From: James Chapman <jchapman@katalix.com>
To: netdev@vger.kernel.org
Cc: kbuild-all@01.org
Subject: [PATCH net-next v2 00/16] l2tp: fix API races discovered by syzbot
Date: Mon, 12 Feb 2018 10:11:04 +0000 [thread overview]
Message-ID: <1518430280-16671-1-git-send-email-jchapman@katalix.com> (raw)
This patch series addresses several races with L2TP APIs discovered by
syzbot. While working on this, it became clear that the L2TP code
needed some work to address object lifetime issues. There are no
functional changes.
The set of patches 1-13 in combination fix the following syzbot reports.
9df43faf0 KASAN: use-after-free Read in pppol2tp_connect
6e6a5ec8d general protection fault in pppol2tp_connect
347bd5acd KASAN: use-after-free Read in inet_shutdown
19c09769f WARNING in debug_print_object
In detail:-
1. Add RCU protection of sk_user_data. Since L2TP hooks on sockets
opened by userspace, we may race with other socket families that
attempt to use the same socket. (patches 1-2)
2. Fix inet_shutdown races when L2TP tunnels close. (patch 3)
3. Refactor code to address internal object lifetime
issues. Previously internal refcounts and socket refcounts were
used inconsistently and led to workarounds to fix specific
bugs. With the changes made here, we can now fetch the
tunnel/session context from its socket sk_user_data and fetch the
socket from the tunnel/session without using other APIs such as
sockfd_lookup. (patches 4-8)
4. Refactor pppol2tp_connect to fix several races and split it up to
improve readability. (patch 9)
5. Refactor session destroy paths to use a workqueue such that all
session cleanup is done using common code, regardless of whether
the session is closed by netlink request or (in the case of ppp)
its socket closed. (patches 10-13)
6. Misc cleanups made possible by the refactoring done in this
series. (patches 14-16)
Changes in v2:-
Fix compile error that would have broken bisect.
James Chapman (16):
l2tp: update sk_user_data while holding sk_callback_lock
l2tp: add RCU read lock to protect tunnel ptr in ip socket destroy
l2tp: don't use inet_shutdown on tunnel destroy
l2tp: refactor tunnel lifetime handling wrt its socket
l2tp: use tunnel closing flag
l2tp: refactor session lifetime handling
l2tp: hide sessions if they are closing
l2tp: hide session from pppol2tp_sock_to_session if it is closing
l2tp: refactor pppol2tp_connect
l2tp: add session_free callback
l2tp: do session destroy using a workqueue
l2tp: simplify l2tp_tunnel_closeall
l2tp: refactor ppp session cleanup paths
l2tp: remove redundant sk_user_data check when creating tunnels
l2tp: remove unwanted error message
l2tp: make __l2tp_session_unhash internal
net/l2tp/l2tp_core.c | 310 ++++++++++++++++++------------------
net/l2tp/l2tp_core.h | 37 ++---
net/l2tp/l2tp_ip.c | 10 +-
net/l2tp/l2tp_ip6.c | 8 +-
net/l2tp/l2tp_ppp.c | 434 ++++++++++++++++++++++++++++++---------------------
5 files changed, 434 insertions(+), 365 deletions(-)
--
1.9.1
next reply other threads:[~2018-02-12 10:11 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-02-12 10:11 James Chapman [this message]
2018-02-12 10:11 ` [PATCH net-next v2 01/16] l2tp: update sk_user_data while holding sk_callback_lock James Chapman
2018-02-12 16:21 ` David Miller
2018-02-12 18:33 ` Guillaume Nault
2018-02-12 10:11 ` [PATCH net-next v2 02/16] l2tp: add RCU read lock to protect tunnel ptr in ip socket destroy James Chapman
2018-02-12 16:22 ` David Miller
2018-02-12 18:35 ` Guillaume Nault
2018-02-12 10:11 ` [PATCH net-next v2 03/16] l2tp: don't use inet_shutdown on tunnel destroy James Chapman
2018-02-12 16:22 ` David Miller
2018-02-12 17:23 ` James Chapman
2018-02-12 18:41 ` Guillaume Nault
2018-02-12 10:11 ` [PATCH net-next v2 04/16] l2tp: refactor tunnel lifetime handling wrt its socket James Chapman
2018-02-12 18:48 ` Guillaume Nault
2018-02-15 8:23 ` kbuild test robot
2018-02-12 10:11 ` [PATCH net-next v2 05/16] l2tp: use tunnel closing flag James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 06/16] l2tp: refactor session lifetime handling James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 07/16] l2tp: hide sessions if they are closing James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 08/16] l2tp: hide session from pppol2tp_sock_to_session if it is closing James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 09/16] l2tp: refactor pppol2tp_connect James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 10/16] l2tp: add session_free callback James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 11/16] l2tp: do session destroy using a workqueue James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 12/16] l2tp: simplify l2tp_tunnel_closeall James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 13/16] l2tp: refactor ppp session cleanup paths James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 14/16] l2tp: remove redundant sk_user_data check when creating tunnels James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 15/16] l2tp: remove unwanted error message James Chapman
2018-02-12 10:11 ` [PATCH net-next v2 16/16] l2tp: make __l2tp_session_unhash internal James Chapman
2018-02-12 18:52 ` [PATCH net-next v2 00/16] l2tp: fix API races discovered by syzbot Guillaume Nault
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1518430280-16671-1-git-send-email-jchapman@katalix.com \
--to=jchapman@katalix.com \
--cc=kbuild-all@01.org \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).