Re: CURRENT [r308668] autofs kernel panic

On 1115T1016, Alex Deiter wrote:
> Hello,
> 
> FreeBSD 12.0-CURRENT-r308668 amd64 crashed on "pkg install” command:
> 
> # cat /usr/local/etc/pkg/repos/local.conf
> local: {
> url: "file:///net/nostromo/var/devel/pkg",
> enabled: yes
> }
> 
> Path /net/nostromo/var/devel/pkg accessible via autofs/automountd:
> 
> # ls /net/nostromo/var/devel/pkg
> All Latest digests.txz meta.txz packagesite.txz
> 
> # mount -t nfs
> nostromo:/var/devel/pkg on /net/nostromo/var/devel/pkg (nfs, nosuid,
> automounted)
> 
> # kgdb /usr/lib/debug/boot/kernel/kernel.debug /var/crash/vmcore.0
> ...
> 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 = 0x100df
> fault code = supervisor read data, page not present
> instruction pointer = 0x20:0xffffffff8063c88d
> stack pointer         = 0x28:0xfffffe085f966520
> frame pointer         = 0x28:0xfffffe085f966570
> 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 = -134214453 (pkg)
> 
> kgdb: kvm_read: invalid address (0x6c6576656c6572)
> kgdb: kvm_read: invalid address (0xf8000ce021d8ffff)
> #0  0x0000000000000000 in ?? ()
> (kgdb) list *0xffffffff8063c88d
> 0xffffffff8063c88d is in hardclock_cnt (/usr/src/sys/kern/kern_clock.c:568).

[..]

It looks a bit weird - the "kvm_read" errors, and also the one-function
traceback that contains only one function, that seems unrelated to autofs.
Are you sure you're using the right kernel?  Could you try the usual

kgdb /boot/kernel/kernel /var/crash/vmcore.0

instead?  Thanks!