Sashiko warns that local attacker can modify the packet
while it is processed by IPVS. Some places read the
IP ihl field multiple times which can cause out-of-bounds
access. One such place is ip_vs_nat_icmp where we
can write after the validated area.
Fix it by providing ciph argument just like it is done for
IPv6 and use ciph->len as offset to the embedded transport
header.
Modify some IPv4 header checks by reading the ihl field
only once.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Link: https://sashiko.dev/#/patchset/20260722101517.36313-1-ja%40ssi.bg
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
IPVS estimator kthread 0 starts with zeroed chain and tick limits until
its initial calculation phase completes. If network namespace teardown
clears ipvs->enable during that phase, ip_vs_est_calc_phase() can return
without installing positive limits.
The kthread can then continue into its main loop and drain
est_temp_list with zero chain_max, tick_max and est_max_count values.
Each enqueue consumes one available tick row, but est_count never
reaches the zero est_max_count value. After all rows are consumed, the
row lookup returns IPVS_EST_NTICKS and ip_vs_enqueue_estimator() writes
past the ticks and tick_len arrays.
Exit kthread 0 after the calculation phase if the kthread is stopping or
IPVS has been disabled. That keeps temporary estimators from being
drained after the limits failed to initialize.
Estimator kthreads can now self-exit before teardown or reload stops
kd->task. Keep an extra task reference after creation and release it
with kthread_stop_put(), so kd->task remains valid until the stop paths
consume that reference.
Fixes: 705dd34440 ("ipvs: use kthreads for stats estimation")
Cc: [email protected]
Reported-by: Vega <[email protected]>
Signed-off-by: Zhiling Zou <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Sashiko warns that ip_vs_nat_icmp() unconditionally mangles the
payload for embedded non-first IPv4 fragments. The problem is
in the very old inverted pp->dont_defrag check which should not
continue when embedded is a non-first TCP/UDP/SCTP fragment.
Check for embedded non-first fragment is also missing from
ip_vs_out_icmp_v6(), it is needed before any connection
lookups that expect ports after the network headers.
Drop the blocking code from ip_vs_in_icmp_v6() which prevents
ICMPv6 from local clients to use non-MASQ forwarding.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Link: https://sashiko.dev/#/patchset/20260720201122.79882-1-ja%40ssi.bg
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
The offsets we use to packet headers and payloads should be
based on skb->data. We even already respect non-zero
network offset in ip_vs_fill_iph_skb() but some places
do it wrongly and support only zero offset which is expected
for the IP layer where IPVS has hooks.
Change all places that instead of skb->data use offsets based
on the network header (skb_network_header, ip_hdr, etc) because
this doubles the network offset as noted by Sashiko.
For ip_vs_nat_icmp_v6() we can even rely on the IPv6 header
parsing done by the caller.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Link: https://sashiko.dev/#/patchset/20260710143733.29741-2-fw%40strlen.de
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
ip_vs_in_icmp_v6() is missing checksum validation for ICMPv6
packets from clients. In fact, as for TCP/UDP we should
validate the checksum for ICMP packets only when we
mangle the packets on MASQ or on reply for tunnel.
Also, Sashiko points out that handle_response_icmp() being
common for IPv4 and IPv6 is missing the pseudo-header
calculation while validating ICMPv6 messages from real
servers which is a problem if checksum is not validated
by the hardware.
Fix the problems by creating ip_vs_checksum_common_check()
helper and use it for TCP/UDP/ICMP both for IPv4 and IPv6.
Rely on the nf_checksum() for validating the ICMP messages
but use it also for TCP and UDP.
Use correct IP offset for IP_VS_DBG_RL_PKT for TCP/UDP/SCTP.
IPVS packets (TCP/UDP/SCTP/ICMP) do not need checksum
validation on LOCAL_OUT (local clients or local real
servers) and on FORWARD (traffic from servers on LAN).
Do it only on LOCAL_IN, in case nf_checksum() is not
called on PRE_ROUTING.
Also, ip_vs_checksum_complete() can be marked static.
Fixes: 2a3b791e6e ("IPVS: Add/adjust Netfilter hook functions and helpers for v6")
Link: https://sashiko.dev/#/patchset/20260708180315.77413-1-ja%40ssi.bg
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Synced conns can be created with one forwarding method
and later updated with different one after the dest
server is configured. This needs adjusting the hashing
for node hn1 because only MASQ supports double hashing.
Modify conn_tab_lock() to support seeking for hash node
hn0 together with adding for hn1. By this way we can
safely modify the forwarding method and hn1.hash_key
under bucket lock for the first node hn0. The forwarding
method is also protected by cp->lock as it is part of
cp->flags.
Fix the usage of stale idx/idx2 values in conn_tab_lock
after jumping to the retry label. Instead, use idx/idx2
values just to order the locking for the old/new tables.
Reported-by: Zhiling Zou <[email protected]>
Link: https://patch.msgid.link/1b914f41d725bc064c9ba9830dc8169329737270.1782540466.git.roxy520tt@gmail.com/
Link: https://sashiko.dev/#/patchset/CALMqdkR704S2BG_QD_bgHTFp2%2B1QCi7n0T4zoZyTo8mDZevYSA%40mail.gmail.com
Fixes: f20c73b046 ("ipvs: use more keys for connection hashing")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Synced connections can be created before their destination exists. When
the destination is later added, ip_vs_bind_dest() copies connection flags
from the destination into cp->flags.
IP_VS_CONN_F_ONE_PACKET connections are not synced. If a synced
connection inherits IP_VS_CONN_F_ONE_PACKET while it is already hashed,
expiry can treat it as a one-packet connection and skip unlinking the
existing conn_tab node, leaving stale hash nodes pointing at a freed
struct ip_vs_conn.
Drop IP_VS_CONN_F_ONE_PACKET from destination flags when binding synced
connections.
Fixes: 26ec037f98 ("IPVS: one-packet scheduling")
Cc: [email protected]
Reported-by: Yuan Tan <[email protected]>
Reported-by: Yifan Wu <[email protected]>
Reported-by: Juefei Pu <[email protected]>
Reported-by: Xin Liu <[email protected]>
Suggested-by: Julian Anastasov <[email protected]>
Signed-off-by: Zhiling Zou <[email protected]>
Signed-off-by: Ren Wei <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Sashiko reports for more incorrect IPv6 transport offsets.
The app code for TCP was assuming IPv4 network header
even after the ipvsh argument was provided. This can
cause problems with apps over IPv6. As for the only
official app in the kernel tree (FTP) this problem is
harmless because we use Netfilter to mangle the FTP
ports and we do not adjust the TCP seq numbers.
Also, provide correct offset of the ICMPV6 header in
ip_vs_out_icmp_v6() for correct checksum checks when
the IPv6 packet has extension headers.
Fixes: d12e12299a ("ipvs: add ipv6 support to ftp")
Fixes: 2a3b791e6e ("IPVS: Add/adjust Netfilter hook functions and helpers for v6")
Cc: [email protected]
Link: https://sashiko.dev/#/patchset/20260706101624.69471-1-zhaoyz24%40mails.tsinghua.edu.cn
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Sashiko points out that after stripping the outer headers
with pskb_pull() we should ensure the inner IP headers
in ICMP errors from tunnels are present in the skb headroom
for functions like ipv4_update_pmtu(), icmp_send() and
IP_VS_DBG().
Also, add more checks for the length of the inner headers.
Fixes: f2edb9f770 ("ipvs: implement passive PMTUD for IPIP packets")
Link: https://sashiko.dev/#/patchset/20260702073430.67680-1-zhaoyz24%40mails.tsinghua.edu.cn
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
set_sctp_state() reads the SCTP chunk header again in order to drive the
IPVS SCTP state table. For IPv6 it computes the offset with
sizeof(struct ipv6hdr), while the surrounding IPVS code uses iph.len from
ip_vs_fill_iph_skb(), where ipv6_find_hdr() has already skipped
extension headers and found the real transport header.
This makes the state machine read from the wrong offset for IPv6 SCTP
packets that carry extension headers. For example, an INIT packet with an
8-byte destination options header can be scheduled correctly by
sctp_conn_schedule(), but set_sctp_state() reads the first byte of the
SCTP verification tag as a DATA chunk type. The connection then moves
from NONE to ESTABLISHED instead of INIT1, gets the longer established
timeout, and updates the active/inactive destination counters
incorrectly. This happens even though the SCTP handshake has not
completed.
Use the parsed transport offset passed down from ip_vs_set_state() for
the SCTP chunk-header lookup. For IPv4 and IPv6 packets without
extension headers this preserves the existing offset.
Fixes: 2906f66a56 ("ipvs: SCTP Trasport Loadbalancing Support")
Cc: [email protected]
Link: https://lore.kernel.org/netdev/[email protected]/
Reported-by: Yizhou Zhao <[email protected]>
Reported-by: Yuxiang Yang <[email protected]>
Reported-by: Ao Wang <[email protected]>
Reported-by: Xuewei Feng <[email protected]>
Reported-by: Qi Li <[email protected]>
Reported-by: Ke Xu <[email protected]>
Assisted-by: Claude Code:GLM-5.2
Signed-off-by: Yizhou Zhao <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
TCP state handling reparses the skb to find the TCP header. For IPv6 it
uses sizeof(struct ipv6hdr), while the surrounding IPVS code already
parsed the packet with ip_vs_fill_iph_skb() and has the real
transport-header offset in iph.len.
This makes TCP state handling look at the wrong bytes when an IPv6
packet carries extension headers. Use the parsed transport offset passed
down from ip_vs_set_state() when reading the TCP header.
For IPv4 and for IPv6 packets without extension headers, the passed
offset matches the previous value.
Fixes: 0bbdd42b7e ("IPVS: Extend protocol DNAT/SNAT and state handlers")
Link: https://lore.kernel.org/netdev/[email protected]/
Reported-by: Yizhou Zhao <[email protected]>
Reported-by: Yuxiang Yang <[email protected]>
Reported-by: Ao Wang <[email protected]>
Reported-by: Xuewei Feng <[email protected]>
Reported-by: Qi Li <[email protected]>
Reported-by: Ke Xu <[email protected]>
Assisted-by: Claude Code:GLM-5.2
Signed-off-by: Yizhou Zhao <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
IPVS callers already parse the packet into struct ip_vs_iphdr before
updating connection state. For IPv6 this records the real
transport-header offset after extension headers in iph.len.
Pass this parsed transport offset through ip_vs_set_state() and the
protocol state_transition() callback so protocol handlers can use the
same packet context as scheduling and NAT handling. This patch only
changes the common callback plumbing and adapts the protocol callback
signatures; TCP and SCTP start using the value in follow-up patches.
Signed-off-by: Yizhou Zhao <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Commit 9a05475ceb ("ipvs: avoid kmem_cache_zalloc in
ip_vs_conn_new") changed ip_vs_conn_new() to allocate an ip_vs_conn
object with kmem_cache_alloc(). The function then initializes many
fields explicitly, but only resets in_seq.delta and out_seq.delta in the
two struct ip_vs_seq members.
That leaves init_seq and previous_delta uninitialized. This is normally
harmless while the corresponding IP_VS_CONN_F_IN_SEQ or
IP_VS_CONN_F_OUT_SEQ flag is clear. For connections learned from a sync
message, however, ip_vs_proc_conn() preserves those flags from
IP_VS_CONN_F_BACKUP_MASK and passes opt=NULL when the message omits
IPVS_OPT_SEQ_DATA. In that case the new connection can be hashed with
SEQ flags set but with the rest of in_seq/out_seq still containing stale
slab data.
When a packet for such a connection is later handled by an IPVS
application helper, vs_fix_seq() and vs_fix_ack_seq() use
previous_delta and init_seq to rewrite TCP sequence numbers. A malformed
sync message can therefore make forwarded packets carry stale slab bytes
in their TCP seq/ack numbers, and can also corrupt the forwarded TCP
flow.
Reset both struct ip_vs_seq members completely before publishing the
connection. This matches the existing "reset struct ip_vs_seq" comment
and keeps the sequence-adjustment gates inactive unless valid sequence
data is installed later.
Fixes: 9a05475ceb ("ipvs: avoid kmem_cache_zalloc in ip_vs_conn_new")
Cc: [email protected]
Reported-by: Yizhou Zhao <[email protected]>
Reported-by: Yuxiang Yang <[email protected]>
Reported-by: Ao Wang <[email protected]>
Reported-by: Xuewei Feng <[email protected]>
Reported-by: Qi Li <[email protected]>
Reported-by: Ke Xu <[email protected]>
Assisted-by: Claude-Code:GLM-5.2
Signed-off-by: Yizhou Zhao <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
When an ICMP Fragmentation Needed error is received for a tunneled IPVS
connection, ip_vs_in_icmp() recomputes the MTU that the original packet
can use by subtracting the tunnel overhead from the reported next-hop
MTU.
The current code always subtracts sizeof(struct iphdr), which is only
the IPIP overhead. For GUE and GRE tunnels, ipvs_udp_decap() and
ipvs_gre_decap() already compute the additional tunnel header length,
but that value is scoped to the decapsulation block and is lost before
the ICMP_FRAG_NEEDED handling. As a result, the ICMP error sent back to
the client advertises an MTU that is too large, so PMTUD can fail to
converge for GUE/GRE-tunneled real servers.
With a reported next-hop MTU of 1400, a GUE tunnel currently returns
1380 to the client. The correct value is 1368:
1400 - sizeof(struct iphdr) - sizeof(struct udphdr) -
sizeof(struct guehdr)
Hoist the tunnel header length into the main ip_vs_in_icmp() scope and
subtract sizeof(struct iphdr) + ulen in the Fragmentation Needed path.
The IPIP path keeps ulen as 0, so its existing 1400 - 20 = 1380 result
is unchanged.
Fixes: 508f744c0d ("ipvs: strip udp tunnel headers from icmp errors")
Cc: [email protected]
Reported-by: Yizhou Zhao <[email protected]>
Reported-by: Yuxiang Yang <[email protected]>
Reported-by: Ao Wang <[email protected]>
Reported-by: Xuewei Feng <[email protected]>
Reported-by: Qi Li <[email protected]>
Reported-by: Ke Xu <[email protected]>
Assisted-by: Claude-Code:GLM-5.2
Signed-off-by: Yizhou Zhao <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
This patch continues the effort to refactor workqueue APIs, which has
begun with the changes introducing new workqueues and a new
alloc_workqueue flag:
commit 128ea9f6cc ("workqueue: Add system_percpu_wq and system_dfl_wq")
commit 930c2ea566 ("workqueue: Add new WQ_PERCPU flag")
The point of the refactoring is to eventually alter the default behavior
of workqueues to become unbound by default so that their workload
placement is optimized by the scheduler.
Before that to happen, workqueue users must be converted to the better
named new workqueues with no intended behaviour changes:
system_wq -> system_percpu_wq
system_unbound_wq -> system_dfl_wq
This way the old obsolete workqueues (system_wq, system_unbound_wq) can
be removed in the future.
This specific work is considered long, so enqueue it using
system_dfl_long_wq instead of system_dfl_wq.
Link: https://lore.kernel.org/all/[email protected]/
Suggested-by: Tejun Heo <[email protected]>
Signed-off-by: Marco Crivellari <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
ip_vs_edit_service() while unbinding the old scheduler clears
the svc->scheduler ptr after the scheduler module initiates
RCU callbacks. This can cause packets to use the old
scheduler at the time when svc->sched_data is already freed
after RCU grace period.
Fix it by clearing the ptr early in ip_vs_unbind_scheduler(),
before the done_service method schedules any RCU callbacks.
Also, if the new scheduler fails to initialize when replacing
the old scheduler, try to restore the old scheduler while still
returning the error code.
Link: https://sashiko.dev/#/patchset/20260519015506.634185-1-rosenp%40gmail.com
Fixes: 05f00505a8 ("ipvs: fix crash if scheduler is changed")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Similar to a few other subsystems: add a new config toggle to
enable netfilter gcov profiling in netfilter, including ebtables,
arptables and so on.
ipset and ipvs gain their own, dedicated toggles.
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Sashiko points out that unprivileged user can frequently
call ip_vs_flush() or ip_vs_del_service() to trigger
svc_table_changes updates that can lead to infinite loop
in ip_vs_dst_event(). This can also happen if the user
triggers frequent table resizing without deleting all
services. We should also consider the possible effects
if the user triggers many NETDEV_DOWN events.
One way to solve it is to hold svc_resize_sem in
ip_vs_dst_event() but this can block the dev notifier
during the whole resizing process.
Instead, use new rw_semaphore svc_replace_sem to protect just
the svc_table replacement which is a short code section.
Then hold svc_replace_sem in ip_vs_dst_event() to serialize
with replacing the svc_table. As result, loop is avoided
as there is no need to repeat the table walking from the
start. By this way changes in svc_table_changes can happen
only when all services are removed and all dev references
dropped which allows us to abort the table walking.
As IP_VS_WORK_SVC_NORESIZE is the flag used to stop the
svc_resize_work under service_mutex, we should check only
this flag often but not while under service_mutex.
To remove the mutex_trylock() for service_mutex in the
second phase where the resizer installs the new table
after rehashing, we will avoid holding the service_mutex
there. As result, the code in configuration context which
is under service_mutex should access ipvs->svc_table under
RCU because it can be replaced at anytime and released
after a RCU grace period. As for ip_vs_zero_all(), it needs
different solution as a table walker which can escape
single RCU read-side critical section: to hold the
svc_replace_sem to prevent table to be replaced.
In ip_vs_status_show() prefer to hold svc_replace_sem
to avoid many loops, just detect if the svc_table is
removed.
Prefer the newly attached table for the u_thresh/l_thresh
checks to know when to grow/shrink while adding or deleting
services because the new table size is based on the latest
parameters.
Link: https://sashiko.dev/#/patchset/20260505001648.360569-1-pablo%40netfilter.org
Fixes: 840aac3d90 ("ipvs: use resizable hash table for services")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
The ip_vs_ctl.c file and the associated ip_vs.h file are the only places
in the kernel where HK_TYPE_KTHREAD cpumask is being retrieved and used.
Now that HK_TYPE_KTHREAD/HK_TYPE_DOMAIN cpumask can be changed at run
time. We need to use RCU to guard access to this cpumask to avoid a
potential UAF problem as the returned cpumask may be freed before it
is being used.
We can replace HK_TYPE_KTHREAD by HK_TYPE_DOMAIN as they are aliases
of each other, but keeping the HK_TYPE_KTHREAD name can highlight the
fact that it is the kthread initiated by ipvs that is being controlled.
Fixes: 03ff735101 ("cpuset: Update HK_TYPE_DOMAIN cpumask from cpuset")
Signed-off-by: Waiman Long <[email protected]>
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Sashiko reports for races and possible crash around
the usage of est_cpulist_valid and sysctl_est_cpulist.
The problem is that we do not lock est_mutex in some
places which can lead to wrong write ordering and
as result problems when calling cpumask_weight()
and cpumask_empty().
Fix them by moving the est_max_threads read/write under
locked est_mutex. Do the same for one ip_vs_est_reload_start()
call to protect the cpumask_empty() usage of sysctl_est_cpulist.
To remove the chance of deadlock while stopping the
estimation kthreads, keep the data structure for kthread 0
even after last estimator is removed and do not hold mutexes
while stopping this task. Now we will use a new flag 'needed'
to know when kthread 0 should run. The kthreads above 0
do not use mutexes, so stop them under est_mutex because
their kthread data still can be destroyed if they do not
serve estimators. Now all kthreads will be started by
the est_reload_work to properly serialize the stop/start
for kthread 0.
Reduce the use of service_mutex in ip_vs_est_calc_phase()
because under est_mutex we can safely walk est_kt_arr to
stop the kthreads above slot 0.
As ip_vs_stop_estimator() for tot_stats should be called
under service_mutex, do it early in the netns exit path
in ip_vs_flush() to avoid locking the mutex again later.
It still should be called in ip_vs_control_net_cleanup_sysctl()
when we are called during netns init error. Use -2 for ktid
as indicator if estimator was already stopped.
Finally, fix use-after-free for kd->est_row in
ip_vs_est_calc_phase(). est->ktrow should simply switch to
a delay value while estimator is linked to est_temp_list.
Link: https://sashiko.dev/#/patchset/20260331165015.2777765-1-longman%40redhat.com
Link: https://sashiko.dev/#/patchset/20260420171308.87192-1-ja%40ssi.bg
Link: https://sashiko.dev/#/patchset/20260422125123.40658-1-ja%40ssi.bg
Link: https://sashiko.dev/#/patchset/20260424175858.54752-1-ja%40ssi.bg
Link: https://sashiko.dev/#/patchset/20260425103918.7447-1-ja%40ssi.bg
Fixes: f0be83d542 ("ipvs: add est_cpulist and est_nice sysctl vars")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
syzbot reports for sleeping function called from invalid context [1].
The recently added code for resizable hash tables uses
hlist_bl bit locks in combination with spin_lock for
the connection fields (cp->lock).
Fix the following problems:
* avoid using spin_lock(&cp->lock) under locked bit lock
because it sleeps on PREEMPT_RT
* as the recent changes call ip_vs_conn_hash() only for newly
allocated connection, the spin_lock can be removed there because
the connection is still not linked to table and does not need
cp->lock protection.
* the lock can be removed also from ip_vs_conn_unlink() where we
are the last connection user.
* the last place that is fixed is ip_vs_conn_fill_cport()
where now the cp->lock is locked before the other locks to
ensure other packets do not modify the cp->flags in non-atomic
way. Here we make sure cport and flags are changed only once
if two or more packets race to fill the cport. Also, we fill
cport early, so that if we race with resizing there will be
valid cport key for the hashing. Add a warning if too many
hash table changes occur for our RCU read-side critical
section which is error condition but minor because the
connection still can expire gracefully. Still, restore the
cport to 0 to allow retransmitted packet to properly fill
the cport. Problems reported by Sashiko.
[1]:
BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48
in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 16, name: ktimers/0
preempt_count: 2, expected: 0
RCU nest depth: 3, expected: 3
8 locks held by ktimers/0/16:
#0: ffffffff8de5f260 (local_bh){.+.+}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163
#1: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163
#2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: spin_lock include/linux/spinlock_rt.h:45 [inline]
#2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: timer_base_lock_expiry kernel/time/timer.c:1502 [inline]
#2: ffff8880b8826360 (&base->expiry_lock){+...}-{3:3}, at: __run_timer_base+0x120/0x9f0 kernel/time/timer.c:2384
#3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:300 [inline]
#3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:838 [inline]
#3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: __rt_spin_lock kernel/locking/spinlock_rt.c:50 [inline]
#3: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rt_spin_lock+0x1e0/0x400 kernel/locking/spinlock_rt.c:57
#4: ffffc90000157a80 ((&cp->timer)){+...}-{0:0}, at: call_timer_fn+0xd4/0x5e0 kernel/time/timer.c:1745
#5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:300 [inline]
#5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:838 [inline]
#5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:315 [inline]
#5: ffffffff8dfc80c0 (rcu_read_lock){....}-{1:3}, at: ip_vs_conn_expire+0x257/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260
#6: ffffffff8de5f260 (local_bh){.+.+}-{1:3}, at: __local_bh_disable_ip+0x3c/0x420 kernel/softirq.c:163
#7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: spin_lock include/linux/spinlock_rt.h:45 [inline]
#7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:324 [inline]
#7: ffff888068d4c3f0 (&cp->lock#2){+...}-{3:3}, at: ip_vs_conn_expire+0xd4a/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260
Preemption disabled at:
[<ffffffff898a6358>] bit_spin_lock include/linux/bit_spinlock.h:38 [inline]
[<ffffffff898a6358>] hlist_bl_lock+0x18/0x110 include/linux/list_bl.h:149
CPU: 0 UID: 0 PID: 16 Comm: ktimers/0 Tainted: G W L syzkaller #0 PREEMPT_{RT,(full)}
Tainted: [W]=WARN, [L]=SOFTLOCKUP
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/18/2026
Call Trace:
<TASK>
dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120
__might_resched+0x329/0x480 kernel/sched/core.c:9162
__rt_spin_lock kernel/locking/spinlock_rt.c:48 [inline]
rt_spin_lock+0xc2/0x400 kernel/locking/spinlock_rt.c:57
spin_lock include/linux/spinlock_rt.h:45 [inline]
ip_vs_conn_unlink net/netfilter/ipvs/ip_vs_conn.c:324 [inline]
ip_vs_conn_expire+0xd4a/0x2390 net/netfilter/ipvs/ip_vs_conn.c:1260
call_timer_fn+0x192/0x5e0 kernel/time/timer.c:1748
expire_timers kernel/time/timer.c:1799 [inline]
__run_timers kernel/time/timer.c:2374 [inline]
__run_timer_base+0x6a3/0x9f0 kernel/time/timer.c:2386
run_timer_base kernel/time/timer.c:2395 [inline]
run_timer_softirq+0xb7/0x170 kernel/time/timer.c:2405
handle_softirqs+0x1de/0x6d0 kernel/softirq.c:622
__do_softirq kernel/softirq.c:656 [inline]
run_ktimerd+0x69/0x100 kernel/softirq.c:1151
smpboot_thread_fn+0x541/0xa50 kernel/smpboot.c:160
kthread+0x388/0x470 kernel/kthread.c:436
ret_from_fork+0x514/0xb70 arch/x86/kernel/process.c:158
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>
Reported-by: [email protected]
Link: https://sashiko.dev/#/patchset/20260415200216.79699-1-ja%40ssi.bg
Link: https://sashiko.dev/#/patchset/20260420165539.85174-4-ja%40ssi.bg
Link: https://sashiko.dev/#/patchset/20260422135823.50489-4-ja%40ssi.bg
Fixes: 2fa7cc9c70 ("ipvs: switch to per-net connection table")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Sashiko warns that the new sysctls vars can be changed
after the hash tables are destroyed and their respective
resizing works canceled, leading to mod_delayed_work()
being called for canceled works.
Solve this in different ways. conn_tab can be present even
without services and is destroyed only on netns exit, so use
disable_delayed_work_sync() to disable the work instead of
adding more synchronization mechanisms.
As for the svc_table, it is destroyed when the services
are deleted, so we must be sure that netns exit is not
called yet (the check for 'enable') and the work is
not canceled by checking all under same mutex lock.
Also, use WRITE_ONCE when updating the sysctl vars as we
already read them with READ_ONCE.
Link: https://sashiko.dev/#/patchset/20260410112352.23599-1-fw%40strlen.de
Fixes: 8d7de5477e ("ipvs: add conn_lfactor and svc_lfactor sysctl vars")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Sashiko reports some problems for the recently added
/proc/net/ip_vs_status:
* ip_vs_status_show() as a table reader may run long after the
conn_tab and svc_table table are released. While ip_vs_conn_flush()
properly changes the conn_tab_changes counter when conn_tab is removed,
ip_vs_del_service() and ip_vs_flush() were missing such change for
the svc_table_changes counter. As result, readers like
ip_vs_dst_event() and ip_vs_status_show() may continue to use
a freed table after a cond_resched_rcu() call.
* While counting the buckets in ip_vs_status_show() make sure we
traverse only the needed number of entries in the chain. This also
prevents possible overflow of the 'count' variable.
* Add check for 'loops' to prevent infinite loops while restarting
the traversal on table change.
* While IP_VS_CONN_TAB_MAX_BITS is 20 on 32-bit platforms and
there is no risk to overflow when multiplying the number of
conn_tab buckets to 100, prefer the div_u64() helper to make
the following dividing safer.
* Use 0440 permissions for ip_vs_status to restrict the
info only to root due to the exported information for hash
distribution.
Link: https://sashiko.dev/#/patchset/20260410112352.23599-1-fw%40strlen.de
Fixes: 9a9ccef907 ("ipvs: add ip_vs_status info")
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Currently, IPVS skips MTU checks for GSO packets by excluding them with
the !skb_is_gso(skb) condition. This creates problems when IPVS tunnel
mode encapsulates GSO packets with IPIP headers.
The issue manifests in two ways:
1. MTU violation after encapsulation:
When a GSO packet passes through IPVS tunnel mode, the original MTU
check is bypassed. After adding the IPIP tunnel header, the packet
size may exceed the outgoing interface MTU, leading to unexpected
fragmentation at the IP layer.
2. Fragmentation with problematic IP IDs:
When net.ipv4.vs.pmtu_disc=1 and a GSO packet with multiple segments
is fragmented after encapsulation, each segment gets a sequentially
incremented IP ID (0, 1, 2, ...). This happens because:
a) The GSO packet bypasses MTU check and gets encapsulated
b) At __ip_finish_output, the oversized GSO packet is split into
separate SKBs (one per segment), with IP IDs incrementing
c) Each SKB is then fragmented again based on the actual MTU
This sequential IP ID allocation differs from the expected behavior
and can cause issues with fragment reassembly and packet tracking.
Fix this by properly validating GSO packets using
skb_gso_validate_network_len(). This function correctly validates
whether the GSO segments will fit within the MTU after segmentation. If
validation fails, send an ICMP Fragmentation Needed message to enable
proper PMTU discovery.
Fixes: 4cdd34084d ("netfilter: nf_conntrack_ipv6: improve fragmentation handling")
Signed-off-by: Yingnan Zhang <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Pablo Neira Ayuso <[email protected]>
Allow the default load factor for the connection and service tables
to be configured.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Add /proc/net/ip_vs_status to show current state of IPVS.
The motivation for this new /proc interface is to provide the output
for the users to help them decide when to tune the load factor for
hash tables, which is possible with the new sysctl knobs coming in
followup patch.
The output also includes information for the kthreads used for stats.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
As conn_tab is per-net, better to show the current hash table size
to users instead of the ip_vs_conn_tab_size (max).
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
When ip_vs_bind_scheduler() succeeds in ip_vs_add_service(), the local
variable sched is set to NULL. If ip_vs_start_estimator() subsequently
fails, the out_err cleanup calls ip_vs_unbind_scheduler(svc, sched)
with sched == NULL. ip_vs_unbind_scheduler() passes the cur_sched NULL
check (because svc->scheduler was set by the successful bind) but then
dereferences the NULL sched parameter at sched->done_service, causing a
kernel panic at offset 0x30 from NULL.
Oops: general protection fault, [..] [#1] PREEMPT SMP KASAN NOPTI
KASAN: null-ptr-deref in range [0x0000000000000030-0x0000000000000037]
RIP: 0010:ip_vs_unbind_scheduler (net/netfilter/ipvs/ip_vs_sched.c:69)
Call Trace:
<TASK>
ip_vs_add_service.isra.0 (net/netfilter/ipvs/ip_vs_ctl.c:1500)
do_ip_vs_set_ctl (net/netfilter/ipvs/ip_vs_ctl.c:2809)
nf_setsockopt (net/netfilter/nf_sockopt.c:102)
[..]
Fix by simply not clearing the local sched variable after a successful
bind. ip_vs_unbind_scheduler() already detects whether a scheduler is
installed via svc->scheduler, and keeping sched non-NULL ensures the
error path passes the correct pointer to both ip_vs_unbind_scheduler()
and ip_vs_scheduler_put().
While the bug is older, the problem popups in more recent kernels (6.2),
when the new error path is taken after the ip_vs_start_estimator() call.
Fixes: 705dd34440 ("ipvs: use kthreads for stats estimation")
Reported-by: Xiang Mei <[email protected]>
Signed-off-by: Weiming Shi <[email protected]>
Acked-by: Simon Horman <[email protected]>
Acked-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Simon Kirby reported long time ago that IPVS connection hashing
based only on the client address/port (caddr, cport) as hash keys
is not suitable for setups that accept traffic on multiple virtual
IPs and ports. It can happen for multiple VIP:VPORT services, for
single or many fwmark service(s) that match multiple virtual IPs
and ports or even for passive FTP with peristence in DR/TUN mode
where we expect traffic on multiple ports for the virtual IP.
Fix it by adding virtual addresses and ports to the hash function.
This causes the traffic from NAT real servers to clients to use
second hashing for the in->out direction.
As result:
- the IN direction from client will use hash node hn0 where
the source/dest addresses and ports used by client will be used
as hash keys
- the OUT direction from NAT real servers will use hash node hn1
for the traffic from real server to client
- the persistence templates are hashed only with parameters based on
the IN direction, so they now will also use the virtual address,
port and fwmark from the service.
OLD:
- all methods: c_list node: proto, caddr:cport
- persistence templates: c_list node: proto, caddr_net:0
- persistence engine templates: c_list node: per-PE, PE-SIP uses jhash
NEW:
- all methods: hn0 node (dir 0): proto, caddr:cport -> vaddr:vport
- MASQ method: hn1 node (dir 1): proto, daddr:dport -> caddr:cport
- persistence templates: hn0 node (dir 0):
proto, caddr_net:0 -> vaddr:vport_or_0
proto, caddr_net:0 -> fwmark:0
- persistence engine templates: hn0 node (dir 0): as before
Also reorder the ip_vs_conn fields, so that hash nodes are on same
read-mostly cache line while write-mostly fields are on separate
cache line.
Reported-by: Simon Kirby <[email protected]>
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Use per-net resizable hash table for connections. The global table
is slow to walk when using many namespaces.
The table can be resized in the range of [256 - ip_vs_conn_tab_size].
Table is attached only while services are present. Resizing is done
by delayed work based on load (the number of connections).
Add a hash_key field into the connection to store the table ID in
the highest bit and the entry's hash value in the lowest bits. The
lowest part of the hash value is used as bucket ID, the remaining
part is used to filter the entries in the bucket before matching
the keys and as result, helps the lookup operation to access only
one cache line. By knowing the table ID and bucket ID for entry,
we can unlink it without calculating the hash value and doing
lookup by keys. We need only to validate the saved hash_key under
lock.
For better security switch from jhash to siphash for the default
connection hashing but the persistence engines may use their own
function. Keeping the hash table loaded with entries below the
size (12%) allows to avoid collision for 96+% of the conns.
ip_vs_conn_fill_cport() now will rehash the connection with proper
locking because unhash+hash is not safe for RCU readers.
To invalidate the templates setting just dport to 0xffff is enough,
no need to rehash them. As result, ip_vs_conn_unhash() is now
unused and removed.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Make the hash table for services resizable in the bit range of 4-20.
Table is attached only while services are present. Resizing is done
by delayed work based on load (the number of hashed services).
Table grows when load increases 2+ times (above 12.5% with lfactor=-3)
and shrinks 8+ times when load decreases 16+ times (below 0.78%).
Switch to jhash hashing to reduce the collisions for multiple
services.
Add a hash_key field into the service to store the table ID in
the highest bit and the entry's hash value in the lowest bits. The
lowest part of the hash value is used as bucket ID, the remaining
part is used to filter the entries in the bucket before matching
the keys and as result, helps the lookup operation to access only
one cache line. By knowing the table ID and bucket ID for entry,
we can unlink it without calculating the hash value and doing
lookup by keys. We need only to validate the saved hash_key under
lock.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Add infrastructure for resizable hash tables based on hlist_bl
which we will use in followup patches.
The tables allow RCU lookups during resizing, bucket modifications
are protected with per-bucket bit lock and additional custom locking,
the tables are resized when load reaches thresholds determined based
on load factor parameter.
Compared to other implementations we rely on:
* fast entry removal by using node unlinking without pre-lookup
* entry rehashing when hash key changes
* entries can contain multiple hash nodes
* custom locking depending on different contexts
* adjustable load factor to customize the grow/shrink process
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Cross-merge networking fixes after downstream PR (net-7.0-rc2).
Conflicts:
tools/testing/selftests/drivers/net/hw/rss_ctx.py
19c3a2a81d ("selftests: drv-net: rss: Generate unique ports for RSS context tests")
ce5a0f4612 ("selftests: drv-net: rss_ctx: test RSS contexts persist after ifdown/up")
include/net/inet_connection_sock.h
858d2a4f67 ("tcp: fix potential race in tcp_v6_syn_recv_sock()")
fcd3d039fa ("tcp: make tcp_v{4,6}_send_check() static")
https://lore.kernel.org/[email protected]
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c
69050f8d6d ("treewide: Replace kmalloc with kmalloc_obj for non-scalar types")
bf4afc53b7 ("Convert 'alloc_obj' family to use the new default GFP_KERNEL argument")
8a96b9144f ("net/mlx5e: Alloc xsk channel param out of mlx5e_open_xsk()")
Adjacent changes:
net/netfilter/ipvs/ip_vs_ctl.c
c59bd9e62e ("ipvs: use more counters to avoid service lookups")
bf4afc53b7 ("Convert 'alloc_obj' family to use the new default GFP_KERNEL argument")
Signed-off-by: Jakub Kicinski <[email protected]>
Change the no_cport counters to be per-net and address family.
This should reduce the extra conn lookups done during present
NO_CPORT connections.
By changing from global to per-net dropentry counters, one net
will not affect the drop rate of another net.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
When new connection is created we can lookup for services multiple
times to support fallback options. We already have some counters
to skip specific lookups because it costs CPU cycles for hash
calculation, etc.
Add more counters for fwmark/non-fwmark services (fwm_services and
nonfwm_services) and make all counters per address family.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
Before now dest->dest_dst is not released when server is moved into
dest_trash list after removal. As result, we can keep dst/dev
references for long time without actively using them.
It is better to avoid walking the dest_trash list when
ip_vs_dst_event() receives dev events. So, make sure we do not
hold dev references in dest_trash list. As packets can be flying
while server is being removed, check the IP_VS_DEST_F_AVAILABLE
flag in slow path to ensure we do not save new dev references to
removed servers.
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
Some places walk the services under mutex but they can just use RCU:
* ip_vs_dst_event() uses ip_vs_forget_dev() which uses its own lock
to modify dest
* ip_vs_genl_dump_services(): ip_vs_genl_fill_service() just fills skb
* ip_vs_genl_parse_service(): move RCU lock to callers
ip_vs_genl_set_cmd(), ip_vs_genl_dump_dests() and ip_vs_genl_get_cmd()
* ip_vs_genl_dump_dests(): just fill skb
Signed-off-by: Julian Anastasov <[email protected]>
Reviewed-by: Dust Li <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
Current ipvs uses one global mutex "__ip_vs_mutex" to keep the global
"ip_vs_svc_table" and "ip_vs_svc_fwm_table" safe. But when there are
tens of thousands of services from different netns in the table, it
takes a long time to look up the table, for example, using "ipvsadm
-ln" from different netns simultaneously.
We make "ip_vs_svc_table" and "ip_vs_svc_fwm_table" per netns, and we
add "service_mutex" per netns to keep these two tables safe instead of
the global "__ip_vs_mutex" in current version. To this end, looking up
services from different netns simultaneously will not get stuck,
shortening the time consumption in large-scale deployment. It can be
reproduced using the simple scripts below.
init.sh: #!/bin/bash
for((i=1;i<=4;i++));do
ip netns add ns$i
ip netns exec ns$i ip link set dev lo up
ip netns exec ns$i sh add-services.sh
done
add-services.sh: #!/bin/bash
for((i=0;i<30000;i++)); do
ipvsadm -A -t 10.10.10.10:$((80+$i)) -s rr
done
runtest.sh: #!/bin/bash
for((i=1;i<4;i++));do
ip netns exec ns$i ipvsadm -ln > /dev/null &
done
ip netns exec ns4 ipvsadm -ln > /dev/null
Run "sh init.sh" to initiate the network environment. Then run "time
./runtest.sh" to evaluate the time consumption. Our testbed is a 4-core
Intel Xeon ECS. The result of the original version is around 8 seconds,
while the result of the modified version is only 0.8 seconds.
Signed-off-by: Jiejian Wu <[email protected]>
Co-developed-by: Dust Li <[email protected]>
Signed-off-by: Dust Li <[email protected]>
Signed-off-by: Julian Anastasov <[email protected]>
Signed-off-by: Florian Westphal <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
This converts some of the visually simpler cases that have been split
over multiple lines. I only did the ones that are easy to verify the
resulting diff by having just that final GFP_KERNEL argument on the next
line.
Somebody should probably do a proper coccinelle script for this, but for
me the trivial script actually resulted in an assertion failure in the
middle of the script. I probably had made it a bit _too_ trivial.
So after fighting that far a while I decided to just do some of the
syntactically simpler cases with variations of the previous 'sed'
scripts.
The more syntactically complex multi-line cases would mostly really want
whitespace cleanup anyway.
Signed-off-by: Linus Torvalds <[email protected]>
This was done entirely with mindless brute force, using
git grep -l '\<k[vmz]*alloc_objs*(.*, GFP_KERNEL)' |
xargs sed -i 's/\(alloc_objs*(.*\), GFP_KERNEL)/\1)/'
to convert the new alloc_obj() users that had a simple GFP_KERNEL
argument to just drop that argument.
Note that due to the extreme simplicity of the scripting, any slightly
more complex cases spread over multiple lines would not be triggered:
they definitely exist, but this covers the vast bulk of the cases, and
the resulting diff is also then easier to check automatically.
For the same reason the 'flex' versions will be done as a separate
conversion.
Signed-off-by: Linus Torvalds <[email protected]>
Julian Anastasov