Created attachment 333691 [details]
logs

Did you install all the updates? What does exectly happen? Just nothing?

 Yes.- and I have updated to kernel 2.6.32.2-0.0.16.d8b32f9-desktop.
The disk load led blinks after issuing powersave -U but nothing else
happens:
> powersave -U
method return sender=:1.0 -> dest=:1.715 reply_serial=2
   int32 1

.

Most likely a kernel issue. S2disk has worked with initial installation. However the default kernel of os11.2 is so buggy that it is impossible to use (not even cups runs with it, ksoftirqd problem). Newer kernels are more stable but lack crucial powersaving features (s2ram & s2disk problems).

What do you mean with newer kernels? Which one do you have installed?

Otherwise, you should create bug reports for the other issues (maybe you have already) where the default 11.2 kernel doesn't work. I can't help much when you use a custom kernel. Does s2disk work with the default 11.2 kernel? If so, this bug is invalid and others should be opened.

  I am not using a custom kernel. I am using the kernel from http://ftp.suse.com/pub/projects/kernel/kotd/master/ which is recommended if you need an update in advance.
  Yes, I have reported all these bugs (crashes of cups due to root cause, several crashes of the default 11.2 kernel, errors of Apparmor kernel module.). It is not my fault if nobody feels responsible to provide updates for known hazards.
  I will urge an update as soon as the most vital issues are resolved. Please do not close but push forward a resolution. Then I will ask for an update. S2disk has probably been working with the default kernel although I have not had the time to test this for a longer period of time.

 Please do have a look at this! S2disk has now been refused by couple of subsequent kernel releases. This is just too bad since the main reason why I prefer Linux over FreeBSD, Mac OS X and Solaris is the good s2disk support. It makes live considerably easier and saves power and time. Having to leave my machine in s2ram over night is an enormous waste of energy especially if many people all over the world have to do so. Besides this s2ram does no memory compression so that all the memory will be used up soon without s2isk even if just a few programs are running (closing all user programs becomes inevitable).

(In reply to comment #3)
>  Yes.- and I have updated to kernel 2.6.32.2-0.0.16.d8b32f9-desktop.
> The disk load led blinks after issuing powersave -U but nothing else
> happens:
> > powersave -U
> method return sender=:1.0 -> dest=:1.715 reply_serial=2
>    int32 1

I just saw this in a separate bug. Let me see if it's the same.

 This is a major flaw while other problems have already been resolved (ksoftirqd 100% cpu, scheduler crashes). Just think of the many annoyed users an the huge power waste! We need to get forward to a stable kernel to distribute a patch for openSUSE 11.2. Please do assign an appropriate priority!

Please attach a dmesg output generated right after a failing hibernation.

Created attachment 337616 [details]
dmesg directly after failed s2disk

method return sender=:1.0 -> dest=:1.462 reply_serial=2
   int32 1

What is the difference between dmesg and /var/log/messages? Why is the provision of log/messages not sufficient?

(In reply to comment #12)
> Created an attachment (id=337616) [details]
> dmesg directly after failed s2disk
> 
> method return sender=:1.0 -> dest=:1.462 reply_serial=2
>    int32 1
> 
> What is the difference between dmesg and /var/log/messages? Why is the
> provision of log/messages not sufficient?

dmesg output contains the kernel messages only.

In your dmesg output there are only messages related to suspend to RAM, so if s2disk fails, that happens in the user space.

Can you try s2disk after a fresh boot and attach the dmesg output generated after the resume (or failure, whichever happens)?

  Ooops, I have overlooked an error:
> cat /var/lib/s2disk.conf
resume device = /dev/sda7
...
> cat /proc/cmdline
root=/dev/disk/by-id/ata-WDC_...-part5 resume=/dev/disk/by-id/ata-WDC_...-part8 splash=silent quiet vga=0x375

 That is why I wanna have the content of /var/run/pm-utils.inhibit in /var/log/pm-suspend.log:
> cat /var/run/pm-utils.inhibit
resume partition '/dev/sda8' not active

 So please add the following to /usr/lib/pm-utils/functions:

	if [ -e "$INHIBIT" ]; then
	  echo $'\n################ suspend INHIBITED !! ################'
	  cat "$INHIBIT";echo; echo

  ... or at best directly distribute a patch.

  Under openSUSE 11.0 I have seen it fail because it thought do_hibernate was no function (the strangest error I have ever seen) - I have patched that, too.

  ... and there remains one other little thing to do:
powersave -U should have a non zero exit value if /usr/sbin/pm-hibernate had so, too.

Created attachment 337956 [details]
suspend-inhibit.patch - log that if suspend has been inhibited

  This will not be resolved just by better logging.
There is something totally messed up with the post-install script of the kernels at http://ftp.suse.com/pub/projects/kernel/kotd/master/.
 They do always install me a wrong suspend device.

resume=/dev/disk/by-id/ata-WDC_WD2500BEVS-22UST0_WD-WXC108372622-part8
instead of 
resume=/dev/disk/by-id/ata-WDC_WD2500BEVS-22UST0_WD-WXC108372622-part7

  Why doesn`t the script just leave things as they are if it can not detect the swap device (swapon -s)?

Sorry, there are no resources for debugging this further. Please open a new bug if the problem persists with openSUSE 11.4.