Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.
Try to plug the MBP 8,2 with Magsafe charger and from a power-off state (no white sleep led) hold down until the Magsafe led will change from orange to green, then release and re-try to power on :

Left Shift+Control+Option+Power button
yeah my freiend did the SMC-Reset quiet often but didn't get the result that I want to :(
 
hmm okay, but it is the first time that this happend like this!
also is there no other possibility to fix that except by doing something at the logic board?

actually the thing really never happend before. and then the MBP just wanted to reboot and then the display just went dark, like no soul into it anymore... . I know that this is pretty random now, but to be honest, it did the same with the white LED on my MBP 13" 7,1. But I could cancel it. Maybe it has something to do with the update procedure by myside or anything else. Its kinda sad that it came like this :( and maybe someone can help me out. :(

If your GPU really is broken, there is no other fix. To be able to disable the GPU from a software side, it has to work at least for the time of the procedure.
@Olivia88 pointed out that you might get it back for a while if you heat up the device carefully. It's somewhere here in this thread. Though unfortunately that didn't work for my MBP with a dead AMD GPU, it never came back.
[doublepost=1554208976][/doublepost]I don't know where you're located in Germany but there are some repair shops that offer a GPU replacement for 100,- to 150,- EUR. But make sure it's a professional service and not just someone who will put your logicboard in the oven.
And make sure you will get a your defective chip back and also some warranty for the repair for at least a year.
I did it and my second GPU lastet exactly 11 month..
 
  • Like
Reactions: jackluke
If your GPU really is broken, there is no other fix. To be able to disable the GPU from a software side, it has to work at least for the time of the procedure.
@Olivia88 pointed out that you might get it back for a while if you heat up the device carefully. It's somewhere here in this thread. Though unfortunately that didn't work for my MBP with a dead AMD GPU, it never came back.
Okay, I get your point and understand it. But it is pretty weird that it had some annomalies and then i fixed them with the software solution once, and just after the update and doing the legacy patch reinstalling it came like this. As I mentioned..it tried to do the same and had the same issue with my MBP 7,1 13" but I could cancel it. So I kinda think that it is broken for sure...but it caused something from the accel or the update. Otherwise I cannot explain why it should be like this. It would be a really really tiny chance that it came just by its own... .
So thats what I try to figure out. Sadly I cannot effort at the moment a repair with the MuxHat and as far as I saw it its kinda difficult. So I guess I need to use just the one MPB 2010 13" 7,1 for now on. :/
Makes me sad actually... . I hope that I can fix it once... .

EDIT :
By getting the MuxHat it should be done forever right?
 
Last edited:
@jklas , @flygbuss and others who own a "bricked" MacBookPro 8,2 15" 2011 , if you get black LCD screen and solid white led sleep after the power-on only two issues could be:

- EFI firmware corruption (I exclude it but also fixable through EEPROM tools)
- AMD/ATI dGPU broken due to factory defects together with intense use and heating within years.

I don't have de-soldering skills but if you have the chance to simply remove the resistor R8911 that's even located on the main front side of the logic board, probably neither needed to disassembly everything, then yours MBPs will boot up again, the only features you'd lose are brightness control and sleep function.

https://www.jeffgeerling.com/sites/...images/resistor-zoomed-in-mbp-2011-radeon.jpg
https://www.jeffgeerling.com/sites/...resistor-gpu-radeon-macbook-pro-highlight.jpg
 
@jklas , @flygbuss and others who own a "bricked" MacBookPro 8,2 15" 2011 , if you get black LCD screen and solid white led sleep after the power-on only two issues could be:

- EFI firmware corruption (I exclude it but also fixable through EEPROM tools)
- AMD/ATI dGPU broken due to factory defects together with intense use and heating within years.

I don't have de-soldering skills but if you have the chance to simply remove the resistor R8911 that's even located on the main front side of the logic board, probably neither needed to disassembly everything, then yours MBPs will boot up again, the only features you'd lose are brightness control and sleep function.

https://www.jeffgeerling.com/sites/...images/resistor-zoomed-in-mbp-2011-radeon.jpg
https://www.jeffgeerling.com/sites/...resistor-gpu-radeon-macbook-pro-highlight.jpg
Well it seems to me that as I mentioned already..it tried and did the same with my MBP 7,1 13". But I could stop it by shutting it down... .
Anyway my point is how could I do this with EFI and give it a try?

Also what I would like to know is...maybe I did something wrong with my procedure... .
After putting those Files into the System/etc folders and so on. I did not run the install.sh was that wrong? do I need to apply this now?
I put all my stuff back into the MBP that it was before.
So actually I could give it a try somehow.
What I also would like to point out and mention is that the display went once online which is kinda annoying and for me it does not make any sens. I kinda feel that it is a software/update issue to be very honest. Since that behaviour never happend before and just happend after the restart immediatly which is pretty weird!
 
Well it seems to me that as I mentioned already..it tried and did the same with my MBP 7,1 13". But I could stop it by shutting it down... .
Anyway my point is how could I do this with EFI and give it a try?

Also what I would like to know is...maybe I did something wrong with my procedure... .
After putting those Files into the System/etc folders and so on. I did not run the install.sh was that wrong? do I need to apply this now?
I put all my stuff back into the MBP that it was before.
So actually I could give it a try somehow.
What I also would like to point out and mention is that the display went once online which is kinda annoying and for me it does not make any sens. I kinda feel that it is a software/update issue to be very honest. Since that behaviour never happend before and just happend after the restart immediatly which is pretty weird!

EFI firmware corruption implies also to desolder the EFI CHIP and programming with external tool then soldering it back, so I'd say removing the R8911 would be an "easier" fix.

Similar thing happened to me with MacBookAir 2009 , but in soldering back with hot air a subatomic sized resistor jumped out and so no more power-on, even if I'd say operating with MacBookAir logic boards is harder than MacBookPro where internal components are "bigger".

I fixed sometimes the MacBookAir failing GPU or onboard RAM exposing only them covering the rest with tinfoil and then putting it into oven for 5-6 minutes at 200 Celsius, and it worked again. But I discourage do this to those who take care of their machines.
 
EFI firmware corruption implies also to desolder the EFI CHIP and programming with external tool then soldering it back, so I'd say removing the R8911 would be an "easier" fix.

Similar thing happened to me with MacBookAir 2009 , but in soldering back with hot air a subatomic sized resistor jumped out and so no more power-on, even if I'd say operating with MacBookAir logic boards is harder than MacBookPro where internal components are "bigger".

I fixed sometimes the MacBookAir failing GPU or onboard RAM exposing only them covering the rest with tinfoil and then putting it into oven for 5-6 minutes at 200 Celsius, and it worked again. But I discourage do this to those who take care of their machines.
Okay, as I saw on the youtube channel from DosDude1 there is something with Firmware flashing. Is that possible to do that at my MBP? Since I cannot figure out the stuff with my issue. I dont know which "fix" would be better to do.

I am kinda very confused since I still think that it doesnt make any sens to me that after the upgrade nothing happend and then updating and doing the steps that dosdude1 and you said...then it just happend after the reboot. this is so strange.

Also if I can flash that thing and it will never happend again..it would be nice though... . And well I dont know whats better to do... . With no response from the display at all...since his issue was the annomaly only
 
Okay, as I saw on the youtube channel from DosDude1 there is something with Firmware flashing. Is that possible to do that at my MBP? Since I cannot figure out the stuff with my issue. I dont know which "fix" would be better to do.

I am kinda very confused since I still think that it doesnt make any sens to me that after the upgrade nothing happend and then updating and doing the steps that dosdude1 and you said...then it just happend after the reboot. this is so strange.

Also if I can flash that thing and it will never happend again..it would be nice though... . And well I dont know whats better to do... . With no response from the display at all...since his issue was the annomaly only

No, firmware flashing on that MBP could be done only through external EEPROM tool, and I guess it's not that the issue, but it's a broken AMD/ATI dGPU that needs to be deceived by removing a resistor.
 
Like I said, I got a bricked AMD GPU in my MBP early 2011 15" as well. It happened in 2017 after 6 years with a lot of audio work (connected Thunderbolt Display most of the time).
I replaced the defetcive chip once but the replacement lasted not even a year.
After that I decided to swap logic boards.
 
Okay, as I saw on the youtube channel from DosDude1 there is something with Firmware flashing. Is that possible to do that at my MBP? Since I cannot figure out the stuff with my issue. I dont know which "fix" would be better to do.

I am kinda very confused since I still think that it doesnt make any sens to me that after the upgrade nothing happend and then updating and doing the steps that dosdude1 and you said...then it just happend after the reboot. this is so strange.

Also if I can flash that thing and it will never happend again..it would be nice though... . And well I dont know whats better to do... . With no response from the display at all...since his issue was the annomaly only
The reason your display is blank is most likely because R8911 was removed, while NOT performing a full permanent hardware bypass of the gMux IC. This is an absolutely terrible solution, as when that NVRAM variable gets reset, this is exactly what happens -- You get no video on the display at all. The BEST possible thing you can do with one of these machines is flash my custom gMux IC firmware, which permanently disables the defective AMD GPU, and of course does NOT rely on any easily-resettable NVRAM variables. I have not yet released the firmware, as I am still working an a flashing program and custom cable/solderless connector to make it easier for most people to actually perform the flash. Currently, though, I do offer mail-in flashing services, if that's something you'd be interested in doing. Flashing this firmware would fix your blank display issue more than likely.
[doublepost=1554217299][/doublepost]
@jklas , @flygbuss and others who own a "bricked" MacBookPro 8,2 15" 2011 , if you get black LCD screen and solid white led sleep after the power-on only two issues could be:

- EFI firmware corruption (I exclude it but also fixable through EEPROM tools)
- AMD/ATI dGPU broken due to factory defects together with intense use and heating within years.

I don't have de-soldering skills but if you have the chance to simply remove the resistor R8911 that's even located on the main front side of the logic board, probably neither needed to disassembly everything, then yours MBPs will boot up again, the only features you'd lose are brightness control and sleep function.

https://www.jeffgeerling.com/sites/...images/resistor-zoomed-in-mbp-2011-radeon.jpg
https://www.jeffgeerling.com/sites/...resistor-gpu-radeon-macbook-pro-highlight.jpg
No, you should NEVER, EVER remove R8911 without doing a FULL, permanent hardware bypass of the dGPU, either by bypassing the gMux IC entirely, or flashing my custom firmware onto it. Removing R8911 WITHOUT doing a full bypass, or setting that "gpu-power-prefs" NVRAM variable will simply result in a display that does not show video.
 
The reason your display is blank is most likely because R8911 was removed, while NOT performing a full permanent hardware bypass of the gMux IC. This is an absolutely terrible solution, as when that NVRAM variable gets reset, this is exactly what happens -- You get no video on the display at all. The BEST possible thing you can do with one of these machines is flash my custom gMux IC firmware, which permanently disables the defective AMD GPU, and of course does NOT rely on any easily-resettable NVRAM variables. I have not yet released the firmware, as I am still working an a flashing program and custom cable/solderless connector to make it easier for most people to actually perform the flash. Currently, though, I do offer mail-in flashing services, if that's something you'd be interested in doing. Flashing this firmware would fix your blank display issue more than likely.
[doublepost=1554217299][/doublepost]
No, you should NEVER, EVER remove R8911 without doing a FULL, permanent hardware bypass of the dGPU, either by bypassing the gMux IC entirely, or flashing my custom firmware onto it. Removing R8911 WITHOUT doing a full bypass, or setting that "gpu-power-prefs" NVRAM variable will simply result in a display that does not show video.

Sorry, I just advised him removing R8911 because I read many success posts about it, anyway @jklas hasn't removed nothing yet, so the R8911 resistor is still there on his logic board, but I understand your explanations, only I am missing something and I do have a question: does the MBP 8,2 15" even with a black screen and solid sleep led work and boot in background or doesn't pass the EFI firmware checks ?

edit:
However, I advised also @jklas in my many posts to follow your guide about MBP 15/17" fixing.
 
Last edited:
Haha. That attitude would make all historians and archeologists obsolete. ;-)
Oh, wait - we are the h&a´s!
[doublepost=1554186063][/doublepost]
My experience in running 10.14.4 on various machines* without tinkering with the transparencies is on the contrary: Very useful, very stable and I had no crashes at all (just on a dual GPU machine I have to use the GT instead of M).
And there are many of us who are happy to have a minimum set of modern apps for daily tasks:
Mail, iTunes, Browsers, iWork or M$Office, some photo viewers...
So my advice would be stop modding the frameworks for optics alone (transparencies) and use the machines a while in that state to see if they really crash as hell on 10.14.4. Mine don´t.

*currently on MacBook 5,2 - MacBook Air 4,1 11" - MacBook Pro 5,2 - MacBook 4,1 - MacBook 7,x and 6,x 13"- MacBook Air 2,1 - MacBook Pro 4,1 17"
Well, wait a bit for 10.14.5, or for 10.15 at the latest to be released and you might change completely your way of thinking. Besides, if you are happy to use a computer just for basic modern apps and browsing, than I suggest you get yourself an Android box for some $60 and it will fulfill all your basic needs plus a lot more. PC/Macs on the other hand represent for many actually a tool of work, creativity and what not. The "basic" stuff won't cut it. And don't come back to me with a cliché of the type: "buy a new mac", cause after its due diligence in research, no intelligent, well-informed person nowadays would fall for the pseudo marketing push "a la" Steve Jobs and buy again all those fancy, over-hipped, expensive, defective and unsupported products, spilled out from Apple year after year.
 
I am very happy to admit that I was wrong about 10.14.4. It seems that I wreaked havoc on my system by abusing Onyx.
I did a clean install of the 10.14.3 and then downloaded 10.14.4 update, at this point I created APFS local snapshot which allowed me to rollback to this point three times without having to re-download the 10.14.4 update(even when I know what to do I seem to miss steps all the time). I absolutely like APFS local snapshots, it takes 3 seconds to get back to where you missed your step. In short a success, I can live without Flurry screen saver(did not even knew it was not working until somebody mentioned it in this thread). Memory management is much better now.

Screen Shot 2019-04-02 at 6.05.21 pm.png
 
I am very happy to admit that I was wrong about 10.14.4. It seems that I wreaked havoc on my system by abusing Onyx.
I did a clean install of the 10.14.3 and then downloaded 10.14.4 update, at this point I created APFS local snapshot which allowed me to rollback to this point three times without having to re-download the 10.14.4 update(even when I know what to do I seem to miss steps all the time). I absolutely like APFS local snapshots, it takes 3 seconds to get back to where you missed your step. In short a success, I can live without Flurry screen saver(did not even knew it was not working until somebody mentioned it in this thread). Memory management is much better now.

View attachment 830022

However, Flurry screensaver does work fine on 10.14.5 beta1 with only Video Framebuffer.

And I add another little adjustment, even using the GPUWrangler.framework from 10.14.4 will allow booting 10.14.5 beta 1 , from my tests "GPUWrangler" must be replaced to reach the loginUI and use Mojave in GUI mode.

Another test: even replacing GPUWrangler.framework from 10.14.3 into 10.14.4 will still get OpenGL Acceleration and will still boot from Dual GPUs Mac.

So in my opinion replacing GPUWrangler.framework doesn't affect nothing bad.

edit:
Funny test: renaming or removing the GLDriver from 10.14.4 will make Flurry screensaver to work again.
 
Last edited:
  • Like
Reactions: avz
Sadly, I have to report, that 10.14.4 is not a daily driver at all. You can goof around with it and beta-test it for the sake of knowledge, but fore serious work you need 10.14.3. So, I was forced to return to 10.14.3, very sad indeed. In my testings 10.14.4 will crash unexpectedly at least once a day. I do not use any of the iOS apps and have no Siri. If one would change appearance elements like "Reduced Transparency" the OS would crash, than it will settle until the next change and crash. Then there are third party programs, which have implemented portions of the new API, which are broken in 10.14.4, but work perfectly fine in 10.14.3. An example for this is Enpass. The support for Mojave's White/Dark themes was implemented as of version 6.0 ( the latest). So that version works outstanding in 10.14.3, but under 10.14.4 it will just open and display a black/white square window depending on the selected theme. So certain programs are rendered by this totally useless. Although at a first glance poucame's 1.4.1 Hybrid mode would seem to work, one can notice some elements having changed appearance and it does not look as pleasing to the eyes and practical for work as with 10.14.3. Given the above, once released, I think none of us will be using 10.15 as a daily driver. By then changes will be so drastic, that no tinkering, nor replacement with older components will manage to rescue the day. We have reached the final station for our old macs. Maybe we should concentrate on making 10.14.3 being able to appear to programs as if of a higher version. In the long run it will be necessary to avoid OS-version restrictions imposed by third party programs. I think in 10.14.3 the support for all features pertinent to Mojave is excellent, this will assure compatibility with future third party apps.
To date my experience with .4 after applying the latest patcher post-install (1.3) has been very positive with or without my hybrid patches. And believe me, I run that mbp 5,3 through its paces with a large variety of applications.
In general, I hate applying retro-patches but so far I've identified the change in the .4 OpenGL and feel comfortable that running a .3 OpenGL+GpuSupport combo on a .4 system won't adversely affect it (at least up until now).

I can't comment on the experience you have running retro hybrid patches as it is not my recommended configuration ( I have advised against it strongly often). I haven't released a hybrid for .4 yet but it is testing well. Given the stability of .3 I may do so too soon. But if you like Dark mode, skip the hybrid patches all together.

The road to 10.15 will be challenging for our unsupported machines...that's for sure.
[doublepost=1554224967][/doublepost]
However, Flurry screensaver does work fine on 10.14.5 beta1 with only Video Framebuffer.

And I add another little adjustment, even using the GPUWrangler.framework from 10.14.4 will allow booting 10.14.5 beta 1 , from my tests "GPUWrangler" must be replaced to reach the loginUI and use Mojave in GUI mode.

Another test: even replacing GPUWrangler.framework from 10.14.3 into 10.14.4 will still get OpenGL Acceleration and will still boot from Dual GPUs Mac.

So in my opinion replacing GPUWrangler.framework doesn't affect nothing bad.

edit:
Funny test: renaming or removing the GLDriver from 10.14.4 will make Flurry screensaver to work again.
So a bootable .5 beta with acceleration is achievable with dude's accelerator patches and a .4 or .3 GpuWrangler?
 
To date my experience with .4 after applying the latest patcher post-install (1.3) has been very positive with or without my hybrid patches. And believe me, I run that mbp 5,3 through its paces with a large variety of applications.
In general, I hate applying retro-patches but so far I've identified the change in the .4 OpenGL and feel comfortable that running a .3 OpenGL+GpuSupport combo on a .4 system won't adversely affect it (at least up until now).

I can't comment on the experience you have running retro hybrid patches as it is not my recommended configuration ( I have advised against it strongly often). I haven't released a hybrid for .4 yet but it is testing well. Given the stability of .3 I may do so too soon. But if you like Dark mode, skip the hybrid patches all together.

The road to 10.15 will be challenging for our unsupported machines...that's for sure.
[doublepost=1554224967][/doublepost]
So a bootable .5 beta with acceleration is achievable with dude's accelerator patches and a .4 or .3 GpuWrangler?

No, .5 beta only bootable with basic Video Framebuffer from any Penryn cpu Mac with a single iGPU adding to the dosdude1's 10.14.4accelpatch a replacement of GPUWrangler.

Not bootable from Dual GPUs Mac, since in my humble opinion apple has done significantly changes to AppleGraphicsControl.kext , I bet mainly to fight this thread, but how they noticed of this thread?

I guess in a simple and effective way: just tracing the source of all the Mojave beta downloads from target machines, simply they noticed thousands of old machines using Mojave (even from iCloud syncs) and taken an action.
 
  • Like
Reactions: ASentientBot
No, .5 beta only bootable with basic Video Framebuffer from any Penryn cpu Mac with a single iGPU adding to the dosdude1's 10.14.4accelpatch a replacement of GPUWrangler.

Not bootable from Dual GPUs Mac, since in my humble opinion apple has done significantly changes to AppleGraphicsControl.kext , I bet mainly to fight this thread, but how they noticed of this thread?

I guess in a simple and effective way: just tracing the source of all the Mojave beta downloads from target machines, simply they noticed thousands of old machines using Mojave (even from iCloud syncs) and taken an action.
So I have 2 gpus on my mbp 5,3, I wonder what would happen if I disabled one? :)
 
  • Like
Reactions: jackluke
So I have 2 gpus on my mbp 5,3, I wonder what would happen if I disabled one? :)

I think it should work, since system will assume it as a single GPU, but from .5 you will achieve Acceleration only if it's Metal, if not, only Framebuffer. Just try and report your results.

edit:
An advise if from .5 verbose mode you get some:
busy timeout [0] (60s) yourMacmodel
busy timeout [1] (60s) yourMacmodel

just wait a bit more they will pass for sure, it's a normal delay I'd say after the update, they will disappear at the next reboot.
 
Last edited:
  • Like
Reactions: pkouame
... but how they noticed of this thread?

I guess in a simple and effective way: just tracing the source of all the Mojave beta downloads from target machines, simply they noticed thousands of old machines using Mojave (even from iCloud syncs) and taken an action.
That would be some really evil empire sh*&t!o_O While they are definitely aware of what's going on (this project and dude1 got some press a while ago) I tend to think they're too busy fixing their own problems than actively scheming to trip us up.
One thing is for sure, they are still changing old "unsupported" frameworks (like OpenGL) . Obviously their internal test suites don't account for our unsupported configurations any more. Hence, the bad side effects we have to suffer with...
 
That would be some really evil empire sh*&t!o_O While they are definitely aware of what's going on (this project and dude1 got some press a while ago) I tend to think they're too busy fixing their own problems than actively scheming to trip us up.
One thing is for sure, they are still changing old "unsupported" frameworks (like OpenGL) . Obviously their internal test suites don't account for our unsupported configurations any more. Hence, the bad side effects we have to suffer with...

Agree, but it's really weird from .4 final build to .5 beta 1 build , no more OpenGL Acceleration for old machines, I've also compared the main unix binaries compiling date by them:

.4 final build binaries: last modified March 21 (to be more clear latest touch to OpenGL Resources 23 March)
.5 beta 1 build binaries: last modified March 26 (latest touch to OpenGL Resources 29 March)

in five days they changed too soon OpenGL APIs, that's suspect.
 
Last edited:
To date my experience with .4 after applying the latest patcher post-install (1.3) has been very positive with or without my hybrid patches. And believe me, I run that mbp 5,3 through its paces with a large variety of applications.
In general, I hate applying retro-patches but so far I've identified the change in the .4 OpenGL and feel comfortable that running a .3 OpenGL+GpuSupport combo on a .4 system won't adversely affect it (at least up until now).

I can't comment on the experience you have running retro hybrid patches as it is not my recommended configuration ( I have advised against it strongly often). I haven't released a hybrid for .4 yet but it is testing well. Given the stability of .3 I may do so too soon. But if you like Dark mode, skip the hybrid patches all together.

The road to 10.15 will be challenging for our unsupported machines...that's for sure.
[doublepost=1554224967][/doublepost]
So a bootable .5 beta with acceleration is achievable with dude's accelerator patches and a .4 or .3 GpuWrangler?
The thing is that I need to use the programs I frequent on a daily basis. All of them work perfectly fine in 10.14.3 with the "retro-hybrid patch" you provided. On the other hand programs like Enpass, Telegram and a growing number of others are totally broken under 10.14.4 even in its plain vanilla state, without any Dark/Light mode modifications. So my point is we have hit the bone, where one has to sacrifice functionality of a growing number of third party, plus some native apps in exchange for the eye-comfort of a point release. And that is a no, no in my book. From now on the situation is only going to get worst with every single new release, that's for sure. We need to modify and perfect all mimicry in 10.14.3, because that is the last version to be ready to cope with anything to come and any modifications there are simpler, than an open fight with Apple's intentions. At the end of every night, one has to wake up and go on.
[doublepost=1554228759][/doublepost]
That would be some really evil empire sh*&t!o_O While they are definitely aware of what's going on (this project and dude1 got some press a while ago) I tend to think they're too busy fixing their own problems than actively scheming to trip us up.
One thing is for sure, they are still changing old "unsupported" frameworks (like OpenGL) . Obviously their internal test suites don't account for our unsupported configurations any more. Hence, the bad side effects we have to suffer with...
You think so, because you have not seen the greed in the eyes of Apple's shareholders. The juvenile, "creative" and the "we do it for you" spirit in Apple's events is just a lousy mask for the uninformed, simpleminded. You seem to forget that at the end of the day Apple is a giant corporation and it won't take long before people realize, that those are the worst of all kind, since it represents the final step of transformation of the capital. Every single "unsupported" mac running Mojave and up is a nail in the eyes of the corporation. What you see, has absolutely nothing to do with the actual reality. Its only a matter of time for it all to become impossible to hide.
 
The reason your display is blank is most likely because R8911 was removed, while NOT performing a full permanent hardware bypass of the gMux IC. This is an absolutely terrible solution, as when that NVRAM variable gets reset, this is exactly what happens -- You get no video on the display at all. The BEST possible thing you can do with one of these machines is flash my custom gMux IC firmware, which permanently disables the defective AMD GPU, and of course does NOT rely on any easily-resettable NVRAM variables. I have not yet released the firmware, as I am still working an a flashing program and custom cable/solderless connector to make it easier for most people to actually perform the flash. Currently, though, I do offer mail-in flashing services, if that's something you'd be interested in doing. Flashing this firmware would fix your blank display issue more than likely.
[doublepost=1554217299][/doublepost]
No, you should NEVER, EVER remove R8911 without doing a FULL, permanent hardware bypass of the dGPU, either by bypassing the gMux IC entirely, or flashing my custom firmware onto it. Removing R8911 WITHOUT doing a full bypass, or setting that "gpu-power-prefs" NVRAM variable will simply result in a display that does not show video.
I know this isn't the ideal solution, but how about a super-simple patcher to create a USB that boots to single-user mode, sets the NVRAM variable, and then reboots? It'd be as easy as restoring a BaseSystem to the drive, setting boot args as -s, and making root's .profile as "nvram ... && reboot".

This would allow for an easy fix when the variable got unset, and not require potentially tricky/expensive hardware fixes. What do you think? I don't have an affected Mac but I could probably implement this if you'd test it.

Alternately, a kernel extension to just set that nvram value at every startup/shutdown could work as well.

I know that software workarounds for hardware issues aren't always the answer, but a lot of users are squeamish about soldering on their MacBooks' logic boards.

Edit: @dosdude1 already did this, and is working on a better firmware-level solution. See his post on the next page.
 
Last edited:
Agree, but it's really weird from .4 final build to .5 beta 1 build , no more OpenGL Acceleration for old machines, I've also compared the main unix binaries compiling date by them:

.4 final build binaries: last modified March 21 (to be more clear latest touch to OpenGL Resources 23 March)
.5 beta 1 build binaries: last modified March 26 (latest touch to OpenGL Resources 29 March)

in five days they changed too soon OpenGL APIs, that's suspect.
Yup, they are still changing "unsupported" frameworks like OpenGL etc.... If you're big enough like Adobe, Microsoft or even SimCity (direct references in the code), Apple will push out updates for you, and break us without an after thought). We knew this would happen...

I actually disassembled the .4 OpenGL binary and compared with .2 (oldes one I have) : I noticed new support for "ExtendedTextures". It is even controllable via a plist (default) setting in com.apple.opengl.plist . There's a whole bunch of new code supporting it - not yet sure how that has affected us.

I suspect more changes are coming when they start merging IOS and macOS this summer.

I would be careful with the modification dates on binaries. Sometimes functionality and logic won't change while the tools they use to build (compilers, linkers etc...) generates new code. I haven't check out the Resource forks...
[doublepost=1554229800][/doublepost]
I think it should work, since system will assume it as a single GPU, but from .5 you will achieve Acceleration only if it's Metal, if not, only Framebuffer. Just try and report your results.

edit:
An advise if from .5 verbose mode you get some:
busy timeout [0] (60s) yourMacmodel
busy timeout [1] (60s) yourMacmodel

just wait a bit more they will pass for sure, it's a normal delay I'd say after the update, they will disappear at the next reboot.
Does anyone know of a software (nvram?) switch to disable a gpu? I know dude has a hardware fix in another thread, but modding my motherboard is out of the question.
 
  • Like
Reactions: jackluke
Is it safe to update my 17" Early 2008 MacBook Pro (MacBookPro4,1) 6GB RAM 512GB SSD (APFS) from Mojave Patcher installed 10.14.3 to 10.14.4 yet? I've already installed all the Mojave Patcher updates when prompted... and macOS updates from 10.14.2 to 10.14.3 were painless when Mojave Patcher post install patches were run when prompted.

Software Update shows" macOS 10.14.4 Update" is available... but unfortunately I don't have spare time or macOS skills to do a bunch of post upgrade hacking / KEXTing.

Thanks!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.