DragonFlyBSD

Hi,

It looks like the system accessed the address 0xdeadc122, which is 0xdeadc0de + 0x44; deadcode is filled in to free()d kernel memory.

The particular dereference that generated this invalid address was the extraction of the mountpoint from the vnode structure in vop_write. This means that vop_write was using an already-kfree()d vnode, which is bad news.

I've experienced the same problem:

Fatal trap 12: page fault while in kernel mode
cpuid = 0; lapic.id = 00000000
fault virtual address = 0xdeadc122
fault code = supervisor read, page not present
instruction pointer = 0x8:0xc028913f
stack pointer = 0x10:0xd62cc6c8
frame pointer = 0x10:0xd62cc7d0
code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 1 (init)
current thread = pri 31
<- SMP: XXX
kernel: type 12 trap, code=0

CPU0 stopping CPUs: 0x00000000
stopped

I have a crash dump available. It's a rather bulky 105MB xz compressed tar ball.

Could you upload the kernel & core to someplace we can access? It'd be great help!