[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20241217132908.38096-1-nicolas.bouchinet@clip-os.org>
Date: Tue, 17 Dec 2024 14:29:05 +0100
From: nicolas.bouchinet@...p-os.org
To: linux-kernel@...r.kernel.org,
linux-fsdevel@...r.kernel.org
Cc: Nicolas Bouchinet <nicolas.bouchinet@....gouv.fr>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Jiri Slaby <jirislaby@...nel.org>,
Alexander Viro <viro@...iv.linux.org.uk>,
Christian Brauner <brauner@...nel.org>,
Jan Kara <jack@...e.cz>,
Luis Chamberlain <mcgrof@...nel.org>,
Kees Cook <kees@...nel.org>,
Joel Granados <j.granados@...sung.com>,
Andrew Morton <akpm@...ux-foundation.org>,
Neil Horman <nhorman@...driver.com>,
Lin Feng <linf@...gsu.com>,
"Theodore Ts'o" <tytso@....edu>
Subject: [PATCH v3 0/2] Fixes multiple sysctl proc_handler usage error
From: Nicolas Bouchinet <nicolas.bouchinet@....gouv.fr>
Hi, while reading sysctl code I encountered two sysctl proc_handler
parameters common errors.
The first one is to declare .data as a different type thant the return of
the used .proc_handler, i.e. using proch_dointvec, thats convert a char
string to signed integers, and storing the result in a .data that is backed
by an unsigned int. User can then write "-1" string, which results in a
different value stored in the .data variable. This can lead to type
conversion errors in branches and thus to potential security issues.
>From a quick search using regex and only for proc_dointvec, this seems to
be a pretty common mistake.
The second one is to declare .extra1 or .extra2 values with a .proc_handler
that don't uses them. i.e, declaring .extra1 or .extra2 using proc_dointvec
in order to declare conversion bounds do not work as do_proc_dointvec don't
uses those variables if not explicitly asked.
This patchset corrects three sysctl declaration that are buggy as an
example and is not exhaustive.
Nicolas
---
Changes since v2:
https://lore.kernel.org/all/20241114162638.57392-1-nicolas.bouchinet@clip-os.org/
* Bound vdso_enabled to 0 and 1 as suggested by Joel Granados.
* Remove patch 3/3 since Greg Kroah-Hartman merged it.
Changes since v1:
https://lore.kernel.org/all/20241112131357.49582-1-nicolas.bouchinet@clip-os.org/
* Take Lin Feng review into account.
---
Nicolas Bouchinet (2):
coredump: Fixes core_pipe_limit sysctl proc_handler
sysctl: Fix underflow value setting risk in vm_table
fs/coredump.c | 4 +++-
kernel/sysctl.c | 3 ++-
2 files changed, 5 insertions(+), 2 deletions(-)
--
2.47.1
Powered by blists - more mailing lists