Information Collection for Laptop LCD hack

[Guest BuildSmart]

Regarding the wireless issue, I've resolved this for every machine I've had wireless issues with and it's nothing more than a proper and complete package because whomever originaly created it missed some required files.

 

With the LCD issue, I have an understanding of what is going on with the locked/hidden partition and I believe it's only purpose is to prevent anyone from accessing a file.

 

I also believes that that partition method currently implemented serves no other purpose than to hide the file from general access and I've managed to back-trace what it's doing and it's basically checking hardware for an active LCD and then skipping any further display checks by returning true for a display match so it doesn't try initializing the VGA port which is similar to the fn-f7 and this can be patched with a hex editor to assume this key combination has been activated.

 

The QE/CI issue is one based on people not implementing all of the files and is a good reason why things need to change.

 

My biggest problem is getting those who are involved with the project to work collectively so until that happens I'm not sure when something will be available for the general public.

 

I'd like to resolve the kernel issues however I have yet to be contacted by anyone who knows enough about the existing kernel patches so you'll have to use whatever bastardized kernel you get.

[Rammjet]

I'd like to resolve the kernel issues however I have yet to be contacted by anyone who knows enough about the existing kernel patches so you'll have to use whatever bastardized kernel you get.

 

As I told Packagefixer, the effort on the kernel is being communicated on the hackint0sh board.

 

Join here: http://www.hackint0sh.org/forum/index.php

 

Thread here: http://www.hackint0sh.org/forum/showthread.php?t=18

[Guest BuildSmart]

As I told Packagefixer, the effort on the kernel is being communicated on the hackint0sh board.

 

Join here: http://www.hackint0sh.org/forum/index.php

 

Thread here: http://www.hackint0sh.org/forum/showthread.php?t=18

I've checked it out, I think I'm above there league, they're working by trial/error using a hex editor to adjust/modify the existing kernel where I'm building the kernel from source and they aren't 100% certain what the actual changes they are making are accomplishing other than bypassing certain routines or forcing certain results.

 

Since I've been going it alone and don't have any AMD cpu's I haven't concerned myself with this aspect.

 

While I think that the JaS kernel patches work for most people, my examination of them doesn't permit me to endorse their effenciency since they end up being modified for specific hardware and have lost most of their dynamic properties.

 

A good example is the SSE2/SSE3 patches.

 

There isn't a requirement for two patches, the kernel should be able to tes tand utilize the appropriate protocols automatically.

 

I was considering breaking down the patches and adjusting the source but this requires the following to be true:

 

a) an unmodified kernel.

a kernel with the SSE2 patch applied.

c) a kernel with the SSE3 patch applied.

 

I have not been able to apply the patches succefully as little documentation exists and I've had no success applying the patches to produce a working kernel so I've shelved it until JaS decides he wants to improve his kernel.

 

Of course the same holds true for anyone working on a kernel who wants to improve functionality.

[Guest BuildSmart]

Digging deeper into the VIAO LCD, it seems that the hidden file is the AppleOnboardDisplay driver binary which was modified to detect the SonyVIAO and I'll have the source for it later this evening.

[guiyoforward]

any news on this buildsmart?

[Neonkoala]

This is some great news - I'm really glad this has got some steam behind it since I've been away (hence no replies). Anyway - as to the 10.4.6 not working for Rainman - I have no problems getting 10.4.6 to work with QE CI on my 1300.. maybe different installer disc?

Perhaps you could dd that hidden partition via Linux by using a Knoppix disc or something? Might work... I've found that Ghost throws back quite a few partitions back at me so trying a different copying program might work - perhaps something on the UBCD might work?

[ultrasoul]

If this works this would be a real breakthough, considering how many GMA 900 (and now 950s) are out there.

 

Sigh. Hate the unseemly look of those paperclips jutting out of my VGA port.

[Guest BuildSmart]

I now have the source for the AppleOnboardDisplay and I can see how they are detecting the video on the GMA900 for the Mac laptops and given time I can decode and obtain the informatyion for the SonyVIAO and then other laptops so a solution is eminent as soon as I can get one.

[Neonkoala]

Excellent, once you work something out I'll be happy to work on it with my GMA900 too.

[Guest BuildSmart]

Well, it may be some time before I can start looking at it.

 

In an effort to fix problems on hardware I have like ethernet/sound on an Intel D915GUXLKR and sound on an Intel D945GPMLKR but I can't seem to find a clear cut solution that provides working sound (even partial working).

 

Once I know what drivers are modified I can fix it at a source level to provide full sound compatability and the ethernet issue I believe I will be able to resolve through driver identification.

 

The issue that seems very apparent is that anyone who claims to have a working solution provides instruction that don't provide a working solution and asking them to send me their fixed files has been refused which leads me to believe that they don't have a working solution or aren't interested in providing any assistance to the cause.

 

Basically this means I'll be reading until someone actually posts the information to provide a working solution that doesn't require any guess-work like telling me I need to enter my vendor ID and such which I don't seem to be able to find at this time and asking me to do anything in windows is like asking me to fly the space shuttle, I don't have windows or an interest in learning how to use it to run some obscure apps that aren't easily accessible and have {censored} instruction on their use.

 

The more work I have to do with little assistance to speed the process means the longer it will take me before I can get to anything else and of course to me the LCD issue isn't a pressing concern because I don't currently own and is more along the lines of something to work on at a later time.

 

I've already started building a dynamic kernel from source to get away from the stupid patched SSE2/SSE3 and AMD/Intel kernels (currently Intel only because I don't have any working AMD stuff at the moment) with great success and am about to move up to a 8.7.1 kernel which should be completed in the week or so as I wait to resolve some of my own hardware issues.

 

Of course if someone else with programming knowledge/experience wants to tackle it now without waiting on me they are more than welcome to do it.

[Guest BuildSmart]

Here are the crucial files, I do suggest you install XCode.

 

If you have an issue with some missing files like "/AppleInternal/xxxxx" I can put together a package of the most common when I have a little more time.

AppleDisplays_146.1.7.tar.gz.zip

[wesley]

Hmm, nothing new yet, huh?

[fireshark]

What does this AppleDisplays source do?

[anarchy99]

anything new?

[Neonkoala]

It seems that the answer is not in AppleOnboardDisplay.kext as this mainly controls brightness and contrast functionality. I need the source for the Intel915GA.kext and similar kexts and bundles. If anyone can help with that please PM me.

[chmielu]

I wont to ask about that Dynamic Kernel stuff. Are there any progress?? Do you have kernel higher than 8.4.1 working on the x86 machine??

Can you publish it somewhere??

[guiyoforward]

hey, on another thread people with inspiron e1505 / gma 950 are reporting that the screen works out of the box without dongles...

[Neonkoala]

@chmielu

 

Keep on thread this is not a thread about kernel development. See hackint0sh for that kind of info.

 

@guiyoforward

 

Do you have a link please? It sounds interesting however I still think that even if the GMA950 has a fix then it is unlikely to apply to the GMA900 so if anyone again has the 915 source then it would be greatly appreciated. I can share the AppleOnboardDisplay.kext source if people PM me although I am pretty sure the problem does not reside there.

[zedzed]

Does anyone have a picture which indicates which holes need to be shorted?

I can't figure out which are pins 1, 2, 6 and 12

 

Update - found it here

http://forum.insanelymac.com/index.php?s=&...st&p=150350

[Rammjet]

Does anyone have a picture which indicates which holes need to be shorted?

I can't figure out which are pins 1, 2, 6 and 12

http://forum.osx86project.org/index.php?ac...ost&id=1522

[Invisible Ink]

After reading this thread: http://forum.osx86project.org/index.php?showtopic=21804 it appears that it may be possible to solve the laptop VGA dongle problem with a simple patch/driver for laptop LCDs.

 

According to the thread the problem with OS X is that it doesn't find any LCD monitor when it boots up, this means it has to have an external monitor (or faked through dongle) to continue the boot process. It may therefore be possible to make it so OS X detects the internal monitor, eliminating the need to a dongle. Possibly a detection patch ahs already been written. We now need to bring any information we have together in order to try and overcome the need for a dongle.

 

Please post any relevant info you have and we will see what we can determine.

 

So far:

1. OS X detects an external monitor on boot up with (possibly?) generic drivers

2. OS X refuses to boot if it detects no monitor

3. OS X *may* not be initially detecting the internal LCD during boot up and stalls accordingly

 

Perhaps OS X does not detect the internal monitor because notebooks do not use standard DVI and VGA connectors, using internal circuits instead?

 

I experienced exactly the same. I suspect that working with old display (no plug-n-play) could give the same results as laptop LCD.

On Linux system I had the same problem, but there is a trick used for Linux systems. A simple kernel (you can find on regular repository) called "915resolution" hacks the 915 VGA video modes adding the desired resolution in a not used video mode (example hex 5b). This kernel module is loaded before the regular VGA driver, so you can set the desired resolution even with that driver. This is the principle I understand for 915resolution hack in Linux. Maybe I'm wrong but hope can be useful. The code of 915 resolution is open. If anyone know how to compile for mac OS X 386 and how to force the system to load it at boot, maybe the problem is solved.

Another idea, but may be it's a stupid thing, is to bluff on kext names, and rename the one that manage the generic VGA with the name of the 915. May it works? But wich is the kext managing the generic VGA monitor? I can't find it ...

[Neonkoala]

Thank you very much for the information Invisible Ink. I will look into the fake kext idea however to have any chance of implementing the 915resolution trick I still need the source for the 915 kext from Mac OS X. SO please if anyone can help me with that thanks.

[spoval]

Guys, I have hp nx6310 laptop with gma950 and there is now way to control brightness of LCD - it's always maximum. How can I solve this problem in 10.4.6?

[benjp2k1]

has this been resolved?

[anarchy99]

bump