From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Mon, 21 Jul 2008 12:47:58 -0700 (PDT)

On Mon, 21 Jul 2008, Larry Finger wrote:

quoted

With an invalid opcode, mine is likely due to stack corruption.

No, invalid opcode is because we use the "ud2" instruction for BUG(), 
which causes an invalid op exception. So any BUG[_ON]() will always cause 
that on x86.

Is there really no more backtrace from that crash message?
It would tell me what driver it's in.

There is some "comm: b43" in the log so I'll check that one.