Alright, I did the hardware test and it said nothing was wrong. But that's not the end all be all, so I'll have to look into the next tip. Thanks for the help so far everyone!
displaced said:So the part of the system that's actually faulting here is pretty close to the metal -- kernel itself, iokit, nmi, mach, etc. Now these bits are pretty well tested, so panics this regular are unlikely to be originally caused by these chunks of code. Now, the backtrace [EDIT: not 'backtrace'... I mean, 'exception chain'] in those logs is a little more interesting. There, the counter's much, much higher, indicating the user-mode app which made the call to the kernel code which led to the crash. Sadly, there's no way for us here to find which program was living at that memory address when the thing blew up.
in2themystik2 said:Okay, since a lot of this lingo is foreign to me, when you say "user-mode app" what exactly do you mean, in terms of which programs that might be?
in2themystik2 said:Okay, before archiving and installing, do I need to make sure all my files are copied on another source so I don't lose them, or will they all still be there. I know I should be backing up, but I just got the computer in December so there are just some school word files on there so far. But reinstalling the OS, does that erase things, or have nothing to do with that? Sorry if this is a stupid question! thanks, you've been very helpful!
*picks remnants of my brain off the ground and hides in a corner*displaced said:No problem... I thought after I'd posted that I was ranting a little, but I hope I'll manage to explain myself
All modern operating systems (OS X, Windows NT, Linux etc.) generally divide memory into two areas: kernel-space and user-space. Memory within kernel space is reserved for use entirely by the core routines of the operating system (input/output, memory management, etc) and by kernel extensions (hardware drivers, etc). Kernel space is special in that it's always in physical memory -- never moved out to the swap-file. Kernel-space code also has complete, unfettered access to the system's hardware. Generally speaking, an unhandled error in kernel-space will result in the panics you've been seeing.
User-space is the rest of memory, and where all other programs on your machine run. Everything from the Aqua user interface to Microsoft Word run in user-space. Of course, these programs often require the kernel-space code to do stuff for them. If MS Word needs to allocate more memory to itself, it asks the kernel-space code to do it. Likewise, if iTunes is playing, eventually the audio data will find itself passed into a kernel-space program before it ends up in your Mac's audio hardware.
The interface between kernel- and user-space is really crucial. Even if a user-space program talks garbage to the kernel-space (for example, sends a picture file to the audio system), this must not cause the kernel code to fall over and die. Rather, the program that sent the invalid data should take the fall.
Most traceable kernel panics are the result of a bad kernel extension. I mentioned a panic I had with a dodgy audio card driver. In this case, the fault was in the kernel extension... and an unhandled error in kernel space will cause a panic.
In your case, the source of the panic does not seem to be within kernel space. Something outside of the kernel is causing a kernel-space error, and thus, a panic.
In your logs, you'll see the PC (program counter -- basically a pointer to the bit of code your processor was running at the time) originally contains a low number (0x000<something>). That's kernel-space. Then, as the log traces the exception (the fault) back to its source, the program counter gets much higher (0x482<something>). That's user-space, and could be any program you're running.
Let's pretend the panic log traces the exception back to address 0x90000000. That's definitely user-mode, since it's so high. But we don't know what program occupied that address at that moment in time. The only way to find out would be to connect a second machine to your Mac and run an external debugger.
As I mentioned, it's really tough for a user-mode program to cause a kernel panic, since the kernel's so strict with how information's passed to it.
That's really, really tough to do. The kernel's servicing thousands of requests by user-mode programs every second (rough guess... possibly an order of magnitude out! ), and not all programs are well behaved... so it has to be incredibly strict about how programs speak to it.
So, I'd imagine the programs you're running aren't to blame. The kernel's falling over either because the kernel itself is damaged (reinstall OS X), or because of a hardware fault (disk corruption, dodgy RAM).
Good luck... I really hope you get this sorted. Fingers crossed, a quick Archive & Install reinstallation of OS X will fix it!
Kingsly said:*picks remnants of my brain off the ground and hides in a corner*
Seriously, thanks for the lesson. Now I know!
In Activity monitor, would kernel space be the RAM that is WIRED, ACTIVE or neither??