Re: CTL_UNNUMBERED (Re: [PATCH] 9p: Don't use binary sysctl numbers.)

["Eric Van Hensbergen" <ericvh@gmail.com>]

On 7/21/07, Eric W. Biederman <ebiederm@xmission.com> wrote:
> Alexey Dobriyan <adobriyan@gmail.com> writes:
>
> >
> > That's separate patch but CTL_UNNUMBERED must die, because it's totally
> > unneeded. If you don't want sysctl(2) interface just SKIP ->ctl_name
> > initialization and save one line for something useful.
>
> As for the 9p code it doesn't seem to need or want a real binary
> interface.  The 9p debug code picking of a semi-random number and not
> patching it into sysctl.h like it should for a binary interface is
> an implementation bug, and a maintenance problem.
>

Now that -rc1 is out, lets talk a bit more about this.  Lucho can you
provide some level of justification of why you went for a sysctl
interface versus something directly accessible within the file system
-- that would seem more on-par with the 9p philosophy.

Perhaps its time for a general cleanup of the debug_level stuff -- it
was always ugly to have it as a global, but there was just no clear
way to have the session structure available everywhere we use it.

               -eric