Re: panic: pmap_l2_to_l3: PA out of range, PA: 0x0

From: Mitchell Horne <mhorne_at_freebsd.org>
Date: Fri, 07 Oct 2022 22:08:06 UTC 

On 10/7/22 02:59, Dennis Clarke wrote:
> 
> Saw this while trying to do buildworld .. which it was for days :
> 
> login: root
> Password:
> Oct  1 21:26:36 ison login[807]: ROOT LOGIN (root) ON ttyu0
> Last login: Sun May 15 17:27:43 on ttyu0
> FreeBSD 14.0-CURRENT #2 main-n255202-0aef8628458-dirty: Sun May  8 
> 20:33:55 UTC 2022     root@ison:/usr/obj/usr/src/riscv.riscv64/sys/GENERIC
> 
> .
> .
> .
> 
> root@ison:~ # sysctl -a hw.ncpu
> hw.ncpu: 8
> root@ison:~ # sysctl -a hw.physmem
> hw.physmem: 34325733376
> 
> root@ison:~ # sysctl -a hw.ncpu | cut -f2 -d\:
>   8
> root@ison:~ # panic: pmap_l2_to_l3: PA out of range, PA: 0x0
> cpuid = 4
> time = 1664870924
> KDB: stack backtrace:
> db_trace_self() at db_trace_self
> db_trace_self_wrapper() at db_trace_self_wrapper+0x38
> kdb_backtrace() at kdb_backtrace+0x2c
> vpanic() at vpanic+0x154
> panic() at panic+0x2a
> pmap_remove_pages() at pmap_remove_pages+0x6c8
> vmspace_exit() at vmspace_exit+0xd6
> exit1() at exit1+0x4ac
> sys_exit() at sys_exit+0x10
> do_trap_user() at do_trap_user+0x206
> cpu_exception_handler_user() at cpu_exception_handler_user+0x72
> --- exception 8, tval = 0
> KDB: enter: panic
> [ thread pid 18297 tid 100640 ]
> Stopped at      kdb_enter+0x4a: sd      zero,0(s1)
> db> nt
> No such command; use "help" to list available commands
> db> bt
> Tracing pid 18297 tid 100640 td 0xffffffc11af8a680
> kdb_enter() at kdb_enter+0x48
> vpanic() at vpanic+0x198
> panic() at panic+0x2a
> pmap_remove_pages() at pmap_remove_pages+0x6c8
> vmspace_exit() at vmspace_exit+0xd6
> exit1() at exit1+0x4ac
> sys_exit() at sys_exit+0x10
> do_trap_user() at do_trap_user+0x206
> cpu_exception_handler_user() at cpu_exception_handler_user+0x72
> --- exception 8, tval = 0
> db>
> 
> Anything worth looking at here or just pull the plug?
> 
> 

Unlucky that you hit this, but the good news is the panic is known and 
recently fixed. I committed a series of changes yesterday which address 
the problem, ending with 1f9cc5ffc505.

Cheers,
Mitchell