[ Upstream commit 87b08913a9 ]
icmp_global_credit was meant to be changed ~1000 times per second,
but if an admin sets net.ipv4.icmp_msgs_per_sec to a very high value,
icmp_global_credit changes can inflict false sharing to surrounding
fields that are read mostly.
Move icmp_global_credit and icmp_global_stamp to a separate
cacheline aligned group.
Fixes: b056b4cd91 ("icmp: move icmp_global.credit and icmp_global.stamp to per netns storage")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260216142832.3834174-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 18fd64d254 ]
Reorganize fast path variables on tx-txrx-rx order.
Fastpath cacheline ends after sysctl_tcp_rmem.
There are only read-only variables here. (write is on the control path
and not considered in this case)
Below data generated with pahole on x86 architecture.
Fast path variables span cache lines before change: 4
Fast path variables span cache lines after change: 2
Suggested-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Wei Wang <weiwan@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Coco Li <lixiaoyan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: 87b08913a9 ("inet: move icmp_global_{credit,stamp} to a separate cache line")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 562b1fdf06 ]
TCP pingpong threshold is 1 by default. But some applications, like SQL DB
may prefer a higher pingpong threshold to activate delayed acks in quick
ack mode for better performance.
The pingpong threshold and related code were changed to 3 in the year
2019 in:
commit 4a41f453be ("tcp: change pingpong threshold to 3")
And reverted to 1 in the year 2022 in:
commit 4d8f24eeed ("Revert "tcp: change pingpong threshold to 3"")
There is no single value that fits all applications.
Add net.ipv4.tcp_pingpong_thresh sysctl tunable, so it can be tuned for
optimal performance based on the application needs.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/1697056244-21888-1-git-send-email-haiyangz@microsoft.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: 87b08913a9 ("inet: move icmp_global_{credit,stamp} to a separate cache line")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 133c4c0d37 ]
This idea came after a particular workload requested
the quickack attribute set on routes, and a performance
drop was noticed for large bulk transfers.
For high throughput flows, it is best to use one cpu
running the user thread issuing socket system calls,
and a separate cpu to process incoming packets from BH context.
(With TSO/GRO, bottleneck is usually the 'user' cpu)
Problem is the user thread can spend a lot of time while holding
the socket lock, forcing BH handler to queue most of incoming
packets in the socket backlog.
Whenever the user thread releases the socket lock, it must first
process all accumulated packets in the backlog, potentially
adding latency spikes. Due to flood mitigation, having too many
packets in the backlog increases chance of unexpected drops.
Backlog processing unfortunately shifts a fair amount of cpu cycles
from the BH cpu to the 'user' cpu, thus reducing max throughput.
This patch takes advantage of the backlog processing,
and the fact that ACK are mostly cumulative.
The idea is to detect we are in the backlog processing
and defer all eligible ACK into a single one,
sent from tcp_release_cb().
This saves cpu cycles on both sides, and network resources.
Performance of a single TCP flow on a 200Gbit NIC:
- Throughput is increased by 20% (100Gbit -> 120Gbit).
- Number of generated ACK per second shrinks from 240,000 to 40,000.
- Number of backlog drops per second shrinks from 230 to 0.
Benchmark context:
- Regular netperf TCP_STREAM (no zerocopy)
- Intel(R) Xeon(R) Platinum 8481C (Saphire Rapids)
- MAX_SKB_FRAGS = 17 (~60KB per GRO packet)
This feature is guarded by a new sysctl, and enabled by default:
/proc/sys/net/ipv4/tcp_backlog_ack_defer
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: 87b08913a9 ("inet: move icmp_global_{credit,stamp} to a separate cache line")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 21d033e472 ]
After the optimization to only perform one GC per jiffy, a new problem
was introduced. If more than 8 new connections are tracked per jiffy the
list won't be cleaned up fast enough possibly reaching the limit
wrongly.
In order to prevent this issue, only skip the GC if it was already
triggered during the same jiffy and the increment is lower than the
clean up limit. In addition, increase the clean up limit to 64
connections to avoid triggering GC too often and do more effective GCs.
This has been tested using a HTTP server and several
performance tools while having nft_connlimit/xt_connlimit or OVS limit
configured.
Output of slowhttptest + OVS limit at 52000 connections:
slow HTTP test status on 340th second:
initializing: 0
pending: 432
connected: 51998
error: 0
closed: 0
service available: YES
Fixes: d265929930 ("netfilter: nf_conncount: reduce unnecessary GC")
Reported-by: Aleksandra Rukomoinikova <ARukomoinikova@k2.cloud>
Closes: https://lore.kernel.org/netfilter/b2064e7b-0776-4e14-adb6-c68080987471@k2.cloud/
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9990ddf47d ]
Make skb_vlan_inet_prepare return the skb drop reasons, which is just
what pskb_may_pull_reason() returns. Meanwhile, adjust all the call of
it.
Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a1356ac774 ]
Move rx_lock from xsk_socket to xsk_buff_pool.
Fix synchronization for shared umem mode in
generic RX path where multiple sockets share
single xsk_buff_pool.
RX queue is exclusive to xsk_socket, while FILL
queue can be shared between multiple sockets.
This could result in race condition where two
CPU cores access RX path of two different sockets
sharing the same umem.
Protect both queues by acquiring spinlock in shared
xsk_buff_pool.
Lock contention may be minimized in the future by some
per-thread FQ buffering.
It's safe and necessary to move spin_lock_bh(rx_lock)
after xsk_rcv_check():
* xs->pool and spinlock_init is synchronized by
xsk_bind() -> xsk_is_bound() memory barriers.
* xsk_rcv_check() may return true at the moment
of xsk_release() or xsk_unbind_dev(),
however this will not cause any data races or
race conditions. xsk_unbind_dev() removes xdp
socket from all maps and waits for completion
of all outstanding rx operations. Packets in
RX path will either complete safely or drop.
Signed-off-by: Eryk Kubanski <e.kubanski@partner.samsung.com>
Fixes: bf0bdd1343 ("xdp: fix race on generic receive path")
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://patch.msgid.link/20250416101908.10919-1-e.kubanski@partner.samsung.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[ Conflict is resolved when backporting this fix. ]
Signed-off-by: Jianqiang kang <jianqkang@sina.cn>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 8e1a1bc4f5 ]
Hamza Mahfooz reports cpu soft lock-ups in
nft_chain_validate():
watchdog: BUG: soft lockup - CPU#1 stuck for 27s! [iptables-nft-re:37547]
[..]
RIP: 0010:nft_chain_validate+0xcb/0x110 [nf_tables]
[..]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_immediate_validate+0x36/0x50 [nf_tables]
nft_chain_validate+0xc9/0x110 [nf_tables]
nft_table_validate+0x6b/0xb0 [nf_tables]
nf_tables_validate+0x8b/0xa0 [nf_tables]
nf_tables_commit+0x1df/0x1eb0 [nf_tables]
[..]
Currently nf_tables will traverse the entire table (chain graph), starting
from the entry points (base chains), exploring all possible paths
(chain jumps). But there are cases where we could avoid revalidation.
Consider:
1 input -> j2 -> j3
2 input -> j2 -> j3
3 input -> j1 -> j2 -> j3
Then the second rule does not need to revalidate j2, and, by extension j3,
because this was already checked during validation of the first rule.
We need to validate it only for rule 3.
This is needed because chain loop detection also ensures we do not exceed
the jump stack: Just because we know that j2 is cycle free, its last jump
might now exceed the allowed stack size. We also need to update all
reachable chains with the new largest observed call depth.
Care has to be taken to revalidate even if the chain depth won't be an
issue: chain validation also ensures that expressions are not called from
invalid base chains. For example, the masquerade expression can only be
called from NAT postrouting base chains.
Therefore we also need to keep record of the base chain context (type,
hooknum) and revalidate if the chain becomes reachable from a different
hook location.
Reported-by: Hamza Mahfooz <hamzamahfooz@linux.microsoft.com>
Closes: https://lore.kernel.org/netfilter-devel/20251118221735.GA5477@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net/
Tested-by: Hamza Mahfooz <hamzamahfooz@linux.microsoft.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 1dbf1d590d upstream.
In ip_output() skb->dev is updated from the skb_dst(skb)->dev
this can become invalid when the interface is unregistered and freed,
Introduced new skb_dst_dev_rcu() function to be used instead of
skb_dst_dev() within rcu_locks in ip_output.This will ensure that
all the skb's associated with the dev being deregistered will
be transnmitted out first, before freeing the dev.
Given that ip_output() is called within an rcu_read_lock()
critical section or from a bottom-half context, it is safe to introduce
an RCU read-side critical section within it.
Multiple panic call stacks were observed when UL traffic was run
in concurrency with device deregistration from different functions,
pasting one sample for reference.
[496733.627565][T13385] Call trace:
[496733.627570][T13385] bpf_prog_ce7c9180c3b128ea_cgroupskb_egres+0x24c/0x7f0
[496733.627581][T13385] __cgroup_bpf_run_filter_skb+0x128/0x498
[496733.627595][T13385] ip_finish_output+0xa4/0xf4
[496733.627605][T13385] ip_output+0x100/0x1a0
[496733.627613][T13385] ip_send_skb+0x68/0x100
[496733.627618][T13385] udp_send_skb+0x1c4/0x384
[496733.627625][T13385] udp_sendmsg+0x7b0/0x898
[496733.627631][T13385] inet_sendmsg+0x5c/0x7c
[496733.627639][T13385] __sys_sendto+0x174/0x1e4
[496733.627647][T13385] __arm64_sys_sendto+0x28/0x3c
[496733.627653][T13385] invoke_syscall+0x58/0x11c
[496733.627662][T13385] el0_svc_common+0x88/0xf4
[496733.627669][T13385] do_el0_svc+0x2c/0xb0
[496733.627676][T13385] el0_svc+0x2c/0xa4
[496733.627683][T13385] el0t_64_sync_handler+0x68/0xb4
[496733.627689][T13385] el0t_64_sync+0x1a4/0x1a8
Changes in v3:
- Replaced WARN_ON() with WARN_ON_ONCE(), as suggested by Willem de Bruijn.
- Dropped legacy lines mistakenly pulled in from an outdated branch.
Changes in v2:
- Addressed review comments from Eric Dumazet
- Used READ_ONCE() to prevent potential load/store tearing
- Added skb_dst_dev_rcu() and used along with rcu_read_lock() in ip_output
Signed-off-by: Sharath Chandra Vurukala <quic_sharathv@quicinc.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20250730105118.GA26100@hu-sharathv-hyd.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
[ Keerthana: Backported the patch to v6.6.y ]
Signed-off-by: Keerthana K <keerthana.kalyanasundaram@broadcom.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 14fb07130c ]
Reject rules where a load occurs from a register that has not seen a store
early in the same rule.
commit 4c905f6740 ("netfilter: nf_tables: initialize registers in
nft_do_chain()")
had to add a unconditional memset to the nftables register space to avoid
leaking stack information to userspace.
This memset shows up in benchmarks. After this change, this commit can
be reverted again.
Note that this breaks userspace compatibility, because theoretically
you can do
rule 1: reg2 := meta load iif, reg2 == 1 jump ...
rule 2: reg2 == 2 jump ... // read access with no store in this rule
... after this change this is rejected.
Neither nftables nor iptables-nft generate such rules, each rule is
always standalone.
This resuts in a small increase of nft_ctx structure by sizeof(long).
To cope with hypothetical rulesets like the example above one could emit
on-demand "reg[x] = 0" store when generating the datapath blob in
nf_tables_commit_chain_prepare().
A patch that does this is linked to below.
For now, lets disable this. In nf_tables, a rule is the smallest
unit that can be replaced from userspace, i.e. a hypothetical ruleset
that relies on earlier initialisations of registers can't be changed
at will as register usage would need to be coordinated.
Link: https://lore.kernel.org/netfilter-devel/20240627135330.17039-4-fw@strlen.de/
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Stable-dep-of: a67fd55f6a ("netfilter: nf_tables: remove redundant chain validation on register store")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit be102eb6a0 ]
When using nf_conncount infrastructure for non-confirmed connections a
duplicated track is possible due to an optimization introduced since
commit d265929930 ("netfilter: nf_conncount: reduce unnecessary GC").
In order to fix this introduce a new conncount API that receives
directly an sk_buff struct. It fetches the tuple and zone and the
corresponding ct from it. It comes with both existing conncount variants
nf_conncount_count_skb() and nf_conncount_add_skb(). In addition remove
the old API and adjust all the users to use the new one.
This way, for each sk_buff struct it is possible to check if there is a
ct present and already confirmed. If so, skip the add operation.
Fixes: d265929930 ("netfilter: nf_conncount: reduce unnecessary GC")
Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5eb902b8e7 ]
FIB6 GC walks trees of fib6_tables to remove expired routes. Walking a tree
can be expensive if the number of routes in a table is big, even if most of
them are permanent. Checking routes in a separated list of routes having
expiration will avoid this potential issue.
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: f72514b3c5 ("ipv6: clear RA flags when adding a static route")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1532ed0d07 ]
Since ehash lookups are lockless, if one CPU performs a lookup while
another concurrently deletes and inserts (removing reqsk and inserting sk),
the lookup may fail to find the socket, an RST may be sent.
The call trace map is drawn as follows:
CPU 0 CPU 1
----- -----
inet_ehash_insert()
spin_lock()
sk_nulls_del_node_init_rcu(osk)
__inet_lookup_established()
(lookup failed)
__sk_nulls_add_node_rcu(sk, list)
spin_unlock()
As both deletion and insertion operate on the same ehash chain, this patch
introduces a new sk_nulls_replace_node_init_rcu() helper functions to
implement atomic replacement.
Fixes: 5e0724d027 ("tcp/dccp: fix hashdance race for passive sessions")
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev>
Signed-off-by: Xuanqiang Luo <luoxuanqiang@kylinos.cn>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20251015020236.431822-3-xuanqiang.luo@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2a198bbec6 ]
This reverts commit f75a2804da.
With all states (whether user or kern) removed from the hashtables
during deletion, there's no need for synchronous destruction of
states. xfrm6_tunnel states still need to have been destroyed (which
will be the case when its last user is deleted (not destroyed)) so
that xfrm6_tunnel_free_spi removes it from the per-netns hashtable
before the netns is destroyed.
This has the benefit of skipping one synchronize_rcu per state (in
__xfrm_state_destroy(sync=true)) when we exit a netns.
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b441cf3f8c ]
The ipcomp fallback tunnels currently get deleted (from the various
lists and hashtables) as the last user state that needed that fallback
is destroyed (not deleted). If a reference to that user state still
exists, the fallback state will remain on the hashtables/lists,
triggering the WARN in xfrm_state_fini. Because of those remaining
references, the fix in commit f75a2804da ("xfrm: destroy xfrm_state
synchronously on net exit path") is not complete.
We recently fixed one such situation in TCP due to defered freeing of
skbs (commit 9b6412e697 ("tcp: drop secpath at the same time as we
currently drop dst")). This can also happen due to IP reassembly: skbs
with a secpath remain on the reassembly queue until netns
destruction. If we can't guarantee that the queues are flushed by the
time xfrm_state_fini runs, there may still be references to a (user)
xfrm_state, preventing the timely deletion of the corresponding
fallback state.
Instead of chasing each instance of skbs holding a secpath one by one,
this patch fixes the issue directly within xfrm, by deleting the
fallback state as soon as the last user state depending on it has been
deleted. Destruction will still happen when the final reference is
dropped.
A separate lockdep class for the fallback state is required since
we're going to lock x->tunnel while x is locked.
Fixes: 9d4139c769 ("netns xfrm: per-netns xfrm_state_all list")
Signed-off-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 094ee6017e ]
Following operations can trigger a warning[1]:
ip netns add ns1
ip netns exec ns1 ip link add bond0 type bond mode balance-rr
ip netns exec ns1 ip link set dev bond0 xdp obj af_xdp_kern.o sec xdp
ip netns exec ns1 ip link set bond0 type bond mode broadcast
ip netns del ns1
When delete the namespace, dev_xdp_uninstall() is called to remove xdp
program on bond dev, and bond_xdp_set() will check the bond mode. If bond
mode is changed after attaching xdp program, the warning may occur.
Some bond modes (broadcast, etc.) do not support native xdp. Set bond mode
with xdp program attached is not good. Add check for xdp program when set
bond mode.
[1]
------------[ cut here ]------------
WARNING: CPU: 0 PID: 11 at net/core/dev.c:9912 unregister_netdevice_many_notify+0x8d9/0x930
Modules linked in:
CPU: 0 UID: 0 PID: 11 Comm: kworker/u4:0 Not tainted 6.14.0-rc4 #107
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.15.0-0-g2dd4b9b3f840-prebuilt.qemu.org 04/01/2014
Workqueue: netns cleanup_net
RIP: 0010:unregister_netdevice_many_notify+0x8d9/0x930
Code: 00 00 48 c7 c6 6f e3 a2 82 48 c7 c7 d0 b3 96 82 e8 9c 10 3e ...
RSP: 0018:ffffc90000063d80 EFLAGS: 00000282
RAX: 00000000ffffffa1 RBX: ffff888004959000 RCX: 00000000ffffdfff
RDX: 0000000000000000 RSI: 00000000ffffffea RDI: ffffc90000063b48
RBP: ffffc90000063e28 R08: ffffffff82d39b28 R09: 0000000000009ffb
R10: 0000000000000175 R11: ffffffff82d09b40 R12: ffff8880049598e8
R13: 0000000000000001 R14: dead000000000100 R15: ffffc90000045000
FS: 0000000000000000(0000) GS:ffff888007a00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000000d406b60 CR3: 000000000483e000 CR4: 00000000000006f0
Call Trace:
<TASK>
? __warn+0x83/0x130
? unregister_netdevice_many_notify+0x8d9/0x930
? report_bug+0x18e/0x1a0
? handle_bug+0x54/0x90
? exc_invalid_op+0x18/0x70
? asm_exc_invalid_op+0x1a/0x20
? unregister_netdevice_many_notify+0x8d9/0x930
? bond_net_exit_batch_rtnl+0x5c/0x90
cleanup_net+0x237/0x3d0
process_one_work+0x163/0x390
worker_thread+0x293/0x3b0
? __pfx_worker_thread+0x10/0x10
kthread+0xec/0x1e0
? __pfx_kthread+0x10/0x10
? __pfx_kthread+0x10/0x10
ret_from_fork+0x2f/0x50
? __pfx_kthread+0x10/0x10
ret_from_fork_asm+0x1a/0x30
</TASK>
---[ end trace 0000000000000000 ]---
Fixes: 9e2ee5c7e7 ("net, bonding: Add XDP support to the bonding driver")
Signed-off-by: Wang Liang <wangliang74@huawei.com>
Acked-by: Jussi Maki <joamaki@gmail.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250321044852.1086551-1-wangliang74@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Rajani Kantha <681739313@139.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c15d5c62ab ]
When a netdev issues a RX async resync request for a TLS connection,
the TLS module handles it by logging record headers and attempting to
match them to the tcp_sn provided by the device. If a match is found,
the TLS module approves the tcp_sn for resynchronization.
While waiting for a device response, the TLS module also increments
rcd_delta each time a new TLS record is received, tracking the distance
from the original resync request.
However, if the device response is delayed or fails (e.g due to
unstable connection and device getting out of tracking, hardware
errors, resource exhaustion etc.), the TLS module keeps logging and
incrementing, which can lead to a WARN() when rcd_delta exceeds the
threshold.
To address this, introduce tls_offload_rx_resync_async_request_cancel()
to explicitly cancel resync requests when a device response failure is
detected. Call this helper also as a final safeguard when rcd_delta
crosses its threshold, as reaching this point implies that earlier
cancellation did not occur.
Signed-off-by: Shahar Shitrit <shshitrit@nvidia.com>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/1761508983-937977-3-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 61fafbee6c ]
The GSO segmentation functions for ESP tunnel mode
(xfrm4_tunnel_gso_segment and xfrm6_tunnel_gso_segment) were
determining the inner packet's L2 protocol type by checking the static
x->inner_mode.family field from the xfrm state.
This is unreliable. In tunnel mode, the state's actual inner family
could be defined by x->inner_mode.family or by
x->inner_mode_iaf.family. Checking only the former can lead to a
mismatch with the actual packet being processed, causing GSO to create
segments with the wrong L2 header type.
This patch fixes the bug by deriving the inner mode directly from the
packet's inner protocol stored in XFRM_MODE_SKB_CB(skb)->protocol.
Instead of replicating the code, this patch modifies the
xfrm_ip2inner_mode helper function. It now correctly returns
&x->inner_mode if the selector family (x->sel.family) is already
specified, thereby handling both specific and AF_UNSPEC cases
appropriately.
With this change, ESP GSO can use xfrm_ip2inner_mode to get the
correct inner mode. It doesn't affect existing callers, as the updated
logic now mirrors the checks they were already performing externally.
Fixes: 26dbd66eab ("esp: choose the correct inner protocol for GSO on inter address family tunnels")
Signed-off-by: Jianbo Liu <jianbol@nvidia.com>
Reviewed-by: Cosmin Ratiu <cratiu@nvidia.com>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 14ad6ed30a ]
Sabrina reported the following splat:
WARNING: CPU: 0 PID: 1 at net/core/dev.c:6935 netif_napi_add_weight_locked+0x8f2/0xba0
Modules linked in:
CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Not tainted 6.14.0-rc1-net-00092-g011b03359038 #996
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Arch Linux 1.16.3-1-1 04/01/2014
RIP: 0010:netif_napi_add_weight_locked+0x8f2/0xba0
Code: e8 c3 e6 6a fe 48 83 c4 28 5b 5d 41 5c 41 5d 41 5e 41 5f c3 cc cc cc cc c7 44 24 10 ff ff ff ff e9 8f fb ff ff e8 9e e6 6a fe <0f> 0b e9 d3 fe ff ff e8 92 e6 6a fe 48 8b 04 24 be ff ff ff ff 48
RSP: 0000:ffffc9000001fc60 EFLAGS: 00010293
RAX: 0000000000000000 RBX: ffff88806ce48128 RCX: 1ffff11001664b9e
RDX: ffff888008f00040 RSI: ffffffff8317ca42 RDI: ffff88800b325cb6
RBP: ffff88800b325c40 R08: 0000000000000001 R09: ffffed100167502c
R10: ffff88800b3a8163 R11: 0000000000000000 R12: ffff88800ac1c168
R13: ffff88800ac1c168 R14: ffff88800ac1c168 R15: 0000000000000007
FS: 0000000000000000(0000) GS:ffff88806ce00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffff888008201000 CR3: 0000000004c94001 CR4: 0000000000370ef0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
gro_cells_init+0x1ba/0x270
xfrm_input_init+0x4b/0x2a0
xfrm_init+0x38/0x50
ip_rt_init+0x2d7/0x350
ip_init+0xf/0x20
inet_init+0x406/0x590
do_one_initcall+0x9d/0x2e0
do_initcalls+0x23b/0x280
kernel_init_freeable+0x445/0x490
kernel_init+0x20/0x1d0
ret_from_fork+0x46/0x80
ret_from_fork_asm+0x1a/0x30
</TASK>
irq event stamp: 584330
hardirqs last enabled at (584338): [<ffffffff8168bf87>] __up_console_sem+0x77/0xb0
hardirqs last disabled at (584345): [<ffffffff8168bf6c>] __up_console_sem+0x5c/0xb0
softirqs last enabled at (583242): [<ffffffff833ee96d>] netlink_insert+0x14d/0x470
softirqs last disabled at (583754): [<ffffffff8317c8cd>] netif_napi_add_weight_locked+0x77d/0xba0
on kernel built with MAX_SKB_FRAGS=45, where SKB_WITH_OVERHEAD(1024)
is smaller than GRO_MAX_HEAD.
Such built additionally contains the revert of the single page frag cache
so that napi_get_frags() ends up using the page frag allocator, triggering
the splat.
Note that the underlying issue is independent from the mentioned
revert; address it ensuring that the small head cache will fit either TCP
and GRO allocation and updating napi_alloc_skb() and __netdev_alloc_skb()
to select kmalloc() usage for any allocation fitting such cache.
Reported-by: Sabrina Dubroca <sd@queasysnail.net>
Suggested-by: Eric Dumazet <edumazet@google.com>
Fixes: 3948b05950 ("net: introduce a config option to tweak MAX_SKB_FRAGS")
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
[ Minor context change fixed. ]
Signed-off-by: Wenshan Lan <jetlan9@163.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 8d59fba493 upstream.
In the parse_adv_monitor_pattern() function, the value of
the 'length' variable is currently limited to HCI_MAX_EXT_AD_LENGTH(251).
The size of the 'value' array in the mgmt_adv_pattern structure is 31.
If the value of 'pattern[i].length' is set in the user space
and exceeds 31, the 'patterns[i].value' array can be accessed
out of bound when copied.
Increasing the size of the 'value' array in
the 'mgmt_adv_pattern' structure will break the userspace.
Considering this, and to avoid OOB access revert the limits for 'offset'
and 'length' back to the value of HCI_MAX_AD_LENGTH.
Found by InfoTeCS on behalf of Linux Verification Center
(linuxtesting.org) with SVACE.
Fixes: db08722fc7 ("Bluetooth: hci_core: Fix missing instances using HCI_MAX_AD_LENGTH")
Cc: stable@vger.kernel.org
Signed-off-by: Ilia Gavrilov <Ilia.Gavrilov@infotecs.ru>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 66048f8b3c ]
During recent testing with the netem qdisc to inject delays into TCP
traffic, we observed that our CLS BPF program failed to function correctly
due to incorrect classid retrieval from task_get_classid(). The issue
manifests in the following call stack:
bpf_get_cgroup_classid+5
cls_bpf_classify+507
__tcf_classify+90
tcf_classify+217
__dev_queue_xmit+798
bond_dev_queue_xmit+43
__bond_start_xmit+211
bond_start_xmit+70
dev_hard_start_xmit+142
sch_direct_xmit+161
__qdisc_run+102 <<<<< Issue location
__dev_xmit_skb+1015
__dev_queue_xmit+637
neigh_hh_output+159
ip_finish_output2+461
__ip_finish_output+183
ip_finish_output+41
ip_output+120
ip_local_out+94
__ip_queue_xmit+394
ip_queue_xmit+21
__tcp_transmit_skb+2169
tcp_write_xmit+959
__tcp_push_pending_frames+55
tcp_push+264
tcp_sendmsg_locked+661
tcp_sendmsg+45
inet_sendmsg+67
sock_sendmsg+98
sock_write_iter+147
vfs_write+786
ksys_write+181
__x64_sys_write+25
do_syscall_64+56
entry_SYSCALL_64_after_hwframe+100
The problem occurs when multiple tasks share a single qdisc. In such cases,
__qdisc_run() may transmit skbs created by different tasks. Consequently,
task_get_classid() retrieves an incorrect classid since it references the
current task's context rather than the skb's originating task.
Given that dev_queue_xmit() always executes with bh disabled, we can use
softirq_count() instead to obtain the correct classid.
The simple steps to reproduce this issue:
1. Add network delay to the network interface:
such as: tc qdisc add dev bond0 root netem delay 1.5ms
2. Build two distinct net_cls cgroups, each with a network-intensive task
3. Initiate parallel TCP streams from both tasks to external servers.
Under this specific condition, the issue reliably occurs. The kernel
eventually dequeues an SKB that originated from Task-A while executing in
the context of Task-B.
It is worth noting that it will change the established behavior for a
slightly different scenario:
<sock S is created by task A>
<class ID for task A is changed>
<skb is created by sock S xmit and classified>
prior to this patch the skb will be classified with the 'new' task A
classid, now with the old/original one. The bpf_get_cgroup_classid_curr()
function is a more appropriate choice for this case.
Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Thomas Graf <tgraf@suug.ch>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Nikolay Aleksandrov <razor@blackwall.org>
Link: https://patch.msgid.link/20250902062933.30087-1-laoar.shao@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 21f82062d0 ]
An exchange with a NFC target must complete within NCI_DATA_TIMEOUT.
A delay of 700 ms is not sufficient for cryptographic operations on smart
cards. CardOS 6.0 may need up to 1.3 seconds to perform 256-bit ECDH
or 3072-bit RSA. To prevent brute-force attacks, passports and similar
documents introduce even longer delays into access control protocols
(BAC/PACE).
The timeout should be higher, but not too much. The expiration allows
us to detect that a NFC target has disappeared.
Signed-off-by: Juraj Šarinay <juraj@sarinay.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://patch.msgid.link/20250902113630.62393-1-juraj@sarinay.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8f12d1137c ]
It is possible for bpf_xdp_adjust_tail() to free all fragments. The
kfunc currently clears the XDP_FLAGS_HAS_FRAGS bit, but not
XDP_FLAGS_FRAGS_PF_MEMALLOC. So far, this has not caused a issue when
building sk_buff from xdp_buff since all readers of xdp_buff->flags
use the flag only when there are fragments. Clear the
XDP_FLAGS_FRAGS_PF_MEMALLOC bit as well to make the flags correct.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://patch.msgid.link/20250922233356.3356453-2-ameryhung@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 751463ceef ]
Periodic advertising enabled flag cannot be tracked by the enabled
flag since advertising and periodic advertising each can be
enabled/disabled separately from one another causing the states to be
inconsistent when for example an advertising set is disabled its
enabled flag is set to false which is then used for periodic which has
not being disabled.
Fixes: eca0ae4aea ("Bluetooth: Add initial implementation of BIS connections")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0d92808024 ]
This fixes the state tracking of advertisement set/instance 0x00 which
is considered a legacy instance and is not tracked individually by
adv_instances list, previously it was assumed that hci_dev itself would
track it via HCI_LE_ADV but that is a global state not specifc to
instance 0x00, so to fix it a new flag is introduced that only tracks the
state of instance 0x00.
Fixes: 1488af7b8b ("Bluetooth: hci_sync: Fix hci_resume_advertising_sync")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit dd831ac822 upstream.
To prevent a potential crash in agg_dequeue (net/sched/sch_qfq.c)
when cl->qdisc->ops->peek(cl->qdisc) returns NULL, we check the return
value before using it, similar to the existing approach in sch_hfsc.c.
To avoid code duplication, the following changes are made:
1. Changed qdisc_warn_nonwc(include/net/pkt_sched.h) into a static
inline function.
2. Moved qdisc_peek_len from net/sched/sch_hfsc.c to
include/net/pkt_sched.h so that sch_qfq can reuse it.
3. Applied qdisc_peek_len in agg_dequeue to avoid crashing.
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com>
Link: https://patch.msgid.link/20250705212143.3982664-1-xmei5@asu.edu
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 21f4d45eba ]
Similarly to ipv4 tunnel, ipv6 version updates dev->needed_headroom, too.
While ipv4 tunnel headroom adjustment growth was limited in
commit 5ae1e9922b ("net: ip_tunnel: prevent perpetual headroom growth"),
ipv6 tunnel yet increases the headroom without any ceiling.
Reflect ipv4 tunnel headroom adjustment limit on ipv6 version.
Credits to Francesco Ruggeri, who was originally debugging this issue
and wrote local Arista-specific patch and a reproducer.
Fixes: 8eb30be035 ("ipv6: Create ip6_tnl_xmit")
Cc: Florian Westphal <fw@strlen.de>
Cc: Francesco Ruggeri <fruggeri05@gmail.com>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Link: https://patch.msgid.link/20251009-ip6_tunnel-headroom-v2-1-8e4dbd8f7e35@arista.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 374d345d9b ]
We currently push everyone to use padding to align 64b values
in netlink. Un-padded nla_put_u64() doesn't even exist any more.
The story behind this possibly start with this thread:
https://lore.kernel.org/netdev/20121204.130914.1457976839967676240.davem@davemloft.net/
where DaveM was concerned about the alignment of a structure
containing 64b stats. If user space tries to access such struct
directly:
struct some_stats *stats = nla_data(attr);
printf("A: %llu", stats->a);
lack of alignment may become problematic for some architectures.
These days we most often put every single member in a separate
attribute, meaning that the code above would use a helper like
nla_get_u64(), which can deal with alignment internally.
Even for arches which don't have good unaligned access - access
aligned to 4B should be pretty efficient.
Kernel and well known libraries deal with unaligned input already.
Padded 64b is quite space-inefficient (64b + pad means at worst 16B
per attr vs 32b which takes 8B). It is also more typing:
if (nla_put_u64_pad(rsp, NETDEV_A_SOMETHING_SOMETHING,
value, NETDEV_A_SOMETHING_PAD))
Create a new attribute type which will use 32 bits at netlink
level if value is small enough (probably most of the time?),
and (4B-aligned) 64 bits otherwise. Kernel API is just:
if (nla_put_uint(rsp, NETDEV_A_SOMETHING_SOMETHING, value))
Calling this new type "just" sint / uint with no specific size
will hopefully also make people more comfortable with using it.
Currently telling people "don't use u8, you may need the bits,
and netlink will round up to 4B, anyway" is the #1 comment
we give to newcomers.
In terms of netlink layout it looks like this:
0 4 8 12 16
32b: [nlattr][ u32 ]
64b: [ pad ][nlattr][ u64 ]
uint(32) [nlattr][ u32 ]
uint(64) [nlattr][ u64 ]
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stable-dep-of: 030e1c4566 ("macsec: read MACSEC_SA_ATTR_PN with nla_get_uint")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d860d1faa6 ]
The 'use' field in struct rose_neigh is used as a reference counter but
lacks atomicity. This can lead to race conditions where a rose_neigh
structure is freed while still being referenced by other code paths.
For example, when rose_neigh->use becomes zero during an ioctl operation
via rose_rt_ioctl(), the structure may be removed while its timer is
still active, potentially causing use-after-free issues.
This patch changes the type of 'use' from unsigned short to refcount_t and
updates all code paths to use rose_neigh_hold() and rose_neigh_put() which
operate reference counts atomically.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Takamitsu Iwai <takamitz@amazon.co.jp>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20250823085857.47674-3-takamitz@amazon.co.jp
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dcb3465902 ]
The current rose_remove_neigh() performs two distinct operations:
1. Removes rose_neigh from rose_neigh_list
2. Frees the rose_neigh structure
Split these operations into separate functions to improve maintainability
and prepare for upcoming refcount_t conversion. The timer cleanup remains
in rose_remove_neigh() because free operations can be called from timer
itself.
This patch introduce rose_neigh_put() to handle the freeing of rose_neigh
structures and modify rose_remove_neigh() to handle removal only.
Signed-off-by: Takamitsu Iwai <takamitz@amazon.co.jp>
Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20250823085857.47674-2-takamitz@amazon.co.jp
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: d860d1faa6 ("net: rose: convert 'use' field to refcount_t")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6bbd0d3f0c ]
Function set_name_sync() uses hdev->dev_name field to send
HCI_OP_WRITE_LOCAL_NAME command, but copying from data to hdev->dev_name
is called after mgmt cmd was queued, so it is possible that function
set_name_sync() will read old name value.
This change adds name as a parameter for function hci_update_name_sync()
to avoid race condition.
Fixes: 6f6ff38a1e ("Bluetooth: hci_sync: Convert MGMT_OP_SET_LOCAL_NAME")
Signed-off-by: Pavel Shpakovskiy <pashpakovskii@salutedevices.com>
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 240fd40552 ]
Add support for the independent control state machine per IEEE
802.1AX-2008 5.4.15 in addition to the existing implementation of the
coupled control state machine.
Introduces two new states, AD_MUX_COLLECTING and AD_MUX_DISTRIBUTING in
the LACP MUX state machine for separated handling of an initial
Collecting state before the Collecting and Distributing state. This
enables a port to be in a state where it can receive incoming packets
while not still distributing. This is useful for reducing packet loss when
a port begins distributing before its partner is able to collect.
Added new functions such as bond_set_slave_tx_disabled_flags and
bond_set_slave_rx_enabled_flags to precisely manage the port's collecting
and distributing states. Previously, there was no dedicated method to
disable TX while keeping RX enabled, which this patch addresses.
Note that the regular flow process in the kernel's bonding driver remains
unaffected by this patch. The extension requires explicit opt-in by the
user (in order to ensure no disruptions for existing setups) via netlink
support using the new bonding parameter coupled_control. The default value
for coupled_control is set to 1 so as to preserve existing behaviour.
Signed-off-by: Aahil Awatramani <aahila@google.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Link: https://lore.kernel.org/r/20240202175858.1573852-1-aahila@google.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Stable-dep-of: 0599640a21 ("bonding: send LACPDUs periodically in passive mode after receiving partner's LACPDU")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6b04716cdc ]
When SAE commit is sent and received in response, there's no
ordering for the SAE confirm messages. As such, don't call
drivers to stop listening on the channel when the confirm
message is still expected.
This fixes an issue if the local confirm is transmitted later
than the AP's confirm, for iwlwifi (and possibly mt76) the
AP's confirm would then get lost since the device isn't on
the channel at the time the AP transmit the confirm.
For iwlwifi at least, this also improves the overall timing
of the authentication handshake (by about 15ms according to
the report), likely since the session protection won't be
aborted and rescheduled.
Note that even before this, mgd_complete_tx() wasn't always
called for each call to mgd_prepare_tx() (e.g. in the case
of WEP key shared authentication), and the current drivers
that have the complete callback don't seem to mind. Document
this as well though.
Reported-by: Jan Hendrik Farr <kernel@jfarr.cc>
Closes: https://lore.kernel.org/all/aB30Ea2kRG24LINR@archlinux/
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20250609213232.12691580e140.I3f1d3127acabcd58348a110ab11044213cf147d3@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
