[REGRESSION] Dell 2650 Notebook Wedges on Boot with Jaunty works with Intrepid

[Pete Graner]

I upgraded a Dell 2650 from Intrepid to Jaunty. After the upgrade the box would wedge on boot.

After some debugging I found it was hanging (see attached photo).

Booting with pci=noacpi works.

2.6.27-11 Boots and works
2.6.28-13 Hangs only will boot with pci=noacpi
2.6.31-3 Hangs only will boot with pci=noacpi

ProblemType: Bug
Architecture: i386
CurrentDmesg:
 [ 36.002135] eth0: setting full-duplex.
 [ 36.400998] NET: Registered protocol family 17
 [ 39.474565] NET: Registered protocol family 10
 [ 39.476180] lo: Disabled Privacy Extensions
 [ 50.328078] eth0: no IPv6 routers present
DistroRelease: Ubuntu 9.04
HibernationDevice: RESUME=UUID=b962c2fd-a8f8-47b6-968d-01762f23b3f4
Lsusb:
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 001 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
MachineType: Dell Computer Corporation Inspiron 2650
NonfreeKernelModules: nvidia
Package: linux-image-2.6.27-11-generic 2.6.27-11.27
ProcCmdLine: root=UUID=76ed4710-32ef-437e-ac47-53211ec25d44 ro quiet splash
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.27-11.27-generic
SourcePackage: linux
UnreportableReason: This is not a genuine Ubuntu package

[Pete Graner (pgraner)]

After talking with rtg in IRC I used a test kernel he made with ACPI_DEBUG turned on:

http://kernel.ubuntu.com/~rtg/linux-image-2.6.31-3-generic_2.6.31-3.19_i386.deb

Booted with the following combo of options:

acpi.debug_layer=0x2 acpi.debug_level=0xffffffff

hpet=disable

clocksource=hpet

 clocksource=pit

They all hung in the same place with no additional info.

[Stefan Bader (smb)]

As I currently got no real clue what might be the cause of this, I would like to try to corner the moment this regression was introduced. For that I prepared a mainline build for 2.6.28-rc1 at https://wiki.ubuntu.com/KernelTeam/MainlineBuilds. If you could try this and report whether it is showing the problem or not. Then we can move into one or the other direction.

[Colin Ian King (colin-king)]

Adding pci=noacpi is a workaround for this.

[Colin Ian King (colin-king)]

I had a look at the machine today and saw that the machine was hanging during the boot phase when trying to mount / rw. This was fixed by removing a bunch of /etc/mtab~* files.

[Stefan Bader (smb)]

Doh, seems to be a duplicate of bug #327499 which seems to have been caused by a change in ACPI initialization.

"ACPI: Attach the ACPI device to the ACPI handle as early as possible"

This got reverted in 2.6.31.6. I will mark this report as duplicate of the other bug.