Leopard Kernel Panic

Looks like a faulty processor but I don'thave any expertise here. I'd be running Hardware test on the machine.

OK..... just had nearly the identical error on a brand new 8-core Mac Pro. Was working all nice in After Effects when WHAM!... KP... thought that was pretty odd. Rebooted and immediately got the flashing power button. Decided to pull all the hard drives except the boot drive. Rebooted and machine came right up. Oh... had a G-RAID attached via FW800 as well... but wasn't doing anything with that drive.

Here is my error report:

Tue Nov 27 20:52:19 2007
Machine-check capabilities (cpu 1) 0x0000000000000806:
6 error-reporting banks
threshold-based error status present
Machine-check status 0x0000000000000005
restart IP valid
machine-check in progress
MCA error-reporting registers:
IA32 MC0STATUS(0x401): 0x1000000020000000 invalid
IA32 MC1STATUS(0x405): 0x0000000000000000 invalid
IA32 MC2STATUS(0x409): 0x0000000000000000 invalid
IA32 MC3STATUS(0x40d): 0x0020000000000000 invalid
IA32 MC4STATUS(0x411): 0x0000000000000011 invalid
IA32 MC5STATUS(0x415): 0xb200001080200e0f valid
MCA error code : 0x0e0f
Model specific error code: 0x8020
Other information : 0x00000010
Threshold-based status : Undefined
Status bits:
Processor context corrupt
Error enabled
Uncorrected error
Machine-check capabilities (cpu panic(cpu 0 call 0xer 0x0000000000000806:0 60er1A756Fo:

Strangely enough, I woke up this morning to find the system had rebooted itself....due to a panic.

Thu Nov 29 07:04:41 2007
panic(cpu 3 caller 0x001A7BED): Kernel trap at 0x00000012, type 14=page fault, registers:
CR0: 0x8001003b, CR2: 0x00000012, CR3: 0x01004000, CR4: 0x00000660
EAX: 0x129da300, EBX: 0x06efb000, ECX: 0x00508444, EDX: 0x090c7d00
CR2: 0x00000012, EBP: 0x5c3efda8, ESI: 0x06ff5d00, EDI: 0x06ff5700
EFL: 0x00010246, EIP: 0x00000012, CS: 0x00000008, DS: 0x00620010
Error code: 0x00000010

Backtrace, Format - Frame : Return Address (4 potential args on stack)
0x5c3efb38 : 0x12b0e1 (0x455670 0x5c3efb6c 0x133238 0x0)
0x5c3efb88 : 0x1a7bed (0x45ea20 0x12 0xe 0x45e1d4)
0x5c3efc68 : 0x19e517 (0x5c3efc80 0x129da300 0x5c3efda8 0x12)
0x5c3efc78 : 0x12 (0xe 0x48 0x5c3e0010 0x400010)
0x5c3efda8 : 0x885d85 (0x6efb000 0x6ff5d00 0x0 0x0)
0x5c3efe48 : 0x886724 (0x6efb000 0x6f6b850 0x0 0x0)
0x5c3efec8 : 0x886779 (0x6efb000 0x0 0x0 0x408ea0)
0x5c3efee8 : 0x87f998 (0x6efb000 0x0 0x5c3eff18 0x87fa7c)
0x5c3eff18 : 0x41d149 (0x6efb000 0x0 0x1 0x19ccc1)
0x5c3eff68 : 0x41c2a6 (0x6f34980 0x781e8fc 0x781e8d4 0x0)
0x5c3eff98 : 0x41bf88 (0x6fe1880 0x0 0x5c3effc8 0x1a6d85)
0x5c3effc8 : 0x19e2ec (0x6fe1880 0x0 0x1a10b5 0x781e8b8)
Backtrace terminated-invalid frame pointer 0
Kernel loadable modules in backtrace (with dependencies):
com.apple.driver.AppleUSBEHCI(3.0.3)@0x87c000->0x88dfff
dependency: com.apple.iokit.IOPCIFamily(2.4)@0x6f1000
dependency: com.apple.iokit.IOUSBFamily(3.0.3)@0x60d000

BSD process name corresponding to current thread: kernel_task

Mac OS version:
9B18

Kernel version:
Darwin Kernel Version 9.1.0: Wed Oct 31 17:46:22 PDT 2007; root:xnu-1228.0.2~1/RELEASE_I386
System model name: MacPro1,1 (Mac-F4208DC8)


Caused by the USB driver by the looks of it. Not quite sure why, as it was asleep at the time. Odd. 😟

By the way, for anyone really interested in kernel panics and CPU exceptions:

http://www.intel.com/design/processor/manuals/253668.pdf

Chapter 5 is very informative. 🙂

OK.... I've been getting the flashing light and inability to boot up problem on a more frequent basis. I'm really starting to suspect a hardware issue. Here is my latest log from when the computer was unable to start up:

Wed Dec 26 05:17:50 2007
Machine-check capabilities (cpu 5) 0x0000000000000806:
6 error-reporting banks
threshold-based error status present
Machine-check status 0x0000000000000005
restart IP valid
machine-check in progress
MCA error-reporting registers:
IA32 MC0STATUS(0x401): 0x1000000020000000 invalid
IA32 MC1STATUS(0x405): 0x0000000000000000 invalid
IA32 MC2STATUS(0x409): 0x0000000000000000 invalid
IA32 MC3STATUS(0x40d): 0x0020000000000000 invalid
IA32 MC4STATUS(0x411): 0x0000000000000011 invalid
IA32 MC5STATUS(0x415): 0xb200001080200e0f valid
MCA error code : 0x0e0f
Model specific error code: 0x8020
Other information : 0x00000010
Threshold-based status : Undefined
Status bits:
Processor context corrupt
Error enabled
Uncorrected error
Manic(cpu 5 hallerh0cx capabilities (cp0


Also... I ran the hardware test and received to sensor errors when running the extended testing:
4SNS/1/40000000:TM2S (first test)
4SNS/1/40000000:TM3S (second test)


If I go in and reseat one or two hard drives many times I'm able to reboot fine.... not sure if it's an SATA bus or controller problem or something completely different. Disk Utility says all my drives are fine.

When I started reading the posts of how others were doing with moving to Leopard, I was concerned about how folks were migrating to Leopard. I decided to take steps to make it easier to isolate any problems when I moved up. I upgraded via Apple's "Up-to-Date" DVD. When I upgraded, I put a new 500 GB drive in another bay, partitioned and formatted it like the system drive, then I copied the data from the data partition and installed Leopard on the new drive. It was not the most painless way to upgrade because I had to re-stack my system from scratch, but it ensured that I had a solid regression path and all my data is backed up. When I was convinced I was ready to complete the transition, I pulled the system drive, put the new Leopard drive into drive bay 1, put the other drive in an enclosure and stored it in a safe place. I did this on 02 Nov, and other than the usual early adopter bugs, I've had no issues or KP's. Like early Tiger, there's a lot to like, and a lot of room for improvement. Fortunately, stability doesn't seem to be an issue ... **Knock on wood**

OK... it appears that the 1GB of RAM that Apple sent with the machine is bad. I had noticed that I had a red light on my Riser A. I pulled the Apple supplied RAM from the board and swapped it out with my Crucial RAM and the light went out. Not sure if that was causing the startup and kernel panics... I'll keep you posted.

Bad RAM can cause fits, especially if there are no visual indicators. That would definitely cause KP's.

MacIntouch "Bad RAM" was a good article and "must read." trouble with bad RAM and k/ps is you can have bad data written to disk as a result. Fixed the problem. Fresh OS, but I had to be careful for months with using Disk Warrior even more often.

This is interesting. I had a few kernel panics over the past few weeks. This was on a PowerMac G5, actually a dual-CPU machine, and it panicked at exactly the same spot, in com.apple.driver.AppleUSBEHCI(3.0.3). Is this coincidence? Completely different hardware, yet the same routine that causes a kernel panic?


Unresolved kernel trap(cpu 0): 0x400 - Inst access DAR=0x00000000E0784000 PC=0x0000000000000060
Latest crash info for cpu 0:
Exception state (sv=0x67e8aa00)
PC=0x00000060; MSR=0x40009030; DAR=0xE0784000; DSISR=0x40000000; LR=0x00575AA0; R1=0x55223A80; XCP=0x00000010 (0x400 - Inst access)
Backtrace:
0x00575A84 0x00570970 0x006ED0DC 0x006EE19C 0x006EE240 0x006E50E0
0x006E5024 0x00355B48 0x00354DAC 0x00354E60 0x000AFE54
Kernel loadable modules in backtrace (with dependencies):
com.apple.driver.AppleUSBEHCI(3.0.3)@0x6e0000->0x6f5fff
dependency: com.apple.iokit.IOPCIFamily(2.4)@0x6ae000
dependency: com.apple.iokit.IOUSBFamily(3.0.3)@0x56f000
com.apple.iokit.IOUSBFamily(3.0.3)@0x56f000->0x5a3fff
Proceeding back via exception chain:
Exception state (sv=0x67e8aa00)
previously dumped as "Latest" state. skipping...
Exception state (sv=0x4c7e1500)
PC=0x00000000; MSR=0x0000D030; DAR=0x00000000; DSISR=0x00000000; LR=0x00000000; R1=0x00000000; XCP=0x00000000 (Unknown)

BSD process name corresponding to current thread: kernel_task

Mac OS version:
9B18

Kernel version:
Darwin Kernel Version 9.1.0: Wed Oct 31 17:48:21 PDT 2007; root:xnu-1228.0.2~1/RELEASE_PPC

I have solved this problem for me, I think. I found this forum thread:

http://discussions.apple.com/thread.jspa?threadID=1220488&tstart=0

A suspicion expressed in there is that kernel panics with USBEHCI in the backtrace can be related to the hub in Cinema displays. And I am using a Cinema display... Unplugging the USB cable from the display and plugging the keyboard directly into the computer seems to have fixed the problem. I haven't seen any freezes since.

And there is a USB cable going into the backside of the Mac Pro out to the ACD. The keyboard itself is also a hub; and the cable can cause issues. But this points to the display as being the culprit.