httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Trawick <>
Subject Re: cvs commit: apache-2.0/src/modules/standard mod_include.c
Date Sat, 16 Sep 2000 00:06:15 GMT writes:

> On 15 Sep 2000 wrote:
> > trawick     00/09/15 14:23:46
> > 
> >   Modified:    src/modules/standard mod_include.c
> >   Log:
> >   Handle a tag that appears at the start of a bucket.
> This is incorrect.  A split at the start of a bucket should create a zero
> length bucket and a bucket with the rest of the data.  If splitting a
> bucket at the zero offset returns an error, then we end up with a lot of
> special case code throughout the server.

When you change the bucket split routines to act like you say they
should here, feel free to change that code back.  (Hint: Look at
simple_split().)  For now, it prevents a segfault and as far as I can
tell it actually works.

> >   Get rid of an endless loop (waiting for dptr2 to become NULL instead of
> This change is also incorrect.  If the find_string does not find the
> starting sequence, then it will return NULL.  With this change, we will
> search for an ending sequence even if we don't have a starting
> sequence.

This check prevents a hard loop which was quite easy to get into.
Simply backing it out would not be cool.  I'll look into that loop
again with your comments in mind and look for the better fix.


Jeff Trawick | | PGP public key at web site:
          Born in Roswell... married an alien...

View raw message