[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180110022954.GA931@zzz.localdomain>
Date: Tue, 9 Jan 2018 18:29:54 -0800
From: Eric Biggers <ebiggers3@...il.com>
To: Kees Cook <keescook@...omium.org>
Cc: "linux-fsdevel@...r.kernel.org" <linux-fsdevel@...r.kernel.org>,
Alexander Viro <viro@...iv.linux.org.uk>,
Joe Lawrence <joe.lawrence@...hat.com>,
Michael Kerrisk <mtk.manpages@...il.com>,
Willy Tarreau <w@....eu>,
Mikulas Patocka <mpatocka@...hat.com>,
"Luis R . Rodriguez" <mcgrof@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Eric Biggers <ebiggers@...gle.com>
Subject: Re: [PATCH 1/7] pipe, sysctl: drop 'min' parameter from
pipe-max-size converter
On Tue, Jan 09, 2018 at 02:20:45PM -0800, Kees Cook wrote:
> On Sun, Jan 7, 2018 at 9:35 PM, Eric Biggers <ebiggers3@...il.com> wrote:
> > From: Eric Biggers <ebiggers@...gle.com>
> >
> > Before validating the given value against pipe_min_size,
> > do_proc_dopipe_max_size_conv() calls round_pipe_size(), which rounds the
> > value up to pipe_min_size. Therefore, the second check against
> > pipe_min_size is redundant. Remove it.
>
> Well, it's not redundant: it provides a hint to anyone trying to tweak
> the sysctl about the minimum value. I think this should stay, but that
> pipe_min_size should be made const.
>
> -Kees
>
It *is* redundant, because it doesn't do anything. round_pipe_size() already
rounds the value up to the minimum.
Eric
Powered by blists - more mailing lists