Touchpad might lockup on some systems

[mfoster]

Due to the same ESD sensitivity documented in Trac #473, it is possible that the touchpad assembly might lockup and not generate cursor movement. If this occurs, the temporary workaround is to power the unit off and on again to reset the er rors induced when static buildup discharges through the touchpad's controller ASIC.

This is a very high priority fix for the B2 systems - details on the fix are noted in Trac #473.

[mfoster]

In addition to the fixes in Trac #473, an additional change will be made to the touchpad's sensor assembly. Currently, an internal ground layer is present in the pad, but this is improperly implemented by the touchpad manufacturer. In B2 systems, a separate metal plate will underlie the touchpad assembly to provide a direct ground path to chassis ground.

[90030501]

Frank:Add grounding on Hinge position and TP. We will get sample on 12/10/06'

[90030501]

Frank: Sorry we can not get sample on time. Cause we stop some tooling modify for 12/03/06' meeting. We will get sample on 12/25/06'.

[jg]

I've had my touchpad lock up on my B2 system once: it was not apparently an EC problem, as a reboot cleared the problem (no battery removal was required). Performing a recalibrate operation did not help (maybe we should suggest to ALPS that recalibrate also try to reset the touch pad chip if possible).

If I figure out how to reproduce this I'll update this note.

[warp]

On jg's lockup. It is possible that this may have been the (now fixed) race condition in the GS to PT switch.

That would have left the touchpad in PT mode, with no indication that this was the case.

Reproduction would involve a build prior to the fix entering the kernel, and a very brief contact with a fingernail or other object that would have been detected as a stylus.

My inclination is to close this and open a new bug if it shows up again on a kernel with the GS to PT switch race fix present. Jim?

[jg]

My report should be separated out; it is clearly different than the ESD error.

So far, so good on B2; I've not heard of any EC lockups (lockups requiring removing power from the system to clear.

[jg]

I think we would have heard of problems by now; it appears that we are through the ESD problems we had in B1, so I'm closing this out.

We do need to do ESD testing at B3 bringup, of course, in case something sneaks in.