httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "William A. Rowe, Jr." <>
Subject Re: svn commit: r208787 - in /httpd/httpd/trunk/modules: http/http_filters.c http/http_protocol.c proxy/mod_proxy.c
Date Thu, 07 Jul 2005 04:53:30 GMT
At 06:21 PM 7/6/2005, Justin Erenkrantz wrote:
>--On July 4, 2005 11:08:18 PM -0500 "William A. Rowe, Jr." <>
>>> It still is not 'correct' until REQUEST_CHUNKED_PASS is reimplemented
>>> and passes some chunk headers, since we aren't echoing the entire
>>> request.  But it gets me further on testing 1.3 -> 2.0 -> 2.1 -> 2.0
>>> 1.3 proxy behaviors.
>As I've said before, REQUEST_CHUNKED_PASS needs to go away in 2.x.  The correct 'fix'
is to remove the CHUNK input filter and then call ap_get_brigade().  But, calling ap_should_client_block,
ap_get_client_block(), etc. is just flat-out wrong for 2.x.  -- justin

Actually, in this odd case you are wrong.  There is no CHUNKED_PASS
in 2.x today, what you desire was already accomplished.

And I agree that a ap_get_brigade structure would be better.  I just
haven't had time to recode to 2.x.

And I agree that in httpd 3.x we eliminate ap_XXX_client_block.

But CHUNKED_PASS is *exactly* the behavior we want if we are
trying to backtrace how a back-end server observed a request
body.  We don't want the pretty output, we want the raw gorey
details, literally.  Roy has previously complained that we do
too much massaging before spitting back out the TRACE result.

View raw message