In the Linux kernel, the following vulnerability has been resolved:
bonding: fix missed rcu protection
When removing the rcu_read_lock in bond_ethtool_get_ts_info() as
discussed [1], I didn't notice it could be called via setsockopt,
which doesn't hold rcu lock, as syzbot pointed:
stack backtrace:
CPU: 0 PID: 3599 Comm: syz-executor317 Not tainted 5.18.0-rc5-syzkaller-01392-g01f4685797a5 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
bond_option_active_slave_get_rcu include/net/bonding.h:353 [inline]
bond_ethtool_get_ts_info+0x32c/0x3a0 drivers/net/bonding/bond_main.c:5595
__ethtool_get_ts_info+0x173/0x240 net/ethtool/common.c:554
ethtool_get_phc_vclocks+0x99/0x110 net/ethtool/common.c:568
sock_timestamping_bind_phc net/core/sock.c:869 [inline]
sock_set_timestamping+0x3a3/0x7e0 net/core/sock.c:916
sock_setsockopt+0x543/0x2ec0 net/core/sock.c:1221
__sys_setsockopt+0x55e/0x6a0 net/socket.c:2223
__do_sys_setsockopt net/socket.c:2238 [inline]
__se_sys_setsockopt net/socket.c:2235 [inline]
__x64_sys_setsockopt+0xba/0x150 net/socket.c:2235
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f8902c8eb39
Fix it by adding rcu_read_lock and take a ref on the real_dev.
Since dev_hold() and dev_put() can take NULL these days, we can
skip checking if real_dev exist.
[1] https://lore.kernel.org/netdev/27565.1642742439@famine/
Metrics
Affected Vendors & Products
References
History
Thu, 01 May 2025 03:00:00 +0000
Type | Values Removed | Values Added |
---|---|---|
Weaknesses | CWE-413 |
Thu, 27 Feb 2025 13:30:00 +0000
Type | Values Removed | Values Added |
---|---|---|
References |
| |
Metrics |
threat_severity
|
cvssV3_1
|
Wed, 26 Feb 2025 02:45:00 +0000
Type | Values Removed | Values Added |
---|---|---|
Description | In the Linux kernel, the following vulnerability has been resolved: bonding: fix missed rcu protection When removing the rcu_read_lock in bond_ethtool_get_ts_info() as discussed [1], I didn't notice it could be called via setsockopt, which doesn't hold rcu lock, as syzbot pointed: stack backtrace: CPU: 0 PID: 3599 Comm: syz-executor317 Not tainted 5.18.0-rc5-syzkaller-01392-g01f4685797a5 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: <TASK> __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106 bond_option_active_slave_get_rcu include/net/bonding.h:353 [inline] bond_ethtool_get_ts_info+0x32c/0x3a0 drivers/net/bonding/bond_main.c:5595 __ethtool_get_ts_info+0x173/0x240 net/ethtool/common.c:554 ethtool_get_phc_vclocks+0x99/0x110 net/ethtool/common.c:568 sock_timestamping_bind_phc net/core/sock.c:869 [inline] sock_set_timestamping+0x3a3/0x7e0 net/core/sock.c:916 sock_setsockopt+0x543/0x2ec0 net/core/sock.c:1221 __sys_setsockopt+0x55e/0x6a0 net/socket.c:2223 __do_sys_setsockopt net/socket.c:2238 [inline] __se_sys_setsockopt net/socket.c:2235 [inline] __x64_sys_setsockopt+0xba/0x150 net/socket.c:2235 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x7f8902c8eb39 Fix it by adding rcu_read_lock and take a ref on the real_dev. Since dev_hold() and dev_put() can take NULL these days, we can skip checking if real_dev exist. [1] https://lore.kernel.org/netdev/27565.1642742439@famine/ | |
Title | bonding: fix missed rcu protection | |
References |
|

Status: PUBLISHED
Assigner: Linux
Published: 2025-02-26T02:13:04.787Z
Updated: 2025-05-04T08:38:08.212Z
Reserved: 2025-02-26T02:08:31.574Z
Link: CVE-2022-49456

No data.

Status : Received
Published: 2025-02-26T07:01:21.950
Modified: 2025-02-26T07:01:21.950
Link: CVE-2022-49456
