Sometimes freeswitch doesn't exit after receiving the SIGTERM signal.
This can be reproduced by sending SIGTERM to a freeswitch instance which
is initializing (which can take quite some time).
Instead of just giving up and exiting - leaving a hung freeswitch
process on the system - this commit adds some lines to the init script
that send SIGKILL to freeswitch in case the attempt to terminate it with
SIGTERM fails.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Applied a patch, submitted upstream at
https://freeswitch.org/jira/browse/FS-11193
that fixes two memory leaks in mod_event_multicast.c
Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
- Bump version from 1.6.20 to 1.8.1 (includes OpenSSL 1.1.0 support).
- Upstream didn't provide a source tarball yet, switch to git to grab
the release.
- freeradius-client got updated to 1.1.7 by upstream, update version and
checksum accordingly.
- tiff, libvpx and pcre patches are no longer needed, drop them.
- Refresh the remaining patches. Only minor changes were needed.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove the hack for iconv on uclibc. Instead include nls.mk which will
sort out the iconv depends.
This changes the iconv patch. AC_SEARCH_LIB doesn't work properly for
openwrt/lede. Instead use AC_CHECK_LIB.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The sed script used on libs/unimrcp is used to set LDFLAGS using
APR_SETVAR. Since nls.mk is included there are LDFLAGS with commas. But
the macro APR_SETVAR uses commas to separate arguments, so now the
LDFLAGS are only getting partially copied.
Solve this by surrounding LDFLAGS with brackets ([...]).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Removes the uClibc iconv hack. Instead nls.mk is included to sort out
iconv depends.
mod_gsmopen isn't able to figure out if iconv's inbuf is const. In
OpenWrt this is the case with libiconv-full, not with libiconv-stub. To
aid mod_gsmopen update 270-fix-uclibc-iconv-in-gsmopen.patch so that we
can pass on if inbuf is const via CFLAGS.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The macro included in apr/apr-util is incorrect. It results in extra
brackets being put into conftest.c, for example:
#include "confdefs.h"
[
#include <stddef.h>
#include <iconv.h>
]
int main(int argc, const char *const *argv) {
[
iconv(0,(char **)0,(size_t *)0,(char **)0,(size_t *)0);
]
return 0; }
The test always fails with:
error: expected identifier or '(' before '[' token
The following checks are affected:
- variable length arrays
- type of inbuf parameter to iconv
Fix from upstream (Apache).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
libmariadb 10.2 needs to be linked in together with iconv. Include nls.mk
to help the build system find the iconv library.
Also update func_iconv's depend. We want to use the variable
$(ICONV_DEPENDS) provided by nls.mk. It resolves to
+BUILD_NLS:libiconv-full. This cannot be handled properly right now, as the
module depends filter cannot handle +<SYMBOL>:<package> depends. So add
another field to the Makefile's module function for complex depends, which
are passed on unfiltered.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
libmariadb 10.2 needs to be linked in together with iconv. Include nls.mk
to help the build system find the iconv library.
Also update func_iconv's depend. We want to use the variable
$(ICONV_DEPENDS) provided by nls.mk. It resolves to
+BUILD_NLS:libiconv-full. This cannot be handled properly right now, as the
module depends filter cannot handle +<SYMBOL>:<package> depends. So add
another field to the Makefile's module function for complex depends, which
are passed on unfiltered.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Since libmariadb 10.2.x requires to be linked in together with iconv,
include nls.mk to help the build system find it.
Also fix a type in PKG_CONFIG_DEPENDS.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently compiles fail with:
CC libfreeswitch_la-switch_regex.lo
src/switch_regex.c: In function 'switch_regex_compile':
src/switch_regex.c:40:9: error: return from incompatible pointer type [-Werror=incompatible-pointer-types]
return pcre_compile(pattern, options, errorptr, erroroffset, tables);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
Makefile:2100: recipe for target 'libfreeswitch_la-switch_regex.lo' failed
make[5]: *** [libfreeswitch_la-switch_regex.lo] Error 1
Fix cherry-picked from upstream.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Switch to use procd for init, with support for jails. Clean up code and
restructure callback processing to be more robust and understandable. This
also fixes a bug processing multiple siproxd config sections, as reported
by Michael Kuron.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
When there is no sounds cache declared the build system will not attempt to
download any sound packs or their SHA1 checksums.
This is to be preferred because:
a) the build may occur offline, causing it to fail
b) plain http is used by the build system for downloading
There is no drawback here because the standard sound packs are included in
the Asterisk source tarball already.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
When there is no sounds cache declared the build system will not attempt to
download any sound packs or their SHA1 checksums.
This is to be preferred because:
a) the build may occur offline, causing it to fail
b) plain http is used by the build system for downloading
There is no drawback here because the standard sound packs are included in
the Asterisk source tarball already.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
When configure is called with '--with-mysql="$(STAGING_DIR)/usr"' it tries
to detect mysql manually instead of using mysql_config. This doesn't always
work:
<snip>
checking for MySQL in /var/lib/buildbot/slaves/slave-lede-builds4/i386_pentium4/build/sdk/staging_dir/target-i386_pentium4_musl/usr... no
<snip>
This commit changes the configure argument to '--with-mysql=yes/no'. When
'yes' is selected configure uses mysql_config and then the detection works
reliably.
<snip>
checking for MySQL using mysql_config... yes 10.1.32
checking whether MYSQL_OPT_RECONNECT is declared... yes
checking whether MYSQL_OPT_READ_TIMEOUT is declared... yes
checking whether MYSQL_OPT_WRITE_TIMEOUT is declared... yes
checking whether mysql_set_character_set is declared... yes
<snip>
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- patches refreshed
- menuselect is bootstrapped from main bootstrap script now, extra call
removed
- drop a module that was removed upstream for being incomplete
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- patches refreshed
- menuselect is bootstrapped from main bootstrap script now, extra call
removed
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Despite all previous efforts chan-lantiq is not built by the build bots.
Checking the logs of a target build reveals that the chan-lantiq builds are
not even attempted.
During the target builds CONFIG_ALL is not set, meaning no regular package
gets selected. CONFIG_ALL_NONSHARED on the other hand is set. But
chan-lantiq still will not be selected, because it depends on "asterisk",
meaning asterisk needs to be selected first for chan-lantiq to become
available.
This commit changes the asterisk depend to "+asterisk", meaning asterisk
will be selected if chan-lantiq is selected.
This was tested with the SDK. A seed config file was added with the
following:
CONFIG_ALL_NONSHARED=y
Running menuselect showed that now chan-lantiq is selected
automatically. This should fix the problem.
The problem with "+asterisk" is that this breaks the asterisk menus for all
asterisk plugins/utilities that follow chan-lantiq alphabetically. To
address this put the chan-lantiq modules into their own sub menu "Telephony
Lantiq". This is not nice, but the alternative would be to rename the
packages to something like asterisk13-z-chan-lantiq. This would make it
harder to find the packages. Maybe somebody has a better idea in the
future.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
chan-lantiq currently is still not built by the build bots. Split the
package in two, one for Asterisk 13, the other for 15, to avoid a VARIANT
build. Maybe the build bots do not support nonshared VARIANT builds.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Tidy up the flags passed on to FS build system for Perl and Python
modules. Some flags seem a bit too random to include - this commit
removes them.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Underscores should not be used in package base names as they're used as
semantic separators by opkg.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
mod-avformat is only built when mod-avcodec is selected. baresip sets it
up like that in mk/modules.mk:
ifneq ($(USE_AVCODEC),)
MODULES += avcodec
ifneq ($(USE_AVFORMAT),)
MODULES += avformat
endif
endif
This commit updates the mod-avformat DEPENDS accordingly.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
There are two pulseaudio packages, pulseaudio-daemon and
pulseaudio-daemon-avahi. Both provide pulseaudio. Depend on pulseaudio
so the user may choose which one to install.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Use explicit ident string ("siproxd") instead of NULL, as the latter is
not guaranteed to prepend the program name (e.g. unspecified in POSIX),
and may result in syslog messages identified only by PID.
Use consistent facility (LOG_DAEMON) across openlog() and syslog() calls.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Explain default configuration options and potential firewall settings.
Provide examples of using documented siproxd configuration directives.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Switch to using callback functions for UCI config processing, and remove
the unnecessary scan_interfaces() call together with related code.
Instead of explicitly handling every possible siproxd config option, use
callbacks to deal with those defined in the UCI config file, and directly
handle only selected options with desired default values. This shrinks
the init code by ~3.5 KB and simplifies updating to future siproxd
versions with new options.
(Note: this change exposed some long-standing, broken aspects of UCI
callback handling, fixed in https://github.com/openwrt/openwrt/pull/805.)
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
- The main package does not actually depend on libpopt. Only the utility
smsq does. This commit reflects that in the dependencies.
- libxslt is only used if generating the XML documentation, which this
package does not use. So remove the bogus dependency.
- app-sms does not depend on libpopt either. It has another dependency
on libstdcpp, which is already covered by the main package. This removes
both dependencies.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- The main package does not actually depend on libpopt. Only the utility
smsq does. This commit reflects that in the dependencies.
- libxslt is only used if generating the XML documentation, which this
package does not use. So remove the bogus dependency.
- app-sms does not depend on libpopt either. It has another dependency
on libstdcpp, which is already covered by the main package. This removes
both dependencies.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently libstdc++ is hardcoded. Change that and use whatever standard
C++ lib has preference.
Note: Two modules, event_zmq and v8, do not compile against libuClibc++.
For them to be available libstdcpp needs to have preference.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The module is never attempted to build in the regular buildbot package
cycle. Stick back the nonshared flag to get at least some log during
the target build cycle.
This also updates the dependency on kmod-ltq-vmmc to be selective (+).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The module list is necessary, unfortunately. But having working
PKG_CONFIG_DEPENDS is nice and the list rarely needs updating.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The module and util lists are necessary, unfortunately. But having
working PKG_CONFIG_DEPENDS is nice and the lists rarely need updating.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove the hack. Instead just call make with .lastclean and afterward
make menuselect again. If .lastclean was not removed by menuselect this
will not do anything, which is nice.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The module and util lists are necessary, unfortunately. But having
working PKG_CONFIG_DEPENDS is nice and the lists rarely need updating.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove the hack. Instead just call make with .lastclean and afterward
make menuselect again. If .lastclean was not removed by menuselect this
will not do anything, which is nice.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Previous revert only works when libstdcpp is the default C++ lib. To
make the compile work in all scenarios LIBS needs to be defined and
MAKE_FLAGS are required instead of MAKE_VARS, otherwise the included
variables are overridden.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This reverts commit b8b9818a13.
In hindsight I have to admit I did not correctly understand the
implications of the uclibc++.mk include.
The include allows a package to follow the user's choice regarding which
C++ library should be the standard. Linking against uClibc++ instead of
libstd++ is not a problem when running musl (which is what I had
incorrectly assumed), as both C++ libs are separate packages. And
uClibc++ is a lot smaller than libstd++, which is probably why it is
even the default C++ lib on OpenWrt currently.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This reverts commit 6857bc21eb.
In hindsight I have to admit I did not correctly understand the
implications of the uclibc++.mk include.
The include allows a package to follow the user's choice regarding which
C++ library should be the standard. Linking against uClibc++ instead of
libstd++ is not a problem when running musl (which is what I had
incorrectly assumed), as both C++ libs are separate packages. And
uClibc++ is a lot smaller than libstd++, which is probably why it is
even the default C++ lib on OpenWrt currently.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently there are two simple patches for libvpx. This commit puts the
second patch into the first.
Also added:
- CPPFLAGS
- a flag to address CVE-2017-0641
- comments
A tiff patch got refreshed due to the above.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently the routine that sets up the modules.conf file resides in
Build/Prepare.
When PKG_CONFIG_DEPENDS causes a reconfiguration, Build/Prepare is not
triggered. So if another module gets selected in menuconfig, this module
would not be added to modules.conf, meaning the rebuild would fail.
Add the routine to Build/Compile to address this.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Some source files need to be copied into PKG_BUILD_DIR. This is
currently done in Build/Prepare.
When PKG_CONFIG_DEPENDS causes a reconfiguration, Build/Prepare is not
triggered again, meaning needed source files would not be copied.
Move these lines to Build/Compile to address this.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Fixes:
AST-2018-004
AST-2018-005
AST-2018-006
- Remove res_pjsip_transport_management.so as it got moved into
res_pjsip's core as per AST-2018-003.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Fixes:
AST-2018-004
AST-2018-005
- Remove res_pjsip_transport_management.so as it got moved into
res_pjsip's core as per AST-2018-003.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Use $(INSTALL_BIN) as the modules are installed by siproxd as
executables. Remove the wildcards from this line as they're not needed.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove LIBLTDL from the Makefile as well as a patch related to LTDL.
Both have in common that they accomplish absolutely nothing.
Also explicitly disable the use of the internal libltdl.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Use the default Build/Compile. Use MAKE_VARS to set up the variables
that are used by pcapsipdump's Makefile.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The include injects a lib depend on libuClibc++.so.0 into the package,
even on musl. Remove it and replace the depend with libstcpp instead.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Yate is the only package in OpenWrt which manually enables RTTI
(run-time type information). There's no explanation in the Makefile as
to why that would be needed or advisable. Remove it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
With this include libuClibc++.so.0 gets added as a lib depend, even on
musl. readelf shows:
Dynamic section at offset 0x10ea60 contains 23 entries:
Tag Type Name/Value
0x0000000000000001 (NEEDED) Shared library: [libuClibc++.so.0]
0x0000000000000001 (NEEDED) Shared library: [libc.so]
0x0000000000000001 (NEEDED) Shared library: [libgcc_s.so.1]
0x000000000000000e (SONAME) Library soname: [libyate.so.6.0.0]
Remove the include from the Makefile and add a depend on libstdcpp
instead.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This reverts commit 4cc2afdcdc which
disabled FORTIFY_SOURCE by removing OpenWrt's CPPFLAGS from the built.
Instead PKG_FORTIFY_SOURCE:=0 is defined, which is the proper way to do
it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
rtpproxy doesn't use OpenWrt's CPPFLAGS (i.e. used for including
fortify-source headers). Add them to CFLAGS.
Run autoreconf. It's usually a good idea.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Version bump
- Refresh debug-disable patch
- Add patch to add missing includes to fix:
rtpp_record.c: In function 'rtpp_record_open':
rtpp_record.c:228:62: error: 'DEFFILEMODE' undeclared (first use in this function)
rrc->fd = open(rrc->spath, O_WRONLY | O_CREAT | O_TRUNC, DEFFILEMODE);
^
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- The libraries and the Berkeley DB recovery utility are currently in
Libraries/Telephony and Utilities/Telephony. That makes little sense
as the are not general purpose libraries/utilities, hence it is better
to move them to where Kamailio 5 sits in menuconfig.
- Rename them by adding prefixes "lib" and "util" to their name.
- Update their titles a bit.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The recently added mod-db-berkeley comes with a companion utility.
Package kambdb_recover so it becomes available.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This commit moves libkamailio-ims & libtrie into their own packages to
keep the main package small.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
kamailio comes with internal libraries. Some are not built when only the
standard group of modules is compiled (i.e. no modules selected in
menuconfig).
The problem is that there is no way to rely on specific internal
libraries being available in the main kamailio package.
To fix this this add some modules (if they are not already selected)
that will kick of the build of the internal libraries in any case.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The INSTALL file from kamailio suggests to make the target "cfg" first.
All settinggs will then be reused when make is called again.
This has the side effect that "$(FPIC)" can be dropped from CFLAGS as
the build system adds the flags automatically now.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
When not specifying "run_target" the kamailio build system assumes
$(DESTDIR)/var/run/kamailio to be the run directory on the target
device. Add the variable to address this.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The kamailio build system provides an RPATH "///usr/lib/kamailio/" when
linking. OpenWrt's rstrip.sh inspects this RPATH and drops it. Now the
modules cannot find the needed libraries. Fix this by specifying a path
that rstrip.sh accepts.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The initial commit for this plugin missed that it comes with a
configuration file and extra files that go into
/usr/share/kamailio/xhttp_pi.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This installs the snmp mibs files provided by mod_snmpstats. A patch is
added to make the kamailio build system install the files in a proper
location in PKG_INSTALL_DIR.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The aim is to add the missing modules that Debian packages, minus the
ones for which depends are missing on OpenWrt.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add install/conffile routine
- use it for main package
- setup BuildKamailio5Module to allow inclusion of configuration files
on a per-module basis (mod-tls is currently the only user)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- dictionary.kamailio & kamailio-basic.cfg are gone, remove from
Makefilei accordingly
- Install the config files from PKG_INSTALL_DIR
- Use $(INSTALL_CONF) for config files
- Remove gratuitous directory creation
- Refresh 050-fix-kamailio-utils.patch
- Remove src/modules/cnxcc/Makefile change from 120-libevent-libs.patch
as $(EXTRA_LIBS) has been removed from Makefile in previous commit
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Removes TLS_HOOKS as this is default anyway
- Removes -DUSE_PTHREAD_MUTEX as on supported arches FAST_LOCK is used
by default, which according to doc/tutorials/locking.txt is fastest.
-DUSE_PTHREAD_MUTEX will be set by the build system automatically in
case FAST_LOCK is unsupported, e.g. on ARC.
- Add OpenWrt LDFLAGS via LD_EXTRA_OPTS.
- With OpenWrt LDFLAGS and CPPFLAGS in place there is no need to declare
PCREDEFS, PCRELIBS nor EXTRA_LIBS. Remove them all.
- With this setup mod-dialplan now links to libpcre as well; add the
depend accordingly.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This commit changes Build/Compile, Build/Install and Build/InstallDev to
prevent building/installing items that are not selected. I.e. when
freeswitch-stable is not selected, only libfreetdm-stable gets built.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This commit renames libfreetdm to libfreetdm-stable. Also the depends
are changed to what they actually are, i.e. there is no depend on
freeswitch-stable anymore (libfreetdm.so is a standalone libary).
The FreeTDM modules get the same treatment.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
ilbcwebrtc and isaccodec currently depend on @!arc because the code does
not compile there. The code does not compile on aarch64 either.
Instead of adding @!aarch64 this commit adds a patch copied from Debian
that turns the error "Please add support for your architecture in
typedefs.h" into a warning, allowing the code to build.
Consequently this commit removes @!arc from the depends as well.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently three copies of every library are installed. Save space by
installing symbolic links instead.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
yate cannot be compiled against fortify-source headers at this moment:
In file included from /home/sk/tmp/lede/staging_dir/toolchain-x86_64_gcc-5.5.0_musl/include/fortify/unistd.h:23:0,
from ../yateclass.h:32,
from ../yatengine.h:29,
from ../yatephone.h:29,
from server/ysnmpagent.cpp:22:
/home/sk/tmp/lede/staging_dir/toolchain-x86_64_gcc-5.5.0_musl/include/fortify/fortify-headers.h:20:42: error: type of 'TelEngine::strcat' is unknown
To work around this remove CPPFLAGS from CFLAGS.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
When gcc 7 is used, -Werror=format-truncation finds snprintf() calls in
mod_cdr_mongodb.c and suspects that they may result in truncation. As a
result the build fails.
Work around this by adding -Wno-error=format-truncation to CFLAGS when
gcc support this.
JIRA-10939 has been raised upstream to follow up on this.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- perl and postgresql do not compile on ARC; disable the modules
accordingly
- ilbcwebrtc and isac also cause build failures on ARC; disable them as
well.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently conffiles is always filled with <module>.conf. This leads to
"no such file and directory" errors when the file actually does not
exist.
This commit adds a 6th column to BuildPlugin() which indicates if there
is a config file or not.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- remove ARGS that are default anyway
- enable sse2 when target is x86_64
- remove --enable-inline as that causes too many warnings:
ilbccodec.cpp:50:7: warning: inlining failed in call to 'virtual {anonymous}::iLBCFactory::~iLBCFactory()': call is unlikely and code size would grow [-Winline]
class iLBCFactory : public TranslatorFactory
^~~~~~~~~~~
ilbccodec.cpp:50:7: note: called from here
class iLBCFactory : public TranslatorFactory
^~~~~~~~~~~
- --enable-internalregex removed as yate can use external regex
- make some choices depending on whether a package is selected; otherwise the
build might fail, i.e. when spandsp is not installed but the Makefiles tells
the build system to use spandsp anyway
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- remove flags to don't have a reason for being there anymore (there is
no comment on why they're needed).
- add CPPFLAGS to CFLAGS so the compiler can use the FORTIFY_SOURCE
headers.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently the Makfile pulls in every dependency mentioned. Fix this by
prefixing the depends with the respective PACKAGE_* items.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Pass CPPFLAGS in the CFLAGS as otherwise the build system will ignore
them.
This way FORTIFY_SOURCE headers are passed to the compiler.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently INCL_MODULES is filled with all modules. This commit changes
this so that only the modules that are selected are added to it.
To further limit the number of modules built group_include="standard" is
added. This way only the modules of the standard group plus the modules
selected in menuconfig are built. The "standard" group only includes
modules without any depends.
This fixes build failures on targets where not all depends are
available, i.e. postgresql on ARC. Now these modules can be properly
prevented from being built.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently the Makefile pulls in depends even for packages which are not
selected. Fix this by prefixing the depends with the respective
PACKAGE_* item.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Adding KAM_MODULES variable is not needed. Remove it and call
BuildPackage from BuildKamailio5Module instead. For this to work
PKG_MAKE_ARGS and Build/Compile need to be moved up, before the module
evals.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The included libtiff is dated and has multiple unfixed CVEs. This commit
applies patches from upstream's master branch to use the system libtiff
instead.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Upstream seems to have updated their libtool. Instead of running
rebootstrap.sh, set PKG_FIXUP:=autoreconf. Otherwise the build would
fail due to libtool version conflicts.
This is actually nice, because now the build is a bit quicker due to
less autoconf and configure runs.
rebootstrap.sh was primarily used because it worked around some other
issues (e.g. not being able to detect how to convert files to the cross
compiler format, i.e. "func_convert_file_noop"), but that doesn't appear
to be needed anymore, which is nice.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- use upstream release tarball instead of git checkout
- update to recently released version 1.4.5
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The source does not change, but reflect the real version (2.2.1) in
PKG_VERSION. Also move from gz to xz to shave off a few Bytes.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This was suggested in PR #241 by @commodo. Python maintainers moved the
.mk files and also want to get rid of the include_mk construct.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Remove 003-disable-ast-xml-docs.patch, it doesn't apply anymore.
Instead change --enable-xmldoc into --disable-xmldoc in the Makefile.
- Refresh patches.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Remove 003-disable-ast-xml-docs.patch, it doesn't apply anymore.
Instead change --enable-xmldoc into --disable-xmldoc in the Makefile.
- Remove 055-ASTERISK-24517.patch again due to multiple hunk failures.
- Refresh patches.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This commit adds mod_freetdm (glue between FreeSWITCH and FreeTDM) and
the following FreeTDM modules:
- ftmod_analog
- ftmod_analog_em
- ftmod_libpri
- ftmod_pritap
- ftmod_skel
- ftmod_zt
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The OpenWrt telephony feed already has DAHDI support. This commit
enables FreeTDM so FreeSWITCH can make use of DAHDI via libpri.
This does not change the existing FreeSWITCH packages as FreeTDM is a
separate library in the FreeSWITCH distribution, hence no revision bump.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
uclibc doesn't include iconv. This commit adds a dependency to
libiconv-full when building against uclibc. iconv can be used by the
included apr and is a hard dependency of mod_gsmopen.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add procd init script
- script starts the daemon as user "baresip" instead of root
- add /etc/default/baresip
- add example configuration into /etc/baresip
- add conffiles definition
- prevent autostart after upgrade by using ENABLE_BARESIP variable in
/etc/default/baresip
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add procd init script
- script starts the daemon as user "restund" instead of root
- add /etc/default/restund
- declare /etc/init.d/restund a configuration file
- prevent autostart after upgrade by using ENABLE_RESTUND variable in
/etc/default/restund
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add CROSS_COMPILE and OS to MAKE_FLAGS; this should help when
compiling on other OSes, e.g. Darwin.
- move RELEASE up (alphabetical sort)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The patch is wrong. It may cause errors of the "illegal instruction" variety.
The libre package fix in commit e40f0ea06f makes
this patch unnecessary anyway. Drop it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The current approach (creating a blocker file) is clumsy. Use a variable
in /etc/default/freeswitch instead.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Add some dependencies to select modules. Dependencies confirmed by
menuselect.
Also fix description of res_monitor.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Add some dependencies to select modules. Dependencies confirmed by
menuselect.
Also fix description of res_monitor.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove the possibility to prevent the dependency on libedit. freeswitch
behaves very quirky without it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Use "-nc -nf" always. When started by procd freeswitch detects the
console to be orphaned and disables it. So disable the interactive
console but force freeswitch to not fork into background.
- Improve readability of command parameters.
- Minor changes.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently baresip often fails to build because the FFmpeg compile was
not successful:
CC [M] build-mipsel/modules/avcodec/avcodec.o
modules/avcodec/avcodec.c:9:32: fatal error: libavcodec/avcodec.h: No such file or directory
compilation terminated.
Address this by adding a configuration symbol that prevents the
FFmpeg-depending modules from being selected by default, except on i386
and x86_64.
It is better having baresip packages without these video modules than
not having baresip packages at all.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
re and baresip builds for 32bit x86 targets are currently failing:
build-i386/sip/auth.o: In function `sip_auth_encode':
/build/lede-snapshots/i386_pentium/build/sdk/build_dir/target-i386_pentium_musl/re-0.5.6/src/sip/auth.c:279: undefined reference to `__stack_chk_fail_local'
The undefined references occur because there is no linking done to ssp.
Fix this by setting LD to TARGET_CC, so ld is not called directly. This
also means one patch is not needed anymore.
Do the same for rem and restund as they basically all share the same
build system.
Compile-tested for x86, x86_64 and mips.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
On the buildbots (and locally, too) the x86 builds currently fail. It's
visible that the Asterisk build system attempts to build menuselect a
second time, which fails because it follows the standard recipes in its
Makefiles. The first build of menuselect on the other hand only works
because the OpenWrt package sets up the build properly.
The second build attempt happens because the menuselect CFLAGS category
is setup to delete a stamp file when anything in it changes. On x86
-march=native is available, so the OpenWrt package disables that via
menuselect, hence this item in the category is changed and the stamp
file automatically removed. The missing stamp file triggers a rebuild of
menuselect (which is not required for OpenWrt's intents and purposes).
Another item in the CFLAGS category with which the package can have the
same issue is LOW_MEMORY, but that is not enabled (changed) by default.
Work around this problem by recreating the stamp file after the
menuselect calls.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
On the buildbots (and locally, too) the x86 builds currently fail. It's
visible that the Asterisk build system attempts to build menuselect a
second time, which fails because it follows the standard recipes in its
Makefiles. The first build of menuselect on the other hand only works
because the OpenWrt package sets up the build properly.
The second build attempt happens because the menuselect CFLAGS category
is setup to delete a stamp file when anything in it changes. On x86
-march=native is available, so the OpenWrt package disables that via
menuselect, hence this item in the category is changed and the stamp
file automatically removed. The missing stamp file triggers a rebuild of
menuselect (which is not required for OpenWrt's intents and purposes).
Another item in the CFLAGS category with which the package can have the
same issue is LOW_MEMORY, but that is not enabled (changed) by default.
Work around this problem by recreating the stamp file after the
menuselect calls.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently the menuselect calls, for each category or item, step into a
directory and call menuselect once. So if all packages are build
menuselect will be called hundreds of times and as many directory
changes will be done.
Instead step into the directory only once, then generate the argument
string by iteratively adding to it and then call menuselect once. The
speedup is big.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently the menuselect calls, for each category or item, step into a
directory and call menuselect once. So if all packages are build
menuselect will be called hundreds of times and as many directory
changes will be done.
Instead step into the directory only once, then generate the argument
string by iteratively adding to it and then call menuselect once. The
speedup is big.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The way CONFIGURE_ARGS are done for srtp and pjproect does not take into
account that more than one package depends on them. This can result in
build failures. So add ifeqs to test properly if --with or --without is
required.
This commit also adds CONFIGURE_ARGS checks for:
alsa (asound)
portaudio
speex
speexdsp
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The way CONFIGURE_ARGS are done for srtp and pjproect does not take into
account that more than one package depends on them. This can result in
build failures. So add ifeqs to test properly if --with or --without is
required.
This commit also adds CONFIGURE_ARGS checks for:
alsa (asound)
portaudio
speex
speexdsp
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
With libcap support Asterisk is able to set the TOS bits even when
dropping its privileges to a non-root user.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
With libcap support Asterisk is able to set the TOS bits even when
dropping its privileges to a non-root user.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
@kochstefan got the module to work on Asterisk 15. Update the Makefile
and include the new variant.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The hack does not work on the buildbots. In the SDK the other packages'
Makefiles and patch directories aren't available, hence the hack fails.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds a mechanism that disables all modules by default and then
enables only the modules needed for the packages selected, resulting in
less build time.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds a mechanism that disables all modules by default and then
enables only the modules needed for the packages selected, resulting in
less build time.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
@kochstefan pointed out that asterisk 15 doesn't load any modules. It
turned out that this happens because of patch that was dropped. Add the
patch again.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- remove some whitespaces
- install only the configuration example from the source instead of the
locally kept files (which are probably outdated anyway)
- add a comment explaining the iostream hack
- use proper install routine instead of "$(CP)"
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The package currently fails to build because librem headers cannot be
found. For reasons unknown baresip searches in SYSROOT instead of
SYSROOT_ALT for them. Fix this by patching the Makefile.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove unneeded variables, whitespaces and trailing slashes, use
specific INSTALL directive instead of "$(CP)".
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Change from custom make call to using the default Build/Compile routine.
This applies the standard memory corruption mitigation methods, e.g.
FORTIFY_SOURCE and RELRO, to the package if they are enabled (default).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove unneeded variables, dependencies, whitespaces and trailing
slashes, use specific INSTALL directive instead of "$(CP)".
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Change from custom make call to using the default Build/Compile routine.
This applies the standard memory corruption mitigation methods, e.g.
FORTIFY_SOURCE and RELRO, to the package if they are enabled (default).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- bump to 1.4.2 (includes Asterisk 15 support)
- clean up the Makefile (whitespaces, remove some leftover defines that
are no longer needed etc.)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Update to latest commit to get support for Asterisk 15
- Upstream renamed the project from chan-sccp-b to chan-sccp. This
commit updates the OpenWrt Makefile accordingly.
- Move the directory from chan-sccp-b to asterisk-chan-sccp (follow the
naming of the other channel drivers' directories).
- OpenWrt Asterisk13 is mistaken for Asterisk15 because it has the
iostream patch applied, fix this with workaround in autoconf script.
- Update upstream URL to the github page.
- Update description of the package.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This commit unpacks the sources of lib_ifxos, drv_tapi and drv_vmmc,
patches them and sets up their header files for the compiler to use when
building chan-lantiq.
The dependency on kmod-ltq-vmmc is dropped so that the package can be
built by the buildbots. Instead, a dependency on mips is added so that
the buildbots will only build the package for the ARCH that Lantiq
actually runs on.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Instead of using local copies pull the source from github. The source
version is selected via CHAN_LANTIQ_VERSION, so there can be a different
version for each VARIANT.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds support for the following calendars:
- CalDAV
- MS Exchange Web Service
- MS Exchange
- iCalendar
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds support for the following calendars:
- CalDAV
- MS Exchange Web Service
- MS Exchange
- iCalendar
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
With the recently updated libsrtp2 asterisk 15 compiles fine, so it
doesn't need to be forcefully prevented anymore from linking against it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
With the recently updated libsrtp2 asterisk 13 compiles fine, so it
doesn't need to be forcefully prevented anymore from linking against it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The previously conceived hack - to make the buildbots actually build
this package during the target phase - does not work. This patch removes
the hack. This includes removing the lines from the asterisk Makefiles
that tar up the asterisk headers.
The upside of removing the hack is that the package name can be reverted
to what it was before, so it can be found in the same place like all the
other asterisk modules, i.e. below its asterisk version.
Another upside is that the Makefile does not pretend to solve the
problem anymore. Issue #193 is still open for the same. Maybe some day
somebody can figure out how to resolve the problem properly.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Several method signatures have changed in bcg729 1.0.1
and higher. This issue was already resolved in upstream.
In the current development branch, there is no compilable
version. Due to those changes, I have disabled linking
against bcg729 since we have bcg729 at version 1.0.2.
By disabling bcg729 the internal g729 rtpproxy API
will be used.
Signed-off-by: Jiri Slachta <jiri@slachta.eu>
I introduce asterisk 15.x as a replacement for asterisk 11.x. Currently
without support for chan-lantiq and sccp-b. Those features are going to
be added soon after this introduction.
Signed-off-by: Jiri Slachta <jiri@slachta.eu>
Asterisk-11.x has been marked as obsolete by upstream, thus we
cannot provide support for this version in telephony repository.
Asterisk at version 11.x is no longer with us, however its successor
is going to be in the repository for a quite a long time. Please use
13.x or newer version.
Signed-off-by: Jiri Slachta <jiri@slachta.eu>
When building using the SDK the $(SOURCE) variable is set differently,
resulting in files not being found. Not using the variable and using
"./" instead works fine.
This fixes issue #190. Thanks to @Bongo79 for reporting and providing
the fix.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Plain git:// doesn't add security. Use https:// instead. See
https://gist.github.com/grawity/4392747.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Cosmetic: change version from 1.0 to git checkout date.
- Remove dependency on asterisk, as otherwise the package won't be
built by the build bots.
- Without asterisk in the staging directory the asterisk headers are
now taken from an archive in $(DL_DIR), which is stored there by the
asterisk package.
- The naming of the package needs to be changed, as without the
dependency on asterisk it would otherwise mess with asterisk's
menuconfig layout.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Store asterisk headers in $(DL_DIR) so chan-lantiq can access them
during the target specific build phase. The headers are tarred up for
"mips_24kc" only because that is currently the only used arch for all
chan-lantiq targets.
Store one archive for each combination of asterisk version (e.g. 13) and
branch (using $(VERSION_NUMBER)). The latter makes sure that a release
build doesn't overwrite the header archive of a master snapshot build
and vice vera.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Currently chan-lantiq doesn't build on the build bots and users end up
having to build it themselves. The reason for this was explained by
Mathias Kresin here:
http://lists.infradead.org/pipermail/lede-dev/2017-September/008886.html.
To fix this put chan-lantiq into its own package and set
PKG_FLAGS:=nonshared. The new package uses variants already, so it is
easily extendable in the future.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Sync module to master.
- Drop big endian patch as it was included upstream.
- This fixes endian issue reported at
https://github.com/wdoekes/asterisk-chan-dongle/issues/44.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds ASTERISK-24517.patch from alpine linux. Without the patch
there are problems with the tcp/tls wrapper when running on a musl
toolchain. This is visible when toying with AMI:
LEDE*CLI> !nc localhost 5038
[Sep 10 12:13:25] WARNING[10315]: tcptls.c:684 handle_tcptls_connection: FILE * open failed!
This fixes issue #185 for asterisk 13 in trunk. The patch cannot be
applied to asterisk 11.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Asterisk 11 cannot deal with a unicode-enabled libedit. Input is not
possible anymore. Use the internal copy. Asterisk 13 works fine with the
external libedit.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The configuration file protection for modules does not work when there
is more than one configuration file. Currently conffiles is populated
with a space separated list. But the items need to be separated by
newlines. With spaces, only the first item gets added to conffiles. The
remaining items are dropped and not protected. Fix this by replacing
spaces with newlines.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Don't add extra TARGET_LDFLAGS for asterisk13-pbx-lua. It is not
needed anymore.
- Stop setting -Wl,-rpath-link in LDFLAGS. There seems to be no point to
do that (anymore).
- Stop calling specific make targets like version.h; make handles the
build properly without it.
- Add AST_FORTIFY_SOURCE to MAKE_FLAGS (seen in Debian rules file,
prevents asterisk's build system to mess with OpenWrt/LEDE flags).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Remove some unneeded parentheses.
- Use CONFIGURE_VARS instead of SITE_VARS.
- Clean up user.mak input, add NDEBUG=1 to avoid debug asserts.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Set data dir to /usr/share/asterisk and use default locations for
everything else.
- This is more in line with FHS and fixes issue #38. Database files are
now written to tmpfs (/var is mounted there). This prevents the
rootfs from filling up. It also prevents the flash to degrade during
constant database rewriting.
- As the data dir changes, so does the location of the sound files. This
commit also takes care of that.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
uclibc doesn't include iconv, so we need to use libiconv-full. This
fixes the ARC builds as they still default to uclibc.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
uclibc doesn't include iconv, so we need to use libiconv-full. This
fixes the ARC builds as they still default to uclibc.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Add new module. Iconv is available in the toolchain, so it doesn't cost
anything. Don't add it when uclibc is used, because there an external
libiconv would be needed.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
libedit now includes a pkgconfig file, meaning asterisk can detect it.
So use the system libedit instead having asterisk include an internal
copy.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Make the Makefile use BUILD_VARIANT so the package can be built for
different Asterisk versions. This is also the reason why the folder is
renamed from asterisk-11.x-chan-dongle to asterisk-chan-dongle.
- Change source URI from upstream to the fork of wdoekes. Upstream
doesn't update the repo anymore, wdoekes does.
- Remove dep on libiconv-full, make the package use toolchain's iconv
(compile-tested with musl).
- Drop patches that are already upstream.
- Update flags.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Add missing modules to curl and odbc packages. This fixes issue #181.
- Add back pbx-lua, which apparently was dropped earlier.
- Revision bump.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Don't add extra TARGET_LDFLAGS for asterisk13-pbx-lua. It is not
needed anymore (apart from that the module isn't available currently,
see next commit).
- Stop setting -Wl,-rpath-link in LDFLAGS. There seems to be no point to
do that (anymore).
- Stop calling specific make targets like version.h; make handles the
build properly without it.
- Use the default compile routine instead of the custom one.
- Add AST_FORTIFY_SOURCE to MAKE_FLAGS (seen in Debian rules file,
prevents asterisk's build system to mess with OpenWrt/LEDE flags).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Remove some unneeded parentheses.
- Remove $(SITE_VARS) as it is not used.
- Remove --without-curl and --with-uuid=... from configure arguments as
they've been removed from configure.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Set data dir to /usr/share/asterisk and use default locations for
everything else.
- This is more in line with FHS and fixes issue #38. Database files are
now written to tmpfs (/var is mounted there). This prevents the
rootfs from filling up. It also prevents the flash to degrade during
constant database rewriting.
- As the data dir changes, so does the location of the sound files. This
commit also takes care of that.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- As suggested by upstream in issue #167, move from svn to git source.
The resulting module is less than half the size in comparison.
- Rework (and rename) previous "drop-Og-optimization" patch. The
previous patch worked fine with -Os in CFLAGS (default). But if numeric
values are used instead (e.g. -O2) this optimization would be completely
removed. The new patch prevents that.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Remove two modules from pjsip. Upstream removed res_pjsip_log_forwarder
and moved res_pjsip_multihomed to res_pjsip.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- The configuration file protection for modules does not work when there
is more than one configuration file. Currently conffiles is populated
with a space separated list. But the items need to be separated by
newlines. With spaces, only the first item gets added to conffiles. The
remaining items are dropped and not protected. Fix this by replacing
spaces with newlines. Revision bump because this could lead to
overwriting files unintentionally.
- Remove a duplicate '/etc/asterisk/modules.conf' from
Package/asterisk13/conffiles.
- Remove '/etc/asterisk/rtp.conf' from Package/asterisk13/conffiles
because the file is installed and protected by res-rtp-asterisk.
- Add missing stasis.conf and pjsip_wizard.conf.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- bumped to 3.5.1
- removed musl patch because it was upstreamed
- add patch to fix ncurses detection
- fixes issue #125
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add some pjsip modules that were missing.
- sort pjsip modules alphabetically.
- add a patch from ASTERISK-26518 to fix the unresolved symbol issue.
The patch is also included in Debian. The patch fixes issue #77 (the
problem was also mentioned in #88).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Make "--disable-floating-point" depend on SOFT_FLOAT
- Remove "--disable-ilbc-coder" as configure no longer recognizes it.
- Align args with the ones used by Asterisk (for their bundled pjproject);
this includes adding "--disable-resample".
- With libresample removed remove the dep from asterisk13 accordingly.
- Note: the disabled items are _not_ used by Asterisk anyway.
- Drop speex dep as pjproject does not link against it the way it is
configured.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Set PKG_BUILD_DIR for each variant. Without it only one variant gets
built and put into both packages. This fixes issue #150.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Set PKG_BUILD_DIR for each variant. Without it only one variant gets
built and put into both packages.
- Add libiconv to build deps. Some iconv m4 files are required. Possibly
fixes issue #121.
- Replace recently added configure args with the ones suggested by upstream
@dkgroot in issue #167. This again shaves some bytes off the package.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This package is just wasting space on the mirrors. Only very few people
will actually want the sound files on their device. And those few who do
will have no problem downloading the files manually.
Get rid of it.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Replace $FS_STABLE_PREFIX_DIR with $FS_STABLE_DATA_DIR. The variable
is set to /usr/share/freeswitch, so it's not the prefix. Data dir is
the correct description.
- Remove lots of lines from CONFIGURE_ARGS. All of them duplicates,
because they're already set by the build system, so we can remove the
clutter.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
In issue #173 a segmentation fault occuring during a simple SIP call was
reported. Compiling with LOW_MEMORY disabled the reproducible segfault
disappeared.
This commits adds a switch for LOW_MEMORY. Default is off. If anybody
wants to debug the segfault s/he can enable it.
The segfault was reported for Asterisk 11. Disable LOW_MEMORY on
Asterisk 13 anyway, better safe than sorry.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Bump to revision 6728.
- Remove '-fi' from PKG_FIXUP. Trust in the build system to do the right
thing.
- Set PKG_INSTALL:=1 so "make install" doesn't need to be called
manually anymore. This also drops LOW_MEMORY from CFLAGS, which is not
needed as chan-sccp-b looks into asterisk's headers to find out if
LOW_MEMORY is enabled or not.
The build output actually looks much more pleasant afterward.
- Disable chan-sccp-b's optimization which it enables by default. It
sets -O3 and all that jazz. To make it simply use the OpenWrt/LEDE
flags optimization needs to be disabled.
- With optimization disabled chan-sccp-b still adds -Og to the end of
the CFLAGS. Add a small patch to prevent that.
- Disable debug which is enabled by default. This brings down the size
of the installed binary from 14 MByte (!) to a little over 2 MByte
(checked on ARM).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Asterisk 13 looks for libsrtp2. Unfortunately it also looks for a header
which libsrtp2 does not make public:
res_srtp.c:44:33: fatal error: srtp2/crypto_types.h: No such file or directory
compilation terminated.
This causes the asterisk build to fail. Fix this by disabling libsrtp2
detection in Asterisk 13.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
dahdi-linux fails to build on aarch64. Disable chan-dahdi for this arch
accordingly, so asterisk can still be built.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
LOW_MEMORY can't be enabled by defining it in ASTCFLAGS. This can be
verified in the headers Asterisk installs: LOW_MEMORY actually is not
defined (it would be if LOW_MEMORY was enabled).
LOW_MEMORY can be enabled successfully with menuselect. This commit
makes menuselect usable and calls it to enable LOW_MEMORY.
This commit also disables BUILD_NATIVE with menuselect. So even if
configure detects that -march=native is available, it won't be used.
This means the configure.ac patch that was added previously can be
dropped.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
menuselect often fails on the buildbots:
menuselect/menuselect: error while loading shared libraries: libxml2.so.2: cannot open shared object file: No such file or directory
This doesn't happen on all builds. The assumption is that on the
particular buildbot where there is no error the buildbots own libxml2.so
is usable (it exists and the version is suitable).
To fix this make the linker add an rpath to the menuselect utility,
pointing to the hostpkg lib dir.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The variable prevents the ifneqs from working on the buildbots. $(SDK)
is set there, so the ifneqs that test for empty will always be true.
libpq for instance doesn't build on aarch64. So the asterisk pgsql
package has a dependency on @!aarch64. An ifneq is used to disable libpq
detection by the configure script, to prevent the build to fail. But
because of $(SDK) being set the ifneq is always true, so libpq detection
is enabled, causing the build to always fail.
There are only two other packages in the tree that actually use the
$(SDK) variable, so it's fair to assume it's not needed for the asterisk
packages and can just be dropped.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Include host-build.mk, otherwise $HOST_CONFIGURE_ARGS will be empty.
- Remove --with-ncurses=PATH as the PATH is wrong. The ncurses host
package does not install anything but 'tic'.
- Use $HOSTCC to compile menuselect. This has the added bonus of using
ccache (if selected).
- Run configure with an unset $CONFIG_SITE so that it doesn't pick a
cross-compile site-script, e.g. for mips.
- Drop '-lxml2' from $LDFLAGS as the menuselect Makefile will add it
anyway.
- Move the variables in front of 'make' as they're not arguments.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
This adds a new Makefile as this module is out-of-tree. The module
dynamically links against bcg729.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Version bump to 13.17.0
- Remove makeopts.embed_rules as the target got removed.
- Refresh patches.
- $(STAGING_DIR)/host doesn't work anymore resulting in libxml2 host
headers not being found. Replace with $(STAGING_DIR_HOSTPGK).
- Prevent Asterisk from adding -march=native when cross-compiling (patch
for configure.ac).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Refresh patches.
- $(STAGING_DIR)/host doesn't work anymore resulting in libxml2 host
headers not being found. Replace with $(STAGING_DIR_HOSTPGK).
- Prevent Asterisk from adding -march=native when cross-compiling (patch
for configure.ac).
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
mod_v8 sporadically failed to build due to a segmentation fault when
generating the heap snapshot. This patch adds a switch to FreeSWITCH to
disable the snapshot and a symbol to toggle it.
Default is to not use a heap snapshot.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Replace main COPTS fixes patch with the updated version sent upstream.
- For the above to work we need to update DEFAULT_VARS in modmake.rules
(new patch)
- mod_pocketsphinx and mod_radius_cdr updated, no more extra COPTS fixes
required there.
- Now zeromq gets to use fortify headers and fails (warning about GCC
extension turns into error), so make it drop '-pedantic'.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Add mod_portaudio, mod_portaudio_stream, mod_radius_cdr and mod_v8.
- mod_v8 takes some time to build and it's quite large, so only build it
for x86_64. Users can force it on via a symbol for targets arm, i386
and mipsel as well.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Add missed dependency on libyuv to libvpx.
- Get rid of gratuitous @ in Config.in.
- Clean up Perl and Python deps.
- Fix installation paths for the ESL packages.
- Don't set ac_cv_lib_perl_perl_alloc=yes, it's not needed.
- Get rid of the build dep on erlang. Have the modules (erlang_event and
kazoo) depend on erlang instead; they both require epmd at runtime.
- Force spandsp and pocketsphinx to use fixed point math when soft float
support is enabled for the target device.
- Make libvpx use OpenWrt/LEDE flags, otherwhise it comes up with
optimized flags on its own.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Fix alphabetical sorting.
- Move some of the blocks around, e.g. move everything together that
adds to $(CONFIGURE_ARGS), move Download define upward to the other
defines etc.
- Replace some strings with variables, e.g. $(FIND) and $(FPIC).
- Be consistent and prefix our own variables with "FS_STABLE_".
- Remove the dev package and use Build/InstallDev instead.
- All in all this change is purely cosmetic.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Use lists for the config entries outbound_domain_{name,host,port}.
This permits multiple outbound proxies and fixes:
https://github.com/openwrt/telephony/issues/127
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
An upstream siproxd update was released after ~5 years, with bugfixes
and new features. This package supports the new plugins and their config
options, and updates required patches.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Use more conventional, clearer naming for config and registration files.
This includes having distinct directories and specific file suffixes.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Using the append_conf_if_set() function removes much repetitive and
error-prone code, reduces the source size by ~30%, and fixes a typo
previously uncaught.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
- Rename PYTHON_PKG_DIR to PYTHON_SITE_DIR, otherwise it's confusing.
- Remove the 'sed' hacks; handle the Python setup like the Perl setup is
already done.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- The builds currently fail with "undefined reference" errors. Fix that
by turning off Pulseaudio support. Also turn of Jack support in case
that gets detected in the future.
- aarch64 fails because this arch isn't recognized by the pocketsphinx
build system. Fix that by running autoreconf in the pocketsphinx build
dirs.
- Fix CFLAGS while we're here.
- Add dep on libsamplerate to mod_pocketsphinx. If it gets detected it
will be linked against, there is no switch to prevent that.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add mod_perl and perl-esl.
- some patching is required to make this work.
- for the modules to find libperl, an rpath is needed, which currently
doesn't work because of an rstrip.sh/patchelf issue. Patch sent to
OpenWrt/LEDE devs (https://patchwork.ozlabs.org/patch/780008/)
- fix some more copts issues that became apparent with Perl enabled.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- add b64, clearmode, cv (broken), dahdi_codec, dialplan_directory,
event_test, event_zmq, pocketsphinx, posix_timer, random, vmd and yuv.
- some of the new modules require extra sources, so add a download
routine for those so we can verify the checksums and prevent FreeSWITCH
build system to download them on its own.
- add two patches to fix some issues with the new modules.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- compile fix for mod_av added, without it the build errors out
(warnings turned to errors). It's still labelled @BROKEN, but once
OpenWrt's ffmpeg maintainers add libavresample to libffmpeg-full
(PR #4490) it should work.
- disable Erlang detection if mod_erlang_event is not selected (build
dep on Erlang unfortunately has to stay)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- make the use of Python conditional and drop the build dep.
- change Python LDFLAGS to the ones suggested by host python-config.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Pull more packages into the misc group (timezones etc.)
- Change Misc generator to not force the misc handle onto packages
- Add possibility to Misc generator to tag packages as arch agnostic
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The previous default logging level is completely silent, making normal
monitoring or troubleshooting impossible. Use the sane upstream default
for logging which captures basic startup, shutdown and error messages.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
Support using standard interface names and make "lan"/"wan" the defaults
for inbound/outbound SIP traffic. This change makes siproxd more portable
but preserves backwards compatibility for upgrades.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
- removed xmlrpc-c todo; I compiled with external version in staging
dir, it didn't make a difference (didn't seem to be used)
- removed swig calls as per suggestion from upstream
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Moved location of example packages from /usr/share/freeswitch/conf to
/usr/share/freeswitch/examples. It makes more sense and is also the
place where the module configuration snippets go.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Some modules include example content, e.g. xml snippets, in their source
directory. Added a symbol that causes these to be included in the
packages.
Disabled by default.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- meant for packages where one only needs to copy files from one
directory to another
- added three new packages using the new builder (dev, fonts, images)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- fixed swig location in all Makefiles in libs/esl and src/mod/languages
- forced a reswig for mod_lua as well, like it is already done for
mod_python
- added patch for mod_lua to properly do the reswig (patch sent
upstream, too)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- if source dir exists install the destination dir as it might not exist
yet
- if source dir does not exist do nothing (not even error out)
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- mod_unimrcp picked up external apr/apr-util if found in the target
directory, which caused the build to fail (undefined references). Fixed
by adding the FreeSWITCH apr/apr-util lib dirs to the front of LDFLAGS.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- added sed scripts in order for mod_python to find the right lib to
link to
- fixed some whitespaces
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- removed $FS_STABLE_DEPS_BASE and $FS_STABLE_DEPS because they aren't
needed to setup dependencies
- changed dependency setups of modules, utils and python-esl so that
their dependencies don't leak to the main package
- fixed dep of mod-hiredis
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- removed python-setuptools build dep as it's not needed
- moved the Python autotools variables to CONFIGURE_VARS
- fixed host swig location
- changed from manually erasing files to force a reswig to calling the
swigclean make targets provided by upstream instead
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- reworded some bools in Config.in
- removed some notes in Makefile which seemed a bit out of touch
- forced parallel build off again, as some module combinations won't
allow it
- fixed modules.conf editing sed script to not uncomment lines
accidentally
- removed wrong dep on db47 for mod_db
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
As suggested by Jo I uncommented the big sound packs, but marked as
BROKEN, meaning they're still available for selection, but not built by
default.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- Fixed sha256 checksums of some of the Russian language prompts
- Commented out non 8 kHz sound packs to not clog the mirrors
- Changed versioning scheme to the one used by upstream
- Made "Sounds" function params more generic
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Make sure that we properly pass down both TARGET_CPPFLAGS and
TARGET_LDFLAGS in order to fix build failures with external toolchains
that don't automatically search for headers and libraries in
$(STAGING_DIR).
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
- FreeSWITCH stable series for OpenWrt/LEDE
- clean, simple Makefiles
- provides 'freeswitch' user on target device, no need for root
- procd init script
- musl and glibc support
- SHA256 for download verification
- hotplug script with extras (condition checks)
- package size reduced
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
Update to 0.5.0
Remove Speex as it's being obsoleted by upstream.
Add Opus to replace Speex
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
Yate ships with a php script (banbrutes.php) which can help block brute
force login/password attacks. Since Yate for OpenWrt doesn't support
php scripts I've written a similar script in perl that's easier to use
with OpenWrt.
The script uses the iptables extension "recent" to keep track of
authentication failures and block users after too many authentication
failures.
This is an alternative to #87. The functionality is mostly the same,
but this is a slightly simplified, more light-weight version since
IP sets aren't used.
Signed-off-by: Robert Högberg <robert.hogberg@gmail.com>
pthread_key_create used to be called when a specific static key object was
initialized, but if another static object was initialized earlier and this
other object needed the key during its initialization yate would crash.
Fixes#99
Signed-off-by: Robert Högberg <robert.hogberg@gmail.com>
Within this commit several modules has been added, such as:
* app_queue
* func_presencestate
Fixes within this commit adds several cfg files to basic
set of files and reduces used space by installing symlinks
(manually merged changes from pr #98 and from pr #103)
Signed-off-by: Jiri Slachta <slachta@cesnet.cz>
Within this commit asterisk has been upgraded to version 11.22.0.
Several build issues has also been fixed (thanks to musl fixes from
Alpine linux).
Signed-off-by: Jiri Slachta <slachta@cesnet.cz>
Within this commit asterisk has been upgraded to version 11.22.0.
Several build issues has also been fixed (thanks to musl fixes from
Alpine linux).
Signed-off-by: Jiri Slachta <slachta@cesnet.cz>
New Features in V4.2.3
- Warning about use of ancient gcc compiler
- New desktop/background images
- New AMI Function:SCCPShowRefcount
- Support for newer/future version of automake
- Small size reduction of chan_sccp.so
Bug Fixes
- Channel being left behind after hangup
- Scheduled Hangup / Schedule Digit Timeout
- Potential Deadlock Fallback Atomic Functions
- MWI loosing track of new voicemails under certain circumstances
- using gnu version of libiconv fixed
- gen_sccp_enum improved / gnu awk extension dependency removed
- set socket option should happen before binding to the socket
- channel monitor/record handling fixed
- Code cleanup / fixed memory leaks / fixed potential null pointer dereferences
- Some smaller patches (Check ChangeLog for details)
Yate.pm that is included in yate-scripts-perl uses perl package Data::Dumper
which is provided by openwrt package perlbase-data. Update dependency to
reflect this.
Signed-off-by: Robert Högberg <robert.hogberg@gmail.com>
echo.sh is a simple yate script which can be used to echo back audio to a
caller. It's useful for testing purposes.
Since the script is a basic shell script it has no specific dependencies.
Signed-off-by: Robert Högberg <robert.hogberg@gmail.com>
1. Removed PROCESS_SHARED locks check from libs/apr/configure.ac.
2. Flagged mod_flite as BROKEN due to lack of flite package.
3. Added some additional modules, i.e. mod_av, mod_esl, and flagged as BROKEN for now.
4. Depricated mod_celt.
Signed-off-by: Mazi Lo <openwrt.mazilo@recursor.net>
1. Added an sqlite2 option for configure in apr-util
2. Moved bootstrap to Prepare stage.
3. Restructured patches directory to include CONFIG_LIBC
4. Fixed some modules dependency issues.
5. Reinstated mod_enum as not BROKEN.
6. Renamed mod_vp8 to mod_vpx according to FS changes.
7. Added some new modules and some are flagged as BROKEN.
i.e. mod_cv, mod_hiredis (BROKEN), mod_kazoo, mod_smpp (BROKEN)
8. Flagged some modules as BROKEN due to insufficient libraries.
i.e. mod_bv, mod_codec2, mod_fsv, mod_ilbc, mod_silk, mod_siren
9. Removed mod_snipe_hunt (depricated by FS)
Signed-off-by: Mazi Lo <openwrt.mazilo@recursor.net>
- upgrade baresip to 0.4.14
- uuid module is now part of the core-baresip
- add building of dtmfio module
Signed-off-By: Alfred E. Heggestad <aeh@db.org>
Acked-by: Jiri Slachta <slachta@cesnet.cz>
Tested-by: Jiri Slachta <slachta@cesnet.cz>
