[GUIDE] Scripted Yosemite/Mavericks Install on Gigabyte Mobos

[Raven-PT]

Hi guys,

 

This may be a dumb question, but I'm a bit confused.

 

I want to install OS X on my target system through a USB drive, so what should I do first ?

 

METHOD #2: Install via Boot Drive

 

or

 

USING THE SCRIPT

 

???

 

Thank you in advance.

[maiatico6]

Hey guys sorry to bump myself, but its been weeks with no reply.

 

Lately i've been having crashing issues with my Adobe CS5 master suite. I think this could be because I am running snow leopard on 32-bit and all the CS5 programs only run in 64. (I could be completely wrong about this, but it seems a likely culrpit)

 

Normally I am just thrown a beach ball and I have to force quit my applications at various times of using them.

 

I am using Cruisar's minimal kext method and have tried making the switch to running SL on a 64 bit kernel but cannot seem to get it to boot. In 64 bit I can hear the OS runing as when I press volume buttons or any other input I can hear the OS functioning, but I only can see black.

 

Anyone know a way to make my video function in SL 64 bit?

 

As a freelance and student filmmaker this is becoming problematic for my workflow and I'd like to resolve this issue as fast as possible,

 

Thanks for any help at all!

 

EDIT: I haven't updated my signature yet, but I am running 10.6.4, not .3

[digital_dreamer]

Hi guys,

 

This may be a dumb question, but I'm a bit confused.

 

I want to install OS X on my target system through a USB drive, so what should I do first ?

 

METHOD #2: Install via Boot Drive

 

or

 

USING THE SCRIPT

 

???

 

Thank you in advance.

All install methods, including installing via boot drive, will require using the script at some point. They are not mutually exclusive. The boot drive method requires one to use the script to create a boot drive.

 

Hey guys sorry to bump myself, but its been weeks with no reply.

 

Lately i've been having crashing issues with my Adobe CS5 master suite. I think this could be because I am running snow leopard on 32-bit and all the CS5 programs only run in 64. (I could be completely wrong about this, but it seems a likely culrpit)

 

Normally I am just thrown a beach ball and I have to force quit my applications at various times of using them.

 

I am using Cruisar's minimal kext method and have tried making the switch to running SL on a 64 bit kernel but cannot seem to get it to boot. In 64 bit I can hear the OS runing as when I press volume buttons or any other input I can hear the OS functioning, but I only can see black.

 

Anyone know a way to make my video function in SL 64 bit?

 

As a freelance and student filmmaker this is becoming problematic for my workflow and I'd like to resolve this issue as fast as possible,

 

Thanks for any help at all!

 

EDIT: I haven't updated my signature yet, but I am running 10.6.4, not .3

1.

64-bit applications will run as 64-bit, regardless of how the kernel was booted. So, if you are booted as 32-bit, you can still run 64-bit applications without any limitations.

 

2.

When you mention "crashing issues" with Adobe software, are you referring to kernel panics, the Adobe applications unexpectedly quitting, or the beachball symptoms you describe later?

If beachball, the system is waiting for a response and not receiving one. The software will not continue until the response is provided, hence the beachball. Most issues with the beachball is hard drive related. What type of drives are you using and what is used for the scratch disk, in the case of Photoshop?

I suspect in your case, it may be GPU related, as the graphics software is waiting for the GPU to process some Quartz graphics and it never completes because QE/CI is not (fully?) enabled on your system. Just a guess. Does this beachball happen after/during specific procedures, like using a photoshop filter?

Do any Adobe and Apple discussion forums mention issues with CS5 and OS X 10.6.4? Or, is the issue specific to our hackintosh build? I want to make sure the issue is not a commonly reported issue with CS5 and 10.6.4, and unique to the hackintosh community.

 

3.

As for the black screen in 64-bit mode, what method are you using to enable your Nvidia 260 GPU in 32-bit/64-bit? EFI strings? Any updated Nvidia drivers? If you are attempting to use an injector (kext) that is not 64-bit ready, then the black screen would be the result.

 

best regards,

MAJ

[maiatico6]

1.

64-bit applications will run as 64-bit, regardless of how the kernel was booted. So, if you are booted as 32-bit, you can still run 64-bit applications without any limitations.

 

2.

When you mention "crashing issues" with Adobe software, are you referring to kernel panics, the Adobe applications unexpectedly quitting, or the beachball symptoms you describe later?

If beachball, the system is waiting for a response and not receiving one. The software will not continue until the response is provided, hence the beachball. Most issues with the beachball is hard drive related. What type of drives are you using and what is used for the scratch disk, in the case of Photoshop?

I suspect in your case, it may be GPU related, as the graphics software is waiting for the GPU to process some Quartz graphics and it never completes because QE/CI is not (fully?) enabled on your system. Just a guess. Does this beachball happen after/during specific procedures, like using a photoshop filter?

Do any Adobe and Apple discussion forums mention issues with CS5 and OS X 10.6.4? Or, is the issue specific to our hackintosh build? I want to make sure the issue is not a commonly reported issue with CS5 and 10.6.4, and unique to the hackintosh community.

 

3.

As for the black screen in 64-bit mode, what method are you using to enable your Nvidia 260 GPU in 32-bit/64-bit? EFI strings? Any updated Nvidia drivers? If you are attempting to use an injector (kext) that is not 64-bit ready, then the black screen would be the result.

 

best regards,

MAJ

 

Hey MAJ,

Thanks for the clarification about 64bit software. I wasn't entirely sure what effects the a kernel running in a different bit than the program would have.

 

As far as the crashing goes, I do not know whether this is a known issue with CS5 and OSX 10.6.4, but after some searching I haven't found any resounding or widespread problems reported.

The actual problems I've been experiencing are a combination of crashing and the beach ball. Initially I was getting beach balled very frequently but after posting my problems on here I discovered it was my windows 7 harddrive that was dying and when it would be probed it would stall. While I do get beach balled every now and then, generally when its processor intensive tasks, the true problem has been with adobe quitting on me mid workflow.

I'll try setting a alternate scratch disc and see what effect that has. Also, how do I check whether QE/CI is enabled fully or not?

 

For my graphics card currently I don't have a kext or an EFI string to make it boot 32. I have the standard nvidia gtx drivers for mac installed and its worked without need for anything else, unless I'm oblivious to something in my extra folder. Would creating an EFI string fix my issues and enable the card?

 

Thanks again for the rapid and helpful reply!

[luh3417]

Hi MAJ, I notice you mention SSD above. SSD is a bit involved even without a Macintosh. Most SSD support TRIM now, but Apple does not support TRIM. Apple will probably refresh the MacBook Air line on 20 October 2010 and probably also release 10.6.5. Even though Apple sells the Air with an SSD option, it doesn't appear (and I wish I were wrong) from developer builds that they will ship TRIM. Windows 7 has TRIM and all I can guess is that the perfect master Steve Jobs is holding something back for 10.7, at which point he'll say he invented TRIM as well as the smart phone...

 

I am using an SSD on my GA-EX58-UD5, and I often get unexplainable slowdowns. Nothing in the log or the activity monitor. So now I suspect my SSD. I think further research on SSD is in order. To understand the need for TRIM, and what goes astray in an OS like OS X that fails to support it.

 

For example, consider this interesting question: why do most SSD come in 50 or 60, and not 64GB capacities? A friend tells me the full 64GB it there, but the missing space is used for TRIM. Now, my OS X SSD boot disk is formatted out to a full 64GB so I may be creating problems for myself. Or maybe just using the full capacity is all you can do if OS X fails to offer TRIM. Anyway, some food for thought.

[RickyTan]

Updates on a variety of topics

 

EX58-UD5 NIC Kexts (or most Realtek NICs)

The EX58-UD5 uses two Realtek 8111D chips for ethernet. Bit Shoveler was gracious enough to provide us with a modified RealtekR1000SL.kext which most of us have been using. However, it does not properly detect the physical link speed and had to be manually specified, especially when using GigE.

 

Lnx2Mac has realeased a new kext based of the open source Realtek linux drivers and integrated changes from RealtekR1000SL into a new driver. This driver supports automatic link speed negotiation. Bonjour works fully. Available here:

 

http://lnx2mac.blogspot.com/p/realtekrtl81xx-osx-driver.html

 

The only downside is that right now the driver is in beta and generates lots of debugging log messages in kernel.log. Performance is about even with RealtekR1000SL.kext. Be sure to remove any old networking kexts first before using the new one.

 

 

Problems with EFI Partition Permissions and Kext Cache

I have been using DigitalDreamer's excellent script to manage my EFI install. I have been having issues installing and rebuilding my EFI kext cache. I determined the problems were due to the EFI partition being mounted with the "noowners" options.

 

This meant I could not change the owner of the kexts to root:wheel. Because ownership was incorrect kextcache would not create my mkext in Extra/Extensions. The solution was to remount with the following command, which included the "owners" option:

 

mount -u -o owners /Volumes/EFI

 

UPDATED 2010-11-15

Courtesy of MAJ, a better option is to use this command to enable owners for the EFI partition:

 

diskutil enableOwnership /Volumes/EFI

 

 

Migrating to ATI Radeon HD 5770

I just purchased a 5770 and wanted to get it up and running. It is the standard GPU on the latest Mac Pros. For a no-hassles install, do the following:

1. Update to 10.6.4
   
2. Install the Snow Leopard Graphics Update
   
3. Update your /Volumes/EFI/boot with Kabyl's new Chameleon build with ATI 5xxx support from this thread:
   http://www.insanelymac.com/forum/index.php?showtopic=231768
   
4. set GraphicsEnabler to "yes" in your com.apple.boot.plist.
   

 

Now I have a 5770 with full CI/QE. Works in FCP and Starcraft 2 like a champ. Performance in Windows still outshines performance in OS X, but that seems to be par for the course these days.

 

Kabyl is supposed to be merging these changes into the Chameleon source trunk sometime soon. The next release should have this code integrated.

 

FYI: I used the GigaByte GV-R577UD-1GD GPU available from Newegg.com for $139:

http://www.newegg.com/Product/Product.aspx...N82E16814125309

 

Both DVIs are working with HDCP displays. Haven't tested DisplayPort or HDMI yet.

 

 

JMicron eSATA Hot Swap

I setup my JMicron SATA ports to use the eSATA adapters that came with my EX58-UD5. However, hot swap was not working with them, I had to reboot to recognize a new drive. Then I found JMicron36xeSATA.kext which allows for hot swapping eSATA drives:

 

http://tonymacx86.blogspot.com/2010/03/jmi...ta-support.html

 

They will show up with orange icons. DO NOT use any fix for this unless it is manually changing the drive icon. Most fixes actually disable hot swapping, which then makes them appear as an internal drive. This kext is great for external hard drives and seems to be very reliable.

 

 

I just wanted to give a little update on some new options for us EX58-UD5 users. Hopefully others can benefit from what I have learned.

[digital_dreamer]

<snip>

Also, how do I check whether QE/CI is enabled fully or not?

 

For my graphics card currently I don't have a kext or an EFI string to make it boot 32. I have the standard nvidia gtx drivers for mac installed and its worked without need for anything else, unless I'm oblivious to something in my extra folder. Would creating an EFI string fix my issues and enable the card?

QE/CI can be determined by having a translucent menubar, with the desktop picture somewhat visible behind it. Also, in Dashboard, you will get "ripples" when you drop a new widget on the desktop.

I suppose that if you do have QE/CI without the use of a injector or EFI string, the bootloader is doing it for you. Which bootloader are you using?

 

Hi MAJ, I notice you mention SSD above. SSD is a bit involved even without a Macintosh. Most SSD support TRIM now, but Apple does not support TRIM. Apple will probably refresh the MacBook Air line on 20 October 2010 and probably also release 10.6.5. Even though Apple sells the Air with an SSD option, it doesn't appear (and I wish I were wrong) from developer builds that they will ship TRIM. Windows 7 has TRIM and all I can guess is that the perfect master Steve Jobs is holding something back for 10.7, at which point he'll say he invented TRIM as well as the smart phone...

 

I am using an SSD on my GA-EX58-UD5, and I often get unexplainable slowdowns. Nothing in the log or the activity monitor. So now I suspect my SSD. I think further research on SSD is in order. To understand the need for TRIM, and what goes astray in an OS like OS X that fails to support it.

 

For example, consider this interesting question: why do most SSD come in 50 or 60, and not 64GB capacities? A friend tells me the full 64GB it there, but the missing space is used for TRIM. Now, my OS X SSD boot disk is formatted out to a full 64GB so I may be creating problems for myself. Or maybe just using the full capacity is all you can do if OS X fails to offer TRIM. Anyway, some food for thought.

Thanks for that info. I understand that some kind of garbage collection can take care of that wear-leveling issue with SSDs. I know the Sandforce controllers do the garbage collection and all, so the OS doesn't need to worry about it. I like to think Apple has a implementation ready to incorporate in a update, but is not releasing it until it knows it to be robust. I really hope we don't have to wait until 10.7, as I personally don't expect 10.7 out until at least another year. I mean the developers haven't even seen any signs of it.

 

Updates on a variety of topics

<SNIP>

I just wanted to give a little update on some new options for us EX58-UD5 users. Hopefully others can benefit from what I have learned.

Thanks for your input!

 

RE: New LAN driivers

I've been using Lnx2Mac's Realtek RTL81xx LAN driver for about a month now and have experienced no issues. In fact, I have it installed in /Extra in the IONetworkingFamily.kext, instead of S/L/E. It requires that the OSBundleRequired string be changed to "Root," of course, for the boot cache to build.

I believe it was based on Chucko's RealtekR1000SL.kext, so that explains why the performance is equivalent. It is faster than Realtek latest release, too. But, Realtek's 2.0.5 release breaks after wake from sleep, so it's a no-go for most of us.

 

RE: EFI permissions

You're not the first to report issues with the boot cache issues on the EFI partition. The other user was using a SSD and experienced all sorts of EFI permission issues, so I dismissed it as an anomaly related to that particular SSD. But, your case brings some new insight here. Thanks for the tip. I don't know why I never ran into this issue myself. Do you have any ideas regarding that and what scenarios make this an issue?

The changes as you suggested will be present in the next release, however.

 

RE: Kabyl's new bootloader with 4xxx and 5xxx GPU detection

I've been watching Kabyl's thread with much interest. Once he gets all the beta testers responding with the correct framebuffers, things should move along quickly. Can't wait until he merges his changes with the rest. I'm currently using revision 582 of RC5 and compile updates somewhat regularly.

 

RE: JMicron eSATA hot-swapping

Thanks for that link. I could use that for sure, as I have a eSATA connection to the JMicron controller, but must make sure I boot with it on to mount.

 

Thanks, again, RickyTan.

 

My next release will support the /Extra/Extensions setup for those wishing to go this route and not worry about boot caches in /Extra. One will be able to switch from the boot cache to /Extensions and back at any time.

 

MAJ

[ronsega]

Nice! Been waiting for something like this. I've usually just put my hackintosh in sleep mode then waking it up rather than the reboot to recognize my eSATA drives but will try these out! How well does it work when you go into sleep mode and wake up from sleep mode? I've had to unmount the eSATA drives first then turn them off after its gone into sleep mode.

 

 

Updates on a variety of topics

 

JMicron eSATA Hot Swap

I setup my JMicron SATA ports to use the eSATA adapters that came with my EX58-UD5. However, hot swap was not working with them, I had to reboot to recognize a new drive. Then I found JMicron36xeSATA.kext which allows for hot swapping eSATA drives:

 

http://tonymacx86.blogspot.com/2010/03/jmi...ta-support.html

 

They will show up with orange icons. DO NOT use any fix for this unless it is manually changing the drive icon. Most fixes actually disable hot swapping, which then makes them appear as an internal drive. This kext is great for external hard drives and seems to be very reliable.

[luh3417]

Thanks for that info. I understand that some kind of garbage collection can take care of that wear-leveling issue with SSDs. I know the Sandforce controllers do the garbage collection and all, so the OS doesn't need to worry about it. I like to think Apple has a implementation ready to incorporate in a update, but is not releasing it until it knows it to be robust. I really hope we don't have to wait until 10.7, as I personally don't expect 10.7 out until at least another year. I mean the developers haven't even seen any signs of it.

Sandforce controllers (probably the 1200 more than the 1500) are great, but the rest of that sentence needs clarification. The controller needs to know which sectors are available in order to perform garbage collection. The controller does not know about the different file systems that may be on the disk. The OS does need to worry about this. TRIM is how the OS to tells the SSD controller which sectors it can mess with. No TRIM, no garbage collection (unless you reformat), adds up to continually degrading performance.

 

It seems unconscionable that Apple has been selling MacBook Air's with SSDs in them for nearly 2 years without TRIM. I expect that within 24 hours we'll see 10.6.5, but we won't see TRIM (no developer has mentioned it, and surely it requires testing). I would be happy to be proved wrong on the TRIM part. We'll soon know.

 

---

 

I'm checking the press on Apple's announcement today, but I see no mention of 10.6.5 nor TRIM. Disappointing that the new MacBook Airs (and the new Mac mini, which was an aborted version of the iTV) use 2-generation-old Core 2 CPUs. I guess that's why we hackintosh. The new MacBook Airs (with the old Core 2 CPUs) have SSD on their motherboards if I'm reading it correctly. I don't understand how Apple thinks their SSD (also an option on the old Air) are immune to TRIM. Maybe their plan is, your $1000 or $2000 Air is going to have a dead non-replaceable battery and a 4-generation old CPU in about 2 years, so it won't matter that the SSD is slow, since the whole unit will be ready for the trash at that point anyway.

 

---

 

OK I talked to a friend who works for Transcend. I haven't seen a real clear exposition of this, but the dirty little secret of ALL flash is that it all requires extra memory for garbage collection. They hide it from you, and some controllers do a better job than others, but its required. Again this correlates with the 80GB Intel SSDs or the 50GB OCZ (and other OEM) sandforce SSDs: they have extra hidden capacity. I believe if we pry open the new MacBook Air, or most SSD, and add up the bits on the chips, we'll also find that extra capacity, that is hidden from us, and used for garbage collection. Not sure about the old MacBook Airs with SSD, what capacity SSD did they have?

 

The big mistake I made was to use a jMicron-controller 64GB Transcend drive. Any SSD with a power-of-2 capacity is asking for trouble, or at least, requires an expert user. This drive gives me no margin for garbage collection. What I should have done was format only 40 or 50GB - the drive does know which parts are unformatted (though I wonder if thats true for GUID!?). Instead, by foolishly formatting it to full capacity, and thinking that Apple would implement TRIM, the drive has been slowing down to the point where its like getting root canal. Even with TRIM, either the drive should ship with extra capacity baked in, or I should format it that way. Still, a question lingers: does the drive see the unformatted capacity when using GUID as opposed to MBR? Needs to be nailed down. Do 50 or 80 GB smart sized SSD actually need TRIM or extra unformatted space?

 

What if I shrink my partition, which GUID allows... will the SSD pick up on it? Well, apparently I can't try...

http://www.macworld.com/article/55274/2007...geekfactor.html

diskutil resizeVolume disk0s2 limits
For device disk0s2 i7b:
	Current size:  63.8 GB (63761768448 Bytes)
	Minimum size:  63.8 GB (63761768448 Bytes)
	Maximum size:  63.8 GB (63761768448 Bytes)

 

OK people there it is. I don't have all the answers but I'm asking the right questions. There are pages upon pages of discussion in the OCZ forums. which is where the real beta testing for consumer-grade SSDs began. It was the highest up OCZ techs that worked with the controller manufacturers and finally figured out that the drives need more cache to make sure they last longer.

 

Yes, I realize few people care about this, but just wanted to write this up for MAJ or anyone considering SSD.

[luh3417]

MAJ,

 

I'm using script version 4.56. From 10.6.4 hackintosh to new SATA volume.

 

Says "1 of 9 kexts in /Extra failed to make it into cache! See logs for details."

 

Logs say...

ATY_Init.kext does not match OSBundleRequired conditions; omitting from mkext.

ATY_Init.kext doesn't support architecture 'x86_64'; omitting from mkext.

 

You mentioned you don't use ATY_Init so maybe you never saw this situation.

I believe I'm doing a fairly stock install and following the guide in the first post.

 

Now I'm trying Combo Build. The Combo Build options is not explained in this thread (at least, search didn't find it when I downloaded the entire thread). I tried Combo Build and the error went away.

 

-----

 

7300GT worked like a charm including QE.

 

I have to say pmVersion=0 in the Chameleon boot. I think I can just add that string to the boot.plist.

 

I get a couple dozen complaints during boot like this... "Refusing new kext com.apple.iokit.IOPCIFamily, v2.6: a loaded copy with a different executable UUID is already present." if anyone has any ideas. This started after the combo update from 10.6.0 to 10.6.4. Also did a Migration Assistant later.

[digital_dreamer]

MAJ,

 

I'm using script version 4.56. From 10.6.4 hackintosh to new SATA volume.

 

Says "1 of 9 kexts in /Extra failed to make it into cache! See logs for details."

 

Logs say...

ATY_Init.kext does not match OSBundleRequired conditions; omitting from mkext.

ATY_Init.kext doesn't support architecture 'x86_64'; omitting from mkext.

 

You mentioned you don't use ATY_Init so maybe you never saw this situation.

I believe I'm doing a fairly stock install and following the guide in the first post.

 

Now I'm trying Combo Build. The Combo Build options is not explained in this thread (at least, search didn't find it when I downloaded the entire thread). I tried Combo Build and the error went away.

 

-----

 

7300GT worked like a charm including QE.

 

I have to say pmVersion=0 in the Chameleon boot. I think I can just add that string to the boot.plist.

 

I get a couple dozen complaints during boot like this... "Refusing new kext com.apple.iokit.IOPCIFamily, v2.6: a loaded copy with a different executable UUID is already present." if anyone has any ideas. This started after the combo update from 10.6.0 to 10.6.4. Also did a Migration Assistant later.

luh3417,

I use the ATY_Init.kext, myself, but it has to be in /System to work properly. If in /Extra, the boot cache won't get built because that kext has a "OSBundleRequired" string of "Safe-Boot" instead of "Root." (Only kexts with the "OSBundleRequired" string of "Root" are included in a boot cache.) But, even if one were to change the "OSBundleRequired" string to "Root" for this ATY_Init.kext, it will cause a hang on boot (at least that was the case in my setup). So, keeping that kext in /Extra is useless - it just wasn't designed to be included in a boot cache, hence the error.

 

Now, the Combo build builds a cache for /Extra that includes all kexts in /Extra and /System. It'll typically be right under 10MB in size.

The script doesn't report on these errors in the combo build, because many kexts in /System don't have a "OSBundleRequired" string of "Root" and that would lead to a lot of false positives. Not all kexts in /System go into a boot cache, anyway. In other words, error reporting for kexts in /System not making into the cache would be too numerous, so the script ignores those so-called "errors." I only log such errors for /Extra, because if it doesn't get in the boot cache, it won't get loaded at all. If it is in /System and it doesn't get in a boot cache, it can still get loaded after boot. It just depends on the "OSBundleRequired" flag.

 

This is why, in a future script update, I'm offering the option to use the /Extra/Extensions setup. In this setup, there is no boot cache. The bootloader looks into each kext and loads the appropriate data. Kexts in /Extra/Extensions can still get loaded by the bootloader after boot, even if it's not suitable for a boot cache.

This setup may be desirable for some who have modified kexts that don't get included in a boot cache, but want it to be in /Extra, not /System. (I prefer just changing the "OSBundleRequired" flag in the plist to "Root", myself.) However, this still won't get around some dependencies, like those required by the ATY_Init.kext. Complicated world, isn't it?

 

BTW, the Super Combo build is the same as the regular combo, but includes all relevant kext data, not just the booting info.

(I have thought about offering a "help" menu. Something like type a "?" at any prompt and it'll explain the options and what it does.)

 

That boot log complaint sounds like a conflict between two identical kexts, but of different versions - perhaps one loading in /Extra and the other in /System.

 

Wish I had more info on the SSD situation. Heh. I wonder if one could boot into a Windows install and let it set for a hour or so and see if it'll take care of the SSD. But, it may only work on Windows formatted MBR. And, even then, wouldn't recognize the HFS+ file system.

 

MAJ

[luh3417]

luh3417,

I use the ATY_Init.kext, myself, but it has to be in /System to work properly. If in /Extra, the boot cache won't get built because that kext has a "OSBundleRequired" string of "Safe-Boot" instead of "Root." (Only kexts with the "OSBundleRequired" string of "Root" are included in a boot cache.) But, even if one were to change the "OSBundleRequired" string to "Root" for this ATY_Init.kext, it will cause a hang on boot (at least that was the case in my setup). So, keeping that kext in /Extra is useless - it just wasn't designed to be included in a boot cache, hence the error.

 

In the first post, you could edit the phrase "Note that some kexts just do not work in /Extra and must be installed into /System (or S/L/E), instead. These include the current selection of audio and networking kexts…" to add "and the graphics kext ATY_Init (if you need it)".

 

Is my reasoning correct that, since my ATY_Init is still in /Extra, and that despite that I have the 7300GT running with QE, that actually I don't need ATY_Init anyway. But wait, I do see this in my console "ATY_Init: The 393 card already has keys set" so I must be using it. I have GraphicsEnabler set in my boot.plist if that matters. Well, I guess its better to have "it works and I don't know why" than "it don't work and I don't know why".

 

I feel (as everyone does) that I did a completely vanilla install. But I seem to be the only one getting errors like the following...

 

Refusing new kext com.apple.iokit.IOPCIFamily, v2.6: a loaded copy with a different executable UUID is already present.

com.apple.iokit.IOPCIFamily, v2.6
com.apple.iokit.IOACPIFamily, v1.3
(these repeat over a dozen times… they do exist in /S/L/E but I don't see them in /Extra)

these similar errors occur only once...
com.apple.kernel.iokit, v7.9.9:
com.apple.kernel.libkern, v7.9.9
com.apple.kernel.mach, v7.9.9

 

Above I've dumped all I know about SSD. The OCZ forums might be a place to research further. I think I now know enough to get them to work properly on the Mac situation. Its just I'm gun shy now after all the pain of the slowdown. I'd say if you go with a non-power-of-2 SSD like the latest Intels or the latest Vertex2, a 50 or 80 or 100GB unit, it will be okay on the Mac. I would avoid the 60 or 64 GB units, since OS X does not understand TRIM.

[jmag999]

I am using Cruisar's fakesmc only method but am having some problems getting ethernet working.

 

Here is my kernel log:

 

MAC Framework successfully initialized
using 16384 buffer headers and 4096 cluster IO buffer headers
IOAPIC: Version 0x20 Vectors 64:87
ACPI: System State [S0 S3 S4 S5] (S3)
AppleIntelCPUPowerManagement: Turbo Ratios 1112
AppleIntelCPUPowerManagement: initialization complete
RTC: Only single RAM bank (128 bytes)
mbinit: done (96 MB memory set for mbuf pool)
From path: "uuid", 
Waiting for boot volume with UUID 5AFF6C8E-99EC-3714-8C7C-83595A713656
Waiting on <dict ID="0"><key>IOProviderClass</key><string ID="1">IOResources</string><key>IOResourceMatch</key><string ID="2">boot-uuid-media</string></dict>
netkas presents fakesmc, a kext which emulates smc device
com.apple.AppleFSCompressionTypeZlib kmod start
Kext com.chucko.driver.RealtekR1000 - library kext com.apple.iokit.IONetworkingFamily not found.
Can't load kext com.chucko.driver.RealtekR1000 - failed to resolve library dependencies.
Kext com.chucko.driver.RealtekR1000 failed to load (0xdc00800e).
Failed to load kext com.chucko.driver.RealtekR1000 (error 0xdc00800e).
Couldn't alloc class "com_chucko_RealtekR1000"
Not loading kext com.chucko.driver.RealtekR1000 - not found and kextd not available in early boot.
com.apple.AppleFSCompressionTypeZlib load succeeded
AppleIntelCPUPowerManagementClient: ready
FireWire runtime power conservation disabled. (2)
Got boot device = IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/SATA@1F,2/AppleAHCI/PRT0@0/IOAHCIDevice@0/AppleAHCIDiskDriver/IOAHCIBlockStorageDevice/IOBlockStorageDriver/WDC WD10EADS-00M2B0 Media/IOGUIDPartitionScheme/Macintosh HD@2
BSD root: disk0s2, major 14, minor 2
Kernel is LP64
FireWire (OHCI) TI ID 8024 PCI now active, GUID 00cb9cdc0000241d; max speed s400.
USBMSC Identifier (non-unique): 12345678901234567890 0x90c 0x6000 0x100
systemShutdown false
Previous Shutdown Cause: 0
NVDANV50HAL loaded and registered.
FakeSMC: key info not found MSDS, length - 6
identified as RTL8168D/8111D
identified as RTL8168D/8111D
Waiting for DSMOS...
DSMOS has arrived
com_chucko_RealtekR1000: Ethernet address 00:24:1d:c6:bd:fb
com_chucko_RealtekR1000: Ethernet address 00:24:1d:c6:be:0b
FakeSMC: key not found BEMB, length - 1
AppleTyMCEDriver::start coreVIDPID = 0x2c418086 Number of packages = 1 Number of cpus = 8 memory monitor trough MCA
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler

 

There are a bunch of error messages relating to the Realtek kext. I did have ethernet working with the boot disc ([url=&quot;http://www.insanelymac.com/forum/topic/279450-why-insanelymac-does-not-support-tonymacx86/&quot;]#####[/url]) that I was using, however it also has stopped working now. Any idea what I might have done?

[jmag999]

I am using Cruisar's fakesmc only method but am having some problems getting ethernet working.

 

Here is my kernel log:

 

MAC Framework successfully initialized
using 16384 buffer headers and 4096 cluster IO buffer headers
IOAPIC: Version 0x20 Vectors 64:87
ACPI: System State [S0 S3 S4 S5] (S3)
AppleIntelCPUPowerManagement: Turbo Ratios 1112
AppleIntelCPUPowerManagement: initialization complete
RTC: Only single RAM bank (128 bytes)
mbinit: done (96 MB memory set for mbuf pool)
From path: "uuid", 
Waiting for boot volume with UUID 5AFF6C8E-99EC-3714-8C7C-83595A713656
Waiting on <dict ID="0"><key>IOProviderClass</key><string ID="1">IOResources</string><key>IOResourceMatch</key><string ID="2">boot-uuid-media</string></dict>
netkas presents fakesmc, a kext which emulates smc device
com.apple.AppleFSCompressionTypeZlib kmod start
Kext com.chucko.driver.RealtekR1000 - library kext com.apple.iokit.IONetworkingFamily not found.
Can't load kext com.chucko.driver.RealtekR1000 - failed to resolve library dependencies.
Kext com.chucko.driver.RealtekR1000 failed to load (0xdc00800e).
Failed to load kext com.chucko.driver.RealtekR1000 (error 0xdc00800e).
Couldn't alloc class "com_chucko_RealtekR1000"
Not loading kext com.chucko.driver.RealtekR1000 - not found and kextd not available in early boot.
com.apple.AppleFSCompressionTypeZlib load succeeded
AppleIntelCPUPowerManagementClient: ready
FireWire runtime power conservation disabled. (2)
Got boot device = IOService:/AppleACPIPlatformExpert/PCI0@0/AppleACPIPCI/SATA@1F,2/AppleAHCI/PRT0@0/IOAHCIDevice@0/AppleAHCIDiskDriver/IOAHCIBlockStorageDevice/IOBlockStorageDriver/WDC WD10EADS-00M2B0 Media/IOGUIDPartitionScheme/Macintosh HD@2
BSD root: disk0s2, major 14, minor 2
Kernel is LP64
FireWire (OHCI) TI ID 8024 PCI now active, GUID 00cb9cdc0000241d; max speed s400.
USBMSC Identifier (non-unique): 12345678901234567890 0x90c 0x6000 0x100
systemShutdown false
Previous Shutdown Cause: 0
NVDANV50HAL loaded and registered.
FakeSMC: key info not found MSDS, length - 6
identified as RTL8168D/8111D
identified as RTL8168D/8111D
Waiting for DSMOS...
DSMOS has arrived
com_chucko_RealtekR1000: Ethernet address 00:24:1d:c6:bd:fb
com_chucko_RealtekR1000: Ethernet address 00:24:1d:c6:be:0b
FakeSMC: key not found BEMB, length - 1
AppleTyMCEDriver::start coreVIDPID = 0x2c418086 Number of packages = 1 Number of cpus = 8 memory monitor trough MCA
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler
Sound assertion "0 != result" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDAWidget.cpp" at line 3800 goto handler
Sound assertion "0 != widget->setUnsolicited ( true )" failed in "/SourceCache/AppleHDA/AppleHDA-187.4.1/AppleHDA/AppleHDACodecGeneric.cpp" at line 1130 goto handler

 

There are a bunch of error messages relating to the Realtek kext. I did have ethernet working with the boot disc ([url=&quot;http://www.insanelymac.com/forum/topic/279450-why-insanelymac-does-not-support-tonymacx86/&quot;]#####[/url]) that I was using, however it also has stopped working now. Any idea what I might have done?

 

I was able to eliminate the error messages by changing the OSBuild in IONetworkingFamily.kext to root. Now the ethernet port turns green in system preferences and get's an IP from the DHCP server but it still isn't allowing internet traffic. Any ideas?

[luh3417]

My advice is, there's safety in numbers. No offense intended, and correct me if I'm wrong, but I don't see much talk about CruisAR's method lately, or many people using it. I switched from CruisAR to MAJ main method and things seem even smoother. Perhaps someone will advise you, but AFAIK the main MAJ method in post 1 is the mainstream now.

 

MAJ states in the original post "All that's really needed to boot into OS X on this board is a disabler (i.e. Disabler.kext or NullCPUPowerManagement.kext.), a decryptor (i.e. fakesmc.kext or dsmos.kext) and graphics support. If your card is one Apple makes available, then it should work OOTB or with EFI strings. That's it. Everything else are little fixes for hardware reporting, updated device IDs, audio, network, etc. In my case, I also needed the ATY_init.kext for ATI graphics card support, as without it I just got video corruption and couldn't see the desktop." Again, no offense intended, and correct me if I'm wrong, but I think that characterizes the CruiSAR method: a disabler and a decrypter. Most people find no trouble and harm, and in fact benefits, to adding the everything else, and you'll be on a more well trodden path if you go that way.

[CruiSAr]

Again, no offense intended, and correct me if I'm wrong, but I think that characterizes the CruiSAR method: a disabler and a decrypter. Most people find no trouble and harm, and in fact benefits, to adding the everything else, and you'll be on a more well trodden path if you go that way.

None Taken, I also started off using DD's mainstream method. The key here is find what best works for you after you have gotten acquainted with how things work under the hood. For me, the less kexts I have the safer I feel.

 

I am using Cruisar's fakesmc only method but am having some problems getting ethernet working.

 

There are a bunch of error messages relating to the Realtek kext. I did have ethernet working with the boot disc ([url=&quot;http://www.insanelymac.com/forum/topic/279450-why-insanelymac-does-not-support-tonymacx86/&quot;]#####[/url]) that I was using, however it also has stopped working now. Any idea what I might have done?

My guess is the ethernet bug we all experienced when we started installing SL on this board. If ethernet was working with the bootdisk before and is no longer working then I suppose its hardware related.

 

The common fix is to shutdown the computer and devoid the motherboard of any power for about 30 seconds. Either switch off the PSU or pull the power chord out. Restart and let me know how it goes. What version of the Realtek kext are you using? The one in DD's latest script works fine.

[luh3417]

None Taken, I also started off using DD's mainstream method. The key here is find what best works for you after you have gotten acquainted with how things work under the hood. For me, the less kexts I have the safer I feel.

Cool. Wondering what minimum set of kexts you settled upon.

[CruiSAr]

Cool. Wondering what minimum set of kexts you settled upon.

Only 2, fakeSMC and Realtek for network. Will soon be dropping the network kext as soon as I get some cash to order an intel card that works OOB.

[Edkroket]

Possible NVidia 8xxx/9xxx Solution for 10.6.3

 

Once you have his install method setup, it is important to remove any EFI strings for your GPU in your com.apple.boot.plist file. You also need to setup the "GraphicsEnabler" option. I would also recommend using the "EthernetBuiltIn" option as well. Here is the relevant snippet from com.apple.boot.plist:

 

<key>GraphicsEnabler</key>

<string>yes</string>

<key>Kernel</key>

<string>mach_kernel</string>

<key>EthernetBuiltIn</key>

<string>Yes</string>

There is no need to use any older NVidia kexts anywhere, and you do not need to add your device IDs to any of the existing kexts. The magic happens with the CruiSAr's new method using a new bootloader (based of Chameleon RC4) that provides automatic EFI string injection and seems to offer better support for the NVidia GPUs in both 10.6.2 and 10.6.3.

 

Once you have this setup, I can confirm that these setups work:

 

Single DVI monitor (HDCP capable)

Dual DVI monitors (both HDCP capable)

Dual monitors with one DVI monitor (HDCP) and one HDMI Sony 1080p HDTV (HDCP)

Single Sony 1080p HDTV (HDCP)

 

My current video card in an NVidia 9600 GT 512MB with Dual DVI outputs. To get HDMI, I used a DVI --> HDMI converter cable.

 

For anyone not using the Gigabyte EX58-UD5 but experiencing the same problem, I suggest using the GraphicsEnabler option from the AsereBLN_v1.1.9 Bootloader available here:

 

http://github.com/aserebln/Chameleon/downloads

 

Hopefully that helps alleviate some frustration. Please let everyone know if this method does or does not work for you.

 

This works like a charm on my 9600 GT 1Gig! Thanks man!

 

Ed

[MikeBuzzsaw]

is there anything differently i need to do for this install if I were to install mac snow leopard server?

 

 

I keep getitng this kernel panic

[digital_dreamer]

is there anything differently i need to do for this install if I were to install mac snow leopard server?

 

 

I keep getitng this kernel panic

Remove the sleepenabler.kext.

It's not compatible with the installed kernel.

 

You could type pmVersion=0 to disable it at the boot prompt, too. Or, add that as a kernel flag in the boot plist.

 

MAJ

[ѕӎѳѳ₮ңӌ]

None Taken, I also started off using DD's mainstream method. The key here is find what best works for you after you have gotten acquainted with how things work under the hood. For me, the less kexts I have the safer I feel.

 

 

My guess is the ethernet bug we all experienced when we started installing SL on this board. If ethernet was working with the bootdisk before and is no longer working then I suppose its hardware related.

 

The common fix is to shutdown the computer and devoid the motherboard of any power for about 30 seconds. Either switch off the PSU or pull the power chord out. Restart and let me know how it goes. What version of the Realtek kext are you using? The one in DD's latest script works fine.

 

I got fed up of doing this so just eBay'd an Apple USB Ethernet Adapter, works wonders and I never have to worry about this any more. It didn't help that my G5 case mod meant that accessing the CLR CMOS button was very difficult!

[CruiSAr]

I got fed up of doing this so just eBay'd an Apple USB Ethernet Adapter, works wonders and I never have to worry about this any more. It didn't help that my G5 case mod meant that accessing the CLR CMOS button was very difficult!

Yeah that is a way I would go. Will be replacing mine soon. Sent you a pm

 

EDIT:Cant send you a pm...did you buy your mod case or you built it yourself?

[rumz]

10.6.5 is out. Users in other threads are reporting that they've had pretty good success just doing the normal software update without any problems.

 

I have yet to attempt a Snow Leopard build, still on 10.5.7 here-- hoping that I can do the 10.6.4 install from the first page of this thread and do a software update to 10.6.5 (though it sounds like SleepEnabler.kext pmVersion=21 is what people are using that works in 10.6.5).

 

Anyone with a UD5 board try 10.6.5 yet?

[jijiji]

updated fine except for support for gtx 470 (it killed it. i can only see blocks of black and white)