You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by rb...@apache.org on 2000/10/16 00:38:28 UTC
Re: cvs commit: apache-2.0/src/include ap_buckets.h
On 15 Oct 2000 gstein@locus.apache.org wrote:
> gstein 00/10/15 15:24:20
>
> Modified: src/ap ap_buckets.c ap_buckets_eos.c ap_buckets_file.c
> ap_buckets_heap.c ap_buckets_mmap.c
> ap_buckets_pipe.c ap_buckets_simple.c
> ap_buckets_socket.c
> src/include ap_buckets.h
> Log:
> const-ify the type stuff and simplify the type registration process
Great set of change. :-) Does this registration process make more sense
now? This is where I wanted to go when I started, but I was taking very
small steps, which I probably shouldn't have done. Now that we are about
99% of the way there (we still need the read/setaside/destroy/split macros
instead of functions) does this make more sense?
If people would still like this removed, I wam willing to remove the
changes, but I personally think that this modification is a good thing,
and since it has been working in FreeBSD without a performance problem, I
don't see any reason not to have it.
Ryan
_______________________________________________________________________________
Ryan Bloom rbb@apache.org
406 29th St.
San Francisco, CA 94131
-------------------------------------------------------------------------------
RE: cvs commit: apache-2.0/src/include ap_buckets.h
Posted by "William A. Rowe, Jr." <wr...@rowe-clan.net>.
> From: rbb@apache.org [mailto:rbb@apache.org]
> Sent: Sunday, October 15, 2000 5:38 PM
>
> On 15 Oct 2000 gstein@locus.apache.org wrote:
>
> > gstein 00/10/15 15:24:20
> >
> > Modified: src/ap ap_buckets.c ap_buckets_eos.c
> ap_buckets_file.c
> > ap_buckets_heap.c ap_buckets_mmap.c
> > ap_buckets_pipe.c ap_buckets_simple.c
> > ap_buckets_socket.c
> > src/include ap_buckets.h
> > Log:
> > const-ify the type stuff and simplify the type
> registration process
+1 - this is great
> (we still need the read/setaside/destroy/split macros
> instead of functions) does this make more sense?
It ain't over till the fat lady sings... let's see that patch :-)
> If people would still like this removed, I wam willing to remove the
> changes
No, please let's identify any remaining performace penalties and
introduce the macros to simplify/optimized this method. It makes
much more sense to me than the original (I wish we had started here :-)
The macros should be 'accessories' - the code we have implemented makes
understanding the filtering much simpler