From owner-freebsd-hackers  Fri Feb  3 13:05:02 1995
Return-Path: hackers-owner
Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id NAA20876 for hackers-outgoing; Fri, 3 Feb 1995 13:05:02 -0800
Received: from gvr.win.tue.nl (root@gvr.win.tue.nl [131.155.210.19]) by freefall.cdrom.com (8.6.9/8.6.6) with ESMTP id NAA20866 for <FreeBSD-hackers@freefall.cdrom.com>; Fri, 3 Feb 1995 13:04:57 -0800
Received: by gvr.win.tue.nl (8.6.9/1.53)
    id WAA20262; Fri, 3 Feb 1995 22:03:51 +0100
From: guido@gvr.win.tue.nl (Guido van Rooij)
Message-Id: <199502032103.WAA20262@gvr.win.tue.nl>
Subject: Re: limits (again)
To: bde@zeta.org.au (Bruce Evans)
Date: Fri, 3 Feb 1995 22:03:50 +0100 (MET)
Cc: FreeBSD-hackers@freefall.cdrom.com
In-Reply-To: <199502030635.RAA31880@godzilla.zeta.org.au> from "Bruce Evans" at Feb 3, 95 05:35:28 pm
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Content-Length: 838       
Sender: hackers-owner@FreeBSD.org
Precedence: bulk

Bruce Evans wrote:
> 
> >Okay, if there are no complaints, I'm gonna implement
> >kern.maxfilesperproc
> 
> One complaint is enough? :-).  I've already complained.

Sorry; I must have missed that one. Could you remail it? 

> 
> >If people do want other such limits implemented, let them shout and I'll
> >try doing that as well.
> 
> A correct implementation seems to require the following: (:-()
> (1) search for and reduce all higher limits in rlim_cur and rlim_max in
> the process table.
> (2) search for and kill all processes that expect the limits that they
> have already checked to remain valid.  See src/lib/libc/gen/sysconf.c.

It depends on how you look at it. Does resetting the value of it means that
at once everything should obey it, or is it a restriction from now on.

I dont see a problem with the latter one.

-Guido