iBook g4 unable to sleep

Bug #26021 reported by Scott Henson
16
Affects Status Importance Assigned to Milestone
linux-source-2.6.12 (Ubuntu)
Invalid
Medium
Unassigned

Bug Description

I have an iBook g4 with a 800MHz cpu and a Radeon M9 9200 Mobility.

This is on a fresh breezy install updated as of today.

The problem comes down to the fact that I cannot get my ibook to sleep. It
slept under hoary with the same setup, but now with breezy, it claims that it is
unable to sleep. Additionally, I can't get the backlight to change in
brightness. This too worked under hoary. The weird thing about it is that I
upgraded the working hoary setup to breezy, and I still had working sleep and
backlight. I wiped the system and installed fresh breezy, and no sleep or
backlight. So I'm convinced the problem is with some config file that was there
with hoary and survived to breezy, but is not there under the fresh breezy install.

pbbcmd sleep does absolutely nothing. When I query pbbcmd to see if sleep is
available, it says it is not.

Looking at bug 23737, it has a command( echo mem | sudo tee /sys/power/state)
that puts my laptop to sleep for a second and then it wakes back up. dmesg says
the following.
[ 746.422371] Stopping tasks:
===========================================================================================|
[ 746.424934] pmac_pm_prepare(3)
[ 746.921174] eth0: suspending, WakeOnLan disabled
[ 747.319240] pmac_pm_enter(3)
[ 747.400098] PCI: Enabling device 0001:10:1b.0 (0000 -> 0002)
[ 747.461099] PCI: Enabling device 0001:10:1b.1 (0000 -> 0002)
[ 747.522100] PCI: Enabling device 0001:10:1b.2 (0000 -> 0002)
[ 747.543101] ehci_hcd 0001:10:1b.2: park 0
[ 747.543236] ehci_hcd 0001:10:1b.2: USB 2.0 restarted, EHCI 1.00, driver 10
Dec 2004
[ 748.576106] eth0: resuming
[ 748.638416] PHY ID: 4061e4, addr: 0
[ 751.038182] eth0: Link is up at 100 Mbps, full-duplex.
[ 751.038368] eth0: Pause is disabled
[ 751.073702] hda: Enabling Ultra DMA 5
[ 751.076896] hdc: Enabling MultiWord DMA 2
[ 751.078078] pmac_pm_finish(3)
[ 751.078176] Restarting tasks... done
[ 751.143356] agpgart: Putting AGP V2 device at 0000:00:0b.0 into 1x mode
[ 751.143379] agpgart: Putting AGP V2 device at 0000:00:10.0 into 1x mode
[ 751.143444] [drm] Loading R200 Microcode

The difference between that bug and this one is that sleep should work on
radeons. Also bug 8675 is related, but that bug is about waking up, I can't
even get to sleep.

I really concider this to be a major regression wrt hoary. Its also a pretty
major bug because it makes my ibook almost useless under Ubuntu(I move around a
lot and having my laptop sleep is very much needed).

Thanks, I tried tracking down the bug myself inside of pbbuttonsd, but it came
down to the kernel telling it that it couldn't sleep, but I know this kernel
sleeps as it did in the upgraded system. Additionally I tried the hoary kernel
with no sucess. I don't think that this is really a bug with pbbuttonsd, but
that is a good starting place. Thanks again.

Tags: apple g4 ppc
Revision history for this message
Scott Henson (scotth) wrote :

Its typical that after I report a bug I find the fix.

I noticed that /proc/cmdline still contained video=ofonly. I had removed it
from my yaboo.conf, but I didn't run ybin. I ran ybin and rebooted and suspend
is working. video=ofonly was needed for the install and was put there by the
installer, maybe this should not be put in the system yaboo.conf. I believe its
needed by the install because radeonfb is a module in the install kernel, but
compiled into the powerpc kernel. Atleast some documentation about how to
enable sleep on an ibook post install would be nice. Thank you.

Revision history for this message
Martin Pitt (pitti) wrote :

(In reply to comment #1)
> Its typical that after I report a bug I find the fix.

Well, it might be a workaround, but not a proper fix. :)

> I noticed that /proc/cmdline still contained video=ofonly. I had removed it
> from my yaboo.conf, but I didn't run ybin. I ran ybin and rebooted and suspend
> is working. video=ofonly was needed for the install and was put there by the
> installer, maybe this should not be put in the system yaboo.conf.

You mean the installer put it there entirely on its own? Or did you specify that
option when you installed Ubuntu?

> I believe its
> needed by the install because radeonfb is a module in the install kernel, but
> compiled into the powerpc kernel. Atleast some documentation about how to
> enable sleep on an ibook post install would be nice. Thank you.

It works fine out of the box on my iBook, OTOH I do not need to specify video=
on installation. So rather than documenting how to fix it, the bug should be
fixed. :)

Revision history for this message
Martin Pitt (pitti) wrote :

Do you still have this problem with the latest Dapper kernel? It works just fine for me.

Can you please attach /var/log/kern.log after a failed sleep attempt if it still does not work?

Changed in pbbuttonsd:
assignee: pitti → kernel-team
status: Confirmed → In Progress
Revision history for this message
Martin Pitt (pitti) wrote :

Benjamin, do you happen to know about sleep problems without a specific video= setting?

Changed in linux-source-2.6.15:
status: In Progress → Needs Info
Revision history for this message
Scott Henson (scotth) wrote :

I don't have my iBook anymore(may it rest in pieces) so I can't say if this bug still exists.

Revision history for this message
Ben Collins (ben-collins) wrote :

For now I'm going to assume it's fixed in dapper, and keep this targeted at breezy, since that's what it was reported against.

Revision history for this message
Andrew Ash (ash211) wrote :

With no activity on this bug in almost a year and Breezy's life coming to a close, I'm going to assume that it's safe to close this bug. If it's still a problem for anyone, feel free to reopen it.

Changed in linux-source-2.6.12:
assignee: kernel-team → nobody
status: Needs Info → Rejected
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.