On Tue, Nov 08, 2011 at 10:50:00PM +0530, Ritesh Raj Sarraf wrote:
> On 10/12/2011 11:39 PM, Alessandro Ghedini wrote:
> > Seems like ulatencyd is being signaled to terminate (either SIGABRT, SIGINT,
> > SIGTERM). That's the only way it could say "abort cleanup" AFAICT.
> >
> > Anyway, this is a rather weird bug, and I cannot reproduce it (I'm not even
> > sure this is actually a bug in ulatencyd itself, or some sort of
> > incompatibility with other software), therefore I can't be much more
> > helpful here.
>
>
> I have some more data points for this, I think.
>
> ├─start_kdeinit
> ├─syndaemon
> ├─udevd───2*[udevd]
> ├─udisks-daemon─┬─udisks-daemon
> │ └─2*[{udisks-daemon}]
> ├─ulatencyd─┬─sh
> │ └─{ulatencyd}
> ├─upowerd───{upowerd}
> ├─vpnc-connect
> ├─wicd───wicd-monitor
> ├─wicd-kde───{wicd-kde}
> └─wpa_supplicant
>
> This is the pstree output. Do you have any idea what that "sh" shell
> process is for. It only is seen during system start-up. If I restart the
> ulatencyd daemon, it is not seen. And as far as I can tell, the silently
> killed problem is seen only with ulatencyd's system start-up process.
> But I could be wrong. I haven't tested it enough of days to conclude.
I see that "sh" too, but I don't know what it is supposed to do.
> Any reasons why you do a test first?
> [...]
Yep, it is used to check whether ulatencyd has been already started and
to return a proper status accordingly (do_start() returns '1' if the daemon
is already running).
> I have parallel boot enabled on my sysvinit with CONCURRENCY=makefile.
>
> Given that the shell process is only seen with system start-up i.e.
> during real parallel execution of init scripts, I have a feel it is
> something wrong there.
I don't know (yet) if it is wrong or not, but it is weird nevertheless.
I'll investigate further. Thanks for your help.
Cheers
--
perl -E'$_=q;$/= @{[@_]};and s;\S+;<inidehG ordnasselA>;eg;say~~reverse'
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
