Boot-Loops on late2008 15" 2,8GHz MacBookPro5,1

[bobesch]

Hi folks, I need some help!
After running MojavePatch on early2008 MBPs flawlessly, yesterday I tried to install MojavePatch onto a late2008 2,8GHz 15" MacBookPro5,1 without any success so far.
Installation of Mojave runs fine, Patcher (with recommended settings for that very model) as well, but when the patched Mojave is to boot for the first time, then the MBP always runs into a Boot-Loop as soon as progress-bar reaches 50% on the light-gray boot-screen with darker grey Apple-logo.
The MBP came with 8GB of RAM, a 250GB Samsung 850 Pro SSD and ElCapitan.
Is there anything I have to consider?

I'm gonna try HighSierra and/or CatalinaPatcher just now and let you know about progress ... But I'd rather stay with Mojave because of 32bit support and dark-mode.

Edit: changed thread-title, because Mojave-Patch or any other patch isn't to blame.
Previous title: "MojavePatch won't install on late2008 15" MacBookPro5,1"

 

[eyoungren]

Did you run the post installation tools?

You probably want to add the APFS patch with that as well.

 

[Amethyst1]

Can you boot in verbose mode? If so, what does it show on the screen just before it reboots?

 

[bobesch]

Thanks both of you!
I tried to install the patches on both HFS+ and APFS partitions (even ran DiskUtility on Leopard/PPC between the changing from HFS+ to APFS and back, to make sure the drive is set up properly ...)
And, yes, ran the post installation tool with each new trial (even twice). And power supply was plugged-in too
Didn't try verbose mode (don't even know anything about how to use that and for what purpose).
There's just that light grey screen with the darker gray Apple logo and the progress bar beneath getting to about 50%, then the screen goes black and the book reboots.

Well, problem seem's to be related to just that very MacBookPro!
After swapping the SSD (that had been set up with HighSierra-Patch on the faulty MBP) into a twin-model (the base-model with just 2,4GHz, 8GB RAM), HighSierra booted fine and also downloaded the missing patches within macOS without any hiccups.
After running HS successfully I proceeded with MojavePatch installation and that did work either well without any problems.
Swapping the SSD back into the faulty book results in a boot loop again and again and again ... (both with patched HS and Mojave).

So the problem apparently is related to that very MacBookPro5,1 (late2008 2.8GHz 8GB RAM)
But how can I find out, what causes the problem? RAM? Any sort of Firmware, that needs to be updated? Airport card?
Actually I have no clue ...

I'd rather use that faster machine in the office to have a copy of important DEVONthink databases at hand, after all other Macs had been migrated from RDP-Clients(OSX) to full Win10Pro clients last spring and so I lost my handy DEVONthink-network at that moment.

Edit: MojavePatch seems to have problems to wake up from hibernation-mode. A second install didn't sort out that problem. So back to HighSierra. ?

 

[Amethyst1]

Didn't try verbose mode (don't even know anything about how to use that and for what purpose).

Hold [Command]-[V] after turning on the machine. Lots of text will start scrolling down the screen as macOS boots. Try to see what the last few lines are before it reboots, perhaps by taking a video of the boot process. This may (hopefully) point to the possible cause of the boot loop.

 

[bobesch]

Hold [Command]-[V] after turning on the machine. Lots of text will start scrolling down the screen as macOS boots. Try to see what the last few lines are before it reboots, perhaps by taking a video of the boot process. This may (hopefully) point to the possible cause of the boot loop.

Thanks!
I'm gonna check that and post the report, but today I'm really fed up from those fruitless attempts to install patched macOS.
Today's goal is to finish installation of patched HighSierra on the "twin"-machine together with all favorite Apps and then further trials will be performed as soon as I have a hard drive-clone at hands.

 

[bobesch]

Well, I have to conclude, that 2,8GHz late2008 15" MBP5,1 has some kind of serious damage.
Even booting from the original ElCapitan-Drive results in reset-loops now.
Internal optical drive isn't working anymore. Trying to boot SL-Install-Disk from external USB-SuperDrive fails (endless rotating gear-wheel). Removed the internal optical drive without any changes.
Swapped RAM modules with the working unit without any effect.
The book gets pretty warm at the bottom in the region of GPU/CPU. Maybe needs repasting?
Before I tried to upgrade from ElCapitan to MojavePatch everything seemed to work fine.
Anything else I could try or check ...

 

[Amethyst1]

Even booting from the original ElCapitan-Drive results in reset-loops now.

Can you boot in Safe Mode? Hold [Shift] right after turning on the machine and keep it held down. This skips loading e.g. the graphics drivers and most other things, so if it boots then, something might be iffy with the GPU or some other component.

 

[bobesch]

Can you boot in Safe Mode? Hold [Shift] right after turning on the machine and keep it held down. This skips loading e.g. the graphics drivers and most other things, so if it boots then, something might be iffy with the GPU or some other component.

Thanks! Yep, booting in Save Mode did work actually.
Here some screenshot about the bug-report etc.

 

[Amethyst1]

Thanks! Yep, booting in Save Mode did work actually.
Here some screenshot about the bug-report etc.

The kernel panic mentions the kext for the NVIDIA GPUs (NVDAResmanTesla). Try booting in Safe Mode again and forcing the built-in 9400M to be used all the time rather than the discrete 9600M GT: System Preferences > Energy Saver > Better battery life.

 

[K two]

This sounds very much like a frayed SATA cable rubbing against the housing. If the cable is not completely hosed a bit of electrial tape where the cable hits will stabilize things. A very common problem.

 

[Amethyst1]

This sounds very much like a frayed SATA cable rubbing against the housing.

The fact the machine boots fine in Safe Mode and the panic log suggest the issue is elsewhere though.

 

[K two]

The fact the machine boots fine in Safe Mode and the panic log suggest the issue is elsewhere though.

Time will tell. A bad SATA cable can appear to be a lot of things, the one constant is symptoms change but most of all whatever appears wrong is interrmittant.

 

[Amethyst1]

Time will tell. A bad SATA cable can appear to be a lot of things, the one constant is symptoms change but most of all whatever appears wrong is interrmittant.

Yep. And a SATA cable is much easier to replace than a GPU

 

[bobesch]

Thanks again!
I replaced the SATA cable with one coming from a defective device, that I bought about 2y ago for 30 bucks meant for screen-replacement of my currently working 2,4GHz unit.
First attempt to boot worked flawless! Second too.
Then I reassembled all parts and replaced the test-drive with the SSD, that is intended for future use.
DANG! Boot-loops again and again - can be only be overcome by booting into Save Mode, but wait, there was one fragile success to boot regularly: I was able to make a screenshot of the crash-report, copy that to a USB-stick and just after ejecting and removing the USB-stick in the right order, the system was frozen ...
Attached are the latest crash-reports.
Is it possible, that the unit, I took the SATA-cable off, also carried a broken cable?
Since I don't dare to disassemble my working 2,4GHz unit I already ordered a replacement SATA-cable and keep my fingers crossed, that this will sort out the problem (well, otherwise I can start trading with SATA cables ... )

 

[Amethyst1]

Attached are the latest crash-reports.

All list the NVDAResmanTesla kext.

Did you try forcing the use of the 9400M?

As a test, you can boot in Safe Mode, go to the /System/Library/Extensions folder (Finder: Go to > Go to folder) and remove all files which have names starting with GeForce or NVDA. Or move them to a different folder if you want to keep a backup.

If the machine starts up reliably (in "normal" mode) after these kexts have been removed, this corroborates my theory that one of the GPUs is to blame.

 

[bobesch]

Update: the SATA-cable did arrive. After replacement the MBP did behave quirky with some booting/rebooting but finally booted several times into macOS. System froze one time. Several attempts to reboot went fine. I swapped in the intended SSD and after some time with tinkering boot-loops were back.
@Amethyst1 guess You were right about the faulty GPU, but I'm a bit fed up and due to lack of time I put the MBP aside to retry on another occasion.
Does disabling the discrete 9600M GT survives rebooting and how is the impact on performance, when only using the discrete 9400M?
I've read about a faulty capacitor on 2010-2011 15" MBP that might also cause problems with the GPU ...

It's a real pity, because this very MBP is one of the fastest of it's kind with large battery-door, easy access to the hard drive and PCExpress-Card-Slot for USB3.0 ?

 

[Amethyst1]

guess You were right about the faulty GPU, but I'm a bit fed up and due to lack of time I put the MBP aside to retry on another occasion.

I feel your pain. The test I suggested (deleting the NVIDIA kexts) will disable graphics acceleration so macOS' GUI will behave like in a VM, i.e. be slow and glitchy. So this isn't a solution by any means.

Does disabling the discrete 9600M GT survives rebooting [...]

Yes, the machine will keep using the built-in 9400M for everything until you change the setting again. However, I don't know when in the boot process the 9600M GT is disabled. If it's only disabled after macOS has loaded the graphics drivers, this may not cure the boot loops. But it's worth a try, assuming the discrete 9600M GT is causing the problems and not the built-in 9400M. (But I'm confident the 9600M GT is to blame - I've not heard of a failing 9400M yet.)

how is the impact on performance, when only using the discrete 9400M?

The 9400M is slower than the 9600M GT of course, but for non-gaming use, the difference doesn't matter.

One thing I noted when I briefly played with a late 2008 15" MBP is that, unlike e.g. on the 2011 MBPs with the Intel iGPU/AMD dGPU combo, external monitors appear to be also driven by the 9400M if it's set to be used all the time. On the 2011, external monitors are always driven by the AMD dGPU so once that is disabled, you cannot use external monitors.

 

[bobesch]

@Amethyst1 So I removed the GeForce* and NVDA* kext files.
And voilà / chapeaux!: MacBook is booting now without booting-loops.
Graphic-performance is decent for Office-Apps, but bloated websites with fast/recurrent re-rendering are a challenge.
At least the Macbook can be used as a thin-client for RDP.
What a shame - that machine was top of crowd at it's time and is still capable of getting 95% of my current tasks done ....
Would Linux encounter the same trouble on booting etc.?
How could this problem be sorted out? GPU-replacement? Bad capacitor?

I made a post on MacUser.de too - mainly because I somehow remembered to read something about a bad capacitor on 15" MBPs there some time ago ...
Got this link http://knownappleissues.com/2019/05/13/2008-2009-15-17-macbook-pro-gpu-power-issue/
But I don't think, that in my case this capacitor is the problem, since the MBP boots fine in Save Mode and that faulty C7771 capacitor is to make the machine not booting at all ...

 

[rampancy]

@Amethyst1 So I removed the GeForce* and NVDA* kext files.
And voilà / chapeaux!: MacBook is booting now without booting-loops.
Graphic-performance is decent for Office-Apps, but bloated websites with fast/recurrent re-rendering are a challenge.
At least the Macbook can be used as a thin-client for RDP.
What a shame - that machine was top of crowd at it's time and is still capable of getting 95% of my current tasks done ....
Would Linux encounter the same trouble on booting etc.?
How could this problem be sorted out? GPU-replacement? Bad capacitor?

I'd imagine that Linux would have the same issue, since we're talking about a hardware-level issue; like with macOS you'd have to disable the 9600M GT (via GRUB) to do that. There's also an Single User Mode command in macOS you could try too.

I was also thinking of a software patcher I'd read about, and a motherboard soldering fix on one of Louis Rossmann's videos, but I think those are for 2011 MacBook Pros with the GT 750M.

 

[Amethyst1]

[...] 2011 MacBook Pros with the GT 750M.

You mean 2012/2013? The 2011s have AMD GPUs.

 

[LightBulbFun]

C7771 capacitor is to make the machine not booting at all

id still try and replace it since its on a main power rail for the MCP79, which is what contains the GeForce 9400M

so if that capacitor is just starting to flake out I can see it causing enough issues to crash the GPU but not take down the entire machine

 

[theMarble]

I've not heard of a failing 9400M yet.

There was a case of failing 9400M's back in 2009, but IIRC that was limited to Mid 2009 13" MBP which were running anything below 10.6.8.

 

[Amethyst1]

There was a case of failing 9400M's back in 2009, but IIRC that was limited to Mid 2009 13" MBP which were running anything below 10.6.8.

So the failure only occurred if running 10.6.7 or older?

 

[theMarble]

So the failure only occurred if running 10.6.7 or older?

Yep! I believe it stopped with a firmware update or something like that. All I remember is that there were near-2011 level glitches until 10.6.8.