Joshua Slive [Wed, 13 Nov 2002 18:20:32 +0000 (18:20 +0000)]
A couple long-overdue changes to configuring:
- Remove the over-detailed and out-of-date description
of Include directories.
- Link Rich's .htaccess tutorial.
Greg Ames [Wed, 13 Nov 2002 15:07:35 +0000 (15:07 +0000)]
core_output_filter: use the current input brigade's pool for the "more"
brigade when we defer network writes due to small data + keepalives. This
prevents leaking resources in the "more" brigade until the end of the
connection.
André Malo [Wed, 13 Nov 2002 00:13:24 +0000 (00:13 +0000)]
- <em> -> <var>
- reformatted examples (no content changes, except for a ytop)
- move the reference to mod_deflate on /top/ of the gzip example
- general markup improvement
- add short descriptions for the directives
Ken Coar [Tue, 12 Nov 2002 18:30:00 +0000 (18:30 +0000)]
Add a SERVER_ADDR keyword to match the CGI environment variable,
to allow conditional setting according to the IP address on
which the server received the request.
Jeff Trawick [Tue, 12 Nov 2002 12:23:42 +0000 (12:23 +0000)]
watch out for the should-not-occur condition where the daemon can't
find the script pid in the hash and returns zero... if we send
sigterm to pid zero, we knock out ourself
change the cleanup to return any error from cleanup_nonchild_process()
Jeff Trawick [Mon, 11 Nov 2002 23:07:29 +0000 (23:07 +0000)]
mod_cgid: Terminate CGI scripts when the client connection
drops.
This is done by mod_cgid daemon remembering the pid, which is queried
by the handler at the end of the request. The handler then
knocks out the CGI script (if it is indeed still running).
It would be slightly easier for the daemon to do this, but naughty
scripts which are hard to kill would hold up all new CGI requests
if the daemon stopped its normal processing to kill a script.
Greg Ames [Mon, 11 Nov 2002 19:57:58 +0000 (19:57 +0000)]
prevent seg faults when running with Electric Fence.
The "more" brigade (created by apr_brigade_split after seeing a FLUSH bucket)
was in the deferred_write_pool. d_w_p is cleared, "more" is copied into b,
then the while loop test for APR_BRIGADE_EMPTY(b) blows up.
André Malo [Mon, 11 Nov 2002 03:13:54 +0000 (03:13 +0000)]
- add a forgotten CSS rule; examples in warnings also
get a border around (and no bgcolor)
- extend mod_deflate documentation
(better example, notes on proxies)
it still needs some fine tuning.
Joshua Slive [Sun, 10 Nov 2002 20:34:48 +0000 (20:34 +0000)]
Add a <parentdocument href="...">...</parentdocument> tag to
allow the docs to have a little more structure. Used for the
"back" link and for the breadcrumb links.
mod_isapi: Fix an issue where HSE_REQ_DONE_WITH_SESSION notification
is received before the HttpExtensionProc() returns HSE_STATUS_PENDING.
[John DeSetto <jdesetto@radiantsystems.com>, William Rowe]
Paul J. Reder [Fri, 8 Nov 2002 22:42:19 +0000 (22:42 +0000)]
Improves the user friendliness of the CacheRoot processing
over my last pass. This version avoids the pool allocations
but doesn't avoid all of the runtime checks. It no longer
terminates during post-config processing. An error is logged
once per worker, indicating that the CacheRoot needs to be set.
[Paul J. Reder]
Greg Ames [Fri, 8 Nov 2002 17:19:10 +0000 (17:19 +0000)]
use a subpool of c->pool for resources which are set aside, then clear it
after writing the data to the network. This closes files sooner with
keepalive connections.
Brian Pane [Fri, 8 Nov 2002 09:24:00 +0000 (09:24 +0000)]
When doing a GET of a proxied URL as a subrequest within
a POSTed request, don't send the original POST's Content-Length
as part of the header for the GET.
Paul J. Reder [Thu, 7 Nov 2002 23:11:10 +0000 (23:11 +0000)]
Added code to process min and max file size directives and to
init the expirychk flag in mod_disk_cache. Added a clarifying
comment to cache_util. [Paul J. Reder]
André Malo [Thu, 7 Nov 2002 03:34:36 +0000 (03:34 +0000)]
- <em> -> <var> (hey, that was easy :))
- add some <module>s
- add <seealso>s to suexec.html
- add explicit <seealso> from mod_cgid to mod_cgi
- add short description of the ScriptSock directive
-> update transformation
Rich Bowen [Thu, 7 Nov 2002 03:03:33 +0000 (03:03 +0000)]
As per discussion on users@httpd.apache.org and a little on IRC, clarify
that "third party" modules can do basically whatever they want, and
typically do it as the "User" uid, so Caveat Administrator
Rule one of winsock and other one-offs (even unix EINTR) ... blocking
isn't necessarily blocking. Should not have changed this in the prior
commit, and adding the same retry to the -1/EAGAIN|EINTR case.
errno? EINTR? what planet was this code on :-? Normalize the
ssl_io_filter_connect code to follow the filter read and write.
Notice that it's buck ugly, but we will extract an rc first from
the input BIO if it was written, and then try the output bio if
it was APR_SUCCESS, during _connect processing.
(I gotta think someone is committing bogus transformations as I can't
figure out how these transformations are valid and why they would be
changing right now. Yet, these need to stay in sync.)
André Malo [Wed, 6 Nov 2002 00:27:17 +0000 (00:27 +0000)]
formal changes:
- <em> -> <var>
- markup improvements
- allow class attribute for <li>
content changes:
- remove the references to apache 1.3 in IndexOptions descriptions
- clarify (prior non-existing) incremental index options example...
Merge the last of the 'filtering' functions into ssl_engine_io.c, merge
ssl_abort into what was ssl_hook_CloseConnection, clean out a bunch of
now-static or private headers from mod_ssl.h, and final fix a very small
but potent segfault if ->pssl is destroyed within our read loop.
Actually, the APR_ECONNABORTED (EOS-only brigade) is the direction we
are contemplating for the next release, not the prior behavior
(which was APR_SUCCESS for c->aborted.)
With a last little bit of help from Justin, this should cause the
appropriate amount of tumolt and turmoil if our client has 'gone away'
on us, sparing us of further processing (and potential 'renegotiations'
with a non-existant client.)
Appears we forgot to check the possibility of errors coming from the
write brigade passed down from the content generator through the body
and protocol filters.
Rename the many flavors of filter_ctx (pRec, fctx etc) to filter_ctx,
wbio to bio_out, BIO new and free to create and destroy (to match OpenSSL),
refactor the bio write code to stash errors in outctx->rc,
fix the blocking read at EOF if we have some data to return,
and preempt the nonblock read on GETLINE if we have the newline already.
Fix for PR 10216 ... we should not return the status unless we want the
core to ap_die for us. If we've handled the response (as determined
by a new response_sent flag) then the server does not need to do so.
Mr. Trawick was dead on, and this revealed a much bigger bug.
Factor out the opts/override merging (since we do it three times)
and eliminate all the nasty goto's. This bug likely caused all
sorts of dir_walk configuration issues including htaccess issues.
Also add a few more docs where things aren't so obvious.
Well that was a little borked, if you were having stack problems using
logio the missing pre_connection hook declartion could have something
to do with it.
projects / apache / log