[GUIDE] Retail OS X Install (10.5.8) on Gigabyte GA-EX58-UD5 (Core i7) Mobo

[iamnicos]

Try this: support.apple.com/downloads/Mac_OS_X_10_5_7_Combo_Update

 

or just do a google search.

 

792Mb...what ? I uploaded it few days ago and it was a 442,8Mo file

 

 

According to Netkas, we were not to update from "Sofware Update" but from the update file, what can be easily understood regarding to our mac model problems...but if they change that file adding like 300Mb ...It may be explaining lots of things...

[airwalk776]

yah my bios has S3 as well, does the other settings in the bios affect this?

strange that me and DD have the same issue though.

 

I have the same issue with sleep on the EVGA x58 SLI mobo set to S3. Updated the bios also (which EVGA claims to have correct S3 issues)

[FUT1L1TY]

792Mb...what ? I uploaded it few days ago and it was a 442,8Mo file

 

 

It is my understanding that the Combo updater will update any version of 10.5. That's why it's bigger. It's typically used by the IT guy/gal at a company to update a lot of different machines.

 

Most mac users simply click software update and get the smaller more specific updater for their system. This updater did not work for me to say the least.

 

The important thing about the combo updater is that you can apply to a drive that is not your boot drive. I highly recommend doing this.

[AdamNJ77]

Here

 

You meant FUT1L1TY's instructions right ?

 

Be sure to remove kernel and system.kext in repository and to add UUID.kext in msc/patches...

 

Let us know of course...

 

Thank you very much

 

And sorry FUT1L1TY, I forgot I don't know how I got the names mixed up. It is your instructions I wish to follow. My bad

 

Update: Looks like I am going to have to wait a few days. The old drive I was using as an external drive with iPC just died so I grabbed a new 80gig 2.5" SATA drive off Ebay for $25 shipped and it will be here in 2-3 days. As soon as I try this out I will post results unless someone beats me to it

[Asterra]

Just to chime in, the non-724MB 10.5.7 combo update I downloaded was from here:

 

http://mac.softpedia.com/progDownload/Appl...load-29834.html

 

With it, I utilized the installation procedure outlined by FUT1L1TY. The end result is an almost completely working 10.5.7 installation on my 4870-based Hack, with resolution change, QE/CI, and better System Profiler information than "can't get info". It's very exciting, I must say.

 

The only remaining issue is the information System Profiler is reporting. It gets the ram type wrong, but I do understand that I ought to be able to easily fix that by changing SMmemtype from 19 to 24.

 

The CPU information remains inaccurate. System Profiler believes it is running at 3000 Mhz (it is not). CPU-X believes it is 3000Mhz x 9. Don't I wish. I figure there is probably something fishy with smbios.plist that can be fixed as easily as the SMmemtype variable. Here are a few suspect lines of text:

 

		<key>SMexternalclock</key>
	 <string>333</string>
	 <key>SMmaximalclock</key>
	 <string>3000</string>

Would it happen to be among those two variables? Obviously the SMmaximalclock is particularly suspect, but I don't want to provide some bad data and cause my house to detonate. I also plan to overclock just as soon as things seem to be working, so if "maximalclock" is supposed to mean the CPU shouldn't be over 3000Mhz, well that's not going to work either.

 

So basically, is System Profiler getting its information from smbios.plist? Is CPU-X? Sure it'll be nice if changing a bit of text makes both of those programs behave, but not at the cost of overclocking flexibility.

[animani]

I just did this:

 

1. I took my mini 9 hackintosh, connected my HDD for the x58-ud5 and opened DD's script version 3.6 and ran step 2,3,and 4.

2.updated to 10.5.7 using the combo update and selected my x58 drive for the install(because my retail disc is 10.5.5.

3.Opened the script again and ran step 5( I deleted the kernel and the natit.kext because I wanted a vanilla install and have no need for the natit.kext).

4. I disconnected my HDD and connected it to my GA-EX58-UD5 and booted from it.

5. Once it booted I downloaded the script and ran the processor info changer and then deleted the script.

6. I ran software update and of course checked my system profiler to make sure everything was working.

I show all 4 cores and in activity monitor all 8.

 

My system is as follows:

GA-EX58-UD5

1 TB SATA HDD

OS X 10.5.7 Retail and VMware 2 Win 7 RC 2

LG dual layer burner

Nvidia Geforce 9800GT 512Mb

ocZ 12GB DDR3 Ram

 

If you have issue's booting make sure you have you bios profile the same as DD's.

 

You said it shows 4 cores but in activity 8 is that a bug still?

 

Is it everything working?

[AdamNJ77]

You said it shows 4 cores but in activity 8 is that a bug still?

 

Is it everything working?

 

 

 

I think he is referring to the i7 having Hyperthreading and each core shows up as two separate threads. This is normal in Windows it shows 8 threads as well. Even though it is a quad core processor each core can multitask and shows as two seperate threads in any CPU monitor. So it is correct in stating it is a 4 core machine and showing 8 activity bars

[randolpholus]

Your BIOS is not set to look at the right drive. This will happen if it's looking at a flash drive, for example, that doesn't have a bootable OS installed.

 

MAJ

 

Thanks DD for the response.

 

I matched my BIOS settings to match your figure on the front page.

 

My setup:

EX58-UD5

Intel 920

6 GB 1600 Ram.

1 5400 RPM 2.5" drive attached via USB

1 7200 RPM 3.5" drive (1.5 TB) attached via onboard SATA.

 

I can boot the 5400 RPM drive with no problem. I used this drive to set up my 1.5 TB drive. I installed on the 1.5 TB using your latest method. I can browse the drive and I see OSX installed correctly, I also see the kernel and the /Extras folder making it seem like the install went ok as well. I then shutdown the box, remove the USB drive and reboot again hoping to boot from the 1.5 TB drive. I still get

 

Verifying DMI Pool Data...................

boot0: GPT

boot0: testing

boot0: testing

boot0: error

 

I also unplugged all hard drives. The only thing plugged in is the hard drive and the DVD drive. Still can't get passed the error.

 

Do you think this has anything to do with my installation setup (using USB instead of on-board SATA)?

[Jory Cameron]

I think he is referring to the i7 having Hyperthreading and each core shows up as two separate threads. This is normal in Windows it shows 8 threads as well. Even though it is a quad core processor each core can multitask and shows as two seperate threads in any CPU monitor. So it is correct in stating it is a 4 core machine and showing 8 activity bars

 

Exacaly, well stated!

[mattrb]

DD, and the UD5 community, time to update to 10.5.7,

 

889a Audio legacy Kexts

Read P.S. below before getting this kext

 

for 10.5.7.

 

VoodooHDA was beginning to suck because no one would tweak it for 889a, consequently I had no SPDIF output.

 

tmongkol is to thank for this kext as well as countless others.

 

P.S. I guess these are not auto headphone switching but the 889a page for getting the correct one is here

[digital_dreamer]

Thanks MAJ

 

What would be the correct method here?

 

Is installing the new bootloader on EFI best done via terminal?

 

Would I have to copy a version of the vanilla kernel to the EFI partition or do I remove the voodoo kernel and let the system use the vanilla kernel at root? - if so that would that need an edit to the boot.plist (that points at voodoo)?

Cosmin Petre,

I almost forgot about this post.

 

The Chameleon team has a installer for the EFI boot. There's also some other scripts that install that bootloader for EFI partition.

Check out their site for a installer: chameleon.osx86.hu/articles/chameleon-20rc1-is-out

(I can't leave URL's here)

 

There's also this:

Modify EFI HFS Boot Partition for Chameleon 2

 

I believe the kernel needs to be in the same location as the bootloader. Not too sure about that, but that's the impression I have. So, you'd want a copy of that same 9.7.0 vanilla kernel installed to EFI. There's no need for Voodoo here (kernel, that is!).

 

MAJ

 

DD, and the UD5 community, time to update to 10.5.7,

 

889a Audio legacy Kexts

 

for 10.5.7.

 

VoodooHDA was beginning to suck because no one would tweak it for 889a, consequently I had no SPDIF output.

 

tmongkol is to thank for this kext as well as countless others.

Thanks, mattrb,

I already had those included, but I just never had the time to sort them out. They're the "no front mic" versions - just "2ins". They'll help get those fMikeyDriver errors out of there. I should just pull out all the "3ins" and just keep these to lesson the confusion.

 

Just a heads up on which to use: IIRC, we normally use the "digital HDA headphone", which offers the auto headphone switching.

 

EDIT: I wanted to say also, thanks for your checking out the VoodooHDA and providing feedback on that. You may never had gotten any response from us, but I enjoyed reading your input on that. If you hadn't tried it, I would have.

Again, thanks for your contributions.

 

regards,

MAJ

[digital_dreamer]

First a BIG thanks to everyone in the thread.... esp. digital_dreamer.

 

I got sleep working on my UD5 rig using

sudo pmset -a hibernatemode 2

i'm using the GA-EX58-UD5 - F8b Bios from http://forums.tweaktown.com/f69/gigabyte-latest-bios-28441/

Thanks for that tip!

 

hibernatemode 2 appears to work great on this system!

Was able to wake just fine.

 

Default is 0, so do a pmset -g to see what your's is set to.

 

regards,

MAJ

[iamnicos]

name='FUT1L1TY' post='1159441' date='May 18 2009, 10:58 PM']It is my understanding that the Combo updater will update any version of 10.5. That's why it's bigger.

 

Yes a figured that out...I was tired.

[animani]

I think he is referring to the i7 having Hyperthreading and each core shows up as two separate threads. This is normal in Windows it shows 8 threads as well. Even though it is a quad core processor each core can multitask and shows as two seperate threads in any CPU monitor. So it is correct in stating it is a 4 core machine and showing 8 activity bars

 

Ohh cool thanks for making that clear.

 

[rama]

hey, I've just updated from .6 to .7

Everything seems to be running ok, except i get a lot more crashes now. especially if i'm away from the computer for like 10+ minutes... then when i come back and try to keep working (especially in Logic 8) it just seizes up and then crashes.

 

Anyone else noticed finicky behavior?

 

Do i need to remove System.kext from the s/l/e folder? or any other kexts? im not using hda cos ive got an apogee duet f/w sound card.. do i need to still replace those hda kexts too?

 

cheers.

[cwall64]

Thanks for that tip!

 

hibernatemode 2 appears to work great on this system!

Was able to wake just fine.

 

Default is 0, so do a pmset -g to see what your's is set to.

 

regards,

MAJ

 

So was everyones sleep working before the 10.5.7 upgrade set at " hibernatemode 0"? I had sleep fine on 10.5.6 at "0"... I will try the mode 2, curious as to what mode 2 is (man pmset didn't say anything about the various states)?

[digital_dreamer]

So was everyones sleep working before the 10.5.7 upgrade set at " hibernatemode 0"? I had sleep fine on 10.5.6 at "0"... I will try the mode 2, curious as to what mode 2 is (man pmset didn't say anything about the various states)?

Mode 2 appears to be tailored to Macbooks:

"0010 (bit 1), in conjunction with bit 0, causes OS X to maintain system state in memory and leave system power on until battery level drops below a near empty threshold (This enables quicker wakeup from memory while battery power is available). Upon nearly emptying the battery, OS X shuts off all system power and hibernates; on wake the system will resume from hibernation image, not from memory."

 

As for the other modes:

0 - Old style sleep mode, with RAM powered on while sleeping, safe sleep disabled, and super-fast wake.

1 - Hibernation mode, with RAM contents written to disk, system totally shut down while “sleeping,” and slower wake up, due to reading the contents of RAM off the hard drive.

3 - The default mode on machines introduced since about fall 2005. RAM is powered on while sleeping, but RAM contents are also written to disk before sleeping. In the event of total power loss, the system enters hibernation mode automatically.

5 - This is the same as mode 1, but it’s for those using secure virtual memory (in System Preferences -> Security).

7 - This is the same as mode 3, but it’s for those using secure virtual memory.

 

mode 3 sounds the most flexible: RAM powered, but contents saved to disk in case of power loss. So, you get fast wakes, or slower wake if power was cut, but at least your work is saved.

I'm not at home now. Has anyone tried mode 1 and 3?

 

MAJ

[joeblough]

So was everyones sleep working before the 10.5.7 upgrade set at " hibernatemode 0"? I had sleep fine on 10.5.6 at "0"... I will try the mode 2, curious as to what mode 2 is (man pmset didn't say anything about the various states)?

 

after some fits and starts i got 10.5.7 up and running on a new disk. it did hang once while applying the first set of software updates. however, wake/sleep into S3 (UD4P running bios F9a) seems to work without changing hibernatemode at all.

 

i ran script v3.6, installing 10.5.0 and then put the 9.6.0 voodoo kernel and system.kext on the disk and rebooted; however midway through the boot it hung. its such an old system i'm not surprised at that. i then booted a working disk and installed 10.5.7 on the disk, and it came up. but i dont know if this is really the right way to go - the 10.5.7 upgrade was supposed to reboot once on its own and that never happened.

 

i wonder if the right way to do this is to use an older script, but install 10.5.0 and then immediately apply 10.5.6, and then continue the script, making sure to install a custom kernel. then once its up, run the new script but instead of running the installer, just run the 10.5.7 updater (the delta should be fine at that point, but i usually use the combo).

[AdamNJ77]

Thanks for that tip!

 

hibernatemode 2 appears to work great on this system!

Was able to wake just fine.

 

Default is 0, so do a pmset -g to see what your's is set to.

 

regards,

MAJ

 

drat...I too did this sudo command but I still have the exact same problem with my computer and display sleeping, and the computer wakes up but not the monitor. I wonder if it has anything to do with my Radeon 4870 video card. Everything else works perfect for the video card without any modified kexts or anything right from the 10.5.7 update (unlike 10.5.6). If anyone out there is using an ATI 4000 series card and has sleep working please let me know HOW in the heck you got it done. As far as my other settings in the "pmset -g" command, sleep, displaysleep, powerbutton, and disksleep are all set to "0". "hibernatemode" is set to 2, and "ttyskeepawake" is at 1. Does that sound about right to everyone or is there anything that should be changed there, or that is different in your settings. Thanks

[iamnicos]

drat...I too did this sudo command but I still have the exact same problem with my computer and display sleeping, and the computer wakes up but not the monitor. I wonder if it has anything to do with my Radeon 4870 video card. Everything else works perfect for the video card without any modified kexts or anything right from the 10.5.7 update (unlike 10.5.6). If anyone out there is using an ATI 4000 series card and has sleep working please let me know HOW in the heck you got it done. As far as my other settings in the "pmset -g" command, sleep, displaysleep, powerbutton, and disksleep are all set to "0". "hibernatemode" is set to 2, and "ttyskeepawake" is at 1. Does that sound about right to everyone or is there anything that should be changed there, or that is different in your settings. Thanks

 

my displaysleep is 15 if that can help. I'm not running an ATI.

[joeblough]

drat...I too did this sudo command but I still have the exact same problem with my computer and display sleeping, and the computer wakes up but not the monitor. I wonder if it has anything to do with my Radeon 4870 video card. Everything else works perfect for the video card without any modified kexts or anything right from the 10.5.7 update (unlike 10.5.6). If anyone out there is using an ATI 4000 series card and has sleep working please let me know HOW in the heck you got it done. As far as my other settings in the "pmset -g" command, sleep, displaysleep, powerbutton, and disksleep are all set to "0". "hibernatemode" is set to 2, and "ttyskeepawake" is at 1. Does that sound about right to everyone or is there anything that should be changed there, or that is different in your settings. Thanks

 

usually when this happens its not that the monitor has not woken up, its that the computer has crashed coming out of sleep. in my case the problem was the bios, which was buggy. i guess you can tell if it's just the monitor by pressing the sleep key on your keyboard followed by "s". if the computer goes back to sleep then in fact it did come out of sleep okay but the video was just dead.

 

if you suspect the video card then you need to get another well-supported video card and see if you can get it to wake okay. something like a geforce 8600 is a good one to try.

[digital_dreamer]

<snip>

i wonder if the right way to do this is to use an older script, but install 10.5.0 and then immediately apply 10.5.6, and then continue the script, making sure to install a custom kernel. then once its up, run the new script but instead of running the installer, just run the 10.5.7 updater (the delta should be fine at that point, but i usually use the combo).

That's the way I've always done it Joe - install from retail, then immediately install combo update. Like you, I feel more comfortable with the combo updates and it has always made the install a two-step process - Retail, plus combo.

 

regards,

MAJ

[joeblough]

That's the way I've always done it Joe - install from retail, then immediately install combo update. Like you, I feel more comfortable with the combo updates and it has always made the install a two-step process - Retail, plus combo.

 

regards,

MAJ

 

sure, but what im getting at is, does it work to apply the 10.5.7 combo update against 10.5.0 before booting the first time? because although people who upgraded from 10.5.6 to 10.5.7 got a double-reboot, this did not happen for me. i'm wondering if because i never booted 10.5.0 and so never did the first-time setup that the 10.5.7 update did not apply properly.

 

the only problem is that your newest script isnt completely compatible with this flow - it looks like it no longer contains the 10.5.7 kernel + system.kext that we used to use with 10.5.6. but i guess the voodoo kernel that's included is good enough to get going.

 

last time around i did what you are saying above - 10.5.0 immediately followed by 10.5.6 combo, and then finished out your script (v3.4 i think). i have another drive with 10.5.6 up and running which i was intending to run the 10.5.7 combo updater against, but before i did that i wanted to make sure i could get 10.5.7 going with your newest script 1) in case something goes wrong with the update on the 10.5.6 disk, and 2) to see if 10.5.7 was stable. so far it does seem stable though after sleeping and waking it overnight with a script, it died on the last wake. i dont know how many times it actually woke up correctly because my script is stupid. i might try that again overnight and keep track of the successful wakes.

[digital_dreamer]

sure, but what im getting at is, does it work to apply the 10.5.7 combo update against 10.5.0 before booting the first time? because although people who upgraded from 10.5.6 to 10.5.7 got a double-reboot, this did not happen for me. i'm wondering if because i never booted 10.5.0 and so never did the first-time setup that the 10.5.7 update did not apply properly.

I think the double-boot comes from applying the update on a live system. Yes, I know what you're saying and I obviously didn't make myself clear is that since I'm installing from another partition, I've always immediately installed the combo without booting into the new install first (after the 10.5.0 install). Has always worked fine. I know in the past I wondered if that would actually work or make a confusing mess of the system, like some things need to be initialized and run before receiving an update, but it has never been an issue. It's like installing straight from a 10.5.6 Retail disk.

I think Apple planned for this type of process, knowing that some systems may not even run on the original retail disk and need an immediate update before booting. On the other hand, Apple could have made this process fail and expected us to buy a newer Retail disk to work with.

 

the only problem is that your newest script isnt completely compatible with this flow - it looks like it no longer contains the 10.5.7 kernel + system.kext that we used to use with 10.5.6. but i guess the voodoo kernel that's included is good enough to get going.

Okay. I can put the vanilla 9.7.0 back in there for those still using 10.5.6. I understand. I think I jumped ahead of myself there and pulled it. I just didn't want people to see the vanilla kernel/System.kext and feel inclined to install them on a vanilla 10.5.7 system. You know how that goes - imply something could be used a certain way and, before you know it, everyone is doing it.

 

last time around i did what you are saying above - 10.5.0 immediately followed by 10.5.6 combo, and then finished out your script (v3.4 i think). i have another drive with 10.5.6 up and running which i was intending to run the 10.5.7 combo updater against, but before i did that i wanted to make sure i could get 10.5.7 going with your newest script 1) in case something goes wrong with the update on the 10.5.6 disk, and 2) to see if 10.5.7 was stable. so far it does seem stable though after sleeping and waking it overnight with a script, it died on the last wake. i dont know how many times it actually woke up correctly because my script is stupid. i might try that again overnight and keep track of the successful wakes.

I haven't really tested sleep enough myself, either. But, I wouldn't be surprised to get a failed-wake. If there's anything unpredictable, it's usually sleep. The biggest complaint I hear is just as you described - it works sometimes/most of the time, but certainly not all of the time.

 

regards,

MAJ

[mattrb]

Okay. I can put the vanilla 9.7.0 back in there for those still using 10.5.6. I understand. I think I jumped ahead of myself there and pulled it. I just didn't want people to see the vanilla kernel/System.kext and feel inclined to install them on a vanilla 10.5.7 system. You know how that goes - imply something could be used a certain way and, before you know it, everyone is doing it.

 

I believe this is the correct route, fellow sage, people aren't going to be using 10.5.6 anymore now that sound works without using headphone-only (until someone hacks it for SPDIF for 889a) voodoohda. 9.7 kernel would not be needed because it was temporary for the time being, now that 10.5.7 is out, it should definitely be omitted from your kext package.