Acknowledged
Created: Jun 18, 2025
Updated: Sep 4, 2026
Resolved Date: Jun 19, 2025
Found In Version: 10.25.33.1
Severity: Standard
Applicable for: Wind River Linux LTS 25
Component/s: Kernel
In the Linux kernel, the following vulnerability has been resolved:EOL][EOL]net: fix refcount bug in sk_psock_get (2)[EOL][EOL]Syzkaller reports refcount bug as follows:[EOL]------------[ cut here ]------------[EOL]refcount_t: saturated; leaking memory.[EOL]WARNING: CPU: 1 PID: 3605 at lib/refcount.c:19 refcount_warn_saturate+0xf4/0x1e0 lib/refcount.c:19[EOL]Modules linked in:[EOL]CPU: 1 PID: 3605 Comm: syz-executor208 Not tainted 5.18.0-syzkaller-03023-g7e062cda7d90 #0[EOL] <TASK>[EOL] __refcount_add_not_zero include/linux/refcount.h:163 [inline][EOL] __refcount_inc_not_zero include/linux/refcount.h:227 [inline][EOL] refcount_inc_not_zero include/linux/refcount.h:245 [inline][EOL] sk_psock_get+0x3bc/0x410 include/linux/skmsg.h:439[EOL] tls_data_ready+0x6d/0x1b0 net/tls/tls_sw.c:2091[EOL] tcp_data_ready+0x106/0x520 net/ipv4/tcp_input.c:4983[EOL] tcp_data_queue+0x25f2/0x4c90 net/ipv4/tcp_input.c:5057[EOL] tcp_rcv_state_process+0x1774/0x4e80 net/ipv4/tcp_input.c:6659[EOL] tcp_v4_do_rcv+0x339/0x980 net/ipv4/tcp_ipv4.c:1682[EOL] sk_backlog_rcv include/net/sock.h:1061 [inline][EOL] __release_sock+0x134/0x3b0 net/core/sock.c:2849[EOL] release_sock+0x54/0x1b0 net/core/sock.c:3404[EOL] inet_shutdown+0x1e0/0x430 net/ipv4/af_inet.c:909[EOL] __sys_shutdown_sock net/socket.c:2331 [inline][EOL] __sys_shutdown_sock net/socket.c:2325 [inline][EOL] __sys_shutdown+0xf1/0x1b0 net/socket.c:2343[EOL] __do_sys_shutdown net/socket.c:2351 [inline][EOL] __se_sys_shutdown net/socket.c:2349 [inline][EOL] __x64_sys_shutdown+0x50/0x70 net/socket.c:2349[EOL] do_syscall_x64 arch/x86/entry/common.c:50 [inline][EOL] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80[EOL] entry_SYSCALL_64_after_hwframe+0x46/0xb0[EOL] </TASK>[EOL][EOL]During SMC fallback process in connect syscall, kernel will[EOL]replaces TCP with SMC. In order to forward wakeup[EOL]smc socket waitqueue after fallback, kernel will sets[EOL]clcsk->sk_user_data to origin smc socket in[EOL]smc_fback_replace_callbacks().[EOL][EOL]Later, in shutdown syscall, kernel will calls[EOL]sk_psock_get(), which treats the clcsk->sk_user_data[EOL]as psock type, triggering the refcnt warning.[EOL][EOL]So, the root cause is that smc and psock, both will use[EOL]sk_user_data field. So they will mismatch this field[EOL]easily.[EOL][EOL]This patch solves it by using another bit(defined as[EOL]SK_USER_DATA_PSOCK) in PTRMASK, to mark whether[EOL]sk_user_data points to a psock object or not.[EOL]This patch depends on a PTRMASK introduced in commit f1ff5ce2cd5e[EOL]("net, sk_msg: Clear sk_user_data pointer on clone if tagged").[EOL][EOL]For there will possibly be more flags in the sk_user_data field,[EOL]this patch also refactor sk_user_data flags code to be more generic[EOL]to improve its maintainability.
CREATE(Triage):(User=lchen-cn) [CVE-2022-49979 (https://nvd.nist.gov/vuln/detail/CVE-2022-49979)