In order to be able to use OpenGL with QT5 with a Rockchip Mali
Bifrost GPU, a device integration provider has to be selected. We are
interested in the KMS integration that works well with Mali Bifrost,
make sure it gets build when using rockchip-mali-bifrost as OpenGL
provider.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Now that we have a libgbm virtual package, it makes sense to use it to
replace the open-coded list of gbm providers in qt5base. All of
mesa3d, gcnano-binaries, ti-sgx-um and imx-gpu-viv are libgbm
providers now.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Interestingly, only the aarch64 blobs provide libgbm support. This
will help in fixing an issue that exists today, where ARM32 +
imx-gpu-viv + wayland + qt5 fails to build, because qt5 is told that
libgbm is available, i.e the following configuration:
BR2_arm=y
BR2_cortex_a9=y
BR2_ARM_ENABLE_VFP=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_TOOLCHAIN_EXTERNAL_BOOTLIN=y
BR2_INIT_NONE=y
BR2_SYSTEM_BIN_SH_NONE=y
# BR2_PACKAGE_BUSYBOX is not set
BR2_PACKAGE_QT5=y
BR2_PACKAGE_QT5BASE_GUI=y
BR2_PACKAGE_QT5BASE_EGLFS=y
BR2_PACKAGE_FREESCALE_IMX=y
BR2_PACKAGE_FREESCALE_IMX_PLATFORM_IMX6Q=y
BR2_PACKAGE_IMX_GPU_VIV=y
BR2_PACKAGE_IMX_GPU_VIV_OUTPUT_WL=y
# BR2_TARGET_ROOTFS_TAR is not set
fails to build during the qt5base configure step with:
ERROR: Feature 'gbm' was enabled, but the pre-condition 'libs.gbm' failed.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Depending on the kernel driver is annoying because it means you can't
build ti-sgx-um without enabling a kernel. This prevents easy testing
of ti-sgx-um, and it also means ti-sgx-um is never built by the
autobuilders.
So instead, we select the kernel driver if a kernel is built by
Buildroot. This requires propagating the BR2_arm dependency from
ti-sgx-km to ti-sgx-um.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The OpenGL ES implementation provides a libgbm implementation, so
provide the corresponding virtual package.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The ti-sgx-um package provides 3 virtual packages, but doesn't have
the TI_SGX_UM_PROVIDES variable. This commit fixes that by adding the
missing variable.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add a new package to build and install the Mali kernel driver for
Midgard and Bifrost GPU.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Continuation lines should be indented one level more than the line the
are the continuation of.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
- Update indentation in hash file (two spaces)
- This bump will fix the following build failure with libest which was
added with commit f6f0e1e581:
In file included from /home/buildroot/autobuild/instance-0/output-1/host/bin/../nios2-buildroot-linux-gnu/sysroot/usr/include/libsafec/safe_lib.h:43:0,
from est_server_http.c:39:
/home/buildroot/autobuild/instance-0/output-1/host/bin/../nios2-buildroot-linux-gnu/sysroot/usr/include/libsafec/safe_types.h:42:9: error: unknown type name 'size_t'
typedef size_t rsize_t;
^~~~~~
https://github.com/rurban/safeclib/blob/v02092020/ChangeLog
Fixes:
- http://autobuild.buildroot.org/results/5a12f1cb4153834078522b6068ff24ac0300d51d
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add "Linux" before kernel in comment to be consistent with other
packages and manual
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Add "Linux" before kernel in comment to be consistent with other
packages and manual
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Add "Linux" before kernel in comment to be consistent with other
packages and manual
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
2.16.12 was the last release of the 2.16 long-time support branch.
Users who want a long-time branch should move to mbedtls-2.28, which
is backward-compatible and will be supported for at least 3 years.
Use LICENSE file as apache-2.0.txt was dropped with
c76a54554chttps://github.com/ARMmbed/mbedtls/releases/tag/v2.28.0
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Drop duplicated sentence from Config.in
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
A module for monitoring memory usage of a python program.
Signed-off-by: Asaf Kahlon <asafka7@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
riscv64gc was made available for BR2_riscv, but RISC-V can be 32-bit
or 64-bit, so we need to restrict it to BR2_RISCV_64. There's no need
to keep the BR2_riscv dependency, as BR2_RISCV_64 can only be true
when BR2_riscv is true.
Also, BR2_PACKAGE_HOST_RUSTC_ARCH needs to be set to riscv64gc to
allow rust-bin to download its pre-compiled standard library
correctly.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
BR2_PACKAGE_HOST_RUSTC_ARCH only had a special value for
BR2_ARM_CPU_ARMV7A, but it also needs a special value for
BR2_ARM_CPU_ARMV5, as the pre-compiled Rust standard library for
ARMv5TE is identified with the "armv5te" architecture name, see
https://doc.rust-lang.org/nightly/rustc/platform-support.html.
We noticed this because Rust binaries wouldn't work on an ARMv5
platform (Illegal instruction). This was due to the usage of the
arm-unknown-linux-gnueabi variant of the Rust standard library, which
is for ARMv6. Thanks to this commit, we correctly use the
armv5te-unknown-linux-gnueabi variant, and Rust binaries work properly
on ARMv5TE.
A better approach would be to do the conversion from architecture
options to Rust tuples in a single string symbol that also defines the
supported architectures, similar to how it's done in e.g. openblas.
However, that's a much bigger change. So for now, just do the easy thing
and fix this one issue.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The host variant was only needed by ca-certificates, which no longer
needs it, so drop this variant.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The host-python-cryptography module is only used by ca-certificates
for a check of the expiration date of certificates, which is only a
warning not even causing the build to abort, i.e something that
Buildroot users are most likely never going to see.
Since the host-python-cryptography dependency would soon require a
dependency on rust, it's a lot simpler to just patch the
certdata2pem.py script to no longer require cryptography, but only
make use of it if available.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The comments were *after* the option they applied to, which was
confusing, so bring back the comments before the option they apply to.
In addition, instead of saying "this option", explicitly name the
option to clear any ambiguity.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
There is already a package adding Mali blobs for Sunxi Mali Utgard
GPUs. Many Rockchip SoCs feature a Mali Bifrost GPU and need specific
libraries provided under binary form with a non-GPL compliant license
by Rockchip. Package these libraries.
While the repository at https://github.com/rockchip-linux/libmali
supports many GPU variants, this package for now only supports the
Bifrost GPU blob. Support for other GPU blobs can be added later.
Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
[Thomas:
- Rename to rockchip-mali, because it contains other blobs than
Bifrost ones
- Add missing entry in DEVELOPERS file
- Change BR2_arm to BR2_ARM_EABIHF as only EABIhf blobs are provided
for arm32
- Add BR2_TOOLCHAIN_HAS_THREADS inherited from libdrm.
- Expose as a libgbm provider.
- Drop the patch, which was not necessary.
- Rework the entire installation logic, to be much simpler.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
python-numpy can be built using fortran compiler if available. Pass the
F90 environment variable to setup.py to make sure it picks up the
cross-compiler. Otherwise, use --fcompiler to make sure it doesn't
accidentally pick up the host compiler.
Don't do anything like this for host compilation. We don't care whether
or not it uses fortran.
Signed-off-by: Guillaume W. Bres <guillaume.bressaix@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Removed patch which was applied upstream:
2dc5bea906
Updated license hash due to typo fixes:
73eca0a9c9
Updated project URL as recommended in the release notes:
https://github.com/unicode-org/icu/releases/tag/release-70-1
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The PYTHON_NUMPY_FIXUP_NPY_PKG_CONFIG_FILES macro was introduced in
commit
2426002d6f ("package/python-numpy: fixup
npymath.ini for numpy extensions").
However, it turns out that by using $(PYTHON3_PATH), we can simplify
the intermediate variables, to the point where they are no longer
needed. Not having the intermediate variables makes the logic a bit
easier to understand.
Also, the logic is changed to use a single sed expression, which we
pass to $(SED).
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The commit [1] added a sed command used to retreive a pattern
to keep only defconfigs whose name start with the pattern.
"<foo>-defconfigs-<pattern>"
The sed command doesn't work as expected if <foo> contains a
single hyphen [2]:
"qemu-6.2.0-defconfigs-qemu"
Update the sed command to ignore completely the part before
"-defconfigs-".
[1] 65d2f04c01
[2] http://lists.busybox.net/pipermail/buildroot/2022-January/632507.html
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Shortly after MSTPd was added, version 0.1.0 was released. Update to the
latest version.
Signed-off-by: Colin Foster <colin.foster@in-advantage.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Drop first patch (already in version)
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Fix build failure raised since the addition of the libexecinfo package
in commit eea8ba446c
Fixes:
- Not autobuilder failures (yet)
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This platform has its own kernel defconfig in Buildroot, but we cannot
get quick idea about how much it diverged from the in-kernel defconfig.
Let's use the upstream arch/arm/config/versatile_defconfig as a base,
and maintain the diff as a merge-config fragment. The same .config is
still generated based on the 5.10.7 kernel.
The diff is quite big, but this is a good start-point for cleanups.
Follow-up works can drop diff lines unless we find a good reason for
divergence.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
[Arnout: rename linux-nommu.config to linux-nommu.fragment]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This platform has its own kernel defconfig in Buildroot, but we cannot
get quick idea about how much it diverged from the in-kernel defconfig.
Let's use the upstream arch/arm/config/versatile_defconfig as a base,
and maintain the diff as a merge-config fragment. The same .config is
still generated (based on the 5.10.7 kernel).
The diff is quite big, but this is a good start-point for cleanups.
Follow-up works can drop diff lines unless we find a good reason for
divergence.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
[Arnout: rename linux.config to linux.fragment]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This defconfig enables sunxi-mali-mainline-driver so we need to disable
Lima in Linux to prevent Mali to not load.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This defconfig enables sunxi-mali-mainline-driver so we need to disable
Lima in Linux to prevent Mali to not load.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This Linux defconfig fragment is needed to disable Lima driver in favour of
Mali(if enabled). This because Lima is enabled by default in Linux's
sunxi_defconfig and if we enable Mali driver in Buildroot it will fail to
load. So let's set CONFIG_DRM_LIMA=n. Of course this fragment is only used
by Buildroot defconfigs that enable Mali driver.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
