RPI-B VM panic

[Hans Petter Selasky]

On 06/11/14 22:51, Michael Tuexen wrote:
> On 11 Jun 2014, at 21:59, Hans Petter Selasky <hps at selasky.org> wrote:
>
>> On 06/06/14 09:41, Hans Petter Selasky wrote:
>>> Hi,
>>>
>>> I'm seeing this with RPI-B:
>>>
>>> panic: vm_page_insert_after: msucc doesn't succeed pindex
>>> KDB: enter: panic
>>> [ thread pid 18 tid 100052 ]
>>> Stopped at      $d:     ldrb    r15, [r15, r15, ror r15]!
>>> db>
>>>
>>>
>>> Any ideas?
> Which revision are you using? What is triggering the panic? I could
> try to reproduce the problem.
>
> I've used r267055 with reverted r266083 for a couple of days and
> it was running stable. It compiled a lot of ports including wireshark.
>

Hi,

I'm running -current with a patch reverted for the CPU counter. It 
happens around growfs. The error is not constant. If I change the boot 
timing by plugging more USB devices, then I sometimes can pass the point 
of error. I think the problem is related to the following commit:

> commit 7d20e37fb658b0e2cd7f3c13dac8022e0e866a21
> Author: alc <alc at FreeBSD.org>
> Date:   Sun May 12 16:50:18 2013 +0000
>
>     Refactor vm_page_alloc()'s interactions with vm_reserv_alloc_page() and
>     vm_page_insert() so that (1) vm_radix_lookup_le() is never called while the
>     free page queues lock is held and (2) vm_radix_lookup_le() is called at most
>     once.  This change reduces the average time that the free page queues lock
>     is held by vm_page_alloc() as well as vm_page_alloc()'s average overall
>     running time.
>
>     Sponsored by:       EMC / Isilon Storage Division
>

Looks like we are trying to grow the stack and then the pages are not in 
the expected order.

--HPS