New upstream sysvinit tarball version 2.87dsf created

[Scott James Remnant]

On Mon, 2009-07-13 at 21:22 +0200, Petter Reinholdtsen wrote:

> Thank you for the interesting feedback.  But some of your comments
> make me suspect you misunderstand what actually happened here when I
> made the new sysvinit release.
> 
> All of the patches now included in the new sysvinit tarball have been
> included in Debian, SuSe and Fedora for several years already.  So,
> nothing really new here, just a new upstream release to provide an
> updated synchronization point between the distributions.
> 
Ah, I see!  I thought you'd merged a whole bunch of SuSE and Fedora
patches recently that changed some behaviours.

I roughly based Upstart's sysv tools off a 2007 version of sysvinit, so
there may be a few differences:

> Yes.  This patch has been in Debian since 2006, fixing #58119.
> 
> >>   * Fix utmp/wtmp updating on 64-bit platforms.  Patch by Bill
> >>     Nottingham and Fedora.
> >> 
> > What was this patch?
> 
> Included in Debian since version 2.86.ds1-40.  See the SVN repository
> for the details.
> 
Looks like I found and fixed this one independently:

http://bazaar.launchpad.net/~scott/upstart/trunk/revision/268

> >>   * Allow 'telinit u' in runlevels 0 and 6.  Patch from Thomas Hood.
> >> 
> > What runlevels are "telinit u" normally allowed in in sysvinit?
> 
> It used to only allow runlevels S12345.  Fixed in debian in version
> 2.86.ds1-9 in 2006, fixing #345719.
> 
I think I just allow it at any point; that being said, it does nothing
right now ;-)

> > I noticed a strange case with killall, it seems to cause the Upstart
> > rc.conf job to get SIGSTOP/SIGCONT repeatedly?
> 
> killall5 signals all processes with SIGSTOP before deciding what to
> kill, to avoid new processes popping up while killall5 is running.
> 
Interesting ... Upstart will send SIGCONT to any supervised process that
raises SIGSTOP, so this is likely working against you here.

I'm not entirely sure why I send SIGCONT tbh.

> >>   * Change reboot/halt to work properly when used as a login shell.
> >>     Patch by Dale R. Worley and Fedora.
> >> 
> > What was this patch?
> 
> See SVN.  Trusted Fedora to know what they were doing, and applied the
> patch without testing how things behave.
> 
I should probably apply this same patch to initctl, reboot, etc.

Scott
-- 
Scott James Remnant
scott at canonical.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
URL: <http://lists.alioth.debian.org/pipermail/initscripts-ng-devel/attachments/20090714/d830188e/attachment.pgp>