Well here I am, and I need your help.

I've posted on Insanelymac.com a few times but it seems everybody is too busy posting about what parts they should buy for their new hack (I've done it too).

So here I am. I browse this forum every now and then and find it to be pretty good. I'm no noob to the scene, built about 6 hacks in my days. Now my magnum opus, the crown jewel of my collection, the opal of my eye is having trouble - and I can track it down to one thing: 8GB of Ram.

A little about the system - It's build on a pretty common p5Ke/Wifi board and an even more common Q6600 proc. I just upgraded to 8GB of 1066 Kingston Hyper X ram, which I am currently running at 5-5-5-18 and 2.2v overclocked (with this board you have to clock up 1066 ram - the ram IS spec'd at that though).

And the problem - constant kernel panics. They happen all day long, maybe three or four times in pretty random settings.

My attempts so far:
1. I have disabled JMicron and removed the jmicron.kext and the appleviaata.kext which are known to cause problems with more than 4GB of ram.
2. I have disabled and removed my add in silicon image 3132 pcie raid card and it's associated kext that is also known to cause problems with 4GB+ ram.

The above two steps seem to have reduced the kernel panics. For example when doing a multicore render in After Effects it would KP every time after about 25 seconds when using the siliconimage3132r5.kext. Without the kext, it will complete the render after about 10 minutes fine. Multitask fine, run memtest and pass fine, use up the full 8 gb of ram fine (if a little laggy?).

However it will KP whilst using Firefox, or KP whilst in Photoshop. The KP's (I'm using the abbreviation now...) are seemingly unrelated to anything I'm actually doing, where as before it would always KP when large amounts of ram were used in AE or memtest.

It was totally fine and stable as a rock when I was using the 4GB of 800 generic ram. To tell you the truth I wouldn't have spent the $200 on the ram if I knew this would happen.

Still there are folks out there who run 8GB of 1066 without any hitches, so that points me to a possible solution. Some things:

What boards are you 8GB 1066 folks running?
What timings are you running?
Any specific kexts you have removed?
Any stability problems?
Any other specific bios settings I should implement?
And please, any advice or help you can offer?

Thanks for reading guys and I'll really appreciate any help!

Have you tried these updated JMicron kexts yet?

http://forum.insanelymac.com/index.php?showtopic=127611

---

 Mac OS X 10.6.4 Retail || Intel Core2 Duo E8200 2.66GHz || Gigabyte G31M-ES2L || GeForce 6600 GT 128MB || Realtek ALC883 Audio || Realtek RTL8169 LAN || Samsung SpinPoint F1 1TB HD || Sony Dual Layer IDE DVD RW DW-D22A || Apple Aluminium Keyboard || Mighty Mouse

 MacBook Pro || 2.4Ghz Core 2 Duo || 200Gb HD || 2Gb Ram || NVIDIA GeForce 8600M GT

 iPhone 4 || 16Gb

Nothing interesting in the Console ?

I have tried the Jmicron kexts and they were incompatible with my board (caused a KP at boot).

Here is the output from the console - I'm not sure if this is the actual panic. The log was written at the same time as the panic, but OSx86 machines don't log to panic.log like real apples do? I have set debug 0x100 so next time it panics I'll take a photo.

Title "SystemUIServer_2008-12-19-215719_Bluemeany.crash

Process: SystemUIServer [744]
Path: /System/Library/CoreServices/SystemUIServer.app/Contents/MacOS/SystemUIServer
Identifier: SystemUIServer
Version: ??? (???)
Code Type: X86 (Native)
Parent Process: launchd [1]

Date/Time: 2008-12-19 21:57:19.374 +1100
OS Version: Mac OS X 10.5.5 (9F33)
Report Version: 6

Exception Type: EXC_BAD_ACCESS (SIGBUS)
Exception Codes: KERN_PROTECTION_FAILURE at 0x0000000000000000
Crashed Thread: 0

Thread 0 Crashed:
0 com.apple.CoreFoundation 0x96ad8688 CFDictionarySetValue + 40
1 com.apple.systemuiserver 0x00020f68 0x1000 + 130920
2 com.apple.systemuiserver 0x00009022 0x1000 + 32802
3 com.apple.systemuiserver 0x0000618a 0x1000 + 20874
4 com.apple.systemuiserver 0x00005cb6 0x1000 + 19638

Thread 1:
0 libSystem.B.dylib 0x9136268e __semwait_signal + 10
1 libSystem.B.dylib 0x9138d36d pthread_cond_wait$UNIX2003 + 73
2 ...e.ImageCaptureNotifications 0x000d2eaa _StaticProcessIOObjectsArray + 95
3 libSystem.B.dylib 0x9138c6f5 _pthread_start + 321
4 libSystem.B.dylib 0x9138c5b2 thread_start + 34

Thread 2:
0 libSystem.B.dylib 0x9138bf66 kevent + 10
1 libSystem.B.dylib 0x9138c6f5 _pthread_start + 321
2 libSystem.B.dylib 0x9138c5b2 thread_start + 34

Thread 0 crashed with X86 Thread State (32-bit):
eax: 0xa08824ec ebx: 0x96ad866b ecx: 0x945cc858 edx: 0x00000010
edi: 0x00000000 esi: 0x00523940 ebp: 0xbffffc68 esp: 0xbffffc10
ss: 0x0000001f efl: 0x00010283 eip: 0x96ad8688 cs: 0x00000017
ds: 0x0000001f es: 0x0000001f fs: 0x00000000 gs: 0x00000037
cr2: 0x00000000


There was a heap of other stuff about binary images too.

Normally I can't deciper the panic dumps on the screen. They all have 14=page fault at the top, and don't point to any kext as the cause.

Thanks for your help guys.

I've decided to come at it all from a new angle: Windows XP

Prime95 seems to be the absolute choice in pushing your machines stability. So I gave it a go with my current bios settings. Low and behold it balked after about 10 minutes.

So back to bios I went - this time with an eye to overclock. So now I'm at 3.2 with a 400x8 multiplier and the ram at 800 vcore at 1.25 and memory at the same 5-5-5-15, 2.2v as before, also chnaged the nb to 1.4v. Ran Prime95 for 3 and half hours with no error and it didn't hit 60 degrees with my Zalman 9700. I'm back on leopard now to see how long it lasts without a KP.

3 days 12 hours uptime with no KP. It's not much I know but the computer has been put to decent levels of stress. Multicore rendering in AE, and hours of Spore on high at 1680x1050.

I am hopeful that the problem has been solved. If so it was related to general instability with the board/cpu/ram clock speed. Run Prime 95 on your system for as long as you can and see if it throws errors. If it does chances are (for me anyway) that OSX with kernel panic.

Might go over to Voodoo now see how stable that is.