No X with nvidia after upgrade from 3.2.0-38 to 3.2.0-39

[Steven "Kreuvf" Koenig]

lsb_release -rd
Description: Ubuntu 12.04.2 LTS
Release: 12.04

uname -a
Linux samus 3.2.0-38-generic #61-Ubuntu SMP Tue Feb 19 12:20:02 UTC 2013 i686 athlon i386 GNU/Linux

After I upgraded from 3.2.0-38 to 3.2.0-39 upon restart X failed to start. The nvidia module does not work anymore. I also tried the quantal kernel 3.5.0-26: same behaviour. The (imho) interesting messages in dmesg are the following:
[ 28.302848] nvidia 0000:01:00.0: can't find IRQ for PCI INT A; please try using pci=biosirq
[ 28.302866] NVRM: Can't find an IRQ for your NVIDIA card!
[ 28.302871] NVRM: Please check your BIOS settings.
[ 28.302876] NVRM: [Plug & Play OS] should be set to NO
[ 28.302881] NVRM: [Assign IRQ to VGA] should be set to YES
[ 28.302900] nvidia: probe of 0000:01:00.0 failed with error -1
[ 28.319027] PCI: setting IRQ 12 as level-triggered
[ 28.319046] snd_ens1371 0000:00:0d.0: found PCI INT A -> IRQ 12
[ 28.319462] NVRM: The NVIDIA probe routine failed for 1 device(s).
[ 28.319473] NVRM: None of the NVIDIA graphics adapters were initialized!

I attached an Xorg.log for 3.2.0-38 and 3.2.0-39, lspci output for 3.2.0-38, 3.2.0-39 and 3.5.0-26, the output of dmesg for 3.2.0-38 and 3.2.0-39 and the output of glxinfo of nvidia-settings. If you need anything else, please tell me.

[Brad Figg (brad-figg)]

This change was made by a bot.

[Joseph Salisbury (jsalisbury)]

Would it be possible for you to test the latest upstream kernel? Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Please test the latest v3.9 kernel[0] (Not a kernel in the daily directory) and install both the linux-image and linux-image-extra .deb packages.

If this bug is fixed in the mainline kernel, please add the following tag 'kernel-fixed-upstream'.

If the mainline kernel does not fix this bug, please add the tag: 'kernel-bug-exists-upstream'.

If you are unable to test the mainline kernel, for example it will not boot, please add the tag: 'kernel-unable-to-test-upstream'.
Once testing of the upstream kernel is complete, please mark this bug as "Confirmed".

Thanks in advance.

[0] http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc3-raring/

[Steven "Kreuvf" Koenig (kreuvf)]

I downloaded the following files:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc3-raring/linux-headers-3.9.0-030900rc3-generic_3.9.0-030900rc3.201303171935_i386.deb
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc3-raring/linux-headers-3.9.0-030900rc3_3.9.0-030900rc3.201303171935_all.deb
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc3-raring/linux-image-3.9.0-030900rc3-generic_3.9.0-030900rc3.201303171935_i386.deb
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc3-raring/linux-image-extra-3.9.0-030900rc3-generic_3.9.0-030900rc3.201303171935_i386.deb

I used gdebi-gtk to install the packages one by one (headers all, headers generic, image, image extra). During installation of the generic headers package I saw the following in the terminal:
 run-parts: executing /etc/kernel/header_postinst.d/dkms 3.9.0-030900rc3-generic /boot/vmlinuz-3.9.0-030900rc3-generic
 Error! Bad return status for module build on kernel: 3.9.0-030900rc3-generic (i686)
 Consult /var/lib/dkms/nvidia-173/173.14.35/build/make.log for more information.

Since the nvidia kernel module did not build I expect to get into trouble, but I will try and reply again in a few minutes. I attached the log (exactly the same output in both cases except for a timestamp).

[Steven "Kreuvf" Koenig (kreuvf)]

Running in low graphics mode for this session. ;) No output, not even the console (?) output I normally see during booting. With 3.2.0-39 and 3.5.0-26 I saw something, but not in the native resolution of the screen, with 3.9rc3 the screen remains black until bulletproof X asks what to do. When I want to log in using the console, the screen stays black. Fortunately, I was able to log in, 'beep' and shutdown blindly. That fits to 'uvesafb: vbe_init() failed with -22'.

I will attach dmesg, Xorg.log and the output of lspci -k -n -vv again. I will also add the tag 'kernel-unable-to-test-upstream' and mark this bug as 'Confirmed'.

[Steven "Kreuvf" Koenig (kreuvf)]

Since there has not been any reply for the past week(s), I wanted to say that, if somebody could prepare intermediate kernel packages from 3.2.0-38 to 3.2.0-39 I would like to help find the commit which introduced the problem.

[Steven "Kreuvf" Koenig (kreuvf)]

I found the following solution to this problem:
* BIOS -> PCI / Plug and Play Setup
 * Allocate IRQ to PCI VGA: Yes (before: No)

Kernels 3.2.0-39, 3.2.0-40, 3.5.0-26 work without problems. Kernel 3.9.0-030900rc3 fails again in uvesafb, so I do not see anything during booting:
kernel: [ 2.438172] uvesafb: Getting VBE info block failed (eax=0x4f00, err=0)
kernel: [ 2.438222] uvesafb: vbe_init() failed with -22
kernel: [ 2.438254] uvesafb: probe of uvesafb.0 failed with error -22

The Xorg.log is attached.

Some more information regarding my system:
* Mainboard: Elitegroup K7S5A
* BIOS: AMIBIOS 1.21.06
* Graphics card: GeForce FX 5200 in an AGP slot

While I can use the latest kernel without problems again, it would probably still be interesting to find out the reason why I have to change a BIOS setting with kernels younger than 3.2.0-38.

[Julian Wiedmann (jwiedmann)]

[ 0.000000] ACPI: no DMI BIOS year, acpi=force is required to enable ACPI
[ 0.000000] ACPI: Disabling ACPI support

Very likely a duplicate of bug 1161134, should be fixed with 3.2.0-41.65 in -proposed.

[Steven "Kreuvf" Koenig (kreuvf)]

This seems to be true. Starting the newer kernels will not automatically turn off the computer when shutting down, I have to do it manually after the 'System halted' message appeared.

[Steven "Kreuvf" Koenig (kreuvf)]

This bug has been fixed with kernel 3.2.0-41. Thanks!