Seems that the header files for the host libffi headers
are installed in the wrong place.
i.e. $(STAGING_DIR_HOSTPKG)/lib/libffi-3.2.1/include
when it should be $(STAGING_DIR_HOSTPKG)/include
Signed-off-by: Alexandru Ardelean <ardeleanalex@gmail.com>
* Switch back to upstream repo
* Follow the same configuration style as ffmpeg and mpg123 for using float vs
fixed point codepaths.
* Remove unneeded cpp variables (musl provides C99 math library)
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
Update (lib)expat to 2.2.3
Remove poor entropy hack, 2.2.3 uses /dev/urandom in worst case
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
- Remove ogg params from speex configure as it doesn't know about them.
- Remove --enable-sse=no from both speex and speexdsp. The configure
scripts do actually not add items to our CFLAGS like previously
assumed.
- When --disable-float-api is used VBR needs to be disabled as well, as
VBR has not been implemented with fixed-point math in speex (yet). So
add --disable-vbr when --disable-float-api is used.
- In speexdsp 1.2rc2 optimizations for NEON were added. Unfortunately
the ASM does not work for 64bit ARM. So force NEON optimizations off
when compiling for AARCH64.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
- speexdsp has been broken out of the speex package by upstream. The
released versions of both aren't aligned (speex is at 1.2.0, speexdsp
at 1.2rc3). Break out speexdsp into its own Makefile accordingly.
- Don't disable VBR anymore as it has its applications (i.e. can be used
with freeswitch-stable).
- Prevent build system from changing our CFLAGS (--enable-sse=no).
- Make integer mode depend on SOFT_FLOAT symbol so it doesn't get
enabled on devices with FPUs.
- Use PKG_INSTALL and remove custom Build/Compile as there is no need
for this.
Signed-off-by: Sebastian Kemper <sebastian_ml@gmx.net>
The patch fixes a problem where an incorrect getline.h would get included
(not gperf's own which resides in ./lib)
Signed-off-by: Espen Jürgensen <espenjurgensen+openwrt@gmail.com>
Add the same workaround as for the expat package in toolchain.
I'm not really sure what made it work for me earlier but this seems to fix it.
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
Update (lib)expat to 2.2.2
Fixes following CVEs: CVE-2017-9233 and CVE-2016-9063 (2.2.2)
Update homepage URL
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
Update (lib)opus to 1.2.1
Compile without floating point on NEON (ARM) capable hardware to enable
performance optimizations.
Discussion about this change:
https://github.com/openwrt/packages/pull/4574
Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
Install include files and libraries into a common place; this
makes compiling depending packages easier.
Signed-off-by: Stefan Tomanek <stefan.tomanek@wertarbyte.de>
avoid accidentally picking paths on the host for pg_config and
odbc_config by supplying them as parameters to configure.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
build odbc_config for host so drivers may use it to figure out build
details, patch odbc_config to return target specific values from
unixodbc_conf.h.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Some toolchains may automatically have sys/params.h which includes
isset() backport upstream commit
675ac7b4b6 to fix that. Also explicitly
link with openssl.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
This allows ODBC drivers to create odbcinst.ini-snippets in
/etc/odbcinst.ini.d/ which will be assembled into
/tmp/etc/odbcinst.ini. /etc/odbcinst.ini is provided as a symlink
pointing to /tmp/etc/odbcinst.ini
Hence the unixODBC-provided PostgreSQL driver was also given an
odbcinst.ini.d snippet.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Extend UCI mechanics to allow pre-populating a newly created
database by executing SQL statements from a scripts file.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
libiio is a library for interfacing with Linux's Industrial Input/Output (IIO)
subsystem. The Linux IIO subsystem is intended to provide support for devices
that in some sense are analog to digital or digital to analog converters.
Signed-off-by: Michael Heimpold <mhei@heimpold.de>