Created attachment 74321 [details] lspci -vvnn Hi, Lenovo S205 fails to wake up from suspend / hibernate. I have to long-press the power button to turn the system off and reboot. Related Ubuntu bug report: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1015642 uname -a: Linux jochen-Ideapad-S205 3.5.0-030500rc3-generic #201206162135 SMP Sun Jun 17 01:36:04 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Created attachment 74331 [details] dmidecode
When I followed https://wiki.ubuntu.com/DebuggingKernelSuspend, in 1 out of 3 runs, this line appeared in dmesg after rebooting: tty tty62: hash matches
Created attachment 74341 [details] /proc/modules
Created attachment 74351 [details] /proc/ioports
Created attachment 74361 [details] /proc/iomem
Please describe what the LEDs do (if there are any) when you suspend, and when you try to provoke a wake with a single power button press. Typically the power or sleep LED will blink periodically when you are asleep, and that will stop when you wake. Re: comment #2 The idea is that you look for the last "hash matches" entry in the dmesg and then unload that device before suspsending the next time to see if that device is what caused the problem and if you can get further. It may be useful to attach the resulting messages files from this process. BTW. do you see the same failure on suspend to RAM and on hibernate to disk?
ping...
Sorry for being so silent. Unfortunately, I don't have access to the device at the moment. I hope to be able to test it at some time, but I cannot tell when this will be.
Ok. Thanks for reply. Mark this bug as unreproducible. When you can reproduce the bug, please feel free to reopen this bug.
Hi Peck: Please check whether the bug still exist in the latest upstream kernel v3.9-rc6 or stable tree 3.8. If it sill existed, please follow Comment #6 to debug. Thanks.
Hi, I tested the latest kernel from the ubuntu mainline archive: Linux ideapad 3.9.0-030900rc6-generic #201304080035 SMP Mon Apr 8 04:36:25 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux Still the same behaviour with suspend and also shutdown (power off) of the computer: I have to long-press the power-button to turn it off, otherwise the fan will go into heavy load after a while. As for the LEDs: all LEDs except the HDD light are on: - power - LED in power button - battery - wireless Any more info that is needed?
Sorry for later reply. busy on other thing. (In reply to comment #11) > As for the LEDs: all LEDs except the HDD light are on: > - power > - LED in power button > - battery > - wireless Do you mean all these leds are on? > Any more info that is needed? Please follow this doc to do basic debug. https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/Documentation/power/basic-pm-debugging.txt From the comment #2, please unload tty driver and test again.
Yes, all these leds are on. I see one "tty tty12: hash matches" line in the /var/log/dmesg logs. Sorry for the basic question, but how do I "unload tty driver"? I'll try follow the instructions from the basic debug document, and get back with the results.
The "tty tty12: hash matches" line only in comment #12 appears only in a kernel 3.5 boot, not in the kernel 3.9rc6 boots that I tested recently, but still show the same buggy behaviour.
I followed the instructions in "basic-pm-debugging.txt", using kernel 3.9-rc8 with CONFIG_PM_DEBUG enabled. I tried both hibernation modes ("platform" and "shutdown"). 1) Using the "platform" hibernation mode: Testing works up to, but not including the "platform" test, i.e., the following fails: # echo platform > /sys/power/pm_test # echo platform > /sys/power/disk # echo disk > /sys/power/state Where "fails" means: - the screen is dark - the cpu fan is running - all leds are on (except HDD) - I cannot connect to the laptop (e.g., via ssh, so processes have been stopped) - I have to long-press the power button to shut down the laptop I notice a clicking sound at the end of the hibernation sequence, probably the hard drive being shut down. 2) Using the "shutdown" hibernation mode: Up to, and including the "core" test are running fine, that means, the laptop is coming back to life after being dark for a few seconds. I.e., the following works: # echo core > /sys/power/pm_test # echo platform > /sys/power/disk # echo disk > /sys/power/state Using the next commands to simulate a real hibernation, however fails partly: # echo core > /sys/power/pm_test # echo platform > /sys/power/disk # echo disk > /sys/power/state It fails partly because it exhibits the same symptoms as above on hibernation (screen dark, leds on etc.). I have to long-press the power-button to forcibly shutdown the laptop. However, after powering the laptop on again and booting, the resume procedure kicks in, and the system restored to the same state that it had just before running the hibernation commands. So, resuming the system from STD seems to work just fine.
Sorry for later reply due to be busy on the other thing. (In reply to comment #15) > I followed the instructions in "basic-pm-debugging.txt", using kernel 3.9-rc8 > with CONFIG_PM_DEBUG enabled. I tried both hibernation modes ("platform" and > "shutdown"). > > 1) Using the "platform" hibernation mode: > > Testing works up to, but not including the "platform" test, i.e., the > following > fails: > > # echo platform > /sys/power/pm_test > # echo platform > /sys/power/disk > # echo disk > /sys/power/state > > Where "fails" means: > - the screen is dark > - the cpu fan is running > - all leds are on (except HDD) > - I cannot connect to the laptop (e.g., via ssh, so processes have been > stopped) > - I have to long-press the power button to shut down the laptop > > I notice a clicking sound at the end of the hibernation sequence, probably > the > hard drive being shut down. > > 2) Using the "shutdown" hibernation mode: > > Up to, and including the "core" test are running fine, that means, the laptop > is coming back to life after being dark for a few seconds. I.e., the > following > works: > > # echo core > /sys/power/pm_test > # echo platform > /sys/power/disk > # echo disk > /sys/power/state > > Using the next commands to simulate a real hibernation, however fails partly: > > # echo core > /sys/power/pm_test > # echo platform > /sys/power/disk > # echo disk > /sys/power/state The previous commands are same so I am not sure which option make the hibernation broken. Is there something I miss? > > It fails partly because it exhibits the same symptoms as above on hibernation > (screen dark, leds on etc.). I have to long-press the power-button to > forcibly > shutdown the laptop. > > However, after powering the laptop on again and booting, the resume procedure > kicks in, and the system restored to the same state that it had just before > running the hibernation commands. So, resuming the system from STD seems to > work just fine. This is very strange.
Sorry, I mixed up the sequences for the shutdown hibernation test. That's the sequence that works: # echo core > /sys/power/pm_test # echo shutdown > /sys/power/disk # echo disk > /sys/power/state And that's the sequence that fails to shut down properly, but restores OK after a forced reboot: # echo none > /sys/power/pm_test # echo shutdown > /sys/power/disk # echo disk > /sys/power/state
With the 3.9.0-rc8 kernel, I still have the problem to power off the device (e.g., using the "poweroff" command) - only long-pressing the power button will work. I think that the primary issue with this device is that it doesn't power off cleanly, and we should focus on this issue first.
Yes, I also think this is power off issue since all the pm tests have passed through. Dose "halt" command work?
(In reply to comment #19) > Yes, I also think this is power off issue since all the pm tests have passed > through. > Dose "halt" command work? The "halt" command has the same effect as "poweroff". "reboot" works, however.
Any chance you can update your BIOS? BIOS Information Vendor: LENOVO Version: 4BCN20WW Release Date: 04/15/2011 latest on lenovo web site is 17 Aug 2011. (unfortunately, it appears to be a .exe that you'd run in Windows -- if you've got a windows partition, simple, otherwise, maybe not simple) The reason that I ask is that the last stages of powerdown are handled by the BIOS, and so sometimes there is a BIOS dependency. BTW. does reboot also fail, or is that more reliable than powerdown?
Hi Len: From Comment #20, reboot is fine.
I don't have Windows installed. I tried this procedure but it didn't work for me: https://s205blog.wordpress.com/2013/01/13/lenovo-s205-bios-recovery-updating-bios-without-installed-windows-os/. I'll give it another try, however, and report back. Talking about BIOS: I'm running the laptop in UEFI mode. Does it make a difference here?
I updated the BIOS to the latest version 4BCN24WW, which itself was not trivial, but doable using the information from my previous comment. I a live CD of Ubuntu Raring which uses kernel 3.8. Using that setup, suspend to ram and resume worked flawlessly. However, I couldn't get it to install on HDD because efibootmgr didn't create the uefi boot entries required to boot from the HDD, probably because of this bug: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1167622 Trying to create the boot entries with an older version of Ubuntu (12.04), I somehow managed to make the laptop unbootable. "Unbootable" means that the computer starts and the display shows the BIOs screen, but immediately becomes unresponsive. It seems like the BIOS crashes (maybe because of faulty boot entries?), any effort of repairing it have failed so far (that includes trying to boot into recovery mode using Fn+R, and removing the CMOS battery). So I'd say this issue can be closed, because there's a solution to it, but unfortunately, the solution has some severe issues itself and is not without risks.
OK. So the newest Bios resolved the power off issue. So close this bug. For your problem, how about using a usb key to install ubuntu? https://help.ubuntu.com/community/Installation/FromUSBStick