Host build fails on updated Arch Linux desktop.
My current /usr/bin/gcc version is 14.1.1 20240522, where
implicit-function-declaration cause build to fail:
2024-06-05T07:03:20 libtool: compile: /home/roy/hymatek/connexi-touchpanel-firmware/mxxf1/output/host/bin/ccache /usr/bin/gcc -DHAVE_CONFIG_H -I. -I.. -I. -I../include -Iinclude -I../src -I/home/roy/hymatek/connexi-touchpanel-firmware/mxxf1/output/host/include -Wall -O2 -I/home/roy/hymatek/connexi-touchpanel-firmware/mxxf1/output/host/include -fexceptions -c ../src/tramp.c -fPIC -DPIC -o src/.libs/tramp.o
2024-06-05T07:03:20 ../src/tramp.c: In function ‘ffi_tramp_get_temp_file’:
2024-06-05T07:03:20 ../src/tramp.c:262:22: error: implicit declaration of function ‘open_temp_exec_file’ [-Wimplicit-function-declaration]
2024-06-05T07:03:20 262 | tramp_globals.fd = open_temp_exec_file ();
2024-06-05T07:03:20 | ^~~~~~~~~~~~~~~~~~~
Patch from master was added to fix build.
Signed-off-by: Roy Kollen Svendsen <roy.kollen.svendsen@akersolutions.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following CVEs:
CVE-2024-24789: archive/zip: mishandling of corrupt central directory record
CVE-2024-24790: net/netip: unexpected behavior from Is methods for IPv4-mapped IPv6 addresses
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When using imx-mkimage 6.1.36_2.1.0 or later, an additional data structure
is inserted in the generated image. The FIT external data position passed
to the uboot mkimage program needs to be adjusted accordingly.
Fixes: https://bugs.busybox.net/show_bug.cgi?id=15973
Fixes: 72de789023 ("package/imx-mkimage: bump version to lf-6.1.36-2.1.0")
Tested-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
[Sébastien: Tested on i.MX8MM EVK and i.MX8MP EVK]
[Sébastien:
- fix subject
- add Tested-by tag
- fix Fixes tags
]
Signed-off-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit baaf7f738a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
It has been well over 10 years since glibc 2.14 was released; the last
Debian version that had an earlier glibc was Wheezy, which Freexian
stopped to maintain as an ELTS in June 2020, 4 years ago, while the
oldest still maintained Ubuntu has glibc 2.21. It is now safe to assume
glibc 2.14 on all major, relevant distributions nowadays.
The distutils module is no longer bundled with python 3.12 so this
eliminates the need to install additional python modules under python
3.12.
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
[yann.morin.1998@free.fr: add Debian and Ubuntu references]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 929a491f40)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When systemd and tpm2-tss with fapi support are enabled, the fakeroot
script fail with:
<stdin>:35: Failed to resolve user 'tss': No such process.
When fapi support is enabled, tpm2-tss package install additional
configuration files that are expecting tss user and group exist [1].
/etc/sysusers.d/tpm2-tss.conf
/etc/tmpfiles.d/tpm2-tss-fapi.conf
The build fail in the fakeroot environment while handling tmpfiles
installed by tpm2-tss with fapi by host-systemd.
tss user and group is currently created by the tpm2-abrmd package but
tpm2-tss package also provide a udev rule file tpm-udev.rules [2] that
set the ownership of dev nodes /dev/tpmX and /dev/tpmrmX to tss
user/group. So tpm2-tss package must define TPM2_TSS_USERS to create
tss user and group, not tpm2-abrmd package.
So, move TPM2_ABRMD_USERS to TPM2_TSS_USERS.
Note: tpm2-abrmd is nowadays deprecated since the in-kernel Resource
Manager (available since kernel 4.12) is preferred [3].
[1] https://github.com/tpm2-software/tpm2-tss/blob/4.1.3/INSTALL.md?plain=1#L184
[2] https://github.com/tpm2-software/tpm2-tss/blob/4.1.3/dist/tpm-udev.rules
[3] https://github.com/tpm2-software/tpm2-abrmd/blob/3.0.0/README.md?plain=1#L39
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 2cdd3d1ccf)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The tarball (with .tar.gz extension but same content) is also available over
https from sourceforge, so use that instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit e5993e6cd4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The ftp server does not respond, so change to https:// instead. Here the
3.2.6 version is under older_versions.
wget ftp://ftp.ncftp.com/ncftp/ncftp-3.2.6-src.tar.gz
--2024-06-03 08:10:44-- ftp://ftp.ncftp.com/ncftp/ncftp-3.2.6-src.tar.gz
=> ‘ncftp-3.2.6-src.tar.gz’
Resolving ftp.ncftp.com (ftp.ncftp.com)... 209.197.102.38
Connecting to ftp.ncftp.com (ftp.ncftp.com)|209.197.102.38|:21... ^C
http redirects to https, so update the help text to match.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 3eb84214a1)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The ftp server does not allow anonymous login, causing the download to fail:
wget --passive-ftp -nd -t 3 -O '/home/peko/source/buildroot/output-qt/build/.lynx2.8.9rel.1.tar.bz2.XlcdCK/output' 'ftp://ftp.invisible-island.net/lynx/tarballs/lynx2.8.9rel.1.tar.bz2'
--2024-06-02 22:21:49-- ftp://ftp.invisible-island.net/lynx/tarballs/lynx2.8.9rel.1.tar.bz2
=> ‘/home/peko/source/buildroot/output-qt/build/.lynx2.8.9rel.1.tar.bz2.XlcdCK/output’
Resolving ftp.invisible-island.net (ftp.invisible-island.net)... 216.194.253.29
Connecting to ftp.invisible-island.net (ftp.invisible-island.net)|216.194.253.29|:21... connected.
Logging in as anonymous ...
Login incorrect.
Luckily the tarball is also available over https://, so use that instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr:
- use https, not http
- drop trailing '/' in _SITE
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit d946d31325)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ftp.astrom.com also serves the files over https://, so use that instead.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr: use https, not http]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 02f6617b68)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
For consistency, adapt the reference in the hash file too.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[yann.morin.1998@free.fr: switch to https as suggested by Baruch]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 1efc1e7b90)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since qt5base was last bumped in 8ab4a0a348 (package/qt5: bump packages
to latest kde submodule versions), the hash for the downloaded tarball
has changed:
$ make qt5base-source
[...]
ERROR: expected: 935d01f5c34903ad9e979431cec7a8a59332ed3fc539e639f5ba87e8d6989b9d
ERROR: got : 3067c4d84ba9927bfe65bf606c17af082199e0a3b22781fbf9bc6c6bc3de26dd
We know the hash was good back when 8ab4a0a348 was applied, because
the tarball has been cached on sources.buildroot.org with the expected
hash:
$ curl 'https://sources.buildroot.net/qt5base/qtbase-da6e958319e95fe564d3b30c931492dd666bfaff.tar.bz2' 2>/dev/null |sha256sum -
935d01f5c34903ad9e979431cec7a8a59332ed3fc539e639f5ba87e8d6989b9d -
But now, the archive generated by the KDE gorge (Gitlab underneath) has
another hash (as seen above). This means that the KDE forge (Gitlab) has
changed the way it generates archives. So, what's the delta? It turns
out that the only changes are about CRLF that were present in the
original archive, and are no longer in the new one. It is to be noted
that the affected files do not have CRLFS in the repository. It further
turns out that the archive was previously generated with .gitattibutes
of the main branch ('dev' in Qt repositories), while now they are
generated with the .gitattibutes of the commit for which they are
generated.
Switch to using the git download method for really reproducible
archives...
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Sebastian Weyer <sebastian.weyer@smile.fr>
Cc: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 75da04c817)
[Peter: adjust for filename/hash used on 2024.02.x]
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently, when we generate archives, e.g. for git, svn, cargo or go, we
use the package _BASENAME_RAW as the root directory of the generated
archive. For example, for package foo at version 1.2.3, that would generate
an archive rooted at foo-1.2.3/.
This is usually what we want, except in one specific condition: when the
package shares its download with another package *and* it is a generated
archive. In that case, the root directory will be different for each of
the two packages, which is incorrect, but was so far benign: we never
had any hash for such generated archives, and they were only generated
in two cases:
- linux and linux-headers
- barebox and barebox-aux
As we skip one directory depth when extracting the archives, we did not
care what the root directory was; whether it was that of one package or
the other was of no consequence.
But now that we can have hashes for archives generated from custom
versions, this breaks the usual case where the headers used for the
toolchains are those of the kernel to build for the target. In this
case, we may end up downloading the linux-headers package before we
download the linux package, so we'd get the hash for an archive rooted
at linux-headers-XXX/, but the one for the linux package the archive
would be rooted at linux-XXX/, or we may end up (e.g. with parallel
builds) downloading the linux package first and linux-headers next.
That would cause conflicts in hashes, as demonstrated by the only defconfig
we have in that situation, olimex_stmp157_olinuxino_lime_defconfig.
_BASENAME_RAW is a construct that is expanded to include the RAWNAME
followed by a dash and the version, if there is a version, or with just
the RAWNAME when there is no version.
We tweak the download macro to use _DL_SUBDIR followed by the version.
This is only used by VCS backends (cvs, git, svn...) and so there will
always be a version string, so no need to duplicate the case without a
version like is done for _BASENAME_RAW
_DL_SUBDIR defaults to _RAWNAME, so this is a noop by default, unless
the package declares it shares its download with another one, in which
case the generated archive will now be rooted as for the shared package.
This was triggered by:
https://patchwork.ozlabs.org/project/buildroot/patch/20240602070634.597337-1-francois.perrad@gadz.org/
Reported-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit ebe238f2b5)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The gcc 13.x. version was bumped to 13.3.0 by commit 5d9c54de0c
(package/gcc: update to 13.3.0), so update gcc-bare-metal to match.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Restarting dnsmasq can cause constant issues: stop works, but start
fails because the new instance can't bind the socket. Another restart
immediately after works just fine:
# /etc/init.d/S80dnsmasq restart
Stopping dnsmasq: OK
Starting dnsmasq:
dnsmasq: failed to create listening socket for 192.168.128.1: Address in use
FAIL
# /etc/init.d/S80dnsmasq restart
Stopping dnsmasq: FAIL
Starting dnsmasq: OK
Solve this by waiting for process to actually stop before returning
from the stop command. Clean up the PID file after to avoid potential
issues with the PID being reused after stop. The wait could also be
placed inside the restart block, but putting it into the stop block
has the advantage that it also avoids similar issues for any other
callers.
Signed-off-by: Fiona Klute <fiona.klute+wiwa@gmx.de>
[yann.morin.1998@free.fr:
- fix shellcheck
- reflow commit log
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 507caef3dd)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This requires python 3.5 or newer but is a bit cleaner than the
previous coroutine method.
This should also fix a python3.12 issue:
[Tue, 28 May 2024 13:09:05] INFO: generate the configuration
Traceback (most recent call last):
File "/home/autobuild/autobuild/instance-0/buildroot/utils/genrandconfig", line 833, in <module>
ret = asyncio.run(gen_config(args))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.12/asyncio/runners.py", line 194, in run
return runner.run(main)
^^^^^^^^^^^^^^^^
File "/usr/lib/python3.12/asyncio/runners.py", line 89, in run
raise ValueError("a coroutine was expected, got {!r}".format(coro))
ValueError: a coroutine was expected, got <generator object gen_config at 0xffff7bd822c0>
[Tue, 28 May 2024 13:09:06] WARN: failed to generate configuration
Signed-off-by: James Hilliard <james.hilliard1@gmail.com>
Reviewed-by: Vincent Fazio <vfazio@xes-inc.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 584ebdea6e)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Most boards use BR2_KERNEL_HEADERS_AS_KERNEL with their custom kernels.
So when creating their custom hash files, the linux-headers.hash is the
same as linux.hash. In this case we symlink linux-headers to linux to
make maintenance easier. Update the add-custom-hashes tool to explicitly
handle this case.
Signed-off-by: Brandon Maier <brandon.maier@collins.com>
[Peter: use cmp -s]
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit d506e232e7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Ccache CMake build system has some conditions that automatically
enables a developer mode (sets CCACHE_DEV_MODE=ON). See [1].
More specifically, if CCACHE_DEV_MODE is unset AND the environment
variable "CI" is set, CCACHE_DEV_MODE is set to "ON".
This situation can happen when Buildroot builds are executed in
Jenkins jobs, for example. Since Buildroot does not set
CCACHE_DEV_MODE and Jenkins sets the "CI" environment variable,
this ccache developer mode can be enabled in an unexpected way
for the Buildroot user. For example, it happened that a Jenkins build
breaks, while the build with the same configuration in the user
session is working.
One of the effects of enabling this ccache developer mode, is to treat
compiler warnings as errors, see [3]. This can lead to build error,
depending on the ccache version and the host compiler being used.
This behavior can be reproduced and observed, with commands:
cat > .config <<EOF
BR2_aarch64=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_CCACHE=y
BR2_PACKAGE_BUSYBOX=y
EOF
make olddefconfig
make clean
make host-ccache
Outputs:
...
-- Ccache version: 4.9.1
-- Ccache dev mode: OFF
-- Setting CMAKE_BUILD_TYPE to Release as none was specified.
...
Whereas:
make clean
CI=true make host-ccache
Outputs:
...
-- Ccache version: 4.9.1
-- Ccache dev mode: ON
-- Setting CMAKE_BUILD_TYPE to Debug as none was specified.
...
For a failure example: on Fedora 40 with host gcc 14.1.1, Buildroot at
tag 2024.02 has ccache 4.8.2. Host ccache can fail, when building with
"CI=true make host-ccache" with output:
/buildroot/output/build/host-ccache-4.8.2/src/third_party/fmt/core.h:3119:44: in 'constexpr' expansion of 'fmt::v8::make_format_args<>(args#0, args#1)'
/buildroot/output/build/host-ccache-4.8.2/src/third_party/fmt/core.h:1706:15: error: possibly dangling reference to a temporary [-Werror=dangling-reference]
1706 | const auto& arg = arg_mapper<Context>().map(std::forward<T>(val));
| ^~~
This commit sets CCACHE_DEV_MODE=OFF to make the ccache behavior more
deterministic in Buildroot, independently of being used in a CI tool
or not.
[1] https://github.com/ccache/ccache/blob/v4.9.1/CMakeLists.txt#L56
[2] https://github.com/jenkinsci/jenkins/blob/jenkins-2.459/core/src/main/java/jenkins/model/CoreEnvironmentContributor.java#L43
[3] https://github.com/ccache/ccache/blob/v4.9.1/cmake/StandardWarnings.cmake#L5
Reported-by: Xavier Roumegue <xroumegue@gmail.com>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 197be7ed87)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This new stable series brings in support for pointer lock, customization
of subprocess launching, and a build fix that allows using EGL
implementations that do not ship a pkg-config module. Version 1.16.x is
recommended for WPE WebKit 2.44.x, which will be updated in a follow-up
patch.
Release notes:
https://wpewebkit.org/release/libwpe-1.16.0.html
Signed-off-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 8845a796bb)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Run-time dependency xcb-cursor found: NO (tried pkgconfig and cmake)
../../br-test-pkg/bootlin-armv7-glibc/build/weston-13.0.0/tests/meson.build:340:2: ERROR: Problem encountered: xcb and xcb-cursor required for running xwayland tests
Fixes:
- https://bugs.busybox.net/show_bug.cgi?id=15766
Signed-off-by: Thomas Devoogdt <thomas@devoogdt.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit fb952e1116)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit 2e349be94a (configs/orangepi_pc_defconfig: bump U-boot to
v2024.01 to fix compatibility with setuptools >= 69) bumped U-Boot, but
forgot to add a dependency on host-openssl. Fix that.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 96e41a849e)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fix the following build failure with kernel >= 6.7:
/home/buildroot/autobuild/instance-0/output-1/build/cryptodev-linux-1.13/./cryptlib.c: In function ‘cryptodev_hash_init’:
/home/buildroot/autobuild/instance-0/output-1/build/cryptodev-linux-1.13/./cryptlib.c:384:28: error: implicit declaration of function ‘crypto_ahash_alignmask’; did you mean ‘crypto_aead_alignmask’? [-Werror=implicit-function-declaration]
384 | hdata->alignmask = crypto_ahash_alignmask(hdata->async.s);
| ^~~~~~~~~~~~~~~~~~~~~~
| crypto_aead_alignmask
Fixes:
- http://autobuild.buildroot.org/results/466360c7baec2edf42dc6f0ad9a8d757dd471c88
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 15ec07403f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fix the following or1k build failure raised since bump to version 0.2.2
in commit 079d992b27:
In file included from /home/buildroot/autobuild/run/instance-0/output-1/host/or1k-buildroot-linux-uclibc/sysroot/usr/include/lua.hpp:6,
from ../src/lua.hh:23,
from ../src/parser.hh:4,
from ../src/ffilib.cc:7:
../src/ffilib.cc: In static member function 'static void ffi_module::setup(lua_State*)':
../src/ffilib.cc:1616:28: error: expected ')' before 'FFI_ARCH_NAME'
1616 | lua_pushliteral(L, FFI_ARCH_NAME);
| ^~~~~~~~~~~~~
Fixes: 079d992b27
- http://autobuild.buildroot.org/results/4e14753732c5b6fe8ba9ecc4050ffb35f471c428
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 9b07b4bf6d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes: https://gitlab.com/buildroot.org/buildroot/-/jobs/6865597706
setuptools.extern.packaging.version.InvalidVersion: Invalid version: 'u-boot-2020.10'
make[5]: *** [scripts/dtc/pylibfdt/Makefile:30: scripts/dtc/pylibfdt/_libfdt.so] Error 1
make[4]: *** [scripts/Makefile.build:419: scripts/dtc/pylibfdt] Error 2
make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:419: scripts/dtc] Error 2
make[2]: *** [Makefile:577: scripts] Error 2
make[1]: *** [package/pkg-generic.mk:283: /builds/buildroot.org/buildroot/output/build/uboot-2020.10/.stamp_built] Error 2
Commit 6b62384e3d (package/python-setuptools: bump to version 69.0.3)
bumped setuptools, which now errors out when passed a version with a prefix
string. This u-boot- prefix string is dropped since U-Boot 2023.01 with
commit c977b1843504 (libfdt: Fix invalid version warning), so bump U-Boot to
v2024.01 to fix this issue.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 2e349be94a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Mutt 2.2.13 was released on March 9, 2024. This is a bug-fix release,
fixing a possible dangling pointer reference in the SMTP client.
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 730de6c00e)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes: https://gitlab.com/buildroot.org/buildroot/-/jobs/6865597717
setuptools.extern.packaging.version.InvalidVersion: Invalid version: 'u-boot-2020.10'
make[5]: *** [scripts/dtc/pylibfdt/Makefile:30: scripts/dtc/pylibfdt/_libfdt.so] Error 1
make[4]: *** [scripts/Makefile.build:419: scripts/dtc/pylibfdt] Error 2
make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:419: scripts/dtc] Error 2
make[2]: *** [Makefile:577: scripts] Error 2
make[1]: *** [package/pkg-generic.mk:283: /builds/buildroot.org/buildroot/output/build/uboot-2020.10/.stamp_built] Error 2
Commit 6b62384e3d (package/python-setuptools: bump to version 69.0.3)
bumped setuptools, which now errors out when passed a version with a prefix
string. This u-boot- prefix string is dropped since U-Boot 2023.01 with
commit c977b1843504 (libfdt: Fix invalid version warning), so bump U-Boot to
v2024.01 to fix this issue.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Romain Naour <romain.naour@smile.fr>
(cherry picked from commit 113eeb55d6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following warning:
orangepi_r1_defconfig:35:warning: override: reassigning to symbol BR2_PACKAGE_HOST_UBOOT_TOOLS
While we're at it, move the HOST_UBOOT_TOOLS_* options where savedefconfig
would put them.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Romain Naour <romain.naour@smile.fr>
(cherry picked from commit 0ba568276a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The python-magic-wormhole runtime test can randomly fail on slow
runners, see [1].
The issue is that the sending command is started first in background
_without_ redirecting its output to /dev/null. The receiving command
is started after, expecting the message to be printed on the first
standard output line. On slower systems, the sending command still
print messages while the test controller expect output from the
receiving command. The expected string finally appear, but not on the
first line. This makes the test fail.
This commit fixes the issue by redirecting all outputs (stdout, stderr)
of the sending command to /dev/null. To help even more, the sleep time
is moved from the emulator to the test controller. The sleep time is
also multiplied by the timeout_multiplier.
Fixes: [1]
[1] https://gitlab.com/buildroot.org/buildroot/-/jobs/6888691508
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Romain Naour <romain.naour@smile.fr>
(cherry picked from commit 9a734700d7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ivi-homescreen needs C++17 filesystem resulting in the following build
failure with gcc < 8 since its addition in commit
9d8497e79d:
In file included from /home/buildroot/instance-0/output-1/build/ivi-homescreen-5ab78a19e95c88cc5d6b173ab1260a211e78cf0a/third_party/flutter/shell/platform/common/path_utils.cc:5:0:
/home/buildroot/instance-0/output-1/build/ivi-homescreen-5ab78a19e95c88cc5d6b173ab1260a211e78cf0a/third_party/./flutter/shell/platform/common/path_utils.h:8:10: fatal error: filesystem: No such file or directory
#include <filesystem>
^~~~~~~~~~~~
Fixes: 9d8497e79d
- http://autobuild.buildroot.org/results/a5e88d5a4264a6165be31a99f1c123af53fa382e
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 898cf04e47)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fix the following python build failure:
In file included from /home/buildroot/instance-0/output-1/host/include/python3.11/Python.h:38,
from /home/buildroot/instance-0/output-1/host/arc-buildroot-linux-gnu/sysroot/usr/include/pybind11/detail/common.h:266,
from /home/buildroot/instance-0/output-1/host/arc-buildroot-linux-gnu/sysroot/usr/include/pybind11/attr.h:13,
from /home/buildroot/instance-0/output-1/host/arc-buildroot-linux-gnu/sysroot/usr/include/pybind11/detail/class.h:12,
from /home/buildroot/instance-0/output-1/host/arc-buildroot-linux-gnu/sysroot/usr/include/pybind11/pybind11.h:13,
from /home/buildroot/instance-0/output-1/build/gr-osmosdr-0.2.4/python/bindings/device_python.cc:1:
/home/buildroot/instance-0/output-1/host/include/python3.11/pyport.h:596:2: error: #error "LONG_BIT definition appears wrong for platform (bad gcc/glibc config?)."
596 | #error "LONG_BIT definition appears wrong for platform (bad gcc/glibc config?)."
| ^~~~~
Fixes:
- http://autobuild.buildroot.org/results/f009958c37902a224512b336fcb431903bdd0b96
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 8759d81b00)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The busybox sysctl applet expects all arguments after -p to be filenames to
read:
sysctl -p file -h
sysctl: -h: No such file or directory
VS:
sysctl -h -p file
sysctl: invalid option -- 'h'
BusyBox v1.36.1 (2024-05-17 15:27:21 CEST) multi-call binary.
Usage: sysctl [-enq] { -a | -p [FILE]... | [-w] [KEY[=VALUE]]... }
Show/set kernel parameters
-e Don't warn about unknown keys
-n Don't show key names
-q Quiet
-a Show all values
-p Set values from FILEs (default /etc/sysctl.conf)
-w Set values
This seems to be the intented behaviour:
https://git.busybox.net/busybox/tree/procps/sysctl.c#n317
Notice: The procps-ng variant is happy with both:
sysctl -p file -h
Usage:
sysctl [options] [variable[=value] ...]
VS:
sysctl -h -p file
Usage:
sysctl [options] [variable[=value] ...]
So pass SYSCTL_ARGS before the -p args so custom sysctl arguments can be
passed.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 60b02eb6c6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit ee59023794 (package/libvncserver: bump to version 0.9.12)
changes the package from autotools to cmake and accidently inverted the
threads logic.
The reason this was not noticed is that the build system verifies if threads
support works and otherwise disables it even when -DWITH_THREADS=ON is
passed, E.G. for a nothread configuration:
cmake ... -DWITH_THREADS=ON ...
..
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - not found
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 7b3c0b19d4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
webkitgtk/wpewebkit needs a host-ruby with yaml support, otherwise the build
fails with errors like:
cd /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore && \
/home/peko/source/buildroot/output/host/bin/ruby \
/home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/GenerateSettings.rb \
--outputDir /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/WebCore/DerivedSources \
--template /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/SettingsTemplates/InternalSettingsGenerated.cpp.erb \
--template /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/SettingsTemplates/InternalSettingsGenerated.idl.erb \
--template /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/SettingsTemplates/InternalSettingsGenerated.h.erb \
--template /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/SettingsTemplates/Settings.cpp.erb \
--template /home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/Scripts/SettingsTemplates/Settings.h.erb \
/home/peko/source/buildroot/output/build/webkitgtk-2.42.2/WTF/Scripts/Preferences/UnifiedWebPreferences.yaml \
/home/peko/source/buildroot/output/build/webkitgtk-2.42.2/Source/WebCore/page/Settings.yaml
/home/peko/source/buildroot/output/host/lib/ruby/3.3.0/yaml.rb:3: warning: It seems your ruby installation is missing psych (for YAML output).
To eliminate this warning, please install libyaml and reinstall your ruby.
We do not currently have any Config.in.host options for ruby or libyaml, but
given how small/fast libyaml is to build just do it unconditionally.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 95093854e4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This is bugfix release of stable libmdbx branch.
The most significant fixes since v0.12.8 (previous version pinned to
buildroot):
- Fixed an major bug inherited from LMDB (Lightning Memory-Mappe
Database) that causes database corruption during use the
MDBX_DUPFIXED mode, and that has existed for more than 10 years.
- Fixed of a false error MDBX_CORRUPTED (-30796) in the scenario of
working in the mode MDBX_DUPFIXED with odd length of multi-values.
- Fixed a bug in adjusting the cursors in the case splitting a page by
adding a new page on the left.
- Troubleshooting an error when opening a database on a read-only file
system.
- A set of C++ API improvements.
The complete ChangeLog:
https://gitflic.ru/project/erthink/libmdbx/blob?file=ChangeLog.md
Signed-off-by: Леонид Юрьев (Leonid Yuriev) <leo@yuriev.ru>
[yann.morin.1998@free.fr: ammend and reflow the commit log]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit aea54eab47)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
FSID daemon and its systemd unit file both depend on
BR2_PACKAGE_NFS_UTILS_RPC_NFSD but they're now always installed. Remove them
both if BR2_PACKAGE_NFS_UTILS_RPC_NFSD is disabled.
Reviewed-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Jan Čermák <sairon@sairon.cz>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit dc3464c4b6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Osmosdr failed to build docs when python support is enabled.
[ 41%] Copying osmosdr docstring templates as pybind headers ...
Traceback (most recent call last):
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/update_pydoc.py", line 22, in <module>
from doxyxml import DoxyIndex, DoxyClass, DoxyFriend, DoxyFunction, DoxyFile
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/doxyxml/__init__.py", line 69, in <module>
from .doxyindex import DoxyIndex, DoxyFunction, DoxyParam, DoxyClass, DoxyFile, DoxyNamespace, DoxyGroup, DoxyFriend, DoxyOther
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/doxyxml/doxyindex.py", line 31, in <module>
from .generated import index
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/doxyxml/generated/index.py", line 13, in <module>
from . import compound
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/doxyxml/generated/compound.py", line 15, in <module>
from . import compoundsuper as supermod
File "/home/martb/Schreibtisch/DiscoSAT/satos/output/build/gr-osmosdr-0.2.4/docs/doxygen/doxyxml/generated/compoundsuper.py", line 15, in <module>
import six
ModuleNotFoundError: No module named 'six'
make[2]: *** [python/bindings/CMakeFiles/osmosdr_docstrings.dir/build.make:73: python/bindings/docstring_status] Fehler 1
make[1]: *** [CMakeFiles/Makefile2:332: python/bindings/CMakeFiles/osmosdr_docstrings.dir/all] Fehler 2
Fixes:
http://autobuild.buildroot.net/results/b60e339e52fbc7ed7607a94381aaab6ec64b7a99/
Signed-off-by: Martin Böh <contact@martb.dev>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit cfda1f0b87)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add -fPIC to CFLAGS to fix the following build failure raised since
commit de6415ad9c:
/home/autobuild/autobuild/instance-5/output-1/host/lib/gcc/x86_64-buildroot-linux-gnu/12.3.0/../../../../x86_64-buildroot-linux-gnu/bin/ld: misc/misc.o: warning: relocation against `stdout@@GLIBC_2.2.5' in read-only section `.text'
/home/autobuild/autobuild/instance-5/output-1/host/lib/gcc/x86_64-buildroot-linux-gnu/12.3.0/../../../../x86_64-buildroot-linux-gnu/bin/ld: log/log.o: relocation R_X86_64_PC32 against symbol `stdout@@GLIBC_2.2.5' can not be used when making a shared object; recompile with -fPIC
Fixes: de6415ad9c
- http://autobuild.buildroot.org/results/0349c6bfd66f5e50429e4a5cc07fb7abf2b07345
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 880066bad6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The patch fixes the following linking failure:
/home/buildroot/instance-0/output-1/host/opt/ext-toolchain/bin/../lib/gcc/i586-buildroot-linux-musl/9.3.0/../../../../i586-buildroot-linux-musl/bin/ld: /home/buildroot/instance-0/output-1/host/i586-buildroot-linux-musl/sysroot/lib/libc.a(getopt.o): in function `getopt':
getopt.c:(.text.getopt+0x0): multiple definition of `getopt'; src/getopt.o:getopt.c:(.text+0x0): first defined here
/home/buildroot/instance-0/output-1/host/opt/ext-toolchain/bin/../lib/gcc/i586-buildroot-linux-musl/9.3.0/../../../../i586-buildroot-linux-musl/bin/ld: /home/buildroot/instance-0/output-1/host/i586-buildroot-linux-musl/sysroot/lib/libc.a(getopt.o):(.data.optind+0x0): multiple definition of `optind'; src/getopt.o:(.data+0x0): first defined here
Fixes:
- http://autobuild.buildroot.net/results/d5b1b4e5e9d9c8eca5e75c345db4d1f3f0cd84ed
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 5a7c40bd1c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Our index.html page still points to git.buildroot.net as the Git
repository, and to the defunct gmane for the mailing list
activity. Fix these by pointing to Gitlab and lore respectively.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 3e3bcd6338)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Builds with GCC 14 print the following error
> zip.h:726:10: error: conflicting types for 'memset'; have 'char *(char *, int, unsigned int)'
This is because with GCC 14, Zip incorrectly detects that the memset functions
exist. Which enables the ZMEM flag and declares its own version of memset.
This is because the ./unix/configure script attempts to compile a C file using
'memset' but it does not include the <string.h>. This was allowed in gnu89, but
in GCC 14 -Werror=implicit-function-declaration is enabled by default[1].
We forcefully set '-std=gnu89' so that Zip will compile everything against
gnu89, which suppresses the warning.
[1] https://gcc.gnu.org/gcc-14/porting_to.html#warnings-as-errors
Signed-off-by: Brandon Maier <brandon.maier@collins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 29c6fe13d3)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
opencv and qt handling can be dropped since switch to an active fork in
commit 456a739831 as they are only used
when examples are enabled resulting in the following warning:
CMake Warning:
Manually-specified variables were not used by the project:
BUILD_DOC
BUILD_DOCS
BUILD_EXAMPLE
BUILD_TEST
BUILD_TESTING
BUILD_TESTS
CMAKE_DISABLE_FIND_PACKAGE_OpenCV
CMAKE_DISABLE_FIND_PACKAGE_Qt5
While at it, also drop BUILD_EXAMPLES which is already passed by
pkg-cmake.mk
Fixes: 456a739831
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit fd64c544d3)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following security issues:
- CVE-2023-36054: lib/kadm5/kadm_rpc_xdr.c in MIT Kerberos 5 (aka krb5)
before 1.20.2 and 1.21.x before 1.21.1 frees an uninitialized pointer
- CVE-2023-39975: kdc/do_tgs_req.c in MIT Kerberos 5 (aka krb5) 1.21 before
1.21.2 has a double free that is reachable if an authenticated user can
trigger an authorization-data handling failure
Signed-off-by: André Zwing <nerv@dawncrow.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit dd1ea28b07)
[Peter: mark as security bump, add CVE details]
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Disable -Werror to fix the following build failure raised at least since
bump to version 3.9.0 in commit e76d9868c9
and
https://github.com/zhaojh329/libuhttpd/commit/81db1fa371bc80f40c1bfe60010e7d64bdf30b46:
/home/autobuild/autobuild/instance-8/output-1/build/libuhttpd-3.14.1/src/file.c: In function '__serve_file':
/home/autobuild/autobuild/instance-8/output-1/build/libuhttpd-3.14.1/src/file.c:322:50: error: format '%llu' expects argument of type 'long long unsigned int', but argument 4 has type '__off_t' {aka 'long int'} [-Werror=format=]
322 | conn->send_header(conn, "Content-Range", "bytes */%" PRIu64, st->st_size);
| ^~~~~~~~~~~ ~~~~~~~~~~~
| |
| __off_t {aka long int}
In file included from /home/autobuild/autobuild/instance-8/output-1/build/libuhttpd-3.14.1/src/file.c:36:
/home/autobuild/autobuild/instance-8/output-1/per-package/libuhttpd/host/mips64el-buildroot-linux-uclibc/sysroot/usr/include/inttypes.h:91:41: note: format string is defined here
91 | # define PRIu64 __PRI64_PREFIX "u"
cc1: all warnings being treated as errors
Fixes: e76d9868c9
- http://autobuild.buildroot.org/results/46349a1409f9cff027b462456078284459e75511
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 5961b21918)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The rock5b_defconfig contained a
# BR2_LINUX_KERNEL_INTREE_DTS_NAME is not set
line. This caused check-dotconfig.py to throw a warning in the
buildroot CI, because the explicit unsetting of the parameter is not
taken on in the actual dotconfig, but instead the dotconfig will
contain the line 'BR2_LINUX_KERNEL_INTREE_DTS_NAME=""'.
This patch removes the parameter from the rock5b_defconfig. The
resulting dotconfig from the original rock5b_defconfig and the
rock5b_defconfig without the parameter is identical, but the
check-dotconfig.py does not throw a warning anymore.
Signed-off-by: Kilian Zinnecker <kilian.zinnecker@mail.de>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 0f0660ad2d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Release notes: https://www.postgresql.org/docs/release/16.2/
The hash of the license file is updated due to a change in copyright
years:
-Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group
+Portions Copyright (c) 1996-2024, PostgreSQL Global Development Group
Signed-off-by: Maxim Kochetkov <fido_max@inbox.ru>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 360d2da3fa)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Buildroot commit f95069814b disabled the
build of test programs unconditionally without considering the option
BR2_PACKAGE_LIBDRM_INSTALL_TESTS.
Signed-off-by: Bernd Kuhls <bernd@kuhls.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 91848e73ae)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since version 2.1, LuaJIT follows a rolling-release scheme, which means
that any commit is as good as any other; LuaJIT uses the comitter's UNIX
timestamp as its semver patch level. It uses the git-attribute
export-subst for the .relver file that contains the %ct placeholder for
git-archive to expand it.
In c9dcd9e459 (package/luajit: bump to version 41fb94defa8f...), we
switched to such an upstream version. There was some confusion around
the handling of the git-attribute and where/when it is generated, and
the first revision of the patch used the git download method, so had to
use post-extract hooks to do the replacement, but the second iteration
kept retrieving the archive generated by github, which has the
replacement already done, but the post-extract hooks were not dropped
although now useless...
With the current code, it is easy to bump the LuaJit version and forget
to update the timestamp stored in the .relver file, which would override
the value that was generated on the github side.
Since the post-extract hook is useless, drop it.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Francois Perrad <fperrad@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
(cherry picked from commit 1bcb51517c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The upstream host, arago-project.org, has vanished, bringing down the
git repository with it.
Switch to another, github-hosted repository, that has the commit we're
interested in.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
(cherry picked from commit b1977d933b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The current version of docker-compose is un-vendorable, because the
dependencies it referenmces (directly or indirectly) are not available:
go: github.com/docker/compose/v2/cmd/compose imports
github.com/moby/buildkit/util/progress/progressui:
github.com/crazy-max/buildkit@v0.7.1-0.20240130133234-d9aa289bd124:
invalid version: unknown revision d9aa289bd124
And indeed, that commit does not exist in that repository. The v0.7.1
tag does exist, but there is not commit that matches the short hash
d9aa289bd124, or even the whole version string. Sigh...
There is no way anyone can vendor the version we currently package, and
all they and us can hope for is that we never lose s.b.o ever.
Bump the version. That one can be vendored. Well, at least it can
_still_ be vendored _now_...
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
(cherry picked from commit 1b189f5491)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
While bumping kodi, we figured out that the kodi-texturepacker and
kodi-jsonschemabuilder were both re-downloading the main Kodi tarball,
even though they contain:
KODI_JSONSCHEMABUILDER_DL_SUBDIR = kodi
KODI_TEXTUREPACKER_DL_SUBDIR = kodi
Both are host packages, and turns out that changing those variables to
HOST_ ones made the download sharing work.
Commit efa7712b09 ("package/pkg-generic:
host variant inherits target download settings") introduced
inheritance of host variables from target variables from a number of
variables, including DL_SUBDIR. But it missed the fact that earlier in
pkg-generic.mk, the following line was defined:
$(2)_DL_SUBDIR ?= $$($(2)_RAWNAME)
So, when this later code kicked in:
ifndef $(2)_DL_SUBDIR
ifdef $(3)_DL_SUBDIR
$(2)_DL_SUBDIR = $$($(3)_DL_SUBDIR)
endif
endif
In fact it never did anything because $(2)_DL_SUBDIR would never be
undefined. This commit fixes this issue by properly adjusting the
logic to inherit the value of the target variable when it exists, or
defaulting to $$($(2)_RAWNAME) otherwise.
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit a1c4d6c884)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The Config.in comment was mentioning both "NPTL" and "threads" as
dependencies, while mentioning only the former is sufficient.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit dabd983c0d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When ImageMagick is selected, the "zbarimg" program is compiled and
installed on target. It allows to decode a QR code from an image file.
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 84ad5c22d0)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As part of the review process of the ivi-homescreen package, which
landed in commit 9d8497e79d, it was
deemed that the dependency on BR2_TOOLCHBAIN_HAS_THREAD_NPTL was not
needed, as it was implied by glibc. According to the commit log:
[yann.morin.1998@free.fr:
- propagate BR2_PACKAGE_HOST_FLUTTER_SDK_BIN_ARCH_SUPPORTS to comments
- drop NPTL, implied by glibc
- reorder dependencies in a more logical way
- reorder comments
- drop undefined BR2_PACKAGE_IVI_HOMESCREEN_HAS_CLIENT
- grammar ("for to change")
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
However, while the BR2_TOOLCHBAIN_HAS_THREAD_NPTL dependency was
removed from the BR2_PACKAGE_IVI_HOMESCREEN option definition, the
corresponding dependency in the Config.in comment was not
dropped. Let's bring things back in sync.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit b9b072f83c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The following defconfig:
BR2_aarch64=y
BR2_cortex_a72=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_INIT_NONE=y
BR2_SYSTEM_BIN_SH_NONE=y
# BR2_PACKAGE_BUSYBOX is not set
BR2_PACKAGE_GSTREAMER1=y
BR2_PACKAGE_GST1_PLUGINS_BASE_LIB_OPENGL=y
BR2_PACKAGE_GST1_PLUGINS_GOOD=y
# BR2_PACKAGE_GST1_PLUGINS_GOOD_PLUGIN_AVI is not set
# BR2_PACKAGE_GST1_PLUGINS_GOOD_PLUGIN_ISOMP4 is not set
# BR2_PACKAGE_GST1_PLUGINS_GOOD_PLUGIN_WAVPARSE is not set
BR2_PACKAGE_GST1_PLUGINS_GOOD_PLUGIN_QMLGL=y
BR2_PACKAGE_MESA3D=y
BR2_PACKAGE_MESA3D_GALLIUM_DRIVER_SWRAST=y
BR2_PACKAGE_MESA3D_OPENGL_EGL=y
BR2_PACKAGE_MESA3D_OPENGL_ES=y
BR2_PACKAGE_QT5=y
BR2_PACKAGE_WAYLAND=y
# BR2_TARGET_ROOTFS_TAR is not set
fails to build in gst1-plugins-good, with:
build/gst1-plugins-good-1.22.9/ext/qt/meson.build:48:4: ERROR: Problem encountered: qt5 qmlglsink plugin is enabled, but qt specific tools were not found
This is due to qt5tools being missing. This commit adds this missing
dependency.
Signed-off-by: Charles Hardin <ckhardin@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit c107dd33e6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The lvm2 package now enables the required Kernel configuration. The
Kernel config fragment included in this test is no longer needed.
This commit removes it.
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 9339343111)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
LVM2 has a hard dependency on the device-mapper presence in the Kernel.
This commit enables those mandatory Kernel configuration by defining
the _LINUX_CONFIG_FIXUPS macro. This will make sure the final system
image will end up in a working configuration.
This was suggested by Arnout in [1].
[1] https://lists.buildroot.org/pipermail/buildroot/2024-April/688776.html
Suggested-by: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 419d39b261)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When building on hosts with libstdc++ without demangle support this error
shows up:
checking for __cxa_demangle in -lstdc++... no
configure: error: __cxa_demangle not found in libstdc++, use --disable-demangler to disable demangler support.
make[1]: *** [package/pkg-generic.mk:273: /home/giuliobenetti/br_reproduce/c2524c7580d97f7387ec22da62be71d77f2ed8ec/output/build/host-elfutils-0.189/.stamp_configured] Error 1
make: *** [Makefile:23: _all] Error 2
So let's disable demangler for host by default.
Fixes: still not showed by autobuilders. Reproduced on Ubuntu 22.04 with
Ubuntu APT g++ 11.4.0
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Reviewed-by: Christian Stewart <christian@aperture.us>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 5f2c9f882d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
pppd fails to start on a systems with buildroot 2024.02.x because of
missing pppd directory in /var/run. There are some logs hinting at this
issue:
Warning: couldn't open ppp database /var/run/pppd/pppd2.tdb
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
Can't create lock file /var/run/pppd/lock/LCK..ppp-tty-fifo1: No such file or directory
The issue has already been detected and fixed upstream (see [1]) and is
expected to be released on a v2.5.1, but this release seems to be stalled
for now (see [2]). Bump on current master, which currently reflects what
will likely be the 2.5.1.
[1] https://github.com/ppp-project/ppp/issues/419
[2] https://github.com/ppp-project/ppp/issues/460
Signed-off-by: Alexis Lothoré <alexis.lothore@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit c1b04a3254)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following security issue:
CVE-2024-24787: cmd/go: arbitrary code execution during build on darwin
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
openocd is not compatible with libgpiod2 resulting in the following
build failure since the addition of libgpiod2 in commit
57391fad2e:
src/jtag/drivers/linuxgpiod.c: In function 'helper_get_line':
src/jtag/drivers/linuxgpiod.c:283:19: error: 'GPIOD_LINE_REQUEST_DIRECTION_INPUT' undeclared (first use in this function); did you mean 'GPIOD_LINE_DIRECTION_INPUT'?
283 | int dir = GPIOD_LINE_REQUEST_DIRECTION_INPUT, flags = 0, val = 0, retval;
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| GPIOD_LINE_DIRECTION_INPUT
Fixes: 57391fad2e
- http://autobuild.buildroot.org/results/7195e8b76350f26ab3b963702d88b254512e6928
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit be3252fe01)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fix the following static build failure raised since bump to version
version 6.0.14 in commit ce17f93e82 and
https://github.com/OISF/suricata/commit/31ba4fd1522dd0d7f933767b9205c67a4e726fe5:
configure:22378: checking for pcap_open_dead in -lpcap
configure:22407: /home/buildroot/autobuild/run/instance-3/output-1/host/bin/x86_64-linux-gcc -o conftest -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O2 -g0 -static -std=c11 -march=native -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -static -rdynamic conftest.c -lpcap -ljansson -lpthread -lyaml -lpcre -latomic -lz >&5
/home/buildroot/autobuild/run/instance-3/output-1/host/opt/ext-toolchain/bin/../lib/gcc/x86_64-buildroot-linux-musl/12.3.0/../../../../x86_64-buildroot-linux-musl/bin/ld: /home/buildroot/autobuild/run/instance-3/output-1/host/x86_64-buildroot-linux-musl/sysroot/usr/lib/../lib64/libpcap.a(pcap-linux.o): in function `nl80211_init':
pcap-linux.c:(.text+0xc34): undefined reference to `nl_socket_alloc'
[...]
checking for pcap_open_dead in -lpcap... no
Fixes: ce17f93e82
- http://autobuild.buildroot.org/results/9e25c2508bd4100d2d1d3180e79060d762361213
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit a3ee47ef22)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The CPIO filesystem generated by TestPerlDBDmysql is too large, and
doesn't fit as an initramfs in the 256MB of RAM available in the
versatilepb machine. This causes a failure while running a basic
test "perl -MDBI -e '1'" since "/usr/lib/perl5", and many files
being missing from the root filesystem, ultimately causing the test
to fail.
Can't locate DBI.pm in @INC (you may need to install the DBI module)
(@INC entries checked: /usr/lib/perl5/site_perl/5.38.2/arm-linux
/usr/lib/perl5/site_perl/5.38.2 /usr/lib/perl5/5.38.2/arm-linux
/usr/lib/perl5/5.38.2).
It would make sense to switch all test cases to use ext2 + a
hard-drive, but for now, let's fix the few test cases that are causing
problems.
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/6735654506
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 8937cd065c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
PIE will be set by buildroot if the user wants it so set
{COMPILE,LINK}_PIE to an empty value to fix the following build failure
raised since the addition of the package in commit
02c818bc5c:
/home/buildroot/autobuild/run/instance-3/output-1/host/bin/m68k-linux-gcc -W -Wall -Waggregate-return -Wcast-align -Wconversion -Wdisabled-optimization -Wmissing-declarations -Wmissing-format-attribute -Wmissing-noreturn -Wmissing-prototypes -Wpointer-arith -Wredundant-decls -Wshadow -Wstrict-prototypes -Wwrite-strings -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -O2 -g0 -std=gnu99 -DLIBEXECDIR=\"/usr/lib\" -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Wl,-z,relro -Wl,-stats -fPIE -pie -Wl,-z,now utempter.c -o utempter
[...]
/home/buildroot/autobuild/run/instance-3/output-1/host/opt/ext-toolchain/bin/../lib/gcc/m68k-buildroot-linux-uclibc/12.3.0/../../../../m68k-buildroot-linux-uclibc/bin/ld: /home/buildroot/autobuild/run/instance-3/output-1/host/m68k-buildroot-linux-uclibc/sysroot/usr/lib/Scrt1.o: in function `lib_main':
(.text+0x4): undefined reference to `__shared_flat_add_library'
Fixes: 02c818bc5c
- http://autobuild.buildroot.org/results/3a5581fd4edf56bbdc48ab111a2351fc70f1c703
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 36b06928c8)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
And enable BR2_DOWNLOAD_FORCE_CHECK_HASHES.
Generated by utils/add-custom-hashes, with the (redundant)
linux-headers.hash replaced by a symlink.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 56ce35d235)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following error/warnings with kconfig from linux-6.9-rc5:
package/x11r7/Config.in:14: syntax error
package/x11r7/Config.in:14:warning: ignoring unsupported character '/'
package/x11r7/Config.in:14:warning: ignoring unsupported character '/'
package/x11r7/Config.in:14:warning: ignoring unsupported character '/'
package/x11r7/Config.in:14:warning: ignoring unsupported character '.'
package/x11r7/Config.in:14: invalid statement
[...]
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit a214c4568b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following error/warnings with kconfig from linux-6.9-rc5:
package/python-pydal/Config.in:2: syntax error
package/python-pydal/Config.in:2: invalid statement
package/python-pydal/Config.in:3: invalid statement
[...]
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 9f8d3f47d7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following error/warnings with kconfig from linux-6.9-rc5:
package/dovecot/Config.in:36: syntax error
package/dovecot/Config.in:36:warning: ignoring unsupported character '/'
package/dovecot/Config.in:36:warning: ignoring unsupported character '/'
package/dovecot/Config.in:36:warning: ignoring unsupported character '.'
package/dovecot/Config.in:36: invalid statement
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit ff70f9ee4c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following error/warnings with kconfig from linux-6.9-rc5:
package/cmocka/Config.in:2: syntax error
package/cmocka/Config.in:2: invalid statement
package/cmocka/Config.in:3: invalid statement
package/cmocka/Config.in:4: invalid statement
[...]
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 4abda655e6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following error/warnings with kconfig from linux-6.9-rc5:
boot/barebox/Config.in:79: syntax error
boot/barebox/Config.in:79:warning: ignoring unsupported character '/'
boot/barebox/Config.in:79:warning: ignoring unsupported character '/'
boot/barebox/Config.in:79:warning: ignoring unsupported character '/'
boot/barebox/Config.in:79:warning: ignoring unsupported character '.'
boot/barebox/Config.in:79: invalid statement
boot/barebox/Config.in:93: syntax error
boot/barebox/Config.in:93:warning: ignoring unsupported character '/'
boot/barebox/Config.in:93:warning: ignoring unsupported character '/'
boot/barebox/Config.in:93:warning: ignoring unsupported character '/'
boot/barebox/Config.in:93:warning: ignoring unsupported character '.'
boot/barebox/Config.in:93: invalid statement
Signed-off-by: Peter Seiderer <ps.report@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit e86aec1663)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
With the bump to version 8.1.1, the patch that disabled the compilation
of the fp-bench test is no longer applicable, even though the package
compilation process does not report any errors in applying the patch
itself. The new patch does not disable the test by default but only if
the file fenv.h is not missing, with the hope that this approach will be
considered acceptable by the maintainer and merged upstream.
The patch is an adaptation of the one sent upstream.
Link: https://lists.nongnu.org/archive/html/qemu-devel/2021-03/msg00492.html
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 2bb0c57ca2)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The http URL redirects to https. This commit updates this URL to
directly use https.
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 2aa3a23bbe)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Wireless regulatory database lists the allowed radio frequencies for
each local jurisdiction. Since linux-4.15 the kernel supports loading
the files regulatory.db/regulatory.db.p7s directly from the
/lib/firmware directory. Currently this package is not enabled and
kernel complains with the following message on every boot:
"""
platform regulatory.0: Direct firmware load for regulatory.db failed
with error -2
cfg80211: failed to load regulatory.db
"""
Add wireless regulatory database package to fix the issue.
Signed-off-by: Konstantin Aladyshev <aladyshev22@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 16e9f51490)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since the toolchain Bootlin update to 2023.11-1 [1], the arm Linux
kernel build is broken with binutils >= 2.41 with:
arch/arm/mm/proc-v7.S: Assembler messages:
arch/arm/mm/proc-v7.S:640: Error: junk at end of line, first unrecognized character is `#'
A similar issue has already be fixed for qemu m68k [2].
Bump to the latest kernel 4.19 that already include the backport
of 790756c7e022 ("ARM: 8933/1: replace Sun/Solaris style flag on section directive")
[1] 7e0e6e3b86
[2] a1ce9474e4
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/6703222383
Signed-off-by: Romain Naour <romain.naour@smile.fr>
(cherry picked from commit 7e126bd38d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
GCC14 now treats implicit int types as error so when check() from
check-lxdialog.sh is called to check whether we can link against ncurses
it will fail silently and the help text indicating to install ncurses is
printed.
However, this is not due to missing ncurses but once the stderr redirect
to /dev/null is removed we can see the root cause:
<stdin>:2:1: error: return type defaults to ‘int’ [-Wimplicit-int]
So, in order for menuconfig to work with GCC14 lets just specify the
return type of main() as int.
Npte that the upstream kconfig in the linux kernel source tree no longer
carries or uses the check-lxdialog.sh script since commit 1c5af5cf9308
(kconfig: refactor ncurses package checks for building mconf and nconf),
so there is no commit we can backport to our kconfig copy.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Reviewed-by: Petr Vorel <petr.vorel@gmail.com>
Tested-by: Petr Vorel <petr.vorel@gmail.com>
[yann.morin.1998@free.fr: add note about upstream kernel]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit a6210d28db)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes the following security issues:
CVE-2021-3575: A heap-based buffer overflow was found in openjpeg in
color.c:379:42 in sycc420_to_rgb when decompressing a crafted .j2k file. An
attacker could use this to execute arbitrary code with the permissions of
the application compiled against openjpeg.
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit ff36bc68cd)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
To ensure the correct prefix is used in the generated tinycbor.pc instead of
/usr/local:
>>> tinycbor 0.6.0 Building
..
sed > tinycbor.pc < tinycbor.pc.in \
-e 's,@prefix@,/usr/local,' \
-e 's,@exec_prefix@,/usr/local,' \
-e 's,@libdir@,/usr/local/lib,' \
-e 's,@includedir@,/usr/local/include,' \
-e 's,@version@,0.6.0,'
>>> tinycbor 0.6.0 Installing to staging directory
..
install -m 644 tinycbor.pc /path/to/buildroot/output/host/aarch64-buildroot-linux-gnu/sysroot/usr/lib/pkgconfig/tinycbor.pc
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit b059e08420)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In _imagingcms.c in Pillow before 10.3.0, a buffer overflow exists because
strcpy is used instead of strncpy.
Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit dfaa34ddd3)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Peter Korsgaard