Freerunner sometimes doesn't wake up

Andy Green andy at
Mon Jun 2 15:37:25 CEST 2008

Hash: SHA1

Somebody in the thread at some point said:

| Sounds like it has to be a race with the jbt6k74 driver because he's the
| only guy that cares (platform calls pcf50633_backlight_resume() when he
| resumes).
| I'll do some work in that area today for you to test rather than apply
| this patch that, according to the logic above, can't be striking at the
| root.

This stuff is on the "andy" branch.  I removed the code that you removed
about reinitializing the backlight to be off, it seems to me that the
pcf50633 somehow comes up with the backlight powered at a low level
before we wrote to it at all.

There's some new code to explicitly manage nasty resume dependencies if
we can identify them... FWIW I have never been able to reproduce the
crash on resume that seems to be associated with leaving the device in
suspend for a long while... I just tried with the last changes leaving
it for an hour and it is still fine here.

So I am curious if this makes any difference yet on your config.

- -Andy
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora -


More information about the openmoko-kernel mailing list