[Devel] Re: Playing with namespaces and bloat-o-meeter
Serge E. Hallyn
serue at us.ibm.com
Wed Sep 26 08:42:36 PDT 2007
Quoting Pavel Emelyanov (xemul at openvz.org):
> Eric W. Biederman wrote:
> > Pavel Emelyanov <xemul at openvz.org> writes:
> >
> >>>
> >>> +config NAMESPACES
> >>> + bool "The namespaces support"
> >>> + help
> >>> + Provides the way to make tasks work with different objects using
> >>> + the same id
> >>> +
> >>> +config NS_UTS
> >>> + bool "Uname namespace"
> >>> + depends on NAMESPACES
> >>> + help
> >>> + The utsname namespace
> >>> +
> >>> +config NS_IPC
> >>> + bool "IPC namespace"
> >>> + depends on NAMESPACES && SYSVIPC
> >>> + help
> >>> + The SYSVIPC ids namespaces
> >>> +
> >>> +config NS_PIDS
> >>> + bool "PID namespace"
> >>> + depends on NAMESPACES
> >>> + help
> >>> + Tasks see only the pids living in the same namespace and in the
> >>> + child namespaces
> >>> +
> >>> +config NS_UID
> >>> + bool "UID namespace"
> >>> + depends on NAMESPACES && EXPERIMENTAL
> >>> + help
> >>> + Support user namespaces. This allows containers, i.e.
> >>> + vservers, to use user namespaces to provide different
> >>> + user info for different servers. If unsure, say N.
> >>> +
> >>> config BLK_DEV_INITRD
> >>> bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
> >>> depends on BROKEN || !FRV
> >
> > The reason we removed these options earlier was a maintenance issue
> > and the fact we could not actually compile out the namespaces.
>
> I do not propose to compile out the namespaces, I just propose
> to compile out the code that does the clone and release of new
> namespaces. This is absolutely painless.
>
> > If we don't cause maintenance complications I think the general
> > idea is fine. But please. This all should show up under
> > CONFIG_EMBEDDED since the only purpose is to save space.
>
> Hm... Ok, but I also try to save the vmlinux size on my home
> PC, so I'd be happy if I could just throw these things out.
>
> Anyway - I will move the CONFIG_NAMESPACES to be selectable
> with the EMBEDDED only.
>
> > While things are experimental there is an additional purpose of
> > not exposing people to broken or partially working code, so it
> > does make sense to have an option there.
>
> Ok, thanks.
>
> So your accolades, can they be transformed into Acked-by-s or
> just mentioned in the patch like "reviewed and approved by ..."?
I for one have no objection to the idea itself. There have been several
good suggestions though so I would like to see one more round here to
which I can add an Acked-by.
thanks,
-serge
_______________________________________________
Containers mailing list
Containers at lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
More information about the Devel
mailing list