Marcel Holtmann wrote:
 > Hi Andreas,
 >
 >
 >>>this should be included in the mainstream kernel source, but the code
 >>>must be a little bit cleaned up. I will check this later and send you a
 >>>patch for testing. After that I'll forward it to the maintainer.
 >>>
 >>
 >>I'll appreciate that, just let me know when you have gone over it and
 >>I'll check that it works as expected.
 >
 >
 > here is my patch for it. I changed some stuff and left some stupid
 > things out. It is now switched on/off with "bluetooth:1"/"bluetooth:0"
 > and please check that I don't break the other stuff like fan or screen
 > brightness.

Ok, first of all I want to say that it works ok for me. I have tried
everything except the "keys" proc entry.

However, I would like to change the behaviour a bit, to behave more
like the other proc entries and more like it was before your rework.
First I'd like a user friendly way of finding out if BT is really
enabled, and second I think you should be able to guess what to "echo" 
into the proc entry.

To enable fan running all the time you use, "echo force_on:1 > fan"
lcd brightness is set through "echo brightness:1 > lcd"

The output from "cat fan lcd" is:
brightness:              6
brightness_levels:       8
running:                 0
force_on:                0

Which means that when you have learnd how they function you should be
able to guess what to do with the bluetooth entry if it for example
looked like this:
#cat bluetooth
bluetooth_available:     15
wireless_switch:         512
bluetooth_on:            0

I have here changed the parameter to echo into "bluetooth" from
bluetooth to bluetooth_on, and added an entry that shows the status of
it. This makes it easy to guess that "echo bluetooth_on:1 > bluetooth"
should enable the device if you know how the other entris work.
Without this you want be able to guess that "bluetooth:1/0" is used
for enabling/disabling the device.

I have attached two patches, one is against "stock" 2.6.0-test9 and the
other is against your previous version on a patched 2.6.0-test9

I had to do some trial and error guessing the value 576, I don't
understand why it changed from 512 as it was in the patch I used
earlier. Maybe you understand and maybe it all should be done totally
different, but you should be able to understand what I want to do and
implement it in a prettier way if thats needed.

Please look at this and give me some feedback on your thoughts about it.

/Andreas