You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Bill Stoddard <st...@raleigh.ibm.com> on 2000/08/23 19:11:55 UTC
Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.c mod_cgid.c mod_include.c mod_mime_magic.c
-1 (that's a veto) on this patch. It breaks Apache on Windows for the sole purpose of
'removing IOLs'. Removing IOLs is not a goal. When enough filtering infrastructure is in
to allow IOLs to be replaced, then go for it.
Bill
> rbb 00/08/22 17:01:58
>
> Modified: src CHANGES
> src/include ap_mpm.h buff.h
> src/main Makefile.in buff.c http_core.c http_protocol.c
> src/modules/mpm/beos beos.c
> src/modules/mpm/dexter dexter.c
> src/modules/mpm/mpmt_beos mpmt_beos.c
> src/modules/mpm/mpmt_pthread mpmt_pthread.c
> src/modules/mpm/perchild perchild.c
> src/modules/mpm/prefork prefork.c
> src/modules/mpm/spmt_os2 spmt_os2.c
> src/modules/mpm/winnt mpm_winnt.c
> src/modules/proxy proxy_connect.c proxy_ftp.c proxy_http.c
> src/modules/standard mod_cgi.c mod_cgid.c mod_include.c
> mod_mime_magic.c
> Log:
> Remove IOLs from Apache. They are no longer necessary, now that we have
> filtering beginning to work. There is a hack that has been repeated
> through this patch, we morph a pipe into a socket, and put the socket
> into the BUFF. Everytime we do that, we are working with a pipe from
> a CGI, and we should be creating a pipe bucket and passing that bucket
> back. Because we don't actually have pipe buckets yet, we are using this
> hack. When we get pipe buckets, this will be fixed
Re: Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.c mod_cgid.c mod_include.c mod_mime_magic.c
Posted by Jeff Trawick <tr...@bellsouth.net>.
Jeff Trawick <tr...@bellsouth.net> writes:
> FWIW, CGIs on Linux generally fail for me today... I haven't gotten to the
> bottom of it yet. It may be timing related, as they work for me when
> I step through with the debugger to see what is going wrong :( It
> remains to be seen whether or not the problem I'm seeing has anything
> to do with Ryan's patch. I suspect so, but it is hard to determine
> how much went in since the last time I did cgis...
>
> (now to add cgi usage to my minimal every-two-hour test bucket...)
>
> --
> Jeff Trawick | trawick@ibm.net | PGP public key at web site:
> http://www.geocities.com/SiliconValley/Park/9289/
> Born in Roswell... married an alien...
Generally (on my system, at least), the httpd process gets around to
trying to read from the cgi *before* the cgi child has even exec-ed
the cgi program. ap_bgets() gets APR_TIMEUP from the read, and an
error is generated at that point ("premature end of script headers").
I presume that APR is confused about the timeout stuff because the
"socket" was not created and/or set up properly. It seems pointless
to debug it further, knowing that everything is zero in the
apr_socket_t but the descriptor and the pool.
--
Jeff Trawick | trawick@ibm.net | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Re: Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.c mod_cgid.c mod_include.c mod_mime_magic.c
Posted by Jeff Trawick <tr...@bellsouth.net>.
"Bill Stoddard" <st...@raleigh.ibm.com> writes:
> -1 (that's a veto) on this patch. It breaks Apache on Windows for the sole purpose of
> 'removing IOLs'. Removing IOLs is not a goal. When enough filtering infrastructure is in
> to allow IOLs to be replaced, then go for it.
>
> Bill
FWIW, CGIs on Linux generally fail for me today... I haven't gotten to the
bottom of it yet. It may be timing related, as they work for me when
I step through with the debugger to see what is going wrong :( It
remains to be seen whether or not the problem I'm seeing has anything
to do with Ryan's patch. I suspect so, but it is hard to determine
how much went in since the last time I did cgis...
(now to add cgi usage to my minimal every-two-hour test bucket...)
--
Jeff Trawick | trawick@ibm.net | PGP public key at web site:
http://www.geocities.com/SiliconValley/Park/9289/
Born in Roswell... married an alien...
Re: Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.cmod_cgid.c mod_include.c mod_mime_magic.c
Posted by Bill Stoddard <st...@raleigh.ibm.com>.
Since you have checked in a pipe_bucket (which is needed to fix this
correctly on Windows), I remove my veto. In the future, I would
appreciate it if you made more of an effort to keep the code buildable.
Removing IOLs w/o having a suitable replacement was not cool.
Bill
----- Original Message -----
From: Bill Stoddard <st...@raleigh.ibm.com>
To: <ne...@apache.org>
Sent: Wednesday, August 23, 2000 2:45 PM
Subject: Re: Fw: cvs commit: apache-2.0/src/modules/standard
mod_cgi.cmod_cgid.c mod_include.c mod_mime_magic.c
> > This patch was posted on Saturday, and nobody complained about it.
How
> > does this break Windows?
>
> You cannot use socket routines to do pipe i/o. That is what your hack
presumes unless i
> am missing something.
>
> > I plan on adding minimal pipe and file buckets in the next day or
two. Can
> > we work around the problems until I can finish those bucket types?
> >
> I don't think so.
>
> Bill
>
>
Re: Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.cmod_cgid.c mod_include.c mod_mime_magic.c
Posted by Bill Stoddard <st...@raleigh.ibm.com>.
> This patch was posted on Saturday, and nobody complained about it. How
> does this break Windows?
You cannot use socket routines to do pipe i/o. That is what your hack presumes unless i
am missing something.
> I plan on adding minimal pipe and file buckets in the next day or two. Can
> we work around the problems until I can finish those bucket types?
>
I don't think so.
Bill
Re: Fw: cvs commit: apache-2.0/src/modules/standard mod_cgi.c
mod_cgid.c mod_include.c mod_mime_magic.c
Posted by rb...@covalent.net.
This patch was posted on Saturday, and nobody complained about it. How
does this break Windows? Is it a compile break or a run-time break? I
plan on adding minimal pipe and file buckets in the next day or two. Can
we work around the problems until I can finish those bucket types?
Ryan
On Wed, 23 Aug 2000, Bill Stoddard wrote:
> -1 (that's a veto) on this patch. It breaks Apache on Windows for the sole purpose of
> 'removing IOLs'. Removing IOLs is not a goal. When enough filtering infrastructure is in
> to allow IOLs to be replaced, then go for it.
>
> Bill
>
> > rbb 00/08/22 17:01:58
> >
> > Modified: src CHANGES
> > src/include ap_mpm.h buff.h
> > src/main Makefile.in buff.c http_core.c http_protocol.c
> > src/modules/mpm/beos beos.c
> > src/modules/mpm/dexter dexter.c
> > src/modules/mpm/mpmt_beos mpmt_beos.c
> > src/modules/mpm/mpmt_pthread mpmt_pthread.c
> > src/modules/mpm/perchild perchild.c
> > src/modules/mpm/prefork prefork.c
> > src/modules/mpm/spmt_os2 spmt_os2.c
> > src/modules/mpm/winnt mpm_winnt.c
> > src/modules/proxy proxy_connect.c proxy_ftp.c proxy_http.c
> > src/modules/standard mod_cgi.c mod_cgid.c mod_include.c
> > mod_mime_magic.c
> > Log:
> > Remove IOLs from Apache. They are no longer necessary, now that we have
> > filtering beginning to work. There is a hack that has been repeated
> > through this patch, we morph a pipe into a socket, and put the socket
> > into the BUFF. Everytime we do that, we are working with a pipe from
> > a CGI, and we should be creating a pipe bucket and passing that bucket
> > back. Because we don't actually have pipe buckets yet, we are using this
> > hack. When we get pipe buckets, this will be fixed
>
_______________________________________________________________________________
Ryan Bloom rbb@apache.org
406 29th St.
San Francisco, CA 94131
-------------------------------------------------------------------------------