From mboxrd@z Thu Jan 1 00:00:00 1970 Reply-To: kernel-hardening@lists.openwall.com Sender: Vasiliy Kulikov Date: Tue, 14 Jun 2011 18:24:29 +0400 From: Vasiliy Kulikov Message-ID: <20110614142429.GB3966@albatros> References: <1307889901-6226-1-git-send-email-segoon@openwall.com> <201106141554.50945.arnd@arndb.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <201106141554.50945.arnd@arndb.de> Subject: [kernel-hardening] Re: [RFC v2 02/04] procfs: add hidepid modes as mount options To: Arnd Bergmann Cc: linux-kernel@vger.kernel.org, kernel-hardening@lists.openwall.com, Andrew Morton , Greg Kroah-Hartman , "David S. Miller" , David Rientjes , Al Viro , Nick Piggin , Miklos Szeredi , Alexey Dobriyan , Frederic Weisbecker , WANG Cong List-ID: On Tue, Jun 14, 2011 at 15:54 +0200, Arnd Bergmann wrote: > On Sunday 12 June 2011, Vasiliy Kulikov wrote: > > +static const match_table_t tokens = { > > + {Opt_hidepid, "hidepid=%u"}, > > + {Opt_gid, "gid=%u"}, > > + {Opt_hidenet, "hidenet"}, > > + {Opt_nohidenet, "nohidenet"}, > > + {Opt_err, NULL}, > > +}; > > I don't really have an opinion on your patch, but it seems that it does more than > the description explains: The hidenet/nohidenet option is in the patch as well, > although it doesn't have much of an effect. Correct, it is just a matter of a patch division granularity. Alexey said the patch should be divided into pid and net parts. I divided it into (pid + all mount opts parsing) and (actual hidenet usage). As both pid and net parts depend on options parsing, they are not fully independent, and cannot be well splitted (or I just don't see how). -- Vasiliy Kulikov