Age | Commit message (Collapse) | Author |
|
Between the binconfig class not being as useful as possible, and this recipe
doing things in a suboptimal way, there was a lot of confusion happening.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
|
|
|
|
Instead of using part of ${base_libdir} to rename scripts, use MLPREFIX. This
is more obvious as MLPREFIX is a common prefix in package names, and as these
filenames are hidden behind alternatives they're rarely used directly.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
The default glob is sufficient now, so we can remove this.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Whilst disabling the binconfig scripts in the sysroot is typically sensible as
these scripts are often not friendly to cross-compilation or sysroots, they tend
to work on the target and more importantly we don't control what is being built
by the user on the target.
Change binconfig-disabled.bbclass to use the logic in binconfig.bbclass to fixup
the packaged binconfig scripts and leave them in the packages, whilst still
disabling the sysroot binconfig scripts.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Unify the behaviour of this class with binconfig-disabled which has far more
sensible behaviour. Currently BINCONFIG_GLOB is a filename glob (defaulting to
*-config) which mangles all *installed* files in the sysroot but files found in
${S} and ${B} for the packages (*not* the installed files in ${D}, which causes
differences if do_install modifies the files). Also FILES_${PN}-dev defaults to
${bindir}/*-config, but the glob matches files in any directory.
The new semantics, matching binconfig-disabled, is that BINCONFIG is a
FILES-style list of globs, defaulting to ${bindir}/*-config. Only files that
are installed to ${D} are processed, not anywhere else.
To help migration to this new model, if BINCONFIG_GLOB is set in a recipe an
error is emitted.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Now that the apr-config scripts are always coming from $D (and not sometimes
from $B) the 'location' replacement is redundant, and explicit removal of
STAGING_DIR_HOST isn't required.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Instead of rummaging around $S and $B for the config scripts, simply look where
do_install put them ($D).
This neatly solves the 'installed vs build tree' differences, and means that the
installed config files match the sysroot config files.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Now that native recipes have crossscripts set to $bindir we need to make sure
that files are correctly written instead of simply truncated, so write to a
temporary file and mv to the destination.
This stops native recipes that use binconfig from shipping both a
/usr/bin/foo-config and /usr/bin/crossscripts/foo-config.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
In the native context we don't use crossscripts, as native.bbclass sets
STAGING_BINDIR_CROSS = "${STAGING_BINDIR_NATIVE}". However, bindir_crossscripts
isn't overridden so some recipes will be writing to /usr/bin/crossscripts in the
native sysroot for no reason.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
|
|
If a config file invokes another config file then there is a chance that it uses
the absolute path of the config file (for example, xslt-config invoking
xml-config). To fix the paths swap STAGING_BINDIR_CROSS with ${bindir}.
DEBUG_PREFIX_MAP and TOOLCHAIN_OPTIONS can often work their way into config
scripts which we don't want, so strip those out entirely.
Fix up some replacements which were not explicitly global.
Finally, add a bbdebug so this operation is in the debug log.
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
|
|
Fix segfault on probing a non-PCI platform device on a system with PCI.
...
at ../../xorg-server-1.20.9/os/log.c:1017
at ../../xorg-server-1.20.9/os/osinit.c:156
at ../../xorg-server-1.20.9/os/osinit.c:110
at ../../../../xorg-server-1.20.9/hw/xfree86/common/xf86platformBus.c:292
argv=argv@entry=0xffffca43c7c8) at ../../../../xorg-server-1.20.9/hw/xfree86/common/xf86Init.c:388
at ../../xorg-server-1.20.9/dix/main.c:193
init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=<optimized out>)
at ../csu/libc-start.c:314
...
Backported from upstream rev e50c85f4e.
(From OE-Core rev: 3b37cbd53219d9c10640b462aa91991d8cbc2a23)
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This CVE is specific to the npm package that can install cmake, so isn't
relevant to our cmake recipe.
(From OE-Core rev: 8e74ed809ec4c1f61264ecf5be4bc319e5e07766)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
(From OE-Core rev: e7cbfd0573b77d7debab3fbf4b05565ac8b33f3a)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
(From OE-Core rev: 3527a361c77ec901a4fa05b7c0835950793b759c)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
To avoid false positives (such as CVE-2010-0734, rubygems:curl), expand
the CVE_PRODUCT list to include all the vendors that have been used.
(From OE-Core rev: bb265122cccea9466405fdd924ad10ce8cda0dec)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Changelog
- Fix issue with uintset and number of bytes copied.
- Fix issue with overflow in DHCP lease T2 computation.
- Fix issue with side channel leak in l_ecc_scalar_new.
- Fix issue with missing MSG_MORE in l_cipher_set_iv.
- Add support for DHCP v6 client implementation.
(From OE-Core rev: 3ad8ca257d40f5041b3ec167e4117c687da448a9)
Signed-off-by: Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Backport the CVE patch from the usptream:
https://gitlab.com/gnutls/gnutls.git
commit 29ee67c205855e848a0a26e6d0e4f65b6b943e0a
(From OE-Core rev: 84b1bc500e318657cb7a8a189b59cc63bc91dca3)
Signed-off-by: Zhixiong Chi <zhixiong.chi@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Its possible some patterns may cause problems with the current path
manipulations, make a small tweak to try and avoid potential pathname
overlap issues.
(From OE-Core rev: 889005dc17d3e3b8eadee907ee2c05b8ff613285)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Similarly to 04ee0e8b95cd8ed890374e0007f976684206b630, ensure only full
build paths are replaced in the environment to avoid breaking buildtools.
(From OE-Core rev: fcd0a9683af1a9155eabbd9056e3b46d4a931b2e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Fix sed: -e expression #1, char 13: unterminated `s' command
Error Message:
| NOTE: Installed into sysroot: []
| NOTE: Skipping as already exists in sysroot: ['pseudo-native', 'glibc', 'patch-native', 'quilt-native', 'gcc-cross-arm', 'gcc-runtime', 'linux-libc-headers', 'libgcc', 'flex-native', 'xz-native', 'libtool-native', 'automake-native', 'binutils-cross-arm', 'zlib-native', 'mpfr-native', 'texinfo-dummy-native', 'autoconf-native', 'libmpc-native', 'gnu-config-native', 'gmp-native', 'attr-native', 'm4-native', 'gettext-minimal-native']
| DEBUG: Python function extend_recipe_sysroot finished
| DEBUG: Executing shell function do_install
| sed: -e expression #1, char 13: unterminated `s' command
| WARNING: exit code 1 from a shell command.
| ERROR: Execution of '/opt/Projects/poky/build/tmp/work/qemux86_64-poky-linux/systemd-serialgetty/1.0-r5/temp/run.do_install.11228' failed with exit code 1:
| sed: -e expression #1, char 13: unterminated `s' command
| WARNING: exit code 1 from a shell command.
|
To Fix this Issue using the strong (single quote) character in sed command.
It is recommend to use quotes. If we have meta-characters in the command, quotes are necessary.
(From OE-Core rev: e2fea05e150dcfec4b7dfbd8edddb53897026bf9)
Signed-off-by: Rahul Kumar <rahulk@mvista.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Support publishing the sdk-update as well as the sdk installer in one
go.
(From OE-Core rev: 97f1544132ba3bb0c85cdafff6a870c7ba0ef2e4)
Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Escaping does not work in my use case. It must be escaped for
python, ssh and shell as well as for different versions of echo.
Let's try it a little less elegant, but hopefully more reliable.
(From OE-Core rev: 5cc1ae332eb6b05d83802c8d64ab2767c7079412)
Signed-off-by: Adrian Freihofer <adrian.freihofer@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Commit 97b439469a45a089431ca9c31893288c855045f4 added a fallback
mechanism for getting the rootfs timestamp. However, it uses curly
braces around the variable name, which causes bitbake resolve the
variable reference, rather than the shell, so the git timestamp
never gets used. Fix the reference to restore the intent of
making it a fallback for when there is no git timestamp to
retrieve.
(From OE-Core rev: fbcf2c1c255b0c61a795c032cf7b67f5db41baa8)
Signed-off-by: Matt Madison <matt@madison.systems>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Otherwise this can interfer with multiprocessing exit handling.
(Bitbake rev: b88816c4c84fa4f5ad39c263f5e75b96476e9768)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
If you make parsing fail (e.g. add something like:
X := "${@d.getVar('MCMACHINES').split()[1]}"
to meson.bbclass, then run "while true; do bitbake -g bash; done"
it will eventually hang. It appears the cancel_join_thread() call the
parsing failure triggers, breaks the results_queue badly enough that it
sits in read() indefintely (called from self.result_queue.get(timeout=0.25)).
The timeout only applies to lock aquisition, not the read call.
I've tried various other approaches such as using cancel_join_thread()
in other places but the only way things don't lock up is to avoid
cancel_join_thread() entirely for results_queue.
I do have a concern that this may adversely affect Ctrl+C handling
but equally, its broken now already and this appears to improve
things.
[YOCTO #14034]
(Bitbake rev: 9c61a1cc7be46c23da1f4ef3bee070fb83c4be57)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
We have an issue where qemu is being starved of resources on our autobuilders.
We can't raise its priority without special capacilties, therefore add a hook
which if present can allow this to happen using an executable
"~/runqemu-renice".
(From OE-Core rev: 141a3c9ce93bc3d526303021ecf0460c6e9fea8a)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
During cooker shutdown, its possible the parser isn't cleaned up. Fix
this (which may partially explain why threads were left hanging around
at exit).
(Bitbake rev: 928609f30f3a20aaa2f88afc18044a4e10199488)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
help2man isn't part of the early build anymore, so there's no point in
referring to it here.
(From meta-yocto rev: c601f5bd5ddfe8e8be709a4541b95c772a0d3b6f)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
1.15 is latest major release changelog is [1] and detailed blog is [2]
Drop hardcoding ldso patch in favor of setting it using GO_LDSO variable
which can be defined in terms of linuxloader defined by OE
Setting GOBUILDMODE to pie is no longer needed
[1] https://golang.org/doc/go1.15
[2] https://blog.golang.org/go1.15
(From OE-Core rev: aa1bfaff4adc9246a2d65592b3a8061d55829086)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This recipe was originally explicitly minimal because it was needed to
bootstrap autotools, but that isn't the case anymore.
Instead of being explicitly native, use BBCLASSEXTEND.
No need to explicitly depend on autoconf-native automake-native, because
the autotools class does that.
As this recipe isn't needed in early build anymore there's no need to
avoid reconfiguring.
(From OE-Core rev: 1e98edf46d0b96da2aea7dabd1d7cf3ce6c5e7bd)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This is the latest release in the 1.7.x series so should be a safe
upgrade, and means we can drop a patch as the AC_TRY_RUN has an
optimistic fallback for cross-compiling now.
(From OE-Core rev: 49ce9c09d86d645f3fa30e41430075b1ffa8c768)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The Kea recipe has PACKAGECONFIG options for boost, openssl, and
log4cplus. However, these are not optional but mandatory dependencies.
Remove the PACKAGECONFIGs and replace with explicit DEPENDS and
EXTRA_OECONF. Also the RDEPENDS in the PACKAGECONFIGs are redundant as
the library dependencies are generated correctly.
(From OE-Core rev: d58dc77ab71ea85f8f12ea4ca2e28e0afee60440)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
kea-msg-compiler is only needed if you alter the messages and the
generated sources need to be rebuilt. When this is the case, there are
better ways to build kea-msg-compiler that don't involve building all of
Kea.
Don't depend on kea-native, remove BBCLASSEXTEND=native, and the target
overrides.
(From OE-Core rev: 5e6174c58d5e3a95eccee773b6670242cb9ab8f4)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Depending on nativesdk- varients in a nativesdk build isn't correct, so
just collapse the DEPENDS down and let bitbake do the right thing (which
is leaving them as -native).
(From OE-Core rev: 4864167ad4ed4c57e49f2aa5e7c58383bddb052b)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
There's no point splitting SRC_URI between the .inc and the .bb when
there's just one recipe and the splitting appeared to be arbitrary.
(From OE-Core rev: f627fa2992420c09e57376fbba62d545512c19f2)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The file path checks are passed an ELF object if the file is an ELF. It
doesn't make a lot of sense to load symlinks to ELFs as if they're in
the same package then the real file will be checked too.
This should speed up do_package_qa slightly as libraries won't be
scanned repeatedly.
(From OE-Core rev: c63af30d3b6350361daff94a59d4f14d7c5395e1)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
The primary Debian archive only contains tarballs which are currently
shipped in a release, so it's easy for a tarball we need to disappear.
Instead, point at snapshot.debian.org to ensure the link remains valid.
(From OE-Core rev: ce48e7d72fc9b747f9c35191d1954a58544ccfe1)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Convert the installed-tests toggle from simply respecting PTEST_ENABLED
to a PACKAGECONFIG, so that it can be turned on/off in the usual manner.
(From OE-Core rev: 08e61ffae6056055b56f93678bcbb9fd71f3303e)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
rpcinfo is also provided vy netkit in meta-networking
(From OE-Core rev: 28183dfd7446de9113773ab89edd0afb4ab82f7e)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Its hard to tell from the server logs whether commands complete or not
(or how long they take). Add extra info to allow more debugging of
server timeouts.
(Bitbake rev: 56285ada585ec1481449522282b335bcb5a2671e)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Knowing which process printed which messages and the timestamp of the
message is useful for debugging, so add this. Ensure the log parsing
isn't affected by using search() instead of match().
(Bitbake rev: 1d043666710df1fa9d9586fd974c0371dd1514b0)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
is started
Showing "leftover process" messages when a new server has started and is being
used by some UI is horrible. Compare the PID data from the lockfile to
avoid this (and the ton of confusing log data it generates).
Also, move the time.sleep() call to be after the first lock attempt, which
reduces noise in the logs significantly.
(Bitbake rev: ce1897a31afb5a14997bc3d2f459b90d43eecb7d)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
(Bitbake rev: ffdb3d3fa690c35e9a96fc451a5811f5131276f3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
This could potentialy account for some of the missing thread cleanup
we're seeing.
(Bitbake rev: 8f2d690428de8934868b406b79c4699a8ebe902c)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|