Bug 113141 - Suspend-to-disk does not work (does not even try resuming)
Summary: Suspend-to-disk does not work (does not even try resuming)
Status: RESOLVED DUPLICATE of bug 112794
Alias: None
Product: SUSE LINUX 10.0
Classification: openSUSE
Component: Installation (show other bugs)
Version: Beta 3
Hardware: x86 All
: P5 - None : Normal
Target Milestone: ---
Assignee: Jiri Srain
QA Contact: Klaus Kämpf
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-08-25 19:39 UTC by Sven Lüppken
Modified: 2005-09-02 08:16 UTC (History)
1 user (show)

See Also:
Found By: Other
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
Powersave logs (1.90 KB, application/x-tgz)
2005-08-25 19:40 UTC, Sven Lüppken
Details
Hwinfo log (304.28 KB, text/plain)
2005-08-25 19:41 UTC, Sven Lüppken
Details
grub config modified by Beta 4 update (968 bytes, text/plain)
2005-08-28 11:12 UTC, Lars Müller
Details
Log while calling powersave -U (2.33 KB, text/plain)
2005-08-28 11:38 UTC, Lars Müller
Details
YaST2 logs (1.05 MB, application/x-gzip)
2005-08-29 17:11 UTC, Sven Lüppken
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Sven Lüppken 2005-08-25 19:39:39 UTC
I tested suspend-to-disk on SUSE 10.0 Beta 3 and it doesn't seem to work on my   
Toshiba A50 Satellite notebook. Booting after suspending seems to be the same   
as without, same messages, same speed.
Comment 1 Sven Lüppken 2005-08-25 19:40:43 UTC
Created attachment 47660 [details]
Powersave logs
Comment 2 Sven Lüppken 2005-08-25 19:41:26 UTC
Created attachment 47661 [details]
Hwinfo log
Comment 3 Olaf Kirch 2005-08-26 07:58:35 UTC
forwarding to pavel 
Comment 4 Pavel Machek 2005-08-26 21:53:36 UTC
I really need to know kernel messages from suspend attempt. Does it poweroff
itself at the end of suspend?

Has suspend ever worked for you?
Comment 5 Sven Lüppken 2005-08-27 10:17:07 UTC
> I really need to know kernel messages from suspend attempt.

Which logs exactly do you need? I'd be happy to provide you with everything
needed to fix this bug.

> Does it poweroff itself at the end of suspend?

Yes it does. And I just saw when rebooting after suspend the filesystems are not
clean and ReiserFS seems to fix plenty of errors in the filesystem.

> Has suspend ever worked for you?

I tried it with another Linux distribution once but it never worked. It only
works on Windows XP.
Comment 6 Pavel Machek 2005-08-27 10:45:26 UTC
I'd like to see messages that scroll on your screen just before poweroff.

Also please take a look at Doc*/power/swsusp.txt in kernel sources. Make sure
you have just one swap partition, etc.

Unclean filesystems are normal after failed suspend; remount them read-only
before attempt to be sure. Also ext2 is more robust against this kind of abuse.
Comment 7 Sven Lüppken 2005-08-27 11:41:07 UTC
> I'd like to see messages that scroll on your screen just before poweroff.

I'm sorry but I have no idea how I should log these to a file before poweroff.
But it seems that the error is fixed: While updating from Beta2 the kernel
parameters weren't updated.
I added resume=/dev/hda3 to /boot/grub/menu.lst and now it resumes to an X
login. The USB mouse needs some time to work but I guess that's another bug.

Thanks for helping. I think the resume=/dev/hdx should be added to the kernel
parameters automatically to avoid such errors.
Comment 8 Pavel Machek 2005-08-27 20:37:49 UTC
Not sure why it works... it should resume from initrd. Anyway if it works with
resume=, it is probably not kernel problem.
Comment 9 Lars Müller 2005-08-28 11:09:50 UTC
Pavel: Shoud resume from disk work without a resume=/dev/hdaX work?

I've updated my workstation at home to Beta3 and no longer have resume=/dev/hda5
in /boot/grub/menu.lst.  But that have been there as I can check from the
menu.lst.old file.
Comment 10 Lars Müller 2005-08-28 11:12:36 UTC
Created attachment 47886 [details]
grub config modified by Beta 4 update
Comment 11 Lars Müller 2005-08-28 11:38:06 UTC
Created attachment 47887 [details]
Log while calling powersave -U

Sven: To capture the output of powersave -U I've configured a serial console in
menu.lst

Pavel: The output gave the hint to use resume= and if I set it again resume
works well. :)
Comment 12 Sven Lüppken 2005-08-28 11:55:01 UTC
Lars: Thanks! 

My powersave -U log looked almost identical. Seems the resume= is definitely  
needed.
Comment 13 Forgotten User ZhJd0F0L3x 2005-08-29 04:05:15 UTC
It can probably not work without resume= parameter because the initrd must setup
the resume device and can only know this from the resume parameter.

This is not a kernel/initrd bug but a bug for whoever hosed the update or forgot
to add the resume= parameter during install
Comment 14 Arvin Schnell 2005-08-29 12:47:00 UTC
Please provide YaST2 logs from installation so we can check why the
resume= parameter was missing in the first place, see:
http://www.opensuse.org/index.php/Bug_Reporting_FAQ#YaST
Comment 15 Sven Lüppken 2005-08-29 17:11:59 UTC
Created attachment 48046 [details]
YaST2 logs

Attaching the output of save_y2logs
Comment 16 Mihnea Istinie 2005-08-30 09:07:45 UTC
Had the same problems with missing the resume=/dev/*** entry in the command line.
It looks like the two machines I updated from Beta2 to Beta3 are missing the
resume line. A fresh installation done afterwards with Beta3 was OK...
Comment 17 Lukas Ocilka 2005-09-02 08:09:47 UTC
Jiri, is this yours? Adding resume=... parameter? See comment #13 and comment #16
Comment 18 Jiri Srain 2005-09-02 08:16:55 UTC
The problem occurred in Beta3 only durign updates, fresh installatino should 
have been fine. It has already been fixed in Beta4. 

*** This bug has been marked as a duplicate of 112794 ***