Richard Purdie [Thu, 24 Feb 2022 14:49:30 +0000 (14:49 +0000)]
python3-pip: Improve reproducibility
Pip installed wheels are not reproducible currently. The direct_url
files encode an installation path and the installed wheels compile
the python files at their location, not their final install location
which is incorrect.
To fix this, simply disable the direct_urls and pass the "root" to
the python compile function to strip that path out of the compiled
files.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Wed, 23 Feb 2022 22:09:10 +0000 (14:09 -0800)]
python3-pip-native: install scripts in ${bindir}
The naïve approach of simply unzipping the built wheel is not
sufficient. 'pip install' would normally have created the
[console_scripts] from entry-points.txt in ${D}${bindir}.
For bootstrapping purposes, hardcode the scripts and install them into
${D}${bindir}. Skip the ${bindir}/pip variant.
[YOCTO #14739]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Feb 2022 12:27:06 +0000 (12:27 +0000)]
python3-pip: Don't change shebang
Patch pip to disable shebang mangling and also force the python executable
to be python3 from the environment when building anything for the target
(or nativesdk).
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Tue, 22 Feb 2022 19:16:33 +0000 (11:16 -0800)]
python3-pip: inherit setuptools_build_meta
Upstream provides a pyproject.toml which declares setuptools.build_meta
as the backend for PEP-517 packaging.
We bootstrap the -native installation by simply unzipping the wheel to
PYTHON_SITEPACKAGES_DIR, so that all other recipes can use pip to
install wheels (as intended by upstream).
Check for presence of ${D}${bindir}/pip and remove if found (this was
originally to make sure Python2 pip was the default).
[YOCTO #14638]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Wed, 23 Feb 2022 22:09:13 +0000 (14:09 -0800)]
python3-wheel-native: install ${bindir}/wheel
pip install would normally install [project.scripts] to ${D}${bindir}
but our naïve bootstrapping only unzipps the wheel to
${D}${PYTHON_SITEPACKAGES_DIR}. Correct this by creating the equivalent
script in do_install:class-native
[YOCTO #14739]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Wed, 23 Feb 2022 12:10:05 +0000 (12:10 +0000)]
pip_install_wheel: Use --ignore-installed for pip
Replace the use of --force-reinstall with --ignore-installed when running pip
install. It can detect currently installed modules in the wrong environment and
try to remove them currently which is not what is desired. Ignoring them is the
correct thing to do.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Tue, 22 Feb 2022 19:16:31 +0000 (11:16 -0800)]
pip_install_wheel.bbclass: add helper class
Provide a helper class to use pip to install wheels built by either
bdist_wheel or a PEP-517 backend.
Set pip install arguments via PIP_INSTALL_ARGS, which can be overriden
by recipes.
Pass --root and --prefix to ensure that pip installs things into the
proper place in sysroot.
By passing --no-deps and --no-index we avoid finicky dependency
checking (pip expects wheels in its cache) and avoid trying to fetch
wheels from pypi.org. This is basically the same behavior we have now,
the dependencies should be declared in the recipe.
Also pass --force-reinstall to make sure built wheels are always installed
so that FILES gets properly populated.
Pass --no-cache to avoid a (harmless) warning about the pip cache in
$HOME be avoiding use of cache. We do not likely want wheels cached
anyway,
pip install changes the python interpreter in scripts installed in
${bindir}, e.g. to #!/usr/bin/nativepython3, correct the behavior after
install to #!/usr/bin/env python3.
[YOCTO #14638]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Tue, 22 Feb 2022 19:16:30 +0000 (11:16 -0800)]
python3-wheel: inherit flit_core
Backport pyproject.toml from flit-backend branch.
Inherit flit_core class to build
Inherit pip_install_wheel to install wheels for target
We need to bootstrap python3-wheel-native in order to have bdist_wheel
available to python3-setuptools-native and the refactored
setuptools3.bbclass. Simply unzip the wheel into
PYTHON_SITEPACKAGES_DIR for class-native.
[YOCTO #14638]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Tim Orling [Tue, 22 Feb 2022 19:16:28 +0000 (11:16 -0800)]
python3-flit-core: add recipe for 3.6.0
This is the core of one of the "new build system" tools--hosted
officially as part of the pypa (Python Packaging Authority)
repositories--it is an increasingly common build tool (e.g.
typing_extensions and tomli) as declared in pyproject.toml for
said packages.
This package provides a very simple bootstrapping method that builds
the source tarball (build_sdist) and a wheel (build_wheel).
Bootstrap -native by simply unzipping the wheel to
PYTHON_SITEPACKAGES_DIR.
Use pip to install the wheel for class-target.
Wheels are the official vehicle for delivering Python packages now.
Eggs (egg-info) are deprecated and will increasingly go away.
[YOCTO #14638]
Signed-off-by: Tim Orling <tim.orling@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
classes: rootfs-postcommands: avoid exception in overlayfs_qa_check
The overlayfs_qa_check in the rootfs-postcommands class throws an
exception if the overlayfs distro feature is enabled but no global
OVERLAYFS_MOUNT_POINT is specifived.
File: '/home/stefan/dalos-linutronix/poky/meta/classes/rootfs-postcommands.bbclass', lineno: 389, function: overlayfs_qa_check
0385: searchpaths = [oe.path.join(imagepath, d.getVar("sysconfdir"), "systemd", "system"),
0386: oe.path.join(imagepath, d.getVar("systemd_system_unitdir"))]
0387:
0388: allUnitExist = True;
*** 0389: for mountPoint in overlayMountPoints:
0390: path = d.getVarFlag('OVERLAYFS_MOUNT_POINT', mountPoint)
0391: unit = mountUnitName(path)
0392:
0393: if not any(os.path.isfile(oe.path.join(dirpath, unit))
Exception: TypeError: 'NoneType' object is not iterable
Signed-off-by: Stefan Herbrechtsmeier <stefan.herbrechtsmeier@weidmueller.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It seems that the homepage has moved (for a while now) from
http://www.mega-nerd.com/libsndfile/
to
https://libsndfile.github.io/libsndfile/
On Github, the development group seems to include Erik de Castro Lopo (the
original author of libsndfile1). Link: https://github.com/orgs/libsndfile/people
All backported CVE patches are in version 1.0.31 (I've checked).
The supported format page [1], mentions that Ogg/Opus is supported since
1.0.29, however that isn't currently added in this patch. And it may
require libopus.
And mp3 is supported at around version 1.1.0, but that version is in beta.
This change focuses solely on updating to 1.0.31 and removing all
backported patches.
Alex Stewart [Tue, 22 Feb 2022 23:44:57 +0000 (17:44 -0600)]
sudo: add /etc/sudoers to sudo-lib conffiles
When OE-core commit 788c95c3bb8ede0d3d6a8f125743ac47c0b3f00e created the
`sudo-lib` subpackage, /etc/sudoers was moved from `sudo` to `sudo-lib`.
The commit didn't update the `CONFFILES:${PN}` assignment in sudo.inc,
however. So the `sudo` base package continued to advertise conffile
ownership of /etc/sudoers, though it did not contain it.
Move the CONFFILES assignment to the sudo.bb file, since it is
packaging-related. Change the package owner to the `sudo-lib`
subpackage, since it is the rightful file-owner.
Signed-off-by: Alex Stewart <alex.stewart@ni.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Wed, 23 Feb 2022 11:03:11 +0000 (19:03 +0800)]
init-system-helpers: upgrade 1.60 -> 1.62
Changelog:
=========
* use fakechroot instead of unsharing the mount namespace and mounting tmpfs
* deb-systemd-invoke: systemctl --machine @<UID> is now available in
v249.10. Adjust the version check accordingly
* Skip build-time tests if DEB_BUILD_OPTIONS=nocheck is set
* Fix typos found by Lintian
* Set Rules-Requires-Root: no
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Wed, 23 Feb 2022 11:03:09 +0000 (19:03 +0800)]
bind: upgrade 9.16.25 -> 9.16.26
Changelog:
==========
Log "quota reached" message when hard quota is reached when accepting a connection.
Add ECS support to the DLZ interface.
A failed view configuration during a named reconfiguration procedure could cause
inconsistencies in BIND internal structures, causing a crash or other unexpected errors.
rndc could crash when interrupted by a signal before receiving a response.
Correctly detect and enable UDP recvmmsg support in all versions of libuv that support it.
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
LiweiSong [Wed, 23 Feb 2022 09:57:55 +0000 (17:57 +0800)]
qemu: add tpm string section to qemu acpi table
This will create _STR and _UID section for TPM in qemu acpi table,
with these section, the missing "description" and "uid" files will
be created for TPM in /sys by kernel.
Signed-off-by: Liwei Song <liwei.song@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Wed, 23 Feb 2022 12:54:31 +0000 (12:54 +0000)]
cve-check: get_cve_info should open the database read-only
All of the function in cve-check should open the database read-only, as
the only writer is the fetch task in cve-update-db. However,
get_cve_info() was failing to do this, which might be causing locking
issues with sqlite.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
wangmy [Wed, 23 Feb 2022 13:08:08 +0000 (21:08 +0800)]
quilt: upgrade 0.66 -> 0.67
Changelog:
=========
- Call pager with original LANG environment variable
- Consistently complain early if no series file is found
- Fix handling of symbolic links by several commands
- Tighten the patch format parsing
- Reuse the shell (performance)
- Document the series file format further
- Document that quilt loads /etc/quilt.quiltrc
- configure: Make stat configurable
- series: Minor optimizations
- setup: Don't obey the settings of any englobing .pc
- setup: Default to fast mode
- quilt.el: Fix documentation of quilt-pc-directory
- quilt.el: Load /etc/quilt.quiltrc if ~/.quiltrc doesn't exist
- quilt.el: Fix quilt-editable when QUILT_PATCHES_PREFIX is set
Signed-off-by: Wang Mingyu <wangmy@fujitsu.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta-openembedded has gtk4 which has its own version of
gtk-update-icon-cache. Allow programs that want to use gtk4
use the proper version of the gtk-update-icon-cache utility
and the proper build and runtime dependencies.
Also use a more readable syntax for conditional values in DEPENDS.
Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Native depmod by default try to find config files from the following
directories:
```
${STAGING_ETCDIR_NATIVE}/depmod.d
/run/depmod.d
/usr/local/lib/depmod.d
/lib/depmod.d
```
but none of them is correct, change to load config files from
${sysconfdir}/depmod.d of basedir.
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com> Signed-off-by: Ming Liu <liu.ming50@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yi Zhao [Thu, 24 Feb 2022 07:41:07 +0000 (15:41 +0800)]
glibc: fix multilib headers conflict for arm
Fixes:
Error: Transaction test error:
file /usr/include/bits/dl_find_object.h conflicts between attempted installs of lib32-libc6-dev-2.35-r0.armv7vet2hf_vfp and libc6-dev-2.35-r0.cortexa57
file /usr/include/bits/rseq.h conflicts between attempted installs of lib32-libc6-dev-2.35-r0.armv7vet2hf_vfp and libc6-dev-2.35-r0.cortexa57
file /usr/include/bits/timesize.h conflicts between attempted installs of lib32-libc6-dev-2.35-r0.armv7vet2hf_vfp and libc6-dev-2.35-r0.cortexa57
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Yi Zhao [Thu, 24 Feb 2022 07:41:06 +0000 (15:41 +0800)]
glibc: unify wordsize.h between arm and aarch64
Make sure this header file is same in arm and aarch64.
Fix the conflict error when enable multilib:
Error: Transaction test error:
file /usr/include/bits/wordsize.h conflicts between attempted installs of lib32-libc6-dev-2.35-r0.armv7vet2hf_vfp and libc6-dev-2.35-r0.cortexa57
Signed-off-by: Yi Zhao <yi.zhao@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Lee Chee Yang [Thu, 24 Feb 2022 02:41:22 +0000 (10:41 +0800)]
libarchive : update to 3.5.3
Libarchive 3.5.3 is a security release
Security Fixes:
extended fix for following symlinks when processing the fixup list (#1566, #1617, CVE-2021-31566)
fix invalid memory access and out of bounds read in RAR5 reader (#1491, #1492, #1493, CVE-2021-36976)
Signed-off-by: Lee Chee Yang <chee.yang.lee@intel.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Murray [Tue, 22 Feb 2022 14:45:34 +0000 (09:45 -0500)]
scripts: fix file writing in convert-spdx-licenses
The convert-spdx-licenses.py script needs the same file closing
fix as was made to convert-variable-renames to ensure modified
file contents get flushed out.
Signed-off-by: Scott Murray <scott.murray@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Martin Jansa [Tue, 22 Feb 2022 13:26:57 +0000 (14:26 +0100)]
icecc.bbclass: fix syntax error
'class' is already taken as keyword and python doesn't like this:
ERROR: Error in compiling python function in oe-core/meta/classes/icecc.bbclass, line 151:
The code lines resulting in this error were:
0020: check_pn = set([pn, bpn])
0021:
0022: class_disable = (d.getVar('ICECC_CLASS_DISABLE') or "").split()
0023:
*** 0024: for class in class_disable:
0025: if bb.data.inherits_class(class, d):
0026: bb.debug(1, "%s: class %s found in disable, disable icecc" % (pn, class))
0027: return "no"
0028:
SyntaxError: invalid syntax (icecc.bbclass, line 151)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Murray [Tue, 22 Feb 2022 04:21:38 +0000 (23:21 -0500)]
scripts: fix file writing in convert-variable-renames
In my test environments (Fedora 35 and Debian 10.10 on AMD 2970WX),
running the convert-variable-renames.py rename script was resulting
in empty files instead of updated ones. From inspection, the new
files are never flushed/closed before moving them into place, which
seems inherently racy. Adding an explicit close to flush the modified
contents out before moving into place fixes the issue for me.
Signed-off-by: Scott Murray <scott.murray@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Thu, 17 Feb 2022 11:43:36 +0000 (11:43 +0000)]
icecc: Improve variables/terminology
The SYSTEM and USER seperation between variables seems odd and not necessary,
drop it. Avoid the use of whitelist/blacklist and also change "packages" to
"recipes" since that misuse causes confusion.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Saul Wold [Thu, 17 Feb 2022 00:32:48 +0000 (16:32 -0800)]
scripts: Add convert-variable-renames script for inclusive language variable renaming
This script searches for a list of variable that have been renamed
and converts them to their more descriptive names. It also searches
for a list of variables that have been removed or deprecated and
prints a message.
It will print a message to inform the user that there are terms that
need to be updated in their files. Many of these changes are context
sensitive and may not be modified as they might be existing calls to
other libraries. This message is informational only.
I have tested this on poky and meta-openembedded so far.
Saul Wold [Fri, 4 Feb 2022 17:01:40 +0000 (09:01 -0800)]
blacklist: Replace class with SKIP_RECIPE variable
Remove the old class and rename VarFlag to SKIP_RECIPE, handling this
in base.bbclass for efficiency. This means a separate inherit is no longer
needed.
This change better describes what the VarFlag is doing since it
is implemeted with the SkipRecipe() function.
By moving this into base.bbclass we simplify the distro inherit.
Signed-off-by: Saul Wold <saul.wold@windriver.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Scott Murray [Thu, 17 Feb 2022 22:12:55 +0000 (17:12 -0500)]
local.conf/oeqa: Update BB_DISKMON_DIRS use
Update the example BB_DISKMON_DIRS definitions in the sample
local.conf files for the rename of the "ABORT" action to "HALT",
and also update related error messages in one selftest to use
the new action name.
Signed-off-by: Scott Murray <scott.murray@konsulko.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
yocto-check-layer: add ability to perform tests from a global bbclass
This is useful when needing to test layer's recipes, where this special
bbclass can define a global python function that gets called on each
recipe parsing during "bitbake -S none world" signature dump and be able
to fail layer's check accordingly.
First test being added is to detect recipes skipping "installed-vs-shipped"
QA check. As "installed-vs-shipped" is a packaging QA check, it happens very
late in the build process and failing it could mean some potential issues
with packaging, especially when recipe uses BBCLASSEXTEND="nativesdk" and
resulting package is used in an SDK.
In OE-Core failing this QA check leads to an error, but other layers can
suppress it or change it to a warning. Detecting weird packaging problems
with SDKs is quite difficult and time consuming. Also, waiting for the
actual "installed-vs-shipped" packaging QA check to fail means that all
recipes in the layer under test have to run through all standard tasks in
the build chain, equivalent to a multi-hour world-build.
Hence yocto-check-layer takes a shortcut and only detects a mere attempt
at skipping "installed-vs-shipped" QA check in the INSANE_SKIP list during
initial parsing when dumping the signature information for the layer.
Signed-off-by: Denys Dmytriyenko <denis@denix.org> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Ross Burton [Thu, 10 Feb 2022 11:52:01 +0000 (11:52 +0000)]
kernel: make kernel-base recommend kernel-image, not depend
As kernel modules depend on the base kernel package, and the base kernel
recipe depends on the kernel image, it's impossible to build file system
images which contain kernel modules but not the kernel itself, such as
an initramfs.
Change the RDEPENDS to RRECOMMENDS so that the disk images can set
PACKAGE_EXCLUDE = "kernel-image-*" to remove the kernel.
Signed-off-by: Ross Burton <ross.burton@arm.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
oeqa/selftest/bblogging: Add logging tests for bb.build.exec_func with shell/python code
The situation regarding logging is different when a function called by
bb.build.exec_func() fails compared to when the task code fails
directly. There is a recent fix in bitbake to solve that and these
tests will hopefully prevent regressions.
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Richard Purdie [Sun, 20 Feb 2022 14:03:39 +0000 (14:03 +0000)]
perl: Improve and update module RPDEPENDS
The perl module RDEPENDS needed refreshing so I updated it with the script. I
also found a ton of issues with missing test2 module dependencies so I've tweaked
the generator script to improve those. For some reason they were previously excluded
but I can't see anything wrong with the generated dependencies so I've added them. A
number of manual dependencies were still needed in addition.
I did notice an issue with trailing ")" characters so the sed is updated to handle
that (it did already handle "(").
This fixes issues seen in the dependecies for libxml-parser-perl's ptest package.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
