Linux Kernel
CVE-2025-39955
HIGH
Severity by source
AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Local unprivileged trigger (AV:L/PR:L) requiring a race against TFO handshake completion (AC:H); observed impact is a kernel WARNING and missed retransmission, so C:N/I:N with only low availability effect.
Primary rating from NVD.
CVSS VectorNVD
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Lifecycle Timeline
2DescriptionNVD
In the Linux kernel, the following vulnerability has been resolved:
tcp: Clear tcp_sk(sk)->fastopen_rsk in tcp_disconnect().
syzbot reported the splat below where a socket had tcp_sk(sk)->fastopen_rsk in the TCP_ESTABLISHED state. [0]
syzbot reused the server-side TCP Fast Open socket as a new client before the TFO socket completes 3WHS:
- accept()
- connect(AF_UNSPEC)
- connect() to another destination
As of accept(), sk->sk_state is TCP_SYN_RECV, and tcp_disconnect() changes it to TCP_CLOSE and makes connect() possible, which restarts timers.
Since tcp_disconnect() forgot to clear tcp_sk(sk)->fastopen_rsk, the retransmit timer triggered the warning and the intended packet was not retransmitted.
Let's call reqsk_fastopen_remove() in tcp_disconnect().
[0]: WARNING: CPU: 2 PID: 0 at net/ipv4/tcp_timer.c:542 tcp_retransmit_timer (net/ipv4/tcp_timer.c:542 (discriminator 7)) Modules linked in: CPU: 2 UID: 0 PID: 0 Comm: swapper/2 Not tainted 6.17.0-rc5-g201825fb4278 #62 PREEMPT(voluntary) Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014 RIP: 0010:tcp_retransmit_timer (net/ipv4/tcp_timer.c:542 (discriminator 7)) Code: 41 55 41 54 55 53 48 8b af b8 08 00 00 48 89 fb 48 85 ed 0f 84 55 01 00 00 0f b6 47 12 3c 03 74 0c 0f b6 47 12 3c 04 74 04 90 <0f> 0b 90 48 8b 85 c0 00 00 00 48 89 ef 48 8b 40 30 e8 6a 4f 06 3e RSP: 0018:ffffc900002f8d40 EFLAGS: 00010293 RAX: 0000000000000002 RBX: ffff888106911400 RCX: 0000000000000017 RDX: 0000000002517619 RSI: ffffffff83764080 RDI: ffff888106911400 RBP: ffff888106d5c000 R08: 0000000000000001 R09: ffffc900002f8de8 R10: 00000000000000c2 R11: ffffc900002f8ff8 R12: ffff888106911540 R13: ffff888106911480 R14: ffff888106911840 R15: ffffc900002f8de0 FS: 0000000000000000(0000) GS:ffff88907b768000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f8044d69d90 CR3: 0000000002c30003 CR4: 0000000000370ef0 Call Trace: <IRQ> tcp_write_timer (net/ipv4/tcp_timer.c:738) call_timer_fn (kernel/time/timer.c:1747) __run_timers (kernel/time/timer.c:1799 kernel/time/timer.c:2372) timer_expire_remote (kernel/time/timer.c:2385 kernel/time/timer.c:2376 kernel/time/timer.c:2135) tmigr_handle_remote_up (kernel/time/timer_migration.c:944 kernel/time/timer_migration.c:1035) __walk_groups.isra.0 (kernel/time/timer_migration.c:533 (discriminator 1)) tmigr_handle_remote (kernel/time/timer_migration.c:1096) handle_softirqs (./arch/x86/include/asm/jump_label.h:36 ./include/trace/events/irq.h:142 kernel/softirq.c:580) irq_exit_rcu (kernel/softirq.c:614 kernel/softirq.c:453 kernel/softirq.c:680 kernel/softirq.c:696) sysvec_apic_timer_interrupt (arch/x86/kernel/apic/apic.c:1050 (discriminator 35) arch/x86/kernel/apic/apic.c:1050 (discriminator 35)) </IRQ>
AnalysisAI
A local privilege-boundary/state-handling flaw in the Linux kernel's TCP stack lets an unprivileged user leave a stale TCP Fast Open request socket (fastopen_rsk) attached to a socket after tcp_disconnect(), which the retransmit timer later dereferences. By abusing accept() → connect(AF_UNSPEC) → connect() on a server-side TFO socket before the 3-way handshake completes, an attacker drives the socket into TCP_ESTABLISHED while retaining fastopen_rsk, triggering a kernel WARNING in tcp_retransmit_timer() and a failed retransmission. There is no public exploit identified at time of analysis, and the issue is not in CISA KEV; it was found by syzbot and resolved upstream by calling reqsk_fastopen_remove() in tcp_disconnect().
Technical ContextAI
The affected component is the IPv4 TCP implementation (net/ipv4/tcp.c and net/ipv4/tcp_timer.c) and specifically the TCP Fast Open (TFO) fast-path, where a server-side socket in TCP_SYN_RECV holds a child request socket pointer in tcp_sk(sk)->fastopen_rsk until the handshake completes. CPE data pins the affected code to the linux_kernel product across 6.17-rc1, rc2 and rc3 (with additional stable-branch backports implied by the multiple git.kernel.org commits). No CWE was assigned in the input; the root-cause class is best characterized as improper resource/pointer cleanup on a state transition (akin to CWE-459 Incomplete Cleanup, with use-after-free/stale-pointer potential, CWE-416), because tcp_disconnect() reset sk_state to TCP_CLOSE and restarted timers without clearing the still-populated fastopen_rsk, leaving the retransmit timer to act on an object that should have been removed.
RemediationAI
Patch available per vendor advisory: apply your distribution's updated kernel package or rebuild from a Linux stable tree that includes the fix commits at https://git.kernel.org/stable/c/17d699727577814198d744d6afe54735c6b54c99 through https://git.kernel.org/stable/c/fa4749c065644af4db496b338452a69a3e5147d9 (the fix calls reqsk_fastopen_remove() in tcp_disconnect()); Debian and other distros should track their security trackers for the rebuilt package version, and Siemens environments should follow SSA-019113 (https://cert-portal.siemens.com/productcert/html/ssa-019113.html). Exact upstream tagged release version is not stated in the input, so confirm the fixed package version with your distribution before deploying. As a compensating control until patched, disable server-side TCP Fast Open where it is not required by setting net.ipv4.tcp_fastopen to 0 (or to a value without the server bit, e.g. 1 for client-only), which removes the TFO code path this bug depends on at the cost of losing TFO latency benefits for accepting services; restricting untrusted local user access to the host also limits who can drive the required socket sequence.
Same technique Information Disclosure
View allShare
External POC / Exploit Code
Leaving vuln.today