Re: cvs commit: apache-1.3/src/main http_main.c

[Dean Gaudet <dg...@arctic.org>]

-1.  I'm backing this out.  Any such problems are problems with 3rd party
modules and libraries, and it is the module implementor's job to fix them. 
Furthermore, this doesn't cover all the crap a library/module could fuck
up, like setrlimit().  Surely you're not going to cover a libraries butt
in that area... just like we don't do anything to close all the
descriptors not listed in a pool.  If a library/module NEEDS this cleanup
then they can damn well register a cleanup.

Dean

On 3 Jan 1999 fielding@hyperreal.org wrote:

> fielding    99/01/03 05:09:49
> 
>   Modified:    src      CHANGES
>                src/main http_main.c
>   Log:
>   In order to prevent some errant library code from setting an alarm
>   on one request and having it affect some later connection, force the
>   alarm to be cleared before each connection even if Apache's own
>   timeouts are being handled by the parent.  This doesn't help for multiple
>   requests on the same connection, but it is extremely unlikely that a
>   request stream would persist longer than a non-Apache errant timeout,
>   and this limits the problem to a single user without impacting performance.
>   
>   Revision  Changes    Path
>   1.1197    +5 -0      apache-1.3/src/CHANGES
>   
>   Index: CHANGES
>   ===================================================================
>   RCS file: /home/cvs/apache-1.3/src/CHANGES,v
>   retrieving revision 1.1196
>   retrieving revision 1.1197
>   diff -u -r1.1196 -r1.1197
>   --- CHANGES	1999/01/03 12:04:34	1.1196
>   +++ CHANGES	1999/01/03 13:09:46	1.1197
>   @@ -1,5 +1,10 @@
>    Changes with Apache 1.3.4
>    
>   +  *) In order to prevent some errant library code from setting an alarm
>   +     on one request and having it affect some later connection, force the
>   +     alarm to be cleared before each connection even if Apache's own
>   +     timeouts are being handled by the parent. [Roy Fielding]
>   +
>      *) Make sure under ELF-based NetBSD (now) and OpenBSD (future) we don't
>         search for an underscore on dlsym() (as it's already the case
>         for FreeBSD 3.0). [Todd Vierling <tv...@pobox.com>] PR#2462
>   
>   
>   
>   1.417     +4 -1      apache-1.3/src/main/http_main.c
>   
>   Index: http_main.c
>   ===================================================================
>   RCS file: /home/cvs/apache-1.3/src/main/http_main.c,v
>   retrieving revision 1.416
>   retrieving revision 1.417
>   diff -u -r1.416 -r1.417
>   --- http_main.c	1999/01/01 19:04:49	1.416
>   +++ http_main.c	1999/01/03 13:09:48	1.417
>   @@ -3610,7 +3610,10 @@
>    	 * (Re)initialize this child to a pre-connection state.
>    	 */
>    
>   -	ap_kill_timeout(0);	/* Cancel any outstanding alarms. */
>   +	ap_kill_timeout(0);	/* Cancel any outstanding alarms */
>   +#ifdef OPTIMIZE_TIMEOUTS
>   +	alarm(0);		/* even if not set by Apache routines */
>   +#endif
>    	current_conn = NULL;
>    
>    	ap_clear_pool(ptrans);
>   
>   
>   
>