Re: [follow-up] Fatal trap 12 in r195146+ in netisr_queue_internal

John Baldwin wrote:
> Can you print out 'owner' as well?  You won't get a panic until you actually 
> dereference 'owner' to get 'owner->td_state' even though gdb will show this 
> as the faulting line (gdb can sometimes get confused by compiler 
> optimization).  You are seeing these values because mtx_lock was changed (due 
> to either a mtx_unlock() or a mtx_init()) while you were spinning.   That 
> value of v is not what I have typically seen in these panics.  Do you also 
> have the original fatal kernel trap messages?
>   
Why does v change to a non-kernel address then? From what I see, it 
should never get assigned a value that's not MTX_UNOWNED or 
0xfff......(address,flags). However, I can reproduce this trap in all 
revisions starting with 195146 up to 195484 (and probably more, didn't 
check yet; at 1956xx these traps stop occurring).

vmcore.51 (all cores starting with .9 are related to mtx_lock_sleep() trap):

Unread portion of the kernel message buffer:
kernel trap 12 with interrupts disabled


Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault virtual address   = 0x14ee288
fault code              = supervisor read data, page not present
instruction pointer     = 0x20:0xffffffff80586255
stack pointer           = 0x28:0xffffff80787115f0
frame pointer           = 0x28:0xffffff8078711620
code segment            = base 0x0, limit 0xfffff, type 0x1b
                        = DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags        = resume, IOPL = 0
current process         = 2438 (iperf)
trap number             = 12
panic: page fault
cpuid = 0
Uptime: 43s
Physical memory: 4014 MB

(kgdb) fr 6
#6  0xffffffff80586255 in _mtx_lock_sleep (m=0xffffffff80e60823, 
tid=18446742977255365296, opts=Variable "opts" is not available.
) at /usr/src/sys/kern/kern_mutex.c:407
407                     owner = (struct thread *)(v & ~MTX_FLAGMASK);
(kgdb) print owner
$1 = (volatile struct thread *) 0x14ee000
(kgdb) print v
$2 = 21946368
(kgdb) print m->mtx_lock
$3 = 4
(kgdb) print owner->td_state
Cannot access memory at address 0x14ee288

vmcore.50:
Unread portion of the kernel message buffer:
kernel trap 12 with interrupts disabled


Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault virtual address   = 0x14ee288
fault code              = supervisor read data, page not present
instruction pointer     = 0x20:0xffffffff80586255
stack pointer           = 0x28:0xffffff80785005f0
frame pointer           = 0x28:0xffffff8078500620
code segment            = base 0x0, limit 0xfffff, type 0x1b
                        = DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags        = resume, IOPL = 0
current process         = 2448 (iperf)
trap number             = 12
panic: page fault
cpuid = 0
Uptime: 53s
Physical memory: 4014 MB

(kgdb) fr 6
#6  0xffffffff80586255 in _mtx_lock_sleep (m=0xffffffff80e60823, 
tid=18446742974555039520, opts=Variable "opts" is not available.
) at /usr/src/sys/kern/kern_mutex.c:407
407                     owner = (struct thread *)(v & ~MTX_FLAGMASK);
(kgdb) print owner
$1 = (volatile struct thread *) 0x14ee000
(kgdb) print m->mtx_lock
$2 = 4
(kgdb) print v
$3 = 21946368
(kgdb) print owner->td_state
Cannot access memory at address 0x14ee288


--
Kamigishi Rei
KREI-RIPE