SSE2 & 3 Retail Leo and Vanilla installs - Boot 132 on pre-Core !
Started by munky, Sep 01 2008 10:31 AM
613 replies to this topic
#21
Posted 02 September 2008 - 10:18 PM
ah well u gots to do ur homework
sorry i cant be of more help...
#22
Posted 02 September 2008 - 10:40 PM
so you can put any kernel in there like a nforce kernel, modbin kernel, if thats true that would be awesome then
#23
Posted 02 September 2008 - 10:43 PM
#24
Posted 03 September 2008 - 07:04 AM
munky, on Sep 2 2008, 08:51 PM, said:
Still looking for some intel sse2 testers!
synapse: draw up a list of kexts ur machine needs to boot a hackintosh and start adding them to the initrd.img. U need to add a clause to the Info.plist to get it to load from the ramdisk. See near the start of the main boot-132 thread for details. I'd post it here but I'm on iPhone with no copy paste ;-)
synapse: draw up a list of kexts ur machine needs to boot a hackintosh and start adding them to the initrd.img. U need to add a clause to the Info.plist to get it to load from the ramdisk. See near the start of the main boot-132 thread for details. I'd post it here but I'm on iPhone with no copy paste ;-)
There is a way to test this in VMware even on Core 2 CPU. I have already done some tests with disabling SSSE3, and should be able to disable SSE3 as well. I will try it and get back to you.
#25
Posted 03 September 2008 - 08:37 AM
#26
Posted 03 September 2008 - 10:35 AM
darkuser:
those are expected behaviour. first attempt was never going to work - you're not specifying the kernel, so it'll try to use the vanilla kernel which wont work on a Pentium D.
in the second example, your machine is waiting for a root device with uuid F8F... which is most likely the initrd ramdisk.
please re-read the first post - you NEED to specify the root device explicitly - either by rd=diskXsY or by boot-uuid=<uuid of the dvd>
you'll need to figure out the values for X and Y, or find the UUID of your DVD. if you have a working hackintosh install on the machine, you can find either by using Disk Utility.
good luck!
woo! thats good news man. i'd like to know the results of that to see if its theoretically possible to boot on SSE2. however, I think real SSE2 machines are still required, as we may have to deal with other deficiencies. i'm not 100% sure what the impact of HPET, PAE, XD etc etc are on booting vanilla - i suspect they're required. can you disable those in your VMWare environment?
imho only Intel chipsets with those features and SSE2 will be likely candidates for booting, unless those deficiencies can be overcome by adding more kexts to the ramdisk. we need, in short, a knowledgeable SSE2 user for testing. any takers?
those are expected behaviour. first attempt was never going to work - you're not specifying the kernel, so it'll try to use the vanilla kernel which wont work on a Pentium D.
in the second example, your machine is waiting for a root device with uuid F8F... which is most likely the initrd ramdisk.
please re-read the first post - you NEED to specify the root device explicitly - either by rd=diskXsY or by boot-uuid=<uuid of the dvd>
you'll need to figure out the values for X and Y, or find the UUID of your DVD. if you have a working hackintosh install on the machine, you can find either by using Disk Utility.
good luck!
Donk, on Sep 3 2008, 08:04 AM, said:
There is a way to test this in VMware even on Core 2 CPU. I have already done some tests with disabling SSSE3, and should be able to disable SSE3 as well. I will try it and get back to you.
woo! thats good news man. i'd like to know the results of that to see if its theoretically possible to boot on SSE2. however, I think real SSE2 machines are still required, as we may have to deal with other deficiencies. i'm not 100% sure what the impact of HPET, PAE, XD etc etc are on booting vanilla - i suspect they're required. can you disable those in your VMWare environment?
imho only Intel chipsets with those features and SSE2 will be likely candidates for booting, unless those deficiencies can be overcome by adding more kexts to the ramdisk. we need, in short, a knowledgeable SSE2 user for testing. any takers?
#27
Posted 03 September 2008 - 11:31 AM
rd(0,1)/mach_kernel.modbin -legacy -v rd=disk2
loading loading and reboot :-(
loading loading and reboot :-(
#28
Posted 03 September 2008 - 11:44 AM
disk2 isnt enough. you need to specify something like disk2s1 which describes the disk number and partition number. if you can boot any old hackintosh Tiger or Leopard on that machine, do so and put the retail DVD in the drive once booted, and open Disk Utility. Select the 'Mac OS X Install DVD' and hit the blue 'Info' button - this should tell you the correct value for the diskXsY and the boot-uuid - either should work.
rd=diskXsY
OR
boot-uuid=ABCDE-123-456-ABCDEF (replace this with your real uuid).
in other news...
an update on my setup - i usually need an older version of IONetworkingFamily and IO80211Family to get Wifi, and I need Natit to get my X1800 to work, and ALCInject for audio.
i placed these on the ramdisk, and booted my vanilla install. result - Natit is doing its thing, I have resolution changing (usually I have QE and CI but no res change), so thats working. Still no wifi, I will need to experiment more. Still no audio.
also, i tried booting the modbin kernel without -legacy. it booted to desktop, but I dont think its running 64-bit mode - certainly Get Info on the Chess application made no mention of 64-bit and Activity Monitor reported Chess.app running as an 'Intel' (rather than Intel 64-bit). Hmm....
rd=diskXsY
OR
boot-uuid=ABCDE-123-456-ABCDEF (replace this with your real uuid).
in other news...
an update on my setup - i usually need an older version of IONetworkingFamily and IO80211Family to get Wifi, and I need Natit to get my X1800 to work, and ALCInject for audio.
i placed these on the ramdisk, and booted my vanilla install. result - Natit is doing its thing, I have resolution changing (usually I have QE and CI but no res change), so thats working. Still no wifi, I will need to experiment more. Still no audio.
also, i tried booting the modbin kernel without -legacy. it booted to desktop, but I dont think its running 64-bit mode - certainly Get Info on the Chess application made no mention of 64-bit and Activity Monitor reported Chess.app running as an 'Intel' (rather than Intel 64-bit). Hmm....
#29
Posted 03 September 2008 - 09:20 PM
was having a think about how best to proceed for my ultimate 'vanilla' setup.
basically what i'd like to do is treat the OS as installed by apple as sacrosanct. i want to get my machine fully working (gfx, audio, wifi etc) without ever having to put anything into /System/Library/Extensions. or at the very least, without having to modify anything thats put there by apple. (putting *extra* stuff in is theoretically ok, cos apple wont overwrite it in a software update).
however, i'd prefer to have it untouched, because i want to share a USB disk between my work machine (a MacBook Pro) and my hackintosh. i have partial success in this so far - I can boot my work machine's clone on the hackintosh and have working gfx (Natit on the ramdisk). If I can have wifi (critical) and audio (nice-to-have) then i'll be set.
ultimately, i'd like to perhaps create a second partition on that disk which will boot the boot-132 routine and then boot the main partition. plugging the same disk into a real mac should just boot the normal apple bootloader from the first partition
basically what i'd like to do is treat the OS as installed by apple as sacrosanct. i want to get my machine fully working (gfx, audio, wifi etc) without ever having to put anything into /System/Library/Extensions. or at the very least, without having to modify anything thats put there by apple. (putting *extra* stuff in is theoretically ok, cos apple wont overwrite it in a software update).
however, i'd prefer to have it untouched, because i want to share a USB disk between my work machine (a MacBook Pro) and my hackintosh. i have partial success in this so far - I can boot my work machine's clone on the hackintosh and have working gfx (Natit on the ramdisk). If I can have wifi (critical) and audio (nice-to-have) then i'll be set.
ultimately, i'd like to perhaps create a second partition on that disk which will boot the boot-132 routine and then boot the main partition. plugging the same disk into a real mac should just boot the normal apple bootloader from the first partition
#30
Posted 03 September 2008 - 11:47 PM
@Dense: sorry, i dunno why your post has ended up halfway thru this topic...
anyway, as you can hopefully see I realised that uuid was unnecessary, as rd=diskXsY works just as well.
im intrigued about the prelinked kernels though... do tell
anyway, as you can hopefully see I realised that uuid was unnecessary, as rd=diskXsY works just as well.
im intrigued about the prelinked kernels though... do tell
#31
Posted 04 September 2008 - 12:32 AM
I have manged to get it work on my Celeron D machine, And I've successfully booted the retail Leopard DVD.
I was having "Still waiting for root device error" but I fixed it by adding AppleVIATA.kext on my BOOT-132 ISO.
I have problems with PS/2 but I'll add its kexts to the BOOT-132 ISO.
Thanks a lot
I was having "Still waiting for root device error" but I fixed it by adding AppleVIATA.kext on my BOOT-132 ISO.
I have problems with PS/2 but I'll add its kexts to the BOOT-132 ISO.
Thanks a lot
#32
Posted 04 September 2008 - 12:51 AM
excellent!! great 3dman, thanks for providing the first independent verification of this method 
i sent you some kexts by PM which may help overcome your PS/2 problems. failing that, go buy a USB keyboard and mouse
i sent you some kexts by PM which may help overcome your PS/2 problems. failing that, go buy a USB keyboard and mouse
#33
Posted 04 September 2008 - 03:24 AM
No PS/2 Problems
Thanks munky for the kexts they fixed PS/2, Now all things are working as they should, I'll install it tomorrow and see what will happen
Thanks
Thanks munky for the kexts they fixed PS/2, Now all things are working as they should, I'll install it tomorrow and see what will happen
Thanks
#34
Posted 04 September 2008 - 09:33 AM
woot!
#35
Posted 04 September 2008 - 10:05 AM
Hi everyone...
i saw something fishy with the 132-BOOT iso`s. Regardless of which iso i boot my pc..after it load the stuff from initrd.img (the whole dots thing ........... ) it says for about a second:
and then continues normaly...
Is that ok to do so? Anyone encountered this thing before?
i saw something fishy with the 132-BOOT iso`s. Regardless of which iso i boot my pc..after it load the stuff from initrd.img (the whole dots thing ........... ) it says for about a second:
System config file '/cam.apple.boot.S/Library/Preferences/SystemConfiguration/com.apple.Boot.plist' not found
and then continues normaly...
Is that ok to do so? Anyone encountered this thing before?
#36
Posted 04 September 2008 - 01:12 PM
synapse: yeah, i noticed that too. doesnt seem to be anything to worry about.
#37
Posted 04 September 2008 - 03:49 PM
munky, on Sep 3 2008, 11:35 AM, said:
woo! thats good news man. i'd like to know the results of that to see if its theoretically possible to boot on SSE2. however, I think real SSE2 machines are still required, as we may have to deal with other deficiencies. i'm not 100% sure what the impact of HPET, PAE, XD etc etc are on booting vanilla - i suspect they're required. can you disable those in your VMWare environment?
imho only Intel chipsets with those features and SSE2 will be likely candidates for booting, unless those deficiencies can be overcome by adding more kexts to the ramdisk. we need, in short, a knowledgeable SSE2 user for testing. any takers?
imho only Intel chipsets with those features and SSE2 will be likely candidates for booting, unless those deficiencies can be overcome by adding more kexts to the ramdisk. we need, in short, a knowledgeable SSE2 user for testing. any takers?
Well my real P4D machine has had a disk crash so won't be testing there until it gets fixed. VMware can mask flags from the CPUID instruction which is how I tested the SSSE3 requirement. It should be able to mask out all the other bits e.g. PAE, XD. HPET is different as is part of the BIOS definition. Whilst I know how to modify a VMware BIOS it is more of a PITA, and would need time to do it.
#38
Posted 04 September 2008 - 10:08 PM
donk: ok man, well if you can manage that'd be cool, otherwise don't sweat it
#39
Posted 05 September 2008 - 06:15 AM
why do you need to put rd(0,1) and boot-uuid. trying to understand how this method works. thanks
#40
Posted 05 September 2008 - 07:35 AM
xfidelity, on Sep 5 2008, 07:15 AM, said:
why do you need to put rd(0,1) and boot-uuid. trying to understand how this method works. thanks
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users



Sign In
Create Account









