sent you some useful infos. Bye Marco Mattiolo
Yes, it mostly is a bug of the nvidia driver package, but there is also a m-a
related part, which expects fixing.
Indeed the solution suggested by Marco Mattiolo worked for me.
I installed nvidia-kernel-dkms from sid. This introduced me to the ease of
dkms, which makes m-a much less useful than it used to be.
For this specific install, I also needed to create a /etc/X11/xorg.conf file
with the following contents:
Section "Device"
Identifier "Default screen"
Driver "nvidia"
EndSection
Philosophical observation:
Probably stable should go for the dkms scheme for non-free drivers of this
sort, that would be even better than the pre-compiled package, and remain
more compatible with future testing.
Ciao, thanx,
Nick