]] Lennart Poettering
| I have decided not to merge this part for now. I'd much prefer if ubuntu
| would adopt the lock group too, since everything else appears to be a
| security nightmare to me. Also note that Ubuntu and Debian are in the
| same boat here, so if we merge some fix for this I wan
]] Kay Sievers
| On Mon, Oct 25, 2010 at 15:24, Andrew Edmunds
| wrote:
|
| > +Where=/var/lock
| > +Type=tmpfs
| > +m4_dnl
| > +m4_ifdef(`TARGET_UBUNTU',
| > +`Options=nosuid,nodev,noexec',
| > +`Options=mode=775,gid=lock')
|
| Are you sure that /var/lock is world-writable? That's something th
Lennart Poettering wrote:
> Do you happen to know why /var/run doesn't have nodev,noexec set too?
No, I'm afraid not.
> I have now added all three options to both mounts for all distros, as I
> think all distros oculd benefit equally from it. I figure people will
> report back if that breaks som
On Tue, 26.10.10 00:44, Michael Biebl (mbi...@gmail.com) wrote:
>
> 2010/10/26 Lennart Poettering :
> > Note that we do not support booting a real system with systemd on PID !=
> > 1. The primary reason for that is that we need the SIGCHLD signals
> > properly, and we only get those as PID 1. Ear
On Mon, Oct 25, 2010 at 4:09 PM, Lennart Poettering
wrote:
> On Mon, 25.10.10 18:20, Lucas De Marchi (lucas.demar...@profusion.mobi) wrote:
>
>>
>> On Mon, Oct 25, 2010 at 5:39 PM, Lennart Poettering
>> wrote:
>> > On Mon, 25.10.10 12:42, Canek Peláez Valdés (can...@gmail.com) wrote:
>> >
>> >> >
2010/10/26 Lennart Poettering :
> Note that we do not support booting a real system with systemd on PID !=
> 1. The primary reason for that is that we need the SIGCHLD signals
> properly, and we only get those as PID 1. Early systemd versions
> supported a mode where we didn't rely on those signals
On Fri, 15.10.10 12:25, Robert Schwebel (r.schwe...@pengutronix.de) wrote:
> We currently only create the private D-Bus server when pid == 1, but
> this has also to be done if the user started us with --system.
> Otherwhise we get this error:
>
> systemd-cgroups-agent[164]: Failed to get D-Bus
On Sat, 16.10.10 11:07, Andrey Borzenkov (arvidj...@gmail.com) wrote:
> (Working on Mandriva cooker), systemd v11.
>
> {pts/0}% sudo rpm -Fvh RPMS/x86_64/*
> Подготовка... ### [100%]
>1:systemd-units
On Sun, 17.10.10 13:39, Ozan Çağlayan (o...@pardus.org.tr) wrote:
Heya,
> 1. swap.target always fails during boot. But once triggered after boot
> with systemctl restart, it succesfully enables the /.swap file. .swap
> file is on the rootfs which is an ext4 filesystem. I'm not too
> powerful with
On Wed, 20.10.10 19:40, enaut (enau...@googlemail.com) wrote:
> Hey,
> as killall.service is not necessary anymore here two patches to remove
> it completely the first removes it from the default units, the second
> removes from the Debian and Fedora specific files. As I neither have
> Debian nor
On Mon, 25.10.10 17:39, Bill Nottingham (nott...@redhat.com) wrote:
> > Hmm, pam_systemd should create a user dir beneath it and set
> > XDG_RUNTIME_DIR accordingly. Doesn't that work for you?
>
> I was looking for non-systemd users - while the user-specific dir
> is created, there's no actual d
Lennart Poettering (lenn...@poettering.net) said:
> > > > They will get 'cleaned' on boot due to being tmpfs anyway.
> > >
> > > Hmm, but at least /var/run/user can be full of user stuff which might
> > > remain around when the user session crashes... Are you sure we don't
> > > need to clena thi
On Thu, 21.10.10 22:37, Bill Nottingham (nott...@redhat.com) wrote:
> They will get 'cleaned' on boot due to being tmpfs anyway.
Merged the subsys part now, but left the user part out, since I think we
should be cleaning this up from time to time during runtime.
Thanks!
> -d /var/lock/subsys 07
Oh indeed the ubuntu packages are another brilliant edition to a brilliant
project! Been so enthusiastic about systemd since your talk @ linuxtag :D
Indeed it could be possible that they see the benefits of a well developed
and well designed alternative to upstart in the future and adopt :P
Paul
On Mon, 25.10.10 16:46, Bill Nottingham (nott...@redhat.com) wrote:
>
> Lennart Poettering (lenn...@poettering.net) said:
> > > They will get 'cleaned' on boot due to being tmpfs anyway.
> >
> > Hmm, but at least /var/run/user can be full of user stuff which might
> > remain around when the use
On Thu, 21.10.10 22:37, Bill Nottingham (nott...@redhat.com) wrote:
Applied. Thanks!
>
> Signed-off-by: Bill Nottingham
> ---
> tmpfiles.d/systemd.conf |3 ++-
> 1 files changed, 2 insertions(+), 1 deletions(-)
>
> diff --git a/tmpfiles.d/systemd.conf b/tmpfiles.d/systemd.conf
> index 41f
On Mon, 25.10.10 18:20, Lucas De Marchi (lucas.demar...@profusion.mobi) wrote:
>
> On Mon, Oct 25, 2010 at 5:39 PM, Lennart Poettering
> wrote:
> > On Mon, 25.10.10 12:42, Canek Peláez Valdés (can...@gmail.com) wrote:
> >
> >> > As for shutdown, yeah, it's as fast as it should be ;-) We're doing
On Mon, 25.10.10 21:22, Paul McDonough (pmc...@gmail.com) wrote:
> I don't think Ubuntu will adopt Systemd as it means them dropping their own
> upstart which is unlikely.
Well, things change. And there are rumours that this move might be much
more likely than people might think. They have made a
Lennart Poettering (lenn...@poettering.net) said:
> > They will get 'cleaned' on boot due to being tmpfs anyway.
>
> Hmm, but at least /var/run/user can be full of user stuff which might
> remain around when the user session crashes... Are you sure we don't
> need to clena this up from time to ti
I don't think Ubuntu will adopt Systemd as it means them dropping their own
upstart which is unlikely.
However I know we are interested in making Aurora Systemd native and have
been playing with the git pulls with Debian patches :D
If you can offer any advice of using Systemd as a replacement for
On Mon, Oct 25, 2010 at 5:39 PM, Lennart Poettering
wrote:
> On Mon, 25.10.10 12:42, Canek Peláez Valdés (can...@gmail.com) wrote:
>
>> > As for shutdown, yeah, it's as fast as it should be ;-) We're doing an
>> > amazing work with systemd! My next hope is to have the cron features
>> > in it and
On Thu, 21.10.10 22:37, Bill Nottingham (nott...@redhat.com) wrote:
> They will get 'cleaned' on boot due to being tmpfs anyway.
Hmm, but at least /var/run/user can be full of user stuff which might
remain around when the user session crashes... Are you sure we don't
need to clena this up from ti
On Fri, 22.10.10 19:11, Tom Gundersen (t...@jklm.no) wrote:
> This is now split into smaller units and shipped with the downstream package.
Applied both patches.
Lennart
--
Lennart Poettering - Red Hat, Inc.
___
systemd-devel mailing list
systemd-dev
On Sun, 24.10.10 16:49, Michael Biebl (mbi...@gmail.com) wrote:
> 2010/10/24 Gustavo Sverzut Barbieri :
> >
> > well, just do the patch, test to see if you handled it right (ie:
> > you're not looping forever in code) and mail, Lennart might like that
> > option you propose.
>
> Sure, proposed pa
On Mon, 25.10.10 06:45, Andrey Borzenkov (arvidj...@gmail.com) wrote:
> As this is used in RPM, it means no getty gets installed on systemd new
> install:
>
> {pts/0}% sudo systemctl enable ge...@.service
> Unit name ge...@.service is not a valid unit name.
> Cannot install unit ge...@.service:
On Mon, 25.10.10 15:24, enaut (enau...@googlemail.com) wrote:
> Am 25.10.2010 07:18, schrieb Canek Peláez Valdés:
> >Is this possible? I don't want anything fancy like bootchart, only the
> >processes (and, if possible, in the order of invocation) from
> >"/bin/systemd udev" to "/sbin/agetty 38400
On Mon, 25.10.10 12:42, Canek Peláez Valdés (can...@gmail.com) wrote:
> > As for shutdown, yeah, it's as fast as it should be ;-) We're doing an
> > amazing work with systemd! My next hope is to have the cron features
> > in it and it's all I ever need to manage my system.
>
> I'm quite happy wit
On Mon, 25.10.10 23:24, Andrew Edmunds (andrew.edmu...@yahoo.com.au) wrote:
> The following entries appear in /lib/init/fstab (Upstart's
> private list of mounts):
> none /var/run tmpfs mode=0755,nosuid,showthrough0 0
> none /var/lock tmpfs nodev,noexec,nosuid,showthrough 0 0
Do you ha
On Mon, 25.10.10 23:21, Andrew Edmunds (andrew.edmu...@yahoo.com.au) wrote:
> Hi,
>
> I have put up some Ubuntu packages for systemd and related things in
> an Ubuntu personal package archive (PPA). It is available here:
> https://launchpad.net/~andrew-edmunds/+archive/ppa
>
> There is a wiki pa
On Mon, 25.10.10 23:23, Andrew Edmunds (andrew.edmu...@yahoo.com.au) wrote:
> The approved method for determining the installed release of
> Ubuntu is to execute "lsb_release". However, this is in /usr/bin and
> is implemented in python so it is not safe to execute at this early
> stage of bootin
On Mon, 25.10.10 23:22, Andrew Edmunds (andrew.edmu...@yahoo.com.au) wrote:
> Previously Ubuntu was treated as being equivalent to Debian, but the two
> distributions require different behaviour in certain places. This commit does
> not change the behaviour of systemd on either distro but it crea
Woohoo, thanks! I'll start playing with it in virtualbox.
On Mon, Oct 25, 2010 at 6:21 AM, Andrew Edmunds
wrote:
>
> Hi,
>
> I have put up some Ubuntu packages for systemd and related things in
> an Ubuntu personal package archive (PPA). It is available here:
> https://launchpad.net/~andrew-edmun
On Mon, Oct 25, 2010 at 6:27 AM, Gustavo Sverzut Barbieri
wrote:
[...]
> Hi, I'm one of the guys helping Systemd to run on Gentoo and had
> similar high-pid counts.
Hi Gustavo.
> The biggest cause of problems was hotplug being installed. Alone it
> would account for over 1000 pid. Removing it, b
On Mon, Oct 25, 2010 at 15:24, Andrew Edmunds
wrote:
> +Where=/var/lock
> +Type=tmpfs
> +m4_dnl
> +m4_ifdef(`TARGET_UBUNTU',
> +`Options=nosuid,nodev,noexec',
> +`Options=mode=775,gid=lock')
Are you sure that /var/lock is world-writable? That's something that
should be fixed on Ubuntu, I guess.
Am 25.10.2010 07:18, schrieb Canek Peláez Valdés:
Is this possible? I don't want anything fancy like bootchart, only the
processes (and, if possible, in the order of invocation) from
"/bin/systemd udev" to "/sbin/agetty 38400 tty1".
Why not bootchart? its quite nice and if you only want the r
The following entries appear in /lib/init/fstab (Upstart's
private list of mounts):
none /var/run tmpfs mode=0755,nosuid,showthrough0 0
none /var/lock tmpfs nodev,noexec,nosuid,showthrough 0 0
The "showthrough" option is Upstart specific but the remaining
options should be honoured.
---
The approved method for determining the installed release of
Ubuntu is to execute "lsb_release". However, this is in /usr/bin and
is implemented in python so it is not safe to execute at this early
stage of booting. This code parses /etc/lsb-release which is
where "lsb_release" looks for the info
Previously Ubuntu was treated as being equivalent to Debian, but the two
distributions require different behaviour in certain places. This commit does
not change the behaviour of systemd on either distro but it creates a
framework for changes to be introduced by later commits.
The following previ
Hi,
I have put up some Ubuntu packages for systemd and related things in
an Ubuntu personal package archive (PPA). It is available here:
https://launchpad.net/~andrew-edmunds/+archive/ppa
There is a wiki page that describes how to install the packages here:
https://wiki.ubuntu.com/systemd
I shou
On Mon, Oct 25, 2010 at 10:20 AM, Kay Sievers wrote:
> On Mon, Oct 25, 2010 at 13:27, Gustavo Sverzut Barbieri
> wrote:
>
>> It boots into my Enlightenment17 with pid around 500,
>> with a hack to ignore some acpid udev modprobes I can get it to 300).
>
> Yeah, it's acpi in /sys (not the deprecat
On Mon, Oct 25, 2010 at 13:27, Gustavo Sverzut Barbieri
wrote:
> It boots into my Enlightenment17 with pid around 500,
> with a hack to ignore some acpid udev modprobes I can get it to 300).
Yeah, it's acpi in /sys (not the deprecated daemon acpid). After you
mentioned it on IRC a while back and
On Mon, Oct 25, 2010 at 3:18 AM, Canek Peláez Valdés wrote:
> Hi; I'm using (with a high success) systemd in my Gentoo system.
> However, even with really few services running my startup time is not
> that different from OpenRC, and my PID count is really high; with an
> emergency boot, it's on th
42 matches
Mail list logo
