Re: RFC: Dynamic group limit

Frank van Maarseveen (F.vanMaarseveen@inter.NL.net)
Wed, 28 Jul 1999 23:16:44 +0200

On Wed, Jul 28, 1999 at 12:05:11PM +0200, Mattias.Gronlund wrote:
> > That would eliminate the need for a sysctl.
>
> Not really, the sysctl is a configurable maximum that the a sysadmin
> could set for some reason. As POSIX has support for a "dynamic" limit
> and it isn't a problem to implement I see no reason not to.
Well, only root (or anyone having a special capability) can call
setgroups() and there are much easier ways for programs to eat all
memory than calling setgroups(). I've seen it before happening with
all sorts of limits: one day someone exceeds that limit and the next
thing which will happen is that the sysadmin has to increase that limit.
I don't see any real benefit from such a limit but on the other hand
no real objection either when it uses the sysctl interface.

> I have checked your NFS-patch and thought a lot about it, but I see that
> as one of the next step problems in this.
Thanks! Finally someone who noticed this patch.

> The problem with changing NGROUPS_MAX is that this is a constant that
> gets
> compiled into binarys. If I buy a program with no source I might get
> trouble.
Yep! Even with source it is a nuisance having to compile it again
because someone used statically compiled limits.

> I do not know enought about searching algorithms to just stef forward
> and say
> which to use, I might just try to implement it with a hash...
Me too. But there exist something called AVL (binary tree, balanced)
which is much more scalable. I believe it is already in the kernel.

[OFF-TOPIC]: Anyone having experience setting
NGROUPS_MAX > 32 on OSF/1 Alpha 4.0? please mail me.

-- 
Frank van Maarseveen                               Driebergen
f.vanmaarseveen@inter.nl.net                  The Netherlands
-------------------------------------------------------------


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/