Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
* [Buildroot] systemd-timesyncd fails to start
@ 2016-10-18  8:55 Petr Kulhavy
  2016-10-18  9:45 ` Thomas Petazzoni
  0 siblings, 1 reply; 3+ messages in thread
From: Petr Kulhavy @ 2016-10-18  8:55 UTC (permalink / raw)
  To: buildroot

Hi,

there is a peculiar problem with systemd NTP synchronization, that in 
the default rootfs the systemd-timesyncd.service always fails to start.
The reason is that systemd-timesyncd requires the /var/tmp to be a 
permanent storage and not just a link to /tmp.
See the second comment on this page: 
https://bbs.archlinux.org/viewtopic.php?id=151937
And really, if I replace the symlink with an empty folder it starts working.

I'm wondering what is the proper way to fix this. By default buildroot 
creates /var/tmp as a link to /tmp, however
on some platforms the rootfs might be read-only or very limited in 
space, so making /var/tmp permanent might break other things...

It would be greatly appreciated if anybody could shed more light on this.

Thanks
Petr

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [Buildroot] systemd-timesyncd fails to start
  2016-10-18  8:55 [Buildroot] systemd-timesyncd fails to start Petr Kulhavy
@ 2016-10-18  9:45 ` Thomas Petazzoni
  2016-10-19 19:38   ` Yann E. MORIN
  0 siblings, 1 reply; 3+ messages in thread
From: Thomas Petazzoni @ 2016-10-18  9:45 UTC (permalink / raw)
  To: buildroot

Hello,

On Tue, 18 Oct 2016 10:55:26 +0200, Petr Kulhavy wrote:

> there is a peculiar problem with systemd NTP synchronization, that in 
> the default rootfs the systemd-timesyncd.service always fails to start.
> The reason is that systemd-timesyncd requires the /var/tmp to be a 
> permanent storage and not just a link to /tmp.
> See the second comment on this page: 
> https://bbs.archlinux.org/viewtopic.php?id=151937
> And really, if I replace the symlink with an empty folder it starts working.
> 
> I'm wondering what is the proper way to fix this. By default buildroot 
> creates /var/tmp as a link to /tmp, however
> on some platforms the rootfs might be read-only or very limited in 
> space, so making /var/tmp permanent might break other things...
> 
> It would be greatly appreciated if anybody could shed more light on this.

I've added Yann E. Morin in Cc. He is working on a revamp of the root
filesystem skeleton for the systemd case, specifically to solve this
sort of problem.

You can have a look at
https://github.com/yann-morin-1998/buildroot/commits/yem/systemd-skeleton,
but I'm not sure it's the latest updated work from Yann.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [Buildroot] systemd-timesyncd fails to start
  2016-10-18  9:45 ` Thomas Petazzoni
@ 2016-10-19 19:38   ` Yann E. MORIN
  0 siblings, 0 replies; 3+ messages in thread
From: Yann E. MORIN @ 2016-10-19 19:38 UTC (permalink / raw)
  To: buildroot

Petr, Thomas, All,

On 2016-10-18 11:45 +0200, Thomas Petazzoni spake thusly:
> On Tue, 18 Oct 2016 10:55:26 +0200, Petr Kulhavy wrote:
> > there is a peculiar problem with systemd NTP synchronization, that in 
> > the default rootfs the systemd-timesyncd.service always fails to start.
> > The reason is that systemd-timesyncd requires the /var/tmp to be a 
> > permanent storage and not just a link to /tmp.
> > See the second comment on this page: 
> > https://bbs.archlinux.org/viewtopic.php?id=151937
> > And really, if I replace the symlink with an empty folder it starts working.
> > 
> > I'm wondering what is the proper way to fix this. By default buildroot 
> > creates /var/tmp as a link to /tmp, however
> > on some platforms the rootfs might be read-only or very limited in 
> > space, so making /var/tmp permanent might break other things...
> > 
> > It would be greatly appreciated if anybody could shed more light on this.
> 
> I've added Yann E. Morin in Cc. He is working on a revamp of the root
> filesystem skeleton for the systemd case, specifically to solve this
> sort of problem.

Yup, still working on it.

> You can have a look at
> https://github.com/yann-morin-1998/buildroot/commits/yem/systemd-skeleton,
> but I'm not sure it's the latest updated work from Yann.

Nope, that is the series I submitted in July, but it "just works". Even
if it is not what's going to ultimately go upstream, I'd still be
interested in getting some feedback for that series, since I'm aiming to
have a similar end-result with the new work. So, if you could give it a
spin and report back, that's be great! ;-)

The revamp taking Thomas' comments into account is:
    https://github.com/yann-morin-1998/buildroot/commits/yem/systemd-skeleton-2

But that is absolutely not finished.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2016-10-19 19:38 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-10-18  8:55 [Buildroot] systemd-timesyncd fails to start Petr Kulhavy
2016-10-18  9:45 ` Thomas Petazzoni
2016-10-19 19:38   ` Yann E. MORIN

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox