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>
We need to prevent systemd-network from managing veth interface when
using platforms like k3s as they control creation and management of the
interfaces.
(From OE-Core rev: b615bb279730ce3e0c4b4098a9e43700f2f095af)
Signed-off-by: Matt Spencer <matt.spencer@arm.com>
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When systemctl fail it would throw an exception and
dump a traceback. Lets use a more controlled workflow.
[Yocto #14395]
(From OE-Core rev: df510ae9a1494bc1be8d6673fbaa43d3f7cc8f40)
Signed-off-by: Armin Kuster <akuster808@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This is not needed as systemd-conf only packages
standalone config files.
(From OE-Core rev: 02bebaaf927e2cf5326bcae1de10cd4a82fd9495)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Drop backports:
48fff0a2af3f62acd446ebec8081b039b72caad8.patch
7c5fd25119a495009ea62f79e5daec34cc464628.patch
0001-analyze-resolve-executable-path-if-it-is-relative.patch
0001-analyze-resolve-executable-path-if-it-is-relative.patch
Drop 0027-proc-dont-trigger-mount-error-with-invalid-options-o.patch
as difficult to rebase; please resend if you feel strongly that
it is needed.
Rebase the big pile of musl patches.
(From OE-Core rev: e8dd5a36bf2f1e645fb2ff15eb3b5e97c04776e6)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
refresh the following patches:
systemd/0001-systemd.pc.in-use-ROOTPREFIX-without-suffixed-slash.patch
systemd/0006-Include-netinet-if_ether.h.patch
(From OE-Core rev: 5db15aec00517e5b6f6585b534a2314c226841bf)
Signed-off-by: Wang Mingyu <wangmy@cn.fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In cases where we configure the IP address and more on the kernel
command line with ip= we should not ask for DHCP with systemd-networkd
later on. We have such a setup with our runqemu script.
With this match in place we can also deploy this unit on qemu systems.
(From OE-Core rev: 8c4981e740c0e630200dbc77c9e3dfd3e43b790b)
Signed-off-by: Stefan Schmidt <stefan.schmidt@huawei.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The systemd-nspawn@.service file uses the following line to try
loading the tun, loop and dm-mod kernel modules, in a non fatal way:
ExecStartPre=-/sbin/modprobe -abq tun loop dm-mod
A non-zero return code from modprobe will not stop the service from
starting, but it would be convenient to inform the user about this
optional support, so add these modules to RRECOMMENDS.
(From OE-Core rev: 9644285dc8a7e7c52b5f7554ef06d9ff53167832)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If systemd is configured with importd, "machinectl import-*" can be
used to add container trees under /var/lib/machines.
But "machinectl import-tar" makes use of "tar --numeric-owner" which
is not supported by the busybox binary, as shown below:
# machinectl import-tar /tmp/container.tar.xz
Enqueued transfer job 1. Press C-c to continue download in background.
Importing '/tmp/container.tar.xz', saving as 'container'.
Imported 0%.
tar: unrecognized option '--numeric-owner'
BusyBox v1.31.1 () multi-call binary.
Usage: tar c|x|t [-ZzJjahmvokO] [-f TARFILE] [-C DIR] [-T FILE] [-X FILE] [FILE]...
Failed to decode and write: Broken pipe
Exiting.
To fix this, recommend the standard tar package.
Cc: Chen Qi <Qi.Chen@windriver.com>
(From OE-Core rev: 801658f0161270c76829a123a4f531ca5a1f73db)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Sort packages in RRECOMMENDS_${PN}-container alphabetically to avoid
subjective sorting and ease future additions.
(From OE-Core rev: 636c048b9d49b0f0bf083add688b4de80f3e37cd)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The comment above the importd config states curl, which does not
exist in PACKAGECONFIG. Adding it results in the following error
(and then misleading compilation failures):
WARNING: systemd-1_244.5-r0 do_configure: QA Issue: systemd: invalid PACKAGECONFIG: curl [invalid-packageconfig]
Support for curl is currently enabled through journal-upload:
PACKAGECONFIG[journal-upload] = "-Dlibcurl=true,-Dlibcurl=false,curl"
While it might be more appropriate to deprecate "journal-upload"
in favor of a "curl" PACKAGECONFIG entry, simply fix the importd
comment for now by stating upload-journal instead of curl.
(From OE-Core rev: e6a0bced605da58ff43ff06ea6a1d67e84c98d91)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The licenses were renamed to match their SPDX names, fix the
references in LIC_FILES_CHKSUM in OE-Core.
(From OE-Core rev: 9af48917cfe583d2db9e1e088c7e396fcf638949)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When polkit is not available, networkd will not have permissions
to call hostnamed's dbus methods, as it runs without privileges.
To solve this, when building without polkit, make a new PACKAGECONFIG
'polkit_hostnamed_fallback' available which changes hostnamed so that
it runs as the 'systemd-network' user, the same as networkd, so that
the authorization works (and also with CAP_SYS_ADMIN since it loses
root).
Also run it with a separate 'systemd-hostname' group which also owns
the bus, to avoid giving the 'systemd-network' additional privileges.
(From OE-Core rev: 07a8f1629662d4d66d056d95be3cb9e749242274)
Signed-off-by: Luca Boccassi <luca.boccassi@microsoft.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These two header fixes are not needed with the current source as the
headers are there through other header files. Drop the patches.
(From OE-Core rev: ece3efe43b6b9de5426c23f29db15dd709bcd749)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As mentioned during review, this patch can be less invasive and hence easier
to maintain. Improve as such.
(From OE-Core rev: 3a1bd768e62a493aa73e82f5c443ca28b108af51)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
include "missing_stdlib.h" is needed for strndupa()
(From OE-Core rev: 87c9ed35fce8c9358d8a5dda20ece0a46cbff325)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Backport a patch to fix the below systemd-analyze issue:
# systemd-analyze --man=false verify /lib/systemd/system/initrd-cleanup.service
initrd-cleanup.service: Command systemctl is not executable: No such file or directory
# which systemctl
/bin/systemctl
# cat /lib/systemd/system/initrd-cleanup.service
[snip]
[Unit]
Description=Cleaning Up and Shutting Down Daemons
DefaultDependencies=no
AssertPathExists=/etc/initrd-release
OnFailure=emergency.target
OnFailureJobMode=replace-irreversibly
After=initrd-root-fs.target initrd-fs.target initrd.target
[Service]
Type=oneshot
ExecStart=systemctl --no-block isolate initrd-switch-root.target
(From OE-Core rev: 2e0259931f71701147039bb8e60251892f67dbcd)
Signed-off-by: Mingli Yu <mingli.yu@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Recent systemd started using ascii args to "hidepid=" mount options
for proc fs - unconditionally -- even though kernels older than v5.8
emit an error message on each attempt:
root@qemux86-64:~# cat /proc/version
Linux version 5.4.87-yocto-standard (oe-user@oe-host) (gcc version 10.2.0 (GCC)) #1 SMP PREEMPT Fri Jan 8 01:47:13 UTC 2021
root@qemux86-64:~# dmesg|grep proc:
[ 29.487995] proc: Bad value for 'hidepid'
[ 43.170571] proc: Bad value for 'hidepid'
[ 44.175615] proc: Bad value for 'hidepid'
[ 46.213300] proc: Bad value for 'hidepid'
root@qemux86-64:~#
Simply ignoring them as the systemd maintainer unconditionally says
is the resolution is clearly not acceptable, given the above.
Add a kernel version check to avoid calling mount with invalid args.
Further details are within the enclosed systemd commit.
Cc: Luca Boccassi <luca.boccassi@microsoft.com>
Cc: Richard Purdie <richard.purdie@linuxfoundation.org>
(From OE-Core rev: 76107fd7372559aa4cd22a89d5517a4dfce9314d)
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meson.build will fall back to greping /etc/login.defs for values of these
if they're not set. Different distros set them (Centos 7/8 does, Ubuntu
does not) so output was not deterministic. Avoid this by setting to the
default values. We now match the vaules from login.defs from shadow.
(From OE-Core rev: 77a6ac0ac266d71e4fe67fd332662081f30cd7bf)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some libraries are now dlopen'ed and skipped at runtime if
not available. Add them to the RRECOMMENDS list, to allow
users to exclude them for smaller images even if generally
enabled at build time.
(From OE-Core rev: 0131a794be1a97f9aef3224c7c66fcf27576ce84)
Signed-off-by: Luca Boccassi <luca.boccassi@microsoft.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Update systemd to v247.2.
Add rule for new oomd dbus conf and for new pam.d
conf directory in /usr/lib|lib64.
Drop selinux-hook-handling-to-enumerate-nexthop.patch,
merged upstream.
Drop 0001-meson-Fix-reallocarray-check.patch,
merged upstream.
Refresh musl patches.
(From OE-Core rev: 1e1d26de68ed13fd53c1a16b9662ac9860dca714)
Signed-off-by: Luca Boccassi <luca.boccassi@microsoft.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
systemd-boot works on 32-bit and 64-bit Arm targets (assuming the
firmware does EFI, obviously), so allow it to be built.
(From OE-Core rev: 3325992e66e8fbd80292beb4b0ffd50beca138d8)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Also build and deploy the EFI stub. SYSTEMD_BOOT_EFI_ARCH can be dropped
as image-uefi.conf now sets EFI_ARCH.
Changes originally taken from meta-intel.
(From OE-Core rev: 00191504bd9321334a5c01b95179ed15b063f1b6)
Signed-off-by: Ross Burton <ross.burton@arm.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In systemd.bbclass, it will replace criterion command 'type systemctl'
with 'systemctl' without any argument to judge whether command systemctl
is available. The change is to fix install/remove package error in
container when command systemctl exists but not callable.
Make native systemctl wrapper prints help mesages if called without any
argument to follow the update.
(From OE-Core rev: 811ebb4adc5cfcc24ac4460fe89fe68d8d03d837)
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If we say we're enabling DHCP on wired/ethernet networks lets be more
specific than trying to catch everything with globbing.
(From OE-Core rev: a2bf422cc498bebfbeedc9382b2cc1533f78bbfd)
Signed-off-by: Jack Mitchell <ml@embed.me.uk>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Allow distros which include other network managers to disable the
auto DHCP setup of interfaces in systemd-networkd.
(From OE-Core rev: ea5e65b578498df7976bbc560bdcd39c2ebcc1d9)
Signed-off-by: Jack Mitchell <ml@embed.me.uk>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
systemd 245 introduced a regression in behavior where they removed
support for non-root users to chvt from a service file. This prevents
running compositors (e.g. weston) as any user other than root. The
intention is for polkit to be used to allow this (and in fact the
default polkit rules that ship with systemd allow this). However, polkit
is a huge dependency to bring in for an embedded system, and isn't
support by OE-core.
The patch has been proposed upstream to restore the previous behavior of
allowing a non-root user to chvt to unbreak the regression without
requiring polkit.
Upstream-Status: Submitted [https://github.com/systemd/systemd/pull/17494]
(From OE-Core rev: ef8cf581b680f8b8ee35d50c75891b4fd8b91f2f)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Per default systemd use a pager for the output of most of its tools
and it expect this pager to be color capable. But that is not the case
when the busybox `less` is used, which lead to output garbled by color
escape sequences.
To fix this issue add a profile frament that disable the systemd pager
when busybox `less` is detected.
(From OE-Core rev: de7e36a7858ebca4615975967fcad1c399eacdb0)
Signed-off-by: Alban Bedel <alban.bedel@aerq.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When selinux is enabled, the call of
manager_rtnl_enumerate_nexthop() fails.
This fix is to facilitate selinux hook handling
for enumerating nexthop.
In manager_rtnl_enumerate_nexthop() there is a check
if "Not supported" is returned by the send_netlink() call.
This check expects that -EOPNOTSUPP is returned,
the selinux hook seems to return -EINVAL instead.
This happens in kernel older than 5.3
(more specificallytorvalds/linux@65ee00a) as it does not support
nexthop handling through netlink.
And if SELinux is enforced in the order kernel, callingRTM_GETNEXTHOP
returns -EINVAL.
Thus adding a call in the manager_rtnl_enumerate_nexthop for the
extra return -EINVAL.
Note: systemd version is different in yocto project (v246.6) and
systemd master(v247) and In systemd verison(246.6)
mac_selinux_enforcing() function is not declared and defined.
(From OE-Core rev: c521df3f0a86521ce78fd5c057293b2c242fba79)
Signed-off-by: Purushottam choudhary <purushottam.choudhary@kpit.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Long ago in commit 473ff65c2f69de4ece3204fadfae7c5cb992149a
(serial-getty service: Add xterm as default TERM), the xterm
became the default for the serial port terminal.
Using the version of vim.tiny in oe-core master with the
serial port connected in xterm version 322 (which is one
of the most widely deployed versions at the current time)
causes artifacts and missed characters.
The example sequence is the following:
* Start vim
* Press "i" to enter input mode
* Type "123"
* Press Escape to enter command mode
* Press "a" to enter append mode
* Type "456"
At this point if you are using xterm less than version 535 you will
see on your screen "12456" instead of "123456".
Changing the TERM variable to "linux" will still allow you to have all
the same functionality with colorization, ansi character escapes
etc..., but will avoid the extra xterm specific escape sequence that
only exists in the most recent versions of xterm.
This patch allows the end user to set the serial terminal type to
something other than the new default of "linux" by changing the
SERIAL_TERM variable in local.conf. For example:
SERIAL_TERM = "xterm"
(From OE-Core rev: 3de72f45135d7c79b43ad25b539f117d040aa246)
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As systemd-boot shares SRCREV with the main systemd recipe, the version
should be kept in sync with it.
(From OE-Core rev: 16ff7f3c9d38898b936340c28fd6820c036c62d5)
Signed-off-by: Alexander Kanavin <alex.kanavin@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch is no longer needed as upstream has added including part of
this include list, perhaps thats all is what was needed to make it
portable
Refresh 0001-Handle-missing-gshadow.patch
(From OE-Core rev: 0f5d9d8cc5d05a7a485e4cb89b37d06ea0d05d23)
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
