pm-suspend freezes computer (though apparently suspends)

humilleitor · 2010-05-19 13:13:49

I have an strange issue (well, I guess it's strange since I haven't found any similar thing in Arch forums, bugs, or even in Google):

When I pm-suspend my computer (no matter whether from the command-line, or with some desktop power management tool; but let's stick to the command line), after some seconds of hard disk working, the screen freezes, the "power on" led remains on, and the computer becomes totally unresponsive. No way of getting it out of that state except for unplugging the mains. Then, after booting it again, and on further examination of the /var/log/pm-suspend.log, to my surprise I realize that the suspend has been theoretically a "success", and I get the typical last line of the process correct: "performing suspend".

It's like if it worked, but it didn't. Or put in other words: it's an aeternal suspend, from which the computer can't be woken up again.

I have kernel 2.6.33.4-1 (but with previous kernel, some 2.6.28, I had the same problem).
I have acpid in my list of daemons to be started at bootup (prior to dbus and hal).
I have all the relevant modules (button, acpi_cpufreq, processor, etc.) loaded at bootup.
I have pm-utils 1.3.0-2 version installed
I have acpi 1.4-2 and acpid 1.0.10-3
I have an Intel 945GM video card

I guess I must be making some BIG mistake somewhere, or missing some essential configuration or package, but I can't figure out where. As far as I can check, I have no broken dependencies.

Any help very appreciated.

Last edited by humilleitor (2010-05-19 13:38:27)

scriptmunkee · 2010-05-22 04:26:00

I am in the same boat. I have all the same app versions, but with an nvidia video card. Suspend had been work fine up to about a week or 2 ago. I have not been able to figure out what application update has caused my laptop to not suspend correctly.

humilleitor · 2010-05-22 11:11:49

I'm totally boggled by this issue. Seems to me totally capricious and random. Yesterday I was for a long while fidgetting with the different files in pm-utils package, and acpid, and also creating new initrd images, and finally (!!) got it working (though I don't know which was the decissive move). I successfully suspended and resumed the laptop two or three times. Then rebooted and suspended/resumed again. Great!

This morning, withouth having changed absolutely anything from yesterday, I'm not able to suspend any more. I have again the bug. I'm starting to think that maybe it depends on some program or script changing or modifying or somehow interfering with the BIOS... I don't know. I'm no expert at all.

scriptmunkee · 2010-05-23 01:38:44

After I posted my "same issue" post, I tried changing packages back to what worked about 2 weeks ago and I didn't get any farther. I'll have to do some more digging. The Pm-utils wiki page provided me no insight.

My next step is to look into upgrading my laptop's BIOS. I have Lenovo T61P. Now sure what hardware you are running, but we can't be the only people having this issue.

humilleitor · 2010-05-23 11:49:24

For sure we're not. But we're just a couple of newbies (I suppose) commenting to each other our mutual ignorance in a dead thread, I'm afraid... ;-)

Good luck to you!

(Anyhow, if I get to some positive and permanent solution, I'll post it here)

allbluedream · 2010-05-23 12:56:59

I am having the same problem. It feels random, but maybe connected to ram usage. When I have just 20% ram usage (1GB ram), hibernation almost always succeed, but with a larger ram consumption (think Firefox) it almost always fails.

joshua.finkbeiner · 2010-05-29 13:35:16

Bump - I'm experiencing the same issue. I'm running on an HP G62-144dx with essentially all of the same kernels/options as the parent; one caveat is that I'm running on an LVM setup with the grub 'resume=' parameter pointing to the LVM swap partition instead of a physical partition. Of course, this shouldn't make much of a difference, but I thought I'd mention it in the interest of fully documenting everything.

The machine can successfully hibernate (to RAM) and resume, and suspend to disk also apparently succeeds. However, it is not able to resume from a suspend-to-disk - resuming from suspend results in a clean reboot of Arch. I've been chasing this issue for a couple of weeks now, but to no avail. I even increased the size of my swap partition to 4GB to see if that would make a difference per allbluedream's hypothesis, but increasing the swap did not resolve the issue.

I guess I can't do much more than a "me too" post right now, but any insight would certainly be appreciated.

-Josh

allbluedream · 2010-06-01 00:08:59

Forgot to mention that with 2.6.32 hibernation always worked. Looked like a regression to me. Waiting for kernel updates...

Kirurgs · 2010-06-03 19:19:12

Hmm, I have very the same problem. I run T500 and I'm pretty sure it's not the video what causes this as I run into the same problem with intel as ATI in use. Also, under heavy load > 4G RAM used it suspends ok, mostly, but there is this random stuff, from which You just don't know when it will fail. I got successful for 1.5 months or so and today it just failed, pressed Fn+F4 and little moon kept blinking forever
To me it seems some PM or Kernel stuff, I'll try to get .34 to work today, if it ever fails I'll notify you

Last edited by Kirurgs (2010-06-03 19:20:25)

eldragon · 2010-06-03 23:05:56

this article helps a lot on suspend issues. give it a thorough read.

http://en.opensuse.org/Suspend_to_RAM

joshua.finkbeiner · 2010-06-10 23:51:09

I've solved my problem, and I'm embarassed to say that I should have read the wiki more closely. I merely added 'resume' to my mkinitpcio.conf hooks (before filesystems but after all other hooks) and it fixed the problem.

See link:http://wiki.archlinux.org/index.php/Pm- … leshooting

-Josh

allbluedream · 2010-06-11 03:13:46

I believe we are talking about s2disk, rather than s2ram? Anyway, the fact that it is random means even if you have the resume hook in place, it wouldn't help all the time.

scriptmunkee · 2010-06-18 05:11:52

I figured out my suspend issue. For a while I've had a SD memory card inserted into the memory card reader of my laptop. I ran across a post mentioning to remove the SD card before trying to suspend or hibernate. I removed my SD card and now suspend is working as expected.

I'm going to look into this situation some more to figure our the true issue and file a proper bug.

scriptmunkee · 2010-06-18 05:32:56

Here is the solution I found to handle unmounting my SD/MMC card before suspending. I added the script to /etc/pm/sleep.d

# Drop to: /etc/pm/sleep.d
# Use this script to prevent data loss on mounted MMC/SD
# cards. It syncs data and umounts all mmcblk devices prior to
# suspend, and cancels suspend if umounting was not possible
# (i.e: something locks a file)
case "${1}" in
    hibernate|suspend)
        /bin/sync
        for drive in $( /bin/ls /dev/mmcblk?p* ); do
            /bin/umount ${drive} > /dev/null
            # If umount failed: abort suspend
            if [ $? -gt 0 ]; then
            # Test if device keeps mounted. Previous command could fail
            # (i.e device was not mounted) with a non-stopper
            # problem for the suspend process.
            /bin/mount | /bin/grep ${drive}
            if [ $? -eq 0 ]; then
                exit 1
            fi
            fi
        done
        ;;
#    resume|thaw)
##       Do nothing. All devices will be automatically mounted again.   
#       ;;
esac

This script was found on the following Ubunut wiki URL https://help.ubuntu.com/community/AspireOne/Ubuntu9.10

Arch Linux

#1 2010-05-19 13:13:49

pm-suspend freezes computer (though apparently suspends)

#2 2010-05-22 04:26:00

Re: pm-suspend freezes computer (though apparently suspends)

#3 2010-05-22 11:11:49

Re: pm-suspend freezes computer (though apparently suspends)

#4 2010-05-23 01:38:44

Re: pm-suspend freezes computer (though apparently suspends)

#5 2010-05-23 11:49:24

Re: pm-suspend freezes computer (though apparently suspends)

#6 2010-05-23 12:56:59

Re: pm-suspend freezes computer (though apparently suspends)

#7 2010-05-29 13:35:16

Re: pm-suspend freezes computer (though apparently suspends)

#8 2010-06-01 00:08:59

Re: pm-suspend freezes computer (though apparently suspends)

#9 2010-06-03 19:19:12

Re: pm-suspend freezes computer (though apparently suspends)

#10 2010-06-03 23:05:56

Re: pm-suspend freezes computer (though apparently suspends)

#11 2010-06-10 23:51:09

Re: pm-suspend freezes computer (though apparently suspends)

#12 2010-06-11 03:13:46

Re: pm-suspend freezes computer (though apparently suspends)

#13 2010-06-18 05:11:52

Re: pm-suspend freezes computer (though apparently suspends)

#14 2010-06-18 05:32:56

Re: pm-suspend freezes computer (though apparently suspends)

Board footer