Most recent kernel where this bug did not occur: n/a Distribution: SLED10 with kernel.org kernel Hardware Environment: Dell Precision M90 laptop Dell D/Dock Expansion Station Problem Description: When repeatedly docking and undocking the Dell M90 from the dock station, the hardware dock eject button will no longer cause the dock to be released cleanly from the dock station. The led remains green and does not indicate that it is safe to dock. This can happen after an indeterminate number of dock/undocks - it is variable. Steps to reproduce: 1. Important - this problem seems to only occur if you boot the laptop inside the dock station, then begin to undock/dock repeatedly. Boot with laptop in dock station. 2. Load dock drivers: modprobe dock; modprobe acpiphp 3. undock and dock repeatedly until failure occurs. This will take a variable number of tries.
I also tested the M65 which has the same problem, however, I was able to duplicate this issue without booting in the dock station first. I expect that I just wasn't patient enough to find it with the M90. I will investigate further to see if I can tell if this is a dsdt problem, or a driver issue.
I beleave I've got a similiar problem with the IBM R52 and the Dock II: After booting in the docking station, everything works, but undocking the machine leaves the docking bridge as 0000:0b:03.0 ffff: Texas Instruments PCI2032 PCI Docking Bridge (rev ff) in lspci. After this, the station refuses to initialize upon docking and the laptop refuses to go into suspend until rebooted properly. After booting outside the docking station, docking works but pressing the eject button won't release the notebook (the docking station stays in powered up mode but the extension slots are removed) until it has been powered down. It may be of interest that the frequency of this happening seems to vary from docking station to docking station - mine shows this behaviour everytime I use it, 2 other stations show it in a seemingly nondeterministic fashion (after a varying number of docking/undocking operations)
Gregor, Thanks for the report! Can you tell me if you are loading the acpiphp (ACPI based PCI hotplug) module in addition to the dock driver? On the Dock II, due to the PCI slots on the station you do need to have both drivers loaded in order to get the PCI bridge/devices inserted and removed properly.
Hi Kristen. thx for the quick response. acpiphp is compiled into the kernel. See: http://graphics.cs.uni-sb.de/~grsch/dock/dot.config http://graphics.cs.uni-sb.de/~grsch/dock/lsmod http://graphics.cs.uni-sb.de/~grsch/dock/kernel.params [State: booted outside the dock, undocked ;) , kernel 2.6.18] Regards, Gregor.
I have similar problem with this bugreport in my ThinkPad A21m. acpiphp module with parameter debug=1 loaded. booting with dock: -[0000:00]-+-00.0 Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX Host bridge +-01.0-[0000:01]----00.0 ATI Technologies Inc Rage Mobility P/M AGP 2x +-02.0 Texas Instruments PCI1450 +-02.1 Texas Instruments PCI1450 +-03.0 Intel Corporation 82557/8/9 [Ethernet Pro 100] +-03.1 Xircom Mini-PCI V.90 56k Modem +-04.0-[0000:08-0e]--+-01.0 Silicon Image, Inc. PCI0648 | +-02.0 Texas Instruments PCI1420 | \-02.1 Texas Instruments PCI1420 +-05.0 Cirrus Logic CS 4614/22/24 [CrystalClear SoundFusion Audio Accelerator] +-07.0 Intel Corporation 82371AB/EB/MB PIIX4 ISA +-07.1 Intel Corporation 82371AB/EB/MB PIIX4 IDE +-07.2 Intel Corporation 82371AB/EB/MB PIIX4 USB \-07.3 Intel Corporation 82371AB/EB/MB PIIX4 ACPI booting without dock and __after docking__ . -[0000:00]-+-00.0 Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX Host bridge +-01.0-[0000:01]----00.0 ATI Technologies Inc Rage Mobility P/M AGP 2x +-02.0 Texas Instruments PCI1450 +-02.1 Texas Instruments PCI1450 +-03.0 Intel Corporation 82557/8/9 [Ethernet Pro 100] +-03.1 Xircom Mini-PCI V.90 56k Modem +-05.0 Cirrus Logic CS 4614/22/24 [CrystalClear SoundFusion Audio Accelerator] +-07.0 Intel Corporation 82371AB/EB/MB PIIX4 ISA +-07.1 Intel Corporation 82371AB/EB/MB PIIX4 IDE +-07.2 Intel Corporation 82371AB/EB/MB PIIX4 USB \-07.3 Intel Corporation 82371AB/EB/MB PIIX4 ACPI
Hi - can I provide you with any more details to help resolving the issue? Regards, Gregor.
Created attachment 9562 [details] kernel.log after power on to slot Hi, i am playing with hotplug and docking. There is new expirience. If I docking, no new devices shown with lspci output. After echo 1 > /sys/bus/pci/slots/1/power PCI devices on the dock are shown and lspci output is OK. But there is many error. lsusb dot not show usb controler on slot. Look at the attachment. Tested with todays linux-git kernel.
On Sat, 18 Nov 2006 15:01:13 -0800 bugme-daemon@bugzilla.kernel.org wrote: > http://bugzilla.kernel.org/show_bug.cgi?id=7048 > > > > > > ------- Additional Comments From goldenfish@bsys.cz 2006-11-18 14:59 ------- > Created an attachment (id=9562) > --> (http://bugzilla.kernel.org/attachment.cgi?id=9562&action=view) > kernel.log after power on to slot > > Hi, > > i am playing with hotplug and docking. There is new expirience. > If I docking, no new devices shown with lspci output. > After > echo 1 > /sys/bus/pci/slots/1/power you can't dock/undock using the sysfs pci slot entries. Use the physical buttons on your dock station. Please post additional questions to the linux-acpi mailing list, as this does not have anything to do with this bugzilla entry. Thanks, Kristen > > PCI devices on the dock are shown and lspci output is OK. > But there is many error. lsusb dot not show usb controler on slot. Look at the > attachment. > Tested with todays linux-git kernel. > > ------- You are receiving this mail because: ------- > You are the assignee for the bug, or are watching the assignee. > You reported the bug, or are watching the reporter. >
On Thu, 16 Nov 2006 12:33:41 -0800 bugme-daemon@bugzilla.kernel.org wrote: > http://bugzilla.kernel.org/show_bug.cgi?id=7048 > > > > > > ------- Additional Comments From grsch-kernel@ran-dom.org 2006-11-16 12:32 ------- > Hi - can I provide you with any more details to help resolving the issue? > > Regards, > Gregor. Gregor, Can you please open a separate bugzilla entry for this? I don't believe it is the same problem. go ahead and put bug 7048 as a dependency though so we can see if they are related. Sorry for the delay, I was on vacation. Kristen > > ------- You are receiving this mail because: ------- > You are the assignee for the bug, or are watching the assignee. > You reported the bug, or are watching the reporter. >
hi kristen. np - but there was an interesting development yesterday when I realized that there was another way for hotswapping ultrabay devices than lt_hotswap and threw it out of my setup: This evening, I couldn't undock even a single time anymore until the next morning... after cutting the power to the dock and rebooting once in the docking station, the problem's (almost) gone (now it seems as if only the pc-card slots were not reinitialized upon reinsertion of the laptop into the dock). So it seems you're right: there seems to have been a problem using lt_hotswap together with acpiphp+dock... thx for the effort and (though I'm still a bit sceptical about the current peace ;) ) maybe some other folks with thinkpads share the same cause for undocking problems... Regards, Gregor.
Oh yes - you absolutely CANNOT use lt_hotswap with the new dock driver, because it tries to do the same thing that dock.ko does - and as you saw from your experience, it does not do it correctly. you don't need it anyway. Kristen On Thu, 30 Nov 2006 04:11:29 -0800 bugme-daemon@bugzilla.kernel.org wrote: > http://bugzilla.kernel.org/show_bug.cgi?id=7048 > > > > > > ------- Additional Comments From grsch-kernel@ran-dom.org 2006-11-30 04:08 ------- > hi kristen. > np - but there was an interesting development yesterday when I realized that > there was another way for hotswapping ultrabay devices than lt_hotswap and threw > it out of my setup: This evening, I couldn't undock even a single time anymore > until the next morning... after cutting the power to the dock and rebooting once > in the docking station, the problem's (almost) gone (now it seems as if only the > pc-card slots were not reinitialized upon reinsertion of the laptop into the dock). > So it seems you're right: there seems to have been a problem using lt_hotswap > together with acpiphp+dock... > > thx for the effort and (though I'm still a bit sceptical about the current peace > ;) ) maybe some other folks with thinkpads share the same cause for undocking > problems... > > Regards, > Gregor. > > > ------- You are receiving this mail because: ------- > You are the assignee for the bug, or are watching the assignee. > You reported the bug, or are watching the reporter. >
Created attachment 9953 [details] full acpi debug from docking TP A21m. size - 8MB uncompressed
Confirmed this problem still exists in 2.6.21-rc3. Also, I will get an OOPs in acpid if the laptop is redocked prior to the completion of the _DCK method, which seems to take an extremely long time on the Dell laptops (in the order of 10s of seconds). Am checking to see if an updated BIOS is available - but it would be nice if acpid did not oops.
BIOS update for M65 fixes the issue with the exception of led flashing.