Since f7725647bb41c3398a867f139efe526efe8aa1b3 when dissecting a disk
image we operate with fds to the device nodes in question wherever we
can. This includes when we fork off fsck, where we pass a /proc/self/fd/
path as argument. This only works if we keep that fd open however and
disable O_CLOEXEC on the fd. Hence do so, and fix fsck this way.
(Without this, all fsck will fail, since the fd path is invalid)
process-util: add new FORK_CLOEXEC_OFF flag for disabling O_CLOEXEC on remaining fds
Often the fds that shall stay around in the child shall be passed
to a process over execve(), hence add an option to explicitly disable
O_CLOEXEC on them in the child.
Luca Boccassi [Thu, 3 Nov 2022 20:10:57 +0000 (20:10 +0000)]
tmpfiles.d: do not fail if provision.conf fails
On a read-only filesystem creating /root/.ssh might fail, but that's ok.
Do not fail the run, as this is only needed to add the credential, which
is a separate step.
Franck Bui [Fri, 4 Nov 2022 11:24:10 +0000 (12:24 +0100)]
tests: make test-execute pass on openSUSE
In my understanding user group "3" (aka "sys") is kept for historical reasons
but not really useful these days. That's probably explained why this group
isn't defined on openSUSE.
Hence let's drop reference to this user group, this shouldn't lessen the
revelance of the test since SupplementaryGroups= is still tested with 2 other
groups.
Yu Watanabe [Thu, 3 Nov 2022 00:39:36 +0000 (09:39 +0900)]
udev: first set properties based on usb subsystem
After 479da1107a0d4e2f7ef5cd938512b87a0e45f180, the usb_id builtin
command does not set ID_SERIAL if ID_BUS is already set.
Before the commit, all properties set based on pci bus were overwritten
by the usb_id, hence now it is sufficient setting them only when ID_BUS is
not set yet.
Samuel Thibault [Tue, 1 Nov 2022 15:53:02 +0000 (16:53 +0100)]
shutdown: Add Xen kexec support
In the Xen case, it's the hypervisor which manages kexec. We thus
have to ask it whether a kernel is loaded, instead of relying on
/sys/kernel/kexec_loaded.
Also, fix a race condition introduced by d16684fe13:
```
[ 16.904218] H testsuite-26.sh[394]: + systemd-run --unit failed.service /bin/false
[ 16.964783] H systemd[845]: failed.service: Executing: /bin/false
[ 16.965062] H systemd[1]: Started failed.service.
[ 16.965462] H testsuite-26.sh[844]: Running as unit: failed.service
[ 16.966390] H testsuite-26.sh[394]: + systemctl is-failed failed.service
[ 16.977970] H testsuite-26.sh[846]: active
[ 16.978403] H systemd[1]: failed.service: Main process exited, code=exited, status=1/FAILURE
[ 16.978478] H systemd[1]: failed.service: Failed with result 'exit-code'.
```
Yu Watanabe [Tue, 1 Nov 2022 21:49:26 +0000 (06:49 +0900)]
test-network: wait for bound interface to be processed by udevd
If another bound interface (dummy98) will be removed before that dummy99
is processed by udevd, then removing dummy98 in the next step makes the
target interface (test1) bring down.
Jörg Thalheim [Thu, 27 Oct 2022 13:59:56 +0000 (15:59 +0200)]
update credentials when reloading a service
Many long-running services will reload configuration and therefore
credentials on reload. Therefore its useful if a service reload will
also update credentials.
Tested in https://github.com/numtide/systemd-vaultd/pull/12
Daan De Meyer [Thu, 27 Oct 2022 11:14:12 +0000 (13:14 +0200)]
namespace: Add hidepid/subset support check
Using fsopen()/fsconfig(), we can check if hidepid/subset are supported to
avoid the noisy logs from the kernel if they aren't supported. This works
on centos/redhat 8 as well since they've backported fsopen()/fsconfig().
Daan De Meyer [Thu, 27 Oct 2022 09:12:10 +0000 (11:12 +0200)]
condition: Check that subsystem is enabled in ConditionSecurity=tpm2
Instead of succeeding when either the firmware reports a TPM device
or we find a TPM device, let's check that the firmware reports a TPM
device and the TPM subsystem is enabled in the kernel.
To check whether the subsystem enabled, we check if the relevant
subdirectory in /sys exists at all.
Repart is growing into an important tool on its own, and users might
want to install newer versions on systems that have older systemd. Let's
make this easier by providing a standalone binary.
Related to https://github.com/systemd/mkosi/issues/1228.
Marek Vasut [Tue, 25 Oct 2022 10:50:09 +0000 (12:50 +0200)]
udev: Handle AMBA bus the same way as generic platform bus
On various ARM systems, it is possible to find devices on AMBA bus.
This is defined in Linux kernel drivers/amba/bus.c as separate bus
type. Udev currently does not recognize this when building ID_PATH
which leads to ID_PATH containing the suffix from next recognized
parent device. On ST STM32MP15xx with SDIO WiFi, the ID_PATH looks
like 'platform-soc', which is not unique and basically useless. On
NXP i.MX8M Plus with SDIO WiFi, ID_PATH is 'platform-30b40000.mmc'
which is far more useful.
Add the 'amba' subsystem handling the same way 'platform' subsystem
is handled to get ID_PATH 'platform-soc-amba-48004000.sdmmc' on the
former, which is far more useful compared to 'platform-soc'.
---
```
$ udevadm info --attribute-walk --path=/devices/platform/soc/48004000.mmc/mmc_host/mmc1/mmc1:0001/mmc1:0001:1/net/wlan0
Udevadm info starts with the device specified by the devpath and then
walks up the chain of parent devices. It prints for every device
found, all possible attributes in the udev rules key format.
A rule to match, can be composed by the attributes of the device
and the attributes from one single parent device.
Charles Hardin [Mon, 10 Oct 2022 20:30:10 +0000 (13:30 -0700)]
udev-builtin-net_id: support getting usb path off the host
To support predictable interface names in various embeeded systems
add support for an additional naming scheming using the USB host
interface. Several asics have usb controllers that are platform
devices and not children of a pci interface. These embedded systems
should be able to enumerate interfaces by udev path as well to support
configurations and policies.
Signed-off-by: Charles Hardin <charles.hardin@chargepoint.com>
Charles Lee [Mon, 31 Oct 2022 11:19:54 +0000 (12:19 +0100)]
po: Translated using Weblate (Chinese (Simplified) (zh_CN))
Currently translated at 100.0% (193 of 193 strings)
Co-authored-by: Charles Lee <lchopn@gmail.com>
Translate-URL: https://translate.fedoraproject.org/projects/systemd/master/zh_CN/
Translation: systemd/main
Takashi Sakamoto [Mon, 31 Oct 2022 03:48:31 +0000 (12:48 +0900)]
hwdb: ieee1394-unit-function: add comment about legacy AV/C devices
It appears that exceptional layout of legacy device requires extra care of
hwdb entry for node device since Linux FireWire subsystem do not pick up
numeric model identifier in vendor directory. In detail, see:
* https://github.com/systemd/systemd/issues/25029
In the case, udev rule without model attribute is used. Thus hwdb entry
for generic AV/C device should match both cases with and without the
attribute. The wildcard added by a commit 5e577da5f824 ("hwdb: drop model
specifier from general entries") satisfies this condition,
Takashi Sakamoto [Mon, 31 Oct 2022 03:48:31 +0000 (12:48 +0900)]
hwdb: ieee1394-unit-function: fullfill for Sony DCR-TRV310
It appeared that Sony DCR-TRV310 has legacy layout of configuration ROM
against 1394 TA standard documentation.
* https://github.com/systemd/systemd/issues/25029
For the case, numeric model identifier and descriptor leaf for model name
are not picked up. This commit fulfill corresponding entry so that
applications can use model name from hardware database.
Yu Watanabe [Wed, 26 Oct 2022 02:08:17 +0000 (11:08 +0900)]
network: try to reconfigure when some information is updated
When at least one of the name, MAC address, udev properties, and so on
for an interface is updated, try to find a matching .network file, and
reconfigure if a new .network file is assigned.
projects / thirdparty / systemd.git / log
