[Bug 34155] sleep and hibernate don't work on Dell C640 (regression from breezy)

[Lucas Nussbaum]

Public bug reported:
https://launchpad.net/malone/bugs/34155

Affects: linux-source-2.6.15 (Ubuntu)
Severity: Normal
Priority: (none set)
Status: Unconfirmed

Description:
Suspend and Hibernate don't work on a Dell C640. Hibernate used to work
perfectly in Breezy. I never tried Suspend with Breezy.

When resuming from Suspend/Hibernate, the screen comes back garbled.

xorg.org device section for the graphic card:
Section "Device"
Identifier "ATI Technologies, Inc. Radeon Mobility M7 LW [Radeon Mo
bility 9000]"
Driver "ati"
BusID "PCI:1:0:0"
EndSection

[Andreas Schildbach]

Public bug report changed:
https://launchpad.net/malone/bugs/34155

Comment:
Garbled with random green rectangles?

Then it's the same on the Dell Latitude X1, which uses an Intel i915GMS.

One time I was able to recover by pressing (Ctrl-)Alt-Fx-Combinations,
but on all other tries this only changed the distribution of the green
rectangles.

[Lucas Nussbaum]

Public bug report changed:
https://launchpad.net/malone/bugs/34155

Comment:
The rectangles are white here, and there are some color lines too.

I never managed to get out of it.

[Lucas Nussbaum]

Hi,

This has been reported a month ago, and is extremly annoying. Are there some guidelines somewhere about helping to debug this ?

** Changed in: linux-source-2.6.15 (Ubuntu)
Sourcepackagename: linux-source-2.6.15 => None

[Erik_Andren]

I've been able to improve the situation by adding vga=0 to the grub boot line.
Please try if this improves the situation for you too.

[Andrew Ash]

** Changed in: Ubuntu
Status: Unconfirmed => Needs Info

** Changed in: Ubuntu
Sourcepackagename: None => linux-source-2.6.15

[Lucas Nussbaum]

I don't have this laptop anymore, so can't help fixing this bug.

[Andrew Ash]

Marking rejected until new information arises. If this continues to
occur on the latest dapper or edgy beta, please feel free to reopen the
bug and continue discussion.

Thanks for the bug report!

** Changed in: linux-source-2.6.15 (Ubuntu)
Status: Needs Info => Rejected

[Erik_Andren]

I'm still experiencing this problem using Edgy current. This is not
something distro specific though. I've been experiencing the same thing
on both Gentoo and Slackware.

[Andrew Ash]

** Changed in: linux-source-2.6.15 (Ubuntu)
Status: Rejected => Needs Info

[Ripper]

Dell Latitude C640, Ubuntu Edgy.

In default configuration, I could not suspend nor hibernate. After
resuming suspend, there were a few lines and a box (mouse cursor) on the
screen, could not restore image in any way. Trying to hibernate simply
shut down the computer like power off.

After adding "vga=0" to boot parametres, the suspend works. At least a few tries worked. Will see in the future.
The hibernation does nothing, but there's a message in console that it couldn't find swap partition or something. I don't have swap partition, so that's maybe my fault. Cannot try hibernation hence.

I also tried follow this thread
http://www.ubuntuforums.org/showthread.php?t=187655
and commented
# SAVE_VBE_STATE=true
# POST_VIDEO=true
lines in /etc/default/acpi-support. This works too, I don't have to add the "vga=0" line. Which is better solution? Are there any drawbacks in either of them?

[Ripper]

Well I have been too optimistic. Does not work in all cases. You have
about 70% chance to succesfully resume after suspend, doesn't depend on
which method you choose to workaround the bug. In other case I got blank
screen or again garbled screen with lines and boxes.

Maybe it is not important, but I noticed, that after suspend resume
there is pc speaker playing, even if it's muted in alsamixer.

[Andrew Ash]

Edgy uses the 2.6.17 kernel, so this bug belongs under linux-
source-2.6.17

** Changed in: linux-source-2.6.15 (Ubuntu)
Sourcepackagename: linux-source-2.6.15 => linux-source-2.6.17

[Andrew Ash]

Could someone please test this with feisty? Now's the time to get it
fixed.

[Erik_Andren]

This is still an issue in feisty current.

[towsonu2003]

per last comment

** Also affects: linux-source-2.6.20 (Ubuntu)
Importance: Undecided
Status: Unconfirmed

** Changed in: linux-source-2.6.20 (Ubuntu)
Status: Unconfirmed => Confirmed

** Changed in: linux-source-2.6.20 (Ubuntu)
Importance: Undecided => Medium

[Tim Gardner]

** Changed in: linux-source-2.6.20 (Ubuntu)
Assignee: (unassigned) => Tim Gardner

[Tim Gardner]

** Changed in: linux-source-2.6.17 (Ubuntu)
Assignee: (unassigned) => Tim Gardner

[Ben Collins]

Confirmed bugs need to be assigned to ubuntu-kernel-team.

** Changed in: linux-source-2.6.20 (Ubuntu)
Assignee: Tim Gardner => Ubuntu Kernel Team

[Erik Andrén]

Still an issue with Gutsy

** Changed in: linux-source-2.6.22 (Ubuntu)
Sourcepackagename: linux-source-2.6.20 => linux-source-2.6.22

[Tim Gardner]

** Changed in: linux-source-2.6.17 (Ubuntu)
Assignee: Tim Gardner => (unassigned)
Status: Incomplete => Won't Fix

[lucio]

I have a acer with Radeon Mobility M9 [Radeon Mobility 9000] and I can't nor hibernate or suspend, but if I disable radeon driver I can hibernate and suspend correctly!
The problem is the almost absent 3d acceleration! I have this problem since the first hoary and I tried gusty and the problem is still the same!

[Erik Andrén]

Lucio: This bug report is only for the Dell C640 with a radeon 7500
mobility. If your hardware is different, open a separate bug.

[DavidYerger]

Problem is buggy Dell BIOS, specifically the ACPI DSDT.

Given there is no hope for Dell to fix this, the workaround is you need
to get a hacked ACPI DSDT table from

http://acpi.sourceforge.net/wiki/index.php/FixedDsdts

, compile it, and slap it on the back of your initrd.

Works for me. I understand if your hardware is different than the one
the hacked DSDT was created from, you may have to extract yours, and
then manually patch vs. the given "vanilla" DSDT.

Now, to the good folks at Ubuntu, you could of course put a post-install
script in with the kernel RPM to do hack the generic initrd thusly if
the system is detected to be a C640, but I fully understand if you
wouldn't want to do that.


** Attachment added: "initrd with patch for C640"
http://launchpadlibrarian.net/11100298/initrd.img-2.6.22-14-generic

[Erik Andrén]

DavidYerger, I installed your initrd and checked that it was
successfully loaded, unfortunately it still doesn't work. Suspending
once usually works but the second suspend ends in the same failure as
stated before.

[Erik Andrén]

** Summary changed:

- sleep and hibernate don't work on Dell C640 (regression from breezy)
+ sleep and hibernate don't work on Dell C640

** Summary changed:

- sleep and hibernate don't work on Dell C640
+ Suspend-To-Ram (S3) doesn't work on Dell Latitude C640

[DavidYerger]

Whoopsie--sorry Erik, I only tested hibernate.

Suspend worked for me twice (with the exception of an external drive),
then locked me up with a striped screen on the third try after waiting
longer.

I think I read somewhere that the problem might be with the video chip.
Maybe I should try the "ati" driver instead of "radeon".
DOUBLE_CONSOLE_SWITCH might help if that's the problem, too.

Also, read here http://ubuntuforums.org/showthread.php?t=211592 on how
to get suspend to work, seems to work fine for me, but it's "cheating"
with regard to the present topic as it drops into S1 not S3 sleep.

HTH

[DavidYerger]

Recently waking up from Standby (S1) gives a screen with a white line,
and the area under it eerily gets brighter, kind of like a mood ring...

I changed the parameter in gconf to hibernate instead:
1. I get a notification that hibernation failed when it wakes up
2. The external drive doesn't wake up
3. The network adapter didn't wake up last time (/etc/init.d/networking restart didn't help), last one might be related to me having a wireless card in, usually this works

[coonj]

I have a similar problem on a Fujitsu P7010D (w/ Intel Extreme graphics
2 chip), and was able to get the video back only if I initiated a
reboot:

- Successful suspend
- on resume, mouse is visible but the screen is white/black/green stripes
- If i bring up the "Quit" dialog (via Alt+Ctrl+Del), I can kind of see where the box is on the screen, and I know where to click for "Shutdown", so I do that
- Immediately, the video comes back! Everything looks great, but I had initiated Shutdown for the to happen, so that is not a solution!

But, I think it is worth mentioning that the video does come back during the Shutdown process, and it is almost immediate, so it is something that runs near the beginning of the shutdown script.

[Erik Andrén]

This is fixed now with the following commit in the xf86-video-ati:
http://gitweb.freedesktop.org/?p=xorg/driver/xf86-video-ati.git;a=commit;h=a0a73208a21546ac120fb9a463261836c9ea7b55



** Changed in: xserver-xorg-driver-ati (Ubuntu)
Sourcepackagename: linux-source-2.6.22 => xserver-xorg-driver-ati

[Erik Andrén]

Seems like I spoke too soon (again).
I suspended the computer successfully 10 times in a row, but when I retried the procedure 30 minutes later it failed.
Back to step 1...

[Tormod Volden]

** Changed in: xserver-xorg-video-ati (Ubuntu)
Sourcepackagename: xserver-xorg-driver-ati => xserver-xorg-video-ati

[Ben Aultowski]

I added linux-source-2.6.24 to the list of affected packages, since this
bug is still present in Hardy.

** Also affects: linux-source-2.6.24 (Ubuntu)
Importance: Undecided
Status: New

[Ben Aultowski]

I don't think this bug is related to X.org's ATI driver at all.

I followed some of the suggestions above, changing /etc/default/acpi-
support and editing Grub. The thing is, I tested everything in recovery
mode, before X.org ~and~ the radeon kernel module are loaded. I was only
able to suspend half the time. I tried removing my wireless module
(ipw2200) and loading the radeon module. It all had the same effect:
suspend-to-ram only resumed half of the time, regardless of the radeon
module's presence. An ACPI problem, perhaps?

(I'm on a Dell C640, as well.)

** Attachment added: "lsmod"
http://launchpadlibrarian.net/12688837/lsmod

** Also affects: acpi (Ubuntu)
Importance: Undecided
Status: New

[Ben Aultowski]

Oh, and I just noticed that this bug just celebrated its second
birthday. I added ACPI to the affected packages as a birthday present,
and invited the Kernel ACPI team to come to the party.

[Leann Ogasawara]

Moving linux-source-2.6.24 task to just linux since beginning with the
Hardy release the kernel source package naming convention changed from
linux-sourc-2.6.xx to just linux. Thanks.

** Changed in: linux (Ubuntu)
Sourcepackagename: linux-source-2.6.24 => linux

[Leann Ogasawara]

Ben Aultowski - Running the latest 2.6.24-12 Hardy Beta kernel, care to
attach your dmesg output after an attempted suspend/resume cycle as
outlined here: https://wiki.ubuntu.com/DebuggingKernelSuspend .
Thanks.

** Changed in: linux (Ubuntu)
Status: New => Incomplete

[Ben Aultowski]

I'm sorry, maybe I don't understand the instructions. I used the command
to simulate a suspend, but when I powered back on, I faced the usual
problem. How am I supposed to get the output from dmesg when my laptop
doesn't boot properly from a suspend? Is there a way to record its
output as it begins the suspend cycle? Or automatically record it when
it tries to resume? Because all I've got is the dmesg from when I had to
do a hard reset after it failed to resume....

(My computer thinks it's 1996 now, so at least that part worked....)

[Erik Andrén]

Ben, you can access the machine via ssh from another computer and
capture the logs.

[Ben Aultowski]

It might be a couple more days before I can follow up on this bug. My
laptop's power adapter was died the other night (the "barely supported
by Ubuntu docking station" comes to to rescue!), and I haven't been able
to hook up my desktop since my most recent move because I can't afford a
the mouse and keyboard to use it...

I will report back before Hardy (a LTS) is launched, as this bug is
making me consider jumping distros. A Dell C640 suspends "out of the
box" on Arch Linux....

[Bryce Harrington]

I don't know if this will fix this specific issue, but I've backported a
bunch of high importance patches that upstream recommended, that fix
problems sort of like this one, so I think it would be worth the time to
test. Please try this .deb and report what you find:

http://people.ubuntu.com/~bryce/Testing/ati/

If we can determine that the patches in this deb fix this bug, they may
be candidates for backporting to Hardy.

[Ben Aultowski]

I tried your package, Bryce, but it didn't fix the suspend issue.

I still haven't been able to investigate this issue using another other
computer to log in, as I still don't have a spare keyboard...

[Bryce Harrington]

Dropping task against -ati as per comment #26. Looks like it's a kernel
issue since the bug can be seen without X running.

** Changed in: xserver-xorg-video-ati (Ubuntu)
Status: Confirmed => Invalid

[Leann Ogasawara]

The Ubuntu Kernel Team is planning to move to the 2.6.27 kernel for the
upcoming Intrepid Ibex 8.10 release. As a result, the kernel team would
appreciate it if you could please test this newer 2.6.27 Ubuntu kernel.
There are one of two ways you should be able to test:

1) If you are comfortable installing packages on your own, the linux-
image-2.6.27-* package is currently available for you to install and
test.

--or--

2) The upcoming Alpha5 for Intrepid Ibex 8.10 will contain this newer
2.6.27 Ubuntu kernel. Alpha5 is set to be released Thursday Sept 4.
Please watch http://www.ubuntu.com/testing for Alpha5 to be announced.
You should then be able to test via a LiveCD.

Please let us know immediately if this newer 2.6.27 kernel resolves the
bug reported here or if the issue remains. More importantly, please
open a new bug report for each new bug/regression introduced by the
2.6.27 kernel and tag the bug report with 'linux-2.6.27'. Also, please
specifically note if the issue does or does not appear in the 2.6.26
kernel. Thanks again, we really appreicate your help and feedback.

** Tags added: cft-2.6.27

[Ben Aultowski]

I tried the 2.6.27 kernel from Intrepid's repos, and the problem remains
unfixed.

[Erik Andrén]

I can confirm that Ubuntu Intrepid Beta still doesn't resolve the
problem.

[Leann Ogasawara]

** Changed in: linux (Ubuntu)
Status: Incomplete => New

[Ben Aultowski]

The problem still remains in Intrepid's Release Candidate.

[Leann Ogasawara]

** Changed in: linux (Ubuntu)
Importance: Undecided => Medium
Assignee: (unassigned) => Ubuntu Kernel Team (ubuntu-kernel-team)
Status: New => Triaged

[Daniel T Chen]

How about in 9.04 with linux-image-2.6.28-1-ub-generic?

[Youri Volkov]

Update: 9.04 with 2.6.27-8-generic works on my end.

[Leann Ogasawara]

Ben Aultowski - would you be able to test the latest pre-release of
Jaunty to confirm if this is resolved for you like it was for Youri.
http://cdimage.ubuntu.com/releases/jaunty/alpha-2/ , you should be able
to test suspend via a LiveCD. Thanks.

** Changed in: linux (Ubuntu)
Status: Triaged => Incomplete

** Tags added: suspend

** Tags added: resume

** Tags added: intrepid

** Tags added: jaunty

[Ben Aultowski]

@Launchpad Janitor: Good to know that top men are working on this. TOP
MEN. Oh, wait, you're just a robot and the developers have pretty much
ignored this bug for three years now. On a related note, I now have two
of these machines (Military surplus), and the second one came with
Windows 2000 installed. I booted it up and closed the lid. A tear ran
down my cheek when I opened the lid and watched a ten year-old OS resume
from suspend.

@Leann Ogasawara: Jaunty Alpha 3 is scheduled to drop tomorrow. I'll
wait for that to be uploaded and report back.

[Erik Andrén]

I've tested with a daily jaunty build from 12 august and the bug can still be reproduced. It may work once or twice in a row.
This leads me to belive that this is a race issue of some kind.

[TJ]

For those that can reproduce this could you provide the following
reports/attachments?

For both hibernate (to disk) and suspend (to RAM) gather the kern.log
files from the current start-up and the previous after a suspend/resume
and hibernate cycle. By providing the previous kern.log file we should
capture useful information even if you had to restart the PC to get to
the logs.

tar -czf kern.logs.tar.gz /var/log/kern.log{,.0}

sudo lspci -vvnn >/tmp/lspci.log

uname -a
lsb_release -a

[Stefan Bader]

Closing the acpi task. There is nobody going to look at that.

** Changed in: acpi (Ubuntu)
Status: New => Invalid

[Stefan Bader]

Can someone confirm this problems on the latest release (Jaunty)? Please
provide the information TJ mentioned as well as the list of modules
loaded. Often those problems are related to loaded modules not working
well on suspend.

** Changed in: linux (Ubuntu)
Assignee: (unassigned) => Stefan Bader (stefan-bader-canonical)

[DavidYerger]

davidy at faustillium:~$ uname -a
Linux faustillium 2.6.28-11-generic #42-Ubuntu SMP Fri Apr 17 01:57:59 UTC 2009 i686 GNU/Linux

davidy at faustillium:~$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 9.04
Release: 9.04
Codename: jaunty


** Attachment added: "Kernel logs as requested"
http://launchpadlibrarian.net/26371357/kern.logs.tar.gz

[DavidYerger]

What happened in the last few boots regarding my attachment earlier:

* standby
* successful resume
* standby
* bad resume (vertical colored stripes, corrupted cursor)
* forced power off and normal start
* hibernate
* locked up on restart
* tried earlier kernel to bypass hibernation stuff, didn't work
* disabled splash and quiet in GRUB boot line, booted.

Also attaching suspend log, maybe that will help.

** Attachment added: "Suspend log from /var/log, might help"
http://launchpadlibrarian.net/26371473/pm-suspend.log

[Nils]

Is there any progress in fixing this bug? My Dell Latitude C640 running
kernel 2.6.30 on Arch Linux, is affected as well. Suspend works only one
time, after that, the screen gets garbled. Very annoying. It would be
great if this bug gets fixed!

[Ben Aultowski]

I now have two of these systems. My old one has a cracked display and
keyboard stopped working, but I will donate it to a developer than can
work on this bug as well as this one:
https://bugs.launchpad.net/ubuntu/+source/hal/+bug/309973

[OkropNick]

the same problem here on dell c640

[Erik Andrén]

Would be interesting if someone gave this a go with the ATI kernel mode setting.
See: http://www2.bryceharrington.org:8080/drupal/node/84

[Ben Aultowski]

@Erik Andr?n

I tried to follow the instructions, but was a little confused, mostly
because configuring GRUB2 is different from the previous version. I
installed the drivers and added the 'radeon.modeset=1' line to GRUB2's
new config file at /etc/default/grub in the area where bootloader
settings are supposed to go. But I just got an error saying that it was
unrecognized and it tossed it out. When my system fully booted, the
display was, well, bizarre. I could make out what I was doing, but
everything seemed like it was being passed through a mosaic filter.
Other pages linked to by the one you provide seem to suggest that this
only works with the 2.6.31-13 kernels and up, but these are not Karmic's
repositories.

I've been following this bug for three years and I really, really want
to get it fixed. Could you provide a little assistance to an ordinary
user that just wants to help?

[Pauli]

That sounds like bad problem with DFS or UTS. You can disable them in
xorg.conf with exa optinos to device context. More details are in man
exa to what can be put to xorg.conf.

Can you attach dmesg and xorg.log when modeset=1 is passed to kernel? It
works but kernel doesn't know about module to handle it before radeon is
loaded so you get that warning for working parameter.

[Ben Aultowski]

My C640 actually resumed from suspend tonight with the latest build of
Karmic, albeit with a kernel oops message. I reported the bug in a
separate report so Apport could upload all the pertinent logs. It can be
found here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/457861

[Ben Aultowski]

** Also affects: linux-meta (Ubuntu)
Importance: Undecided
Status: New

[Andy Whitcroft]

This is a kernel issue, the packaging is not affected. Closing out the
linux-meta task only. The linux task where the work is ongoing remains
ope.

** Changed in: linux-meta (Ubuntu)
Status: New => Invalid

[DavidYerger]

Tried again with Lucid, got corrupted screen upon restore.

[Ubuntu ix86]

I'm attempting to get standby/resume to work on my c640 running Ubuntu 10.04.
I'm pretty sure that it had been working in past releases, but isn't working reliably now...

I tried modifying /etc/X11/xorg.conf with the instructions on the quirks page here
https://wiki.ubuntu.com/X/Quirks#ATI%20Radeon%20Driver%20Quirks

resulting in

Section "Device"
...
Option "AGPMode" "x"
EndSection

where x is some number 1, 2, 4 or 8.

I found that when using AGPMode 4:

I can suspend and resume with the power plugged in selecting standby
from the shutdown menu and resume by pressing the power button.

If I select suspend from the menu, and then close the lid, the computer
will turn back on, and the backlight will turn on, but the screen
remains blank and black. There is no response to keyboard or mouse
input, no ctrl+alt+F#. The only option that works is to hold down the
power button, and reboot.

When I shut the lid, the laptop goes into standby, and when the lid is
opened again the laptop resumes, but the video is all garbled into
various colors and strips. The area where the mouse pointer is
occasionally has a blotchy square where it should be. I can switch to a
terminal and that works fine, no graphical issues there. But the only
way to get back to X11 is to use ctr+alt+backspace or run "sudo service
gdm restart" from a terminal.

I can't think of anything else to tell...

I've tried attaching the xorg.conf file that I'm using.

** Attachment added: "xorg.conf file that somewhat works"
http://launchpadlibrarian.net/49965734/xorg.conf

[Ubuntu ix86]

Forgot to add,
Using AGPMode 1 and 8 resuming from either lid or menu standby results in the blank screen, backlight is on, but no disk activity or anything, requiring a hardboot.

In AGPMode 2 almost the same as AGPMode 4, but there are random
occurrences of the above symptoms.

[Jeremy Foshee]

** Changed in: linux (Ubuntu)
Status: Incomplete => Triaged

** Tags added: kernel-needs-review kernel-power