You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Nathan J Kurz <na...@tripod.tripod.com> on 1997/07/09 21:45:16 UTC

assert() without -DNDEBUG

I'm confused by the use of assert's in Apache.  In bprintf.c there are
a couple used as sanity checks.  This is fine, but should the CFLAGS
then default to include -DNDEBUG?  Or are the assert's necessary to
the code?  But why would abort() in the middle of a print be a good
thing in the distributed version?

nate@tripod.com
http://www.tripod.com



Re: assert() without -DNDEBUG

Posted by Rasmus Lerdorf <ra...@lerdorf.on.ca>.
> Which reminds me - anyone tracking henry spencer's package to make sure we
> get bug/portability fixes?

I don't think Henry has made any changes to this package for years.  

-Rasmus


Re: assert() without -DNDEBUG

Posted by Brian Behlendorf <br...@organic.com>.
At 01:55 PM 7/9/97 -0600, you wrote:
>There are no assert()s in anywhere except the regex code in the current
>source.  There are ap_assert()s.
>
>I'm not sure what you are talking about WRT NDEBUG.  It is in regex
>becaues it is in regex.

Which reminds me - anyone tracking henry spencer's package to make sure we
get bug/portability fixes?

	Brian


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
"Why not?" - TL           brian@organic.com - hyperreal.org - apache.org

Re: assert() without -DNDEBUG

Posted by Marc Slemko <ma...@worldgate.com>.
There are no assert()s in anywhere except the regex code in the current
source.  There are ap_assert()s.

I'm not sure what you are talking about WRT NDEBUG.  It is in regex
becaues it is in regex.

In 1.3, ap_assert logs it then does an abort().  That is a good thing
because that is the way assert() behaves.  Something is the way it
shouldn't be, and that could impact anything.  Best solution is to kill
the process.

On Wed, 9 Jul 1997, Nathan J Kurz wrote:

> I'm confused by the use of assert's in Apache.  In bprintf.c there are
> a couple used as sanity checks.  This is fine, but should the CFLAGS
> then default to include -DNDEBUG?  Or are the assert's necessary to
> the code?  But why would abort() in the middle of a print be a good
> thing in the distributed version?
> 
> nate@tripod.com
> http://www.tripod.com
> 
>