[Dell Studio XPS 1340] nouveau driver causes kernel panic when entering graphics mode

This change was made by a bot.

Installing nvidia-current allows this system to boot without needing the nouveau blacklisting hack (because nv overtakes nouveau, I presume).

Note that this system has a long history of nouveau-related problems, see for instance https://bugs.launchpad.net/ubuntu/+source/linux/+bug/754711.

Is this the first time Ubuntu has been installed on this system? If so, it would be good to know if there is an earlier release or kernel version that did not exhibit this bug.

Can you test the following kernels and see if they also have this bug?

v3.5.7: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.5.7-quantal/
v3.6 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.6-quantal/
v3.7 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.7-raring/
v3.8-rc1: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.8-rc1-raring/

You don't have to test every kernel, just up until the kernel that first has this bug.

Thanks in advance!

Hi Joseph,

Wow, this is a new one :) This system worked fine with Ubuntu 12.10, it only started exhibiting this particular problem on Raring.

I'll test the kernels you suggest and update accordingly.

Thanks!

OK, here are results of the testing:

v3.5.7: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.5.7-quantal/ This kernel works OK.
v3.6 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.6-quantal/ This kernel works OK.
v3.7 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.7-raring/ This kernel works OK.
v3.8-rc1: http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.8-rc1-raring/ This kernel panics as originally described.

So apparently the problem was introduced at some point after 3.7 final.

I started a kernel bisect between v3.7 final and v3.8-rc1. The kernel bisect will require testing of about 7-10 test kernels.

I built the first test kernel, up to the following commit:
6be35c700f742e911ecedd07fcc43d4439922334

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Thanks in advance

Hi Joseph!

The first test kernel tested as GOOD, the system boots up fine and loads the nouveau module.

The tested kernel is: Linux version 3.7.0-030700-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201301301508 SMP Wed Jan 30 20:09:35 UTC 2013

I built the next test kernel, up to the following commit:
027c0a6af42efa4f2f6034421349bd26a3ca4923

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi!

Second test kernel is also GOOD, the system boots without any issues and nouveau is loaded.

Linux version 3.7.0-030700-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201301311229 SMP Thu Jan 31 17:31:47 UTC 2013

I built the next test kernel, up to the following commit:
ea88eeac0cb8328014b53d80ca631e8dc0dc18dc

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hello!

OK, this kernel is also GOOD, system boots up fine and nouveau is happily loaded.

Linux version 3.7.0-030700-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201301311522 SMP Thu Jan 31 20:23:45 UTC 2013

Hi Joseph,

Apologies, I need to amend my last comment, this kernel is BAD, the report was based on my ability to SSH into the system, but I wasn't looking at the screen at the moment. The screen shows that nouveau crashed, I'm attaching to this coment a screenshot showing the crash and the kernel version (3.7.0-030700-generic #201301311522) is clearly visible here.

Hello, so based on my previous comment I went back and retested the two first kernels, their good results stand, so we have:

6be35c700f742e911ecedd07fcc43d4439922334 - GOOD
027c0a6af42efa4f2f6034421349bd26a3ca4923 - GOOD
ea88eeac0cb8328014b53d80ca631e8dc0dc18dc - BAD :(

I built the next test kernel, up to the following commit:
f562146a3daf6aa0bbf2a1bc4b6b7da031ed5dcd

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

Latest kernel is BAD, it shows the kernel panic and the GUI doesn't appear. I'm able to SSh into the system though, here's the kernel string:

Linux version 3.7.0-030700-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302011757 SMP Fri Feb 1 22:58:12 UTC 2013

I built the next test kernel, up to the following commit:
f7960736d0b8647fd149687d15deb79e47122d69

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

Latest test kernel is BAD, it also produces the panic on the laptop's screen.

Version number as follows:

Linux version 3.7.0-030700rc4-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302051502 SMP Tue Feb 5 20:03:57 UTC 2013

im also getting this on my macbook 5,1 im have been trying for the past few weeks but neither the i386 or x64 work just ends in the same kernel panic as you

I built the next test kernel, up to the following commit:
9a76b1c68ffb843ad6f31b746141365fdf262b97

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

Latest test kernel is BAD again, I got the kernel panic on boot.

Linux version 3.7.0-030700rc2-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302121334 SMP Tue Feb 12 18:35:21 UTC 2013

Thanks!

I built the next test kernel, up to the following commit:
397fe15715ef1457d89f52666d0e249eb5eae64c

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

Yay, string of bad kernels! This is the 7th tested kernel, the last 5 have been bad.

Latest is BAD again, I got a nice kernel panic that has a lot of nouveau-related data.

Linux version 3.7.0-030700rc2-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302131529 SMP Wed Feb 13 20:30:04 UTC 2013

Thanks!

I built the next test kernel, up to the following commit:
d7d4eeddb8f72342f70621c4b3cb718af9361712

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi,

Sorry for the delay in testing this.

Yay, we got a good kernel!

This one boots fine and the GUI and desktop are visible and usable. Nouveau is loaded without problems.

Linux version 3.6.0-030600rc7-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302140913 SMP Thu Feb 14 14:14:53 UTC 2013

I built the next test kernel, up to the following commit:
36ec8f877481449bdfa072e6adf2060869e2b970

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

This kernel also seems good, system booted up OK with the desktop visible and noveau loaded nicely.

Linux version 3.6.0-030600rc7-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302191655 SMP Tue Feb 19 21:56:33 UTC 2013

Thanks!

I built the next test kernel, up to the following commit:
1d508706ea848e32ff20bb311f4325896c6eb7b9

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

This last kernel is BAD, I got the nasty panic and the GUI did not appear.

Linux version 3.7.0-030700rc2-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302211159 SMP Thu Feb 21 17:01:06 UTC 2013

Thanks!

I built the next test kernel, up to the following commit:
29b99b4841cc15854475a299aceaa018e7a468f1

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph,

This last kernel is BAD, I get a kernel panic when booting and the GUI does not appear.

Linux version 3.7.0-030700rc2-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302251707 SMP Mon Feb 25 22:08:25 UTC 2013

I built the next test kernel, up to the following commit:
4d47e4f57f0a821f5ba84b8e101ee7e44e4ffae6

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

Hi Joseph, the latest test kernel is GOOD, the desktop came up fine and nouveau is loaded.

Linux version 3.6.0-030600rc7-generic (jsalisbury@gomeisa) (gcc version 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1) ) #201302261425 SMP Tue Feb 26 19:26:46 UTC 2013

I built the next test kernel, up to the following commit:
c2fb7916927e989ea424e61ce5fe617e54878827

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

This is close to the last kernel to test before we identify the offending commit.

Hi Joseph,

Glad to hear we're near the end!

The last kernel is BAD, I got the kernel panic and no GUI.

Linux 200912-4906 3.7.0-030700rc2-generic #201302271112 SMP Wed Feb 27 16:13:53 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

I built the next test kernel, up to the following commit:
29de6ce574870a0d3fd157afdbf51c0282e2bf63/

The test kernel can be downloaded from:
http://people.canonical.com/~jsalisbury/lp1109309

Can you test that kernel and report back if it has the bug or not. I will build the next test kernel based on your test results.

This is close to the last kernel to test before we identify the offending commit.

Hi Joseph,

Latest test kernel is GOOD, graphical desktop came up and works fine.

Linux version 3.6.0-030600rc7-generic (root@gomeisa) (gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5) ) #201303071326 SMP Thu Mar 7 18:28:26 UTC 2013

c2fb7916927e989ea424e61ce5fe617e54878827 is the first bad commit:

commit c2fb7916927e989ea424e61ce5fe617e54878827
Merge: 29de6ce 6f0c058
Author: Daniel Vetter <email address hidden>
Date: Mon Oct 22 14:34:51 2012 +0200

    Merge tag 'v3.7-rc2' into drm-intel-next-queued

    Linux 3.7-rc2

This is a pretty large patch set, and cannot be reverted easily. I will probably need to contact upstream about this.

Before I contact upstream, can you test the latest mainline kernel, just to confirm it's not fixed yet:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc2-raring/

Hi Joseph,

Hm,, results with 3.9 are inconclusive :/ As it pertains to this bug, I *don't* get a kernel panic first thing after booting. But I see the following:

- On boot-up, the system which is set to auto-login is unable to do so. After a while it dumps me back to lightdm. If I try to login, I see a plain background with movable cursor, after a while it goes back to lightdm, looking like compiz/unity are unable to start.

- I see a bunch of messages like this on dmesg:

[ 133.045072] nouveau E[ PGRAPH][0000:03:00.0] TRAP_M2MF NOTIFY
[ 133.047648] nouveau E[ PGRAPH][0000:03:00.0] TRAP_M2MF 00304041 00004360 00000000 06000432
[ 133.049065] nouveau E[ PGRAPH][0000:03:00.0] TRAP
[ 133.049065] nouveau E[ PGRAPH][0000:03:00.0] ch 2 [0x000fb35000 Xorg[1394]] subc 0 class 0x5039 mthd 0x0328 data 0x00000000
[ 133.049065] nouveau E[ PFB][0000:03:00.0] trapped write at 0x00422f0000 on channel 0x0000fb35 [Xorg[1394]] PGRAPH/PROP/DST2D reason: PAGE_NOT_PRESENT
[ 133.057807] nouveau E[ PGRAPH][0000:03:00.0] TRAP_TPDMA_2D - TP 0 - Unknown fault at address 00422f0000
[ 133.060373] nouveau E[ PGRAPH][0000:03:00.0] TRAP_TPDMA_2D - TP 0 - e0c: 00000000, e18: 00000000, e1c: 00000000, e20: 00000011, e24: 0c030000
[ 133.061749] nouveau E[ PGRAPH][0000:03:00.0] TRAP
[ 133.061749] nouveau E[ PGRAPH][0000:03:00.0] ch 2 [0x000fb35000 Xorg[1394]] subc 2 class 0x502d mthd 0x060c data 0x00000320
[ 133.061749] nouveau E[ PFB][0000:03:00.0] trapped write at 0x0042416b00 on channel 0x0000fb35 [Xorg[1394]] PGRAPH/PROP/DST2D reason: PAGE_NOT_PRESENT

All in all, it's hard to tell; I'd say the original bug is no longer present but nouveau has some other bugs that cause this system to still be unusable.

It might be worthwhile to test the v3.9-rc1 kernel:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc1-raring/

Hi Joseph,

3.9.0-rc1 behaves the same as 3.9.0-rc2, see my prior comment for behavior. Kernel version as follows:

Linux version 3.9.0-030900rc1-generic (apw@gomeisa) (gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5) ) #201303060659 SMP Wed Mar 6 12:00:04 UTC 2013

Can you perform one additional upstream test? The v3.9-rc4 kernel is now available. I will ping upstream if this bug still exists in that kernel:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.9-rc4-raring/

Thanks again for all the testing you've done.

Hi Joseph,

Interestingly, 3.9.0-rc4 does work well; there's no kernel panic on startup, and I get to a usable desktop. Nouveau is loaded fine.

Here's the version I tested:

Linux version 3.9.0-030900rc4-generic (apw@gomeisa) (gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5) ) #201303232035 SMP Sun Mar 24 00:36:21 UTC 2013

Hi. I think iv got the same problem with 13.04. I can use the kernel 3.5 but not the default kernel 3.8. Iv got the same laptop, but not the same graphic card: here it is an ATI.

lspci | grep VGA
01:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI RV635/M86 [Mobility Radeon HD 3670]

Then it looks like it is not a problem with Nouveau. Unfortunatly i dont have skill to help you more.

Upstream did not have any ideas on what may have introduced this. The next best step may be to perform a Reverse bisect since we know the bug is fixed in 3.9.0-rc4.

Before starting the reverse bisect, can you test the latest 3.8 stable kernel, to confirm it still exists there:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.8.10-raring/

Upgrading Quantal to Raring using the update manager.
The upgrade went flawlessly. The first reboot was OK.
But since the next reboot, I have a kernel oops when nouveau is loaded. From there the system is blocked (continues to load drivers for Wifi but no display manager starting).

Hardware is: Lenovo W520 which has Intel Core i7 quadcore, 8 GB RAM, Intel + nvidia GPU (hybrid).
Ubuntu running in 32bit.

I tried the v3.9 stable and v3.9 rc4 as suggested earlier. None of them works for me.

The v3.10-rc4 kernel is now available. Can you see if that kernel still exhibits this bug:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.10-rc4-saucy/

Joseph,
I tried the 3.10-rc4-saucy kernel as advised.
The system boots. During the boot, the splash screen does not appear (screen remains black after the purple background of grub until the display manager kicks in). There is no kernel oops and Wifi is operational.
BUT...
It is impossible to properly shutdown the machine. When I shutdown from X, I get a black screen, lots of HDD activity and then nothing but it stays up.
I tried from a text console. There I get some messages, the last one being system is going for HALT.

Might be worth to know: I installed nouveau-firmware 20091212-0ubuntu1, linux-firmware 1.106 and linux-firmware-nonfree 1.14ubuntu1. Also, I uninstalled Oracle VirtualBox drivers.

However, in dmesg, I see the following worrying messages:

1)
[ 2.882217] [drm] Wrong MCH_SSKPD value: 0x16040307
[ 2.882218] [drm] This can cause pipe underruns and display issues.
[ 2.882219] [drm] Please upgrade your BIOS to fix this.

2)
[ 30.815712] WARNING: at /home/apw/COD/linux/drivers/gpu/drm/i915/intel_display.c:8207 intel_modeset_stage_output_state.isra.28+0x364/0x3b0 [i915]()
[ 30.815714] Modules linked in: parport_pc ppdev rfcomm bnep ext2 dm_crypt joydev arc4 iwldvm mac80211 intel_powerclamp coretemp snd_hda_codec_conexant uvcvideo snd_hda_intel snd_hda_codec snd_hwdep videobuf2_vmalloc snd_pcm videobuf2_memops snd_page_alloc thinkpad_acpi kvm iwlwifi psmouse videobuf2_core nvram snd_seq_midi snd_seq_midi_event videodev snd_rawmidi snd_seq snd_seq_device snd_timer btusb lp cdc_mbim cdc_acm cdc_wdm snd bluetooth parport lpc_ich mac_hid soundcore cdc_ncm serio_raw cfg80211 usbnet microcode tpm_tis hid_generic usbhid hid nouveau e1000e crc32_pclmul ptp i915 aesni_intel aes_i586 firewire_ohci xts mxm_wmi lrw ttm pps_core ahci i2c_algo_bit gf128mul video firewire_core drm_kms_helper ablk_helper libahci sdhci_pci cryptd drm sdhci wmi crc_itu_t
[ 30.815743] CPU: 2 PID: 1403 Comm: Xorg Not tainted 3.10.0-031000rc4-generic #201306020443
[ 30.815744] Hardware name: LENOVO 428449G/428449G, BIOS 8BET58WW (1.38 ) 07/06/2012
[ 30.815746] f87f8f30 f87f8f30 ea6b9dc8 c15fb5e8 ea6b9df0 c1038da4 c17d593f f87f8f30
[ 30.815750] 0000200f f87b79d4 f87b79d4 e9d69a00 f5f7a980 e9d69a09 ea6b9e00 c1038de2
[ 30.815753] 00000009 00000000 ea6b9e44 f87b79d4 0000044c 0000046a 00000048 0000000a
[ 30.815757] Call Trace:
[ 30.815763] [<c15fb5e8>] dump_stack+0x16/0x18
[ 30.815766] [<c1038da4>] warn_slowpath_common+0x64/0x80
[ 30.815776] [<f87b79d4>] ? intel_modeset_stage_output_state.isra.28+0x364/0x3b0 [i915]
[ 30.815784] [<f87b79d4>] ? intel_modeset_stage_output_state.isra.28+0x364/0x3b0 [i915]
[ 30.815786] [<c1038de2>] warn_slowpath_null+0x22/0x30
[ 30.815794] [<f87b79d4>] intel_modeset_stage_output_state.isra.28+0x364/0x3b0 [i915]
[ 30.815801] [<f87b7c87>] ? intel_set_config_compute_mode_changes.isra.37+0x107/0x170 [i915]
[ 30.815811] [<f87c11c4>] intel_crtc_set_config+0x114/0x300 [i915]
[ 30.815816] [<c12d8e60>] ? copy_to_user+0x40/0x60
[ 30.815825] [<f8543e95>] drm_mode_set_config_internal+0x25/0x50 [drm]
[ 30.815829] [<f8485e2d>] drm_fb_helper_restore_fbdev_mode+0x2d/0x50 [drm_kms_helper]
[ 30.815841]...

Hi again,
I did a quick test of different kernel versions from the PPA site:
- 3.8.0.23 (official kernel of 13.04)
-- no splash screen
-- sound OK, wifi OK
-- seen: error in intel driver related to drm
-- seen in a second try: error of vmalloc during swapon followed by a number of vmalloc errors
-- switching to tty1-6 does not work, coming back to X works
-- halting the system does not work

- 3.8.13
-- no splash screen
-- no sound, no wifi
-- vmalloc error (as seen in 3.8.0.23)
-- halting is not OK

- 3.9
-- no splash screen
-- sound OK, wifi OK
-- error intel/drm as in 3.8.0.23
-- halting is not OK

If there's any test I can run to help you investigate, don't hesitate.
(I've copies of the dmesg logs for all those runs.)

Olivier

Hi Joseph,

In preparation for testing the mainline kernel you suggested, I started by launching a Raring install (using the released 13.04 image with 3.8.0-19.29 kernel). I expected this to bomb at install time, and having to add the kernel parameter as originally reported.

To my surprise, it installed and booted without a hitch, using the nouveau driver.

Linux 200912-4906 3.8.0-19-generic #29-Ubuntu SMP Wed Apr 17 18:16:28 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

I guess, then, that whatever it is that fixes this problem was included in this released Ubuntu kernel.

Just to be thorough, I did install the 3.10-rc4-saucy kernel you suggested. This also boots fine and I can reach a usable desktop.

Linux 200912-4906 3.10.0-031000rc4-generic #201306020443 SMP Sun Jun 2 08:43:56 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

Let me know how you'd like to proceed, I think we could set this to Fix Released, at least for the XPS 1340. I'd like to know more precisely where the fix came from but realistically we may not have a lot of time to do that.

Thanks for the update, Daniel. I'll set this to fix released for now. Thanks for the work you put in on this bug.