Thinkpads Forum

Hi folks,

I have an issue with a T430u since a few weeks.

The On-site service was here twice, exchanging the system board and palmrest/touchpad. The technician could not fix the reboots.
Me and the technician could also not explain why this happens.

Disconnecting the touchpad from the system board stops this issue.
Under Ubuntu it's rather hard to make it rebooting, but it still tends to reboot randomly from time to time. It can run stable under load for many hours.
In fact, if I don't touch it in memtest86+, it will also run many hours without errors.

I now agreed with Lenovo to send the T430u to their repair center, so hopefully a defect will be found.
Already tried testing with removing all components, like PCIe cards, hard drive, keyboard, Changing RAM slots, using different RAM etc.
Only disconnecting the Touchpad from the system board seems to help. And keep in mind they already exchanged system board and touchpad.
The new touchpad seems to be even more sensitive, before I needed to apply a little pressure sometimes to make it reboot, now just touching seems to be enough.

I made a video of the issue I'm having:
http://www.youtube.com/watch?v=UVJ-7-sqMeM



Thinkpad T430u Users:

Please test what I did and boot into Memtest86+ (download from http://www.memtest.org/ ) and touch or apply a little pressure to the point I marked on my Touchpad in the video.

No one?

Please realize that the majority on this Forum dislikes anything that is newer than T420/T520/W420/W520/X220 (all 2012 models).
They are no longer considered real Thinkpads, thanks to the royal screwups from Lenovo, notably the horrible keyboards and screens.
And because of the above, very few people on the forum have such a machine, hence the loud silence!

If both the touchpad and the board were replaced, and the problem persists, then the only thing I can think of (besides an amazing coincidence of two boards/pads with the same uncommon defect), is something short-circuiting in the case.

RealBlackStuff wrote:They are no longer considered real Thinkpads, thanks to the royal screwups from Lenovo, notably the horrible keyboards and screens.

Well, the screens have actually improved in some models (notable, IPS is more available since the *40 series), and the keyboards are far from horrible. Just the layout is.

Just FYI,

I found another T430u user and it happens to him too. So this seems to be some kind of firmware bug? As it is stable under Windows.

If it's stable under Windows, try also a different Linux distribution beyond Ubuntu, like Linux Mint, Debian or whatever and on a different graphic environment. See if the issue persists.

Went back to your video, just saw that you're running memtest86+.

AFAIK, memtest86+ is a stand-alone program, so I would venture that
Ubuntu / Linux has nothing to do with the problem.

bit_twiddler wrote:Went back to your video, just saw that you're running memtest86+.

AFAIK, memtest86+ is a stand-alone program, so I would venture that Ubuntu / Linux has nothing to do with the problem.

Ya like I said, it might be some sort of firmware bug that is addressed in the Windows drivers, so it doesn't cause issues with Windows.

Memtest86+ is very low level C code as far as I know, and this is the very first computer where I ever saw it crashing for a different reason than hardware failure.

One way to test your hypothesis would be to turn off
the touchpad in the BIOS - if you still get the reboot
under memtest86+, then it's probably not firmware,
it might be an electrical problem.

bit_twiddler wrote:One way to test your hypothesis would be to turn off the touchpad in the BIOS - if you still get the reboot under memtest86+, then it's probably not firmware, it might be an electrical problem.

The problem with this is, that you can't disable it in the T430u BIOS, was surprised myself, but they did not implement it. Disconnecting the Touchpad cable from the Motherboard stops it.

Also this doesn't happen under Windows at all, so I really doubt it's a electrical problem. They also exchanged the System Board and Touchpad+cables already.

I don't know a whole lot about clickpads, but I notice that you are hitting the clickpad in the same place each time to produce the behavior. Does this happen when you depress the clickpad elsewhere? Does your clickpad emulate mouse buttons or have some kind of adjustable behavior w/respect to the position where you touch?

You might be able to get around the problem by turning off any mouse button emulation by the clickpad. I'm not sure what memtest86+ is telling us except that it thinks that it has received some kind of input event.

Also, it looks like it is going back to the BIOS, and then immediately back to memtest86+. Because of that, I'm confused as to exactly what is happening. What is your boot sequence? Are you booting off a USB drive directly into memtest86+, or through GRUB? We don't see the GRUB prompt, so perhaps you could explain how you have set things up to boot.

I don't know about emulation, there's really not much you can set in BIOS, would need to doble check, but I don't think so.

I removed all kinds of components to make sure it's not caused by those. I booted from a bootable Memtest86+ USB-stick directly, created by the installer you can find on memtest.org, it's very minimal, using syslinux for booting I think.

It does not happen at all positions, this one I showed is particulary easy.
Before the Touchpad was exchanged for a new one by the Lenovo technician, it was actually needed to apply a little more pressure, now just almost touching it is enough to cause a reboot.
So it kind a got "worse" with the new one.
It could be some kind of input that the Touchpad receives and that can cause Memtest to reboot...

Have you noticed this happen with the trackpad assembly separated from the system? I'm thinking this may be an electrical issue also, and probably that Windows is not affected because of something technical, maybe because it runs in protected mode, whereas Memtest86 runs in real mode.

It took me some time getting a freeze-frame from your YouTube video on the exact moment your system re-booted. It appears that memtest86+ receives an ESCAPE KEY sequence. There is a "Halting" message just before the screen goes black. This is exactly what memetest86+ will do if you press the Escape key.

Now, WHY does touching the touchpad do this? I don't know. Is it possible that there's some pre-programmed swipe sequence that simulates an Escape?

One way to verify this would be to get into Windows and open up any screen/window which is designed to automatically close when the Escape key is pressed. One such window is the Clock. Double click the clock/time on the Taskbar for Windows XP or single click for Windows 7. That should launch the Clock applet. As long as it has keyboard focus, it will close if it detects an Escape key.

rkawakami wrote:It took me some time getting a freeze-frame from your YouTube video on the exact moment your system re-booted. It appears that memtest86+ receives an ESCAPE KEY sequence. There is a "Halting" message just before the screen goes black. This is exactly what memetest86+ will do if you press the Escape key.

Now, WHY does touching the touchpad do this? I don't know. Is it possible that there's some pre-programmed swipe sequence that simulates an Escape?

One way to verify this would be to get into Windows and open up any screen/window which is designed to automatically close when the Escape key is pressed. One such window is the Clock. Double click the clock/time on the Taskbar for Windows XP or single click for Windows 7. That should launch the Clock applet. As long as it has keyboard focus, it will close if it detects an Escape key.

Wow, never would have thought that video contains useful evidence...thanks!

Will check on that ESC key thing and report back.

Okay,


Checked in Windows, pressing the same spot that causes the reboot in Memtest did not "emulate" the ESC key anywhere,

Not in Clock, not in the Lock screen or anywhere else.
Also booted into freedos and Linux, but no ESC key press seems to be detected.

Since freedos does not load any mouse drivers and shold be pretty low level too, I thought at least there it should behave in this way?

Maybe if I unload the mouse driver in Linux and tail the event device of the Touchpad?

I don't have time to look at the code for memtest86+, but I would bet that
it sets up interrupts to handle keyboard events and the like.

The behavior that you get when you hit the touchpad in that one
spot may be getting translated into something that memtest86+ is
handling (such as keyboard input) or it could just fire off an
interrupt which memtest86+ is not handling.

In other words, you have a default set of interrupts that your
bios is responsible for setting up, and then whatever program
is loaded is able to override that setup. So, it's perfectly
reasonable that Windows is overriding whatever your
bios has set up.

Just out of interest, is your bios set to boot in legacy (i.e., non-uefi) mode?

Yes it's booting non-uefi.

What happens if you go into the bios (to do setup, or whatever) and
you hit the trackpad at the same spot? Does it reboot?

bit_twiddler wrote:What happens if you go into the bios (to do setup, or whatever) and
you hit the trackpad at the same spot? Does it reboot?

No it does not.

No it does not.

That's too bad. I'm not sure what the behavior under memtest86+ indicates.
You could try reporting it as a bug to Canonical (Ubuntu), and see if they
can track it down. The fact that another T430u user is having the same
problem should be brought to their attention.

The T430 was apparently certified to run with Ubuntu 12.04. You might try
an install of that, and if the problem disappears, then the Ubuntu
folks might be able to bisect the problem if it is related to a single
component, such as the Synaptics driver that they are shipping.