Yun Zhou [Mon, 22 Jun 2026 07:43:50 +0000 (15:43 +0800)]
net: mvneta: re-enable percpu interrupt on resume
On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu
IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls
disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules
NAPI poll, which calls enable_percpu_irq() on completion to unmask.
If suspend occurs while NAPI poll is pending (between
disable_percpu_irq in the ISR and enable_percpu_irq in poll
completion), the interrupt is never re-enabled:
1. mvneta_percpu_isr: disable_percpu_irq() + napi_schedule()
=> MPIC masked, percpu_enabled cpumask bit cleared
2. NAPI poll does not complete before suspend proceeds
(on PREEMPT_RT this is highly likely since softirqs run in
ksoftirqd which gets frozen; on non-RT it can happen when
softirq processing is deferred to ksoftirqd)
3. mvneta_stop_dev => napi_disable(): cancels the pending poll
without executing the completion path
4. suspend_device_irqs => IRQCHIP_MASK_ON_SUSPEND: masks MPIC
(already masked, but records IRQS_SUSPENDED)
5. Resume: mpic_resume checks irq_percpu_is_enabled() => false
(bit was cleared in step 1) => skips unmask
6. mvneta_start_dev only restores device-level INTR_NEW_MASK,
does not touch the MPIC per-CPU mask
Result: MPIC per-CPU interrupt stays masked permanently. The NIC
generates interrupts (INTR_NEW_CAUSE != 0) but the CPU never
receives them, causing complete loss of network connectivity.
Fix by calling on_each_cpu(mvneta_percpu_enable) in the resume path
to unconditionally unmask the MPIC per-CPU interrupt regardless of
pre-suspend state.
Fixes: 12bb03b436da ("net: mvneta: Handle per-cpu interrupts") Signed-off-by: Yun Zhou <yun.zhou@windriver.com> Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Link: https://patch.msgid.link/20260622074350.1666290-1-yun.zhou@windriver.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ratheesh Kannoth [Mon, 22 Jun 2026 03:42:29 +0000 (09:12 +0530)]
octeontx2-af: fix CGX debugfs RVU AF PCI reference leaks
CGX per-lmac debugfs seq readers obtained struct rvu via
pci_get_drvdata(pci_get_device(..., PCI_DEVID_OCTEONTX2_RVU_AF, ...)),
which leaks a PCI device reference on every read. Store rvu and the CGX
handle in debugfs inode private data when creating stats, mac_filter,
and fwdata files (one context per CGX), and use debugfs aux numbers for
fwdata so lmac_id matches the other CGX debugfs entries.
Fixes: f967488d095e ("octeontx2-af: Add per CGX port level NIX Rx/Tx counters") Fixes: dbc52debf95f ("octeontx2-af: Debugfs support for DMAC filters") Fixes: 49f02e6877d1 ("Octeontx2-af: Debugfs support for firmware data") Cc: Linu Cherian <lcherian@marvell.com> Reported-by: Yuho Choi <dbgh9129@gmail.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com> Link: https://patch.msgid.link/20260622034229.2254145-1-rkannoth@marvell.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
NIX maximum number of LFs can be set via devlink command
but that can be done before assigning any LFs to a PF/VF.
The condition used to check whether any LFs are assigned is
incorrect. This patch fixes that condition.
Haoxiang Li [Sun, 21 Jun 2026 03:17:14 +0000 (11:17 +0800)]
net: wwan: t7xx: destroy DMA pool on CLDMA late init failure
t7xx_cldma_late_init() creates md_ctrl->gpd_dmapool before
initializing the TX and RX rings. If any ring initialization
fails, the error path frees the already initialized rings but
leaves the DMA pool allocated.
Destroy md_ctrl->gpd_dmapool on the late-init failure path
to avoid leaking the DMA pool.
Fixes: 39d439047f1d ("net: wwan: t7xx: Add control DMA interface") Cc: stable@vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com> Reviewed-by: Loic Poulain <loic.poulain@oss.qualcomm.com> Link: https://patch.msgid.link/20260621031714.3605022-1-haoxiang_li2024@163.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Lorenzo Bianconi [Sat, 20 Jun 2026 15:04:51 +0000 (17:04 +0200)]
net: airoha: fix BQL underflow in shared QDMA TX ring
When multiple netdevs share a QDMA TX ring and one device is stopped,
netdev_tx_reset_subqueue() zeroes that device's BQL counters while its
pending skbs remain in the shared HW TX ring. When NAPI later completes
those skbs via netdev_tx_completed_queue(), the already-zeroed
dql->num_queued counter underflows.
Fix the issue:
- Remove netdev_tx_reset_subqueue() from airoha_dev_stop() so pending
skbs are completed naturally by NAPI with proper BQL accounting.
- Rework airoha_qdma_tx_cleanup() to disable TX DMA, flush BQL
counters, DMA-unmap and free all pending skbs while skb->dev
references are still valid. Use a per-queue flushing flag checked
under q->lock in airoha_dev_xmit() to prevent races between teardown
and transmit. Call airoha_qdma_stop_napi() before
airoha_qdma_tx_cleanup() at the call sites.
- Move DMA engine start into probe. Split DMA teardown so TX DMA is
disabled in airoha_qdma_tx_cleanup() and RX DMA in
airoha_qdma_cleanup().
- Remove qdma->users counter since DMA lifetime is now tied to
probe/cleanup rather than per-netdev open/stop.
Fixes: a9c2ca61fec7 ("net: airoha: Support multiple net_devices for a single FE GDM port") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260620-airoha-bql-fixes-v3-1-76b95374e63e@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jan Klos [Sat, 20 Jun 2026 01:19:53 +0000 (03:19 +0200)]
net: phy: realtek: Clear MDIO_AN_10GBT_CTRL_ADV10G bit
On RTL8127A connected to a link partner that advertises 10000baseT
speed cannot be changed to anything other than 10000baseT as 10GbE
is always advertised regardless of any setting. Fix this by
clearing MDIO_AN_10GBT_CTRL_ADV10G bit in rtl822x_config_aneg()'s
call to phy_modify_mmd_changed().
Fixes: 83d962316128 ("net: phy: realtek: add RTL8127-internal PHY") Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Signed-off-by: Jan Klos <honza.klos@gmail.com> Link: https://patch.msgid.link/20260620011956.37181-1-honza.klos@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ratheesh Kannoth [Fri, 19 Jun 2026 09:51:00 +0000 (15:21 +0530)]
octeontx2-af: npc: cn20k: Fix subbank free list indexing for search order
subbank_srch_order[i] is the physical subbank at search-order slot i,
so each subbank's arr_idx must be i (its slot), not
subbank_srch_order[sb->idx]. The old logic mis-keyed xa_sb_free
and broke allocation traversal order.
Populate arr_idx and xa_sb_free in a single pass over the search
order after subbank structs are initialized.
Fixes: 7ac9d4c4075c ("octeontx2-af: npc: cn20k: add subbank search order control") Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260619095100.1864440-1-rkannoth@marvell.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: mana: Fall back to standard MTU when PF reports adapter_mtu of 0
Commit d7709812e13d ("net: mana: hardening: Validate adapter_mtu from
MANA_QUERY_DEV_CONFIG") rejected any adapter_mtu value smaller than
ETH_MIN_MTU + ETH_HLEN, including 0, returning -EPROTO and failing
mana_probe().
Some older PF firmware versions still in the field report
adapter_mtu as 0 in the MANA_QUERY_DEV_CONFIG response. With the
hardening check in place, the MANA VF driver now fails to load on
those hosts, breaking networking entirely for guests.
MANA hardware always supports the standard Ethernet MTU. Treat a
reported adapter_mtu of 0 as "the PF did not advertise a value" and
fall back to ETH_FRAME_LEN, the same value used for the pre-V2
message version path. Only jumbo frames remain unavailable until
the PF reports a valid MTU.
Other small-but-nonzero bogus values are still rejected, preserving
the original protection against the unsigned-subtraction wrap that
would otherwise let ndev->max_mtu underflow to a huge value.
Fixes: d7709812e13d ("net: mana: hardening: Validate adapter_mtu from MANA_QUERY_DEV_CONFIG") Signed-off-by: Erni Sri Satya Vennela <ernis@linux.microsoft.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260619055348.467224-1-ernis@linux.microsoft.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
dsa_unregister_switch() frees the dsa_port objects. If a CRC error
schedules the work during teardown it can run after the ports have been
freed and dereference freed memory.
Guard the port walk with MXL862XX_FLAG_WORK_STOPPED, which is already set
before dsa_unregister_switch(). DSA tears the ports down under
rtnl_lock(), so checking the flag under rtnl_lock() means the work either
runs before teardown and sees valid ports, or runs afterwards, observes
the flag and skips the walk. This mirrors the host_flood_work handler,
which skips torn-down ports under rtnl_lock().
Daniel Golle [Fri, 19 Jun 2026 03:39:25 +0000 (04:39 +0100)]
net: dsa: mxl862xx: avoid unaligned 16-bit access in api_wrap
The MXL862XX_API_* macros pass the address of a stack-allocated, __packed
firmware-ABI struct to mxl862xx_api_wrap() as a void *. The struct has an
alignment of 1, so the compiler is free to place it at an odd address.
mxl862xx_api_wrap() reinterprets that buffer as a __le16 * and accesses it
with data[i], for which the compiler assumes the natural 2-byte alignment
of __le16 and emits aligned 16-bit loads/stores (e.g. lhu/sh on MIPS).
When the buffer lands on an odd address these fault on architectures that
do not support unaligned access, such as MIPS32.
-Waddress-of-packed-member does not catch this: the packed origin is
laundered through the void * parameter, so the cast inside api_wrap looks
alignment-safe to the compiler and no warning is emitted.
Use get_unaligned_le16()/put_unaligned_le16() for the three 16-bit word
accesses. The byte accesses (*(u8 *)&data[i], crc16()) are already safe
and are left unchanged.
David Yang [Thu, 18 Jun 2026 14:01:55 +0000 (22:01 +0800)]
net: dsa: realtek: fix memory leak in rtl8366rb_setup_led()
led_classdev_register_ext() only reads init_data.devicename - it never
stores the pointer. However, the caller allocated devicename with
kasprintf() but never freed it, leaking the string memory.
Fix it with a stack buffer to avoid dynamic buffers completely.
Fixes: 32d617005475 ("net: dsa: realtek: add LED drivers for rtl8366rb") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Linus Walleij <linusw@kernel.org> Link: https://patch.msgid.link/20260618140200.1888707-1-mmyangfl@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ixp4xx_hss_probe() allocates two HDLC netdevs. The first one is stored
in ndev, initialized, and registered with register_hdlc_device(). The
second one is stored in port->netdev and later used by the remove path
for unregister_hdlc_device() and free_netdev().
This means that the registered netdev is not the same object that is
unregistered and freed on remove. It also leaks the first allocation if
the second alloc_hdlcdev() call fails, and the first allocation is not
checked before ndev is used.
Older code allocated the HDLC netdev only once and stored the same object
in both the local variable and port->netdev. The buggy conversion split
this into two alloc_hdlcdev() calls. A later rename changed the local
variable name to ndev, but the underlying mismatch remained.
Fix this by allocating the HDLC netdev only once and assigning the same
object to port->netdev.
Fixes: 99ebe65eb9c0 ("net: ixp4xx_hss: move out assignment in if condition") Cc: stable@vger.kernel.org Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com> Reviewed-by: Linus Walleij <linusw@kernel.org> Link: https://patch.msgid.link/20260622043015.643637-1-haoxiang_li2024@163.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Lorenzo Bianconi [Fri, 19 Jun 2026 11:37:14 +0000 (13:37 +0200)]
net: airoha: fix netif_set_real_num_tx_queues for sparse QoS channels
airoha_tc_htb_alloc_leaf_queue() assigns queue IDs based on the channel
index (opt->qid = AIROHA_NUM_TX_RING + channel), but updates
real_num_tx_queues with a simple increment (num_tx_queues + 1). When QoS
channels are allocated sparsely (e.g., channels 0 and 3 without 1 and
2), the returned qid can exceed real_num_tx_queues, causing out-of-bounds
accesses in the networking stack.
For example, allocating channel 0 then channel 3 results in
real_num_tx_queues = 34 but qid = 35, which is out of range [0, 34).
Fix this by computing real_num_tx_queues based on the highest active
channel index rather than using a simple counter, in both the allocation
and deletion paths.
Lorenzo Bianconi [Fri, 19 Jun 2026 11:37:13 +0000 (13:37 +0200)]
net: airoha: Fix off-by-one in airoha_tc_remove_htb_queue()
airoha_tc_htb_alloc_leaf_queue() computes the HTB QoS channel index
as opt->classid % AIROHA_NUM_QOS_CHANNELS and stores it in qos_sq_bmap.
However, airoha_tc_remove_htb_queue() clears the HTB configuration
using queue + 1 as the channel index, causing an off-by-one error.
Use queue directly as the QoS channel index to match the allocation
logic.
Jakub Kicinski [Mon, 22 Jun 2026 15:47:53 +0000 (08:47 -0700)]
eth: fbnic: fix ordering of heartbeat vs ownership
When requesting ownership of the NIC (MAC/PHY control), we set up
the heartbeat to look stale:
/* Initialize heartbeat, set last response to 1 second in the past
* so that we will trigger a timeout if the firmware doesn't respond
*/
fbd->last_heartbeat_response = req_time - HZ;
fbd->last_heartbeat_request = req_time;
The scheme is a bit odd, but it should work in principle.
Fix the ordering of operations. We have to set up the stale heartbeat
before we send the message. Otherwise if the response is very fast
we will override it. This triggers on QEMU if we run on the core
that handles the IRQ, and results in ndo_open failing with ETIMEDOUT.
The change in ordering doesn't impact releasing the ownership.
Both ndo_stop and heartbeat check are under rtnl_lock.
Fixes: 20d2e88cc746 ("eth: fbnic: Add initial messaging to notify FW of our presence") Reviewed-by: Alexander Duyck <alexanderduyck@fb.com> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com> Link: https://patch.msgid.link/20260622154753.827506-1-kuba@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
ipv6: fix error handling in disable_ipv6 sysctl
While working on a different IPv6 patch series I have spotted multiple
minor bugs around sysctl error handling and notifications. In general,
they are not serious issues.
In addition, there is one more issue in forwarding sysctl as it does not
check for CAP_NET_ADMIN for the namespace. I am keeping that patch out
of this series and I am aiming it at the net-next tree once it re-opens.
During v3, Ido's pointed out that it is unnecessary to reset the
position pointer when the return value is negative as at
new_sync_write() the ppos is only advanced when ret return value is
positive. That means we can get rid of that operation in ipv4/ipv6
sysctls. That is going to be sent to net-next too.
====================
ipv6: fix missing notification for ignore_routes_with_linkdown
When changing the ignore_routes_with_linkdown sysctl for a specific
interface, the RTM_NEWNETCONF netlink notification was not being emitted
to userspace. Fix this by emitting the notification when needed.
In addition, fix bogus return value for successful "all" and specific
interface write operation leading to a wrong reset of the position
pointer.
Fixes: 35103d11173b ("net: ipv6 sysctl option to ignore routes when nexthop link is down") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-7-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ipv6: fix state corruption during proxy_ndp sysctl restart
When handling proxy_ndp, if rtnl_net_trylock() fails, the operation is
retried but as the value was already modified by the initial
proc_dointvec() call, the restarted syscall will read the newly modified
value as the 'old' state.
Fix this by taking the RTNL lock before parsing the input value if the
operation is a write.
Fixes: c92d5491a6d9 ("netconf: add support for IPv6 proxy_ndp") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-6-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When writing to the disable_policy sysctl, if proc_dointvec() fails to
parse the input, it returns a negative error code. The current
implementation is resetting the position argument even if an error
occurred during proc_dointvec() and not only during sysctl restart.
Fix this by checking the return value of proc_dointvec() and returning
early on failure.
Fixes: df789fe75206 ("ipv6: Provide ipv6 version of "disable_policy" sysctl") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-5-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When writing to the forwarding sysctl, if proc_dointvec() fails to parse
the input, it returns a negative error code. The current implementation
is overwriting that error for write operations.
This results in a silent failure, it returns a successful write although
the configuration was not modified at all. When modifying the "all"
variant it can also modify the configuration of existing interfaces to
the wrong value.
Fix this by checking the return value of proc_dointvec() and returning
early on failure. In addition, adjust return code of
addrconf_fixup_forwarding() for successful operation.
Fixes: b325fddb7f86 ("ipv6: Fix sysctl unregistration deadlock") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-4-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ipv6: fix error handling in ignore_routes_with_linkdown sysctl
When writing to the ignore_routes_with_linkdown sysctl, if
proc_dointvec() fails to parse the input, it returns a negative error
code. The current implementation is overwriting that error for write
operations.
This results in a silent failure, it returns a successful write although
the configuration was not modified at all. When modifying the "all"
variant it can also modify the configuration of existing interfaces to
the wrong value.
Fix this by checking the return value of proc_dointvec() and returning
early on failure.
Fixes: 35103d11173b ("net: ipv6 sysctl option to ignore routes when nexthop link is down") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-3-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When writing to the disable_ipv6 sysctl, if proc_dointvec() fails to
parse the input, it returns a negative error code. The current
implementation is overwriting that error for write operations.
This results in a silent failure, it returns a successful write although
the configuration was not modified at all. When modifying the "all"
variant it can also modify the configuration of existing interfaces to
the wrong value.
Fix this by checking the return value of proc_dointvec() and returning
early on failure.
Fixes: 56d417b12e57 ("IPv6: Add 'autoconf' and 'disable_ipv6' module parameters") Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Signed-off-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260622130857.5115-2-fmancera@suse.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Runyu Xiao [Fri, 19 Jun 2026 15:18:16 +0000 (23:18 +0800)]
net: au1000: move free_irq out of the close-time spinlocked section
au1000_close() calls free_irq() while aup->lock is still held with
spin_lock_irqsave(). free_irq() can sleep because it takes the IRQ
descriptor request mutex, so it does not belong inside the close-time
spinlocked section.
This was found by our static analysis tool and then confirmed by manual
review of the in-tree au1000_close() .ndo_stop path. The reviewed path
keeps aup->lock held across the MAC reset, queue stop and
free_irq(dev->irq, dev).
A directed runtime validation kept that ndo_stop carrier and the same
free_irq(dev->irq, dev) operation under the driver lock. Lockdep reported
"BUG: sleeping function called from invalid context" and "Invalid wait
context" while free_irq() was taking desc->request_mutex, with
au1000_close() and free_irq() on the stack.
Drop aup->lock before freeing the IRQ. The protected close-time work still
stops the device and queue before IRQ teardown, but the sleepable IRQ core
path now runs outside the spinlocked section.
Xin Long [Sat, 20 Jun 2026 15:48:54 +0000 (11:48 -0400)]
sctp: fix err_chunk memory leaks in INIT handling
When sctp_verify_init() encounters unrecognized parameters, it allocates an
err_chunk to report them. However, this chunk is leaked in several code
paths:
1. In sctp_sf_do_5_1B_init(), if security_sctp_assoc_request() fails after
sctp_verify_init() has populated err_chunk, the function returns
immediately without freeing it.
2. In sctp_sf_do_unexpected_init(), the same leak occurs on the
security_sctp_assoc_request() failure path.
3. In sctp_sf_do_unexpected_init(), on the success path after copying
unrecognized parameters to the INIT-ACK, the function returns without
freeing err_chunk, unlike sctp_sf_do_5_1B_init() which properly frees
it.
Fix all three leaks by adding sctp_chunk_free(err_chunk) calls before
returning in the error paths and on the success path in
sctp_sf_do_unexpected_init().
Fixes: c081d53f97a1 ("security: pass asoc to sctp_assoc_request and sctp_sk_clone") Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Sashiko <sashiko-bot@kernel.org> Signed-off-by: Xin Long <lucien.xin@gmail.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/0656704f1b0158287c98aec09ba36c83e4a537ab.1781970534.git.lucien.xin@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jamal Hadi Salim [Sat, 20 Jun 2026 13:07:49 +0000 (09:07 -0400)]
net/sched: cls_api: Handle TC_ACT_CONSUMED in tcf_qevent_handle
tcf_classify() can return TC_ACT_CONSUMED while the skb is held by the
defragmentation engine (e.g. act_ct on out-of-order fragments). When
that happens the skb is no longer owned by the caller and must not be
touched again.
tcf_qevent_handle() did not handle TC_ACT_CONSUMED: it fell through the
switch and returned the skb to the caller as if classification had
passed. The only qdisc that wires up qevents today is RED, via three call sites
(qe_mark on RED_PROB_MARK/HARD_MARK, qe_early_drop on congestion_drop)
red_enqueue() was continuing to operate on an skb it no longer owns in this
case -- enqueueing it, dropping it, or updating statistics. Resulting in a UAF.
(with ct defrag enabled and traffic that produces out-of-order
fragments, e.g. a fragmented UDP stream)
Handle TC_ACT_CONSUMED in tcf_qevent_handle() the same way the ingress
and egress fast paths do: treat it as stolen and return NULL without
touching the skb. Unlike the TC_ACT_STOLEN case, the skb must not be
dropped/freed here, as it is no longer owned by us.
Fixes: 3f14b377d01d ("net/sched: act_ct: fix skb leak and crash on ooo frags") Reported-by: Zero Day Initiative <zdi-disclosures@trendmicro.com> Tested-by: Victor Nogueira <victor@mojatatu.com> Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com> Link: https://patch.msgid.link/20260620130749.226642-1-jhs@mojatatu.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Sitnicki [Fri, 19 Jun 2026 17:09:29 +0000 (19:09 +0200)]
selftests/bpf: Add LWT encap tests for skb metadata
Test that an LWT encapsulation does not silently corrupt XDP metadata
sitting in the skb headroom. Exercise all three LWT dispatch paths:
- BPF LWT xmit prog reserves headroom on the LWT .xmit redirect,
- mpls pushes an MPLS label on the LWT .xmit redirect,
- seg6 in encap mode runs on the LWT .input redirect,
- ioam6 encap inserts an IOAM Hop-by-Hop option on LWT .output redirect.
Jakub Sitnicki [Fri, 19 Jun 2026 17:09:28 +0000 (19:09 +0200)]
net: lwtunnel: Drop skb metadata before LWT encapsulation
skb metadata is meant for passing information between XDP and TC. It lives
in the skb headroom, immediately before skb->data. LWT programs cannot
access the __sk_buff->data_meta pseudo-pointer to metadata.
However, LWT encapsulation prepends outer headers, moving skb->data back
over the headroom where the metadata sits. On an RX-originated (forwarded)
packet that still carries XDP metadata this goes wrong in two different
ways, depending on the encap type:
1. Non-BPF LWT encaps (mpls, seg6, ioam6 ...) call skb_push()/skb_pull()
and silently overwrite the metadata that sits in the headroom.
2) BPF LWT xmit calls bpf_skb_change_head(), which uses skb_data_move().
That helper expects metadata immediately before skb->data. But since
the IP output path runs LWT xmit before neighbour output has built
the outgoing L2 header, for forwarded packets skb->data points at the
L3 header while skb_mac_header() still points at the old L2 header.
skb_data_move() sees metadata ending at skb_mac_header(), not before
skb->data, warns and clears metadata:
Every encap funnels through the three LWT dispatch helpers, so drop the
metadata there, right before handing the skb to the encap op. This
single chokepoint covers all encap types and all three redirect modes:
Alternatively, we could clear the metadata right after TC ingress hook.
That would require a compromise, however. Metadata would become
inaccessible from TC egress (in setups where it actually reaches the
hook it tact, that is without any L2 tunnels on path).
1) xfrm: use compat translator only for u64 alignment mismatch
Gate the XFRM_USER_COMPAT translator on COMPAT_FOR_U64_ALIGNMENT
so 32-bit compat tasks on arches whose 32-bit ABI already matches
the native 64-bit layout are no longer rejected with -EOPNOTSUPP.
From Sanman Pradhan.
2) net: af_key: initialize alg_key_len for IPComp states
Initialize the alg_key_len to 0 in the IPComp branch of
pfkey_msg2xfrm_state() so an uninitialized value cannot drive
xfrm_alg_len() into a slab-out-of-bounds kmemdup during
XFRM_MSG_MIGRATE. From Zijing Yin.
3) xfrm: Fix dev use-after-free in xfrm async resumption
Stash the original skb->dev and extend the RCU critical section
across xfrm_rcv_cb() and transport_finish() to prevent a
tunnel-device UAF and original-device refcount leak when a
callback replaces skb->dev. From Dong Chenchen.
4) xfrm: Fix xfrm state cache insertion race
Move the state-validity check inside xfrm_state_lock in the
input state cache insertion path so a state cannot be killed
between the check and the insert. From Herbert Xu.
5) xfrm: annotate data-races around xfrm_policy_count[] and xfrm_policy_default[]
Add READ_ONCE()/WRITE_ONCE() annotations on xfrm_policy_count
and xfrm_policy_default to silence the KCSAN data race reported
on net->xfrm.policy_count. From Eric Dumazet.
6) espintcp: use sk_msg_free_partial to fix partial send
Replace the manual skmsg accounting in espintcp with
sk_msg_free_partial() so the skmsg stays consistent on every
iteration and the partial-send accounting bugs go away.
From Sabrina Dubroca.
7) xfrm: validate selector family and prefixlen during match
Reject mismatched address families in xfrm_selector_match() and
bound prefixlen in addr4_match()/addr_match() to prevent the
shift-out-of-bounds syzbot reported when an AF_UNSPEC selector
with a large prefixlen is matched against an IPv4 flow.
From Eric Dumazet.
* tag 'ipsec-2026-06-22' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec:
xfrm: validate selector family and prefixlen during match
espintcp: use sk_msg_free_partial to fix partial send
xfrm: annotate data-races around xfrm_policy_count[] and xfrm_policy_default[]
xfrm: Fix xfrm state cache insertion race
xfrm: Fix dev use-after-free in xfrm async resumption
net: af_key: initialize alg_key_len for IPComp states
xfrm: use compat translator only for u64 alignment mismatch
====================
Nicolai Buchwitz [Mon, 22 Jun 2026 10:29:11 +0000 (12:29 +0200)]
net: usb: lan78xx: restore VLAN and hash filters after link up
Configured VLANs intermittently stop receiving traffic after a link
down/up cycle, e.g. when the network cable is unplugged and plugged back
in. VLAN filtering stays enabled but all VLAN-tagged frames are dropped
until a VLAN is added or removed again.
The LAN7801 datasheet (DS00002123E) states:
"A portion of the MAC operates on clocks generated by the Ethernet
PHY. During a PHY reset event, this portion of the MAC is designed to
not be taken out of reset until the PHY clocks are operational"
(section 8.10, MAC Reset Watchdog Timer)
"After a reset event, the RFE will automatically initialize the
contents of the VHF to 0h."
(section 7.1.4, VHF Organization)
Thus a link down/up cycle stops and restarts the PHY clock, resets the
PHY-clocked portion of the MAC, and the RFE clears its VLAN/DA hash
filter (VHF) memory. The VHF holds both the VLAN filter table and the
multicast hash table, but the driver never reprograms either from its
shadow copy once the link is back, so both stay empty.
Reprogram the VLAN filter and multicast hash tables on link up.
Reported-by: Sven Schuchmann <schuchmann@schleissheimer.de> Closes: https://lore.kernel.org/netdev/BEZP281MB224501E38B30BFDC4BD3D364D9E32@BEZP281MB2245.DEUP281.PROD.OUTLOOK.COM/T/#u Tested-by: Sven Schuchmann <schuchmann@schleissheimer.de> Fixes: 55d7de9de6c3 ("Microchip's LAN7800 family USB 2/3 to 10/100/1000 Ethernet device driver") Signed-off-by: Nicolai Buchwitz <nb@tipi-net.de> Link: https://patch.msgid.link/20260622102911.484045-1-nb@tipi-net.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet [Mon, 22 Jun 2026 11:18:25 +0000 (11:18 +0000)]
veth: fix NAPI leak in XDP enable error path
During XDP enablement in veth, if xdp_rxq_info_reg() or
xdp_rxq_info_reg_mem_model() fails, the driver rolls back the changes.
However, the rollback loop:
for (i--; i >= start; i--) {
decrements the loop index 'i' before the first iteration. This
correctly skips unregistering the rxq for the failed index 'i' (as
registration failed or was already cleaned up), but it also
erroneously skips calling netif_napi_deli() for rq[i].xdp_napi.
Since netif_napi_add() was already called for index 'i', this leaves
a dangling napi_struct in the device's napi_list. When the veth
device is later destroyed, the freed queue memory (which contains the
leaked NAPI structure) can be reused.
The subsequent device teardown iterates the NAPI list and
corrupts the reallocated memory, leading to UAF.
Fix this by explicitly deleting the NAPI association for the failed
index 'i' before rolling back the successfully configured queues.
Fixes: b02e5a0ebb17 ("xsk: Propagate napi_id to XDP socket Rx path") Reported-by: Guenter Roeck <groeck@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Björn Töpel <bjorn.topel@intel.com> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com> Link: https://patch.msgid.link/20260622111825.88337-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Meghana Malladi [Thu, 18 Jun 2026 10:03:48 +0000 (15:33 +0530)]
net: ti: icssg: Fix XSK zero copy TX during application wakeup
emac_xsk_xmit_zc() handles tx xmit for zero copy and gets called
inside napi context. User application wakes up the kernel while
initiating the transmit which triggers napi to start processing
the tx packets. The num_tx check inside emac_tx_complete_packets()
returns early if no packet transfer happen hindering the call
to emac_xsk_xmit_zc(). Remove this check to let application
wakeup initiate zero copy xmit traffic.
Add __netif_tx_lock() to ensure that the TX queue is protected
from concurrent access during the transmission of XDP frames.
This fixes netdev watchdog timeout for long runs.
net: dsa: sja1105: round up PTP perout pin duration
pin_duration is converted from the user-provided period to SJA1105
clock ticks and is later passed as the cycle_time argument to
future_base_time().
Very small period values may become zero after the conversion,
which can lead to a division by zero in future_base_time().
Round zero pin_duration up to 1 tick so that the smallest unsupported
periods use the minimum non-zero hardware duration instead of passing
zero to future_base_time().
Fixes: 747e5eb31d59 ("net: dsa: sja1105: configure the PTP_CLK pin as EXT_TS or PER_OUT") Signed-off-by: Aleksandrova Alyona <aga@itb.spb.ru> Link: https://patch.msgid.link/20260618110508.53094-1-aga@itb.spb.ru Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet [Mon, 22 Jun 2026 11:01:08 +0000 (11:01 +0000)]
net: do not acquire dev->tx_global_lock in netdev_watchdog_up()
Marek Szyprowski reported a deadlock during system resume when virtio_net
driver is used.
The deadlock occurs because netif_device_attach() is called while holding
dev->tx_global_lock (via netif_tx_lock_bh() in virtnet_restore_up()).
netif_device_attach() calls __netdev_watchdog_up(), which now also tries
to acquire dev->tx_global_lock to synchronize with dev_watchdog().
This recursive lock acquisition results in a deadlock.
Fix this by removing the tx_global_lock acquisition from netdev_watchdog_up().
The critical state (watchdog_timer and watchdog_ref_held) is already
protected by dev->watchdog_lock, which was introduced in the blamed commit.
Fixes: 8eed5519e496 ("net: watchdog: fix refcount tracking races") Reported-by: Marek Szyprowski <m.szyprowski@samsung.com> Closes: https://lore.kernel.org/netdev/a443376e-5187-4268-93b3-58047ef113a8@samsung.com/ Signed-off-by: Eric Dumazet <edumazet@google.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260622110108.69541-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Xiang Mei [Sat, 20 Jun 2026 20:15:31 +0000 (13:15 -0700)]
net, bpf: check master for NULL in xdp_master_redirect()
xdp_master_redirect() dereferences the result of
netdev_master_upper_dev_get_rcu() without a NULL check, but that helper
returns NULL when the receiving device has no upper-master adjacency.
The reach guard only checks netif_is_bond_slave(). On bond slave release
bond_upper_dev_unlink() drops the upper-master adjacency before clearing
IFF_SLAVE, so an XDP_TX reaching xdp_master_redirect() in that window
still passes netif_is_bond_slave() while master is already NULL, and
faults on master->flags at offset 0xb0:
The missing check dates back to the original code; commit 1921f91298d1
("net, bpf: fix null-ptr-deref in xdp_master_redirect() for down master")
later added the master->flags read where the fault now lands but kept the
unconditional deref. Check master for NULL before use; a NULL master is
treated the same as one that is not up.
Fixes: 879af96ffd72 ("net, core: Add support for XDP redirection to slave device") Reported-by: Weiming Shi <bestswngs@gmail.com> Signed-off-by: Xiang Mei <xmei5@asu.edu> Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev> Link: https://patch.msgid.link/20260620201531.180123-1-xmei5@asu.edu Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
selftests/xsk: stabilize timeout test behavior
This series improves AF_XDP selftests by making timeout handling
explicit and fixing sources of non-determinism in xsk timeout tests.
Patch 1 introduces test_spec::poll_tmout and removes implicit
dependence on RX UMEM setup state for timeout behavior.
Patch 2 fixes thread harness sequencing by attaching XDP programs
before worker startup, removing signal-based termination, and using
barrier synchronization only for dual-thread runs.
Patch 3 restores shared_umem after POLL_TXQ_FULL so test-local
configuration does not leak into subsequent cases on shared-netdev
runs.
Together these changes make timeout handling easier to follow and
improve selftest stability, especially on real NIC runs.
====================
Tushar Vyavahare [Tue, 16 Jun 2026 15:49:52 +0000 (21:19 +0530)]
selftests/xsk: make poll timeout mode explicit
Stop inferring timeout behavior from RX UMEM initialization state.
That ties timeout semantics to setup internals and obscures intent.
Use test_spec::poll_tmout as the explicit timeout-mode selector in
TX and RX paths.
In RX, treat poll timeout as expected only in timeout mode.
In TX, let send_pkts() own loop completion in non-timeout mode
and use __send_pkts() only for progress and timeout detection.
This makes timeout logic explicit and keeps control flow predictable.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com> Signed-off-by: Tushar Vyavahare <tushar.vyavahare@intel.com> Reviewed-by: Jason Xing <kerneljasonxing@gmail.com> Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Tested-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Link: https://patch.msgid.link/20260616154955.1492560-2-tushar.vyavahare@intel.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The net-next-hw spinners on netdev.bots.linux.dev observe failing
so-txtime-py tests. A review of stdout shows most failures to be
due to exceeding the 4ms grace period. All I saw were within 8ms.
So increase to that.
Double the bounds from 4 to 8ms. This is still is small enough to
differentiate the delays programmed by the test, 10 and 20ms.
Fixes: 5c6baef3885c ("selftests: drv-net: convert so_txtime to drv-net") Reported-by: Jakub Kicinski <kuba@kernel.org> Closes: https://lore.kernel.org/netdev/20260610170651.1b644001@kernel.org/ Signed-off-by: Willem de Bruijn <willemb@google.com> Link: https://patch.msgid.link/20260621200137.1564776-1-willemdebruijn.kernel@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In airoha_qdma_set_chan_tx_sched(), the loop clearing queue mask was
using AIROHA_NUM_TX_RING (32) instead of AIROHA_NUM_QOS_QUEUES (8).
Each channel has 8 queues, and TXQ_DISABLE_CHAN_QUEUE_MASK(channel, i)
computes BIT(i + (channel * 8)). With i ranging 0..31, this causes:
- channel 0: clears bit 0..31 (all 4 channels) instead of 0..7
- channel 1: clears bit 8..31 (channels 1-3) instead of 8..15
- channel 2: clears bit 16..31 (channels 2-3) instead of 16..23
- channel 3: clears bit 24..31 (channel 3 only) - correct by accident
While BIT(32+) on arm64 produces 64-bit values truncated to 0 in u32
mask parameter, the loop still incorrectly clears queues within the
same channel beyond queue 7.
Even though this is functionally harmless (the register resets to 0
and is only ever cleared, never set — so clearing extra bits is a
no-op), the loop bound is semantically wrong and should be fixed for
correctness and clarity.
Fix by using AIROHA_NUM_QOS_QUEUES (8) as the loop upper bound.
Abdun Nihaal [Sat, 20 Jun 2026 06:23:50 +0000 (11:53 +0530)]
bnx2x: fix potential memory leak in bnx2x_alloc_mem_bp()
If the allocation of fp[i].tpa_info fails, the error path will not free
the struct bnx2x_fastpath allocated earlier, as it is not linked to the
bp structure yet. Fix that by linking it immediately after allocation.
Cc: stable@vger.kernel.org Fixes: 15192a8cf8a8 ("bnx2x: Split the FP structure") Signed-off-by: Abdun Nihaal <nihaal@cse.iitm.ac.in> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260620062402.89549-1-nihaal@cse.iitm.ac.in Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ipv4: fib: Don't ignore error route in local/main tables.
When CONFIG_IP_MULTIPLE_TABLES is enabled but no rule is added,
fib_lookup() performs route lookup directly on two tables.
Since the first lookup does not properly bail out, the result
of an error route in the merged local/main table could be
overwritten by another route in the default table:
# unshare -n
# ip link set lo up
# ip route add 192.168.0.0/24 dev lo table 253
# ip route add unreachable 192.168.0.0/24
# ip route get 192.168.0.1
192.168.0.1 dev lo table default uid 0
cache <local>
Once a random rule is added, the error route is respected:
# ip rule add table 0
# ip rule del table 0
# ip route get 192.168.0.1
RTNETLINK answers: No route to host
Let's fix the inconsistent behaviour.
Fixes: f4530fa574df ("ipv4: Avoid overhead when no custom FIB rules are installed.") Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260619212753.3367244-1-kuniyu@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Fri, 19 Jun 2026 19:15:38 +0000 (12:15 -0700)]
eth: bnxt: improve the timing of stats
Kernel selftests wait 1.25x of the promised stats refresh time
(as read from ethtool -c). bnxt reports 1sec by default, but
the stats update process has two steps. First device DMAs the
new values, then the service task performs update in full-width
SW counters. So the worst case delay is actually 2x.
Note that the behavior is different for ring stats and port stats.
Port stats are fetched synchronously by the service worker, so
there's no risk of doubling up the delay there.
The problem of stale stats impacts not only tests but real workloads
which monitor egress bandwidth of a NIC. The inaccuracy causes double
counting in the next cycle and spurious overload alarms.
Try to read from the DMA buffer more aggressively, to mitigate
timing issues between DMA and service task. The SW update should
be cheap.
Fixes: 51f307856b60 ("bnxt_en: Allow statistics DMA to be configurable using ethtool -C.") Reviewed-by: Michael Chan <michael.chan@broadcom.com> Link: https://patch.msgid.link/20260619191538.104165-1-kuba@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Xiang Mei [Fri, 19 Jun 2026 04:53:34 +0000 (21:53 -0700)]
ipv6: Fix null-ptr-deref in fib6_nh_mtu_change().
fib6_nh_mtu_change() re-fetches idev via __in6_dev_get(arg->dev) and
dereferences idev->cnf.mtu6 without a NULL check. addrconf_ifdown()
clears dev->ip6_ptr with RCU_INIT_POINTER() after rt6_disable_ip() has
released tb6_lock, so the RA-driven MTU walk can observe a NULL idev and
oops. The caller rt6_mtu_change_route() guards its own __in6_dev_get(),
but this re-fetch is unguarded; nexthop-backed routes survive
addrconf_ifdown()'s flush, so the walk still reaches it after ip6_ptr is
nulled.
Return 0 when idev is NULL, matching rt6_mtu_change_route() and the
fib6_mtu() fix in commit 5ad509c1fdad ("ipv6: Fix null-ptr-deref in
fib6_mtu().").
Oops: general protection fault, ... KASAN: null-ptr-deref in range
[0x00000000000002a8-0x00000000000002af]
RIP: 0010:fib6_nh_mtu_change+0x203/0x990
rt6_mtu_change_route+0x141/0x1d0
__fib6_clean_all+0xd0/0x160
rt6_mtu_change+0xb4/0x100
ndisc_router_discovery+0x24b5/0x2cb0
icmpv6_rcv+0x12e9/0x1710
ipv6_rcv+0x39b/0x410
Fixes: c0b220cf7d80 ("ipv6: Refactor exception functions") Reported-by: Weiming Shi <bestswngs@gmail.com> Signed-off-by: Xiang Mei <xmei5@asu.edu> Reviewed-by: Fernando Fernandez Mancera <fmancera@suse.de> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20260619045334.2427073-1-xmei5@asu.edu Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fan Wu [Wed, 17 Jun 2026 02:05:18 +0000 (02:05 +0000)]
hdlc_ppp: sync per-proto timers before freeing hdlc state
Each PPP control protocol (LCP/IPCP/IPV6CP) embedded in struct ppp
registers a timer via timer_setup(). That struct ppp is the
hdlc->state allocation, which detach_hdlc_protocol() frees with kfree()
in both teardown paths: unregister_hdlc_device() and the re-attach inside
attach_hdlc_protocol().
The ppp proto never registered a .detach callback, so
detach_hdlc_protocol() performs no timer synchronization before the
kfree(). The only cancel, timer_delete(&proto->timer) in ppp_cp_event(),
is partial (it does not wait for a running callback) and only runs on the
->CLOSED transition; ppp_stop()/ppp_close() do not sync either. A
ppp_timer callback already executing (blocked on ppp->lock) survives the
kfree and then dereferences proto->state / ppp->lock in freed memory,
leading to a use-after-free.
Fix this by adding a .detach helper that calls timer_shutdown_sync() on
every per-proto timer. detach_hdlc_protocol() invokes proto->detach(dev)
before kfree(hdlc->state), so timer_shutdown_sync()
now runs on both free paths.
timer_shutdown_sync() is used instead of timer_delete_sync() because the
keepalive path re-arms the timer through add_timer()/mod_timer() and
shutdown blocks any re-activation during teardown.
Initialize the per-protocol timers in ppp_ioctl() when the protocol is
attached, and remove the now-redundant timer_setup() from ppp_start(), so
that the timers are initialized exactly once at attach time and
ppp_timer_release() never operates on uninitialized timer_list
structures. attach_hdlc_protocol() uses kmalloc() (not kzalloc), so
struct ppp's protos[i].timer is uninitialized garbage until the first
timer_setup(); without this init-at-attach, attaching the PPP protocol
without ever bringing the device up would leave timer_shutdown_sync()
operating on uninitialized memory in .detach. Moving the init out of
ppp_start() (which only runs on NETDEV_UP) into the attach path makes the
initialization unconditional and avoids initializing the same timer_list
twice.
icssg_ndo_get_stats64() unconditionally calls emac_get_stat_by_name()
with FW PA stat names regardless of whether the PA stats block is
present on the hardware. emac_get_stat_by_name() already guards the
PA stats lookup with `if (emac->prueth->pa_stats)`; when that pointer
is NULL the lookup falls through to netdev_err() and returns -EINVAL.
Because ndo_get_stats64 is polled regularly by the networking stack
this produces thousands of log entries of the form:
A secondary consequence is that the int(-EINVAL) return value is
implicitly widened to a near-ULLONG_MAX unsigned value when accumulated
into the __u64 fields of rtnl_link_stats64, silently corrupting the
rx_errors, rx_dropped and tx_dropped counters reported by `ip -s link`.
Every other PA-aware code path in the driver is already guarded with
the same `if (emac->prueth->pa_stats)` check. Apply the same guard
here.
Ziran Zhang [Tue, 16 Jun 2026 01:32:45 +0000 (09:32 +0800)]
rocker: Fix memory leak in ofdpa_port_fdb()
In ofdpa_port_fdb(), the hash_del() only unlinks the node from
hash table, but does not free it.
Fix this by adding kfree(found) after the !found == removing check,
where the pointer value is no longer needed.
Found by Coccinelle kfree script.
Cc: <stable+noautosel@kernel.org> # rocker is a test harness, it's never loaded on production systems Signed-off-by: Ziran Zhang <zhangcoder@yeah.net> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20260616013245.7098-1-zhangcoder@yeah.net Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Mon, 22 Jun 2026 17:33:38 +0000 (10:33 -0700)]
Merge tag 'nf-26-06-21' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net. This batches
fixes for real crashes with trivial/correctness fixes. There is too
a rework of the conntrack expectation timeout strategy to deal with
a possible race when removing an expectation.
1) Fix the incorrect flowtable timeout extension for entries in
hw offload, from Adrian Bente. This is correcting a defect in
the functionality, no crash.
2) Hold reference to device under the fake dst in br_netfilter,
from Haoze Xie. This is fixing a possible UaF if the device
is removed while packet is sitting in nfqueue.
3) Reject template conntrack in xt_cluster, otherwise access to
uninitialize conntrack fields are possible leading to WARN_ON
due to unset layer 3 protocol. From Wyatt Feng.
4) Make sure the IPv6 tunnel header is in the linear skb data
area before pulling. While at it remove incomplete NEXTHDR_DEST
support. From Lorenzo Bianconi. This possibly leading to crash
if IPv4 header is not in the linear area.
5) Use test_bit_acquire in ipset hash set to avoid reordering
of subsequent memory access. This is addressing a LLM related
report, no crash has been observed. From Jozsef Kadlecsik.
6) Use test_bit_acquire in ipset bitmap set too, for the same
reason as in the previous patch, from Jozsef Kadlecsik.
7) Call kfree_rcu() after rcu_assign_pointer() to address a
possible UaF if kfree_rcu() runs inmediately, which to my
understanding never happens. Never observed in practise,
reported by LLM. Also from Jozsef Kadlecsik.
8) Use disable_delayed_work_sync() instead cancel_delayed_work_sync()
to avoid that ipset GC handler re-queues work as reported by LLM.
From Jozsef Kadlecsik. This is for correctness.
9) Restore the check in nft_payload for exceeding payloda offset
over 2^16. From Florian Westphal. This fixes a silent truncation,
not a big deal, but better be assertive and reject it.
10) Validate NFT_META_BRI_IIFHWADDR can only run from bridge
prerouting. From Florian Westphal. Harmless but it could allow
to read bytes from skb->cb.
11) Zero out destination hardware address during the flowtable
path setup, also from Florian. This is a correctness fix, LLM
points that possible infoleak can happen but topology to achieve
it is not clear.
12) Skip IPv4 options if present when building the IPV4 reject reply.
Otherwise bytes in the IPv4 options header can be sent back to
origin where the ICMP header is being expected. Again from
Florian Westphal.
13) Replace timer API for expectation by GC worker approach. This
is implicitly fixing a race between nf_ct_remove_expectations()
which might fail to remove the expectation due to timer_del()
returning false because timer has expired and callback is
being run concurrently. This fix is addressing a crash that has
been already reported with a reproducer.
14) Check if br_vlan_get_pvid_rcu() fails, otherwise possible stack
infoleak of 4-bytes. From Florian Westphal.
* tag 'nf-26-06-21' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: nft_meta_bridge: fix NFT_META_BRI_IIFPVID stack leak
netfilter: nf_conntrack_expect: use conntrack GC to reap expectations
netfilter: nf_reject: skip iphdr options when looking for icmp header
netfilter: nft_flow_offload: zero device address for non-ether case
netfilter: nft_meta_bridge: add validate callback for get operations
netfilter: nft_payload: reject offsets exceeding 65535 bytes
netfilter: ipset: make sure gc is properly stopped
netfilter: ipset: fix order of kfree_rcu() and rcu_assign_pointer()
netfilter: ipset: Don't use test_bit() in lockless RCU readers in bitmap types
netfilter: ipset: Don't use test_bit() in lockless RCU readers in hash types
netfilter: flowtable: fix and simplify IP6IP6 tunnel handling
netfilter: xt_cluster: reject template conntracks in hash match
netfilter: nf_queue: pin bridge device while NFQUEUE holds fake dst
netfilter: flowtable: fix offloaded ct timeout never being extended
====================
Ioana Ciornei [Thu, 18 Jun 2026 09:28:12 +0000 (12:28 +0300)]
dpaa2-switch: do not accept VLAN uppers while bridged
The dpaa2-switch driver does not support VLAN uppers while its ports are
bridged. This scenario tried to be prevented by rejecting a bridge join
while VLAN uppers exist but the reverse order was still possible.
This patches adds a check so that the dpaa2-switch also does not accept
VLAN uppers while bridged.
Jiayuan Chen [Thu, 18 Jun 2026 10:43:35 +0000 (18:43 +0800)]
ipv6: ioam: fix type confusion of dst_entry
IOAM uses a dummy dst_entry(null_dst) to mark that the destination should
not be changed after the transformation. This dst is stored in the IOAM lwt
state and may be passed to dst_cache_set_ip6().
However, the IPv6 dst cache path eventually calls rt6_get_cookie(), which
treats the dst_entry as part of a struct rt6_info. Since the null_dst was
embedded directly as a struct dst_entry in struct ioam6_lwt, this resulted
in an invalid cast and rt6_get_cookie() reading fields from the wrong
object.
In practice, the wrong cookie is not used while dst->obsolete is zero, but
rt6_get_cookie() may also access per-cpu value when rt->sernum is
zero. In this case, rt->sernum aliases ioam6_lwt::cache::reset_ts, which
can become zero, making this a potential invalid pointer access.
Fix this by embedding a full struct rt6_info for the dummy IPv6 route and
passing its dst member to the dst APIs.
datalen already includes fraggap (datalen = length + fraggap). When
fraggap is non-zero, this is not the first skb and transhdrlen is zero.
The fraggap bytes carried over from the previous skb are copied just past
the fragment headers in the new skb's linear area. The linear area is
therefore undersized by fraggap bytes while pagedlen is overstated by the
same amount, and the copy writes past skb->end into the trailing
skb_shared_info.
An unprivileged user can trigger this via a UDPv6 socket using
MSG_MORE together with MSG_SPLICE_PAGES.
The bad accounting was introduced by commit 773ba4fe9104 ("ipv6:
avoid partial copy for zc"). Before commit ce650a166335 ("udp6: Fix
__ip6_append_data()'s handling of MSG_SPLICE_PAGES"), the negative
copy value caused -EINVAL to be returned. That later commit allowed
MSG_SPLICE_PAGES to proceed in this case, making the corruption
triggerable.
The non-paged branch sets alloclen to fraglen, which already accounts
for fraggap because datalen does. Bring the paged branch in line by
adding fraggap to alloclen and subtracting it from pagedlen.
After this adjustment, copy no longer collapses to -fraggap on the
paged path, so remove the stale comment describing that old arithmetic.
Since a negative copy is no longer expected for a valid MSG_SPLICE_PAGES
case, remove the MSG_SPLICE_PAGES exception from the negative copy check.
Fixes: 773ba4fe9104 ("ipv6: avoid partial copy for zc") Signed-off-by: Jungwoo Lee <jwlee2217@gmail.com> Signed-off-by: Wongi Lee <qw3rtyp0@gmail.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/ajFTqRljatR17fFy@DESKTOP-19IMU7U.localdomain Signed-off-by: Jakub Kicinski <kuba@kernel.org>
datalen already includes fraggap, but the fraggap bytes carried over
from the previous skb are copied into the new skb's linear area at
offset transhdrlen by the subsequent skb_copy_and_csum_bits(). The
linear area is therefore undersized by fraggap bytes while pagedlen is
overstated by the same amount.
The non-paged branch sets alloclen to fraglen, which already accounts
for fraggap because datalen does. Bring the paged branch in line by
adding fraggap to alloclen and subtracting it from pagedlen.
After this adjustment, copy no longer collapses to -fraggap on the
paged path, so remove the stale comment describing that old arithmetic.
Fixes: 8eb77cc73977 ("ipv4: avoid partial copy for zc") Signed-off-by: Jungwoo Lee <jwlee2217@gmail.com> Signed-off-by: Wongi Lee <qw3rtyp0@gmail.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/ajFR1eLAIs42TN3g@DESKTOP-19IMU7U.localdomain Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Xingquan Liu [Fri, 19 Jun 2026 15:13:48 +0000 (11:13 -0400)]
selftests/tc-testing: Add DualPI2 GSO backlog accounting test
Add a regression test for DualPI2 GSO backlog accounting when it is
used as a child qdisc of QFQ.
The test sends one UDP GSO datagram through a QFQ class with DualPI2 as
the leaf qdisc. DualPI2 splits the skb into two segments. After the
traffic drains, both QFQ and DualPI2 must report zero backlog and zero
qlen.
On kernels with the broken accounting, QFQ can keep a stale non-zero
qlen after all real packets have been dequeued.
Signed-off-by: Xingquan Liu <b1n@b1n.io> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Reviewed-by: Victor Nogueira <victor@mojatatu.com> Link: https://patch.msgid.link/20260619151447.223640-2-b1n@b1n.io Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Xingquan Liu [Fri, 19 Jun 2026 15:13:47 +0000 (11:13 -0400)]
net/sched: dualpi2: fix GSO backlog accounting
When DualPI2 splits a GSO skb into N segments, it propagates N
additional packets to its parent before returning NET_XMIT_SUCCESS.
The parent then accounts for the original skb once more, leaving its
qlen one larger than the number of packets actually queued.
With QFQ as the parent, after all real packets are dequeued, QFQ still
has a non-zero qlen while its in-service aggregate has no active
classes. qfq_choose_next_agg() returns NULL and qfq_dequeue() passes
the result to qfq_peek_skb(), causing a NULL pointer dereference.
Follow the same pattern used by tbf_segment() and taprio: count only
successfully queued segments, propagate the difference between the
original skb and those segments, and return NET_XMIT_SUCCESS whenever
at least one segment was queued.
Fixes: 8f9516daedd6 ("sched: Add enqueue/dequeue of dualpi2 qdisc") Cc: stable@vger.kernel.org Signed-off-by: Xingquan Liu <b1n@b1n.io> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Reviewed-by: Victor Nogueira <victor@mojatatu.com> Link: https://patch.msgid.link/20260619151447.223640-1-b1n@b1n.io Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Weiming Shi [Wed, 17 Jun 2026 06:55:13 +0000 (14:55 +0800)]
ipv6: ndisc: fix NULL deref in accept_untracked_na()
accept_untracked_na() re-fetches the inet6_dev with __in6_dev_get(dev)
and dereferences idev->cnf.accept_untracked_na without a NULL check,
even though its only caller ndisc_recv_na() already fetched and
NULL-checked idev for the same device.
Both reads of dev->ip6_ptr run in the same RCU read-side critical
section, but a concurrent addrconf_ifdown() can clear dev->ip6_ptr
between them: lowering the MTU below IPV6_MIN_MTU calls addrconf_ifdown()
without the synchronize_net() that orders the unregister path, so the
re-fetch returns NULL and oopses:
It is reachable by an unprivileged user via a network namespace.
Pass the caller's already validated idev instead of re-fetching it; the
idev stays alive for the whole RCU critical section, so it is safe even
after dev->ip6_ptr has been cleared.
Fixes: aaa5f515b16b ("net: ipv6: new accept_untracked_na option to accept na only if in-network") Reported-by: Xiang Mei <xmei5@asu.edu> Signed-off-by: Weiming Shi <bestswngs@gmail.com> Reviewed-by: Jiayuan Chen <jiayuan.chen@linux.dev> Link: https://patch.msgid.link/20260617065512.2529757-2-bestswngs@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Maoyi Xie [Thu, 18 Jun 2026 07:08:17 +0000 (15:08 +0800)]
net: sit: require CAP_NET_ADMIN in the device netns for changelink
ipip6_changelink() operates on at most two netns, dev_net(dev) and the
tunnel link netns t->net. They differ once the device is created in or
moved to a netns other than the one the request runs in. The rtnl
changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a
caller privileged there but not in t->net can rewrite a tunnel that
lives in t->net.
Gate ipip6_changelink() on rtnl_dev_link_net_capable() at its top,
before any attribute is parsed. sit was the one tunnel type not covered
by the recent series that added this check to the other changelink()
handlers.
implicitly converts to unsigned arithmetic: when skb->priority is 0
(the default for unclassified traffic), (0u - 1u) wraps to UINT_MAX,
and UINT_MAX % 8 = 7, routing default best-effort packets to the
highest-priority QoS queue. This causes QoS inversion where the
majority of traffic on a PON gateway starves actual high-priority
flows (VoIP, gaming, etc.).
The "- 1" offset was a leftover from the ETS offload implementation
that has since been removed. The correct mapping is a direct modulo:
queue = skb->priority % AIROHA_NUM_QOS_QUEUES;
This maps priority 0 → queue 0 (lowest), priority 7 → queue 7
(highest), with higher priorities wrapping around. This is the
standard Linux sk_prio → HW queue mapping used by other drivers.
net/sched: act_ct: fix nf_connlabels leak on two error paths
tcf_ct_fill_params() calls nf_connlabels_get() (setting put_labels) when
TCA_CT_LABELS is present, but two later error sites use a bare return
instead of "goto err", skipping the err: nf_connlabels_put() cleanup.
They also precede the "p->put_labels = put_labels" assignment, so the
tcf_ct_params_free() fallback does not release the count either. Each
failed RTM_NEWACTION on these paths leaks one nf_connlabels reference:
net->ct.labels_used is incremented and never released. The action is
reachable with CAP_NET_ADMIN over the netns, i.e. from an unprivileged
user namespace on default-userns kernels.
Impact: an unprivileged user with CAP_NET_ADMIN over a network namespace
(e.g. via user namespaces) leaks one nf_connlabels reference per failed
RTM_NEWACTION on the two error paths; net->ct.labels_used is never
released.
The err: label is safe to reach from both sites: p->tmpl is still NULL
there (kzalloc'd, not yet assigned) and nf_ct_put(NULL) is a no-op, so
no inline release is needed.
Fixes: 70f06c115bcc ("sched: act_ct: switch to per-action label counting") Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Link: https://patch.msgid.link/20260617215708.1115818-1-michael.bommarito@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Rosen Penev [Thu, 18 Jun 2026 02:34:05 +0000 (19:34 -0700)]
net: emac: Fix NULL pointer dereference in emac_probe
Move devm_request_irq() after devm_platform_ioremap_resource() so that
dev->emacp is mapped before the interrupt handler can fire. An early
interrupt hitting emac_irq() would dereference the NULL dev->emacp and
crash.
Also remove redundant error message. devm_platform_ioremap_resource()
already returns an error message with dev_err_probe().
Fixes: dcc34ef7c834 ("net: ibm: emac: manage emac_irq with devm") Signed-off-by: Rosen Penev <rosenp@gmail.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Link: https://patch.msgid.link/20260618023405.415644-1-rosenp@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Sun, 21 Jun 2026 21:53:06 +0000 (14:53 -0700)]
Merge tag 'batadv-net-pullrequest-20260619' of https://git.open-mesh.org/batadv
Simon Wunderlich says:
====================
Here are some batman-adv bugfixes, all by Sven Eckelmann:
- gw: don't deselect gateway with active hardif
- TTL fixes in bcast/mcast and fragmentation handling (4 patches)
- BATMAN V: prevent OGM aggregation on disabled hardif
- tp_meter: concurrency and range fixes, add cap to unacked list
(4 patches)
- tt: fix unchecked VLAN ID in tt for local add and roam (2 patches)
- dat: check VLAN ID in the hash
- tvlv: enforce 2-byte alignment
- tvlv: avoid race of cifsnotfound handler state
* tag 'batadv-net-pullrequest-20260619' of https://git.open-mesh.org/batadv:
batman-adv: tvlv: avoid race of cifsnotfound handler state
batman-adv: tvlv: enforce 2-byte alignment
batman-adv: dat: prevent false sharing between VLANs
batman-adv: tt: track roam count per VID
batman-adv: tt: don't merge change entries with different VIDs
batman-adv: tp_meter: handle overlapping packets
batman-adv: tp_meter: prevent parallel modifications of last_recv
batman-adv: tp_meter: annotate last_recv_time access with READ/WRITE_ONCE
batman-adv: tp_meter: restrict number of unacked list entries
batman-adv: v: prevent OGM aggregation on disabled hardif
batman-adv: frag: avoid underflow of TTL
batman-adv: frag: ensure fragment is writable before modifying TTL
batman-adv: fix (m|b)cast csum after decrementing TTL
batman-adv: ensure bcast is writable before modifying TTL
batman-adv: gw: don't deselect gateway with active hardif
====================
Geetha sowjanya [Tue, 16 Jun 2026 19:00:20 +0000 (00:30 +0530)]
octeontx2-pf: mcs: Fix mcs resources free on PF shutdown
On PF shutdown, the current driver free mcs hardware
resources though mcs resources are not allocated to it.
This patch checks the mcs resources status and if resources
are allocated then only sends mailbox message to free them.
Ratheesh Kannoth [Wed, 17 Jun 2026 10:21:49 +0000 (15:51 +0530)]
octeontx2-af: npc: cn20k: fix NPC defrag
npc_defrag_alloc_free_slots() always passed NPC_MCAM_KEY_X2 into
__npc_subbank_alloc(), which must match sb->key_type, so defrag never
allocated replacement slots on X4 banks. Pass the subbank key type for
bank 0, and only extend the search into bank 1 for X2 (X4 MCAM indices
are confined to b0b..b0t).
Fixes: 645c6e3c1999 ("octeontx2-af: npc: cn20k: virtual index support") Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260617102149.1309913-1-rkannoth@marvell.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Wayen Yan [Wed, 17 Jun 2026 05:48:13 +0000 (13:48 +0800)]
net: ethernet: mtk_ppe: Fix rhashtable leak in mtk_ppe_init error paths
In mtk_ppe_init(), when accounting is enabled, the error paths for
dmam_alloc_coherent(mib) and devm_kzalloc(acct) failures return NULL
directly, bypassing the err_free_l2_flows label that destroys the
rhashtable initialized earlier.
While this leak only occurs during probe (not runtime) and the leaked
memory is minimal (an empty rhash table), fixing it ensures proper
error path cleanup consistency.
Fix by changing the two return NULL statements to goto err_free_l2_flows.
Fixes: 603ea5e7ffa7 ("net: ethernet: mtk_eth_soc: fix memory leak in error path") Signed-off-by: Wayen Yan <win847@gmail.com> Acked-by: Lorenzo Bianconi <lorenzo@kernel.org> Link: https://patch.msgid.link/178167550101.2217645.14579307712717502425@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ilya Maximets [Tue, 16 Jun 2026 10:03:29 +0000 (12:03 +0200)]
net: dst_metadata: fix false-positive memcpy overflow in tun_dst_unclone
kmalloc_flex() in metadata_dst_alloc() sets __counted_by for the
structure to the options_len, which is then initialized to zero.
Later, we're initializing the structure by copying the tunnel info
together with the options, and this triggers a warning for a potential
memcpy overflow, since the compiler estimates that the options can't
fit into the structure, even though the memory for them is actually
allocated.
memcpy: detected buffer overflow: 104 byte write of buffer size 96
WARNING: CPU: X PID: Y at lib/string_helpers.c:1036 __fortify_report
skb_tunnel_info_unclone+0x179/0x190
geneve_xmit+0x7fe/0xe00
The issue is triggered when built with clang and source fortification.
Fix that by doing the copy in two stages: first - the main data with
the options_len, then the options. This way the correct length should
be known at the time of the copy.
It would be better if the options_len never changed after allocation,
but the allocation code is a little separate from the initialization
and it would be awkward and potentially dangerous to return a struct
with options_len set to a non-zero value from the metadata_dst_alloc().
Another option would be to use ip_tunnel_info_opts_set(), but it is
doing too many unnecessary operations for the use case here.
Fixes: 69050f8d6d07 ("treewide: Replace kmalloc with kmalloc_obj for non-scalar types") Reported-by: Johan Thomsen <write@ownrisk.dk> Closes: https://lore.kernel.org/netdev/CAKv6aAM8_EWgXScnKmKYm_4SwGDVBK++dzfP+Y6msUXbp99QUw@mail.gmail.com/ Signed-off-by: Ilya Maximets <i.maximets@ovn.org> Link: https://patch.msgid.link/20260616100332.1308294-1-i.maximets@ovn.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Weiming Shi [Wed, 17 Jun 2026 13:57:45 +0000 (21:57 +0800)]
tipc: fix use-after-free of the discoverer in tipc_disc_rcv()
bearer_disable() frees b->disc with tipc_disc_delete()'s plain kfree(),
but tipc_disc_rcv() still dereferences b->disc in RX softirq under
rcu_read_lock() (tipc_udp_recv -> tipc_rcv -> tipc_disc_rcv).
L2 bearers are safe thanks to the synchronize_net() in
tipc_disable_l2_media(), but the UDP bearer defers that call to the
cleanup_bearer() workqueue, so the discoverer is freed with no grace
period:
BUG: KASAN: slab-use-after-free in tipc_disc_rcv (net/tipc/discover.c:149)
Read of size 8 at addr ffff88802348b728 by task poc_tipc/184
<IRQ>
tipc_disc_rcv (net/tipc/discover.c:149)
tipc_rcv (net/tipc/node.c:2126)
tipc_udp_recv (net/tipc/udp_media.c:391)
udp_rcv (net/ipv4/udp.c:2643)
ip_local_deliver_finish (net/ipv4/ip_input.c:241)
</IRQ>
Freed by task 181:
kfree (mm/slub.c:6565)
bearer_disable (net/tipc/bearer.c:418)
tipc_nl_bearer_disable (net/tipc/bearer.c:1001)
The bearer is freed with kfree_rcu(); free the discoverer the same way.
Add an rcu_head to struct tipc_discoverer and free it and its skb from an
RCU callback.
Because the RCU callback (tipc_disc_free_rcu) lives in module text, a
call_rcu() that is still pending when the tipc module is unloaded would
invoke a freed function. Add an rcu_barrier() to tipc_exit() after the
bearer subsystem has been torn down, so all pending discoverer callbacks
have run before the module text goes away.
Reachable from an unprivileged user namespace: the TIPCv2 genl family is
netnsok and its bearer commands have no GENL_ADMIN_PERM. Needs CONFIG_TIPC
and CONFIG_TIPC_MEDIA_UDP.
Fixes: 25b0b9c4e835 ("tipc: handle collisions of 32-bit node address hash values") Reported-by: Xiang Mei <xmei5@asu.edu> Signed-off-by: Weiming Shi <bestswngs@gmail.com> Reviewed-by: Tung Nguyen <tung.quang.nguyen@est.tech> Link: https://patch.msgid.link/20260617135744.3383175-3-bestswngs@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ruoyu Wang [Wed, 17 Jun 2026 19:32:28 +0000 (03:32 +0800)]
net: marvell: prestera: initialize err in prestera_port_sfp_bind
prestera_port_sfp_bind() returns err after walking the ports node. If no
child node matches the port's front-panel id, err is never assigned.
Initialize err to 0 because absence of a matching optional port device
tree node is not an error. In that case no phylink is created and port
creation should continue with port->phy_link left NULL. Errors from
malformed matched nodes and phylink_create() still propagate.
Fixes: 52323ef75414 ("net: marvell: prestera: add phylink support") Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Acked-by: Elad Nachman <enachman@marvell.com> Link: https://patch.msgid.link/20260617193228.1653582-1-ruoyuw560@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
netfilter: nf_conntrack_expect: use conntrack GC to reap expectations
This patch replaces the timer API by GC worker approach for
expectations, as it already happened in many other subsystems.
Use the existing conntrack GC worker to iterate over the local list of
expectations in the master conntrack to reap expired expectations.
Check IPS_HELPER_BIT to run GC for expectations, set it on for nft_ct
expectation which nevers sets it. Hold the expectation spinlock while
iterating over the master conntrack expectation list to synchronize with
nf_ct_remove_expectations(). This also performs runtime packet path
garbage collection through the expectation insertion and lookup
functions while walking over one of the chains of the global expectation
hashtables. Unconfirmed conntrack entries are skipped since ct->ext can
be reallocated and dying are skipped since those will be gone soon.
Set on IPS_HELPER_BIT if the helper ct extension is added, then the new
GC worker does not need to bump the ct refcount to check if the ct->ext
helper is available.
This removes the extra bump on the refcount for expectation timers, this
allows to remove several nf_ct_expect_put() calls after the unlink,
after this update only refcount remains at 1 while on the expectation
hashes.
This patch implicitly addresses a race with the existing timer API
allowing an expectation to access a stale exp->master pointer which has
been already released when expectation removal loses races with an
expiring timer, ie. timer_del() reporting false.
Add a new NF_CT_EXPECT_DEAD flag to reap this expectation via GC. This
is needed by nf_conntrack_unexpect_related() which is called in error
paths to invalidate newly created expectations that has been added into
the hashes. These expectactions cannot be inmediately released as GC or
nf_ct_remove_expectations() could race to make it. On expectation
insert, the runtime GC reaps stale expectations before checking the
expectation limit set by policy.
Set current timestamp in nf_ct_expect_alloc(), then add the expectation
policy timeout (or custom timeout specified added on top of this) to
specify the expectation lifetime.
Fixes: bffcaad9afdf ("netfilter: ctnetlink: ensure safe access to master conntrack") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Thu, 18 Jun 2026 08:49:24 +0000 (10:49 +0200)]
netfilter: nf_reject: skip iphdr options when looking for icmp header
Not a big deal but this hould have used the real ip header length and not the
base header size. As-is, if there are options then
nf_skb_is_icmp_unreach() result will be random.
Fixes: db99b2f2b3e2 ("netfilter: nf_reject: don't reply to icmp error messages") Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Thu, 18 Jun 2026 06:25:47 +0000 (08:25 +0200)]
netfilter: nft_flow_offload: zero device address for non-ether case
LLM points out that the skip causes unitialised stack array to
propagate down into dev_fill_forward_path(). Its not clear to me that
there is a guarantee that a later ctx.dev->netdev_ops->ndo_fill_forward_path()
would always fix this up.
Cc: Felix Fietkau <nbd@nbd.name> Fixes: 45ca3e61999e ("netfilter: nft_flow_offload: skip dst neigh lookup for ppp devices") Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Thu, 18 Jun 2026 06:16:18 +0000 (08:16 +0200)]
netfilter: nft_meta_bridge: add validate callback for get operations
Blamed commit added NFT_META_BRI_IIFHWADDR to the set validate callback,
yet this is a get operation.
Add a get validate callback and move the NFT_META_BRI_IIFHWADDR key
there.
AFAICS this is harmless, NFT_META_BRI_IIFHWADDR can deal with a NULL
input device and the set handler ignores a NFT_META_BRI_IIFHWADDR
operation, but it allows to read 4 bytes off bridge skb->cb[].
Large offsets were rejected based on netlink policy, but blamed commit
removed the policy without updating nft_payload_inner_init() to use the
truncation-check helper.
Silent truncation is not a problem, but not wanted either, so add a
check.
Fixes: 077dc4a27579 ("netfilter: nft_payload: extend offset to 65535 bytes") Signed-off-by: Florian Westphal <fw@strlen.de> Reviewed-by: Fernando Fernandez Mancera <fmancera@suse.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Jozsef Kadlecsik [Wed, 17 Jun 2026 08:41:26 +0000 (10:41 +0200)]
netfilter: ipset: make sure gc is properly stopped
Sashiko noticed that when destroying a set,
cancel_delayed_work_sync() was called while gc
calls queue_delayed_work() unconditionally which
can lead not to properly shutting down the gc.
Jozsef Kadlecsik [Wed, 17 Jun 2026 08:41:24 +0000 (10:41 +0200)]
netfilter: ipset: fix order of kfree_rcu() and rcu_assign_pointer()
Sashiko pointed out that kfree_rcu() was called before
rcu_assign_pointer() in handling the comment extension.
Fix the order so that rcu_assign_pointer() called first.
Fixes: b57b2d1fa53f ("netfilter: ipset: Prepare the ipset core to use RCU at set level") Signed-off-by: Jozsef Kadlecsik <kadlec@netfilter.org> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Jozsef Kadlecsik [Wed, 17 Jun 2026 08:41:23 +0000 (10:41 +0200)]
netfilter: ipset: Don't use test_bit() in lockless RCU readers in bitmap types
The pair of the patch "netfilter: ipset: Don't use test_bit() in lockless
RCU readers in hash types" for the bitmap types.
Fixes: 02a3231b6d82 ("netfilter: nf_conntrack_expect: store netns and zone in expectation") Fixes: b0da3905bb1e ("netfilter: ipset: Bitmap types using the unified code base") Signed-off-by: Jozsef Kadlecsik <kadlec@netfilter.org> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Jozsef Kadlecsik [Wed, 17 Jun 2026 08:41:22 +0000 (10:41 +0200)]
netfilter: ipset: Don't use test_bit() in lockless RCU readers in hash types
Sashiko pointed out that there are a few lockless RCU readers
using test_bit() which is a relaxed atomic operation and
provides no memory barrier guarantees. Use test_bit_acquire()
instead where the operation may run parallel with add/del/gc,
i.e. is not one from the next cases
- protected by region lock
- in a set destroy phase
- in a new/temporary set creation phase
Fixes: 18f84d41d34f ("netfilter: ipset: Introduce RCU locking in hash:* types") Signed-off-by: Jozsef Kadlecsik <kadlec@netfilter.org> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
ieee802154: allow legacy LLSEC ADD/DEL ops to pass strict validation
The LLSEC ADD/DEL doit handlers under the legacy IEEE802154_NL family
consume IEEE802154_ATTR_LLSEC_KEY_BYTES and
IEEE802154_ATTR_LLSEC_KEY_USAGE_COMMANDS, both declared in
net/ieee802154/nl_policy.c as bare length entries with no .type
(defaulting to NLA_UNSPEC). Generic netlink strict validation rejects
all NLA_UNSPEC attributes via validate_nla(), so every LLSEC_ADD_KEY,
LLSEC_DEL_KEY, LLSEC_ADD_DEV, LLSEC_DEL_DEV, LLSEC_ADD_DEVKEY,
LLSEC_DEL_DEVKEY, LLSEC_ADD_SECLEVEL, and LLSEC_DEL_SECLEVEL request
fails at the dispatcher with "Unsupported attribute" before reaching
the handler.
The doit path has been silently dead since strict validation became
the default for genl families that do not opt out. The dump path is
unaffected because dump requests carry no LLSEC attributes to
validate, which is why the LLSEC_LIST_KEY read remained reachable
(patch 1/2). Introduce IEEE802154_OP_RELAXED() mirroring
IEEE802154_OP() but with .validate = GENL_DONT_VALIDATE_STRICT, and
use it for the eight legacy LLSEC mutate ops so admin-driven LLSEC
configuration via the legacy interface works again.
Fixes: 3e9c156e2c21 ("ieee802154: add netlink interfaces for llsec") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com> Link: https://lore.kernel.org/20260520141640.1149513-3-michael.bommarito@gmail.com Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
In net/ieee802154/netlink.c, the legacy IEEE802154_NL family ops table
builds the LLSEC dump entries (LLSEC_LIST_KEY, LLSEC_LIST_DEV,
LLSEC_LIST_DEVKEY, LLSEC_LIST_SECLEVEL) with IEEE802154_DUMP() which
sets no .flags, so generic netlink runs them ungated. The modern
nl802154 family admin-gates the equivalent reads via
NL802154_CMD_GET_SEC_KEY and friends with .flags = GENL_ADMIN_PERM.
Any local uid that can open AF_NETLINK / NETLINK_GENERIC can resolve
the "802.15.4 MAC" family and dump LLSEC_LIST_KEY on any wpan netdev
that has an LLSEC key installed; the dump handler writes the raw
16-byte AES-128 key bytes (IEEE802154_ATTR_LLSEC_KEY_BYTES, copied
verbatim from struct ieee802154_llsec_key.key) into the reply.
Recovering the AES key compromises 802.15.4 LLSEC link confidentiality
and authenticity, since LLSEC uses CCM* and the same key authenticates
and encrypts frames.
Impact: any local uid with no capabilities can read the raw 16-byte
AES-128 LLSEC key from the kernel keytable on any wpan netdev that has
an administrator-installed LLSEC key, by issuing an LLSEC_LIST_KEY
dump on the legacy IEEE802154_NL generic-netlink family.
Introduce IEEE802154_DUMP_PRIV() mirroring IEEE802154_DUMP() but
setting .flags = GENL_ADMIN_PERM, and use it for the four LLSEC dump
entries. LIST_PHY and LIST_IFACE retain IEEE802154_DUMP() because the
modern nl802154 family exposes their equivalents to unprivileged
readers by design (NL802154_CMD_GET_WPAN_PHY and
NL802154_CMD_GET_INTERFACE carry "can be retrieved by unprivileged
users" annotations).
Fixes: 3e9c156e2c21 ("ieee802154: add netlink interfaces for llsec") Cc: stable@vger.kernel.org Assisted-by: Claude:claude-opus-4-7 Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com> Link: https://lore.kernel.org/20260520141640.1149513-2-michael.bommarito@gmail.com Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
mac802154: Prevent overwrite return code in mac802154_perform_association()
When assoc_status not equal to IEEE802154_ASSOCIATION_SUCCESSFUL, the
return value assigned to either "-ERANGE" or "-EPERM" but this return
value will be overwritten to 0 after exiting the conditional scope.
So, jump to clear_assoc label to preserve the return value when
assoc_status not equal to IEEE802154_ASSOCIATION_SUCCESSFUL.
This is reported by Coverity Scan as "Unused value".
Fixes: fefd19807fe9 ("mac802154: Handle associating") Signed-off-by: Robertus Diawan Chris <robertusdchris@gmail.com> Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com> Link: https://lore.kernel.org/20260602054133.470293-1-robertusdchris@gmail.com Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
Aleksandr Nogikh [Wed, 27 May 2026 20:18:18 +0000 (20:18 +0000)]
ieee802154: fix kernel-infoleak in dgram_recvmsg()
KMSAN reported a kernel-infoleak in move_addr_to_user():
BUG: KMSAN: kernel-infoleak in instrument_copy_to_user
include/linux/instrumented.h:131 [inline]
BUG: KMSAN: kernel-infoleak in _inline_copy_to_user
include/linux/uaccess.h:205 [inline]
BUG: KMSAN: kernel-infoleak in _copy_to_user+0xcc/0x120
lib/usercopy.c:26
instrument_copy_to_user include/linux/instrumented.h:131 [inline]
_inline_copy_to_user include/linux/uaccess.h:205 [inline]
_copy_to_user+0xcc/0x120 lib/usercopy.c:26
copy_to_user include/linux/uaccess.h:236 [inline]
move_addr_to_user+0x2e7/0x440 net/socket.c:302
____sys_recvmsg+0x232/0x610 net/socket.c:2925
...
Uninit was stored to memory at:
ieee802154_addr_to_sa include/net/ieee802154_netdev.h:369 [inline]
dgram_recvmsg+0xa09/0xbe0 net/ieee802154/socket.c:739
The issue occurs because the `pan_id` field of `struct ieee802154_addr`
is left uninitialized when the address mode is `IEEE802154_ADDR_NONE`.
The execution flow is as follows:
1. `__ieee802154_rx_handle_packet()` declares a local `struct
ieee802154_hdr hdr` on the stack.
2. `ieee802154_hdr_pull()` calls `ieee802154_hdr_get_addr()` to parse
the source and destination addresses into this structure.
3. If the address mode is `IEEE802154_ADDR_NONE`,
`ieee802154_hdr_get_addr()` previously only set the `mode` field,
leaving the `pan_id` field containing uninitialized stack memory.
4. This uninitialized `pan_id` is later copied into a `struct
sockaddr_ieee802154` in `dgram_recvmsg()` via `ieee802154_addr_to_sa()`.
5. Finally, `move_addr_to_user()` copies the socket address structure to
user space, leaking the uninitialized bytes.
Fix this by using `memset` to zero out the address structure in
`ieee802154_hdr_get_addr()` when the mode is `IEEE802154_ADDR_NONE`.
Fixes: 94b4f6c21cf5 ("ieee802154: add header structs with endiannes and operations") Assisted-by: Gemini:gemini-3.1-pro-preview Gemini:gemini-3-flash-preview syzbot Reported-by: syzbot+346474e3bf0b26bd3090@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=346474e3bf0b26bd3090 Link: https://syzkaller.appspot.com/ai_job?id=a507a109-d683-4a2c-bc03-93394f491b17 Signed-off-by: Aleksandr Nogikh <nogikh@google.com> Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com> Link: https://lore.kernel.org/62795fd9-fc0c-48eb-bb82-05ffc5a57104@mail.kernel.org Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
Doruk Tan Ozturk [Tue, 26 May 2026 18:37:26 +0000 (20:37 +0200)]
mac802154: llsec: add skb_cow_data() before in-place crypto
llsec_do_encrypt_unauth(), llsec_do_encrypt_auth(),
llsec_do_decrypt_unauth(), and llsec_do_decrypt_auth() all perform
in-place cryptographic transformations on skb data. They build a
scatterlist with sg_init_one() pointing into the skb's linear data area
and then pass the same scatterlist as both src and dst to the crypto API
(e.g. crypto_skcipher_encrypt/decrypt, crypto_aead_encrypt/decrypt).
On the RX path, __ieee802154_rx_handle_packet() clones the received skb
before handing it to each subscriber via ieee802154_subif_frame(). The
cloned skb shares the same underlying data buffer via reference
counting. When llsec_do_decrypt() subsequently modifies this shared
buffer in place, it corrupts data that other clones -- potentially
belonging to other sockets or subsystems -- still reference.
On the TX path, similar data sharing can occur when an skb's head has
been cloned (skb_cloned() returns true).
The fix is to call skb_cow_data() before performing any in-place crypto
operation. skb_cow_data() ensures that the skb's data area is not
shared: if the skb head is cloned or the data spans multiple fragments,
it copies the data into a private buffer that can be safely modified in
place. This is the same pattern used by:
Without this guard, in-place crypto on shared skb data leads to:
- Silent data corruption of other skb clones
- Use-after-free when the crypto API scatterwalk writes through a
page that has already been freed by another clone's kfree_skb()
- Kernel crashes under concurrent 802.15.4 traffic with security
enabled (KASAN/KMSAN reports slab-use-after-free)
Found by 0sec (https://0sec.ai) using automated source analysis.
Fixes: 4c14a2fb5d14 ("mac802154: add llsec decryption method") Fixes: 03556e4d0dbb ("mac802154: add llsec encryption method") Cc: stable@vger.kernel.org Reported-by: Doruk Tan Ozturk <doruk@0sec.ai> Closes: https://lore.kernel.org/linux-wpan/20260525161806.96158-1-doruk@0sec.ai/ Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com> Signed-off-by: Doruk Tan Ozturk <doruk@0sec.ai> Closes: <link to your mail on lore> Link: https://lore.kernel.org/20260526183726.56100-1-doruk@0sec.ai Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
ieee802154: ca8210: fix pointer truncation in kfifo on 64-bit
ca8210_test_int_driver_write() and ca8210_test_int_user_read() exchange
a kmalloc'd buffer pointer through a struct kfifo, but pass a literal
'4' as the byte count to kfifo_in()/kfifo_out().
This is correct on 32-bit (pointer = 4 bytes), but on 64-bit only the
low 4 bytes of the 8-byte pointer are written into the FIFO. The reader
then reads back 4 bytes into an 8-byte local pointer variable, leaving
the upper 4 bytes uninitialized stack data. The first dereference of
the reconstructed pointer (fifo_buffer[1]) accesses an arbitrary kernel
address and generally results in an oops.
Use sizeof(fifo_buffer) so the byte count matches pointer width on every
architecture.
The driver has no architecture restriction in Kconfig, so any 64-bit
build with CONFIG_IEEE802154_CA8210_DEBUGFS=y is exposed. Issue has
been latent since the driver was added in 2017 because it is most
commonly deployed on 32-bit MCUs.
Found via a custom Coccinelle semantic patch hunting for short-byte
kfifo I/O on byte-mode kfifos used to shuttle pointers.
ieee802154: ca8210: fix cas_ctl leak on spi_async failure
ca8210_spi_transfer() allocates cas_ctl with kzalloc_obj(GFP_ATOMIC)
and relies entirely on the SPI completion callback
ca8210_spi_transfer_complete() to free it.
The spi_async() API only invokes the completion callback on successful
submission. On failure it returns a negative error code without ever
queuing the callback, which leaves cas_ctl and its embedded spi_message
and spi_transfer orphaned. Every kfree(cas_ctl) in the driver is
inside the completion callback, so there is no other reclamation path.
ca8210_spi_transfer() is called from ca8210_spi_exchange(), the
interrupt handler ca8210_interrupt_handler(), and from the retry path
inside the completion callback itself. The exchange and interrupt
handler paths loop on -EBUSY, so under sustained SPI bus contention
every retry iteration leaks a fresh cas_ctl (~600 bytes per
occurrence).
Fix it by freeing cas_ctl on the spi_async() error path. While here,
correct the misleading error string: the function calls spi_async(),
not spi_sync().
Ivan Abramov [Thu, 3 Apr 2025 10:19:34 +0000 (13:19 +0300)]
ieee802154: Remove WARN_ON() in cfg802154_pernet_exit()
There's no need to call WARN_ON() in cfg802154_pernet_exit(), since
every point of failure in cfg802154_switch_netns() is covered with
WARN_ON(), so remove it.
Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
Fixes: 66e5c2672cd1 ("ieee802154: add netns support") Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com> Signed-off-by: Ivan Abramov <i.abramov@mt-integration.ru> Link: https://lore.kernel.org/20250403101935.991385-4-i.abramov@mt-integration.ru Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
Ivan Abramov [Thu, 3 Apr 2025 10:19:33 +0000 (13:19 +0300)]
ieee802154: Avoid calling WARN_ON() on -ENOMEM in cfg802154_switch_netns()
It's pointless to call WARN_ON() in case of an allocation failure in
dev_change_net_namespace() and device_rename(), since it only leads to
useless splats caused by deliberate fault injections, so avoid it.
Found by Linux Verification Center (linuxtesting.org) with Syzkaller.
Lorenzo Bianconi [Mon, 15 Jun 2026 09:18:25 +0000 (11:18 +0200)]
netfilter: flowtable: fix and simplify IP6IP6 tunnel handling
Fix nf_flow_ip6_tunnel_proto() to use pskb_may_pull() instead of
skb_header_pointer() to ensure the outer IPv6 header is in the skb
headroom, which is required for subsequent packet processing. Move
ctx->offset update inside the IPPROTO_IPV6 conditional block since it
should only be adjusted when an IP6IP6 tunnel is actually detected.
Simplify the rx path by removing ipv6_skip_exthdr() and checking
ip6h->nexthdr directly, as the flowtable fast path only handles simple
IP6IP6 encapsulation without extension headers.
Drop the tunnel encapsulation limit destination option support from the
tx path to match, since the rx path no longer handles extension headers.
Remove the encap_limit parameter from nf_flow_offload_ipv6_forward(),
nf_flow_tunnel_ip6ip6_push() and nf_flow_tunnel_v6_push(), along with
the ipv6_tel_txoption struct and related headroom/MTU adjustments.
Fixes: d98103575dcdd ("netfilter: flowtable: Add IP6IP6 rx sw acceleration") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Wyatt Feng [Thu, 11 Jun 2026 07:21:42 +0000 (15:21 +0800)]
netfilter: xt_cluster: reject template conntracks in hash match
xt_cluster_mt() treats any non-NULL nf_ct_get() result as a fully
initialized conntrack and passes it to xt_cluster_hash().
This causes a state confusion bug when the raw table CT target attaches
a template conntrack to skb->_nfct before normal conntrack processing.
Templates carry IPS_TEMPLATE status but do not have a valid tuple for
hashing yet, so xt_cluster_hash() can hit its WARN_ON() path on the
zeroed l3num field.
Reject template conntracks before hashing them. This matches existing
netfilter handling for template objects and avoids hashing incomplete
conntrack state.
Haoze Xie [Mon, 8 Jun 2026 05:43:44 +0000 (13:43 +0800)]
netfilter: nf_queue: pin bridge device while NFQUEUE holds fake dst
The br_netfilter fake rtable is embedded in struct net_bridge and is
attached to bridged packets with skb_dst_set_noref(). If such a packet is
queued to NFQUEUE, __nf_queue() upgrades that fake dst with
skb_dst_force().
At that point the queued skb can hold a real dst reference after bridge
teardown has started. The problem is not that every bridged packet needs
its own dst reference. The problem is that NFQUEUE can keep the bridge
private fake dst alive after unregister begins.
Fix this by keeping the bridge fake dst model unchanged and pinning the
bridge master device only while the packet sits in NFQUEUE. Record the
bridge device in nf_queue_entry when the queued skb carries a bridge fake
dst, take a device reference for the queue lifetime, and drop it when the
queue entry is freed.
Also make sure queued entries are reaped when that bridge device goes
down, and drop the redundant nf_bridge_info_exists() test from the fake
dst detection.
This keeps netdev_priv(br->dev) alive until verdict completion, so the
embedded fake rtable and its metrics backing storage cannot be freed out
from under dst_release(). It also avoids the constant refcount bump and
avoids using ipv4-specific dst helpers for IPv6 bridge traffic.
Fixes: 34666d467cbf ("netfilter: bridge: move br_netfilter out of the core") Cc: stable@kernel.org Reported-by: Yuan Tan <yuantan098@gmail.com> Reported-by: Yifan Wu <yifanwucs@gmail.com> Reported-by: Juefei Pu <tomapufckgml@gmail.com> Reported-by: Xin Liu <bird@lzu.edu.cn> Signed-off-by: Haoze Xie <royenheart@gmail.com> Signed-off-by: Ren Wei <n05ec@lzu.edu.cn> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Adrian Bente [Thu, 28 May 2026 07:08:51 +0000 (10:08 +0300)]
netfilter: flowtable: fix offloaded ct timeout never being extended
OpenWrt has recently migrated many platforms to kernel 6.18. On the
MediaTek platform, which supports hardware network offloading, WiFi
connections accelerated via the WED path were observed to drop after
roughly 300 seconds.
After several debugging sessions, assisted by the Claude LLM, the
problem was narrowed down as follows:
nf_flow_table_extend_ct_timeout() extends ct->timeout for offloaded
flows using:
cmpxchg(&ct->timeout, expires, new_timeout);
'expires' comes from nf_ct_expires(ct) and is a relative value, while
ct->timeout holds an absolute timestamp. The two are never equal, so
the cmpxchg always fails and the timeout is never extended.
This goes unnoticed for most flows, but a long-lived hardware (WED)
offloaded flow on MediaTek MT7986 eventually has ct->timeout decay to
zero, the conntrack entry is reaped and the connection breaks.
Open-code the relative value from a single READ_ONCE(ct->timeout)
snapshot and compare against that same absolute snapshot in the
cmpxchg, so the timeout extension actually takes effect while the
datapath remains authoritative if it updates ct->timeout concurrently.
Daniel Zahka [Wed, 17 Jun 2026 10:39:49 +0000 (03:39 -0700)]
eth: fbnic: take netif_addr_lock_bh() around rx mode address programming
When __fbnic_set_rx_mode() is called from contexts other than
.ndo_set_rx_mode_async(), the uc and mc addr lists are accessed
without the addr lock that __hw_addr_sync_dev() and
__hw_addr_unsync_dev() require. Wrap these unprotected accesses with
netif_addr_lock_bh(). fbnic_clear_rx_mode() has similar issues.
Doruk Tan Ozturk [Wed, 17 Jun 2026 07:58:18 +0000 (09:58 +0200)]
tipc: fix slab-use-after-free Read in tipc_aead_decrypt_done
tipc_aead_decrypt() goes straight from tipc_bearer_hold(b) to
crypto_aead_decrypt(req) without taking a reference on the netns, unlike
the encrypt path. When crypto_aead_decrypt() is offloaded asynchronously
(e.g. the SIMD aead wrapper queuing to cryptd), the cryptd worker runs
tipc_aead_decrypt_done() later. If the bearer's netns is torn down in the
meantime, cleanup_net() -> tipc_exit_net() -> tipc_crypto_stop() frees the
per-netns tipc_crypto, and the completion then reads it:
tipc_aead_decrypt_done() dereferences aead->crypto->stats and
aead->crypto->net, and tipc_crypto_rcv_complete() dereferences
aead->crypto->aead[] and the node table -- reading freed memory.
This is the same class of bug that commit e279024617134 ("net/tipc: fix
slab-use-after-free Read in tipc_aead_encrypt_done") fixed for the encrypt
side. The encrypt path takes maybe_get_net(aead->crypto->net) before
crypto_aead_encrypt() and drops it with put_net() on the synchronous
return paths and in tipc_aead_encrypt_done(); the -EINPROGRESS/-EBUSY
return keeps the reference for the async callback to release. The decrypt
path was left without the equivalent guard.
Mirror the encrypt-side fix on the decrypt path: take a net reference
before crypto_aead_decrypt() (failing with -ENODEV and the matching
bearer put if it cannot be acquired), keep it across the
-EINPROGRESS/-EBUSY async return, and drop it with put_net() on the
synchronous success/error return and at the end of
tipc_aead_decrypt_done().
Reproduced under KASAN on v7.1-rc7: a UDP bearer with a cluster key is
flooded with crafted encrypted frames from an unknown peer (driving the
cluster-key decrypt path) while the bearer's netns is repeatedly torn
down. The completion must run asynchronously to outlive
tipc_crypto_stop(); on x86 the stock aesni gcm(aes) now decrypts
synchronously, so the async path was exercised via cryptd offload. The
unguarded aead->crypto dereference in tipc_aead_decrypt_done() is the
unpatched upstream path; tipc_aead_decrypt() still lacks
maybe_get_net(aead->crypto->net), so the completion can outlive the free
on any config where crypto_aead_decrypt() goes async.
Found by 0sec automated security-research tooling (https://0sec.ai).
Fixes: fc1b6d6de220 ("tipc: introduce TIPC encryption & authentication") Cc: stable@vger.kernel.org Signed-off-by: Doruk Tan Ozturk <doruk@0sec.ai> Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com> Reviewed-by: Tung Nguyen <tung.quang.nguyen@est.tech> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20260617075818.37431-1-doruk@0sec.ai Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: rds: check cmsg_len before reading rds_rdma_args in size pass
rds_rm_size() handles RDS_CMSG_RDMA_ARGS after only CMSG_OK() and then
calls rds_rdma_extra_size(), which reads args->local_vec_addr and
args->nr_local without first checking that cmsg_len covers struct
rds_rdma_args. The other two RDS_CMSG_RDMA_ARGS consumers already guard
this: rds_rdma_bytes() in rds_sendmsg() and rds_cmsg_rdma_args() in
rds_cmsg_send() both reject cmsg_len < CMSG_LEN(sizeof(struct
rds_rdma_args)). Add the same check to rds_rm_size() so all three RDMA
args passes are consistent.
This is a consistency and hardening change with no behavioral effect for
well-formed senders and no reachable bug today: rds_rdma_bytes() runs
before rds_rm_size() in rds_sendmsg() and already rejects a short
RDS_CMSG_RDMA_ARGS, so the size pass is not reached with an undersized
cmsg. But rds_rm_size() reads the args independently of that earlier
pass, and nothing in rds_rm_size() itself records or enforces the
precondition, so a reader or a future refactor of the size pass cannot
tell the cmsg has already been length-checked. Applying the same
cmsg_len guard in all three RDS_CMSG_RDMA_ARGS consumers keeps that
invariant local to each and robust to reordering.
projects / linux / log