Jump to content

HOW TO: Installing 10.4.3 in VMWare

DeathChill

By DeathChill,
in OSx86 10.4 (Tiger)

 Share
216 posts in this topic

Recommended Posts

JaS
JaS
Posted
Posted
Could you explain what you see in -v mode (or make a screen shot). Also, what's your exact CPU spec?

Also, can you boot with -v and look for a message "CPU extended features:" on the Darwin screen. What does it say?

 

I have an Intel P4 3.0E Ghz 1MB L2 w/HTT 32 bit (pics included in the post to show the instructions it can use)

 

I never even get to the "CPU extended features:" on the Darwin screen.

 

Edit: The cpu does use sse3,I used Twekxp to show the CPU info and It doesnt show it in the list.

It might be this unknown instruction listed here in this pic

Link to comment
Share on other sites
kuchdawg
kuchdawg
Posted
Posted

Maxxxus have you tried:

 

# mv /var/db/netinfo/local.nidb /var/db/netinfo/local.nidb.bad

#rm /var/db/.AppleSetupDone

#reboot

 

that used to get people passed the non-loading loading screen...i hope that works for ya....

 

and can the kext also be from PPC version...cause i think i can get that...

Link to comment
Share on other sites
kuruu
kuruu
Posted
Posted
A problem however is that the ApplePS2Controller kext is missing in 10.4.3. So in single user mode you could not use your keyboard. from The 10.4.1 version (originally from Darwin 8.0.1) does not work as this picture shows:

 

Has anyone a idea? We would need the 10.4.3 (Darwin 8.3) kext, compiled on a 10.4.3 or maybe Darwin 8.3 system. Can anybody provide this?

 

 

JaS said on win2osx (link) that he knew someone on irc that could compile with 10.4.3...

Link to comment
Share on other sites
maxxuss
maxxuss
Posted
Posted
I have an Intel P4 3.0E Ghz 1MB L2 w/HTT 32 bit (pics included in the post to show the instructions it can use)

I never even get to the "CPU extended features:" on the Darwin screen.

I'll try to focus on this problem now. But just to make things clear, as I cannot observe the problem and not from your logs (thanks anyway), I need to know:

 

- Do have NX/DX/DEP security feature in your CPU and is it enabled? (i simply check is to use e.g. Everest tool, navigate to Motherboard -> CPUID -> scroll down to Security features)

 

- What is visible in vmware until the crash occurs? Or try to pause the vmware (best with -v boot option) just the moment before the crash and make a screen shot. Specifically: Do you reach "Starting Darwin/x86"?

 

Maxxxus have you tried: ... that used to get people passed the non-loading loading screen...i hope that works for ya....

 

and can the kext also be from PPC version...cause i think i can get that...

Thanks for tip; however the 100% cpu load still remains. And no, the PPC version cannot be loaded.

Link to comment
Share on other sites
sbeehre
sbeehre
Posted
Posted

When the first Phenix DVD was released there was a script that someone created which upgraded darwin 8 to 8.1 i think.... does anyone know where this is now. Could it be adapted to update an 8.2 system to 8.3 from the sources?

Link to comment
Share on other sites
JaS
JaS
Posted
Posted
I'll try to focus on this problem now. But just to make things clear, as I cannot observe the problem and not from your logs (thanks anyway), I need to know:

 

- Do have NX/DX/DEP security feature in your CPU and is it enabled? (i simply check is to use e.g. Everest tool, navigate to Motherboard -> CPUID -> scroll down to Security features)

 

- What is visible in vmware until the crash occurs? Or try to pause the vmware (best with -v boot option) just the moment before the crash and make a screen shot. Specifically: Do you reach "Starting Darwin/x86"?

 

Well Everest says under CPUID-->Sercurity Feartures that Data Execution Prevention(DEP,NX,ED8) Unsupported.

Link to comment
Share on other sites
maxxuss
maxxuss
Posted
Posted
Well Everest says under CPUID-->Sercurity Feartures that Data Execution Prevention(DEP,NX,ED8) Unsupported.

OK. This is different to me. But what about the other question? I need it to check whether it's the boot loader, the kernel or a kext.

Link to comment
Share on other sites
EtherealRemnant
EtherealRemnant
Posted
Posted
I installed the update, I thought everything went well, but after I rebooted nothing was changed. When I clicked on about this mac it still said that I have OSX 10.4.1 I dont know what to do. I cant reload the update, because it took up all the space that I had... and I think I deleated the install file....I am an idiot :unsure:

 

Interesting... I patched the ISO as instructed and booted it in VMWare...

 

It says "Unable to find driver for this platform: "ACPI"

 

And then it panics on the platform start.

 

Booted with platform=X86PC and now have the same kernel dump that you did, Maxxuss.

Link to comment
Share on other sites
EtherealRemnant
EtherealRemnant
Posted
Posted

Me?

 

No, I just modified the ISO (converted from DMG) with hex workshop as described by Maxxuss to disable the SSE3 check.

 

It gave me a completely different panic booting with just -v though, it basically said ACPI wasn't supported so I had to use platform=X86PC which got me to the same panic that Maxxuss had.

 

The only difference of the actual panics is that my frame pointer is 0xc73c3fd4, whereas his was 0xc7953fd4

 

Also - UltraISO wouldn't read the DMG... I used TransMac to do it instead. That's with UltraISO 7 but it was media edition so I don't know if that's why.

Link to comment
Share on other sites
JaS
JaS
Posted
Posted
OK. This is different to me. But what about the other question? I need it to check whether it's the boot loader, the kernel or a kext.

 

I am reinstalling it again in vmware since I thought it might have been something I was doing.So I will post a Screenshot in about 20 min. but I know where it panics.I can set the -v boot option and it starts to load the Extensions.mkext,It appears that after loading the /system/library/extensions/unsupported it panics.

 

I will post a shot of it when the install is done.

Link to comment
Share on other sites
maxxuss
maxxuss
Posted
Posted
It gave me a completely different panic booting with just -v though, it basically said ACPI wasn't supported so I had to use platform=X86PC which got me to the same panic that Maxxuss had.

 

The only difference of the actual panics is that my frame pointer is 0xc73c3fd4, whereas his was 0xc7953fd4

Two things:

 

- Did you try the mach_kernel I posted here?

- Can you make a screenshot of the panic screen (running with -v).

Link to comment
Share on other sites
EtherealRemnant
EtherealRemnant
Posted
Posted

Well, I haven't actually messed with the install yet. VMWare is way too slow for me so I'm directly working on making the install ISO boot. I'll try that though.

 

One question... is there anything other than UltraISO... preferably something that will write changes directly to the ISO instead of making a new one?

 

And if you want the panic, yeah, I can do that. But its the exact same thing you got, minus the difference in CPU and the frame pointer. Its the commpage_populate panic.

 

Maybe I'll try the install later on tonight... I don't have anything better to do.

Link to comment
Share on other sites
maxxuss
maxxuss
Posted
Posted
One question... is there anything other than UltraISO... preferably something that will write changes directly to the ISO instead of making a new one?

Use Hex Workshop. Open Options -> Preference -> Configuration. Set "Enable direct file editing". You can now load the ISO file and work on it.

 

And if you want the panic, yeah, I can do that. But its the exact same thing you got, minus the difference in CPU and the frame pointer. Its the commpage_populate panic.

My mach_kernel.test patches the commpage_populate SSE3 check. If any adress in the trace is different, this is important and might indicate other locations requiring a patch here. (commpage routines are enhanced in 10.4.3 compared to Darwin 8.3, e.g. they used SSE2 enhanced bzero, bcopy etc.)

Link to comment
Share on other sites
EtherealRemnant
EtherealRemnant
Posted
Posted

I was talking more with working on the individual files. I already did the hex workshop thing.

 

Just patched the mach_kernel on the ISO and it booted with a new kernel dump. One second and I'll post the image.

 

BTW, I found that TransMac will immediately write changes to files to the ISO and it keeps it as a hybrid image, whereas UltraISO has to write a new ISO and it deletes the hybrid, just making it an ISO, which crashes VMWare.

 

EDIT: I may have screwed something up doing the hex edit beforehand so I'm ripping a new ISO off the burned DVD I made and I'm going to try patching it again and see if I get any further. I'm still learning all of this so it makes more sense in my mind for me to play with the ISO first and see if I can get that to boot before I start spending hours installing through VMWare because I'll have to install a fresh 10.4.1 on an image, then make a new image for 10.4.3 and install 10.4.3 onto that, THEN try to work through the rest.

 

Anyway, image-wise, here you go:

post-11287-1131243781_thumb.jpg

Link to comment
Share on other sites
maxxuss
maxxuss
Posted
Posted
maxxuss since I couldnt get you a screenshot of exactly where it panics I just used vmware to make a small movie of the boot up :unsure: you can download it from here http://www.mytempdir.com/247015

JaS, Nice idea. So before the panic, there is no "Starting Darwin/x86"? In the video I see "Loading Darwin/x86", followed by the loading of the kexts into kernel space, but not the last display message from the boot loader saying "Starting Darwin/x86". (Sorry if this redundant question, but it's hard to handle different issues all at once.) If this last message does not get displayed it might indicate that you're stuck in the boot loader.

Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...