USB resume problem with IBM dock

Hi,

I hope I've come to the right place; I'm going based on the information in my kernel's MAINTAINERS file.

I have been using suspend2 with a 2.6.18 kernel, my Thinkpad T60p, and a Thinkpad Advanced dock. When resuming (from disk), suspend2 puts all the USB devices into a suspended state so they're consistent with the state at suspend time. If I have any devices plugged into the dock's USB ports, it's very common that the devices fail to enter a suspended state (or so the APIs report). When that happens, suspend2 forces a segfault and all heck breaks loose (log excerpt appended) -- suffice it to say I lose my suspended state, all the filesystems need to be checked, and often it takes an additional reboot just to make everything healthy again -- so it's really important to me to address this problem. Please let me know if there's anything I can do to help track this down.

Thanks,
Dave

--- /var/log/syslog excerpt ---

  Oct 11 22:11:40 pereiro kernel: [    7.004000] Atomic restore preparation
  Oct 11 22:11:40 pereiro kernel: [    7.008000] Freezing cpus ...
  Oct 11 22:11:40 pereiro kernel: [    7.080000] usb 5-6.1: new low speed USB device using ehci_hcd and address 5
  Oct 11 22:11:40 pereiro kernel: [    7.128000] CPU 1 is now offline
  Oct 11 22:11:40 pereiro kernel: [    7.128000] SMP alternatives: switching to UP code
  Oct 11 22:11:40 pereiro kernel: [    7.128000] CPU1 is down
  Oct 11 22:11:40 pereiro kernel: [    7.192000] usb 5-6.1: configuration #1 chosen from 1 choice
  Oct 11 22:11:40 pereiro kernel: [    7.392000] usb 5-6.2: new full speed USB device using ehci_hcd and address 6
  Oct 11 22:11:40 pereiro kernel: [    7.484000] usb 5-6.2: configuration #1 chosen from 1 choice
  Oct 11 22:11:40 pereiro kernel: [    7.484000] suspend_device(): usb_generic_suspend+0x0/0x15c [usbcore]() returns -16
  Oct 11 22:11:40 pereiro kernel: [    7.484000] usbcore: registered new driver hiddev
  Oct 11 22:11:40 pereiro kernel: [    7.484000] Could not suspend device 5-6: error -16
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Some devices failed to suspend
  Oct 11 22:11:40 pereiro kernel: [    7.492000] ------------[ cut here ]------------
  Oct 11 22:11:40 pereiro kernel: [    7.492000] kernel BUG at kernel/power/atomic_copy.c:496!
  Oct 11 22:11:40 pereiro kernel: [    7.492000] invalid opcode: 0000 [#1]
  Oct 11 22:11:40 pereiro kernel: [    7.492000] PREEMPT SMP 
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Modules linked in: usbhid dm_mod ide_generic ehci_hcd uhci_hcd usbcore piix generic sd_mod thermal processor fan fbcon tileblit font bitblit softcursor
  Oct 11 22:11:40 pereiro kernel: [    7.492000] CPU:    0
  Oct 11 22:11:40 pereiro kernel: [    7.492000] EIP:    0060:[suspend_atomic_restore+487/544]    Not tainted VLI
  Oct 11 22:11:40 pereiro kernel: [    7.492000] EFLAGS: 00010246   (2.6.18-local-20060823 #2) 
  Oct 11 22:11:40 pereiro kernel: [    7.492000] EIP is at suspend_atomic_restore+0x1e7/0x220
  Oct 11 22:11:40 pereiro kernel: [    7.492000] eax: 00004010   ebx: 00000000   ecx: 00000000   edx: 00000002
  Oct 11 22:11:40 pereiro kernel: [    7.492000] esi: 00000001   edi: 00000001   ebp: c0393270   esp: dfcadef0
  Oct 11 22:11:40 pereiro kernel: [    7.492000] ds: 007b   es: 007b   ss: 0068
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Process busybox (pid: 2428, ti=dfcac000 task=c21ed290 task.ti=dfcac000)
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Stack: c034fc04 c034cd8f 00000001 00000001 c014c6ee 00000000 c014c921 00000000 
  Oct 11 22:11:40 pereiro kernel: [    7.492000]        c014c7a7 c014e9f3 00000000 c13fb860 fffa8278 00000000 00000000 00000001 
  Oct 11 22:11:40 pereiro kernel: [    7.492000]        df802a40 00000001 00000001 df802a54 c01cc84c c03931b8 c0393270 dfdc3000 
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Call Trace:
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [suspend_resume+110/272] suspend_resume+0x6e/0x110
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [suspend_start_anything+81/192] suspend_start_anything+0x51/0xc0
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [__suspend_try_resume+23/48] __suspend_try_resume+0x17/0x30
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [suspend2_attr_store+99/688] suspend2_attr_store+0x63/0x2b0
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [sysfs_write_file+156/240] sysfs_write_file+0x9c/0xf0
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [vfs_write+222/448] vfs_write+0xde/0x1c0
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [sys_write+75/128] sys_write+0x4b/0x80
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  [syscall_call+7/11] syscall_call+0x7/0xb
  Oct 11 22:11:40 pereiro kernel: [    7.492000] Code: 16 00 0f 0b 13 02 74 cd 34 c0 83 c4 08 b8 01 00 00 00 5e 5f c3 c7 04 24 04 fc 34 c0 e8 23 5f fd ff a1 04 2d 39 c0 f6 c4 10 75 17 <0f> 0b f0 01 74 cd 34 c0 e9 6a fe ff ff e8 37 82 ff ff e9 47 fe 
  Oct 11 22:11:40 pereiro kernel: [    7.492000] EIP: [suspend_atomic_restore+487/544] suspend_atomic_restore+0x1e7/0x220 SS:ESP 0068:dfcadef0
  Oct 11 22:11:40 pereiro kernel: [    7.492000]  <6>JFS: nTxBlock = 8192, nTxLock = 65536

Dave Abrahams
Boost Consulting
www.boost-consulting.com

Hi Dave,
Does this happen when you have booted inside the dock station and then suspend to disk, then resume? Or is it something like, boot outside the dock station suspend to disk, resume inside dock? Can you be more specific about how to duplicate the problem?

Thanks,
Kristen

Hi Kristen,

Yes.

As a matter of fact, I can get resume to succeed by first unplugging all the USB devices while the machine is suspended.

Tried above. If you want more specificity than that, I think I'll need some more specific questions, because at this point I don't know what more to tell you.

I very much appreciate your attention to this issue :)

Dave Abrahams
Boost Consulting
www.boost-consulting.com