[945GM] X hangs on startup

[Kevin]

Binary package hint: xserver-xorg-video-intel

When KMS is on, Plymouth appears, then disappears. X then appears to start, and then hangs with either a black screen, or a purple screen. The keyboard usually appears to be working, (caps lock light can turn on and off) however it will not switch to a VT. Looking through the logs, this is what appears in the kernel messages when the hang happens:

Mar 10 23:31:26 9876-laptop kernel: [ 110.962921] Process Xorg (pid: 1735, threadinfo ffff88002a6b6000, task ffff88001cbd96f0)
Mar 10 23:31:26 9876-laptop kernel: [ 110.962927] ffff88002a6b7b08 ffffffffff0000aa ffff88002a6b7b68 ffff88002b894000
Mar 10 23:31:26 9876-laptop kernel: [ 110.962935] <0> ffff88002b894478 00000000ffffffff ffff88002a6b7dd8 ffff88003af57000
Mar 10 23:31:26 9876-laptop kernel: [ 110.962944] <0> ffff88002a6b7c38 ffffffffa0109985 ffff88002a6b7bd8 00000003a00b0700
Mar 10 23:31:26 9876-laptop kernel: [ 110.962980] [<ffffffffa0109985>] intel_tv_detect+0xa5/0x190 [i915]
Mar 10 23:31:26 9876-laptop kernel: [ 110.962994] [<ffffffffa00d3f3e>] drm_helper_probe_single_connector_modes+0x2de/0x330 [drm_kms_helper]
Mar 10 23:31:26 9876-laptop kernel: [ 110.963005] [<ffffffff8155579e>] ? mutex_lock+0x1e/0x50
Mar 10 23:31:26 9876-laptop kernel: [ 110.963029] [<ffffffffa009d8da>] drm_mode_getconnector+0x34a/0x410 [drm]
Mar 10 23:31:26 9876-laptop kernel: [ 110.963048] [<ffffffffa0090e1a>] drm_ioctl+0x27a/0x480 [drm]
Mar 10 23:31:26 9876-laptop kernel: [ 110.963069] [<ffffffffa009d590>] ? drm_mode_getconnector+0x0/0x410 [drm]
Mar 10 23:31:26 9876-laptop kernel: [ 110.963079] [<ffffffff81083580>] ? autoremove_wake_function+0x0/0x40
Mar 10 23:31:26 9876-laptop kernel: [ 110.963088] [<ffffffff81150162>] vfs_ioctl+0x22/0xa0
Mar 10 23:31:26 9876-laptop kernel: [ 110.963094] [<ffffffff81150411>] do_vfs_ioctl+0x81/0x380
Mar 10 23:31:26 9876-laptop kernel: [ 110.963101] [<ffffffff81140132>] ? vfs_write+0x132/0x1a0
Mar 10 23:31:26 9876-laptop kernel: [ 110.963107] [<ffffffff81150791>] sys_ioctl+0x81/0xa0
Mar 10 23:31:26 9876-laptop kernel: [ 110.963114] [<ffffffff815574be>] ? do_device_not_available+0xe/0x10
Mar 10 23:31:26 9876-laptop kernel: [ 110.963123] [<ffffffff810131f2>] system_call_fastpath+0x16/0x1b
Mar 10 23:31:26 9876-laptop kernel: [ 110.963221] RSP <ffff88002a6b7ac8>

FYI, using Lucid updated as of March 11, with the 2.6.32-16 kernel.

ProblemType: Bug
Architecture: amd64
Date: Wed Mar 10 23:56:20 2010
DistroRelease: Ubuntu 10.04
DkmsStatus: Error: [Errno 2] No such file or directory
MachineType: Apple Inc. MacBook2,1
Package: xserver-xorg-video-intel 2:2.9.1-1ubuntu12
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-16-generic root=UUID=0d2bdb6a-458e-4cd7-bca5-0ba9de840fd7 ro quiet splash nomodeset
ProcEnviron:
 LANG=en_CA.utf8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-16.25-generic
SourcePackage: xserver-xorg-video-intel
Uname: Linux 2.6.32-16-generic x86_64
dmi.bios.date: 06/27/07
dmi.bios.vendor: Apple Inc.
dmi.bios.version: MB21.88Z.00A5.B07.0706270922
dmi.board.asset.tag: Base Board Asset Tag
dmi.board.name: Mac-F4208CAA
dmi.board.vendor: Apple Inc.
dmi.board.version: PVT
dmi.chassis.asset.tag: Asset Tag
dmi.chassis.type: 10
dmi.chassis.vendor: Apple Inc.
dmi.chassis.version: Mac-F4208CAA
dmi.modalias: dmi:bvnAppleInc.:bvrMB21.88Z.00A5.B07.0706270922:bd06/27/07:svnAppleInc.:pnMacBook2,1:pvr1.0:rvnAppleInc.:rnMac-F4208CAA:rvrPVT:cvnAppleInc.:ct10:cvrMac-F4208CAA:
dmi.product.name: MacBook2,1
dmi.product.version: 1.0
dmi.sys.vendor: Apple Inc.
system:
 distro: Ubuntu
 codename: lucid
 architecture: x86_64
 kernel: 2.6.32-16-generic

[lspci]
00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller [8086:27a2] (rev 03)
     Subsystem: Intel Corporation Device [8086:7270]

[Kevin (campbell-kc)]

This appears so far to have been fixed in the recent xserver updates. I've only restarted a couple of times, but it was nearly 100% repeatable before, so it's looking good.

[Bryce Harrington (bryce)]

Okay thanks for letting us know it looks fixed now; feel free to reopen if it comes back.

[Kevin (campbell-kc)]

Unfortunately, it is still occurring, just not 100% of the time anymore. In general, it requires anywhere between 1 and 10 tries to start up the X server successfully. Somewhat oddly, it seems to always start up successfully immediately after installing an xserver package update, as well as after booting into OSX. (I'm on a Macbook...) After that, it is back to being random.

Also, while the behaviour is the same as before, I haven't been able to locate a similar segmentation fault as the one I originally posted in any more logs. Also interestingly, every time it freezes, the caps lock key still responds, however, if I try and switch VT (Ctrl-alt-f1) then it really locks up and nothing responds.

FYI this is now after the 2.6.32.18 kernel updates and everything else as of Mar 30th.

[Kevin (campbell-kc)]

So, if the Xorg.0.log.old file works as I assume (the .old file is the previous boot...) then I have attached here a successful boot (the Xorg.0.log file) and an unsuccessful boot (the Xorg.0.log.old file).

The most obvious difference is these two lines at the end of the unsuccessful boot:

(WW) "Apple Computer Apple Internal Keyboard / Trackpad": unable to handle keycode 464
(II) AIGLX: Suspending AIGLX clients for VT switch

However, I have a suspicion that could be the full on lockup that occurs when I tried to switch to VT1 after it got stuck.

There are a couple more differences between the files, but I'll leave that to the experts to look at...

[Kevin (campbell-kc)]

A little more interesting details I've figured out here... sorry for all the posts but I've been trying to figure out what's causing this...

The odd time it does boot up properly with KMS on, (to be clear here, it boots up ok if I set nomodeset at grub...) it will crash in the exact same way if I switch to another VT and then try and go back to VT7. The last line written in the Xorg.0.log file is:

(II) AIGLX: Suspending AIGLX clients for VT switch

When I have disabled KMS, the process of switching VT's and back leaves these lines:

(II) AIGLX: Suspending AIGLX clients for VT switch
(II) Open ACPI successful (/var/run/acpid.socket)
(II) AIGLX: Resuming AIGLX clients after VT switch

So it would appear something is happening when trying to open ACPI. Now, in the files I attached before, the time my boot was successful, Xorg.0.log reports an error:

(WW) Open ACPI failed (/var/run/acpid.socket) (No such file or directory)

Perhaps the successful boots are only occurring when there is some sort of error here? My unsuccessful boot log has a success line here...

[Bryce Harrington (bryce)]

(WW) Open ACPI failed (/var/run/acpid.socket) (No such file or directory)

That's a red herring. Ignore the ACPI error messages, they're completely irrelevant.

[Kevin (campbell-kc)]

I tried updating to the newest intel driver using the xorg-edgers ppa, and the problems still happened. So I assume the problem was in the kernel, not the x driver. Regardless, I have reverted to everything in lucid, and the 2.6.32-19 kernel update seems to have resolved the problems. I've been testing it out for a couple days to make sure, but everything is rock solid on multiple reboots and mode swtiches. So it appears you can mark this one solved!

For the record, from what I remember, the problems started appearing after the 2.6.33 drm backports landed, which would make sense.

[Chris Halse Rogers (raof)]

Marking as fixed as per the previous comment. If you think this is incorrect, please feel free to reopen.