[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: radeonfb never redraws screen on wake



Hi,

To refresh our memories, I get a black display when my iBook wakes
from sleep.  Under OSX the machine can sleep and wake with a working
display.

The thread started here:

http://lists.debian.org/debian-powerpc/2003/debian-powerpc-200311/msg00283.html

After a while we got my kernel better configured, started trying to
get the new 2.6 radeon driver to work, etc.  The last message that I
am responding to was a month ago, you can read it here:

http://lists.debian.org/debian-powerpc/2003/debian-powerpc-200311/msg00512.html

On Fri, Nov 21, 2003 at 11:02:08AM +1100, Benjamin Herrenschmidt wrote:
> Try without pbbuttonsd, I don't know what it's mucking...

pbbuttonsd is now gone.

> > http://lists.debian.org/debian-powerpc/2003/debian-powerpc-200308/msg00414.html
> 
> well, it helps if you set CONFIG_PMAC_BACKLIGHT :)

now enabled and I have upgraded to 2.6.0-ben1.

Hitting the backlight down key (f1) has no noticable effect until the
15th hit at which point the display goes completely dark (backlight
off, no features on the display).  I can hit backlight up key (f2)
once and now the display is back to full brightness.  OSX has this
same problem (first few taps of backlight do nothing, then it turns
completely off.  The OSX install is an old version, it asks me to
upgrade, but I don't use it and am spooked it may make Debian
unbootable if I upgrade it).

Now that the backlight keys do something (that is,
CONFIG_PMAC_BACKLIGHT=y) I can test what they do after the machine
wakes and the screen is black.  Hitting the keys does the same thing
described above.  The screen remains black (but backlit) until the
15th hit at which point the screen goes completely black.  I can hit
the backlight up key (f2) to turn the backlight back on, but the
display remains black.

I looked around in the darwin source at
:pserver:anonymous@anoncvs.opendarwin.org:/Volumes/src/cvs/od but have
yet to find the radeon hardware access that control sleeping so as to
compare with the Linux driver.

Thanks for any more clues,

David



Reply to: