macOS 10.14 Mojave on Unsupported Macs Thread

[ASentientBot]

Boot up on a "prohibitory or stop sign"

I have this current situation on an unsupported MacbookPro 15" mid-2009 (MacbookPro5,3)

1. Used @dosdude1's Mojave Patcher v1.1.1. Successfully installed and patched on an external USB bootable drive.
2. Rebooted on my MBP5,3, and got the prohibitory sign after showing the apple logo (on gray screen).
3. Tried the USB boot drive on my MBP5,5. Same failed result.
4. Tried the USB boot drive on my unsupported MBAir4,2. This one worked (booted up fine).
5. Additional note: my MBP5,3 does NOT have any internal drives (will install one later). My MBP5,5 does have an internal drive. But still failed result.

Any suggestion folks?

Appreciate it.

Open the USB drive /Library/Preferences/SystemConfiguration/com.apple.Boot.plist

Change the kernel flags to -v -no_compat_check

See if that helps.

 

[TimothyR734]

Boot up on a "prohibitory or stop sign"

I have this current situation on an unsupported MacbookPro 15" mid-2009 (MacbookPro5,3)

1. Used @dosdude1's Mojave Patcher v1.1.1. Successfully installed and patched on an external USB bootable drive.
2. Rebooted on my MBP5,3, and got the prohibitory sign after showing the apple logo (on gray screen).
3. Tried the USB boot drive on my MBP5,5. Same failed result.
4. Tried the USB boot drive on my unsupported MBAir4,2. This one worked (booted up fine).
5. Additional note: my MBP5,3 does NOT have any internal drives (will install one later). My MBP5,5 does have an internal drive. But still failed result.

Any suggestion folks?

Appreciate it.

did you format your drive with HFS or APFS I have a MacBook 5,2 and tried APFS on a external usb got the same "prohibitory or stop sign as you so I tried the APFS bot rom tool it was not a supported device so I installed macOS Mohave on my internal hdd apfs still would not boot so formatted to hfs and no issues

 

[visberry]

I would

Oh yeah - for sure. My Light Mode on macpro 3,1 + GTX680 is fine without transparency. The mbp5,3 however requires it (I increased contrast too) and it has a GeForce 2400M 256MB. So it's GPU/acceleration/metal (shader?) related. I think @jackluke tried swapping out the entire Skylight framework but it doesn't boot correctly. No surprises there as Skylight is a critical private framework retooled between HS and Mojave (per my previous comment)
At this point, I think the only option is to trace through the code (as @ASentientBot) suggested and glean a patch.
If it occurs on supported systems, Apple may respond to a radar - but I don't think it does (unless I'm mistaken) so fat chance of that happening.

Peeling though the object code I did find symbols related to shaders associated with ui components in dark (like toolbar) and a new addition from the HS base :

00000000002d32f0 b __ZL34_placeholder_selection_system_dark

A class dump of the Skylight binary also indicates the offending code is probably in WindowServer or a lower level framework or dylib. The Skylight class-dump looks like this:

//
//     Generated by class-dump 3.5 (64 bit).
//
//     class-dump is Copyright (C) 1997-1998, 2000-2001, 2004-2013 by Steve Nygard.
//

#pragma mark -

//
// File: /System/Library/PrivateFrameworks/SkyLight.framework/Versions/A/SkyLight
// UUID: 25DCE5F7-0737-301F-8A27-CAB3A1AA298C
//
//                           Arch: x86_64
//                Current version: 600.0.0
//          Compatibility version: 64.0.0
//                 Source version: 336.8.2.0.0
//
// Objective-C Garbage Collection: Unsupported
//

@protocol MTLResource <NSObject>
@property(readonly) unsigned long long allocatedSize;
@property(readonly) id <MTLHeap> heap;
@property(readonly) unsigned long long storageMode;
@property(readonly) unsigned long long cpuCacheMode;
@property(readonly) id <MTLDevice> device;
@property(copy) NSString *label;
- (BOOL)isAliasable;
- (void)makeAliasable;
- (unsigned long long)setPurgeableState:(unsigned long long)arg1;
@end

@protocol MTLResourceSPI <MTLResource>
@property(readonly) MTLResourceAllocationInfo *cachedAllocationInfo;
@property(readonly) MTLResourceAllocationInfo *sharedAllocationInfo;
@property int responsibleProcess;
- (BOOL)doesAliasAnyResources:(const id *)arg1 count:(unsigned long long)arg2;
- (BOOL)doesAliasAllResources:(const id *)arg1 count:(unsigned long long)arg2;
- (BOOL)doesAliasResource:(id <MTLResource>)arg1;
- (void)waitUntilComplete;
- (BOOL)isComplete;
- (BOOL)isPurgeable;
@end

@protocol NSObject
@property(readonly, copy) NSString *description;
@property(readonly) Class superclass;
@property(readonly) unsigned long long hash;
- (struct _NSZone *)zone;
- (unsigned long long)retainCount;
- (id)autorelease;
- (oneway void)release;
- (id)retain;
- (BOOL)respondsToSelector:(SEL)arg1;
- (BOOL)conformsToProtocol:(Protocol *)arg1;
- (BOOL)isMemberOfClass:(Class)arg1;
- (BOOL)isKindOfClass:(Class)arg1;
- (BOOL)isProxy;
- (id)performSelector:(SEL)arg1 withObject:(id)arg2 withObject:(id)arg3;
- (id)performSelector:(SEL)arg1 withObject:(id)arg2;
- (id)performSelector:(SEL)arg1;
- (id)self;
- (Class)class;
- (BOOL)isEqual:(id)arg1;

@optional
@property(readonly, copy) NSString *debugDescription;
@end

@interface SLSessionOwner : NSObject
{
    unsigned int _port;
    NSNumber *_auditSessionID;
    NSNumber *_sessionID;
}

+ (id)sessionOwnerForNewSessionWithAuditSessionID:(id)arg1;
+ (id)sessionOwnerForNewSessionWithAuditSessionID:(id)arg1 launchData:(id)arg2;
+ (id)sessionOwnerBySettingLoginwindowConnection:(unsigned int)arg1;
+ (id)sessionOwnerWithXPCSerialization:(id)arg1;
+ (id)sessionOwnerWithPort:(unsigned int)arg1 auditSessionID:(int)arg2 cgSessionID:(unsigned int)arg3;
@property(retain, nonatomic) NSNumber *sessionID; // @synthesize sessionID=_sessionID;
@property(retain, nonatomic) NSNumber *auditSessionID; // @synthesize auditSessionID=_auditSessionID;
@property(nonatomic) unsigned int port; // @synthesize port=_port;
@property(readonly, nonatomic, getter=isValid) BOOL valid;
- (id)createXPCSerializationAndInvalidate;
- (id)initWithXPCSerialization:(id)arg1;
- (id)initWithPort:(unsigned int)arg1 auditSessionID:(int)arg2 cgSessionID:(unsigned int)arg3;
- (void)dealloc;

@end

So some higher level code that establishes connections and sessions...

I would be careful of a radar.
[doublepost=1533418122][/doublepost]Maybe try swap out the Sky framework with a 2012 one slowly.

 

[pkouame]

I would

I would be careful of a radar.
[doublepost=1533418122][/doublepost]Maybe try swap out the Sky framework with a 2012 one slowly.

I know and you're right - i would only submit one for a supported machine with similar Mojave glitches - which is not our case. Radars are pretty useless anyway, as are the developer forums of late...Apple has crippled all that open source sharing mantra a while ago...

 

[TimothyR734]

maybe another possible reason for not being able to update software updates and macOS installations might be prevented if more than one volume has the same name :0

 

[ASentientBot]

maybe another possible reason for not being able to update software updates and macOS installations might be prevented if more than one volume has the same name :0

If two volumes have the same name, one of them gets "1" appended to it in the file system. Like:

$ ls /Volumes
Macintosh HD
Untitled
Untitled 1

But in Finder this is hidden.

However, this shouldn't trip up any scripts since they actually have different names.

 

[visberry]

If two volumes have the same name, one of them gets "1" appended to it in the file system. Like:

$ ls /Volumes
Macintosh HD
Untitled
Untitled 1

But in Finder this is hidden.

However, this shouldn't trip up any scripts since they actually have different names.

It’s a known issue in Mojave as per release notes.

 

[pkouame]

maybe another possible reason for not being able to update software updates and macOS installations might be prevented if more than one volume has the same name :0

While I'm not sure what happens in that case, I didn't have conflicting logical volume names.

At least it downloads the entire installer consistently.

So 50% of OTA system update is achieved. When I'm in a hurry, I just manually patch and run the OSInstaller.mpkg in InstallESD.dmg - post-patching necessary of course. Only thing is, quick and dirty doesn't create the latest recovery partition of course. But there are ways around that and the latest and greatest recovery partition is not always necessary (for me at least) We'll see what happens with devB6.

 

[TimothyR734]

I still think the possibility is that the software update has a radar that scans for the version we have installed and the version of the install macOS Mojave beta so if you have 18A336e installed the install macOS Mojave beta should be the same 18A336e thus you get the update if you don't have it or doesn't match you get the full installer
[doublepost=1533427534][/doublepost]

If two volumes have the same name, one of them gets "1" appended to it in the file system. Like:

$ ls /Volumes
Macintosh HD
Untitled
Untitled 1

But in Finder this is hidden.

However, this shouldn't trip up any scripts since they actually have different names.

I read that in the known issues for macOS Mojave release notes

 

[pkouame]

I still think the possibility is that the software update has a radar that scans for the version we have installed and the version of the install macOS Mojave beta so if you have 18A336e installed the install macOS Mojave beta should be the same 18A336e thus you get the update if you don't have it or doesn't match you get the full installer
[doublepost=1533427534][/doublepost]
I read that in the known issues for macOS Mojave release notes

Yup. Software Update reads /S/L/CoreServices/SystemVersion.plist for your build and OS info. The seed process also reads /L/Preferences.com.apple.seeding.plist to figure out which seed (public or dev) you are qualified for (part of the enrollment process) So yes, it's smart enough to determine what you should get full or patches to individual products. Normal OTA system updates are coming through fine for me (like my xcode updates) . All is good there.

The only issue I find puzzling is that dB3 to dB4 OTA did that, but also kicked off the install immediately (see the screenshots I posted earlier) So it "supported" my system without manual intervention. db4 to db5 downloaded fine but was basically the same image one gets if you download from a supported machine. Normally the software update patch and faker dylib intercept the installation checks(by replacing VMM with FPU). This time it didn't. Strange.

 

[thegurgz]

when I successfully upgraded to DP5 on my internal hdd on my iMac 9,1 I had version b12 of dosdude's Mojave patcher which is version 14014 18A3363 DP4 so when DP5 came out I originally had Mojave enrolled in the public beta so I used SWU Switcher and switched to developers beta checked the software update and successfully updated from 14014 to 14015 DP5 I think the software update scans for a install macOS Mojave beta and which version it is so if you don't have a install macOS Mojave beta it downloads the full installer. Both my macOS Mojave patcher and install macOS Mojave beta are in my applications folder

I used the SWU Switcher and switched to dev beta, running 18A336e. After restart, had it run for updates, still saying mac is up to date. Why won't it show that build 18A347e is available?

 

[TimothyR734]

I used the SWU Switcher and switched to dev beta, running 18A336e. After restart, had it run for updates, still saying mac is up to date. Why won't it show that build 18A347e is available? View attachment 774419

one issue could be you have automatically keep my Mac up to date unchecked and we really haven't narrowed down why it updates on some mac's with the software update but I downloaded version 1.1.1 of the patcher and macOS Mojave beta 5 with the patcher and installed on top DP4 with no issues
[doublepost=1533430504][/doublepost]

Yup. Software Update reads /S/L/CoreServices/SystemVersion.plist for your build and OS info. The seed process also reads /L/Preferences.com.apple.seeding.plist to figure out which seed (public or dev) you are qualified for (part of the enrollment process) So yes, it's smart enough to determine what you should get full or patches to individual products. Normal OTA system updates are coming through fine for me (like my xcode updates) . All is good there.

The only issue I find puzzling is that dB3 to dB4 OTA did that, but also kicked off the install immediately (see the screenshots I posted earlier) So it "supported" my system without manual intervention. db4 to db5 downloaded fine but was basically the same image one gets if you download from a supported machine. Normally the software update patch and faker dylib intercept the installation checks(by replacing VMM with FPU). This time it didn't. Strange.

maybe a radar

 

[pkouame]

one issue could be you have automatically keep my Mac up to date unchecked and we really haven't narrowed down why it updates on some mac's with the software update but I downloaded version 1.1.1 of the patcher and macOS Mojave beta 5 with the patcher and installed on top DP4 with no issues
[doublepost=1533430504][/doublepost]
maybe a radar

I'm not sweating this too much right now TIMTOWTDI More betas to come...

 

[jackluke]

In attempting deleting one by one UserEventPlugins to luckily boot with SkyLight from HS I didn't succeed, but booting in verbose mode with the SkyLight from Mojave, I've replaced on-the-fly the SkyLight HS, then logged out, logging out on the verbose logging out I read "process crashed mdworker_shared" (it should be spotlight) and it stucks there without logging out user. So I am near to drop.


BUT in meanwhile this is an experiment just for playtime:

- You can delete all the 45/46 plugins into /System/Library/UserEventPlugins/
Except this one: com.apple.remoted.plugin
Without this one you will stuck on black screen after stage2 apple logo.

Leaving only that one as if by magic Mojave will boot and perfectly working.

I've done the experiment on a Macbook7,1 but I guess it will work on any C2D Mac and maybe any unsupported Mac.

Of course try just for fun, better leaving inside all in.

 

[pkouame]

In attempting deleting one by one UserEventPlugins to luckily boot with SkyLight from HS I didn't succeed, but booting in verbose mode with the SkyLight from Mojave, I've replaced on-the-fly the SkyLight HS, then logged out, logging out on the verbose logging out I read "process crashed mdworker_shared" (it should be spotlight) and it stucks there without logging out user. So I am near to drop.


BUT in meanwhile this is an experiment just for playtime:

- You can delete all the 45/46 plugins into /System/Library/UserEventPlugins/
Except this one: com.apple.remoted.plugin
Without this one you will stuck on black screen after stage2 apple logo.

Leaving only that one as if by magic Mojave will boot and perfectly working.

I've done the experiment on a Macbook7,1 but I guess it will work on any C2D Mac and maybe any unsupported Mac.

Of course try just for fun, better leaving inside all in.

Play and fun are good, we all learn from it. Just curious though, what do you think the relation between UserEventPlugins and the Skylight framework is? With respect to the light mode glitch of course...

 

[ASentientBot]

Play and fun are good, we all learn from it. Just curious though, what do you think the relation between UserEventPlugins and the Skylight framework is? With respect to the light mode glitch of course...

I doubt that there is any relation.

If I had to guess, I'd say the problem is connected to AppKit.framework, OpenGL.framework, or one of their dependencies. Since this happens on all unsupported graphics hardware, it's probably a product of rendering the UI using OpenGL without Metal.

Playing with the Swift file I uploaded earlier, different NSVisualEffectView.material settings cause either normal behavior or just grey. All the code for that is in AppKit. I disassembled that and I've been trying to understand how it draws the transparent window backgrounds. I found higher-level functions for redrawing the window but I haven't yet traced it to the source.

If anybody else wants to look into this, try opening AppKit in Hopper Disassembler and maybe start with [NSVisualEffectView drawRect]. The variable accessibilityDisplayShouldReduceTransparency might also provide some useful hints.

I'll check back in after a day or two if I make any progress. Good luck everybody!

 

[thegurgz]

Run into a large problem. Redownloaded patcher, redownloaded Mojave, rewrote it to a drive, reinstalled Mojave. Ran post install, and now it refuses to boot. It was booting perfectly before reinstall. After “Waiting for DSMOS” it just gives back “operation failed” followed by “unexpected session” followed by busy timeouts. I’ve ran into this issue with high Sierra in the past, and the only solution I’ve found is a clean install. Any advice?

 

[TimothyR734]

Run into a large problem. Redownloaded patcher, redownloaded Mojave, rewrote it to a drive, reinstalled Mojave. Ran post install, and now it refuses to boot. It was booting perfectly before reinstall. After “Waiting for DSMOS” it just gives back “operation failed” followed by “unexpected session” followed by busy timeouts. I’ve ran into this issue with high Sierra in the past, and the only solution I’ve found is a clean install. Any advice?

did you try force rebuild caches after you did the post install

 

[thegurgz]

did you try force rebuild caches after you did the post install

Yes, same result.

 

[TimothyR734]

did you check to make sure your SIP is disabled is your hdd or sad formatted HFS or APFS

 

[thegurgz]

did you check to make sure your SIP is disabled is your hdd or sad formatted HFS or APFS

Don’t know what happened, but after letting it sit there for an hour, it magically started running code in verbose and booted up. Fixed for now I guess.

 

[TimothyR734]

Don’t know what happened, but after letting it sit there for an hour, it magically started running code in verbose and booted up. Fixed for now I guess.

Let the OS settle a bit after setup it will improve as you use it

 

[DaniloGeekDude]

Open the USB drive /Library/Preferences/SystemConfiguration/com.apple.Boot.plist

Change the kernel flags to -v -no_compat_check

See if that helps.

No dice. I even included the -keepsyms on the kernel flags along with the ones above.
It got along further, verbose boot messages received and then gray screen with the prohibitory sign. But boot messages continued. After a while, I just powered off. Still looking at other options. Thanks for the suggestion.
[doublepost=1533447626][/doublepost]

did you format your drive with HFS or APFS I have a MacBook 5,2 and tried APFS on a external usb got the same "prohibitory or stop sign as you so I tried the APFS bot rom tool it was not a supported device so I installed macOS Mohave on my internal hdd apfs still would not boot so formatted to hfs and no issues

The USB drive is formatted HFS+. Thanks.

 

[ASentientBot]

No dice. I even included the -keepsyms on the kernel flags along with the ones above.
It got along further, verbose boot messages received and then gray screen with the prohibitory sign. But boot messages continued. After a while, I just powered off. Still looking at other options. Thanks for the suggestion.
[doublepost=1533447626][/doublepost]

The USB drive is formatted HFS+. Thanks.

Ohhhh, I thought that the prohibitory sign was a "not supported" prohibitory sign. This sounds more like a "still waiting for root device" prohibitory sign. Any chance that's the last message that shows up in the log?
[doublepost=1533448186][/doublepost]

Run into a large problem. Redownloaded patcher, redownloaded Mojave, rewrote it to a drive, reinstalled Mojave. Ran post install, and now it refuses to boot. It was booting perfectly before reinstall. After “Waiting for DSMOS” it just gives back “operation failed” followed by “unexpected session” followed by busy timeouts. I’ve ran into this issue with high Sierra in the past, and the only solution I’ve found is a clean install. Any advice?

I'm 90% sure the "operation failed" and "unexpected session" messages are normal. The busy timeouts happen to me too (very rarely) since Sierra or High Sierra. Never figured out the source. I just force a reboot and it works fine the second time.

 

[mat77_mac]

In first post desdude say amd video card Radeon 6xxx no accelleration in iMac mid 2011 but system work all or not work? If possibile use al the function or not? in short, I would like to know exactly what's wrong and what's wrong! can I use internet office programs or not? give me a list of things that go without acceleration and things that need acceleration thanks