imx6ulpico has a BCM4339 Wifi chip. Add Wifi support by default
to allow a better customer experience.
The dts patch has already been sent to the linux-arm-kernel list
and we can remove it once it reaches a mainline kernel (in
version 4.9 probably).
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Let the wireless packages be selected by default to make the
usage of Wifi a bit simpler.
Suggested-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
After fixing the <bits/local_lim.h> issue by using <limits.h>, the
build still fail with udefined __WORDSIZE.
Adding <sys/user.h> to define __WORDSIZE reveal that xenomai use
<error.h> which is not provided by musl.
The discussion on the musl mailing list [1] about glibc error reporting
functions concluded to not add this support in the C library.
For now, disable Xenomai user space support for musl toolchains since
it require several patches to build correctly and needs to remove each
glibc error reporting functions.
Users interested in musl support for Xenomai can work with upstream to
fix these issues and revert this patch.
[1] http://www.openwall.com/lists/musl/2014/06/29/8
Fixes:
http://autobuild.buildroot.net/results/e36/e36a21c1df33bdd3fbc61d516a3e8f7c5f7c41af
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Following include/asm-arm/features.h, Xenomai doesn't support cortex-M
cpus.
error "Could not find current ARM architecture"
Fixes:
http://autobuild.buildroot.net/results/06a/06a52b2c2c23df595f58feb6b95437fc65a86d31
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Introduce BR2_PACKAGE_XENOMAI_ARCH_SUPPORTS to hold the
architectures than can build and use Xenomai on the target.
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
There is no need for the 'exit' statement in the post-image scripts,
so just remove it.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Provide a more complete set of instructions on how to get Wifi
working on the warpboard.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
grub2 assumes the strip command will generate output and the output should
always be stripped - so, just use the $(TARGET_CROSS)strip to make sure that
the build succeeds regardless of the buildroot strip configuration.
Signed-off-by: Charles Hardin <ckhardin@exablox.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
iozone requires thread support since commit
06b53255cf, so having code to handle the
non-threaded case in the .mk file is useless.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The existent patch is not enough, as pthread_barrier_wait is used, which
isn't implemented in uClibc-ng Linuxthreads.
Remove the patch.
Fixes:
http://autobuild.buildroot.net/results/e5bbb80c81f4f170cf48d375e016e87a296ff754/
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The autofs configure script verifies on the build system if /proc is
provided by a Linux system. This might fail if /proc is not mounted,
but does ont mean that procfs will not be available in the target.
Therefore, this commit adds a patch improving the AF_LINUX_PROCFS
autoconf macro to allow passing an autoconf cache variable. autofs.mk
then uses that to force the fact that /proc will be available on the
target.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit adds a patch to the tinycbor package that fixes the
installation process with old make version.
This fixes the build on autobuilders using an old make, where tinycbor
was creating empty directories instead of installing files.
Fixes:
http://autobuild.buildroot.net/results/7605c301f460398c2435f6b73d6d07c7d1022e8c/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The readme.txt for the three boards are almost the same, so merge
them into board/zynq/readme.txt.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
U-Boot v2016.01 in the mainline. It includes proper ps7_init_gpl.c/h
in it, so builds working SPL without any manual intervention.
BR2_TARGET_UBOOT_BOARD_NAME should be changed to zynq_zc706 since
U-Boot now has separate defconfig files for ZC702 and ZC706 boards.
BR2_TARGET_UBOOT_SPL_NAME should be changed to spl/boot.bin since
the Zynq image support for the mkimage tool was upstreamed and it
now generates boot.bin under spl/ directory.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
Linux v4.4 in the mainline.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
U-Boot v2016.01 in the mainline. To sync with the Zedboard updates,
switch over to the Xilinx GitHub repository.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
Linux v4.4 in the mainline.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
U-Boot v2016.01 in the mainline.
Commit 46d6a561be ("zedboard: Generate BOOT.BIN automatically")
switched from the Xilinx GitHub repository to the U-Boot mainline,
but this commit gets it back to the Xilinx GitHub repository for the
following reasons:
- This defconfig still relies on the Xilinx's local repository for
the kernel. The vendor does the best test for the combination of
U-Boot and the kernel with the same release tag (xilinx-v2016.2
in this case).
- At the time of commit 46d6a561be, the u-boot-xlnx still needed
manual copy of ps7_init(_gpl).c/h in order to build a working SPL
image. So, the mainline U-Boot had advantage in that point of
time. However, the improvement in the mainline U-Boot was merged
into the u-boot-xlnx at the xilinx-v2015.3 release. Now, the
mainline and the u-boot-xlnx are even in this point of view.
- The mainline U-Boot defaults to boot FIT, so something must be
done; either patch environments around with a local patch to
switch to uImage booting (current solution), or build an FIT
with a post build script (chromebook snow does this). On the
other hand, the Xilinx repository defaults to uImage booting,
so it is straightforward, and needs no addition care.
This commit does:
- Switch to the Xilinx custom git repository, and stick to the
xilinx-v2016.2 tag.
- Delete the local patch board/avnet/zedboard/uboot/0001...
since the Xilinx custom repository can boot uImage by default.
- Enable BR2_TARGET_UBOOT_FORMAT_IMG instead of ..._FORMAT_DTB_IMG
since all the Zynq boards in U-Boot enable CONFIG_OF_EMBED.
- Replace BR2_TARGET_UBOOT_ZYNQ_IMAGE with BR2_TARGET_UBOOT_SPL_NAME
since U-Boot can natively generate the Zynq boot image now. The
Zynq image support for mkimage tool was upstreamed at v2016.01
(so xilinx-v2016.1 as well), so no additional tool is needed
any more.
- Update readme.txt
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This is the latest release tag in the Xilinx repository, based on
Linux v4.4 in the mainline.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This reverts commit 6f8162cf8c. This is
causing too many problems that are not easy to solve.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
SSDP managing daemon.
Designed to work with miniupnpc, miniupnpd, minidlna, etc.
http://miniupnp.free.fr/
Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
[Thomas:
- remove patch for _GNU_SOURCE, pass it from the .mk file instead
- add dependency on BR2_USE_MMU, fork() is used.
- rename $IF variable in init script/systemd unit file to $IFACE, for
clarity.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
GUPnP Tools are free replacements of Intel UPnP tools that use
GUPnP. They provides client and server side tools which enable
one to easily test and debug one's UPnP devices and control
points.
http://www.gupnp.org/
Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
[Thomas:
- use "depends on BR2_PACKAGE_LIBGTK3" instead of a select, and
simplify the Config.in comments consequently.
- move from "Development tools" to "Networking applications"
- license is GPLv2+, not LGPLv2+.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
GtkSourceView is a portable C library that extends the
standard GTK+ framework for multiline text editing with
support for configurable syntax highlighting, unlimited
undo/redo, search and replace, a completion framework,
printing and other features typical of a source code editor.
https://projects.gnome.org/gtksourceview
Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
[Thomas:
- simplify the Config.in by using "depends on BR2_PACKAGE_LIBGTK3"
instead of a select. It is pretty obvious that Gtk is needed if you
want to use gtksourceview.
- license is LGPLv2.1+, not LGPLv2+.
- change location in package/Config.in, gtksourceview is a library,
it should not go under "Development tools".]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
GUPnP DLNA is a small utility library that aims to ease the
DLNA-related tasks such as media profile guessing, transcoding
to a given profile, etc.
http://www.gupnp.org/
Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
tinydtls is a library for Datagram Transport Layer Security
(DTLS) covering both the client and the server state machine.
It is implemented in C and provides support for the mandatory
cipher suites specified in CoAP.
https://sourceforge.net/projects/tinydtls
Signed-off-by: Fabrice Fontaine <fabrice.fontaine@orange.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Version B.02.17 was found to be unstable on recent HW.
Status of the patches:
- The patch allowing to add extra values to the LIBS variable has
been refreshed, and changed to a Git formatted patch.
- The two patches from git.alpinelinux.org were needed for lshw to
build with the musl C library, but they have been merged upstream
(commit cd690bff1516b40fecd5ec4a7f6619e5bffc3cf0).
- The last patch was taken from upstream, and therefore already
merged, and now part of B.02.18.
This patch was tested with kernel 4.4.16.
Signed-off-by: Hubert Sokolowski <hubert.sokolowski@intel.com>
[Thomas:
- better explanation about patches
- re-add patch from Gustavo about LIBS, since it is really needed.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Some parameters displayed in the CSV reports are declared
as uint64_t, but the printf format doesn't reflect this.
Submitted bug: https://sourceforge.net/p/iperf/bugs/66/
Signed-off-by: Matt Poduska <matt.poduska@rockwellcollins.com>
Signed-off-by: Atul Singh <atul.singh.mandla@rockwellcollins.com>
Signed-off-by: Matt Weber <matthew.weber@rockwellcollins.com>
Reviewed-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
iperf2 is a fork of the original unmaintained iperf project, that
continues the development of the iperf 2.x series. Update homepage
link, and download location.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
[Thomas:
- change the download location to use downloads.sourceforge.net
- turn the patch into a Git-formatted one, since upstream uses Git as
its version control system
- as noticed by Khem Raj, a C99 compiler is needed for the bool type,
so added AC_PROG_CC_C99 to configure.ac, and consequently added
IPERF_AUTORECONF = YES.
- removed two <pkg>_CONF_ENV variables related to the bool type, they
were added by an older commit
e13ac0ec87 back when we had a
config.cache shared between packages (which is no longer the case
today)]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Selecting the driver CONFIG_BRCMFMAC as built-in causes some probe issues
as we cannot guarantee that the brcmfmac driver will be probed after the
rootfs has been mounted. The brcmfmac driver retrieves the firmware and
nvram file from the rootfs, so the rootfs should be mounted first.
To avoid such issues let the CONFIG_BRCMFMAC be built as module, which is
the original option in imx_v6_v7_defconfig.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
squashfs-tools development has been moved to git.kernel.org from
Sourceforge.
The hash was selected to fix a threading issue observed when
creating images on a build machine with at least 64 cores.
(Fixes random corrupted images that would fail to boot)
Signed-off-by: Atul Singh <atul.singh.mandla@rockwellcollins.com>
Signed-off-by: Matt Weber <matthew.weber@rockwellcollins.com>
[Thomas: use https:// instead of git://.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
KMS++ is a suite of library and test tools to interact with KMS drivers in
the linux kernel.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
[Thomas:
- rename prompt to kms++, suggested by Yann E. Morin
- fixup the thread dependency comment
- remove the mention of the python wrapper in the Config.in help
text, since they are not installed
- fix the Config.in comment to mention the C++ and gcc >= 4.8
dependencies
- use = instead of += when appropriate
- use a loop to install the test programs
- use a loop to install the libraries
- add installation to staging as well, both the libraries and header
files
- add missing dependency on host-pkgconf
- add hash file.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
wavemon is an ncurse-based wireless device monitoring application allowing to
watch signal and noise levels, packet statistics, device configuration and
network parameters of wireless network hardware.
[Peter: needs host-pkgconf, explain CC override]
Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Now that we have introduced the support for ARM no-MMU in Buildroot,
we need to update the dependencies of the musl external toolchain. It
supports only MMU-capable ARM cores, so it must depend on BR2_USE_MMU,
at least for the ARM platforms.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ARM big-endian is different from ARMv4/5/6 and ARMv7. Big-endian on
ARMv4/5/6 is BE-32 while big-endian on ARMv7 is BE-8, which are not
compatible.
Therefore, the musl big endian toolchain that is built for ARMv4
cannot work for ARMv7, it can only work for ARMv4/5/6.
This commit updates the musl toolchain dependency accordingly.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The "--no-patch" option used by the git downloader appeared on git
1.8.4. Systems with older git versions show an error and fall back to
the wget downloader, which isn't suitable for all the cases.
Signed-off-by: Enrique Ocaña González <eocanha@igalia.com>
Tested-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Tested-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
