kernel/include/net/inet_timewait_sock.h, branch linux-6.18.y

tcp: Update bind bucket state on port release

2025-09-23T08:12:15Z

Today, once an inet_bind_bucket enters a state where fastreuse >= 0 or fastreuseport >= 0 after a socket is explicitly bound to a port, it remains in that state until all sockets are removed and the bucket is destroyed. In this state, the bucket is skipped during ephemeral port selection in connect(). For applications using a reduced ephemeral port range (IP_LOCAL_PORT_RANGE socket option), this can cause faster port exhaustion since blocked buckets are excluded from reuse. The reason the bucket state isn't updated on port release is unclear. Possibly a performance trade-off to avoid scanning bucket owners, or just an oversight. Fix it by recalculating the bucket state when a socket releases a port. To limit overhead, each inet_bind2_bucket stores its own (fastreuse, fastreuseport) state. On port release, only the relevant port-addr bucket is scanned, and the overall state is derived from these. Signed-off-by: Jakub Sitnicki Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20250917-update-bind-bucket-state-on-unhash-v5-1-57168b661b47@cloudflare.com Signed-off-by: Paolo Abeni

net: tcp: allow tcp_timewait_sock to validate skbs before handing to device

2025-09-18T10:32:06Z

Provide a callback to validate skb's originating from tcp timewait socks before passing to the device layer. Full socks have a sk_validate_xmit_skb member for checking that a device is capable of performing offloads required for transmitting an skb. With psp, tcp timewait socks will inherit the crypto state from their corresponding full socks. Any ACKs or RSTs that originate from a tcp timewait sock carrying psp state should be psp encapsulated. Reviewed-by: Willem de Bruijn Signed-off-by: Daniel Zahka Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20250917000954.859376-8-daniel.zahka@gmail.com Signed-off-by: Paolo Abeni

net: modify core data structures for PSP datapath support

2025-09-18T10:32:06Z

Add pointers to psp data structures to core networking structs, and an SKB extension to carry the PSP information from the drivers to the socket layer. Reviewed-by: Willem de Bruijn Signed-off-by: Jakub Kicinski Co-developed-by: Daniel Zahka Signed-off-by: Daniel Zahka Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20250917000954.859376-4-daniel.zahka@gmail.com Signed-off-by: Paolo Abeni

tcp: Measure TIME-WAIT reuse delay with millisecond precision

2024-12-12T04:17:33Z

Prepare ground for TIME-WAIT socket reuse with subsecond delay. Today the last TS.Recent update timestamp, recorded in seconds and stored tp->ts_recent_stamp and tw->tw_ts_recent_stamp fields, has two purposes. Firstly, it is used to track the age of the last recorded TS.Recent value to detect when that value becomes outdated due to potential wrap-around of the other TCP timestamp clock (RFC 7323, section 5.5). For this purpose a second-based timestamp is completely sufficient as even in the worst case scenario of a peer using a high resolution microsecond timestamp, the wrap-around interval is ~36 minutes long. Secondly, it serves as a threshold value for allowing TIME-WAIT socket reuse. A TIME-WAIT socket can be reused only once the virtual 1 Hz clock, ktime_get_seconds, is past the TS.Recent update timestamp. The purpose behind delaying the TIME-WAIT socket reuse is to wait for the other TCP timestamp clock to tick at least once before reusing the connection. It is only then that the PAWS mechanism for the reopened connection can detect old duplicate segments from the previous connection incarnation (RFC 7323, appendix B.2). In this case using a timestamp with second resolution not only blocks the way toward allowing faster TIME-WAIT reuse after shorter subsecond delay, but also makes it impossible to reliably delay TW reuse by one second. As Eric Dumazet has pointed out [1], due to timestamp rounding, the TW reuse delay will actually be between (0, 1] seconds, and 0.5 seconds on average. We delay TW reuse for one full second only when last TS.Recent update coincides with our virtual 1 Hz clock tick. Considering the above, introduce a dedicated field to store a millisecond timestamp of transition into the TIME-WAIT state. Place it in an existing 4-byte hole inside inet_timewait_sock structure to avoid an additional memory cost. Use the new timestamp to (i) reliably delay TIME-WAIT reuse by one second, and (ii) prepare for configurable subsecond reuse delay in the subsequent change. We assume here that a full one second delay was the original intention in [2] because it accounts for the worst case scenario of the other TCP using the slowest recommended 1 Hz timestamp clock. A more involved alternative would be to change the resolution of the last TS.Recent update timestamp, tw->tw_ts_recent_stamp, to milliseconds. [1] https://lore.kernel.org/netdev/CANn89iKB4GFd8sVzCbRttqw_96o3i2wDhX-3DraQtsceNGYwug@mail.gmail.com/ [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b8439924316d5bcb266d165b93d632a4b4b859af Signed-off-by: Jakub Sitnicki Reviewed-by: Eric Dumazet Reviewed-by: Jason Xing Link: https://patch.msgid.link/20241209-jakub-krn-909-poc-msec-tw-tstamp-v2-1-66aca0eed03e@cloudflare.com Signed-off-by: Jakub Kicinski

tcp: populate XPS related fields of timewait sockets

2024-11-30T21:00:52Z

syzbot reported that netdev_core_pick_tx() was reading an uninitialized field [1]. This is indeed hapening for timewait sockets after recent commits. We can copy the original established socket sk_tx_queue_mapping and sk_rx_queue_mapping fields, instead of adding more checks in fast paths. As a bonus, packets will use the same transmit queue than prior ones, this potentially can avoid reordering. [1] BUG: KMSAN: uninit-value in netdev_pick_tx+0x5c7/0x1550 netdev_pick_tx+0x5c7/0x1550 netdev_core_pick_tx+0x1d2/0x4a0 net/core/dev.c:4312 __dev_queue_xmit+0x128a/0x57d0 net/core/dev.c:4394 dev_queue_xmit include/linux/netdevice.h:3168 [inline] neigh_hh_output include/net/neighbour.h:523 [inline] neigh_output include/net/neighbour.h:537 [inline] ip_finish_output2+0x187c/0x1b70 net/ipv4/ip_output.c:236 __ip_finish_output+0x287/0x810 ip_finish_output+0x4b/0x600 net/ipv4/ip_output.c:324 NF_HOOK_COND include/linux/netfilter.h:303 [inline] ip_output+0x15f/0x3f0 net/ipv4/ip_output.c:434 dst_output include/net/dst.h:450 [inline] ip_local_out net/ipv4/ip_output.c:130 [inline] ip_send_skb net/ipv4/ip_output.c:1505 [inline] ip_push_pending_frames+0x444/0x570 net/ipv4/ip_output.c:1525 ip_send_unicast_reply+0x18c1/0x1b30 net/ipv4/ip_output.c:1672 tcp_v4_send_reset+0x238d/0x2a40 net/ipv4/tcp_ipv4.c:910 tcp_v4_rcv+0x48f8/0x5750 net/ipv4/tcp_ipv4.c:2431 ip_protocol_deliver_rcu+0x2a3/0x13d0 net/ipv4/ip_input.c:205 ip_local_deliver_finish+0x336/0x500 net/ipv4/ip_input.c:233 NF_HOOK include/linux/netfilter.h:314 [inline] ip_local_deliver+0x21f/0x490 net/ipv4/ip_input.c:254 dst_input include/net/dst.h:460 [inline] ip_sublist_rcv_finish net/ipv4/ip_input.c:578 [inline] ip_list_rcv_finish net/ipv4/ip_input.c:628 [inline] ip_sublist_rcv+0x15f3/0x17f0 net/ipv4/ip_input.c:636 ip_list_rcv+0x9ef/0xa40 net/ipv4/ip_input.c:670 __netif_receive_skb_list_ptype net/core/dev.c:5715 [inline] __netif_receive_skb_list_core+0x15c5/0x1670 net/core/dev.c:5762 __netif_receive_skb_list net/core/dev.c:5814 [inline] netif_receive_skb_list_internal+0x1085/0x1700 net/core/dev.c:5905 gro_normal_list include/net/gro.h:515 [inline] napi_complete_done+0x3d4/0x810 net/core/dev.c:6256 virtqueue_napi_complete drivers/net/virtio_net.c:758 [inline] virtnet_poll+0x5d80/0x6bf0 drivers/net/virtio_net.c:3013 __napi_poll+0xe7/0x980 net/core/dev.c:6877 napi_poll net/core/dev.c:6946 [inline] net_rx_action+0xa5a/0x19b0 net/core/dev.c:7068 handle_softirqs+0x1a0/0x7c0 kernel/softirq.c:554 __do_softirq kernel/softirq.c:588 [inline] invoke_softirq kernel/softirq.c:428 [inline] __irq_exit_rcu+0x68/0x180 kernel/softirq.c:655 irq_exit_rcu+0x12/0x20 kernel/softirq.c:671 common_interrupt+0x97/0xb0 arch/x86/kernel/irq.c:278 asm_common_interrupt+0x2b/0x40 arch/x86/include/asm/idtentry.h:693 __preempt_count_sub arch/x86/include/asm/preempt.h:84 [inline] kmsan_virt_addr_valid arch/x86/include/asm/kmsan.h:95 [inline] virt_to_page_or_null+0xfb/0x150 mm/kmsan/shadow.c:75 kmsan_get_metadata+0x13e/0x1c0 mm/kmsan/shadow.c:141 kmsan_get_shadow_origin_ptr+0x4d/0xb0 mm/kmsan/shadow.c:102 get_shadow_origin_ptr mm/kmsan/instrumentation.c:38 [inline] __msan_metadata_ptr_for_store_4+0x27/0x40 mm/kmsan/instrumentation.c:93 rcu_preempt_read_enter kernel/rcu/tree_plugin.h:390 [inline] __rcu_read_lock+0x46/0x70 kernel/rcu/tree_plugin.h:413 rcu_read_lock include/linux/rcupdate.h:847 [inline] batadv_nc_purge_orig_hash net/batman-adv/network-coding.c:408 [inline] batadv_nc_worker+0x114/0x19e0 net/batman-adv/network-coding.c:719 process_one_work kernel/workqueue.c:3229 [inline] process_scheduled_works+0xae0/0x1c40 kernel/workqueue.c:3310 worker_thread+0xea7/0x14f0 kernel/workqueue.c:3391 kthread+0x3e2/0x540 kernel/kthread.c:389 ret_from_fork+0x6d/0x90 arch/x86/kernel/process.c:147 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244 Uninit was created at: __alloc_pages_noprof+0x9a7/0xe00 mm/page_alloc.c:4774 alloc_pages_mpol_noprof+0x299/0x990 mm/mempolicy.c:2265 alloc_pages_noprof+0x1bf/0x1e0 mm/mempolicy.c:2344 alloc_slab_page mm/slub.c:2412 [inline] allocate_slab+0x320/0x12e0 mm/slub.c:2578 new_slab mm/slub.c:2631 [inline] ___slab_alloc+0x12ef/0x35e0 mm/slub.c:3818 __slab_alloc mm/slub.c:3908 [inline] __slab_alloc_node mm/slub.c:3961 [inline] slab_alloc_node mm/slub.c:4122 [inline] kmem_cache_alloc_noprof+0x57a/0xb20 mm/slub.c:4141 inet_twsk_alloc+0x11f/0x9d0 net/ipv4/inet_timewait_sock.c:188 tcp_time_wait+0x83/0xf50 net/ipv4/tcp_minisocks.c:309 tcp_rcv_state_process+0x145a/0x49d0 tcp_v4_do_rcv+0xbf9/0x11a0 net/ipv4/tcp_ipv4.c:1939 tcp_v4_rcv+0x51df/0x5750 net/ipv4/tcp_ipv4.c:2351 ip_protocol_deliver_rcu+0x2a3/0x13d0 net/ipv4/ip_input.c:205 ip_local_deliver_finish+0x336/0x500 net/ipv4/ip_input.c:233 NF_HOOK include/linux/netfilter.h:314 [inline] ip_local_deliver+0x21f/0x490 net/ipv4/ip_input.c:254 dst_input include/net/dst.h:460 [inline] ip_sublist_rcv_finish net/ipv4/ip_input.c:578 [inline] ip_list_rcv_finish net/ipv4/ip_input.c:628 [inline] ip_sublist_rcv+0x15f3/0x17f0 net/ipv4/ip_input.c:636 ip_list_rcv+0x9ef/0xa40 net/ipv4/ip_input.c:670 __netif_receive_skb_list_ptype net/core/dev.c:5715 [inline] __netif_receive_skb_list_core+0x15c5/0x1670 net/core/dev.c:5762 __netif_receive_skb_list net/core/dev.c:5814 [inline] netif_receive_skb_list_internal+0x1085/0x1700 net/core/dev.c:5905 gro_normal_list include/net/gro.h:515 [inline] napi_complete_done+0x3d4/0x810 net/core/dev.c:6256 virtqueue_napi_complete drivers/net/virtio_net.c:758 [inline] virtnet_poll+0x5d80/0x6bf0 drivers/net/virtio_net.c:3013 __napi_poll+0xe7/0x980 net/core/dev.c:6877 napi_poll net/core/dev.c:6946 [inline] net_rx_action+0xa5a/0x19b0 net/core/dev.c:7068 handle_softirqs+0x1a0/0x7c0 kernel/softirq.c:554 __do_softirq kernel/softirq.c:588 [inline] invoke_softirq kernel/softirq.c:428 [inline] __irq_exit_rcu+0x68/0x180 kernel/softirq.c:655 irq_exit_rcu+0x12/0x20 kernel/softirq.c:671 common_interrupt+0x97/0xb0 arch/x86/kernel/irq.c:278 asm_common_interrupt+0x2b/0x40 arch/x86/include/asm/idtentry.h:693 CPU: 0 UID: 0 PID: 3962 Comm: kworker/u8:18 Not tainted 6.12.0-syzkaller-09073-g9f16d5e6f220 #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/13/2024 Workqueue: bat_events batadv_nc_worker Fixes: 79636038d37e ("ipv4: tcp: give socket pointer to control skbs") Fixes: 507a96737d99 ("ipv6: tcp: give socket pointer to control skbs") Reported-by: syzbot+8b0959fc16551d55896b@syzkaller.appspotmail.com Link: https://lore.kernel.org/netdev/674442bd.050a0220.1cc393.0072.GAE@google.com/T/#u Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: Brian Vazquez Link: https://patch.msgid.link/20241125093039.3095790-1-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: remove volatile qualifier on tw_substate

2024-08-29T00:08:16Z

Using a volatile qualifier for a specific struct field is unusual. Use instead READ_ONCE()/WRITE_ONCE() where necessary. tcp_timewait_state_process() can change tw_substate while other threads are reading this field. Signed-off-by: Eric Dumazet Reviewed-by: Jason Xing Link: https://patch.msgid.link/20240827015250.3509197-2-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: move inet_twsk_schedule helper out of header

2024-06-10T10:54:18Z

Its no longer used outside inet_timewait_sock.c, so move it there. Reviewed-by: Eric Dumazet Signed-off-by: Florian Westphal Signed-off-by: David S. Miller

net: tcp/dccp: prepare for tw_timer un-pinning

2024-06-10T10:54:18Z

The TCP timewait timer is proving to be problematic for setups where scheduler CPU isolation is achieved at runtime via cpusets (as opposed to statically via isolcpus=domains). What happens there is a CPU goes through tcp_time_wait(), arming the time_wait timer, then gets isolated. TCP_TIMEWAIT_LEN later, the timer fires, causing interference for the now-isolated CPU. This is conceptually similar to the issue described in commit e02b93124855 ("workqueue: Unbind kworkers before sending them to exit()") Move inet_twsk_schedule() to within inet_twsk_hashdance(), with the ehash lock held. Expand the lock's critical section from inet_twsk_kill() to inet_twsk_deschedule_put(), serializing the scheduling vs descheduling of the timer. IOW, this prevents the following race: tcp_time_wait() inet_twsk_hashdance() inet_twsk_deschedule_put() del_timer_sync() inet_twsk_schedule() Thanks to Paolo Abeni for suggesting to leverage the ehash lock. This also restores a comment from commit ec94c2696f0b ("tcp/dccp: avoid one atomic operation for timewait hashdance") as inet_twsk_hashdance() had a "Step 1" and "Step 3" comment, but the "Step 2" had gone missing. inet_twsk_deschedule_put() now acquires the ehash spinlock to synchronize with inet_twsk_hashdance_schedule(). To ease possible regression search, actual un-pin is done in next patch. Link: https://lore.kernel.org/all/ZPhpfMjSiHVjQkTk@localhost.localdomain/ Reviewed-by: Eric Dumazet Signed-off-by: Valentin Schneider Co-developed-by: Florian Westphal Signed-off-by: Florian Westphal Signed-off-by: David S. Miller

tcp/dccp: do not care about families in inet_twsk_purge()

2024-04-02T04:27:58Z

We lost ability to unload ipv6 module a long time ago. Instead of calling expensive inet_twsk_purge() twice, we can handle all families in one round. Also remove an extra line added in my prior patch, per Kuniyuki Iwashima feedback. Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/netdev/20240327192934.6843-1-kuniyu@amazon.com/ Reviewed-by: Kuniyuki Iwashima Link: https://lore.kernel.org/r/20240329153203.345203-1-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: Remove dead code and fields for bhash2.

2023-12-22T22:15:35Z

Now all sockets including TIME_WAIT are linked to bhash2 using sock_common.skc_bind_node. We no longer use inet_bind2_bucket.deathrow, sock.sk_bind2_node, and inet_timewait_sock.tw_bind2_node. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller