Hi again, thanks for your input > we have *intentionally* avoided doing writes
I see, I was unaware of that policy here, I think it is reasonable. > but I don't know what that code would look like I think I better make sure if adt(4) in particular is used somewhere beyond macppc and if so, perhaps an #ifdef approach would be a good solution. > noticed its fan starts to run as soon It's quite revealing what Nuno mentions here. Do you happen to know if the powermac 6,4 has an adt(4) device that might be driving this behavior? A dmesg would be nice, even if offlist to keep it neat here. > I'm not fluent in openfirmware I don't think there is a way but I'd have to triple check. I believe it gets reset every time you reboot regardless. > Alternatively, maybe in Jon's machine case Yes indeed I mentioned in my first post I based my adt(4) diff on that project, I'm quite aware of it. In fact, it's what inspired my suggestion of perhaps adding a sysctl for this - so that we are able to tweak these things from userspace like these kind of tools permit. However, in light of theo@'s reply, I'd be quite content with adding it as macppc only config, so long as it makes sense for all the machines that possess the chip. So far, I have tested it on a Powerbook 5,8 and 5,9, I'd dare say that for all powerbook 15" and 17" inches it should work fine. We could add runtime checks for e.g. power state and/or AC energy state to set this threshold. By the way, another small note regarding G4FanControl: It offers separate controls for each one of the 3 temperature targets whereas I am setting all three to the same value here. I played around a bit with G4FanControl on OSX leopard and it seems to be reasonable to do such a thing. In any case, I'd be happy to change this behavior to anything else that might make more sense.
