From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Tejun Heo <tj@kernel.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Herbert Xu <herbert@gondor.apana.org.au>,
"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 4.1 26/29] netlink: Fix autobind race condition that leads to zero port ID
Date: Thu, 1 Oct 2015 11:32:02 +0200 [thread overview]
Message-ID: <20151001093146.852831059@linuxfoundation.org> (raw)
In-Reply-To: <20151001093145.730759857@linuxfoundation.org>
4.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Herbert Xu <herbert@gondor.apana.org.au>
[ Upstream commit 1f770c0a09da855a2b51af6d19de97fb955eca85 ]
The commit c0bb07df7d981e4091432754e30c9c720e2c0c78 ("netlink:
Reset portid after netlink_insert failure") introduced a race
condition where if two threads try to autobind the same socket
one of them may end up with a zero port ID. This led to kernel
deadlocks that were observed by multiple people.
This patch reverts that commit and instead fixes it by introducing
a separte rhash_portid variable so that the real portid is only set
after the socket has been successfully hashed.
Fixes: c0bb07df7d98 ("netlink: Reset portid after netlink_insert failure")
Reported-by: Tejun Heo <tj@kernel.org>
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
net/netlink/af_netlink.c | 11 ++++++-----
net/netlink/af_netlink.h | 1 +
2 files changed, 7 insertions(+), 5 deletions(-)
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -1017,7 +1017,7 @@ static inline int netlink_compare(struct
const struct netlink_compare_arg *x = arg->key;
const struct netlink_sock *nlk = ptr;
- return nlk->portid != x->portid ||
+ return nlk->rhash_portid != x->portid ||
!net_eq(sock_net(&nlk->sk), read_pnet(&x->pnet));
}
@@ -1043,7 +1043,7 @@ static int __netlink_insert(struct netli
{
struct netlink_compare_arg arg;
- netlink_compare_arg_init(&arg, sock_net(sk), nlk_sk(sk)->portid);
+ netlink_compare_arg_init(&arg, sock_net(sk), nlk_sk(sk)->rhash_portid);
return rhashtable_lookup_insert_key(&table->hash, &arg,
&nlk_sk(sk)->node,
netlink_rhashtable_params);
@@ -1105,7 +1105,7 @@ static int netlink_insert(struct sock *s
unlikely(atomic_read(&table->hash.nelems) >= UINT_MAX))
goto err;
- nlk_sk(sk)->portid = portid;
+ nlk_sk(sk)->rhash_portid = portid;
sock_hold(sk);
err = __netlink_insert(table, sk);
@@ -1117,10 +1117,11 @@ static int netlink_insert(struct sock *s
err = -EOVERFLOW;
if (err == -EEXIST)
err = -EADDRINUSE;
- nlk_sk(sk)->portid = 0;
sock_put(sk);
}
+ nlk_sk(sk)->portid = portid;
+
err:
release_sock(sk);
return err;
@@ -3167,7 +3168,7 @@ static inline u32 netlink_hash(const voi
const struct netlink_sock *nlk = data;
struct netlink_compare_arg arg;
- netlink_compare_arg_init(&arg, sock_net(&nlk->sk), nlk->portid);
+ netlink_compare_arg_init(&arg, sock_net(&nlk->sk), nlk->rhash_portid);
return jhash2((u32 *)&arg, netlink_compare_arg_len / sizeof(u32), seed);
}
--- a/net/netlink/af_netlink.h
+++ b/net/netlink/af_netlink.h
@@ -25,6 +25,7 @@ struct netlink_ring {
struct netlink_sock {
/* struct sock has to be the first member of netlink_sock */
struct sock sk;
+ u32 rhash_portid;
u32 portid;
u32 dst_portid;
u32 dst_group;
next prev parent reply other threads:[~2015-10-01 9:33 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-01 9:31 [PATCH 4.1 00/29] 4.1.10-stable review Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 01/29] ip6_gre: release cached dst on tunnel removal Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 02/29] vxlan: re-ignore EADDRINUSE from igmp_join Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 03/29] cls_u32: complete the check for non-forced case in u32_destroy() Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 04/29] usbnet: Get EVENT_NO_RUNTIME_PM bit before it is cleared Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 05/29] sock, diag: fix panic in sock_diag_put_filterinfo Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 06/29] ipv6: fix exthdrs offload registration in out_rt path Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 07/29] net: fec: clear receive interrupts before processing a packet Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 08/29] net: eth: altera: fix napi poll_list corruption Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 09/29] net/ipv6: Correct PIM6 mrt_lock handling Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 10/29] net: dsa: bcm_sf2: Fix ageing conditions and operation Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 11/29] ipv6: fix multipath route replace error recovery Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 12/29] net: dsa: bcm_sf2: Fix 64-bits register writes Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 13/29] netlink, mmap: transform mmap skb into full skb on taps Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 14/29] sctp: fix race on protocol/netns initialization Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 16/29] net/mlx4_en: really allow to change RSS key Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 17/29] macvtap: fix TUNSETSNDBUF values > 64k Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 18/29] openvswitch: Zero flows on allocation Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 19/29] tcp: add proper TS val into RST packets Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 20/29] net: revert "net_sched: move tp->root allocation into fw_init()" Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 21/29] fib_rules: fix fib rule dumps across multiple skbs Greg Kroah-Hartman
2015-10-01 9:31 ` [PATCH 4.1 22/29] net: dsa: bcm_sf2: Do not override speed settings Greg Kroah-Hartman
2015-10-02 0:47 ` Ben Hutchings
2015-10-02 1:22 ` David Miller
2015-10-03 11:38 ` Greg KH
2015-10-01 9:31 ` [PATCH 4.1 23/29] net: phy: fixed_phy: handle link-down case Greg Kroah-Hartman
2015-10-01 9:32 ` [PATCH 4.1 24/29] of_mdio: add new DT property managed to specify the PHY management type Greg Kroah-Hartman
2015-10-01 9:32 ` [PATCH 4.1 25/29] mvneta: use inband status only when explicitly enabled Greg Kroah-Hartman
2015-10-01 9:32 ` Greg Kroah-Hartman [this message]
2015-10-01 9:32 ` [PATCH 4.1 27/29] netlink: Replace rhash_portid with bound Greg Kroah-Hartman
2015-10-01 9:32 ` [PATCH 4.1 28/29] zram: fix possible use after free in zcomp_create() Greg Kroah-Hartman
2015-10-01 9:32 ` [PATCH 4.1 29/29] hp-wmi: limit hotkey enable Greg Kroah-Hartman
[not found] ` <20151001093146.372268230@linuxfoundation.org>
2015-10-01 23:13 ` [PATCH 4.1 15/29] bridge: fix igmpv3 / mldv2 report parsing Linus Lüssing
2015-10-02 1:29 ` [PATCH 4.1 00/29] 4.1.10-stable review Guenter Roeck
2015-10-02 6:23 ` Sudip Mukherjee
[not found] ` <560e9a58.e127b40a.bc060.fffffb97@mx.google.com>
2015-10-02 15:04 ` Kevin Hilman
2015-10-02 15:41 ` Shuah Khan
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=20151001093146.852831059@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=davem@davemloft.net \
--cc=herbert@gondor.apana.org.au \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.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).