Best practice when building a -STABLE branch kernel

2002-10-18 Thread joe
Is it a best practice to include debuging options in a -STABLE tree 
kernel? 

I have read the following article which suggests that a business-as- 
usual practice should be to include debugging options.

http://www.onlamp.com/pub/a/bsd/2002/03/21/Big_Scary_Daemons.html?page=1

The kernel developer's handbook 
(http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/kerneldebug.html)
 
does not offer an opinion

Is there some alternative thinking?

---
Joe Sotham
---
Christianity got over the difficulty of furious opposites by keeping 
them both and keeping them furious.
  - G.K. Chesterton

To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-questions in the body of the message



Re: Best practice when building a -STABLE branch kernel

2002-10-18 Thread Matt Smith
List, please correct me if I am wrong:

The business-as-usual practice would be to not run -Stable, but rather
run a -Release.  -Stable, although more stable than -Current, should not
be run in business-production, although my hunch is that many small
environments do.

If you are running -Stable, then chances are you have some technical
knowledge, and could contribute back to the project, in which case
including debugging options could be helpful.

Any other thoughts?
-Matt

On Fri, 2002-10-18 at 12:10, joe wrote:
 Is it a best practice to include debuging options in a -STABLE tree 
 kernel? 
 
 I have read the following article which suggests that a business-as- 
 usual practice should be to include debugging options.
 
 http://www.onlamp.com/pub/a/bsd/2002/03/21/Big_Scary_Daemons.html?page=1
 
 The kernel developer's handbook 
 
(http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/kerneldebug.html)
 
 does not offer an opinion
 
 Is there some alternative thinking?
 
 ---
 Joe Sotham
 ---
 Christianity got over the difficulty of furious opposites by keeping 
 them both and keeping them furious.
   - G.K. Chesterton
 
 To Unsubscribe: send mail to [EMAIL PROTECTED]
 with unsubscribe freebsd-questions in the body of the message
 



To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-questions in the body of the message



Re: Best practice when building a -STABLE branch kernel

2002-10-18 Thread Gary W. Swearingen
Matt Smith [EMAIL PROTECTED] writes:

 The business-as-usual practice would be to not run -Stable, but rather
 run a -Release.  -Stable, although more stable than -Current, should not
 be run in business-production, although my hunch is that many small
 environments do.

IMO, the best practice is to build your OS and thoroughly test it as you
will be using it before putting it into important use.  If you do this,
you might as well use -STABLE.  Otherwise, unless you really need
something in -STABLE, you should use -RELEASE except that you should
review security fixes, etc, and maybe upgrade using a cvs tag like
RELENG_4_7 which has only important fixes for -RELEASE-4.7.

 If you are running -Stable, then chances are you have some technical
 knowledge, and could contribute back to the project, in which case
 including debugging options could be helpful.

AFAIK, including the debugging options is not risky or performance-
harming (except maybe using more memory?).  But for most people, it
doesn't make much sense to use it unless you also prepare your OS to
save crash dumps.  Most will do all this only so they can help OS
development by giving decent reports about OS crashes.  I know of no
good reason not to do it for any OS version, except to avoid the setup
effort.  It doesn't take much technical knowledge that can't be learned
by reading the FAQ about kernel panics (and maybe a few manuals
starting with crash(8)).

To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-questions in the body of the message