Re: [systemd-devel] Help with rules to configure internal USB network to BMC as link-local

23.11.2025 00:32, Rebecca Cran wrote: I'm trying to fix an issue on Ampere-based systems such as those with Altra and AmpereOne CPUs. When running a desktop such as GNOME in Ubuntu 25.10 there's a pop-up message every few minutes saying that it failed to configure the network interface. That's h

Re: [systemd-devel] Systemd-executor in 'u' state ?

Here is the lsfd output if lsof is not terse $ sudo lsfd -p 1615 12:59 [65/373] COMMANDPID USER ASSOC XMODETYPE SOURCE MNTID INODE NAME (sd-pam) 1615 bcvexe -- REG dm-1 0

Re: [systemd-devel] Systemd-executor in 'u' state ?

Hi I can confirm (atleast a few instances now) for sure that it is systemd-executor What I found was that systemd-executor runs as sd-pam and sd-pam is the one which is opening the root in rw This is what I get COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME (sd-pam) 1615

Re: [systemd-devel] Systemd-executor in 'u' state ?

Thanks. I am still investigating which process is opening / in RW mode. I will update you. Thanks again for lsfd. I will use this On Thu, Nov 20, 2025 at 10:13 PM Cristian Rodríguez wrote: > On Thu, Nov 20, 2025 at 12:26 PM Bhasker C V > wrote: > > > > is there a reason why systemd has opened >

Re: [systemd-devel] Systemd-executor in 'u' state ?

20.11.2025 20:04, Andrei Borzenkov wrote: 20.11.2025 18:26, Bhasker C V wrote: Hi all, I run my rootfs in 'ro' state and occasionally when i have to do an install I switch to remount,rw ... do stuff and then go back to remount,ro However, remount,ro many a times fails since someone has open

Re: [systemd-devel] Systemd-executor in 'u' state ?

20.11.2025 18:26, Bhasker C V wrote: Hi all, I run my rootfs in 'ro' state and occasionally when i have to do an install I switch to remount,rw ... do stuff and then go back to remount,ro However, remount,ro many a times fails since someone has opened the rootfs in "rw" state While investiga

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

Turns out it's good to '--verify' every once in a while. *journal* had one issue in my case: ... 31e6ff8: Invalid object: Bad message File corruption detected at ... Now cleaned and all *-F* units list as expected.

Re: [systemd-devel] Mapped device not becoming active

start. Thanks, Fredrik Hugosson From: Mantas Mikulėnas Sent: Monday, October 7, 2024 11:26 AM To: Lennart Poettering Cc: Fredrik Hugosson ; Fredrik Hugosson ; [email protected] Subject: Re: [systemd-devel] Mapped device not becoming active On Mon, Oct 7, 2024 at 11:23 AM Lenn

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

sorry, I assumed it'd have to be explicitly mentioned if/when non-root operates, otherwise I assumed that the default, was the *root* user. It's uid 0 doing and getting what I showed. thanks, L.

Re: [systemd-devel] bootloader=enabled just to get systemd-tpm2-setup?

On Fri, Nov 7, 2025 at 10:51 AM Lennart Poettering wrote: > On Fr, 07.11.25 10:34, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > Is it not about using grub at all? It's about building and supporting > > systemd features that accomodate tpm2 devices for other bootloaders

Re: [systemd-devel] bootloader=enabled just to get systemd-tpm2-setup?

On Fr, 07.11.25 10:51, Lennart Poettering ([email protected]) wrote: > As mentioned, pcrextend measures various things at boot into various > nvindexes we took possession off. i.e. [email protected], sorry, i meant: "… into various *pcrs* …" Lennart -- Lennart Poettering, Berlin

Re: [systemd-devel] bootloader=enabled just to get systemd-tpm2-setup?

On Fr, 07.11.25 10:34, Itxaka Serrano Garcia ([email protected]) wrote: > Is it not about using grub at all? It's about building and supporting > systemd features that accomodate tpm2 devices for other bootloaders that > conform to the specification? Like which one? The boot loade

Re: [systemd-devel] bootloader=enabled just to get systemd-tpm2-setup?

On Fri, Nov 7, 2025 at 10:24 AM Lennart Poettering wrote: > On Fr, 07.11.25 10:02, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > Hey folks, > > > > I'm having a go at building systemd myself and I can't get my head around > > this. > > > > If I disable the bootloader part

Re: [systemd-devel] bootloader=enabled just to get systemd-tpm2-setup?

On Fr, 07.11.25 10:02, Itxaka Serrano Garcia ([email protected]) wrote: > Hey folks, > > I'm having a go at building systemd myself and I can't get my head around > this. > > If I disable the bootloader part, because I don't want systemd-boot, I also > dont get nice services like sys

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

On Thu, Nov 6, 2025 at 10:45 AM lejeczek wrote: > > probably does not relate to my original question - but in case it does - I > get no output, yet command exits with 0 > > -> $ journalctl -F _SYSTEMD_UNIT > This shows only entries you can access which normally means user journal. Check that the

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

probably does not relate to my original question - but in case it does - I get no output, yet command exits with 0 -> $ journalctl -F _SYSTEMD_UNIT this is via an ssh session if it might matter. Everything else with regards to systemd/journald seems ok, logs, states, statuses, etc. same _no_ o

Re: [systemd-devel] Status update: NoNewPrivs/no setuid by default

On 11/5/25 03:51, Thorsten Kukuk wrote: > Hi, > > All Systems Go! is already long ago, so a status update where I'm are > using a Linux distribution with NoNewPrivs set by default. > > I renamed my "pwaccess" PoC to "account-utils" > (https://github.com/thkukuk/account-utils), which should contai

Re: [systemd-devel] Setting OOMScoreAdjust on scope unit

OOMScoreAdjust corresponds to a per-process parameter, not a cgroup parameter (it is set through /proc/PID/oom_score_adj and inherited by child processes). When creating a .service, systemd is always in charge of creating (forking/exec'ing) the process, but when creating a .scope, systemd only rec

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

On Fri, Oct 31, 2025 at 10:28:07AM +0100, lejeczek wrote: > Hi guys. > > I have my 'journalctl' to complete bash "tab-tab" taking minutes, like so: > -> $ journalctl -lf -o cat -u ... > I've not tried every args combinations but those I have seems to "misbehave" > the same - to complete unit name

Re: [systemd-devel] journalctl bash complete unit name extremely slow - why?

On Fri, Oct 31, 2025 at 12:28 PM lejeczek wrote: > > Hi guys. > > I have my 'journalctl' to complete bash "tab-tab" taking minutes, like so: > -> $ journalctl -lf -o cat -u ... > I've not tried every args combinations but those I have seems to "misbehave" > the same - to complete unit name takes

Re: [systemd-devel] Can RDP when NetwrokManager active; Can NOT RDP when systemd-networkd active

And (as this is plain text email) I forget to provide link where I describe more my issue. https://discussion.fedoraproject.org/t/can-rdp-when-netwrokmanager-active-can-not-rdp-when-systemd-networkd-active/163442 But if you need more details, or post something in this email thread - please let

Re: [systemd-devel] Can RDP when NetwrokManager active; Can NOT RDP when systemd-networkd active

> Try to stop the firewall. I tried # systemctl stop firewalld.service # firewall-cmd --state not running with no luck

Re: [systemd-devel] Can RDP when NetwrokManager active; Can NOT RDP when systemd-networkd active

On Tue, Oct 21, 2025 at 2:15 PM Tomasz Drożdż wrote: > > Hello all. > > I have an issue as I described here: > Can RDP when NetwrokManager active; Can NOT RDP when systemd-networkd active > Try to stop the firewall. > Can anyone help, assist ? > If You need more info please let me know. > If I s

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

To this original question back to me: any EFI and XBOOTLDR partitions used in this way must exist on the same block device as the intended root and usr partitions. Dracut does a fantastic job of including any tools needed to boot most any setup, including RAID devices. Most of my experience com

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On Mo, 20.10.25 16:29, Demi Marie Obenour ([email protected]) wrote: > > This is all anchored on the drive the firmware first boots from: > > systemd-boot searches for UKI on that drive, and then invokes the UKI > > from that drive. The UKI stub code than passes a reference to the > > drive to

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On Mo, 20.10.25 16:28, Demi Marie Obenour ([email protected]) wrote: > > LoaderDevicePartUUID reports the partition UUID of the ESP > > systemd-boot or systemd-stub first were invoked from. > > > > StubDevicePartUUID reports the partition UUID of the partition the UKI > > was invoked from (whi

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 16:13, Lennart Poettering wrote: > On Sa, 18.10.25 21:49, Demi Marie Obenour ([email protected]) wrote: > >> This isn't systemd-specific, but I know that at least some systemd >> developers recommend using UEFI secure boot + dm-verity, which leads >> to this problem. I also don't

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 16:21, Lennart Poettering wrote: > On Mo, 20.10.25 14:02, Demi Marie Obenour ([email protected]) wrote: > >> On 10/20/25 13:57, Andrei Borzenkov wrote: >>> 20.10.2025 20:33, Demi Marie Obenour wrote: On 10/19/25 11:36, Feli Flitzberg wrote: > Hi, long time watcher, first t

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On Mo, 20.10.25 15:45, Demi Marie Obenour ([email protected]) wrote: > >> Unfortunately this isn't what is needed. Looks like UEFI variable > >> storage is the best option. > > > > Partition UUID is (supposed to be) unique for every partition, so it > > most certainly allows one to identify t

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On Mo, 20.10.25 14:02, Demi Marie Obenour ([email protected]) wrote: > On 10/20/25 13:57, Andrei Borzenkov wrote: > > 20.10.2025 20:33, Demi Marie Obenour wrote: > >> On 10/19/25 11:36, Feli Flitzberg wrote: > >>> Hi, long time watcher, first time poster. If the bootloader supports the > >>>

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On Sa, 18.10.25 21:49, Demi Marie Obenour ([email protected]) wrote: > This isn't systemd-specific, but I know that at least some systemd > developers recommend using UEFI secure boot + dm-verity, which leads > to this problem. I also don't know a better place to ask for help > on this. > > H

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 15:10, Andrei Borzenkov wrote: > 20.10.2025 21:22, Demi Marie Obenour wrote: >> On 10/20/25 14:20, Andrei Borzenkov wrote: >>> 20.10.2025 21:02, Demi Marie Obenour wrote: On 10/20/25 13:57, Andrei Borzenkov wrote: > 20.10.2025 20:33, Demi Marie Obenour wrote: >> On 10/19/25

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

20.10.2025 21:22, Demi Marie Obenour wrote: On 10/20/25 14:20, Andrei Borzenkov wrote: 20.10.2025 21:02, Demi Marie Obenour wrote: On 10/20/25 13:57, Andrei Borzenkov wrote: 20.10.2025 20:33, Demi Marie Obenour wrote: On 10/19/25 11:36, Feli Flitzberg wrote: Hi, long time watcher, first time

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 14:20, Andrei Borzenkov wrote: > 20.10.2025 21:02, Demi Marie Obenour wrote: >> On 10/20/25 13:57, Andrei Borzenkov wrote: >>> 20.10.2025 20:33, Demi Marie Obenour wrote: On 10/19/25 11:36, Feli Flitzberg wrote: > Hi, long time watcher, first time poster. If the bootloader supp

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

20.10.2025 21:02, Demi Marie Obenour wrote: On 10/20/25 13:57, Andrei Borzenkov wrote: 20.10.2025 20:33, Demi Marie Obenour wrote: On 10/19/25 11:36, Feli Flitzberg wrote: Hi, long time watcher, first time poster. If the bootloader supports the Discoverable Partitions Specification, all that'

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 13:57, Andrei Borzenkov wrote: > 20.10.2025 20:33, Demi Marie Obenour wrote: >> On 10/19/25 11:36, Feli Flitzberg wrote: >>> Hi, long time watcher, first time poster. If the bootloader supports the >>> Discoverable Partitions Specification, all that's needed is the correct >>> partiti

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

20.10.2025 20:33, Demi Marie Obenour wrote: On 10/19/25 11:36, Feli Flitzberg wrote: Hi, long time watcher, first time poster. If the bootloader supports the Discoverable Partitions Specification, all that's needed is the correct partition GUID assigned to every partition. After that, you don'

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/19/25 11:36, Feli Flitzberg wrote: > Hi, long time watcher, first time poster. If the bootloader supports the > Discoverable Partitions Specification, all that's needed is the correct > partition GUID assigned to every partition. After that, you don't need to > pass any partitions or use /

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

On 10/20/25 05:57, killermoehre wrote: > Am Samstag, dem 18.10.2025 um 21:49 -0400 schrieb Demi Marie Obenour: > >> This isn't systemd-specific, but I know that at least some systemd >> developers recommend using UEFI secure boot + dm-verity, which leads >> to this problem. I also don't know a be

Re: [systemd-devel] With UEFI secure boot and dm-verity, how does one find the user data partition?

Am Samstag, dem 18.10.2025 um 21:49 -0400 schrieb Demi Marie Obenour: > This isn't systemd-specific, but I know that at least some systemd developers recommend using UEFI secure boot + dm-verity, which leads to this problem. I also don't know a better place to ask for help on this. > > How do OS

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

Also a small thing, there is nothing on the docs on how to select a profile via Type 1 entries, is that even possible? Does it go into the efi stanza like EFI_FILE@PROFILE_NUMBER or something? Cheers, Itxaka On Mon, Sep 22, 2025 at 4:35 PM Itxaka Serrano Garcia < [email protected]> w

Re: [systemd-devel] timedatectl set-ntp

Thanks, it became clear On Wed, Oct 1, 2025 at 11:10 AM Mantas Mikulėnas wrote: > On Wed, Oct 1, 2025 at 6:56 AM Alipour Alipour > wrote: > >> Hi, >> >> Is there a config file to control (enable/disable) the `timedatectl` NTP >> service status without using the `timedatectl set-ntp` command lin

Re: [systemd-devel] DHCPv6 configuration with NixOS

24.09.2025 19:29, Mark VanDyke wrote: >>> The following relevant portion in my working networking config. >> >> You did not show any of the systemd/networkd conf files. > > True. Let me start over. > > In the "old" situation I configure the system with "legacy scripts" as > they call it, which do

Re: [systemd-devel] DHCPv6 configuration with NixOS

> On 23 Sep 2025, at 21:47, Mark VanDyke wrote: > > The following relevant portion in my working networking config. > > You did not show any of the systemd/networkd conf files. What software are you using for dhcp? In my case my ISP provides a PPPoE interface that does a lot of the heavy

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Mon, Sep 22, 2025 at 4:18 PM Lennart Poettering wrote: > On Mo, 22.09.25 15:01, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > Hello folks, > > > > we are investigating a bit around multi-profile UKIs and cmdlines and > while > > this works as expected, we noticed that

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Tue, Sep 23, 2025 at 4:11 PM Lennart Poettering wrote: > On Di, 23.09.25 09:26, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > Also a small thing, there is nothing on the docs on how to select a > profile > > via Type 1 entries, is that even possible? Does it go into th

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On 9/24/25 00:38, Lennart Poettering wrote: > On Di, 23.09.25 18:26, Demi Marie Obenour ([email protected]) wrote: > >> On 9/23/25 17:56, Luca Boccassi wrote: >>> On Tue, 23 Sept 2025 at 22:45, Ian Pilcher wrote: This was discussed in this issue[1], but the issue was closed without

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On 9/23/25 17:56, Luca Boccassi wrote: > On Tue, 23 Sept 2025 at 22:45, Ian Pilcher wrote: >> This was discussed in this issue[1], but the issue was closed without >> any real resolution. (Giving a confined service access to everything >> labeled var_run_t is most definitely not acceptable.) > >

Re: [systemd-devel] timedatectl set-ntp

On Wed, Oct 1, 2025 at 6:56 AM Alipour Alipour wrote: > Hi, > > Is there a config file to control (enable/disable) the `timedatectl` NTP > service status without using the `timedatectl set-ntp` command line? > > I.e. where does `timedatectl` save its NTP service status when you use the > `timedat

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Tue, Sep 23, 2025 at 1:07 PM Lennart Poettering wrote: > On Mo, 22.09.25 16:35, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > > > Would be nice to know if this is a wanted feature on both types, or > type > > > 2 > > > > only or what. I think it would be nice to have o

Re: [systemd-devel] Non-idempotent RUN+= scripts

On Wed, Sep 24, 2025 at 8:27 PM Demi Marie Obenour wrote: > There are cases where a RUN+= script needs to do something > exactly once each time a device appears, such as binding a > different driver to the device. If the udev rule matches > based on a property (such as PCI device information) th

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On 9/23/25 11:29 PM, Lennart Poettering wrote: On Di, 23.09.25 15:44, Ian Pilcher ([email protected]) wrote: > (The symlinks under /tmp/keys are needed, because stunnel doesn't support any form of variable expansion in its configuration file.) Uff, the /tmp/keys thing is an instant DoS, I

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Di, 23.09.25 09:26, Itxaka Serrano Garcia ([email protected]) wrote: > Also a small thing, there is nothing on the docs on how to select a profile > via Type 1 entries, is that even possible? Does it go into the efi stanza > like EFI_FILE@PROFILE_NUMBER or something? You have to

Re: [systemd-devel] Systemd timers run repeatedly after time skip

On 10/10/25 20:18, Jeremy Herbert wrote: > Hi, > > I am not sure if this is a bug or just me not understanding the behaviour > correctly (the latter is more likely). Perhaps related to > https://github.com/systemd/systemd/issues/6036 > > I have a system that doesn't have access to time or an RTC

Re: [systemd-devel] DHCPv6 configuration with NixOS

> On 24 Sep 2025, at 19:38, Mark VanDyke wrote: > > Sorry, I don't understand what you mean? I get a /48 from my ISP and > want to use it to delegate to my LAN clients - this part works with > radvd and dhcpcd, but not if I try to get this done with systemd. > And, about the address... This is

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On 9/24/25 13:20, Ian Pilcher wrote: > On 9/23/25 11:29 PM, Lennart Poettering wrote: >> On Di, 23.09.25 15:44, Ian Pilcher ([email protected]) wrote: > > >>> (The symlinks under /tmp/keys are needed, because stunnel doesn't >>> support any form of variable expansion in its configuration file.)

Re: [systemd-devel] Non-idempotent RUN+= scripts

24.09.2025 23:46, Demi Marie Obenour wrote: On 9/24/25 13:46, Mantas Mikulėnas wrote: On Wed, Sep 24, 2025 at 8:27 PM Demi Marie Obenour wrote: There are cases where a RUN+= script needs to do something exactly once each time a device appears, such as binding a different driver to the device.

Re: [systemd-devel] Non-idempotent RUN+= scripts

On 9/24/25 13:46, Mantas Mikulėnas wrote: > On Wed, Sep 24, 2025 at 8:27 PM Demi Marie Obenour > wrote: > >> There are cases where a RUN+= script needs to do something >> exactly once each time a device appears, such as binding a >> different driver to the device. If the udev rule matches >> bas

Re: [systemd-devel] DHCPv6 configuration with NixOS

On 2025-09-24 20:08, Andrei Borzenkov wrote: > 24.09.2025 19:29, Mark VanDyke wrote: The following relevant portion in my working networking config. >>> >>> You did not show any of the systemd/networkd conf files. >> >> True. Let me start over. (..) >> interface wan >> nooption interface_mtu

Re: [systemd-devel] DHCPv6 configuration with NixOS

>> The following relevant portion in my working networking config. > > You did not show any of the systemd/networkd conf files. True. Let me start over. In the "old" situation I configure the system with "legacy scripts" as they call it, which do not use networkd beyond putting .link units in /

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On Di, 23.09.25 18:26, Demi Marie Obenour ([email protected]) wrote: > On 9/23/25 17:56, Luca Boccassi wrote: > > On Tue, 23 Sept 2025 at 22:45, Ian Pilcher wrote: > >> This was discussed in this issue[1], but the issue was closed without > >> any real resolution. (Giving a confined service

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On Di, 23.09.25 15:44, Ian Pilcher ([email protected]) wrote: > I recently discovered systemd credentials, and I was excited to start > using them to pass private keys (for TLS certificates) into my stunnel > service. > > LoadCredential=ampersand.key:/etc/pki/tls/private/stunnel/ampersand.key

Re: [systemd-devel] Credentials and SELinux - Am I missing something?

On Tue, 23 Sept 2025 at 22:45, Ian Pilcher wrote: > This was discussed in this issue[1], but the issue was closed without > any real resolution. (Giving a confined service access to everything > labeled var_run_t is most definitely not acceptable.) Sorry, but this is a self-imposed restriction t

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Tue, Sep 23, 2025 at 6:32 PM Lennart Poettering wrote: > On Di, 23.09.25 14:46, Itxaka Serrano Garcia ( > [email protected]) wrote: > > > > > In our case it is due to needing to differentiate the "state", like > we > > > have > > > > an active/passive/recovery but the actual conte

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Di, 23.09.25 14:46, Itxaka Serrano Garcia ([email protected]) wrote: > > > In our case it is due to needing to differentiate the "state", like we > > have > > > an active/passive/recovery but the actual content of the efi files are > > the > > > same (at least initially) so we ide

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Di, 23.09.25 16:25, Itxaka Serrano Garcia ([email protected]) wrote: > On Tue, Sep 23, 2025 at 4:11 PM Lennart Poettering > wrote: > > > On Di, 23.09.25 09:26, Itxaka Serrano Garcia ( > > [email protected]) wrote: > > > > > Also a small thing, there is nothing on the

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Mo, 22.09.25 16:35, Itxaka Serrano Garcia ([email protected]) wrote: > > > Would be nice to know if this is a wanted feature on both types, or type > > 2 > > > only or what. I think it would be nice to have on both so the same files > > > can be either dropped for autodiscovery or

Re: [systemd-devel] Multiprofile UKIs on Type 1 entries

On Mo, 22.09.25 15:01, Itxaka Serrano Garcia ([email protected]) wrote: > Hello folks, > > we are investigating a bit around multi-profile UKIs and cmdlines and while > this works as expected, we noticed that on Type 1 entries, the Title ID is > not read from the .profile of the entr

Re: [systemd-devel] [PATCHv5 06/12] kexec: Integrate with the introduced bpf kfuncs

On Thu, Sep 18, 2025 at 9:43 PM Philipp Rudo wrote: > > Hi Pingfan, > > On Tue, 16 Sep 2025 14:52:38 +0800 > Pingfan Liu wrote: > > > On Mon, Sep 01, 2025 at 04:30:42PM +0200, Philipp Rudo wrote: > > > Hi Pingfan, > > > > > > > > > On Tue, 19 Aug 2025 09:24:22 +0800 > > > Pingfan Liu wrote: > >

Re: [systemd-devel] [PATCHv5 00/12] kexec: Use BPF lskel to enable kexec to load PE format boot image

Hi Pingfan, On Tue, 16 Sep 2025 10:00:41 +0800 Pingfan Liu wrote: > On Mon, Sep 01, 2025 at 04:29:29PM +0200, Philipp Rudo wrote: > Hi Philipp, > > Thank you for deep insight, please see the comments > > > Hi Pingfan, > > > > thanks for sharing the updated version of the series. There are a f

Re: [systemd-devel] [PATCHv5 06/12] kexec: Integrate with the introduced bpf kfuncs

Hi Pingfan, On Tue, 16 Sep 2025 14:52:38 +0800 Pingfan Liu wrote: > On Mon, Sep 01, 2025 at 04:30:42PM +0200, Philipp Rudo wrote: > > Hi Pingfan, > > > > > > On Tue, 19 Aug 2025 09:24:22 +0800 > > Pingfan Liu wrote: > > > > > This patch does two things: > > > First, register as a listener

Re: [systemd-devel] [PATCHv5 05/12] kexec: Introduce kexec_pe_image to parse and load PE file

On Mon, Sep 01, 2025 at 04:30:20PM +0200, Philipp Rudo wrote: > Hi Pingfan, > > a few nits in addition to what is mentioned in the cover letter. > Besides the following comment, as we agree on your suggestion, many of the logic in this file will be moved to kimage_file_prepare_segments(). > On

Re: [systemd-devel] Is systemd.io down?

Hi, Glad to see it up again! Have a great day. On 9/17/25 9:28 PM, Zbigniew Jędrzejewski-Szmek wrote: On Wed, Sep 17, 2025 at 07:10:38PM +0600, Ar Rakin wrote: Hi there, The systemd website at shows a 404 Not Found error for all pages. I don't know if something has cha

Re: [systemd-devel] Is systemd.io down?

On Wed, Sep 17, 2025 at 07:10:38PM +0600, Ar Rakin wrote: > Hi there, > > The systemd website at shows a 404 Not Found error for > all pages. I don't know if something has changed, just wanted to let you > all know that it is not working. It's fixed now. I guess it was a pr

Re: [systemd-devel] [PATCHv5 06/12] kexec: Integrate with the introduced bpf kfuncs

On Mon, Sep 01, 2025 at 04:30:42PM +0200, Philipp Rudo wrote: > Hi Pingfan, > > > On Tue, 19 Aug 2025 09:24:22 +0800 > Pingfan Liu wrote: > > > This patch does two things: > > First, register as a listener on bpf_copy_to_kernel() > > Second, in order that the hooked bpf-prog can call the sleepa

Re: [systemd-devel] [PATCHv5 00/12] kexec: Use BPF lskel to enable kexec to load PE format boot image

On Mon, Sep 01, 2025 at 04:29:29PM +0200, Philipp Rudo wrote: Hi Philipp, Thank you for deep insight, please see the comments > Hi Pingfan, > > thanks for sharing the updated version of the series. There are a few > small nits you can find in my comments to the individual patches. > > I also to

Re: [systemd-devel] [PATCHv5 08/12] kexec: Factor out routine to find a symbol in ELF

On Mon, Sep 01, 2025 at 04:31:07PM +0200, Philipp Rudo wrote: > On Tue, 19 Aug 2025 09:24:24 +0800 > Pingfan Liu wrote: > > > The routine to search a symbol in ELF can be shared, so split it out. > > > > Signed-off-by: Pingfan Liu > > Cc: Baoquan He > > Cc: Dave Young > > Cc: Andrew Morton >

Re: [systemd-devel] A start job for unit has failed

Thank you all for the assistance. It turns out that I was chasing ghosts - The original crash was due to a corrupted file. I fixed that pretty quickly, but my follow up issue was the consistent booting into emergency mode that I could not understand. The error message I transcribed for this email w

Re: [systemd-devel] A start job for unit has failed

Usually .device units are needed by /etc/fstab entries (or some other .mount). From the shape of the UUID it looks like it's the /boot or /efi mount for your EFI System Partition (or *some* FAT32 partition in any case), which might have gotten corrupted. In the emergency mode, compare any UUID= or

Re: [systemd-devel] A start job for unit has failed

10.09.2025 21:21, Christian M. wrote: Greetings, I'd like to ask for help with my personal machine if that is okay. I'll list the errors first then explain how I think it occured, and then what I have done so far. The system of mine is stuck in emergency mode. The error from the journalctl is t

Re: [systemd-devel] Arranging groups of services

On Sa, 16.08.25 19:11, Demi Marie Obenour ([email protected]) wrote: > I'm working on Spectrum OS (https://spectrum-os.org/) and am > currently porting it from s6 (https://skarnet.org/software/s6-linux-init/) > to systemd. > > Spectrum OS's host (which is what is being ported) is rather > diff

Re: [systemd-devel] Bringing up ssh after fsck failures

On Thu, Sep 4, 2025 at 10:19 AM Dike, Jeff wrote: > > Hi, > > Is there a canonical way to modify the stock Ubuntu systemd configuration so > that it continues booting enough to bring up sshd after other service > failures? In theory, you could have something like dropbear configured to take pub

Re: [systemd-devel] Arranging groups of services

Hello. On Sat, Aug 16, 2025 at 07:11:32PM -0400, Demi Marie Obenour wrote: > If the Cloud Hypervisor instance is stopped or exits, the others > should be stopped automatically, as they have no other use. > Having BindsTo=, After=, PropagatesStopTo=, and PropagatesReloadTo= > should handle most c

Re: [systemd-devel] Bringing up ssh after fsck failures

On Thu, 4 Sept 2025 at 15:18, Dike, Jeff wrote: > > This is for remote servers, so the default failure mode of running an > emergency shell on the console isn't useful. > > > Apologies for not answering your question, but remote servers, do you not have IDRAC/ILO/KVM/virtual console Those are the

Re: [systemd-devel] [PATCHv5 06/12] kexec: Integrate with the introduced bpf kfuncs

Hi Pingfan, On Tue, 19 Aug 2025 09:24:22 +0800 Pingfan Liu wrote: > This patch does two things: > First, register as a listener on bpf_copy_to_kernel() > Second, in order that the hooked bpf-prog can call the sleepable kfuncs, > bpf_handle_pefile and bpf_post_handle_pefile are marked as > KF_SL

Re: [systemd-devel] Starting programs before udev is running

> On 2 Sep 2025, at 10:09, Demi Marie Obenour wrote: > > The reason I ask is that I'm using systemd-udevd with a > non-systemd service manager (s6 + s6-rc) and it has an > incompatible method for readiness notification (writing > a newline to a file descriptor). Two thoughts. Wrap in a progra

Re: [systemd-devel] [PATCHv5 08/12] kexec: Factor out routine to find a symbol in ELF

On Tue, 19 Aug 2025 09:24:24 +0800 Pingfan Liu wrote: > The routine to search a symbol in ELF can be shared, so split it out. > > Signed-off-by: Pingfan Liu > Cc: Baoquan He > Cc: Dave Young > Cc: Andrew Morton > Cc: Philipp Rudo > To: [email protected] > --- > include/linux/kexec.

Re: [systemd-devel] [PATCHv5 05/12] kexec: Introduce kexec_pe_image to parse and load PE file

Hi Pingfan, a few nits in addition to what is mentioned in the cover letter. On Tue, 19 Aug 2025 09:24:21 +0800 Pingfan Liu wrote: > As UEFI becomes popular, a few architectures support to boot a PE format > kernel image directly. But the internal of PE format varies, which means > each parser

Re: [systemd-devel] [PATCHv5 00/12] kexec: Use BPF lskel to enable kexec to load PE format boot image

Hi Pingfan, thanks for sharing the updated version of the series. There are a few small nits you can find in my comments to the individual patches. I also took an other look at the bigger picture. The way I see it the series contains two major changes. 1. A generic mechanism to parse and run bpf

Re: [systemd-devel] [PATCH] initrd: support erofs as initrd

On Sat, Aug 30, 2025 at 03:49:48PM +0400, Askar Safin wrote: > On Thu, 28 Aug 2025 21:14:34 +0400 Gao Xiang > wrote --- > > Which part of the running system check the cpio signature. > > You mean who checks cpio signature at boot? > Ideally, bootloader should do this. The kernel should

Re: [systemd-devel] Issues with udev/systemd/lvm and initramfs

Okay I have figured it out. It was something "hidden" in the udev rules, and a mistake on my part. LFS doesn't seem to follow the "everything into /usr/bin", and so they were moving "lvm" into /usr/sbin (a separate folder). Inspecting the udev rules led me to notice that it was calling "/usr

Re: [systemd-devel] networkd "Could not set hostname" on boot

Nevermind, that log was apparently from shutdown and not boot. The issues I'm experiencing were actually because ifupdown was still being used for network-online.target. I've disabled that and now everything properly waits for systemd-networkd. On Thu, Aug 28, 2025 at 12:15 PM Sanjay Vasandani w

Re: [systemd-devel] Issues with udev/systemd/lvm and initramfs

Hi ! Indeed I'm 100% aware that this is not the "ideal setup", I'm simply curious as to why it isn't working right. I'd like to get my small init to work well with lvm/udev/systemd. As per your advice, I added -vvv to vgchange and does seem to confirm your idea that it is waiting for the cookie.

Re: [systemd-devel] RestrictSUIDSGID causing unexpected error

On 8/25/25 6:57 AM, Michal Sekletar wrote: openat2() is disabled due to the seccomp filter applied as a consequence of RestrictSUIDSGID=yes. Rationale for this behavior is described in code comment here, https://github.com/systemd/systemd/blob/main/src/shared/seccomp- util.c#L2311

Re: [systemd-devel] RestrictSUIDSGID causing unexpected error

On Sun, Aug 24, 2025 at 11:59 PM Ian Pilcher wrote: > Any suggestions on what might be going on here or what my program should > be doing differently to make this call work would be appreciated. This > error is preventing me from setting DynamicUser=true, because it implies > RestrictSUIDSGID=tru

Re: [systemd-devel] PAM conversation function via varlink?

On Mon, Aug 25, 2025 at 10:34 AM Lennart Poettering wrote: > > On Fr, 22.08.25 14:32, Thorsten Kukuk ([email protected]) wrote: > > > Hi, > > > > Would it be possible or has anybody ever tried to implement the PAM > > conversation function (https://manpages.opensuse.org/pam_conv.3) via > > varlink? >

Re: [systemd-devel] PAM conversation function via varlink?

On Fr, 22.08.25 14:32, Thorsten Kukuk ([email protected]) wrote: > Hi, > > Would it be possible or has anybody ever tried to implement the PAM > conversation function (https://manpages.opensuse.org/pam_conv.3) via > varlink? > > To get rid of setuid binaries (here passwd, chsh and chfn) my idea was >

Re: [systemd-devel] RestrictSUIDSGID causing unexpected error

The way I’ve read the internal operation of this function, is its calling SECCOMP internally in Linux such that calls that allow changing ownership or permissions are disallowed. Though not expressly stated for the function call you listed, it under the covers would be calling those which should

Re: [systemd-devel] systemd-logind fails to start if /etc is read-only

On Mi, 20.08.25 14:50, Demi Marie Obenour ([email protected]) wrote: > >> On a system where /etc is read-only, systemd-logind fails to start. > >> I have tried making / a writable overlayfs without any success so > >> far. The code is at https://github.com/DemiMarie/spectrum (branch > >> b4/s

Re: [systemd-devel] systemd-logind fails to start if /etc is read-only

On 8/18/25 05:02, Lennart Poettering wrote: > On Sa, 16.08.25 22:47, Demi Marie Obenour ([email protected]) wrote: > >> On a system where /etc is read-only, systemd-logind fails to start. >> I have tried making / a writable overlayfs without any success so >> far. The code is at https://githu

  1   2   3   4   5   6   7   8   9   10   >