On Fri, Mar 15, 2019 at 11:48:43AM +0100, Andrea Bolognani wrote:
On Fri, 2019-03-15 at 10:26 +0000, Daniel P. Berrangé wrote:
> On Fri, Mar 15, 2019 at 10:23:52AM +0000, Daniel P. Berrangé wrote:
> > On Thu, Mar 14, 2019 at 12:31:08PM +0100, Andrea Bolognani wrote:
> > > env:
> > > - IMAGE="centos-7"
> > > - -
DISTCHECK_CONFIGURE_FLAGS="--with-init-script=upstart"
> > > + -
DISTCHECK_CONFIGURE_FLAGS="--with-init-script=redhat"
> >
> > Heh, this was always kind of wrong since RHEL 6 was the one
> > with upstart
True, but the argument IIRC was that we would get coverage for that
code path regardless of what distribution we'd run it on... It's not
like we actually test whether init can bring up libvirtd and friends
in our 'make check' anyway :)
> Actually on second thoughts, this is not desirable.
>
> We should be actually purging the traditional init script too. RHEL-7
> is systemd based and so are all Fedora's. The "redhat" initscript
> was last used in RHEL-6. Even if other distros use classis sysvinit
> I don't think they'll use the Red Hat variant initscript.
Yeah, I was thinking about that yesterday too...
Honestly I just didn't spend time checking whether the "redhat" init
scripts actually have anything RHEL-specific or they would work on
other SysV-init based distribution, so I decided to go only for the
obvious low-hanging fruit at first.
IME the only thing common about sysvinit scripts is that they're
all written in shell and all buggy. I would just kill it as the
name we gave it always indicated it was only for redhat distro
variants.
Regards,
Daniel
--
|:
https://berrange.com -o-
https://www.flickr.com/photos/dberrange :|
|:
https://libvirt.org -o-
https://fstop138.berrange.com :|
|:
https://entangle-photo.org -o-
https://www.instagram.com/dberrange :|