Re: a Kernel hack for a Powerbook thermal sensor gone crazy?
site_archiver@lists.apple.com Delivered-To: darwin-kernel@lists.apple.com For example (PBG4_DPSCtrlLoop.cpp): I thought this was the right place... isn't it? Fabio Cionini On 25 lug 2006, at 18:56, Michael Smith wrote: On Jul 25, 2006, at 9:39 AM, Fabio Cionini wrote: = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.a... thanks for your answer... applecare says that the guarantee period is over so I have to pay. No more no less. By the way, I thought the files I mentioned were kind of sensor drivers, and modifying the driver would have modified the relationship between the kernel and the sensor itself. So how the thing is working? How kernel receives values from the sensor? And what does these files I mentioned do if not checking the sensor values? 482 // fetch the current trackpad temperature state 483 if (safeGetState(fTPadTempSensor) >= 1) { 484 newMetaState = gIOPPluginOne; 485 CTRLLOOP_DLOG( "PBG4_DPSCtrlLoop::updateMetaState trackpad non- zero temp state 0x%lx\n", safeGetState(fTPadTempSensor)); 486 } The reason why I'm writing all this to a kernel developing mailing list is that I thought about the issue and maybe there could be a way to fix it without replacing any hardware. I'm kinda newbie in that, but I browsed the darwin sources using this darwingrok site http://darwingrok.physics.ox.ac.uk:8080/source/xref/ and found that this file http://darwingrok.physics.ox.ac.uk:8080/source/xref/ AppleMacRISC4PE/IOPlatformPlugins/PBG4_PlatformPlugin/CtrlLoops/ PBG4_DPSCtrlLoop.cpp (lines 97 and 483), among with others in the PBG4_PlatformPlugin of the AppleMacRISC4PE project, there are references to that specific thermal sensor, as long as methods to get the sensor value. Do you think it's possible to fake the kernel by modifying these lines, to send a hardcoded value just to avoid system craziness, or just comment the lines to avoid that specific sensor checking? The other three sensors are working ok so I don't think this hack would lead to any danger. There's no code in the kernel per se that does anything specific with that sensor, so there's nothing you would be able to change about it. I tried to compile the AppleMacRISC4PE project but I failed (I think I'm too newbie) even without modifying anything, so I'm asking help to you with this issue. Maybe it's just a small modification. In your shoes, I would be looking for a second-hand top case, after talking to AppleCare to see what they had to say. This email sent to site_archiver@lists.apple.com
participants (1)
-
Fabio Cionini