From 8de9e7b15d1caa560aca38d191f736184703b172 Mon Sep 17 00:00:00 2001 From: Chris Darroch Date: Mon, 7 Jul 2008 19:42:42 +0000 Subject: [PATCH] propose a bit of a jumbo patch git-svn-id: https://svn.apache.org/repos/asf/httpd/httpd/branches/2.2.x@674610 13f79535-47bb-0310-9956-ffa450edef68 --- STATUS | 38 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 38 insertions(+) diff --git a/STATUS b/STATUS index e17321df642..1f1f091b1fc 100644 --- a/STATUS +++ b/STATUS @@ -168,6 +168,44 @@ PATCHES PROPOSED TO BACKPORT FROM TRUNK: http://svn.apache.org/viewvc?rev=673999&view=rev to this because otherwise the perl test framework breaks. + * core, authn/z: Avoid calling access control hooks for internal requests + with configurations which match those of initial request. Revert to + original behaviour (call access control hooks for internal requests + with URIs different from initial request) if any access control hooks or + providers are not registered as permitting this optimization. + Introduce wrappers for access control hook and provider registration + which can accept additional mode and flag data. Convert common + provider version strings to macros. + The core purpose of this pile of patches is to avoid unnecessary + authn/z hooks when a single request spawns large numbers of internal + requests to which an identical set of httpd configurations apply. + This permits modules such as mod_authn_dbd and mod_dav to work together + acceptably. + Because certain external modules such as mod_authz_svn rely on the old + behaviour, this optimization can be made only when all authn/z hooks and + providers are registered with the appropriate flag. + It would be excellent if Windows and NetWare people could ensure this + builds correctly. + In particular, mod_auth.h must be included into request.c and I've left + mod_auth.h under modules/aaa rather than try to replicate wrowe's work + in trunk moving all the include files around. + I'm open to suggestions that this remain in trunk only, but in that case, + it would be very helpful to know whether most people expect a 2.4 branch + or just a 3.0 branch to be next. If 3.0, some of the backwards + compatibility work could potentially be ditched. + Trunk version of patches: + http://svn.apache.org/viewvc?view=rev&revision=644525 + http://svn.apache.org/viewvc?view=rev&revision=644562 (trunk MMN bump) + http://svn.apache.org/viewvc?view=rev&revision=645395 + http://svn.apache.org/viewvc?view=rev&revision=645472 + http://svn.apache.org/viewvc?view=rev&revision=645540 + http://svn.apache.org/viewvc?view=rev&revision=646445 (reverted by r659160) + http://svn.apache.org/viewvc?view=rev&revision=658046 + http://svn.apache.org/viewvc?view=rev&revision=659160 + Backport version for 2.2.x of patch: + http://people.apache.org/~chrisd/patches/walk_cache/walk_cache-2.2.x.patch + +1: chrisd + PATCHES/ISSUES THAT ARE STALLED * beos MPM: Create pmain pool and run modules' child_init hooks when -- 2.47.2