PATCHES ACCEPTED TO BACKPORT FROM TRUNK:
[ start all new proposals below, under PATCHES PROPOSED. ]
+ *) When more than one module calls APACHE_CHECK_OPENSSL, MOD_LDFLAGS and MOD_CFLAGS
+ remain unset for second or later caller. Fix this by caching the values for
+ MOD_LDFLAGS and MOD_CFLAGS created during the first run and set them in the
+ cached case. Change by rpluem
+ trunk patch: http://svn.apache.org/r1725325
+ 2.4.x patch: trunk works
+ +1: icing, jim, ylavic
+
+ *) mod_ssl: Fix missing TLS Upgrade/Connection headers on OPTIONS * requests
+ (Note that this is simply a necessary band-aid, and still won't allow h2c
+ to peacefully coexist with tls/n.n upgrades, because Upgrade phase needs
+ to come earlier, not as a handler, and the bogus connection-close behavior
+ must be removed before the Protocol API can handle TLS upgrade.)
+ PR58688
+ trunk patch: http://svn.apache.org/r1717816
+ +1: wrowe, icing, ylavic
+
+ *) mod_ssl: handle APR_TIMEUP on empty input by keeping connection state valid
+ for later retries.
+ trunk patch: http://svn.apache.org/r1725940
+ +1: icing, jim, ylavic
+
+ *) Add APLOGNO, first chunk (those that were detected by coccinelle).
+ There are some more but they are easier to backport once these here are applied.
+ trunk patch: http://svn.apache.org/r1725392 (modules)
+ http://svn.apache.org/r1725394 (server/mpm)
+ http://svn.apache.org/r1725395 (server others)
+ http://svn.apache.org/r1725468 (remove APLOGNO from mod_log_debug custom messages)
+ 2.4.x patch: http://people.apache.org/~rjung/patches/aplogno-backport-r1725392.patch
+ http://people.apache.org/~rjung/patches/aplogno-backport-r1725394.patch
+ http://people.apache.org/~rjung/patches/aplogno-backport-r1725395.patch
+ http://people.apache.org/~rjung/patches/aplogno-backport-r1725468.patch
+ +1: rjung, jim, ylavic
+ icing: needs http://svn.apache.org/r1725468
+ rjung: You are right, I added r1725468 and restarted the vote.
+
+ *) core: Prevent a server crash in case of an invalid CONNECT request with
+ a custom error page for status code 400 that uses server side includes.
+ PR 58929
+ Trunk version of patch:
+ http://svn.apache.org/r1727544
+ Backport version for 2.4.x of patch:
+ Trunk version of patch works modulo CHANGES
+ +1: rpluem, covener, ylavic
+
+ *) core: Add expression support to SetHandler (someone wanted this in
+ a well-articulated serverfault post).
+ trunk patch: http://svn.apache.org/r1725149
+ http://svn.apache.org/r1725151
+ http://svn.apache.org/r1726233
+ http://svn.apache.org/r1729374
+ 2.4.x patch: MMN/changes only: http://people.apache.org/~covener/patches/2.4.x-sethandler-expr-2.diff
+ + http://svn.apache.org/r1729374
+ +1: covener, jim, ylavic
+
+ *) core: introducing new hook "pre_close_connection" to give protocols other
+ than http/1.1 a chance to send one last frame before close.
+ Requires MMN bump
+ trunk patch: http://svn.apache.org/r1727071
+ http://svn.apache.org/r1727573
+ 2.4.x patch: trunk works (modulo CHANGES)
+ +1: icing, covener, ylavic
+ covener: needs MMN bump, but sorting out previous MMN question first.
PATCHES PROPOSED TO BACKPORT FROM TRUNK:
to silence this warning
jailletc36: This has been done in r1711919
ylavic: Also, I don't think the changes on worker/event are needed,
- prefork only requires the accept mutex in ONE_PROCESS mode.
-
- *) mod_ssl: Fix missing TLS Upgrade/Connection headers on OPTIONS * requests
- (Note that this is simply a necessary band-aid, and still won't allow h2c
- to peacefully coexist with tls/n.n upgrades, because Upgrade phase needs
- to come earlier, not as a handler, and the bogus connection-close behavior
- must be removed before the Protocol API can handle TLS upgrade.)
- PR58688
- trunk patch: http://svn.apache.org/r1717816
- +1: wrowe, icing
- ylavic: how about adding !ap_request_has_body(r) to the test then?
- wrowe responds: nonstarter, as discussed on list.
+ only prefork requires the accept mutex in ONE_PROCESS mode.
*) mod_ssl: Free dhparams when getting DH params. This fixes issue when
SSLCryptoDevice does not get unregistered because of non-zero refcount
trunk patch: http://svn.apache.org/r1720129
2.4.x patch: http://jkaluza.fedorapeople.org/httpd-2.4.x-dhparams-free.patch
+1: jkaluza
+ ylavic: +1 with r1723295 (likewise for ecparams).
*) mod_cache_socache: Fix a possible cached entity body corruption when it
is received from an origin server in multiple batches and forwarded by
2.4.x patch: trunk works, module CHANGES
+1: ylavic, jim
- *) core: Add expression support to SetHandler (someone wanted this in
- a well-articulated serverfault post).
- trunk patch: http://svn.apache.org/r1725149
- http://svn.apache.org/r1725151
- http://svn.apache.org/r1726233
- http://svn.apache.org/r1729374
- 2.4.x patch: MMN/changes only: http://people.apache.org/~covener/patches/2.4.x-sethandler-expr-2.diff
- + http://svn.apache.org/r1729374
- +1: covener, jim
-
- *) Add APLOGNO, first chunk (those that were detected by coccinelle).
- There are some more but they are easier to backport once these here are applied.
- trunk patch: http://svn.apache.org/r1725392 (modules)
- http://svn.apache.org/r1725394 (server/mpm)
- http://svn.apache.org/r1725395 (server others)
- http://svn.apache.org/r1725468 (remove APLOGNO from mod_log_debug custom messages)
- 2.4.x patch: http://people.apache.org/~rjung/patches/aplogno-backport-r1725392.patch
- http://people.apache.org/~rjung/patches/aplogno-backport-r1725394.patch
- http://people.apache.org/~rjung/patches/aplogno-backport-r1725395.patch
- http://people.apache.org/~rjung/patches/aplogno-backport-r1725468.patch
- +1: rjung, jim
- icing: needs http://svn.apache.org/r1725468
- rjung: You are right, I added r1725468 and restarted the vote.
-
- *) mod_ssl: handle APR_TIMEUP on empty input by keeping connection state valid
- for later retries.
- trunk patch: http://svn.apache.org/r1725940
- +1: icing, jim
-
- *) When more than one module calls APACHE_CHECK_OPENSSL, MOD_LDFLAGS and MOD_CFLAGS
- remain unset for second or later caller. Fix this by caching the values for
- MOD_LDFLAGS and MOD_CFLAGS created during the first run and set them in the
- cached case. Change by rpluem
- trunk patch: http://svn.apache.org/r1725325
- 2.4.x patch: trunk works
- +1: icing, jim
-
- *) core: introducing new hook "pre_close_connection" to give protocols other
- than http/1.1 a chance to send one last frame before close.
- Requires MMN bump
- trunk patch: http://svn.apache.org/r1727071
- http://svn.apache.org/r1727573
- 2.4.x patch: trunk works (modulo CHANGES)
- +1: icing, covener
- covener: needs MMN bump, but sorting out previous MMN question first.
-
- *) core: Prevent a server crash in case of an invalid CONNECT request with
- a custom error page for status code 400 that uses server side includes.
- PR 58929
- Trunk version of patch:
- http://svn.apache.org/r1727544
- Backport version for 2.4.x of patch:
- Trunk version of patch works modulo CHANGES
- +1: rpluem, covener
-
*) mod_ssl: Add SSLOCSPProxyURL to add the possibility to do all queries
to OCSP responders through a HTTP proxy.
Trunk version of patch:
projects / thirdparty / apache / httpd.git / commitdiff
