Ryan Bloom [Tue, 13 Nov 2001 07:15:36 +0000 (07:15 +0000)]
This allows modules to add socket descriptors to the pollset. I have
only added this to the perfork MPM, and the others work without it.
Tomorrow I will add it to the other MPMs.
Ryan Bloom [Tue, 13 Nov 2001 06:04:25 +0000 (06:04 +0000)]
Remove commas from the end of the macros that define
directives that are used by MPMs. Previous to this patch,
you would use these macros without commans, which was unlike
any other directives. Now, after the macro, you must have
a comma. This makes the macros look more like the rest of the
directives.
I know this is cosmetic, and I was going to leave it alone, but when
I found out that it bothered Cliff too, I decided to fix it after all.
Ryan Bloom [Tue, 13 Nov 2001 02:09:07 +0000 (02:09 +0000)]
Cleanup some code that was created during the abstration. This basically
takes the old ap_new_connection, and puts into the new core_create_conn
function. There is no good reason to have two functions to do this.
Ryan Bloom [Tue, 13 Nov 2001 01:24:42 +0000 (01:24 +0000)]
Remove more network logic from the MPMs. Essentially, each MPM can define
AP_MPM_DISABLE_NAGLE_ACCEPTED_SOCK, which tells Apache to disable_nagle
on the accepted socket.
Ryan Bloom [Mon, 12 Nov 2001 23:49:08 +0000 (23:49 +0000)]
Begin to abstract out the underlying transport layer.
The first step is to remove the socket from the conn_rec,
the server now lives in a context that is passed to the
core's input and output filters. This forces us to be very
careful when adding calls that use the socket directly,
because the socket isn't available in most locations.
Doug MacEachern [Mon, 12 Nov 2001 22:01:14 +0000 (22:01 +0000)]
fix segv triggered by recent ap_lingering_close change
need to set SSLFilterRec.pssl = NULL when ssl_hook_CloseConnection is called
otherwise, ap_lingering_close -> ap_flush_conn will call ssl_io_filter_Output
which thinks it can still use the SSLFilterRec.pssl
PR:
Obtained from:
Submitted by:
Reviewed by:
Jeff Trawick [Mon, 12 Nov 2001 17:50:59 +0000 (17:50 +0000)]
Really reset the MaxClients value in worker when the
configured value is not a multiple of the number of threads
per child. We said we did previously but we forgot to.
Aaron Bannert [Mon, 12 Nov 2001 08:40:33 +0000 (08:40 +0000)]
Keep track of this so I don't forget. Not entirely sure how to go about
doing this properly, but I'll figure it out unless someone else beats
me to it.
Cliff Woolley [Mon, 12 Nov 2001 04:34:13 +0000 (04:34 +0000)]
Port Justin's mod_cgid change to mod_cgi.
mod_cgi wasn't actually broken in the way that mod_cgid was, but
it's still good do have these checks be against APR_SUCCESS rather
than 0 for clarity.
Cliff Woolley [Mon, 12 Nov 2001 03:23:12 +0000 (03:23 +0000)]
Remove the returned-length "w" parameter from apr_bucket_heap_create()
and apr_bucket_heap_make(). It was useless, since the length is invariant
from the length passed in and from the resulting bucket's b->length. This
takes care of a long-standing issue first brought up in February and
discussed on the dev@apr list. (Issue #2 from the "Bucket API Cleanup
Issues" thread.)
See http://marc.theaimsgroup.com/?l=apr-dev&m=98324983126666&w=2
Ian Holsman [Sun, 11 Nov 2001 22:31:04 +0000 (22:31 +0000)]
This patch changes the apr_table_elts macro so that it provides
access to the internals of an apr_table_t via a const pointer
instead of the current non-const pointer.
Submitted by: Brian Pane <BPane@pacbell.net>
Reviewed by: Ian Holsman
Ian Holsman [Sun, 11 Nov 2001 22:12:25 +0000 (22:12 +0000)]
This patch zero-fills just the integer and pointer fields in
the structure, plus the first byte of each of the string buffers.
This updated version of the patch doesn't allocate space for
the error_str and time_str buffers in the mod_include filter
context until/unless they're actually needed.
--Brian
Submitted by: Brian Pane <bpane@pacbell.net>
Reviewed by: Ian Holsman
apr_file_gets returns an apr_status_t not a char* and it returns
APR_SUCCESS when it reads something (which is 0).
Two of the cases were doing while apr_file_gets > 0 which would cause
it to loop when it returned APR_EOF. So, the valid check here is to
loop while we are receiving APR_SUCCESS.
Fix all of the other apr_file_gets to check APR_SUCCESS explicitly so
that it is obvious that we are checking an apr_status_t. Yes,
0 == APR_SUCCESS, but it obviously wasn't clear to someone what it
was returning.
Ryan Bloom [Sat, 10 Nov 2001 21:07:13 +0000 (21:07 +0000)]
Fix the Windows MPM. Windows doesn't always use the lingering close
function. If it can re-use the socket, we are better off not calling
the function. To fix this, we re-expose the lingering_close function,
and we allow the MPM to remove the cleanup.
Ryan Bloom [Sat, 10 Nov 2001 18:26:30 +0000 (18:26 +0000)]
Remove ap_lingering_close from all of the MPMs. This is now done as
a cleanup registered with the connection_pool. I have also turned
ap_lingering_close into a static function, because it is only used
in connection.c. This is the next step to consolidating all of the
socket function calls. ap_lingering_close will only be added if the
core is dealing with a standard socket.
Stat, don't lstat the final target. This means we may double-stat the
final target file if check for symlinks reveals it's an APR_LNK, but this
is preferable to the convoluted code required to 'reuse' the original stat.
Reintroduce the 'one stat dir_walk', with singificantly more sensible
behavior than the old path_info logic.
If the stat() of r->filename succeeds ignore every segment that requires
no symlink check (and on Win32/OS2/Netware the name matches the canonical
name, assuring us that the case/aliases match the true name.)
This optimization can be improved by establishing a second cache of the
actual partial filenames that _are_ tested for symlinks (or canonical
filename case), and then skiping such tests when the conditions match on
successive passes for subrequests or redirects.
Clean up the declarations and comments of some of the variables in
ap_get_client_block.
There were two len_read vars (each of different types), so the outer one
was shadowed - move it to the correct scope.
Move apr_bucket* old declaration inside of the loop so it is obvious
what its scope is.
Martin Kraemer [Thu, 8 Nov 2001 14:29:37 +0000 (14:29 +0000)]
Fixed previous patch to reflect what Apache-1.3 did regarding self-
referential uri's, and updated the manual accordingly.
XXX There is a kludge here: XXX the port number from the client's Host:
header used to be tossed, and there is no clean mechanism to pass it
(in the request_rec) to other consumers. As the unparsed_uri structure
(which could avoid repeated parsing of URI, Host, Port etc) seems to be
mostly unused currently, I used that to pass the port.
Martin Kraemer [Thu, 8 Nov 2001 12:34:21 +0000 (12:34 +0000)]
When no port is given in a "ServerName host" directive, the
server_rec->port is now set to zero, not 80. That allows for
run-time deduction of the correct server port (depending on
SSL/plain, and depending also on the current setting of
UseCanonicalName). This change makes redirections
work, even with https:// connections. [Martin Kraemer]
David Reid [Thu, 8 Nov 2001 00:54:00 +0000 (00:54 +0000)]
This doesn't feel quite right, but it doesn't cause any problems
and stops an irritation about the standard conf file for
BeOS which barfs if the User/Group is present.
Submitted by: Brad Froehle <epEnochRoot@yahoo.com>
Reviewed by: David Reid <dreid@apache.org>
Aaron Bannert [Wed, 7 Nov 2001 19:59:51 +0000 (19:59 +0000)]
Please do not mix spaces and tabs within the same blocks, and please do not
arbitrarily and inconsistently wrap lines. I apologise ahead of time for
mucking up the cvslog, but this needed to be done.
Greg Ames [Wed, 7 Nov 2001 15:41:43 +0000 (15:41 +0000)]
Insure that ap_get_client_block _always_ sets the length parm positive when
calling ap_get_brigade. This takes care of the loop with short request bodies
seen on 2.0.27.
This eliminates the use of r->remaining within ap_get_client_block.
http_filter has the equivalent in its context. r->remaining is
still used in ap_setup_client_block for a "body too long" check. But that
check can't cover chunked bodies, so it probably should happen elsewhere.
Doug MacEachern [Wed, 7 Nov 2001 05:41:22 +0000 (05:41 +0000)]
ap_getline() is useful to modules outside of the core..
change declaration from AP_CORE_DECLARE to AP_DECLARE so it can be used
used outside the core
PR:
Obtained from:
Submitted by:
Reviewed by:
Joshua Slive [Tue, 6 Nov 2001 19:09:51 +0000 (19:09 +0000)]
Put the ssl docs in a format where they will be maintainable without Ralf's
tools. Mostly I've only touched the top and the bottom of each doc.
I also got rid of the old cover page and the overview, since they seemed
obsolete.
Once again, I apologize for wrecking Ralf's nice work.
htdbm provides a compiled interface to dbm authentication databases,
using the compiled-in dbm manager and auth schema. Resolves the hassles
of grabbing the appropriate cpan package for md5 and fixing the dbm
engine for dbmmanage.pl.
Joshua Slive [Tue, 6 Nov 2001 16:06:13 +0000 (16:06 +0000)]
I am just moving the reference chapter of the ssl docs into the modules directory
to be the baseline. The next commit will put it in the correct format.
projects / apache / log