[i845g] Login screen doesn't appear

[Simon Lewis]

Binary package hint: xserver-xorg-video-intel

When the machine boots, the splash screen appears as normal. The display then goes black, except for the mouse pointer.

The login 'drum beat' sound plays and the monitor backlight
turns off and back on again. This cycle then repeats several times until machine hangs completely. Sometimes the display shows two flickering mouse pointers rather than one.

However, I can login to the machine by booting into recovery mode, choosing 'resume normal boot' option, then Ctrl+Alt+F1 to bring up a login prompt. After logging in, 'startx' displays the desktop, although graphics performance is sluggish.

This may be related to bug #446641, as some of the bios and mobo details are the same (HP machine, Gamila/Giovani/Neon series mobo, v.3.26 bios), with same issue of /dev/dri/card0 not being created.

Also of possible relevance is the message “ACPI Error: A valid RSDP was not found”, although this is an ACPI-capable machine.

When booting from the LiveCD, the desktop appears correctly, although the splash screen is text-based.

ProblemType: Bug
Architecture: i386
CurrentDmesg:
 [ 24.511393] ppdev: user-space parallel port driver
 [ 26.538661] eth0: no IPv6 routers present
Date: Wed Mar 31 09:33:15 2010
DistroRelease: Ubuntu 10.04
DkmsStatus: Error: [Errno 2] No such file or directory
GdmLog:

GdmLog1:

GdmLog2:

InstallationMedia: Ubuntu 10.04 "Lucid Lynx" - Beta i386 (20100318)
Lsusb:
 Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
MachineType: HP Pavilion 061 PC069A-ABU a509.uk
Package: xserver-xorg-video-intel 2:2.9.1-1ubuntu18
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-18-generic root=UUID=a38c7fa7-9959-4f24-bcc5-13639dff077e ro single
ProcEnviron:
 SHELL=/bin/bash
 LANG=en_GB.UTF-8
ProcVersionSignature: Ubuntu 2.6.32-18.27-generic 2.6.32.10+drm33.1
SourcePackage: xserver-xorg-video-intel
Tags: lucid lucid
Uname: Linux 2.6.32-18-generic i686
XorgLogOld:

dmi.bios.date: 05/09/2005
dmi.bios.version: 3.26
dmi.board.name: Gamila/Giovani/Neon series
dmi.board.vendor: MICRO-STAR INTERNATIONAL CO., LTD
dmi.board.version: 030
dmi.chassis.type: 3
dmi.modalias: dmi:bvn&oenixTechnologies,LTD:bvr3.26:bd05/09/2005:svnHPPavilion061:pnPC069A-ABUa509.uk:pvr0d'0211RE101NEON10:rvnMICRO-STARINTERNATIONALCO.,LTD:rnGamila/Giovani/Neonseries:rvr030:cvn:ct3:cvr:
dmi.product.name: PC069A-ABU a509.uk
dmi.product.version: 0d'0211RE101NEON 10
dmi.sys.vendor: HP Pavilion 061
glxinfo: Error: [Errno 2] No such file or directory
system:
 distro: Ubuntu
 codename: lucid
 architecture: i686
 kernel: 2.6.32-18-generic

[lspci]
00:02.0 VGA compatible controller [0300]: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device [8086:2562] (rev 03)
     Subsystem: Micro-Star International Co., Ltd. Device [1462:5778]

[Bryce Harrington (bryce)]

Please test if turning off KMS makes the issue go away. See https://wiki.ubuntu.com/X/KernelModeSetting for directions.

If that works around the issue, then maybe we'll consider this as evidence we should blacklist KMS for i845 (see also bug #541492)

[Chris Halse Rogers (raof)]

While it's worth checking if turning off KMS will fix this, I suspect it won't. The problem here appears to be that drm doesn't even *attempt* to load.

The output of running “sudo modprobe -vvv i915” in a terminal would be instructive.

[Simon Lewis (simonski)]

Bryce: Turning off KMS made no difference. I haven't experienced the symptoms in bug #541492, but that might be my next issue once this one is resolved.

Chris: Output from modprobe:
$ sudo modprobe -vvv i915

insmod /lib/modules/2.6.32-19-generic/kernel/drivers/acpi/video.ko

WARNING: Error inserting video (/lib/modules/2.6.32-19-generic/kernel/drivers/acpi/video.ko): No such device

WARNING: Error inserting i2c_algo_bit (/lib/modules/2.6.32-19-generic/kernel/drivers/i2c/algos/i2c-algo-bit.ko): No such device

WARNING: Error inserting agpgart (/lib/modules/2.6.32-19-generic/kernel/drivers/char/agp/agpgart.ko): No such device

WARNING: Error inserting drm (/lib/modules/2.6.32-19-generic/kernel/drivers/gpu/drm/drm.ko): No such device

WARNING: Error inserting drm_kms_helper (/lib/modules/2.6.32-19-generic/kernel/drivers/gpu/drm/drm_kms_helper.ko): No such device

FATAL: Error inserting i915 (/lib/modules/2.6.32-19-generic/kernel/drivers/gpu/drm/i915/i915.ko): No such device

It may not be relevant, but I also got an Xorg error this morning:
X: ../../intel/intel_bufmgr_fake.c:1447: drm_intel_fake_bo_exec: Assertion 'ret == 0' failed.

[Chris Halse Rogers (raof)]

Ok. There seems to be the same underlying problem as bug #446641 - ACPI fails, causing a cascade of failures which prevents drm from loading.

Here, though, there's also the black screen problem. Could you please try booting with “blacklist=vga16fb” added to your kernel command line, after “quiet splash”?

[Simon Lewis (simonski)]

Using “blacklist=vga16fb” made no difference.

The reason ACPI fails on this machine is that the 'extended checksum' field (and the 'length' field for that matter) in the RSDP table is wrong, so it is not being recognised as a valid table. (Incidentally, ACPI works correctly under Windows XP, so I assume XP only validates the original 'checksum' field to determine whether the table is valid. Linux validates the 'extended checksum' field as well if the table is ACPI v. 2+ format.)

I have been able to extract the ACPI tables from the BIOS (by amending acpidump from pmtools to ignore the invalid checksum) and the output is attached.