Sreejith Ravi [Wed, 26 Jun 2024 14:05:00 +0000 (14:05 +0000)]
package.py: Add Requires.private field in process_pkgconfig
Currently, the process_pkgconfig() only handles the "Requires" key
field from the .pc file and ignores the "Requires.private" field while
generating the dev dependency chain. This results in a broken dependency
list and requires the installation of recommended packages to resolve
the build dependencies when using the dev IPKs. This increases the
hard disk space usage and download time and installs many unnecessary
packages as part of the recommendations.
This patch ensures that the "Requires.private" field is also considered
when creating the dependency list for the dev IPKs. With this, the dev
IPK will have the proper dependencies listed, eliminating the need to
install recommended packages to resolve the build time dependencies.
When trying to resolve the build time dependency with libical
package using “-dev” ipk generated, it will throw the below error.
-----------
Package icu-i18n was not found in the pkg-config search path.
Perhaps you should add the directory containing `icu-i18n.pc'
to the PKG_CONFIG_PATH environment variable
Package 'icu-i18n', required by 'libical', not found
-----------
This patch will fix the broken dependency list.
-------
libical-dev depends field generated with this patch
Depends: glib-2.0-dev, icu-dev, libical (= 3.0.7-r0)
-------
Other examples of packages generated with broken dev dependency.
libflac-dev : https://packages.debian.org/sid/libflac-dev
Without patch:
Depends: flac (= 1.3.3-r0), libflac, libflac++
with patch:
Depends: flac (= 1.3.3-r0), libflac, libflac++, libogg-dev
Joshua Watt [Mon, 24 Jun 2024 19:10:59 +0000 (13:10 -0600)]
classes/spdx-common: Move SPDX_SUPPLIER
Move the SPDX_SUPPLIER variable to create-spdx-2.2 since it's format
only has meaning in SPDX 2.2 (SPDX 3 uses SPDX_PACKAGE_SUPPLIER with a
different format)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Joshua Watt [Mon, 24 Jun 2024 19:10:55 +0000 (13:10 -0600)]
binutils-cross-testsuite: Rename to binutils-testsuite
This recipe needs to be renamed because the "-cross-" substring in the
name triggers the cross architecture detection in sstate, but this
recipe is not actually a cross recipe.
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jan Vermaete [Sat, 22 Jun 2024 06:14:38 +0000 (08:14 +0200)]
python3-websockets: added python3-zipp as RDEPENDS
Because the Python package 'zipfile' was missing.
root@qemux86-64:~# python3 clTraceback (most recent call last):
File "/home/root/server.py", line 4, in <module>
from websockets.server import serve
File "/usr/lib/python3.12/site-packages/websockets/__init__.py", line 6, in <module>
from .version import version as __version__ # noqa: F401
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/lib/python3.12/site-packages/websockets/version.py", line 3, in <module>
import importlib.metadata
File "/usr/lib/python3.12/importlib/metadata/__init__.py", line 8, in <module>
import zipfile
ModuleNotFoundError: No module named 'zipfile'
Signed-off-by: Jan Vermaete <jan.vermaete@gmail.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Kai Kang [Tue, 25 Jun 2024 02:02:49 +0000 (10:02 +0800)]
glibc: fix fortran header file conflict for arm
There is a file conflict for arm when multilib enabled:
| Error: Transaction test error:
| file /usr/include/finclude/math-vector-fortran.h conflicts between attempted installs of
lib32-libc6-dev-2.39+git0+312e159626-r0.armv7at2hf_neon and libc6-dev-2.39+git0+312e159626-r0.cortexa72
Install math-vector-fortran.h to the gfortran default search directory
which is arch specific to avoid the conflict.
Signed-off-by: Kai Kang <kai.kang@windriver.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
* after:
https://git.openembedded.org/openembedded-core/commit/?id=44fc7aa1468ff042739cc5a91c84ef5c2a09e0a3
nativesdk-libgfortran is pulled as dependency of nativesdk-gcc so this affects
more people who didn't explicitly use nativesdk-libgfortran before
* the INHIBIT_DEFAULT_DEPS and gcc-runtime was there since gcc-4.8:
https://git.openembedded.org/openembedded-core/commit/?id=a5e7ee5770b9e0cf719c573efffd874440f74289
Signed-off-by: Martin Jansa <martin.jansa@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Daniel Semkowicz [Thu, 20 Jun 2024 13:03:31 +0000 (15:03 +0200)]
os-release: Fix VERSION_CODENAME in case it is empty
If DISTRO_CODENAME was not set, VERSION_CODENAME field was populated
with unparsed string. This resulted in the following line in os-release
file:
VERSION_CODENAME="${DISTRO_CODENAME}"
According to systemd documentation, this field is optional.
Fix the problem by setting VERSION_CODENAME conditionally, only if
DISTRO_CODENAME was set.
Signed-off-by: Daniel Semkowicz <dse@thaumatec.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Wang Mingyu [Fri, 21 Jun 2024 02:16:45 +0000 (10:16 +0800)]
rng-tools: upgrade 6.16 -> 6.17
Changelog:
===========
- mix data fed to the kernel byte-wise from multiple sources
- added option to attempt more persistent use of slow entropy sources
- fix some missing m4 quotes
- improved debug output to show FIPS failures more clearly
- added a named pipe entropy source
- adjusted linux poolsize
- fixed some pkcs11 error messages
- fixed ignorefail and random_step options
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Wang Mingyu [Fri, 21 Jun 2024 02:16:44 +0000 (10:16 +0800)]
pciutils: upgrade 3.12.0 -> 3.13.0
Changelog:
===========
* lspci decodes CXL 1.1 device link status information.
This requires a recent kernel which exports rcd_* atributes via
sysfs.
* Further development of the pcilmr (the link margining utility)
* Dump parsing supports 6-digit domain numbers.
* Bug fixes in PCIe link state reporting.
* Decode more fields in PCIe AER capability.
* Fixed build on Linux systems with musl libc.
* Updated pci.ids.
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Wang Mingyu [Fri, 21 Jun 2024 02:16:40 +0000 (10:16 +0800)]
libjitterentropy: upgrade 3.4.1 -> 3.5.0
Lisence-Update: Copyright year updated to 2024.
Changelog:
===========
* add distinction between intermittent and permanent health failure
* add compile time option to allow configuring a mask to reduce the size of
the time stamp used for the APT
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Wang Mingyu [Fri, 21 Jun 2024 02:16:30 +0000 (10:16 +0800)]
diffoscope: upgrade 267 -> 271
Changelog:
===========
* Drop Build-Depends on liblz4-tool. Thanks, Chris Peterson.
* Update tests to support zipdetails version 4.004 shipped with Perl 5.40.
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Wang Mingyu [Fri, 21 Jun 2024 02:16:34 +0000 (10:16 +0800)]
git: upgrade 2.45.1 -> 2.45.2
Changeglog:
============
- send-email: drop FakeTerm hack
- send-email: avoid creating more than one Term::ReadLine object
- ci: drop mention of BREW_INSTALL_PACKAGES variable
- ci: avoid bare "gcc" for osx-gcc job
- ci: stop installing "gcc-13" for osx-gcc
- hook: plug a new memory leak
- init: use the correct path of the templates directory again
- Revert "core.hooksPath: add some protection while cloning"
- tests: verify that `clone -c core.hooksPath=/dev/null` works again
- clone: drop the protections where hooks aren't run
- Revert "Add a helper function to compare file contents"
- Revert "fsck: warn about symlink pointing inside a gitdir"
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Clément Péron [Fri, 21 Jun 2024 10:10:34 +0000 (12:10 +0200)]
openssl: Remove patch already upstreamed
Since the bump to OpenSSL 3.3.1, riscv32 doesn't build anymore due to
the folowing error:
crypto/riscv32cpuid.s:77: Error: symbol `riscv_vlen_asm' is already
defined
systemd: bpf-framework: pass 'recipe-sysroot' to BPF compiler
Pass the "recipe-sysroot" path via the CFLAGS=--sysroot= to the
compiler used by systemd to build the BPF, so that it can find the
needed system includes.
Signed-off-by: Johannes Schneider <johannes.schneider@leica-geosystems.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
systemd: bpf-framework: 'propagate' the '--sysroot=' for crosscompilation
The eBPFs are pre-compiled during the systemd-build with a different
compiler than the cross-compiler used to build systemd itself.
This is either a 'clang-native' or a gcc (bpf-unknown-none) which do
not see the BUILD_CFLAGS, that point to the correct include search
patch. To address this have systemd's meson.build "propagate" the
--system from the C_FLAGS into the BPF compiler call.
Signed-off-by: Johannes Schneider <johannes.schneider@leica-geosystems.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
The bpf-framework is used to pre-compile eBPFs that required for the
systemd.resource-control features RestrictFileSystems=[1] and
RestrictNetworkInterfaces=[2] to work.
Apart from 'clang-native' to compile the eBPFs, the required kernel
switches are described in [3].
- Treat methodName="runTest" similar to unittest.TestCase,
fixes compatibility with pytest 8.3. (Natanael Copa, #372)
- Format with ``ruff format``. (Jelmer Vernooij)
- Use ruff for linting. (Jelmer Vernooij)
- Fix compatibility with Python 3.12.1. (Matthew Treinish)
- Deprecate SkippedTest exception. (Stephen Finucane)
- Drop support for Python 3.7. (Jelmer Vernooij)
6.103.2 - 2024-06-14
- This patch improves our deduplication tracking across all strategies (pull request #4007). Hypothesis is now less likely to generate the same input twice.
6.103.1 - 2024-06-05
- Account for time spent in garbage collection during tests, to avoid flaky DeadlineExceeded errors as seen in issue #3975.
- Also fixes overcounting of stateful run times, a minor observability bug dating to version 6.98.9 (pull request #3890).
10.0.8:
- Fixed compile without ARP
- Fixed closefrom test for glibc
- Fixed spelling of ADVERTISEMENT
10.0.7:
- DHCP: use request_time, fallback_time and ipv4ll_time rather than reboot timeout
- DHCP6: Wait for IRT to elapse before requesting advertisments
- DHCPv6: Don't re-INFORM if the RA changes
- privsep: Reduce fd use
- dhcpcd: Add support for arp persist defence by @pradeep-brightsign in #273
- Move dhcp(v4) packet size check earlier by @pemensik in #295
- Define the Azure Endpoint and other site-specific options by @lparkes in #299
- add RFC4191 support by @goertzenator in #297
- dhcpcd: Respect IPV6_PREFERRED_ONLY flag regardless of state by @taoyl-g in #307
- Fix time_offset to be int to match RFC-2132 by @ColinMcInnes in #319
- hooks/30-hostname: Exit with 0 if setting hostname is not needed by @bdrung in #320
Jose Quaresma [Wed, 19 Jun 2024 15:47:43 +0000 (16:47 +0100)]
go: drop the old 1.4 bootstrap C version
Bootstrap [1]
As mentioned in the Go 1.20 release notes, Go 1.22 now requires the final point release of Go 1.20 or later for bootstrap.
We expect that Go 1.24 will require the final point release of Go 1.22 or later for bootstrap.
The default recipe for bootstrap is the go-binary-native as can be seen in:
meta/conf/distro/include/tcmode-default.inc:68:PREFERRED_PROVIDER_go-native ?= "go-binary-native"
Currently if we change it to use the old go-native and compile the go1.4-bootstrap-20170531
it fails:
| Building Go cmd/dist using /build/workdir/tmp-glibc/work/x86_64-linux/go-native/1.22.3-r0/go1.4/go. (go1.4-bootstrap-20170531 linux/amd64)
| can't load package: package ./cmd/dist: found packages build.go (main) and notgo120.go (building_Go_requires_Go_1_20_6_or_later) in /build/workdir/tmp-glibc/work/x86_64-linux/go-native/1.22.3-r0/go/src/cmd/dist
This has been broken for some time but as we used go-binary-native by default it went unnoticed.
[1] https://go.dev/doc/go1.22#bootstrap
Signed-off-by: Jose Quaresma <jose.quaresma@foundries.io> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
automake: mark new_rt_path_for_test-driver.patch as Inappropriate
The patch is specific to how oe-core runs autotools-generated tests:
by cherry-picking needed bits from builddir and srcdir, then hacking
Makefile with sed until it runs.
As GNU is not interested in installable tests, they wouldn't be
interested in this patch either; and if they become interested,
it's probably going to be done in a whole different way.
Signed-off-by: Alexander Kanavin <alex@linutronix.de> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As the comment says this was used by very old kernels
(pre 3.10), and slang has been the only option for a
long time. Also, correct the option to disable slang TUI.
Signed-off-by: Alexander Kanavin <alex@linutronix.de> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Tue, 18 Jun 2024 11:14:13 +0000 (11:14 +0000)]
expect: update code for Tcl channel implementation
Tcl 8.4.0 changed the channel implementation, take a patch submitted
upstream to update the code for the new interface and remove the silencing
of the incompatible assignment error which was due to this issue.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Hundeb?ll [Mon, 17 Jun 2024 13:36:40 +0000 (15:36 +0200)]
ofono: upgrade 2.7 -> 2.8
Changelog:
* Release 2.8
* build: Require at least version 0.66 when building with external ELL
* qmi: Remove unused shutdown members in qmi_device_qrtr
* drivers: Use the new license header format
* plugins: Use the new license header format
* isimodem: Use the new license header format
* rilmodem: Use the new license header format
* mbimmodem: Use the new license header format
* unit: Use the new license header format
* tools: Use the new license header format
* include: Use the new license header format
* dundee: Use the new license header format
* core: Use the new license header format
* qmimodem: Use the new license header format
* atmodem: Use the new license header format
* build: Add notifylist.[ch] from ell
* udevng: Fix detection of USB attached tty devices
* udevng: Don't crash for non-QMI devices
* qmimodem: Remove the create exclusive service API
* voicecall: Fix use after free
* atmodem: gprs-context: use default PPP ACCM for Quectel serial modems
* atmodem: sim: Fix CRSM result handling
* qmi: gprs-context: support bind_mux for pcie devices
* udevng: Add mhi subsystem detection
* udevng: add and use get_ifname() for netdev nodes
* gobi: ensure required properties are provided
* gobi: Rename KernelDriver to NetworkInterfaceDriver
* gobi: add / use DeviceProtocol property
* qmi: Use l_basename instead of basename
* log: Use l_basename instead of basename
* build: link dundee with ell
* log: Handle dladdr failure
License-Update: license header replaced with spdx identifier Signed-off-by: Martin Hundebøll <martin@geanix.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/runtime: fix race-condition in minidebuginfo test
Fix this error where 'coredumpctl info' warns that the coredump is still being
processed:
```
AssertionError: 1 != 0 : MiniDebugInfo Test failed: No match found.
-- Notice: 1 systemd-coredump@.service unit is running, output may be incomplete.
```
The variable contains the suffix of the CVE JSON manifest file.
By default, this variable is set to 'json', so the current behavior
is not changed, but enables developers to use some other suffix,
e.g., cve.json (similar to spdx.json).
Signed-off-by: Aleksandar Nikolic <an010@live.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Jörg Sommer [Sun, 9 Jun 2024 14:59:07 +0000 (16:59 +0200)]
classes/kernel: No symlink in postinst without KERNEL_IMAGETYPE_SYMLINK
The commit “Use a copy of image for kernel*.rpm if fs doesn't support
symlinks” [1] added postinst and postrm scripts to the kernel package which
create a symlink after package installation. This should not happen if
`KERNEL_IMAGETYPE_SYMLINK` is not `1`.
Background: The u-boot implementation of jffs2 does not support symlinks.
Using a hardlink or removing `${KERNEL_VERSION}` from the file name fails,
because the current postinst script replaces the file with the symlink.
Benjamin Szőke [Fri, 14 Jun 2024 07:37:29 +0000 (09:37 +0200)]
archiver.bbclass: Fix work-shared checking for kernel recipes
Source dir can be a symbolic link in some BSP's linux kernel recipe
which points to work-shared path (like linux-fslc in meta-freescale).
Change to use os.path.realpath() in order to get real path of
source dir.
Signed-off-by: Benjamin Szőke <egyszeregy@freemail.hu> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yi Zhao [Sun, 16 Jun 2024 11:33:37 +0000 (19:33 +0800)]
debianutils: upgrade 5.17 -> 5.19
ChangeLog:
5.19:
* update-shells: Avoid duplicate lines when package shells contain both
aliased and canonical shells.
5.18
* po4a/po/fr.po: Updated french translations for manpages.
* d/control: Bump standards version from 4.6.2 to 4.7.0.
* acinclude.m4: Bump DEBIANUTILS_VERSION from 5.17 to 5.18.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* fix calculations for time differences
* fix extension for zip compression
* fix omitted copy for logs with mail and rotate 0
* fix wrongly skipping copy with copytruncate and compress
* fix ambiguities between mode, UID and GID parsing when not specifying
all options
* fix hang when encountering a named pipe
* on prerotate failure logs are preserved instead of rotated
* in case a configuration file was skipped due to unsafe permissions
the exit status after rotattion will be 1
* the state is no longer written to non-regular files
* the systemd timer now correctly utilizes load distribution
* add dateformat specifier %z for timezone offsets
* change default mode for created olddir directories to 0755
* support quoted user and group names in su, create, and createolddir
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Dmitry Baryshkov [Sat, 15 Jun 2024 11:11:57 +0000 (14:11 +0300)]
ffmpeg: backport patch to fix errors with GCC 14
On ARMv7 compilation of ffmpeg breaks if Vulkan support is enabled.
Backport a patch from the trunk to fix compilation issues:
| src/libavcodec/vulkan_av1.c: In function 'vk_av1_create_params':
| src/libavcodec/vulkan_av1.c:214:43: error: initialization of 'long long unsigned int' from 'void *' makes integer from pointer without a cast [-Wint-conversion]
| 214 | .videoSessionParametersTemplate = NULL,
| | ^~~~
| src/libavcodec/vulkan_av1.c:214:43: note: (near initialization for '(anonymous).videoSessionParametersTemplate')
| make: *** [/oe/build/tmp-rpb_wayland-glibc/work/armv7at2hf-neon-linaro-linux-gnueabi/ffmpeg/6.1.1/ffmpeg-6.1.1/ffbuild/common.mak:81: libavcodec/vulkan_av1.o] Error 1
| make: *** Waiting for unfinished jobs....
| src/libavcodec/vulkan_decode.c: In function 'ff_vk_decode_prepare_frame':
| src/libavcodec/vulkan_decode.c:191:26: error: assignment to 'VkImageView' {aka 'long long unsigned int'} from 'void *' makes integer from pointer without a cast [-Wint-conversion]
| 191 | vkpic->img_view_ref = NULL;
| | ^
| src/libavcodec/vulkan_decode.c:192:26: error: assignment to 'VkImageView' {aka 'long long unsigned int'} from 'void *' makes integer from pointer without a cast [-Wint-conversion]
| 192 | vkpic->img_view_out = NULL;
| | ^
| src/libavcodec/vulkan_decode.c:193:26: error: assignment to 'VkImageView' {aka 'long long unsigned int'} from 'void *' makes integer from pointer without a cast [-Wint-conversion]
| 193 | vkpic->img_view_dest = NULL;
| | ^
| make: *** [/oe/build/tmp-rpb_wayland-glibc/work/armv7at2hf-neon-linaro-linux-gnueabi/ffmpeg/6.1.1/ffmpeg-6.1.1/ffbuild/common.mak:81: libavcodec/vulkan_decode.o] Error 1
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changqing Li [Thu, 30 May 2024 09:02:34 +0000 (17:02 +0800)]
systemd: fix wrong path of tmp.mount
According to meson.build, tmp.mount is installed under
"prefixdir/'lib/systemd/system'", but for 64bit system, rootlibdir is
/usr/lib64/systemd/system, this make tmp.mount not removed, and /tmp
still mounted as tmpfs filesystem. Fixed by using rootlibexecdir, which
is /usr/lib/systemd/system.
Signed-off-by: Changqing Li <changqing.li@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 12 Jun 2024 17:57:34 +0000 (18:57 +0100)]
oeqa/sdk/case: Ensure DL_DIR is populated with artefacts if used
Where we're using DL_DIR in sdk archive to try and cache testing artefacts,
copy into the cache so that it gets populated and this doesn't have to be done
manually. Currently we're making a lot of repeat requests to github as this
wasn't being populated.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 12 Jun 2024 11:06:59 +0000 (11:06 +0000)]
curl: rewrite ptest installation
The latest libtool upgrade appears to have resulted in intermediate files
containing build paths. This wouldn't normally be a problem but the
curl-ptest package is populated by copying ${B}/tests/ which includes all
of the intermediate objects, so this causes buildpath warnings.
Rewrite the ptest installation to install just the pieces we need: the
test harness, utility scripts, and the test data. We do not need the
libcurl unit tests nor the HTTP server tests, as we don't run those.
Remove all of the explicitly disabled tests, as many of these were for
tests that run curl-config or scan the source code, neither of which are
available at ptest time. Instead use keyword exclusions to skip them.
Tell the test runner to use the system curl instead of symlinking in the
binary.
Don't copy curl-config, skipping these minor tests is acceptable.
Remove the RDEPENDS on bash, nothing needs this now we're not shipping a
build tree.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 12 Jun 2024 11:06:58 +0000 (11:06 +0000)]
scripts/makefile-getvar: add script to get values from Makefiles
There is often a need to extract a value from a Makefile, and standard
GNU Make doesn't provide a way to do this. This script lets you access
values from Makefiles directly:
Tim Orling [Tue, 11 Jun 2024 15:55:15 +0000 (08:55 -0700)]
oe-selftest: add RECIPE_UPDATE_EXTRA_TASKS test
Add test_devtool_upgrade_recipe_update_extra_tasks test case
to test upgrade of python3-guessing-game from v0.1.0 to v0.2.0
which will exercise the update_crates task during the upgrade.
Add python3-guessing-game_git.bb.upgraded and
python3-guessing-game-crates.inc.upgraded which are the 0.2.0
variants.
Check that the new recipe file has the expected differences.
Check that the new -crates.inc file has the expected differences,
which should be reproducible because of Cargo.lock.
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Tue, 11 Jun 2024 15:55:14 +0000 (08:55 -0700)]
meta-selftest: add python3-guessing-game
Add v0.1.0 of python3-guessing-game which is used as the baseline
for an upgrade to v0.2.0 in test_devtool_upgrade_recipe_update_extra_tasks
test case.
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When we upgrade a recipe that inherits cargo-update-recipe-crates and
the upstream Cargo.toml/Cargo.lock have been changed, we need to run
the update_crates task or else the devtool upgrade (and therefore
AUH upgrade) will fail.
Add "do_update_crates" task to RECIPE_UPDATE_EXTRA_TASKS for all
recipes that inherit this class.
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Wed, 12 Jun 2024 17:28:49 +0000 (10:28 -0700)]
devtool upgrade: enable RECIPE_UPDATE_EXTRA_TASKS
For some recipes, such as those that inherit cargo-update-recipe-crates,
we need to run additional tasks once the new sources have been unpacked.
Introduce a new variable RECIPE_UPDATE_EXTRA_TASKS which is a space-
delimited list of tasks to run after the new sources have been
unpacked in scripts/lib/devtool/upgrade.py ugrade() method.
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Dan McGregor [Thu, 28 Mar 2024 15:43:40 +0000 (09:43 -0600)]
gcc: Allow using libc++
With the addition of the C++ runtime setting added recently, allow
gcc to use libc++ as its runtime. There's some minor fixes still
required, such as allowing setting the unwinder library. But this
allows for testing libc++ with gcc.
Signed-off-by: Daniel McGregor <daniel.mcgregor@vecima.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The "Test installation" step fails with some harmless error messages
(see [1]). This can however make a user think that the buildtools
have not been installed correctly.
Two reasons for the error messages:
- some envvars in the environment-setup-<arch>-pokysdk-linux file
start and end with double quotes (e.g., PATH) and are as such
written into python os.environ. This leads that their usage is
not valid later when testing the installation. This patch removes
the double quotes before writing, if they are present.
- if installation directory (install_dir), given through the option
--directory, is given as a relative path, checking if the path to
a tool (e.g., gcc) in buildtools starts it will always fail. This
patch converts the install_dir variable to an absolute path.
[1]
ERROR: Something went wrong: tar not found in ./build-tools
ERROR: Something went wrong: installation failed
Signed-off-by: Aleksandar Nikolic <aleksandar.nikolic@zeiss.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
projects / thirdparty / openembedded / openembedded-core-contrib.git / log
