According to the NEWS message from the main OVS site:
http://openvswitch.org/releases/NEWS-2.3.0
Excerpt:
- ovs-controller has been renamed test-controller. It is no longer
packaged or installed by default, because too many users assumed
incorrectly that ovs-controller was a necessary or desirable part
of an Open vSwitch deployment.
Current feed was updated based on a feed that packaged OVS 1.9,
and subsequently updated.
For now, we'll remove it, since it's not required.
When I tested the feed, I tested it on a setup that has been configured
some time ago, and does not have that package enabled.
So, I'll remove it to prevent confusion.
Signed-off-by: Alexandru Ardelean ardeleanalex@gmail.com
- [PATCH 01/13] DOC: clearly state that the "show sess" output format
- [PATCH 02/13] MINOR: stats: fix minor typo fix in
- [PATCH 03/13] MEDIUM: Improve signal handling in systemd wrapper.
- [PATCH 04/13] MINOR: Also accept SIGHUP/SIGTERM in systemd-wrapper
- [PATCH 05/13] DOC: indicate in the doc that track-sc* can wait if
- [PATCH 06/13] MEDIUM: http: enable header manipulation for 101
- [PATCH 07/13] BUG/MEDIUM: config: propagate frontend to backend
- [PATCH 08/13] MEDIUM: config: properly propagate process binding
- [PATCH 09/13] MEDIUM: config: make the frontends automatically bind
- [PATCH 10/13] MEDIUM: config: compute the exact bind-process before
- [PATCH 11/13] MEDIUM: config: only warn if stats are attached to
- [PATCH 12/13] MEDIUM: config: report it when tcp-request rules are
- [PATCH 13/13] MINOR: config: detect the case where a tcp-request
Signed-off-by: Thomas Heil <heil@terminal-consulting.de>
By default the package detects the presence of pdflatex and builds documentation automatically.
During the OpenWrt build this behavior is not acceptable. The added patch introduces a configure
option to explicitly enable the build of the documentation on request.
Signed-off-by: Johannes Morgenroth <morgenroth@ibr.cs.tu-bs.de>
Add a patch to prevent use of LD_LIBRARY_PATH when checking for BDB libs
This caused build failure when host = target
Signed-off-by: Nicolas Thill <nico@openwrt.org>
rsync, by default, includes a patched zlib within its source tarball. The patch
enables a better compression ratio for rsync streams. However, the client and
server need the same featureset in their zlib in order to understand the
compression stream, or the server will abort the transfer. Older versions have
no understanding of the new compression standard used.
This patch allows the builder to choose whether to use the bundled or system
zlib. It defautls to using the system zlib, as is the default before this
patch.
Signed-off-by: Ian Leonard <antonlacon@gmail.com>
This package adds the mac-telnet server, client, ping and discovery utilities.
See https://github.com/aouyar/MAC-Telnet for details.
This package uses the fork from https://github.com/jow-/MAC-Telnet as source,
the code there has a number of bugfixes and results in smaller binaries since
most core functionality has been ported to libubox facilities provided by
OpenWrt.
Signed-off-by: Jo-Philipp Wich <jow@openwrt.org>
- BUG: config: error in http-response replace-header number of arguments
- BUG/MINOR: Fix search for -p argument in systemd wrapper.
- BUG/MEDIUM: auth: fix segfault with http-auth and a configuration with an unknown encryption algorithm
- BUG/MEDIUM: config: userlists should ensure that encrypted passwords are supported
- MEDIUM: connection: add new bit in Proxy Protocol V2
- BUG/MINOR: server: move the directive #endif to the end of file
- BUG/MEDIUM: http: tarpit timeout is reset
- BUG/MAJOR: tcp: fix a possible busy spinning loop in content track-sc*
- BUG/MEDIUM: http: fix inverted condition in pat_match_meth()
- BUG/MEDIUM: http: fix improper parsing of HTTP methods for use with ACLs
- BUG/MINOR: pattern: remove useless allocation of unused trash in pat_parse_reg()
- BUG/MEDIUM: acl: correctly compute the output type when a converter is used
- CLEANUP: acl: cleanup some of the redundancy and spaghetti after last fix
- BUG/CRITICAL: http: don't update msg->sov once data start to leave the buffer
Signed-off-by: Thomas Heil <heil@terminal-consulting.de>
The git-daemon command currently doesn't work and displays the following
error whenever a repository is cloned:
error: cannot run daemon: No such file or directory
[10920] unable to fork
On the client side the connection is simply terminated. The problem is,
that git-daemon tries to start a new instance of itself for every
new client that is connecting. It expects argv[0] to contain
"git-daemon", but since it is converted into a builtin command, argv[0]
only contains "daemon", which does not exist and causes the above error.
The fix simply prepends "git" to the list of arguments, so that the
resulting call looks something like "git daemon --serve ..."
Signed-off-by: Andreas Rohner <andreas.rohner@gmx.net>
W. Michael Petullo
Bruno Randolf
Damiano Renfer
Jo-Philipp Wich
Alexandru Ardelean
Steven Barth
Alexandru Ardelean
Thomas Heil
Daniel Salzman
Nuno Goncalves
Hauke Mehrtens
Rupan
Etienne CHAMPETIER
Adze1502
Matthew Sykes
Johannes Morgenroth
Nicolas Thill
Ian Leonard
Liviu Pislaru
Andreas Rohner
Nikos Mavrogiannopoulos
Peter Wagner
Artem Makhutov