mountlate being too mount-happy

[Mateusz Guzik]

On Sun, Nov 18, 2012 at 02:28:28PM +0000, Chris Rees wrote:
> On 18 November 2012 06:09, Hiroki Sato <hrs at freebsd.org> wrote:
> > Mateusz Guzik <mjguzik at gmail.com> wrote
> >   in <20121118002245.GB15055 at dft-labs.eu>:
> >
> > mj> On Sun, Nov 18, 2012 at 07:43:25AM +0900, Hiroki Sato wrote:
> > mj> > Chris Rees <utisoft at gmail.com> wrote
> > mj> >   in <CADLo839wqzAPenuQDOVpQ74yjCMkPQNceKpvs_N9XNwMLrkC1A at mail.gmail.com>:
> > mj> >
> > mj> > ut> On 2 November 2012 14:21, Eitan Adler <lists at eitanadler.com> wrote:
> > mj> > ut> > On 2 November 2012 09:56, Chris Rees <utisoft at gmail.com> wrote:
> > mj> > ut> >> I'll take a look.
> > mj> > ut> >
> > mj> > ut> > untested:
> > mj> > ut>
> > mj> > ut> Based on Eitan's patch, I've tested this one, and documented it in mount(8) too:
> > mj> > ut>
> > mj> > ut> http://www.bayofrum.net/~crees/patches/mountonlylate.diff
> > mj> > ut>
> > mj> > ut> Does anyone have any suggestions/objections/urge to approve it?
> > mj> >
> > mj> >  Is the original problem due to backgrounding of NFS mount only?  If
> > mj> >  so, implementing prevention of duplicate invocation into mount(8)
> > mj> >  would be more reasonable, I think.
> > mj> >
> > mj>
> > mj> We have 2 distinct scripts that try to mount same set of filesystems.
> > mj> I think this is the real bug here and proposed patches makes it go away in
> > mj> an IMHO acceptable way.
> >
> >  I just wanted to make sure if the case is limited to background NFS
> >  mount or not.
> >
> >  rc.d/mountlate just tries to mount the filesystems that are not
> >  mounted yet at that time in addition to the "late" ones, not always
> >  to mount the same set twice.  If it is a bug, it is better to simply
> >  fix -l to exclude not-yet-mounted ones without "late" keyword than
> >  adding another option.
> 
> I don't think it's a bug as such-- -l option is clearly labelled in
> the manpage (emphasis mine):
> 
> When used in conjunction with the -a option, *also* mount those
> file systems which are marked as ``late''.
> 
> I think that for POLA and to avoid changing behaviour of an option
> that's been there a long time we need the -L option.
> 
> I disagree with Mateusz here-- split operations in rc makes two
> scripts necessary; mount and mountlate are two separate operations,
> done at different times.
> 

So it turns out I expressed myself incorrectly.

I'm fine with two separate scripts doing mounts. I'm not fine with the
fact that both scripts may try to mount very same filesystem (which is
the case now with background nfs).

crees's patch fixes this problem in a way that I think is ok.

-- 
Mateusz Guzik <mjguzik gmail.com>