svn commit: r226405 - in stable/9: . etc/defaults lib/libc/stdlib sys/amd64/conf sys/i386/conf sys/ia64/conf sys/pc98/conf sys/powerpc/conf sys/sparc64/conf

[Sergey Kandaurov]

On 17 October 2011 18:40, Ken Smith <kensmith at buffalo.edu> wrote:
> On Sun, 2011-10-16 at 09:24 +0800, Adrian Chadd wrote:
>> Hi!
>>
>> Would you consider leaving these in but just commenting them out?
>>
>> That way if a user hits a deadlock or memory corruption issue, we can
>> just say "uncomment the section marked _debugging_ in your config
>> file."
>>
>> Thanks,
>
> I'm not completely against it.  My *slight* hesitation in doing it would
> be that from my experience these knobs aren't a one-size-fits-all thing
> for people trying to debug stuff in production.  So you sorta need to
> know a bit more about what you're doing when fiddling with them.
>
> Turning them all on is probably OK for a workstation user but less
> likely to be OK for a server admin.  Some may want/need to add in
> KDB_UNATTENDED depending on their circumstances, etc.
>

Hi.

In 8-STABLE (and 8.2) both KDB and KDB_TRACE present in GENERIC
kernels (added in r214326, I guess they are needed for printing a stack
trace on panic). Were these options removed intentionally in this commit?

-- 
wbr,
pluknet