Passing params as numbers to hwclock is broken in util-linux 2.39.3 due
to wrong pointer handling. So backport the fix from upstream included
since util-linux 2.41.
(From OE-Core rev: 3d8f88906f5560286462eaf55226b872e2805df7)
Signed-off-by: Bastian Krause <bst@pengutronix.de>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
This change already exists on master, but it was made as part of the
larger migration to ${UNPACKDIR} and is not cherry-pickable.
See: d73595df696 (recipes: Update WORKDIR references to UNPACKDIR)
(From OE-Core rev: 378f87f087651bacdb6efc6b98168bc6ba865070)
Signed-off-by: Chris Laplante <chris.laplante@agilent.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
During ptest execution, util-linux adds mount entries in /etc/fstab
and runs `mount -a`, which mounts all available entries from
/etc/fstab. This can cause unintended mounts that are unrelated to
the test, leading to incorrect test behavior.
To avoid this, upstream util-linux introduced a mechanism using
CUSTOM_FSTAB,which isolates test-specific fstab entries. Only entries
listed in CUSTOM_FSTAB are mounted during test execution, ensuring
tests do not interfere with or depend on the system's /etc/fstab.
This commit backports below upstream changes to use CUSTOM_FSTAB.
ed3d33faffb1580bd7606aa8d17b6b
(From OE-Core rev: e7420db0d77611140149ccfefefc8becfad4f34b)
Signed-off-by: Virendra Thakur <virendra.thakur@kpit.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
The 2.39 version of util-linux took new file descriptors based mount
kernel API into use. In relation to this change, the upstream release
notes in
https://github.com/util-linux/util-linux/blob/v2.39/Documentation/releases/v2.39-ReleaseNotes#L14-L21
mention that
This change is very aggressive to libmount code, but hopefully, it does not introduce regressions in traditional mount(8) behavior.
After observing following failure when booting a board using a bit
older 6.1 series kernel together with initramfs rootfs based boot flow
[FAILED] Failed to start Remount Root and Kernel File Systems.
See 'systemctl status systemd-remount-fs.service' for details.
closer inspection revealed:
demoboard ~ # systemctl status -l systemd-remount-fs.service
x systemd-remount-fs.service - Remount Root and Kernel File Systems
Loaded: loaded (/usr/lib/systemd/system/systemd-remount-fs.service; enabled-runtime; preset: disabled)
Active: failed (Result: exit-code) since Wed 2024-08-14 14:53:48 UTC; 1min 22s ago
Docs: man:systemd-remount-fs.service(8)
https://www.freedesktop.org/wiki/Software/systemd/APIFileSystems
Process: 76 ExecStart=/usr/lib/systemd/systemd-remount-fs (code=exited, status=1/FAILURE)
Main PID: 76 (code=exited, status=1/FAILURE)
Aug 14 14:53:48 demoboard systemd-remount-fs[76]: /usr/bin/mount for / exited with exit status 32.
Aug 14 14:53:48 demoboard systemd-remount-fs[81]: mount: /: mount point not mounted or bad option.
Aug 14 14:53:48 demoboard systemd-remount-fs[81]: dmesg(1) may have more information after failed mount system call.
Aug 14 14:53:48 demoboard systemd[1]: systemd-remount-fs.service: Main process exited, code=exited, status=1/FAILURE
Aug 14 14:53:48 demoboard systemd[1]: systemd-remount-fs.service: Failed with result 'exit-code'.
Aug 14 14:53:48 demoboard systemd[1]: Failed to start Remount Root and Kernel File Systems.
also consequentially, 'systemctl status' reported:
State: degraded
When issuing 'strace -ff mount -o remount /' the failure occurred at
mount_setattr(3, "", AT_EMPTY_PATH, {attr_set=MOUNT_ATTR_RDONLY|MOUNT_ATTR_NOATIME|MOUNT_ATTR_NODIRATIME, attr_clr=MOUNT_ATTR_NOSUID|MOUNT_ATTR_NODEV|MOUNT_ATTR_NOEXEC|MOUNT_ATTR_NOATIME|MOUNT_ATTR_STRICTATIME|MOUNT_ATTR_NOSYMFOLLOW|0x40, propagation=0 /* MS_??? */, userns_fd=0}, 32) = -1 EINVAL (Invalid argument)
After further investigation, The issue was pinpointed to lack of Linux
kernel commit
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=95de4ad173ca0e61034f3145d66917970961c210
("fs: relax mount_setattr() permission checks") in the kernel version
that was being used. Above mitigation was discussed in email related to
then-rejected CVE-2024-26821:
https://lore.kernel.org/linux-cve-announce/2024051606-imaging-entrench-b327@gregkh/T/
After testing with qemuarm64 machine different linux-yocto versions,
it was observed that the issue impacts following versions of currently
supported LTS kernels:
- 6.6.17 (fixed since 6.6.18 i.e. mount_setattr() returns 0)
- 6.1.78 (fixed since 6.1.79 i.e. mount_setattr() returns 0)
- 5.15.164 which is currently the newest of 5.15.y series (i.e. no
known working version)
Taking the above findings into consideration, add a new PACKAGECONFIG
option removing which enables users to opt-out from using the feature
which can cause issues with a bit older kernels. The option is enabled
only for class-target here, since it otherwise causes following error
during util-linux-native's do_configure task on Debian 11 build host
(mountfd_api requirement fails):
| configure: error: libmount_mountfd_support selected, but required mount FDs based API not available
Versions 5.10.223, 5.4.279 and 4.10.317 were also tested with qemuarm64
but the issue was not reproduced with those versions - using strace
showed that the mount_setattr call associated with the new mount API
problem was not issued with these LTS kernel versions, which seemed to
be confirmed also by following libmount debug message in these cases:
415: libmount: HOOK: [0x7fa115e818]: failed to init new API
Note: In addition to the aforementioned, this change was tested also
briefly using the current latest kernel versions 6.1.104, 6.6.45 and
6.10.3 that using the old mount API with newest kernels did not
introduce any observable regression to the boot flow.
(From OE-Core rev: dc086d9a8613143607af3583c72ed892e20b4d66)
(From OE-Core rev: c0487c8141ae2b9242447b53b71052769db6338f)
Signed-off-by: Niko Mauno <niko.mauno@vaisala.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
wall in util-linux through 2.40, often installed with setgid
tty permissions, allows escape sequences to be sent to other
users' terminals through argv. (Specifically, escape sequences
received from stdin are blocked, but escape sequences received
from argv are not blocked.) There may be plausible scenarios
where this leads to account takeover.
References:
https://nvd.nist.gov/vuln/detail/CVE-2024-28085
(From OE-Core rev: b40a77416f73955833faeddf6091a99ff9837199)
Signed-off-by: Soumya Sambu <soumya.sambu@windriver.com>
Signed-off-by: Steve Sakoman <steve@sakoman.com>
This avoids problems if BSD-4-Clause is in INCOMPATIBLE_LICENSE since
util-linux-fcntl-lock is now a dependency of run-postinsts.
(From OE-Core rev: c2c7e1624e8080fb8fa1f14d252f20e85b232eb7)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a version of flock that uses the fnctl based lockf locking instead of
flock based locks. This allows us to take the same lock that opkg would
use from a shell script. The two different locking mechanisms operate
independently of each other.
Inserting this C file into the util-linux build seems like the easiest/best
place to insert the code. At this point it hasn't been discussed with upstream.
(From OE-Core rev: d2b784110e2c3df8a0a41e4819cf2de9003f9fa3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The MIT license was missing from the license list for util-linux. Add
a patch, submitted to upstream which adds the missing license mentions.
(From OE-Core rev: 24078c0542626f74c93203d17ab4b2bb2f9b5630)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This utility conflicts with script from busybox with rich
configuration.
(From OE-Core rev: c30437939e0af6fa243cbe6a305491119726f1a1)
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Now that we've solved the util-linux - gtk-doc - xmlto - util-linux
dependency loop, we can enable gtk-doc in util-linux.
We explicitly disable it in util-linux-libuuuid as this doesn't have API
documentation.
(From OE-Core rev: 6999719143df04816c74fbc46d4a97b377bc0b36)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
License-Update:
===============
GPL-1.0-or-later is added. fdisk.c is licensed under it.
0001-lscpu-Use-4K-buffer-size-instead-of-BUFSIZ.patch is dropped as
it's been in the new version.
(From OE-Core rev: 9a0bd46e6988b2da446f89864131639dde63c6d0)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These tests are marked as known failures upstream as well [1]
if we delete them then we can drop dependency on mdadm for ptests which
can further reduce the dependency chain for ptest enabled minimal images
[1] 7519c3edab
(From OE-Core rev: 4f3b4b4f27924f9093dc6b4dd949cabe5cd2891b)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This was needed for uclibc, which is long gone in OE
This reverts commit d3ffa39e2e.
(From OE-Core rev: fdf81adb8fc705f3c989244ca93067da7db88311)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These tests need strptime with %s which is glibc specific extention
(From OE-Core rev: 0190b3eb67d02ecf42531a7c087a5a6f75ee7a5e)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When enabling ipcs and ipcrm configuration into busybox, both tools are
built and then deployed during do_rootfs. These operation lead to below
issue (similar behavior happens for ipcs):
do_rootfs: Postinstall scriptlets of ['busybox'] have failed. If the intention is to defer them to first boot,
then please place them into pkg_postinst_ontarget:${PN} ().
update-alternatives: Error: not linking .../build/tmp/work/board-poky-linux/board-image/1.0-r0/rootfs/usr/bin/ipcrm
to /bin/busybox since .../build/tmp/work/board-poky-linux/board-image/1.0-r0/rootfs/usr/bin/ipcrm exists and is not a link
Binaries enter in conflict with same named util-linux utilities during
do_rootfs step.
Adding ALTERNATIVE_LINK_NAME for both tools fix the issue.
(From OE-Core rev: e4d60408b869c9cc2ccff794d4e271d993ec8a97)
Signed-off-by: Benjamin Bouvier <benjamin.bouvier@ekinops.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Check whether magic.h header exists before defining HAVE_MAGIC.
Despite library availability there still can be missing header.
Current test doesn't cover that possibility which will lead compilation
to fail in case of separate sysroot.
Upstream-Status: Backport
[263381ddd4]
(From OE-Core rev: a93e5c4aba34996fdb9b253a4a8529f33ec114f5)
Signed-off-by: Mateusz Marciniec <mateuszmar2@gmail.com>
Signed-off-by: Tomasz Dziendzielski <tomasz.dziendzielski@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
glibc 2.36 has added sys/pidfd.h and APIs for
pidfd_send_signal and pidfd_open, therefore check
for this header and include it if it exists
(From OE-Core rev: c72d4ad4d3c41e6bc36e7d7542a42db04b5a2e97)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
busybox also installs findfs but into base_sbindir which works out to be
ok when sbindir != base_sbindir but with usrmerge distro feature enabled
this starts to cause trouble because busybox's postinst is trying to
create a symlink for findfs applet in base_sbindir which is same as
sbindir now and there already is binary from util-linux and image fails
to build
do_rootfs: Postinstall scriptlets of ['busybox'] have failed. If the intention is to defer them to first boot,
then please place them into pkg_postinst_ontarget:${PN} ().
The real reason is burried in do_rootfs logs
update-alternatives: Error: not linking /mnt/b/yoe/master/build/tmp/work/beaglebone-yoe-linux-gnueabi/yoe-sdk-image/1.0-r0/rootfs/usr/sbin/findfs to /usr/bin/busybox.suid since /mnt/b/yoe/master/build/tmp/work/beaglebone-yoe-linux-gnueabi/yoe-sdk-image/1.0-r0/rootfs/usr/sbin/findfs exists and is not a link
Creating proper u-a for findfs in util-linux fixes the issue
(From OE-Core rev: 211ae2db1ab8fec1ed678170f9d8cbca2cc27ef3)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
An automated conversion using scripts/contrib/convert-spdx-licenses.py to
convert to use the standard SPDX license identifiers. Two recipes in meta-selftest
were not converted as they're that way specifically for testing. A change in
linux-firmware was also skipped and may need a more manual tweak.
(From OE-Core rev: ceda3238cdbf1beb216ae9ddb242470d5dfc25e0)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Sort the list of files to ensure the pkgdata output is deterministic.
(From OE-Core rev: 82e683f8f9ae630dea46ec6be6e636e498579835)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
raw.h has been dropped in linux-libc-headers-5.14 leading to:
configure: error: raw selected, but required raw.h header file not available
WARNING: exit code 1 from a shell command.
(From OE-Core rev: 7f577c10913104860121f682b9b3754870c4db23)
Signed-off-by: MarkusVolk <f_l_k@t-online.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There was a time that util-linux requres asciidoctor to be there
to have man pages. However, now the tarball ships generated man
pages and will use them. So add back the related settings. Otherwise,
we get the following error when enabling 'doc-pkgs' image feature.
+ update-alternatives --install /usr/share/man/man1/kill.1 kill.1 /usr/share/man/man1/kill.1.procps 200
update-alternatives: Error: not linking /.../tmp/work/qemux86_64-poky-linux/core-image-minimal/1.0-r0/rootfs/usr/share/man/man1/kill.1 \
to /usr/share/man/man1/kill.1.procps since /.../tmp/work/qemux86_64-poky-linux/core-image-minimal/1.0-r0/rootfs/usr/share/man/man1/kill.1 exists and is not a link
(From OE-Core rev: 3d66fcc425495b5b68caf1a63c7118c692236cc4)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is the result of automated script conversion:
scripts/contrib/convert-overrides.py <oe-core directory>
converting the metadata to use ":" as the override character instead of "_".
(From OE-Core rev: 42344347be29f0997cc2f7636d9603b1fe1875ae)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
close_range expects 3 parameters, thus far it was not implemented in
glibc so it was hidden but with glibc 2.34+ it will start to show up
(From OE-Core rev: 0d6f638230a96808e22438d9925aec6383d136a7)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
They are also provided by shadow-native e.g. when building native
recipes and packages where they depend on both shadow-native and
util-linux-native, this can conflict
Enable it only when pam is enabled in distro
Fixes
ERROR: systemd-1_248.3-r0 do_prepare_recipe_sysroot: The file /usr/bin/chsh is installed by both util-linux-native and shadow-native, aborting
(From OE-Core rev: 336625fe433100fb57ea0dda4d9ff5fcb835dc95)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Cc: Ross Burton <ross.burton@arm.com>
Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Instead of duplicating logic, we can use the run.sh that the test suite
installs.
Unless told otherwise, the util-linux test suite assumes that it is
running in a build tree and looks for binaries to run, but we're on
target so pass --use-system-commands (drops ~180 skips to tens).
Add --show-diff so that debugging can be done from the logs alone.
Remove redundant path manipulation that is now done upstream.
If PAM is disabled, delete the chfn test: it will run the chfn from
shadow which has different output, and the test will fail.
[ YOCTO #14244 ]
(From OE-Core rev: 720660b936b5523ec48a7c04712d10c88198e12c)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Backport a number of fixes so the test suite passes.
(From OE-Core rev: 398b9dab8675829f3428a4f00bffcf84525e0a46)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The kernel modules loop and sd-mod are needed for the block device tests.
GNU find is used in some tests.
(From OE-Core rev: 9350c391f11bf7587fe46a54c93dc6e4bf3b51db)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As we now split util-linux per-binary there's no disadvantage to not
building chfn and chsh.
This also means that the tests pass instead of failing when the test
uses the binaries from shadow.
(From OE-Core rev: 804c6b5bd3d398d5ea2a45d6bcc23c76e328ea3f)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop manpages support, as it requires asciidoctor
(written in ruby, isn't in oe-core).
(From OE-Core rev: 6cb58dd4feae12a064fad3ab30b1d44e159fe7e2)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Changing BPN to be "util-linux" regardless if it is the util-linux
recipe or the util-linux-libuuid recipe that is being built was an
easy way to allow ${BPN} and ${BP} to be used in the SRC_URI for both
recipes. However, it causes problems for native.bbclass where there
are expectations that ${BPN} and ${PN} match each other.
Use "util-linux" directly in the SRC_URI instead to avoid the problem.
(From OE-Core rev: 701ef7ff0e1b17150424a64acfafbb83f518f9c9)
Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rename the recipe from util-linux-uuid to util-linux-libuuid which means
we can drop the custom PACKAGES and FILES defintions which simplifies
things. Also move the LICENSE setting to the libuuid recipe so that
it is correctly applied to the right packages.
This means the standard definitions from bitbake.conf are used, avoiding
errors from situations where users have customised settings causing
failures.
(From OE-Core rev: 65efd76198ad805060fe28714765cd423fa748dc)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
util-linux-uuid does not need libtcolors.a and libcommon.a but
they get build regardless. Backport a patch from upstream to
skip them and save some compilation time.
c65953d72b
(From OE-Core rev: 9353c4338acd51e1bc4172554fca018c4044829f)
Signed-off-by: Luca Boccassi <luca.boccassi@microsoft.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Recently util-linux gained an (optional) build dependency on libcryptsetup.
But libcryptsetup build-depends on util-linux for blkid (optional, can be disabled)
and uuid (mandatory).
Split out util-linux-uuid in a different recipe to break the cycle.
https://github.com/karelzak/util-linux/pull/898
(From OE-Core rev: 5f204796e73f37ee67d0a86512ce3ab6f19f9ed0)
Signed-off-by: Luca Boccassi <luca.boccassi@microsoft.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
0001-build-sys-do-not-build-plymouth-ctrl.c-w-disabled-pl.patch
0001-hwclock-do-not-assume-__NR_settimeofday_time32.patch
removed since they are included in 2.36.2
(From OE-Core rev: 4db0babf6a69a90df14363103f5f6b643f208b12)
Signed-off-by: Wang Mingyu <wangmy@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
