1 APACHE 2.3 STATUS: -*-text-*-
2 Last modified at [$Date$]
4 The current version of this file can be found at:
6 * http://svn.apache.org/repos/asf/httpd/httpd/trunk/STATUS
8 Documentation status is maintained seperately and can be found at:
10 * docs/STATUS in this source tree, or
11 * http://svn.apache.org/repos/asf/httpd/httpd/trunk/docs/STATUS
13 Consult the following STATUS files for information on related projects:
15 * http://svn.apache.org/repos/asf/apr/apr/trunk/STATUS
16 * http://svn.apache.org/repos/asf/apr/apr-util/trunk/STATUS
18 Patches considered for backport are noted in their branches' STATUS:
20 * http://svn.apache.org/repos/asf/httpd/httpd/branches/1.3.x/STATUS
21 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.0.x/STATUS
22 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.2.x/STATUS
26 [NOTE that x.{odd}.z versions are strictly Alpha/Beta releases,
27 while x.{even}.z versions are Stable/GA releases.]
29 2.3.7 : In development.
30 2.3.6 : Tagged June 11, 2010.
31 2.3.5 : Released on January 26, 2010.
32 2.3.4 : Released on December 8, 2009.
33 2.3.3 : Tagged on November 11, 2009, not released.
34 2.3.2 : Tagged on March 23, 2009, not released.
35 2.3.1 : Tagged on January 2, 2009, not released.
36 2.3.0 : Tagged on December 6, 2008, not released.
38 Contributors looking for a mission:
40 * Just do an egrep on "TODO" or "XXX" in the source.
42 * Review the bug database at: http://issues.apache.org/bugzilla/
44 * Review the "PatchAvailable" bugs in the bug database:
46 https://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
48 After testing, you can append a comment saying "Reviewed and tested".
50 * Open bugs in the bug database.
53 CURRENT RELEASE NOTES:
60 * Modules that are not ready for production use should be marked as
61 experimental. Candidates:
67 * Modules without documentation need to be moved to experimental or be
70 * Not all MPMs are updated to set conn_rec::current_thread correctly.
71 (Prefork, Worker, Event, Simple are updated).
72 jim sez: Then we just ship with those... mark any others as
75 * Fix or remove ap_note_auth_failure():
76 There are two incompatible sets of *note_*_auth_failure functions, one in
77 server/protocol.c, the other in mod_auth_*.c. The set in server/protocol.c
78 should be axed and ap_note_auth_failure() must either call the functions in
79 mod_auth_*.c or must be removed, too.
84 OLD ISSUES THAT WERE THOUGHT TO BE SHOWSTOPPERS FOR 2.2 BUT OBVIOUSLY WEREN'T:
86 * Handling of non-trailing / config by non-default handler is broken
87 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105451701628081&w=2
88 jerenkrantz asks: Why should this block a release?
89 wsanchez agrees: this may be a change in behavior, but isn't
90 clearly wrong, and even if so, it doesn't seem like a
93 * the edge connection filter cannot be removed
94 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105366252619530&w=2
96 jerenkrantz asks: Why should this block a release?
98 stas replies: because it requires a rewrite of the filters stack
99 implementation (you have suggested that) and once 2.2 is
100 released you can't do that anymore.
102 pgollucci: this affects mod_perl I'm pretty sure.
106 * If the parent process dies, should the remaining child processes
107 "gracefully" self-terminate. Or maybe we should make it a runtime
108 option, or have a concept of 2 parent processes (one being a
110 See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
112 Self-destruct: Ken, Martin, Lars, sctemme (parent shouldn't die, ever)
113 Not self-destruct: BrianP, Ian, Cliff, BillS
114 Make it runtime configurable: Aaron, jim, Justin, wrowe, rederpj, nd, pgollucci
116 /* The below was a concept on *how* to handle the problem */
117 Have 2 parents: +1: jim
118 -1: Justin, wrowe, rederpj, nd, pgollucci
119 +0: Lars, Martin (while standing by, could it do
122 * Make the worker MPM the default MPM for threaded Unix boxes.
123 +1: Justin, Ian, Cliff, BillS, striker, wrowe, nd, pgollucci, sctemme
124 +0: BrianP, Aaron (mutex contention is looking better with the
125 latest code, let's continue tuning and testing), rederpj, jim
128 pquerna: Do we want to change this for *2.4*?
131 * Name the Server (version 2.4 or 3.0, depending on the final call)
132 Recent discussion indicates we should designate a (short name).
133 This is not yet a [Vote] - Your nominations please:
134 * Apache HTTP Server (httpd)
135 +1: sctemme (why mess with it?)
137 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
139 * Patches submitted to the bug database:
140 http://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
142 * Filter stacks and subrequests, redirects and fast redirects.
143 There's at least one PR that suffers from the current unclean behaviour
144 (which lets the server send garbage): PR 17629
145 nd says: Every subrequest should get its own filter stack with the
146 subreq_core filter as bottom-most. That filter does two things:
147 - swallow EOS buckets
148 - redirect the data stream to the upper request's (rr->main)
149 filter chain directly after the subrequest's starting
151 Once we have a clean solution, we can try to optimize
152 it, so that the server won't be slow down too much.
154 * RFC 2616 violations.
155 Closed PRs: 15852, 15857, 15859, 15861, 15864, 15869, 15870, 16120,
156 16125, 16135, 16136, 16137, 16138, 16139, 16140, 16518,
158 Open PRs: 15865, 15866, 15868, 16126, 16133, 16142, 16521
159 jerenkrantz says: need to decide how many we need to backport and/or
160 if these rise to showstopper status.
161 wrowe suggests: it would be nice to see "MUST" v.s. "SHOULD" v.s. "MAY"
162 out of this list, without reviewing them individually.
164 * There is a bug in how we sort some hooks, at least the pre-config
165 hook. The first time we call the hooks, they are in the correct
166 order, but the second time, we don't sort them correctly. Currently,
167 the modules/http/config.m4 file has been renamed to
168 modules/http/config2.m4 to work around this problem, it should moved
169 back when this is fixed.
171 OtherBill offers that this is a SERIOUS problem. We do not sort
172 correctly by the ordering arguments passed to the register hook
173 functions. This was proven when I reordered the open_logs hook
174 to attempt to open the error logs prior to the access logs. Possibly
175 the entire sorting code needs to be refactored.
177 * pipes deadlock on all platforms with limited pipe buffers (e.g. both
178 Linux and Win32, as opposed to only Win32 on 1.3). The right solution
179 is either GStein's proposal for a "CGI Brigade", or OtherBill's proposal
180 for "Poll Buckets" for "Polling Filter Chains". Or maybe both :-)
182 * All handlers should always send content down even if r->header_only
183 is set. If not, it means that the HEAD requests don't generate the
184 same headers as a GET which is wrong.
186 * exec cmd and suexec arg-passing enhancements
187 Status: Patches proposed
188 Message-ID: <20020526041748.A29148@prodigy.Redbrick.DCU.IE>
189 (see the "proc.patch" and "suexec-shell.patch" links in this message)
191 * The 2.0.36 worker MPM graceless shutdown changes work but are
192 a bit clunky on some platforms; eg, on Linux, the loop to
193 join each worker thread seems to hang, and the parent ends up
194 killing off the child with SIGKILL. But at least it shuts down.
196 chrisd: Has this been fixed by the changes for PR 38737?
198 * We do not properly substitute the prefix-variables in the configuration
199 scripts or generated-configs. (i.e. if sysconfdir is etc,
200 httpd-std.conf points to conf.)
202 * If any request gets through ap_process_request_internal() and is
203 scheduled to be served by the core handler, without a flag that this
204 r->filename was tested by dir/file_walk, we need to 500 at the very
205 end of the ap_process_request_internal() processing so sub_req-esters
206 know this request cannot be run. This provides authors of older
207 modules better compatibility, while still improving the security and
210 Status: still need to decide where this goes, OtherBill comments...
211 Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
212 [Deleted comments regarding the ap_run_handler phase, as irrelevant
213 as BillS points out that "common case will be caught in
214 default_handler already (with the r->finfo.filetype == 0 check)"
215 and the issue is detecting this -before- we try to run the req.]
217 gregames says: can this happen somehow without a broken module
218 being involved? If not, why waste cycles trying to defend against
219 potential broken modules? It seems futile.
220 wrowe counters: no, it shouldn't happen unless the module is broken.
221 But the right answer is to fail the request up-front in dir/file
222 walk if the path was entirely invalid; and we can't do that either
223 UNTIL 2.1 or we break modules that haven't hooked map_to_storage.
225 * Can a static httpd be built reliably?
226 Message-ID: <20020207142751.T31582@clove.org>
228 * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
231 <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
232 Jeff wonders if we still care about this. It is no longer an
233 API issue but simply an extra trip through the brigade.
235 * Try to get libtool inter-library dependency code working on AIX.
236 Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
238 Justin says: If we get it working on AIX, we can enable this
239 on all platforms and clean up our build system somewhat.
240 Jeff says: I thought I tested a patch for you sometime in
241 January that you were going to commit within a few days.
243 * Handling of %2f in URIs. Currently both 1.3 and 2.0
244 completely disallow %2f in the request URI path (see
245 ap_unescape_url() in util.c). It's permitted and passed
246 through in the query string, however. Roy says the
247 original reason for disallowing it, from five years ago,
248 was to protect CGI scripts that applied PATH_INFO to
249 a filesystem location and which might be tricked by
250 ..%2f..%2f(...). We *should* allow path-info of the
251 form 'http://foo.com/index.cgi/path/to/path%2finfo'.
252 Since we've revamped a lot of our processing of path
253 segments, it would be nice to allow this, or at least
254 allow it conditionally with a directive.
256 OtherBill adds that %2f as the SECOND character of a multibyte
257 sequence causes the request to fail! This happens notably in
260 * There is increasing demand from module writers for an API
261 that will allow them to control the server à la apachectl.
262 Reasons include sole-function servers that need to die if
263 an external dependency (e.g., a database) fails, et cetera.
264 Perhaps something in the (ever more abused) scoreboard?
266 On the other hand, we already have a pipe that goes between parent
267 and child for graceful shutdown events, along with an API that
268 can be used to send a message down that pipe. In threaded MPMs,
269 it is easy enough to make that one pipe be used for graceful
270 and graceless events, and it is also easy to open that pipe
271 to both parent and child for writing. Then we just need to
272 figure out how to do graceless on non-threaded MPMs.
274 * Allow the DocumentRoot directive within <Location > scopes? This
275 allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
276 by a <Directory /somepath/foo> to become simply
277 <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
278 and in-your-face.) DocumentRoot unset would be accepted [and would
279 not permit content to be served, only virtual resources such as
280 server-info or server-status.
281 This proposed change would _not_ depricate Alias.
282 striker: See the thread starting with Message-ID:
283 JLEGKKNELMHCJPNMOKHOGEEJFBAA.striker@apache.org.
285 * Win32: Rotatelogs sometimes is not terminated when Apache
286 goes down hard. FirstBill was looking at possibly tracking the
287 child's-child processes in the parent process.
288 stoddard: Shared scoreboard might offer a good way for the parent
289 to keep track of 'other child' processes and whack them if the child
291 Other thoughts on walking the process chain using the NT kernel
292 have also been proposed on APR.
294 * Eliminate unnecessary creation of pipes in mod_cgid
296 * Combine log_child and piped_log_spawn. Clean up http_log.c.
299 * Platforms that do not support fork (primarily Win32 and AS/400)
300 Architect start-up code that avoids initializing all the modules
301 in the parent process on platforms that do not support fork.
303 * There are still a number of places in the code where we are
304 losing error status (i.e. throwing away the error returned by a
305 system call and replacing it with a generic error code)
307 * Mass vhosting version of suEXEC.
309 * All DBMs suffer from confusion in support/dbmmanage (perl script) since
310 the dbmmanage employs the first-matched dbm format. This is not
311 necessarily the library that Apache was built with. Aught to
312 rewrite dbmmanage upon installation to bin/ with the proper library
313 for predictable mod_auth_dbm administration.
314 Questions; htdbm exists, time to kill dbmmanage, or does it remain
315 useful as a perl dbm management example? If we keep it,
316 do we address the issue above?
319 Some additional items remaining:
320 - case_preserved_filename stuff
321 (use the new canonical name stuff?)
322 - find a new home for ap_text(_header)
323 - is it possible to remove the DAV: namespace stuff from util_xml?
325 * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
326 are a bit wonky. The function should probably be exposed as a utility
327 function (such as ap_translate_url2fs() or ap_validate_fs_url() or
328 something). Another approach would be a new hook phase after
329 "translate" which would allow the module to munge what the
330 translation has decided to do.
331 Status: Greg +1 (volunteers)
333 * Explore use of a post-config hook for the code in http_main.c which
334 calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
335 ap_sort_hooks() [to reduce the logic in main()]
337 * read the config tree just once, and process N times (as necessary)
339 * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
341 * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
342 into a VirtualHost container) to 2.0.
344 * shift stuff to mod_core.h
346 * callers of ap_run_create_request() should check the return value
347 for failure (Doug volunteers)
349 * Fix the worker MPM to use POD to kill child processes instead
350 of ap_os_killpg, regardless of how they should die.
352 chrisd: Is this done, by any chance? See r92598 and r93358.
354 * Scoreboard structures could be changed in the future such that
355 proper alignment is not maintained, leading to segfaults on
356 some systems. Cliff posted a patch to deal with this issue but
357 later recanted. See this message to dev@apr.apache.org:
359 <Pine.LNX.4.44.0203011354090.16457-200000@deepthought.cs.virginia.edu>
361 * APXS either needs to be fixed completely for use when apr is out of tree,
362 or it should drop query mode altogether, and we just grow an
363 httpd-config or similar arrangement.
364 To quote a discussion in STATUS earlier:
366 thommay: this doesn't fix all the problems with apxs and out of
367 tree apr/apr-util, but it's a good start. There's still the
368 query cases; but I'm beginning to think that in these cases
369 the app should be querying ap{r,u}-config directly
370 deprecate -q: add htpd-config: gstein, pquerna, minfrin, pgollucci
373 TODO ISSUES REMAINING IN MOD_SSL:
375 * Do we need SSL_set_read_ahead()?
377 * the ssl_expr api is NOT THREAD SAFE. race conditions exist:
378 -in ssl_expr_comp() if SSLRequire is used in .htaccess
379 (ssl_expr_info is global)
380 -is ssl_expr_eval() if there is an error
381 (ssl_expr_error is global)
383 * SSLRequire directive (parsing of) leaks memory
385 * Diffie-Hellman-Parameters for temporary keys are hardcoded in
386 ssl_engine_dh.c, while the comment in ssl_engine_kernel.c says:
387 "it is suggested that keys be changed daily or every 500
388 transactions, and more often if possible."
390 * ssl_var_lookup could be rewritten to be MUCH faster
392 * CRL callback should be pluggable
394 * session cache store should be pluggable
396 * init functions should return status code rather than ssl_die()
398 * ssl_engine_pphrase.c needs to be reworked so it is generic enough
399 to also decrypt proxy keys
401 * output warning when allowing SSL v2.0 ? its so old
404 * mod_proxy: Ability to run SSL over proxy gateway connections,
405 encrypting (or reencrypting) at the proxy.
407 * mod_cache: Handle ESI tags.
409 * mod_cache: Resolve issue of how to cache page fragments (or perhaps
410 -if- we want to cache page fragments). Today, mod_cache/mod_mem_cache
411 will cache #include 'virtual' requests (but not #include 'file'
412 requests). This was accomplished by making CACHE_IN a
413 CONTENT_SET-1 filter to force it to run before the SUBREQ_CORE
414 filter. But now responses cannot be cached that include the
415 effects of having been run through CONTENT_SET filters
416 (mod_deflate, mod_expires, etc). We could rerun all the
417 CONTENT_SET filters on the cached response, but this will not
418 work in all cases. For example, mod_expires relies on installing
419 the EXPIRATION filter during fixups. Contents served out of
420 mod_cache (out of the quick_handler) bypass -all- the request
421 line server hooks (Ryan really hated this. It is great for
422 performance, but bad because of the complications listed above).
424 mod_cache/mod_mem_cache/mod_disk_cache:
426 * mod_mem_cache: Consider adding a RevalidateTimeout directive to
427 specify time at which local cached content is to be revalidated
428 (ie, underlying file stat'ed to see if it has changed).
430 * mod_cache: CacheEnable/CacheDisable should accept regular expressions.
431 jerenkrantz says: Too slow. Get regexs away from speedy caches by
432 default. Introduce a new CacheEnableRegex if you want.
434 * mod_mem_cache/mod_disk_cache: Need to be able to query cache
435 status (num of entries, cache object properties, etc.).
436 mod_status could be extended to query optional hooks defined
437 by modules for the purpose of reporting module status.
438 mod_cache (et. al.) could define optional hooks that are called
439 to collect status. Status should be queryable by
441 jerenkrantz says: Yawn. Who cares.
443 * MaxRequestsPerChild measures connections, not requests.
444 Until someone has a better way, we'll probably just rename it
445 "MaxConnectionsPerChild".
447 * Regex containers don't work in an intutive way
448 Status: No one has come up with an efficient way to fix this
449 behavior. Dean has suggested getting rid of regex containers
451 OtherBill suggests: We at least seem to agree on eliminating
452 the <Container ~ foo> forms, and using only
453 <ContainerMatch foo> semantics.
455 * orig_ct in the byterange/multipart handling may not be
456 needed. Apache 1.3 just never stashed "multipart" into
457 r->content_type. We should probably follow suit since the
458 byterange stuff doesn't want the rest of the code to see the
459 multipart content-type; the other code should still think it is
460 dealing with the <orig_ct> stuff.
461 Status: Greg volunteers to investigate (esp. since he was most
462 likely the one to break it :-)
464 EXPERIMENTAL MODULES:
466 Experimental modules should eventually be be promoted to fully supported
467 status or removed from the repository entirely (ie, the
468 'experiment' failed). This section tracks what needs to happen to
469 get the modules promoted to fully supported status.