Jump to content

nForce OS X Mavericks (10.9) APP STORE DOWNLOAD INSTALL GUIDE on a Series 6 or 7 nForce chipset / Intel CPU MOBO


verdant
 Share

181 posts in this topic

Recommended Posts

@ Beheaded

 

Your MOBO LAN chipset is the Realtek 8211CL (0x076010de), so the nForceLAN.0.64.5.kext without any Info.plist Device ID insertion should work properly if it loading OK:

 

Supported PCI ids:

0x01c310de, 0x006610de, 0x00d610de, 0x008610de, 0x008c10de, 0x00e610de, 0x00df10de,0x005610de, 0x005710de, 0x003710de, 0x003810de, 0x026810de, 0x026910de, 0x037210de,0x037310de, 0x03e510de, 0x03e610de, 0x03ee10de, 0x03ef10de, 0x045010de, 0x045110de,0x045210de, 0x045310de, 0x054c10de, 0x054d10de, 0x054e10de, 0x054f10de, 0x07dc10de,0x07dd10de, 0x07de10de, 0x07df10de, 0x076010de, 0x076110de, 0x076210de, 0x076310de,0x0ab010de, 0x0ab110de, 0x0ab210de, 0x0ab310de, 0x0d7d10de

 

Do the following:

 

1. Go to /System/Library/Extensions/IONetworkingFamily.kext/Contents/PlugIns/nvenet.kext

 

2. Rename nvenet.kext

 

In Terminal, type

cd /System/Library/Extensions/IONetworkingFamily.kext/Contents/PlugIns
ls
sudo mv nvenet.kext nvenet.kext.bad
ls
sudo touch /System/Library/Extensions/
exit

"ls will list the files in PlugIns"

 

3. The v0.64.5 kext for Leopard should be installed manually into /S/L/E (v0/64.6 is not recommended as many have problems with using it).

 

4. Then reset the ownwership and permissions of /S/L/E/ via Terminal:

sudo chown -R root:wheel /System/Library/Extensions
sudo chmod -R 755 /System/Library/Extensions
sudo touch /System/Library/Extensions
exit

5. Then go to /System/Library/Caches/com.apple.kext.caches/Startup and delete these files within Startup….

 

IOKitPersonalities_i386.ioplist.gz

IOKitPersonalities_x86_64.ioplist.gz

kernelcache

KextPropertyValues_OSBundleHelper_i386.plist.gz

KextPropertyValues_OSBundleHelper_x86_64.plist.gz

 

6. Reboot and check if the nForceLAN.kext is being loaded by booting with -v and also by entering the kextstat command in Terminal.....

@ Fred Coelho

 

Glad to help and that now OSX Mavericks is up and running well for you..... :)

Welcome to the light side....you are enjoying using OS 110.9.x so much compared to Windows 8.1........ ;)

 

Regarding installing kexts from S/L/E in the USB boot pen to the S/L/E of Mavericks, see 4. and 5. above in this post to Beheaded about manually setting kext ownership and permissions......

I did that multiple times. the kext itself is being loaded, it detects it, and I also have it forcing a MAC address (tried it without as well). The sole problem is that the kernel isn't spitting out an interface name for it. Deleted the preferences plist for interface names, even manually set the interface myself for it. It almost seems like a kernel issue, going to try new r7 kernel later today. And yes, nvenet is removed, and the device id is in the Info.plist and permissions have been set multiple times along with running kextwizard.

 

Been working on it for nearly 3 days straight with no avail. Debating recompiling kernel and drivers from source. From what I understand the kext is loading, then requests an interface name from a slot from the kernel by passing a call to it from the InterfaceNamer class of configd. Configd spits out usually 7-8 errors of the same being "Unable to name known/new interface." So something is happening in the middle of the kext and kernel talking to each other.

Link to comment
Share on other sites

@ Beheaded

 

In this reply, based on your first post, I am also assuming that you did the all the following without success:

 

1. Go to System Preferences > Network  and in LHS panel delete all services via [-] button

 

2. Go to /Library/Preferences/SystemConfiguration and delete

 

com.apple.network.eapolclient.configuration.plist
NetworkInterfaces.plist
preferences.plist

 

3. Reboot

 

However, if you did/do generate a new Ethernet service, then via the Service Order list, ensure that your "Display Ethernet" interface is the top of the list .

 

If not, then  given what you have tried and say, it would appear to be kernel rather than kext related.......did you have Ethernet woking on your MOBO with pre Mavericks OS X patched kernel? If so, I am assuming you compared all relevant logs from it with those from OS X Mavericks.....

Link to comment
Share on other sites

@ Beheaded

 

In this reply, based on your first post, I am also assuming that you did the all the following without success:

 

1. Go to System Preferences > Network  and in LHS panel delete all services via [-] button

 

2. Go to /Library/Preferences/SystemConfiguration and delete

 

com.apple.network.eapolclient.configuration.plist

NetworkInterfaces.plist

preferences.plist

 

3. Reboot

 

However, if you did/do generate a new Ethernet service, then via the Service Order list, ensure that your "Display Ethernet" interface is the top of the list .

 

If not, then  given what you have tried and say, it would appear to be kernel rather than kext related.......did you have Ethernet woking on your MOBO with pre Mavericks OS X patched kernel? If so, I am assuming you compared all relevant logs from it with those from OS X Mavericks.....

 

Indeed with no success. I'm leaning more and more to it being kernel related. It can't generate the new Ethernet device because it never gets assigned a name from the kernel via InterfaceNamer. I did have it working in Lion and Snow leopard at one point, this was ages ago though and recently decided to give the hackintosh route on the desktop a go again, I have a Lenovo G780 that's running OSX flawlessly right now (minus the non-use of the switchable graphics, can't win them all i guess). From digging through Darwin source code It seems that once the kext is loaded, it requests a non-used name from the kernel to get assigned to (e.g. en0, en1, wlan0, etc)  via configd and ISServices, but the kernel never spits one out for it. Is there a way to enable more debugging from the kernel itself with re-compiling it? I have the debug log verbosity set in the kext to the highest possible.

Link to comment
Share on other sites

@ Beheaded

 

Check out here and here and here......in conjunction had you thought of re-installing OS X 10.6.8, OS X 10.7.5 and OS X 10.8.5 alongside OS X 10.9.3 at the same time to see what AMD patched kernels and OS X versions work with nForceLAN.kext.....?

Link to comment
Share on other sites

@ Beheaded

 

Check out here and here and here......in conjunction had you thought of re-installing OS X 10.6.8, OS X 10.7.5 and OS X 10.8.5 alongside OS X 10.9.3 at the same time to see what AMD patched kernels and OS X versions work with nForceLAN.kext.....?

Thought about it? Yes. The question is do I REALLY want to do it... LOL. Im going to try out a couple of kernels here tonight to see if there's any difference and if that doesn't work then Ill go the long route, and if that doesn't work then Im just going to hit up Newegg and grab a compatible lan card and call it a day >_<.

 

and Mountain Lion is going to be a no go since im running an Nvidia card anyways.

Link to comment
Share on other sites

I'm experiencing something similar with my system.  The system boots fine and everything.  After following the instructions in the last few posts to make sure that the nForceLan.kext is loaded at the system boot, the kext loads fine as shown by the below lines from the dmesg:

 

nForceLAN: found nVidia MCP61 Ethernet Controller device 10de:03ef with flags 0x0101774a
nForceLAN: Version 0.64.5 starting

 

However, when the system comes up, I have no internet access.  I've attached the full dmesg with this post in case there's anything interesting in it.

 

I can also run the system in a virtual machine I've set up in VirtualBox in Linux with access to the raw partition which provides a different Ethernet controller (in Linux, the lspci command in the virtual machine (also attached) shows it to be an Intel 82543GC Gigabit Ethernet Controller (Copper) (rev 02)).  After startup in the virtual machine, there is no internet access.  (It looks like it should work because in the /System/Library/Extensions/IONetworkingFamily.kext/Contents/PlugIns directory there is a kext called "AppleIntel8254XEthernet.kext" which looks like a plugin that should support the controller.  I also have a Mountain Lion installation that I can run in the same virtual machine which recognizes the controller without problems which leads me to believe that there is something wrong in the Mavericks system.  I'm using the mach_10.9_2_rc7_fixed kernel from a recent post in the AMD kernel development thread with a Mavericks 10.9.3 system.  The system runs perfectly fine, except for the networking problem describred in the last few posts.)

dmesg.06-15-14.txt

Virtual Machine Device List (lspci).txt

Link to comment
Share on other sites

  • 2 weeks later...

Hey guys,

 

I know this is slightly off-topic. I'm a video journalist working for the Wall Street Journal on a story about people building their own Macs from scratch. Basically we want to see who's doing it, what are some of the coolest builds, how challenging is it, and how much can you save?

 

If any of you might be interested in participating, please PM me. Especially if anyone is California-based. Thanks!

 

Best,

Michael

Link to comment
Share on other sites

  • 5 weeks later...

I made a patched IOATAFamily.kext for 10.9.4 that should hopefully speed up boot time when using the AppleNForceATA.kext. I did the same thing ar4er did here, just decreased the timeout. I also had to add a missing header file to get it to build. The original timeout is 3100ms so I made a 100ms one and a 50ms one.

Here is the difference in boot time for me, using a 7200RPM hard drive. Someone using an SSD might notice a bigger difference.

-Stock IOATAFamily: 1:18 to wallpaper

-100ms IOATAFamily: 1:06 to wallpaper

-50ms IOATAFamily: 1:00 to wallpaper

 

Anyone want to test it with an SSD?

 

Here is my source: https://github.com/jamiethemorris/IOATAFamily

 

If anybody has weird problems such as drives unmounting after boot I can increase the timeout a bit.

IOATAFamily_50ms.kext.zip

IOATAFamily_100ms.kext.zip

Link to comment
Share on other sites

When I get a chance I will try it with my 7200 rpm HDDs.

 

What is your SSD boot time with the Stock IOATAFamily.kext?

Link to comment
Share on other sites

When I get a chance I will try it with my 7200 rpm HDDs.

 

What is your SSD boot time with the Stock IOATAFamily.kext?

It's also 1:18, the same as with a HDD, I actually just tested it. BTW this seems to improve boot times on everything, not just NForce boards. I now have it installed on girlfriend's Macbook and on my own MacBook and my Z68 build I'm autopatching it with Clover. I managed to find out how to change the times with a binary patch using Hopper. There are 3 instances of this in the stock IOATAFamily: 81 FB 1C 0C replace them all with this: 81 FB 32 00. That changes the 3100 ms timeouts to 50 ms. If you wanted 100 ms then you would use 81 FB 64 00. Here is the Clover patch for anyone that is interested:

            <dict>
                <key>Comment</key>
                <string>IOATAFamily Faster Boot</string>
                <key>Find</key>
                <data>
                gfscDA==
                </data>
                <key>Name</key>
                <string>IOATAFamily</string>
                <key>Replace</key>
                <data>
                gfsyAA==
                </data>
            </dict>            
Link to comment
Share on other sites

  • 2 weeks later...

@verdant

 

Are you familiar or have you heard of random freezes/reboots on these nforce boards? This is a HTPC/home server so it's on 24/7. It either reboots or freezes about once every week or 2. The reboots mostly happened with Chameleon, I would turn on the TV and see that it had obviously rebooted at some point, I looked in the console log and there were no errors or anything, it just said it had been rebooted normally. Using Clover it seems to lock up rather than reboot. No kernel panic or console messages or anything, the only evidence I can find of a problem are messages like these:

 

8/6/14 8:07:26.599 PM com.apple.time[222]: Interval maximum value is 946100000 seconds (specified value: 9223372036854775807).

repeated over and over again before the freeze.

 

FSB speed seems to be detected correctly... Processor interconnect speed in system profiler seems to be correct at 4.27 GT/s. No TSC syncing messages in the console log or anything.

 

So aside from the com.apple.time messages spamming the console right before the freeze, nothing looks out of the ordinary. Any suggestions?

Link to comment
Share on other sites

@ jamiethemorris

 

Presuming that your Dell XPS 720 MOBO has a nForce 680i chipset, then there are reports of it randomly freezing..........for example, iLeopod stopped using his 680i; see here..........

 

I think that the 680i suffers from temperature instabilities due to its Southbridge and Northbridge running very hot.........even on my 650i I have small fans cooling my Southbridge and Northbridge......which affect the CPU / RAM interacting during SATA data transfers, particularly large data transfers......

 

Also see here and here and here

 

Finally, see the posts starting here between myself and Nick Savage, indicating that setting maxmem=3456 greatly aids stability......also see two posts by djdubs here.....

Link to comment
Share on other sites

@ jamiethemorris

 

Presuming that your Dell XPS 720 MOBO has a nForce 680i chipset, then there are reports of it randomly freezing..........for example, iLeopod stopped using his 680i; see here..........

 

I think that the 680i suffers from temperature instabilities due to its Southbridge and Northbridge running very hot.........even on my 650i I have small fans cooling my Southbridge and Northbridge......which affect the CPU / RAM interacting during SATA data transfers, particularly large data transfers......

 

Also see here and here and here

 

Finally, see the posts starting here between myself and Nick Savage, indicating that setting maxmem=3456 greatly aids stability......also see two posts by djdubs here.....

Thanks verdant, that is very helpful. 

Link to comment
Share on other sites

Glad to have helped.....:)

 

BTW re my DIY chipset cooling, my 650i Northbridge chipset has a Scythe Mini-Kase Ultra 40mm Quiet Cooling fan mounted on its finned heatsink to blow air down on it (running full speed), while the Southbridge chipset has a Rasurbo 40mm Quiet Cooling fan (with its 4 mounting screws partially screwed into the fan frame) mounted on its flat heatsink to blow air down on it (running in Q-Fan mode). These fans are mounted using "blobs" of Butyl rubber adhesive with very good holding power against gravity [but still removable] and vibration damping behaviour, with a maximum service temperature of 93 Celcius.

Link to comment
Share on other sites

Glad to have helped..... :)

 

BTW re my DIY chipset cooling, my 650i Northbridge chipset has a Scythe Mini-Kase Ultra 40mm Quiet Cooling fan mounted on its finned heatsink to blow air down on it (running full speed), while the Southbridge chipset has a Rasurbo 40mm Quiet Cooling fan (with its 4 mounting screws partially screwed into the fan frame) mounted on its flat heatsink to blow air down on it (running in Q-Fan mode). These fans are mounted using "blobs" of Butyl rubber adhesive with very good holding power against gravity [but still removable] and vibration damping behaviour, with a maximum service temperature of 93 Celcius.

And I take it this has solved any possible freezing issues? You didn't say whether you experienced them yourself but i can see it's a common problem... It seems odd to me that that would only happen every few weeks, especially since it typically happens when the machine is idle. As you can see in the picture this thing seems to have some huge fans, no idea how much cfm they push/pull though.Since it sounds like a hardware issue, I take it this happens in other operating systems as well? How do I know if my northbridge and southbridge are actually getting hot? As you know HWMonitor doesn't see them, unless there is a DSDT edit I can do to have ACPISensors access the temps.

 

By the way, did you ever test that kext I posted to see if it improves your boot times?

post-969691-0-53975400-1407485949_thumb.jpg

Link to comment
Share on other sites

My Northbridge and Southbridge fans are in addition to a 250 mm side fan blowing air into the case and the PSU fan, and a rear and from 120 mm fans drawing air out of the case (i.e. it has a moving positive internal air pressure),

in conjunction with the AppleNForceATA.imk64bit.kext allow me to run the 650i with 8 GB of DDR2 800 MHz RAM and a Q6600 OC'd to 3.0 GHz with the MOBO and CPU temps running below 40 Celcius, with the CPU PWM cooler fan running at < 900 rpm to achieve this low CPU temperature.

 

I experienced them mainly when doing large data transfers but the original workaround was to boot with cpus=1 and maxmem=2048, and later just with maxmem=3456 and all cores running.....

 

I will be testing the kext shortly.......I have been busy sorting out ALC1200 and AD2000B audio under OX from 10.5.8 thru to 10.9.4.

 

BTW the amount of dust in your fan intakes will greatly impede the fan flow over the heatsinks...... :shock:

Link to comment
Share on other sites

BTW the amount of dust in your fan intakes will greatly impede the fan flow over the heatsinks...... :shock:

Lol I know... I don't take care of this computer as much as I should. Probably because I got it for free :D Too many computers to clean on a regular basis lol

 

My CPU core temps are typically 40 degrees or less, do you think the north and south bridge could still be getting hot?

Link to comment
Share on other sites

It depends on how much free air flow there is around and over the North and South bridge chipset heatsinks........needing heatsinks says a lot about how hot they get........you can get quite inexpensive handheld  IR laser spotting thermometers on ebay to check your internal case temperatures while the PC is running.....see inside my PC case:

 

post-200327-0-25903000-1407527697_thumb.jpg                                           post-200327-0-32475600-1407527650_thumb.jpg                                                      post-200327-0-47289800-1407527674_thumb.jpg

Link to comment
Share on other sites

I'm pretty sure my dad has one of those thermometers you're talking about so I'll check that out. Perhaps it would be best for me to just get rid of the current cooling setup entirely and replace it with some more modern stuff. I may also try undervolting it a little bit with clover and see if that helps. Although we both know what undervolting can do for stability.

Link to comment
Share on other sites

Just a suggestion:

 

1. Measure North and South bridge chipset heatsinks temps and other spot temps with IR thermometer

 

2. Remove dust from all key areas using compressed air

 

3. Re-measure North and South bridge chipset heatsinks temps and other spot temps with IR thermometer

 

4. If required then put a small fan on North and on South bridge chipset heatsinks

 

5. Re-measure North and South bridge chipset heatsinks temps and other spot temps with IR thermometer

 

This way you may not need to spend much money nor not risk stability further by undervolting.....you may achieve around a 10 C temp drop.......see here.....

Link to comment
Share on other sites

@ Kilopopo

 

I will be looking at installing the Yosemite release version when it arrives........:)

Link to comment
Share on other sites

  • 2 weeks later...

It's still freezing. I guess I'm going to try maxmem=3456. I would like to use all 6gb of my RAM but I suppose it's better than freezing. Why 3456? Is it just an arbitrary number that happens to work?

Link to comment
Share on other sites

 Share

×
×
  • Create New...