httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rainer Jung <>
Subject Re: [VOTE] Release httpd-2.3.14 as beta
Date Mon, 08 Aug 2011 08:00:36 GMT
On 01.08.2011 18:58, Jim Jagielski wrote:
> The tarballs for httpd-2.3.14 are available at:
> Please VOTE on whether to release these as Apache httpd-2.3.14,
> beta.

 +1 to 2.3.14-beta.

- Sigs and hashes OK
- contents of tarballs identical
- contents of tag and tarballs identical
  except for expected deltas

Build on Solaris 8+10 Sparc, SuSE 10 32 and 64, and RHEL 5 64
- with default (shared) and static modules
- with module sets none, few, most, all, reallyall and default
- against bundled APR/APU, external APR/APU 1.4.5/1.3.12
  and apr trunk r1152907

- using expat 2.0.1, pcre 8.12, openssl 1.0.0d, lua 5.1.4
- Tool chain:
    - platform gcc except for Solaris 8+10 (gcc 4.1.2)
    - CFLAGS: -O2 -g -Wall -fno-strict-aliasing
              (and -mpcu=v9 on Solaris)

Passed test framework on all those platforms for all available MPMs (did
not test simple)
- with shared and static modules using all of the above module sets
- with bundled APR/APU, external APR/APU 1.4.5/1.3.12 and trunk apr

Problems encountered and exceptions to test framework passes:

- module set reallyall does not build with apr trunk, because
  of missing kdap support.
  Fixed by SF in httpd trunk by correctly checking ldap support
  when choosing module sincluded in reallyall.

- module set all does not build as static modules
  (link error of mod_heartbeat against mod_watchdog).
  Again fixed by SF in httpd trunk.

- module mod_watchdog and mod_proxy_fdpass always built shared,
  even in static builds.
  Again fixed by SF in httpd trunk.

- module mod_proxy_fdpass contained in builds for module sets
  few and none though no mod_proxy included.
  Again fixed by SF in httpd trunk.

- Crashes during statically linking httpd on SLES 32 / 64.
  Reason yet unknown, it looks like the crash is happening in the
  libtool script, likely due to the commandline being to long for
  the big module sets (most and above).

- I can't run test test suite for Solaris on the statically linked
  reallyall, since then also mod_privileges in included and I have
  problems running the test suite with it.

- I can't run the test suite for the "none" module set, because of
  unknown directives, but I'd say that test case is moot.

None of those looks like a showstopper.



View raw message