libunwind configure script explicitly links libunwind against
libgcc_s. libgcc_s is only guaranteed to be available for toolchains
that supports dynamic linking: pure static linking toolchains only
have libgcc.a, not libgcc_s.so.
Therefore, let's make libunwind unavailable on toolchains that lack
dynamic linking support. We could potentially support linking with
libgcc, but switching to libgcc_s was done upstream because libgcc was
lacking some symbols on ARM
(https://lists.nongnu.org/archive/html/libunwind-devel/2014-06/msg00024.html). Even
though recent gcc versions seem to provide such symbols in libgcc.a,
having libunwind available on static linking configurations is not a
useful enough use-case to do the necessary research to find when this
issue was fixed in gcc.
Since libunwind is not used as a mandatory dependency in any package,
adding this !BR2_STATIC_LIBS dependency is trivial and nicely avoids
the problematic situation.
This fixes two different autobuilder failures:
- Gstreamer 1.x programs failing to link, because libunwind links
against libgcc_s that isn't available (static linking
configuration):
http://autobuild.buildroot.net/results/9d4fbf7167e9afce0eef5c9e0cfd42c966ecba36/
- Gmrender-resurrect, which fails to link, because GStreamer 1.x uses
some libunwind functionality, but does not take into account the
libunwind dependency in its .pc files (static linking
configuration):
http://autobuild.buildroot.net/results/0a3a2485c187a000482c178f1e9c64dd716a858f/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit is similar to 350941e31d
("python: remove target Python packages from PYTHONPATH") but for
python3.
We currently have
$(TARGET_DIR)/usr/lib/python$(PYTHON3_VERSION_MAJOR)/site-packages/
inside the PYTHON3_PATH variable, which gets used to define
PYTHONPATH, passed to the host Python interpreter when
building/installing target packages.
However, this is terribly wrong, as it causes the host interpreter to
potentially import target Python packages. This is wrong for several
reasons:
- Some Python packages might need some Python modules to be installed
on the host (described in setup_requires in setup.py), but their
installation currently works because by luck the corresponding
Python module is installed for the target. Some of those cases were
happening for real, and fixed by previous patches.
- Some Python packages include some native code, therefore built for
a specific CPU architecture. When you point the host Python
interpreter to native libraries built for the target, you get nice
build failures, such as the one affecting the python-cffi related
packages.
This change fixes the following build failures:
http://autobuild.buildroot.net/results/9005b89407e46b537a54cac6cc0c69dcac4dc5ea/
(python-cryptography)
http://autobuild.buildroot.net/results/395682d33d02fdcaa39d3c0326355bd9ea3d6feb/
(python-pynacl)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com>
Tested-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit backports three patches that are already upstream in
kvmtool fixing build warnings with musl. Those are not strictly needed
for the build to succeed, they just reduce the amount of warning
noise.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
In Linux 4.12, the header <asm/msr-index.h> has been removed from the
set of headers exported to userspace. Therefore, it cannot be used by
kvmtool anymore. This commit takes the simple approach of duplicating
inside kvmtool the MSR_* definitions that were used from this
<asm/msr-index.h> header.
This fixes:
x86/kvm-cpu.c:7:27: fatal error: asm/msr-index.h: No such file or directory
#include <asm/msr-index.h>
Which is the second part of:
http://autobuild.buildroot.net/results/4459a909e735343d1cf768d30466bc3c57eca19e/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit backports an upstream patch that fixes the build of
kvmtool with musl:
In file included from builtin-balloon.c:9:0:
include/kvm/kvm.h:22:0: warning: "PAGE_SIZE" redefined
#define PAGE_SIZE (sysconf(_SC_PAGE_SIZE))
Fixes one part of:
http://autobuild.buildroot.net/results/4459a909e735343d1cf768d30466bc3c57eca19e/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
i.MX1/L boards require an old Linux kernel version to run, which is not
compatible with latest Buildroot needs.
Will fix:
https://gitlab.com/buildroot.org/buildroot/-/jobs/27873568
Signed-off-by: Julien BOIBESSOT <julien.boibessot@armadeus.com>
[Thomas: update .gitlab-ci.yml file.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit adds a patch to the norm package that fixes the build with
gcc 7.x. Many thanks to Romain Naour for pointing out the solution to
this C++ build problem.
Fixes:
http://autobuild.buildroot.net/results/c79dc84cdc34d62199099eb4438b1aed3e7459bb/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
[Thomas: add information that the patch has been submitted upstream
and accepted.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Our new WiFi 802.11ac BT4.1 module (BD-SDMAC [1]) driver allows to
override the wlan MAC address using a module parameter.
Since its driver is now included in our external repository [2], update
the bootscript so it sets the parameter for that driver too.
As a FYI, $wlmac is based on Ethernet MAC address located in fuses.
[1] https://boundarydevices.com/product/bd_sdmac_wifi/
[2] https://github.com/boundarydevices/buildroot-external-boundary
Signed-off-by: Gary Bisson <gary.bisson@boundarydevices.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since 392b0a26f5 (toolchain-external: default BR2_TOOLCHAIN_EXTERNAL_PATH
to empty), calling 'make clean' or similar can yield a spurious stderr
message:
dirname: missing operand
Try 'dirname --help' for more information.
Which is definitely baffling and unsettling...
It turns out that it is pretty trivial to reproduce, and this defconfig
is just enough:
$ cat my-defconfig
BR2_TOOLCHAIN_EXTERNAL=y
$ make BR2_DEFCONFIG=$(pwd)/my-defconfig defconfig
$ make clean
dirname: missing operand
Try 'dirname --help' for more information.
[--snip--]
This is because the cross-compiler is not found in the PATH (and for
good reasons, I don't have it in the PATH, not even at all).
So, when the cross-compiler is not found in the path, we simply
continue as if all was good, and postpone the check to much later,
when we try to copy the toolchain libs...
So, use a make construct rather than calling to the shell: $(dir ...)
does not whine if passed nothing.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since Toradex V2.7 SDK, U-Boot image has changed from u-boot.imx to
SPL + u-boot.img.
Also fix U-Boot build error with GCC6.
Signed-off-by: Sergio Prado <sergio.prado@e-labworks.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Fix a build failure in case (a non functional) sphinx documentation
system is installed on the host (reported [1] and fix tested [2]
by grunpferd@netscape.net).
Fixes:
sphinx-build -b html -d ./doctrees . ./html
Error: Source directory doesn't contain a conf.py file.
[1] http://lists.busybox.net/pipermail/buildroot/2017-August/200021.html
[2] http://lists.busybox.net/pipermail/buildroot/2017-August/200267.html
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Cc: grunpferd@netscape.net
[Arnout:
- use --disable-sphinx-doc instead of a cv variable
- remove the comment, it speaks for itself]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit 2c8de6c4 (gcc 4.9.1: add patch for PR60102) removed the SPE
condition becasue of said PR, but forgot to remove the associated
comment, which has been tagging along all this time...
Remove it, it is no longer valid and causes confusion.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The VC4 GPU does not support full GL, it only provides EGL.
Currently, it is possible to build the VC4 backend without EGL support,
but that does not make sense in the slighest.
So, forcibly enable EGL with VC4, like is done for etnaviv and virgl.
Reported-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This adds a patch cherry-picked from libepoxy 1.4.2, which adds missing
NULL-pointer checks and avoids segmentation faults when using libepoxy
under X11, when the server does not have the GLX extension, or it is
disabled -- and applications can still use EGL.
Signed-off-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The github repo is more informative than the list Marco's software on
the previous link.
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
When we use the cross-compiler to build syslinux with a recent binutils
version, it fails with:
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld -Bsymbolic -pie -E --hash-style=gnu -T
/builds/arnout/buildroot/output/build/syslinux-6.03/core/i386/syslinux.ld -M -o ldlinux.elf ldlinux.o \
--start-group libcom32.a --whole-archive /builds/arnout/buildroot/output/build/syslinux-6.03/bios/com32/lib/libcom32core.a libldlinux.a --end-group -N
--no-omagic \
> ldlinux.map
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld: ldlinux.elf: Not enough room for program headers, try linking with -N
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld: final link failed: Bad value
/builds/arnout/buildroot/output/build/syslinux-6.03/core/Makefile:167: recipe for target 'ldlinux.elf' failed
Backport an upstream patch that reorganises the i386 bios build by
removing some symbols and making others hidden. To simplify the
backport, an additional patch that also touches the link script is
also included - it anyway looks like that patch could be relevant as
well.
Partially fixes: https://gitlab.com/arnout/buildroot/-/jobs/28979377
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Also the last two are regenerated, their context has changed due to
the patches that have been removed earlier.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit fixes the build of libspatialindex on the Microblaze
architecture by working around a gcc bug, by passing -O0 as the
optimization level.
A patch to configure.ac is needed to not have -O2 be forced by the
libspatialindex build system, and therefore AUTORECONF=YES is now
needed.
Fixes:
http://autobuild.buildroot.net/results/bbba2a2c97dbec21340c7fd07162a316a411cba4/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
In case we're using an initrd, we create an empty "root" directory that
will contain only the bootloader stuff, not the actual root filesystem,
because it is in an initrd (standalone or initramfs).
We have to ensure that the directory is empty before assembling the
filesystem (to avoid any file lingering from a previous run, like the
sequence "make; make"). So we first remove it before we create it, so
that on each build (especially not-from-scratch builds) we get the exact
expected content without any leftover.
However, the macro responsible for that, although defined since 7080eef9,
was never called.
Fix that by registering it as a pre-gen hook.
Note: the directory need not be created, as there are quite a few
"install -D" commands that ensure it is created. Yet, we prefer to
create it explicitly to avoid any confusion.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Although the issue can very well occur with low-paralle builds, or even
with non-parallel builds, the conditions are so strict that the ocasion
it breaks is extremely rare, to the point where a failure would go
unnoticed.
Fixes#10141.
Reported-by: Clayton Shotwell <clayton.shotwell@rockwellcollins.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Clayton Shotwell <clayton.shotwell@rockwellcollins.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
elf2flt contains a program called ld-elf2flt, which gets installed as
a replacement for ld, with ld renamed to ld.real. This program
therefore calls ld.real internally.
The logic to find ld.real worked fine for:
- bin/TARGET_ALIAS-ld
- TARGET_ALIAS/bin/ld
However, it failed badly if bin/SIMPLER_TARGET_ALIAS-ld is used, as
symlink to bin/TARGET_ALIAS-ld. For example:
$ ./output/host/bin/arm-buildroot-uclinux-uclibcgnueabi-ld
/home/thomas/buildroot/buildroot/output/host/bin/../arm-buildroot-uclinux-uclibcgnueabi/bin/ld.real: no input files
$ ./output/host/bin/arm-linux-ld
arm-linux-ld (ld-elf2flt): error trying to exec '/home/thomas/buildroot/buildroot/output/host/bin/ld.real': execvp: No such file or directory
$ ./output/host/arm-buildroot-uclinux-uclibcgnueabi/bin/ld
/home/thomas/buildroot/buildroot/output/host/arm-buildroot-uclinux-uclibcgnueabi/bin/ld.real: no input files
This commit fixes that by adding a patch that adjusts the ld-elf2flt
logic to properly cope with this situation.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Fixes serious usability issues (Debian changelog):
* Fixed whois referrals for .com, .net, .jobs, .bz, .cc and .tv,
broken by an ICANN-mandated output change:
https://www.icann.org/resources/pages/rdds-labeling-policy-2017-02-01-en
Use snapshot.debian.org for the .dsc reference, since files tend to
disappear from the official Debian ftp site over time.
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Disable O2 and add O0. The suggested flags in the gcc report
are not enough for gcc >= 6.
Fixes:
http://autobuild.buildroot.net/results/3686cd3f3e7b6aee84f4377bd2dded1115321fb3/
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
[Thomas: use Git to format the patch, improve patch description.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Recent glibc have deprecated the implicit include of sys/sysmacros.h
from sys/types.h. That means that the macros major and minor are no
longer defined unless this header is included.
This problem was observed for host-e2fsprogs when building on a host
with recent glibc.
Add an upstream patch that includes sys/sysmacros.h when needed.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Dagg Stompler <daggs@gmx.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Tested-by: Dagg Stompler <daggs@gmx.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Unbreak failing defconfig.
Resulting kernel and rootfs booted on ci20 device.
While at it, drop the useless BR2_TARGET_UBOOT_VERSION variable, this
variable is a hidden variable, so it is useless to mention it in a
defconfig.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
[Thomas: drop BR2_TARGET_UBOOT_VERSION.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The lua staging and host installation commands generate a file in
usr/lib/pkgconfig, without first making sure that this directory
exists, which causes build failures if it doesn't. This commit adjusts
those installation commands to create this directory if needed.
Fixes:
http://autobuild.buildroot.net/results/101c89e1d6aee942a0b1c4e4f3daf8ac2414a56c/
Based on investigation and initial (more complicated) fix provided by
Francois Perrad <francois.perrad@gadz.org>.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit adds a patch to the sysvinit package that fixes various
build issues against musl due to missing header includes.
Signed-off-by: Dagg Stompler <daggs@gmx.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Probably not noticed before because dtc was already built or it was
installed on the host.
Fixes: https://gitlab.com/arnout/buildroot/-/jobs/28979369
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Michał Łyszczek <michal.lyszczek@bofc.pl>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The skeletons are based on the selection of BR2_INIT_*, so add init- to
the package name to make this clearer. The name skeleton-common implies
that it is common to all skeletons, yet it does not apply to
skeleton-custom. It is only common to the skeleton-init-* packages, so
name it the same way.
Signed-off-by: Cam Hutchison <camh@xdna.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The skeletons are based on the selection of BR2_INIT_*, so add init- to
the package name to make this clearer. While skeleton-sysv is relatively
clear, skeleton-common and skeleton-none are less clear on their
relationship to BR2_INIT_*. So rename skeleton-sysv to conform to a
clearer pattern.
Signed-off-by: Cam Hutchison <camh@xdna.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The skeletons are based on the selection of BR2_INIT_*, so add init- to
the package name to make this clearer. While skeleton-systemd is
relatively clear, skeleton-common and skeleton-none are less clear on
their relationship to BR2_INIT_*. So rename skeleton-systemd to conform
to clearer pattern.
Signed-off-by: Cam Hutchison <camh@xdna.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The skeletons are based on the selection of BR2_INIT_*, so add init- to
the package name to make this clearer. The name skeleton-none implies no
skeleton at all, not a base skeleton with no init-specific files.
Signed-off-by: Cam Hutchison <camh@xdna.net>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This release contains only two real changes to fix two bugs. One is
only relevant for Windows; the other one fixes handling of empty
comment lines.
libconfuse is used by genimage, and we have one defconfig,
linksprite_pcduino, that has empty comment lines in its genimage.cfg.
Thus, it fixes https://gitlab.com/buildroot.org/buildroot/-/jobs/27873642
While we're at it, also change the upstream URL in Config.in to github.
The old URL on nongnu was just a download directory and didn't have
anything newer than 2010, while github has a nice README.md.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Frank Hunleth <fhunleth@troodon-software.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This reverts commit c69088b8c3, which is
no longer needed now that alsa-lib has been fixed properly.
Indeed, this commit was made because alsa-utils failed to build with:
/home/buildroot/build/instance-0/output/host/usr/xtensa-buildroot-linux-uclibc/sysroot/usr/lib/libasound.a(pcm_dmix.o): In function `snd_pcm_dmix_htimestamp':
pcm_dmix.c:(.text+0x5c0): warning:
/home/buildroot/build/instance-0/output/host/usr/xtensa-buildroot-linux-uclibc/sysroot/usr/lib/libasound.a(rawmidi_symbols.o):(.data.rel+0x4): undefined reference to `_snd_module_rawmidi_virt'
collect2: error: ld returned 1 exit status
Which is exactly the problem that we fixed in alsa-lib by making sure
that rawmidi_symbols.c doesn't reference symbols from rawmidi_virt.c
when SEQ support is not enabled.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This commit adds a patch to alsa-lib that fixes an incorrect code
dependency: some code built when MIDI support is enabled is
unconditionally referencing code that is built when SEQ support is
enabled, both options being independent.
This causes build failures such as:
/home/buildroot/buildroot-test/instance-1/output/host/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libasound.a(pcm_dmix.o): In function `snd_pcm_dmix_sync_ptr':
pcm_dmix.c:(.text+0x83c): warning:
/home/buildroot/buildroot-test/instance-1/output/host/arm-buildroot-linux-uclibcgnueabi/sysroot/usr/lib/libasound.a(rawmidi_symbols.o):(.data+0x4): undefined reference to `_snd_module_rawmidi_virt'
Fixes:
http://autobuild.buildroot.net/results/5964e7028c31a8f75b38a27270410f0b3c5c8422/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Arnout: add upstream link to the patch]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
The beaglebone_qt5_defconfig uses an older kernel version which does not
provide a device tree for the beagle bone green.
post-image.sh now selects genimage_linux41.cfg if am335x-bonegreen.dtb is
not build and genimage.cfg otherwise.
Signed-off-by: Lothar Felten <lothar.felten@gmail.com>
[Thomas: minor tweaks.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since the bump to 7.55.0, libcurl fails to build on a number of
uncommon architectures (ARC, OpenRISC, etc.). This is due to upstream
commit 73a2fcea0b4adea6ba342cd7ed1149782c214ae3 ("includes: remove
curl/curlbuild.h and curl/curlrules.h"), which makes libcurl rely on
more architecture-specific related defines in include/curl/system.h.
This commit therefore adds a patch that fixes the 32-bit vs. 64-bit
detection for all architecture, using gcc's __SIZEOF_LONG__
definition. It has been tested successfully with test-pkg on all 47
toolchain configurations.
Fixes:
http://autobuild.buildroot.net/results/bf26c08cf3267214278674472f931603f69951ae/
(and many similar issues)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The linux 4.1 kernel does not compile with gcc 6.x.
The compiler version is now fixed to 5.x until a new kernel is used.
Signed-off-by: Lothar Felten <lothar.felten@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
