Re: Multiple issues for Linksys NSLU2 (evbarm, armeb) NetBSD in HEAD date since June 2014

[Nick Hudson <nick.hudson%gmx.co.uk@localhost>]

> On 28 May 2015, at 22:47, Eddy Petrișor <eddy.petrisor+netbsd.org%gmail.com@localhost> wrote:
> 
> 2015-05-29 0:20 GMT+03:00 Nick Hudson <skrll%netbsd.org@localhost>:
>>> On 05/28/15 20:49, Eddy Petrișor wrote:
>>> 
>>> 2015-05-28 0:40 GMT+03:00 Nick Hudson <skrll%netbsd.org@localhost>:
>>>> 
>>>>> On 05/27/15 21:52, Eddy Petrișor wrote:
>>>>> 
>>>>> The last time I tried the problem below manifested. The problem below
>>>>> is also present in  HEAD, I just tested.
>>>> 
>>>> You mean the libcrypto problem? This is fixed now, but you can build just
>>>> a
>>>> kernel with
>>>> 
>>>>     ./build.sh kernel=NSLU2
>>>> 
>>>> 
>>>>> For issue 2 or 3, I wasn't yet able to get a trace.
>>>>> 
>>>>> Is there a conflict between any of the following options? (I already
>>>>> have DDB and DDB_HISTORY_SIZE and DDB_KEYCODE=0x40 from
>>>>> sys/arch/evbarm/conf/NSLU2)
>>>>> 
>>>>> options DDB_TEE_MSGBUF=1
>>>>> options DDB_COMMANDONENTER="trace;trace/u;trace/t;frame;show registers"
>>>>> options DDB_VERBOSE_HELP
>>>>> options DDB_ONPANIC=2
>>>>> makeoptions    DEBUG="-g"
>>>>> options VERBOSE_INIT_ARM
>>>> 
>>>> 
>>>> What do you mean by conflict?
>>>> 
>>>> A kernel with these options just built for me
>>> 
>>> When building with those options and the kernel panics I don't get a
>>> trace.
>>> I was wondering why?
>> 
>> 
>> No idea... what do you get?
> 
> I guess I assumed it was a kernel panic, I only something as light as:
> 
> [..]
> init: copying out path `/sbin/init' 11
> panic: kernel diagnostic assertion "*ptep == opte" failed: file
> "./arm/arm32/pmap.h", line 575 0x11dd605b [*0xc000a004] != 0
> Stopped in pid 1.1 (init) at    netbsd:cpu_Debugger+0x4:        bx      r14
> db>
> 
> Nothing else. I guess I assumed I should get a trace even in userland.
> 

Not sure onenter isn't working. Can't you just type "trace"?

>> Nick