Subject: Kernel panic (double fault trap) on X2
To: None <>
From: Dan McGregor <>
List: tech-kern
Date: 08/22/2005 00:24:56
I've been experinecing kernel panics on -current with a new AMD X2.=20
It's pretty intermittent and as far as I can tell it's related to the
sk(Marvell Gigabit) network adapter that I have on my motherboard.  It
occurs in the ifconfig process and it seems that one of the functions
in the driver or related to the driver is causing this fault.  It only
occurs with an SMP enabled kernel.  Looks like a uvm_fault that it's
causing, as I can't get it to back trace or make a core dump from
within DDB.

Am I correct in my belief that the old stack base pointer on an AMD64
is stored in $rbp - 0x8 on the stack of the interrupt service routine?
 That's what the AMD documentation seemed to suggest.