Suspend to RAM broken with 2.6.35 kernels on T500

[ReneS]

I tried the latest Maverick 2.6.35 kernels for regression testing on my Lenovo Thinkpad T500 with Ubuntu 9.10. All kernels including 2.6.35.4 show a suspend to RAM problem. The system simply freezes when suspending. Screen is still on, disk off, cpu fan off, but the moon still flashes. There is no way out besides a hard power down.

I attached to suspend logs from a success with 2.6.35.4 and a failure with 2.6.35.4. Suspend seems to work fine one or two times, but when doing it for the third time, the machines freezes.

No other log contains any information because some of the suspend information seems to be written after waking the machine up and therefore do not make it into the log when a hard restart is necessary.

This is from the kernel log, when the machine suspends correctly (but writes some of the log after waking up).

Aug 30 18:23:28 linpicard kernel: [ 8717.973146] PM: Syncing filesystems ... done.
Aug 30 18:23:28 linpicard kernel: [ 8717.997241] PM: Preparing system for mem sleep
Aug 30 20:52:43 linpicard kernel: [ 8719.061031] Freezing user space processes ... (elapsed 0.02 seconds) done.
Aug 30 20:52:43 linpicard kernel: [ 8719.090077] Freezing remaining freezable tasks ... (elapsed 0.01 seconds) done.
Aug 30 20:52:43 linpicard kernel: [ 8719.110057] PM: Entering mem sleep
Aug 30 20:52:43 linpicard kernel: [ 8719.110121] Suspending console(s) (use no_console_suspend to debug)
Aug 30 20:52:43 linpicard kernel: [ 8719.262302] sd 1:0:0:0: [sdb] Synchronizing SCSI cache
Aug 30 20:52:43 linpicard kernel: [ 8719.262511] sd 0:0:0:0: [sda] Synchronizing SCSI cache
Aug 30 20:52:43 linpicard kernel: [ 8719.263019] sd 0:0:0:0: [sda] Stopping disk
Aug 30 20:52:43 linpicard kernel: [ 8719.270256] sd 1:0:0:0: [sdb] Stopping disk
Aug 30 20:52:43 linpicard kernel: [ 8719.313497] ACPI handle has no context!
Aug 30 20:52:43 linpicard kernel: [ 8719.313505] sdhci-pci 0000:15:00.2: PCI INT C disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.313511] ACPI handle has no context!
Aug 30 20:52:43 linpicard kernel: [ 8719.318576] ACPI handle has no context!
Aug 30 20:52:43 linpicard kernel: [ 8719.318665] uhci_hcd 0000:00:1d.2: PCI INT C disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.318690] uhci_hcd 0000:00:1d.1: PCI INT B disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.318716] uhci_hcd 0000:00:1d.0: PCI INT A disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.318719] pciehp 0000:00:1c.4:pcie04: pciehp_suspend ENTRY
Aug 30 20:52:43 linpicard kernel: [ 8719.318723] pciehp 0000:00:1c.3:pcie04: pciehp_suspend ENTRY
Aug 30 20:52:43 linpicard kernel: [ 8719.318726] pciehp 0000:00:1c.2:pcie04: pciehp_suspend ENTRY
Aug 30 20:52:43 linpicard kernel: [ 8719.318730] pciehp 0000:00:1c.0:pcie04: pciehp_suspend ENTRY
Aug 30 20:52:43 linpicard kernel: [ 8719.318761] uhci_hcd 0000:00:1a.2: PCI INT C disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.318787] uhci_hcd 0000:00:1a.1: PCI INT B disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.318813] uhci_hcd 0000:00:1a.0: PCI INT A disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.319085] e1000e 0000:00:19.0: PCI INT A disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.319091] e1000e 0000:00:19.0: PME# enabled
Aug 30 20:52:43 linpicard kernel: [ 8719.319098] e1000e 0000:00:19.0: wake-up capability enabled by ACPI
Aug 30 20:52:43 linpicard kernel: [ 8719.330033] ehci_hcd 0000:00:1d.7: PCI INT D disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.330050] ehci_hcd 0000:00:1a.7: PCI INT D disabled
Aug 30 20:52:43 linpicard kernel: [ 8719.330089] pciehp 0000:00:1c.1:pcie04: pciehp_suspend ENTRY
Aug 30 20:52:43 linpicard kernel: [ 8719.330104] i915 0000:00:02.0: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8719.420338] HDA Intel 0000:00:1b.0: PCI INT B disabled
Aug 30 20:52:43 linpicard kernel: [ 8720.321272] PM: suspend of devices complete after 1210.730 msecs
Aug 30 20:52:43 linpicard kernel: [ 8720.321274] PM: suspend devices took 1.210 seconds
Aug 30 20:52:43 linpicard kernel: [ 8720.411296] ehci_hcd 0000:00:1d.7: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8720.471303] uhci_hcd 0000:00:1d.0: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8720.551303] ehci_hcd 0000:00:1a.7: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8720.611303] uhci_hcd 0000:00:1a.2: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8720.671303] uhci_hcd 0000:00:1a.0: power state changed by ACPI to D3
Aug 30 20:52:43 linpicard kernel: [ 8720.671342] PM: late suspend of devices complete after 350.064 msecs
Aug 30 20:52:43 linpicard kernel: [ 8720.771298] ACPI: Preparing to enter system sleep state S3
Aug 30 20:52:43 linpicard kernel: [ 8721.121324] PM: Saving platform NVS memory
Aug 30 20:52:43 linpicard kernel: [ 8721.133139] Disabling non-boot CPUs ...
Aug 30 20:52:43 linpicard kernel: [ 8721.250032] CPU 1 is now offline

[ReneS (mail-03146f06)]

Just to add that all kernels before 2.6.35, such as the currently running 2.6.34.6 are running fine in terms of suspend to RAM.

[ReneS (mail-03146f06)]

Found that as potential suspect: http://linux.derkeiler.com/Mailing-Lists/Kernel/2010-07/msg06421.html

[peterfalter (peter-sowosama)]

I have a quite similar experience with my T40 ThinkPad:
suspend works only when i dont have my WLAN USB stick attached.
With the stick the moon keeps flashing and i have to hardreset.
And: a mounted 1GB Flashstick keeps suspend also working.

In 10.04 suspend is ok together with WLAN stick.

[Miguel J (mjulier)]

Same symptoms for me. It used to work fine, but now it works just the 2 or 3 first times (after a reboot) and then the last try gives a freeze (screen on, but nothing responding, have to hard reboot).

This is an ASUS X51L, with Intel dual-core CPU.
2.6.32-25-generic #44-Ubuntu SMP Fri Sep 17 20:26:08 UTC 2010 i686 GNU/Linux

[Akagi (akagi010)]

I can confirm this problem on my Lenovo T500 with a fresh installed Ubuntu 10.10. It seems that the process is somehow interrupted. There is a command line visible on screen and the "Sleep" LED blinks, but the system does not sleep.

[Alin Trăistaru (alindt)]

Probably https://bugzilla.kernel.org/show_bug.cgi?id=16256

After unloading tpm_tis my T500 suspends just fine.

[Akagi (akagi010)]

@alindt: Thanks for your suggestion. But it seems to be a different problem:

I tried:
Remove tpm_tis - Suspend failed
Remove all tpm modules - Suspend failed
Deactivate TPM via BIOS - Suspend failed

Funny. Back in Ubuntu 9.04 (IIRC) I also could not suspend because I deactivated the TPM in BIOS (an inactive TPM worked)

Best regards!

[howessc (steve-stevenhowes)]

Worked fine on lucid. Worked fine on maverick until two days ago (perhaps more recent). My experience is essentially same as what is documented above. I installed clean from ubuntu-10.10-beta-alternate-amd64.iso in early September.

[gdi2k (gdi2k)]

This affects me too on a Lenovo X200. Syslog when failing to suspend is attached.

Notes:
 - I suspended at 06:47, which failed, so I powered off.
 - I powered back on 7:56. Looks like the suspend tried to continue at that point.

I had no issues suspending with 10.04 and previous versions of Ubuntu.

[pliz (pliz)]

@Akagi, thanks for the solution!
sudo rmmod tpm_tis
solved the problem for me - now I need to make it permanent

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.