New package which provides the driver and binary blob libraries for
the STM32MP157 vivante gcnano gpu. The precompiled libraries depend
on wayland and libdrm.
Signed-off-by: Jens Kleintje <scooby22@web.de>
[Thomas:
- add missing dependency on glibc
- add missing dependencies on threads (wayland, libdrm) and !static
(wayland)
- introduce GCNANO_BINARIES_LIB_VERSION,
GCNANO_BINARIES_DRIVER_VERSION and GCNANO_BINARIES_USERLAND_VERSION
variables to simplify code a bit
- greatly simplify the installation, by using more the available
libraries/files. This also avoids the installation of duplicate
release/debug libraries]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Fixes:
In file included from lnx_init.c:33:
../../../../hw/xfree86/common/compiler.h:767:10: fatal error: sys/io.h: No such file or directory
#include <sys/io.h>
^~~~~~~~~~
compilation terminated.
The ARM sys/io.h has been removed from upstream glibc, which is in
buildroot. This causes the xorg-server build to fail on ARM when using
the glibc toolchain. See [1], [2].
The following patches ([3], [4]) from upstream xserver fix this, but
have not yet been released.
[1] https://sourceware.org/glibc/wiki/Release/2.30#A.3Csys.2BAC8-io.h.3E_removed_on_32-bit_Arm
[2] https://gitlab.freedesktop.org/xorg/xserver/issues/840
[3] 6a2ce6c5da
[4] fe4cd0e7f5
Signed-off-by: Thomas Preston <thomas.preston@codethink.co.uk>
Reviewed-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The two helper programs TesturePackager and JsonSchemaBuilder are built
out of the same source tree as Kodi, so to avoid downloading the same
archive three times, let's them share the same download directory.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
A sub-component of Kodi (cpluff) needs to be autoreconf-ed, and uses
gettext, so calls to autopoint:
[ 2%] Performing autoreconf step for 'libcpluff'
autoreconf: Entering directory `.'
autoreconf: running: autopoint --force
Can't exec "autopoint": No such file or directory at [...]/host/share/autoconf/Auto4te/FileUtils.pm line 345.
autoreconf: failed to run autopoint: No such file or directory
autoreconf: autopoint is needed because this package uses Gettext
make[4]: *** [CMakeFiles/libcpluff.dir/build.make:121: build/cpluff/src/libcpluff-stamp/libcpluff-autoreconf] Error 1
make[3]: *** [CMakeFiles/Makefile2:615: CMakeFiles/libcpluff.dir/all] Error 2
Add a dependency to host-gettext to bring an appropriate variant that
provides autotpoint (gettext-tiny's autopoint works like a charm, for
the curious).
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When a textdomain is specified, we ignore it twice: we shift args, and
we printf ${2}, which would yield an empty printf.
Fix that by not shifting, and just printf ${2}; this is nicer.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
gettext-tiny also installs a fake autopoint, but it still needs to be
pointed at the m4 macros, which the template autopoint expects to be in
@datadir@, where datadir is the traditional autotools datadir, derived
from datarootdir, itself derived from prefix.
So, pass prefix so all the locations are properly pointing to HOST_DIR,
instead of the default /usr/local.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Vadim Kochan <vadim4j@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Updated patch 1 to handle changed context.
Removed patch 2 which is now upstream.
The update allows QEMU's configure script to successfully detect libfdt.
Signed-off-by: Sam Bobroff <sbobroff@linux.ibm.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The libfdt headers are installed at $(HOST_DIR)/include/libfdt now, to
prevent conflicts with the kernel dtc code. Pass that path to the qemu
configure script, otherwise it uses its bundled copy of libfdt.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Installing the libfdt headers at $(HOST_DIR)/include interferes with the
build of the kernel, as in beaglebone:
2019-11-01T22:23:16 /usr/bin/make -f ./scripts/Makefile.build obj=scripts
2019-11-01T22:23:16 /usr/bin/make -f ./scripts/Makefile.build obj=scripts/dtc need-builtin=
2019-11-01T22:23:16 (cat /dev/null; ) > scripts/dtc/modules.order
2019-11-01T22:23:16 /usr/bin/gcc -O2 -I/work/beaglebone/host/include -L/work/beaglebone/host/lib -Wl,-rpath,/work/beaglebone/host/lib -Wp,-MD,scripts/dtc/libfdt/.fdt_ro.o.d -Wall -Wmissing-prototypes -Wstrict-p>
2019-11-01T22:23:16 scripts/dtc/libfdt/fdt_ro.c:92:10: error: redefinition of ‘fdt_get_max_phandle’
2019-11-01T22:23:16 uint32_t fdt_get_max_phandle(const void *fdt)
2019-11-01T22:23:16 ^~~~~~~~~~~~~~~~~~~
2019-11-01T22:23:16 In file included from scripts/dtc/libfdt/fdt_ro.c:54:
2019-11-01T22:23:16 /work/beaglebone/host/include/libfdt.h:384:24: note: previous definition of ‘fdt_get_max_phandle’ was here
2019-11-01T22:23:16 static inline uint32_t fdt_get_max_phandle(const void *fdt)
2019-11-01T22:23:16 ^~~~~~~~~~~~~~~~~~~
2019-11-01T22:23:16 make[4]: *** [scripts/Makefile.host:107: scripts/dtc/libfdt/fdt_ro.o] Error 1
2019-11-01T22:23:16 make[3]: *** [scripts/Makefile.build:544: scripts/dtc] Error 2
2019-11-01T22:23:16 make[2]: *** [Makefile:1066: scripts] Error 2
2019-11-01T22:23:16 make[1]: *** [package/pkg-generic.mk:241: /work/beaglebone/build/linux-4dae378bbe721277b08699d1d88ffae12acc9b09/.stamp_built] Error 2
We cannot simply disable dtc support and patching the kernel makefiles
is risky and error-prone, so let's just install the host-dtc headers at
$(HOST_DIR)/include/libfdt to circumvent the problem.
Packages that depend on those headers (e.g. host-qemu) must be updated
to user the correct path but this is small change.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This adds support for different platforms where the binaries have
different extensions.
Signed-off-by: Michael Schenk <michael.schenk@noser.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Update configs with latest linux4sam and at91bootstrap.
Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
neardal is licensed under LGPL-2.0, which is (most likely) not license
compatible with readline (GPL-3.0+), so always use the libedit backend.
The choice is done by ./configure, which first checks for libedit, and
uses readline only as a fallback. Since we do build libedit before
neardal, that's what is going to be picked up.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr:
- add blurb about preference in ./configure
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
COPYING states LGPL-2.0, not GPL-2.0:
GNU LIBRARY GENERAL PUBLIC LICENSE
Version 2, June 1991
So use that for the license tag.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Quoting
https://lists.freedesktop.org/archives/mesa-announce/2019-November/000555.html
"This is an emergency release, to fix a critical bug found in the
19.2.3 release which causes incomplete rendering on all mesa drivers."
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
{python-,}readline is (no longer) a required dependency of kodi, and
readline (GPL-3.0+) is not license compatible with kodi (GPL-2.0), so drop
the dependency.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
In commit 6136765b23 ("toolchain:
generate check-headers program under $(BUILD_DIR)"), the
check_kernel_headers_version function was simplified to not check the
return value of the check-kernel-headers.sh script, assuming that
"make" does bail out on the first failing command.
However, check_kernel_headers_version when used in $(2)_CONFIGURE_CMDS
from pkg-toolchain-external.mk, is called in a sequence of commands,
where the return value of each command is not checked. Therefore, a
failure of check-kernel-headers.sh no longer aborts the build.
Since all other macros are using this principle of calling "exit 1",
we revert back to the same for check_kernel_headers_version, as it was
done prior to 6136765b23.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Carlos Santos <unixmania@gmail.com>
Reviewed-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
oracle-mysql is licensed under GPL-2.0, which is not license compatible with
modern readline (GPL-3.0+), so instead use the bundled older version
(GPL-2.0+ licensed) of readline instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
socat is licensed under GPL-2.0, which is not license compatible with
readline (GPL-3.0+), so drop the optional dependency and add a comment
explaining why.
This also matches how socat is packaged in Debian, where the man page has
the following snippet added:
READLINE
Uses GNU readline and history on stdio to allow editing and reusing input lines (example).
Due to licensing restrictions the readline feature is disabled in Debian. See BUGS.
You can use STDIO instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Reviewed-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
mariadb is licensed under GPL-2.0, which is not license compatible with
modern readline (GPL-3.0+), so instead use the bundled older version
(GPL-2.0+ licensed) of readline instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
lvm2 is licensed under GPL-2.0, which is not license compatible with
readline (GPL-3.0+), so drop the optional dependency and add a comment
explaining why.
Notice: The readline support is only used when the raw lvm tool is called
without arguments.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The tarball unfortunately does not include a dedicated license file, so
instead use the main source file for the setkey command.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
chrony is licensed under GPL-2.0, which is not license compatible with
readline (GPL-3.0+), so remove the optional readline handling and replace
with libedit instead.
While we are at it, also explicitly disable the libedit backend when not
available.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Dropwatch links with readline, which is GPL-3.0+, so not compatible with
GPL-2.0. When asked about this, upstream has clarified that the license
really is GPL-2.0+:
https://github.com/nhorman/dropwatch/issues/14
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr: add a coomet as suggested by Baruc]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Baruch Siach <baruch@tkos.co.il>
While the core connman code is licensed under GPL-2.0, the client code is
GPL-2.0+ for compatibility with readline (which is GPL-3.0+).
Extend the _LICENSE with this info to clarify that linking against
readline is OK licensing wise.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Use the github macro instead of git, to save download time.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
A C++ library for interacting with sd-bus and a C++ bindings generator.
The host build produces a code generation tool, sdbus++ while the
target build gives libsdbusplus.
The code generator requires host-python-pyyaml, host-python-inflection,
and host-python-mako. Since it isn't built for the target, the target
build does not require them.
The host package includes a code and documentation generation program,
which is useful for including in a buildroot SDK, even if no target
package needs it, so make it a user-selectable host package.
host-autoconf-archive is added as a dependency since configure.ac
uses the AX_PKG_CHECK_MODULES macro.
Signed-off-by: John Faith <jfaith@impinj.com>
Signed-off-by: Trent Piepho <tpiepho@impinj.com>
[Thomas: use host-python3 when BR2_PACKAGE_PYTHON3=y]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When the rcw tools is built, it is probably to generate the RCW (Reset
Configuration Word) of the target. So it makes sense to do this directly
in the Buildroot infra and install it in the images directory, just like
we do for e.g. device tree blobs.
Add an option BR2_PACKAGE_HOST_RCW_CUSTOM_PATH that allows to specify
the source for the RCW. And while we're at it, make it possible to
specify multiple .rcw and .rcwi files, just like for device trees.
Signed-off-by: Jared Bents <jared.bents@rockwellcollins.com>
[Arnout:
* Reword help text and commit message
* Remove definition of RCW_DELIVERY_FILE
* Define and use RCW_FILES
* Enclose everything in ifneq($(RCW_FILES),)
* Enclose error in ifeq($(BR_BUILDING),y)
* Add actual error message
* Don't limit to only %.rcwi files
* Don't delete *.bin
]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Thanks to the introduction of GPT partition table support in genimage,
this commit improves the pc_x86_64_efi_defconfig to remove the use of
the custom script creating the image.
Tested in QEMU, not on a physical device.
So:
- revert commit fee29b05bb7db25e37c8a5175ce00dc712554edf[1]
- add GPT support
- tweak shell script to add the correct UUID in genimage config.
[1]: https://git.buildroot.net/buildroot/commit/?id=fee29b05bb7db25e37c8a5175ce00dc712554edf
[2]: https://git.buildroot.net/buildroot/commit/?id=79b8540d624ac4846ba341b1b9691eccacf0bc05
Signed-off-by: Alexandre PAYEN <alexandre.payen@smile.fr>
Cc: Carlos Santos <casantos@datacom.com.br>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
[Thomas:
- drop commented code in post-build.sh
- take into account comments made by Carlos Santos in
http://patchwork.ozlabs.org/patch/1143502/]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
ninja depends on python3 specifically, but the configure.py file
simply uses "env python". Where no python is selected for the target
you simply won't get a python symlink in $(HOST_DIR)/usr/bin, so the
configure.py script fails to run since it can't find "python".
Notice that in order to reproduce the issue, you must not have
python2 installed on your host machine.
Signed-off-by: Avi Shukron <avraham.shukron@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This corresponds to -mcu=hs38 with mpy-option=9 (64-bit multiplier)
Signed-off-by: Vineet Gupta <vgupta@synopsys.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This includes the following changes:
3cebf12 dev_table: add STM32G07xxx/08xxx
5ad1f87 Makefile: Check if CC and AR are defined
ee5b009 Fix comparison between signed and unsigned integer
74bfe1a serial_w32: remove always false if-then
d302abf Add comment on switch fall-through
12f06e3 Fix warning for unused parameters with mingw-gcc
4957909 Fix compile warning for unused parameters
475da89 Fix signed/unsigned comparison
68cc278 Fix dev_table for Android compilation
2e09c51 Fix indentation warning from GCC 7.1.1
c4293d3 Flip PEMPTY bit before SW reset on STM32L452
acf15a2 dev_table: add device STM32L01xxx/02xxx
59d08c2 dev_table: add device family H7
12bfd33 dev_table: fix flash layout of STM32F72xxx/73xxx
0c84669 dev_table: add STM32L45xxx/46xxx
e2a4f65 dev_table: fix option byte end address for STM32L43xxx/44xxx
76da7c4 dev_table: L4: reorder with AN2606 rev 31
cc70634 dev_table: F0: reorder with AN2606 rev 31
8a0d423 dev_table: F4: reorder and align names with AN2606 rev 31
bc0876d dev_table: fix RAM size for STM32F412xx
2783493 Add some more F4/F7/L4 devices
27626f5 Add autotools support
6b6b6e5 Let user decide length for unknown memory areas
671561e Allow read/write to full option byte area and system memory
843c6f5 Don't write diagnostics messages to stdout unconditionally
3a07489 Don't always print banner to stdout
cc97bca Treat filenames starting with hyphen correctly
a1b2fe4 Improve validation of GPIO sequence
7683db8 Allow specifying more than 255 pages to erase
033aaa8 Report RAM and flash size as maximal values
1f10b4e Serial POSIX: Check if tty is already opened by another stm32flash
8aa685c Add ctrl-C signal handler
3daa90e Print message on failing read/write protect/unprotect
7b5b33a Verbose error message on GPIO signal change failure
d5df420 Add message and fix return code on failed boot entry sequence
a285a09 Flush port after boot GPIO sequence
ee7c963 Fix return code from GPIO bootloader exit sequence
de7332e Fix return code from GPIO bootloader entry sequence
efeab97 Allow extra delay or no delay at all in GPIO sequences
d202bd0 Make termios timeout settable via TERMIOS_TIMEOUT_MS define
0242da6 Ensure correct precedence of bit and logic comparisons
26854ef Fix return value in case of successful protect/unprotect operation
3772e95 Fix return code in case of flash failure
7a7b4c9 Always run GPIO exit sequence if present
d581ce8 Put Linux specific GPIO code under conditional compile
8c4aa65 dev_table: Mark 0x417, 0x429, 0x427 for no mass-erase
2381ce3 Fix parity setting/checking on *BSD and MacOSX
5361ed8 Fix for device 0x442: System memory start address
Also add hash for license file
Signed-off-by: Pierre-Jean Texier <pjtexier@koncepto.io>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Install QtWayland only if the latest Qt version is selected. It does not
exist in version 5.6.
Signed-off-by: Carlos Santos <unixmania@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Docker fails to start with "Devices cgroup isn't mounted" as of systemd 243.
According to the systemd documentation:
systemd now defaults to the "unified" cgroup hierarchy setup during
build-time, i.e. -Ddefault-hierarchy=unified is now the build-time default.
Previously, -Ddefault-hierarchy=hybrid was the default. [...] Downstream
production distributions might want to continue to use
-Ddefault-hierarchy=hybrid (or even =legacy) for their builds as unfortunately
the popular container managers have not caught up with the kernel API changes.
Changing this option to "hybrid" or "legacy" fixes the Docker startup.
Reference: https://github.com/opencontainers/runc/issues/654
Signed-off-by: Christian Stewart <christian@paral.in>
Tested-by: Jérémy Rosen <jeremy.rosen@smile.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Build can sometimes fails on:
src/svgtiny.c:21:10: fatal error: autogenerated_colors.c: No such file or directory
#include "autogenerated_colors.c"
^~~~~~~~~~~~~~~~~~~~~~~~
because svgtiny.c does not properly depends on autogenerated_colors.c
that is built by gperf. So, just disable parallel build instead of
trying to fix this issue especially because libsvgtiny uses the netsurf
buildsystem
Fixes:
- http://autobuild.buildroot.org/results/48e7a7f7c72634d59cca817778d31661bfe8e72f
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Just some bugfixes, including an important one to fix long indication
messages that are split into multiple fragments.
Signed-off-by: Aleksander Morgado <aleksander@aleksander.es>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
