On Mon, Feb 13, 2006 at 10:07:03AM +0100, Michael Gernoth wrote:
> It happened to me, too. I didn't have debugging settings when this
> happened, so I had no chance to find out what is wrong. (Magic-SYSRQ via
> serial console should still work)
> 2.4.31 seems to be more stable in this regard, but I can't see what
> could cause this.
When this happened, the machine Oopsed. I was able to capture the Oops
via the serial console and find the bug in the keyboard-driver, which
was responsible for this. In previous version it seems that this did not
lead to a NULL pointer dereference but to an incorrectly repeated key
after resume.
linux-2.4.32-j720-8 which fixes this and should save more power (as the
speaker amplifier is not always switched on after boot) in now available
at:
http://wwwcip.informatik.uni-erlangen.de/~simigern/jornada-7xx/linux-2.4.32-j720/
Changes:
* Fix global speaker handling: Make sure the speaker is
really off after boot and use correct logic to turn it
on during playback
* Fix keyboard driver delivering not enough bytes on read
after suspend/resume. This fixes the crash after "resume"
* Remove some silly debug messages
* Give the MCU more time to answer
Regards,
Michael
Received on Tue Feb 14 2006 - 11:15:05 EST
This archive was generated by hypermail 2.2.0 : Tue Feb 14 2006 - 11:15:14 EST