A few packets may still be sent out during the termination of iperf
processes. These late packets cause failures in rss_ctx.py when they
arrive on queues expected to be empty.
Example failure observed:
Check failed 2 != 0 traffic on inactive queues (context 1):
[0, 0, 1, 1, 386385, 397196, 0, 0, 0, 0, ...]
Check failed 4 != 0 traffic on inactive queues (context 2):
[0, 0, 0, 0, 2, 2, 247152, 253013, 0, 0, ...]
Check failed 2 != 0 traffic on inactive queues (context 3):
[0, 0, 0, 0, 0, 0, 1, 1, 282434, 283070, ...]
To avoid such failures, wait until all client sockets for the requested
port are either closed or in the TIME_WAIT state.
Fixes: 847aa551fa ("selftests: drv-net: rss_ctx: factor out send traffic and check")
Signed-off-by: Nimrod Oren <noren@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Carolina Jubran <cjubran@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250722122655.3194442-1-noren@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Marc Kleine-Budde says:
====================
pull-request: can 2025-07-22
The patch is by me and fixes a potential NULL pointer deref in the CAN
device driver infrastructure. It can be triggered from user space.
* tag 'linux-can-fixes-for-6.16-20250722' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: netlink: can_changelink(): fix NULL pointer deref of struct can_priv::do_set_mode
====================
Link: https://patch.msgid.link/20250722110059.3664104-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The test is supposed to observe that the 'clash_resolve' stat counter
incremented (i.e., the code path was covered).
This check was incorrect, 'conntrack -S' needs to be called in the
revevant namespace, not the initial netns.
The clash resolution logic in conntrack is only exercised when multiple
packets with the same udp quadruple race. Depending on kernel config,
number of CPUs, scheduling policy etc. this might not trigger even
after several retries. Thus the script eventually returns SKIP if the
retry count is exceeded.
The udpclash tool with also exit with a failure if it did not observe
the expected number of replies.
In the script, make a note of this but do not fail anymore, just check if
the clash resolution logic triggered after all.
Remove the 'single-core' test: while unlikely, with preemptible kernel it
should be possible to also trigger clash resolution logic.
With this change the test will either SKIP or pass.
Hard error could be restored later once its clear whats going on, so
also dump 'conntrack -S' when some packets went missing to see if
conntrack dropped them on insert.
Fixes: 78a5883635 ("selftests: netfilter: add conntrack clash resolution test case")
Signed-off-by: Florian Westphal <fw@strlen.de>
Link: https://patch.msgid.link/20250721223652.6956-1-fw@strlen.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2025-07-21 (i40e, ice, e1000e)
For i40e:
Dennis Chen adjusts reporting of VF Tx dropped to a more appropriate
field.
Jamie Bainbridge fixes a check which can cause a PF set VF MAC address
to be lost.
For ice:
Haoxiang Li adds an error check in DDP load to prevent NULL pointer
dereference.
For e1000e:
Jacek Kowalski adds workarounds for issues surrounding Tiger Lake
platforms with uninitialized NVMs.
* '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
e1000e: ignore uninitialized checksum word on tgp
e1000e: disregard NVM checksum on tgp when valid checksum bit is not set
ice: Fix a null pointer dereference in ice_copy_and_init_pkg()
i40e: When removing VF MAC filters, only check PF-set MAC
i40e: report VF tx_dropped with tx_errors instead of tx_discards
====================
Link: https://patch.msgid.link/20250721173733.2248057-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Andrei Lalaev reported a NULL pointer deref when a CAN device is
restarted from Bus Off and the driver does not implement the struct
can_priv::do_set_mode callback.
There are 2 code path that call struct can_priv::do_set_mode:
- directly by a manual restart from the user space, via
can_changelink()
- delayed automatic restart after bus off (deactivated by default)
To prevent the NULL pointer deference, refuse a manual restart or
configure the automatic restart delay in can_changelink() and report
the error via extack to user space.
As an additional safety measure let can_restart() return an error if
can_priv::do_set_mode is not set instead of dereferencing it
unchecked.
Reported-by: Andrei Lalaev <andrey.lalaev@gmail.com>
Closes: https://lore.kernel.org/all/20250714175520.307467-1-andrey.lalaev@gmail.com
Fixes: 39549eef35 ("can: CAN Network device driver and Netlink interface")
Link: https://patch.msgid.link/20250718-fix-nullptr-deref-do_set_mode-v1-1-0b520097bb96@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
might_sleep could be trigger in the atomic context in qfq_delete_class.
qfq_destroy_class was moved into atomic context locked
by sch_tree_lock to avoid a race condition bug on
qfq_aggregate. However, might_sleep could be triggered by
qfq_destroy_class, which introduced sleeping in atomic context (path:
qfq_destroy_class->qdisc_put->__qdisc_destroy->lockdep_unregister_key
->might_sleep).
Considering the race is on the qfq_aggregate objects, keeping
qfq_rm_from_agg in the lock but moving the left part out can solve
this issue.
Fixes: 5e28d5a3f7 ("net/sched: sch_qfq: Fix race condition on qfq_aggregate")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/4a04e0cc-a64b-44e7-9213-2880ed641d77@sabinyo.mountain
Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://patch.msgid.link/20250717230128.159766-1-xmei5@asu.edu
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
gve_tx_timeout was calculating missed completions in a way that is only
relevant in the GQ queue format. Additionally, it was attempting to
disable device interrupts, which is not needed in either GQ or DQ queue
formats.
As a result, TX timeouts with the DQ queue format likely would have
triggered early resets without kicking the queue at all.
This patch drops the check for pending work altogether and always kicks
the queue after validating the queue has not seen a TX timeout too
recently.
Cc: stable@vger.kernel.org
Fixes: 87a7f321bb ("gve: Recover from queue stall due to missed IRQ")
Co-developed-by: Tim Hostetler <thostet@google.com>
Signed-off-by: Tim Hostetler <thostet@google.com>
Signed-off-by: Praveen Kaligineedi <pkaligineedi@google.com>
Signed-off-by: Harshitha Ramamurthy <hramamurthy@google.com>
Link: https://patch.msgid.link/20250717192024.1820931-1-hramamurthy@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On ASP versions v2.x we need to program the TX map vector register to
properly exercise end-to-end flow control, otherwise the TX engine can
either lock-up, or cause the hardware calculated checksum to be
wrong/corrupted when multiple back to back packets are being submitted
for transmission. This register defaults to 0, which means no flow
control being applied.
Fixes: e9f31435ee ("net: bcmasp: Add support for asp-v3.0")
Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://patch.msgid.link/20250718212242.3447751-1-florian.fainelli@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Matthieu Baerts says:
====================
selftests: mptcp: connect: cover alt modes
mptcp_connect.sh can be executed manually with "-m <MODE>" and "-C" to
make sure everything works as expected when using "mmap" and "sendfile"
modes instead of "poll", and with the MPTCP checksum support.
These modes should be validated, but they are not when the selftests are
executed via the kselftest helpers. It means that most CIs validating
these selftests, like NIPA for the net development trees and LKFT for
the stable ones, are not covering these modes.
To fix that, new test programs have been added, simply calling
mptcp_connect.sh with the right parameters.
The first patch can be backported up to v5.6, and the second one up to
v5.14.
v1: https://lore.kernel.org/20250714-net-mptcp-sft-connect-alt-v1-0-bf1c5abbe575@kernel.org
====================
Link: https://patch.msgid.link/20250715-net-mptcp-sft-connect-alt-v2-0-8230ddd82454@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The "mmap" and "sendfile" alternate modes for mptcp_connect.sh/.c are
available from the beginning, but only tested when mptcp_connect.sh is
manually launched with "-m mmap" or "-m sendfile", not via the
kselftests helpers.
The MPTCP CI was manually running "mptcp_connect.sh -m mmap", but not
"-m sendfile". Plus other CIs, especially the ones validating the stable
releases, were not validating these alternate modes.
To make sure these modes are validated by these CIs, add two new test
programs executing mptcp_connect.sh with the alternate modes.
Fixes: 048d19d444 ("mptcp: add basic kselftest for mptcp")
Cc: stable@vger.kernel.org
Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20250715-net-mptcp-sft-connect-alt-v2-1-8230ddd82454@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As described by Vitaly Lifshits:
> Starting from Tiger Lake, LAN NVM is locked for writes by SW, so the
> driver cannot perform checksum validation and correction. This means
> that all NVM images must leave the factory with correct checksum and
> checksum valid bit set.
Unfortunately some systems have left the factory with an uninitialized
value of 0xFFFF at register address 0x3F (checksum word location).
So on Tiger Lake platform we ignore the computed checksum when such
condition is encountered.
Signed-off-by: Jacek Kowalski <jacek@jacekk.info>
Tested-by: Vlad URSU <vlad@ursu.me>
Fixes: 4051f68318 ("e1000e: Do not take care about recovery NVM checksum")
Cc: stable@vger.kernel.org
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Vitaly Lifshits <vitaly.lifshits@intel.com>
Tested-by: Mor Bar-Gabay <morx.bar.gabay@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
As described by Vitaly Lifshits:
> Starting from Tiger Lake, LAN NVM is locked for writes by SW, so the
> driver cannot perform checksum validation and correction. This means
> that all NVM images must leave the factory with correct checksum and
> checksum valid bit set. Since Tiger Lake devices were the first to have
> this lock, some systems in the field did not meet this requirement.
> Therefore, for these transitional devices we skip checksum update and
> verification, if the valid bit is not set.
Signed-off-by: Jacek Kowalski <jacek@jacekk.info>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Vitaly Lifshits <vitaly.lifshits@intel.com>
Fixes: 4051f68318 ("e1000e: Do not take care about recovery NVM checksum")
Cc: stable@vger.kernel.org
Tested-by: Mor Bar-Gabay <morx.bar.gabay@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
When the PF is processing an Admin Queue message to delete a VF's MACs
from the MAC filter, we currently check if the PF set the MAC and if
the VF is trusted.
This results in undesirable behaviour, where if a trusted VF with a
PF-set MAC sets itself down (which sends an AQ message to delete the
VF's MAC filters) then the VF MAC is erased from the interface.
This results in the VF losing its PF-set MAC which should not happen.
There is no need to check for trust at all, because an untrusted VF
cannot change its own MAC. The only check needed is whether the PF set
the MAC. If the PF set the MAC, then don't erase the MAC on link-down.
Resolve this by changing the deletion check only for PF-set MAC.
(the out-of-tree driver has also intentionally removed the check for VF
trust here with OOT driver version 2.26.8, this changes the Linux kernel
driver behaviour and comment to match the OOT driver behaviour)
Fixes: ea2a1cfc3b ("i40e: Fix VF MAC filter removal")
Signed-off-by: Jamie Bainbridge <jamie.bainbridge@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Currently the tx_dropped field in VF stats is not updated correctly
when reading stats from the PF. This is because it reads from
i40e_eth_stats.tx_discards which seems to be unused for per VSI stats,
as it is not updated by i40e_update_eth_stats() and the corresponding
register, GLV_TDPC, is not implemented[1].
Use i40e_eth_stats.tx_errors instead, which is actually updated by
i40e_update_eth_stats() by reading from GLV_TEPC.
To test, create a VF and try to send bad packets through it:
$ echo 1 > /sys/class/net/enp2s0f0/device/sriov_numvfs
$ cat test.py
from scapy.all import *
vlan_pkt = Ether(dst="ff:ff:ff:ff:ff:ff") / Dot1Q(vlan=999) / IP(dst="192.168.0.1") / ICMP()
ttl_pkt = IP(dst="8.8.8.8", ttl=0) / ICMP()
print("Send packet with bad VLAN tag")
sendp(vlan_pkt, iface="enp2s0f0v0")
print("Send packet with TTL=0")
sendp(ttl_pkt, iface="enp2s0f0v0")
$ ip -s link show dev enp2s0f0
16: enp2s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 3c:ec:ef:b7:e0:ac brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
0 0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
0 0 0 0 0 0
vf 0 link/ether e2:c6:fd:c1:1e:92 brd ff:ff:ff:ff:ff:ff, spoof checking on, link-state auto, trust off
RX: bytes packets mcast bcast dropped
0 0 0 0 0
TX: bytes packets dropped
0 0 0
$ python test.py
Send packet with bad VLAN tag
.
Sent 1 packets.
Send packet with TTL=0
.
Sent 1 packets.
$ ip -s link show dev enp2s0f0
16: enp2s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 3c:ec:ef:b7:e0:ac brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
0 0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
0 0 0 0 0 0
vf 0 link/ether e2:c6:fd:c1:1e:92 brd ff:ff:ff:ff:ff:ff, spoof checking on, link-state auto, trust off
RX: bytes packets mcast bcast dropped
0 0 0 0 0
TX: bytes packets dropped
0 0 0
A packet with non-existent VLAN tag and a packet with TTL = 0 are sent,
but tx_dropped is not incremented.
After patch:
$ ip -s link show dev enp2s0f0
19: enp2s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 3c:ec:ef:b7:e0:ac brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
0 0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
0 0 0 0 0 0
vf 0 link/ether 4a:b7:3d:37:f7:56 brd ff:ff:ff:ff:ff:ff, spoof checking on, link-state auto, trust off
RX: bytes packets mcast bcast dropped
0 0 0 0 0
TX: bytes packets dropped
0 0 2
Fixes: dc645daef9 ("i40e: implement VF stats NDO")
Signed-off-by: Dennis Chen <dechen@redhat.com>
Link: https://www.intel.com/content/www/us/en/content-details/596333/intel-ethernet-controller-x710-tm4-at2-carlsville-datasheet.html
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
In the original design, it is assumed local and peer eswitches have the
same number of vfs. However, in new firmware, local and peer eswitches
can have different number of vfs configured by mlxconfig. In such
configuration, it is incorrect to derive the number of vfs from the
local device's eswitch.
Fix this by updating the peer miss rules add and delete functions to use
the peer device's eswitch and vf count instead of the local device's
information, ensuring correct behavior regardless of vf configuration
differences.
Fixes: ac004b8321 ("net/mlx5e: E-Switch, Add peer miss rules")
Signed-off-by: Shahar Shitrit <shshitrit@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/1752753970-261832-3-git-send-email-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The fsl_mc_get_endpoint() function uses device_find_child() for
localization, which implicitly calls get_device() to increment the
device's reference count before returning the pointer. However, the
caller dpaa2_switch_port_connect_mac() fails to properly release this
reference in multiple scenarios. We should call put_device() to
decrement reference count properly.
As comment of device_find_child() says, 'NOTE: you will need to drop
the reference with put_device() after use'.
Found by code review.
Cc: stable@vger.kernel.org
Fixes: 84cba72956 ("dpaa2-switch: integrate the MAC endpoint support")
Signed-off-by: Ma Ke <make24@iscas.ac.cn>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250717022309.3339976-3-make24@iscas.ac.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The fsl_mc_get_endpoint() function uses device_find_child() for
localization, which implicitly calls get_device() to increment the
device's reference count before returning the pointer. However, the
caller dpaa2_eth_connect_mac() fails to properly release this
reference in multiple scenarios. We should call put_device() to
decrement reference count properly.
As comment of device_find_child() says, 'NOTE: you will need to drop
the reference with put_device() after use'.
Found by code review.
Cc: stable@vger.kernel.org
Fixes: 7194792308 ("dpaa2-eth: add MAC/PHY support through phylink")
Signed-off-by: Ma Ke <make24@iscas.ac.cn>
Tested-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250717022309.3339976-2-make24@iscas.ac.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull networking fixes from Jakub Kicinski:
"Including fixes from Bluetooth, CAN, WiFi and Netfilter.
More code here than I would have liked. That said, better now than
next week. Nothing particularly scary stands out. The improvement to
the OpenVPN input validation is a bit large but better get them in
before the code makes it to a final release. Some of the changes we
got from sub-trees could have been split better between the fix and
-next refactoring, IMHO, that has been communicated.
We have one known regression in a TI AM65 board not getting link. The
investigation is going a bit slow, a number of people are on vacation.
We'll try to wrap it up, but don't think it should hold up the
release.
Current release - fix to a fix:
- Bluetooth: L2CAP: fix attempting to adjust outgoing MTU, it broke
some headphones and speakers
Current release - regressions:
- wifi: ath12k: fix packets received in WBM error ring with REO LUT
enabled, fix Rx performance regression
- wifi: iwlwifi:
- fix crash due to a botched indexing conversion
- mask reserved bits in chan_state_active_bitmap, avoid FW assert()
Current release - new code bugs:
- nf_conntrack: fix crash due to removal of uninitialised entry
- eth: airoha: fix potential UaF in airoha_npu_get()
Previous releases - regressions:
- net: fix segmentation after TCP/UDP fraglist GRO
- af_packet: fix the SO_SNDTIMEO constraint not taking effect and a
potential soft lockup waiting for a completion
- rpl: fix UaF in rpl_do_srh_inline() for sneaky skb geometry
- virtio-net: fix recursive rtnl_lock() during probe()
- eth: stmmac: populate entire system_counterval_t in get_time_fn()
- eth: libwx: fix a number of crashes in the driver Rx path
- hv_netvsc: prevent IPv6 addrconf after IFF_SLAVE lost that meaning
Previous releases - always broken:
- mptcp: fix races in handling connection fallback to pure TCP
- rxrpc: assorted error handling and race fixes
- sched: another batch of "security" fixes for qdiscs (QFQ, HTB)
- tls: always refresh the queue when reading sock, avoid UaF
- phy: don't register LEDs for genphy, avoid deadlock
- Bluetooth: btintel: check if controller is ISO capable on
btintel_classify_pkt_type(), work around FW returning incorrect
capabilities
Misc:
- make OpenVPN Netlink input checking more strict before it makes it
to a final release
- wifi: cfg80211: remove scan request n_channels __counted_by, it's
only yielding false positives"
* tag 'net-6.16-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (66 commits)
rxrpc: Fix to use conn aborts for conn-wide failures
rxrpc: Fix transmission of an abort in response to an abort
rxrpc: Fix notification vs call-release vs recvmsg
rxrpc: Fix recv-recv race of completed call
rxrpc: Fix irq-disabled in local_bh_enable()
selftests/tc-testing: Test htb_dequeue_tree with deactivation and row emptying
net/sched: Return NULL when htb_lookup_leaf encounters an empty rbtree
net: bridge: Do not offload IGMP/MLD messages
selftests: Add test cases for vlan_filter modification during runtime
net: vlan: fix VLAN 0 refcount imbalance of toggling filtering during runtime
tls: always refresh the queue when reading sock
virtio-net: fix recursived rtnl_lock() during probe()
net/mlx5: Update the list of the PCI supported devices
hv_netvsc: Set VF priv_flags to IFF_NO_ADDRCONF before open to prevent IPv6 addrconf
phonet/pep: Move call to pn_skb_get_dst_sockaddr() earlier in pep_sock_accept()
Bluetooth: L2CAP: Fix attempting to adjust outgoing MTU
netfilter: nf_conntrack: fix crash due to removal of uninitialised entry
net: fix segmentation after TCP/UDP fraglist GRO
ipv6: mcast: Delay put pmc->idev in mld_del_delrec()
net: airoha: fix potential use-after-free in airoha_npu_get()
...
Pull power management fixes from Rafael Wysocki:
"These address three issues introduced during the current development
cycle and related to system suspend and hibernation, one triggering
when asynchronous suspend of devices fails, one possibly affecting
memory management in the core suspend code error path, and one due to
duplicate filesystems freezing during system suspend:
- Fix a deadlock that may occur on asynchronous device suspend
failures due to missing completion updates in error paths (Rafael
Wysocki)
- Drop a misplaced pm_restore_gfp_mask() call, which may cause swap
to be accessed too early if system suspend fails, from
suspend_devices_and_enter() (Rafael Wysocki)
- Remove duplicate filesystems_freeze/thaw() calls, which sometimes
cause systems to be unable to resume, from enter_state() (Zihuan
Zhang)"
* tag 'pm-6.16-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
PM: sleep: Update power.completion for all devices on errors
PM: suspend: clean up redundant filesystems_freeze/thaw() handling
PM: suspend: Drop a misplaced pm_restore_gfp_mask() call
Luiz Augusto von Dentz says:
====================
bluetooth pull request for net:
- hci_sync: fix connectable extended advertising when using static random address
- hci_core: fix typos in macros
- hci_core: add missing braces when using macro parameters
- hci_core: replace 'quirks' integer by 'quirk_flags' bitmap
- SMP: If an unallowed command is received consider it a failure
- SMP: Fix using HCI_ERROR_REMOTE_USER_TERM on timeout
- L2CAP: Fix null-ptr-deref in l2cap_sock_resume_cb()
- L2CAP: Fix attempting to adjust outgoing MTU
- btintel: Check if controller is ISO capable on btintel_classify_pkt_type
- btusb: QCA: Fix downloading wrong NVM for WCN6855 GF variant without board ID
* tag 'for-net-2025-07-17' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
Bluetooth: L2CAP: Fix attempting to adjust outgoing MTU
Bluetooth: btusb: QCA: Fix downloading wrong NVM for WCN6855 GF variant without board ID
Bluetooth: hci_dev: replace 'quirks' integer by 'quirk_flags' bitmap
Bluetooth: hci_core: add missing braces when using macro parameters
Bluetooth: hci_core: fix typos in macros
Bluetooth: SMP: Fix using HCI_ERROR_REMOTE_USER_TERM on timeout
Bluetooth: SMP: If an unallowed command is received consider it a failure
Bluetooth: btintel: Check if controller is ISO capable on btintel_classify_pkt_type
Bluetooth: hci_sync: fix connectable extended advertising when using static random address
Bluetooth: Fix null-ptr-deref in l2cap_sock_resume_cb()
====================
Link: https://patch.msgid.link/20250717142849.537425-1-luiz.dentz@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
David Howells says:
====================
rxrpc: Miscellaneous fixes
Here are some fixes for rxrpc:
(1) Fix the calling of IP routing code with IRQs disabled.
(2) Fix a recvmsg/recvmsg race when the first completes a call.
(3) Fix a race between notification, recvmsg and sendmsg releasing a call.
(4) Fix abort of abort.
(5) Fix call-level aborts that should be connection-level aborts.
====================
Link: https://patch.msgid.link/20250717074350.3767366-1-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When a call is released, rxrpc takes the spinlock and removes it from
->recvmsg_q in an effort to prevent racing recvmsg() invocations from
seeing the same call. Now, rxrpc_recvmsg() only takes the spinlock when
actually removing a call from the queue; it doesn't, however, take it in
the lead up to that when it checks to see if the queue is empty. It *does*
hold the socket lock, which prevents a recvmsg/recvmsg race - but this
doesn't prevent sendmsg from ending the call because sendmsg() drops the
socket lock and relies on the call->user_mutex.
Fix this by firstly removing the bit in rxrpc_release_call() that dequeues
the released call and, instead, rely on recvmsg() to simply discard
released calls (done in a preceding fix).
Secondly, rxrpc_notify_socket() is abandoned if the call is already marked
as released rather than trying to be clever by setting both pointers in
call->recvmsg_link to NULL to trick list_empty(). This isn't perfect and
can still race, resulting in a released call on the queue, but recvmsg()
will now clean that up.
Fixes: 17926a7932 ("[AF_RXRPC]: Provide secure RxRPC sockets for use by userspace and kernel both")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Junvyyang, Tencent Zhuque Lab <zhuque@tencent.com>
cc: LePremierHomme <kwqcheii@proton.me>
cc: Simon Horman <horms@kernel.org>
cc: linux-afs@lists.infradead.org
Link: https://patch.msgid.link/20250717074350.3767366-4-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If a call receives an event (such as incoming data), the call gets placed
on the socket's queue and a thread in recvmsg can be awakened to go and
process it. Once the thread has picked up the call off of the queue,
further events will cause it to be requeued, and once the socket lock is
dropped (recvmsg uses call->user_mutex to allow the socket to be used in
parallel), a second thread can come in and its recvmsg can pop the call off
the socket queue again.
In such a case, the first thread will be receiving stuff from the call and
the second thread will be blocked on call->user_mutex. The first thread
can, at this point, process both the event that it picked call for and the
event that the second thread picked the call for and may see the call
terminate - in which case the call will be "released", decoupling the call
from the user call ID assigned to it (RXRPC_USER_CALL_ID in the control
message).
The first thread will return okay, but then the second thread will wake up
holding the user_mutex and, if it sees that the call has been released by
the first thread, it will BUG thusly:
kernel BUG at net/rxrpc/recvmsg.c:474!
Fix this by just dequeuing the call and ignoring it if it is seen to be
already released. We can't tell userspace about it anyway as the user call
ID has become stale.
Fixes: 248f219cb8 ("rxrpc: Rewrite the data and ack handling code")
Reported-by: Junvyyang, Tencent Zhuque Lab <zhuque@tencent.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
cc: LePremierHomme <kwqcheii@proton.me>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Simon Horman <horms@kernel.org>
cc: linux-afs@lists.infradead.org
Link: https://patch.msgid.link/20250717074350.3767366-3-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The rxrpc_assess_MTU_size() function calls down into the IP layer to find
out the MTU size for a route. When accepting an incoming call, this is
called from rxrpc_new_incoming_call() which holds interrupts disabled
across the code that calls down to it. Unfortunately, the IP layer uses
local_bh_enable() which, config dependent, throws a warning if IRQs are
enabled:
WARNING: CPU: 1 PID: 5544 at kernel/softirq.c:387 __local_bh_enable_ip+0x43/0xd0
...
RIP: 0010:__local_bh_enable_ip+0x43/0xd0
...
Call Trace:
<TASK>
rt_cache_route+0x7e/0xa0
rt_set_nexthop.isra.0+0x3b3/0x3f0
__mkroute_output+0x43a/0x460
ip_route_output_key_hash+0xf7/0x140
ip_route_output_flow+0x1b/0x90
rxrpc_assess_MTU_size.isra.0+0x2a0/0x590
rxrpc_new_incoming_peer+0x46/0x120
rxrpc_alloc_incoming_call+0x1b1/0x400
rxrpc_new_incoming_call+0x1da/0x5e0
rxrpc_input_packet+0x827/0x900
rxrpc_io_thread+0x403/0xb60
kthread+0x2f7/0x310
ret_from_fork+0x2a/0x230
ret_from_fork_asm+0x1a/0x30
...
hardirqs last enabled at (23): _raw_spin_unlock_irq+0x24/0x50
hardirqs last disabled at (24): _raw_read_lock_irq+0x17/0x70
softirqs last enabled at (0): copy_process+0xc61/0x2730
softirqs last disabled at (25): rt_add_uncached_list+0x3c/0x90
Fix this by moving the call to rxrpc_assess_MTU_size() out of
rxrpc_init_peer() and further up the stack where it can be done without
interrupts disabled.
It shouldn't be a problem for rxrpc_new_incoming_call() to do it after the
locks are dropped as pmtud is going to be performed by the I/O thread - and
we're in the I/O thread at this point.
Fixes: a2ea9a9072 ("rxrpc: Use irq-disabling spinlocks between app and I/O thread")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Junvyyang, Tencent Zhuque Lab <zhuque@tencent.com>
cc: LePremierHomme <kwqcheii@proton.me>
cc: Simon Horman <horms@kernel.org>
cc: linux-afs@lists.infradead.org
Link: https://patch.msgid.link/20250717074350.3767366-2-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
htb_lookup_leaf has a BUG_ON that can trigger with the following:
tc qdisc del dev lo root
tc qdisc add dev lo root handle 1: htb default 1
tc class add dev lo parent 1: classid 1:1 htb rate 64bit
tc qdisc add dev lo parent 1:1 handle 2: netem
tc qdisc add dev lo parent 2:1 handle 3: blackhole
ping -I lo -c1 -W0.001 127.0.0.1
The root cause is the following:
1. htb_dequeue calls htb_dequeue_tree which calls the dequeue handler on
the selected leaf qdisc
2. netem_dequeue calls enqueue on the child qdisc
3. blackhole_enqueue drops the packet and returns a value that is not
just NET_XMIT_SUCCESS
4. Because of this, netem_dequeue calls qdisc_tree_reduce_backlog, and
since qlen is now 0, it calls htb_qlen_notify -> htb_deactivate ->
htb_deactiviate_prios -> htb_remove_class_from_row -> htb_safe_rb_erase
5. As this is the only class in the selected hprio rbtree,
__rb_change_child in __rb_erase_augmented sets the rb_root pointer to
NULL
6. Because blackhole_dequeue returns NULL, netem_dequeue returns NULL,
which causes htb_dequeue_tree to call htb_lookup_leaf with the same
hprio rbtree, and fail the BUG_ON
The function graph for this scenario is shown here:
0) | htb_enqueue() {
0) + 13.635 us | netem_enqueue();
0) 4.719 us | htb_activate_prios();
0) # 2249.199 us | }
0) | htb_dequeue() {
0) 2.355 us | htb_lookup_leaf();
0) | netem_dequeue() {
0) + 11.061 us | blackhole_enqueue();
0) | qdisc_tree_reduce_backlog() {
0) | qdisc_lookup_rcu() {
0) 1.873 us | qdisc_match_from_root();
0) 6.292 us | }
0) 1.894 us | htb_search();
0) | htb_qlen_notify() {
0) 2.655 us | htb_deactivate_prios();
0) 6.933 us | }
0) + 25.227 us | }
0) 1.983 us | blackhole_dequeue();
0) + 86.553 us | }
0) # 2932.761 us | qdisc_warn_nonwc();
0) | htb_lookup_leaf() {
0) | BUG_ON();
------------------------------------------
The full original bug report can be seen here [1].
We can fix this just by returning NULL instead of the BUG_ON,
as htb_dequeue_tree returns NULL when htb_lookup_leaf returns
NULL.
[1] https://lore.kernel.org/netdev/pF5XOOIim0IuEfhI-SOxTgRvNoDwuux7UHKnE_Y5-zVd4wmGvNk2ceHjKb8ORnzw0cGwfmVu42g9dL7XyJLf1NEzaztboTWcm0Ogxuojoeo=@willsroot.io/
Fixes: 512bb43eb5 ("pkt_sched: sch_htb: Optimize WARN_ONs in htb_dequeue_tree() etc.")
Signed-off-by: William Liu <will@willsroot.io>
Signed-off-by: Savino Dicanosa <savy@syst3mfailure.io>
Link: https://patch.msgid.link/20250717022816.221364-1-will@willsroot.io
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Do not offload IGMP/MLD messages as it could lead to IGMP/MLD Reports
being unintentionally flooded to Hosts. Instead, let the bridge decide
where to send these IGMP/MLD messages.
Consider the case where the local host is sending out reports in response
to a remote querier like the following:
mcast-listener-process (IP_ADD_MEMBERSHIP)
\
br0
/ \
swp1 swp2
| |
QUERIER SOME-OTHER-HOST
In the above setup, br0 will want to br_forward() reports for
mcast-listener-process's group(s) via swp1 to QUERIER; but since the
source hwdom is 0, the report is eligible for tx offloading, and is
flooded by hardware to both swp1 and swp2, reaching SOME-OTHER-HOST as
well. (Example and illustration provided by Tobias.)
Fixes: 472111920f ("net: bridge: switchdev: allow the TX data plane forwarding to be offloaded")
Signed-off-by: Joseph Huang <Joseph.Huang@garmin.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250716153551.1830255-1-Joseph.Huang@garmin.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Assuming the "rx-vlan-filter" feature is enabled on a net device, the
8021q module will automatically add or remove VLAN 0 when the net device
is put administratively up or down, respectively. There are a couple of
problems with the above scheme.
The first problem is a memory leak that can happen if the "rx-vlan-filter"
feature is disabled while the device is running:
# ip link add bond1 up type bond mode 0
# ethtool -K bond1 rx-vlan-filter off
# ip link del dev bond1
When the device is put administratively down the "rx-vlan-filter"
feature is disabled, so the 8021q module will not remove VLAN 0 and the
memory will be leaked [1].
Another problem that can happen is that the kernel can automatically
delete VLAN 0 when the device is put administratively down despite not
adding it when the device was put administratively up since during that
time the "rx-vlan-filter" feature was disabled. null-ptr-unref or
bug_on[2] will be triggered by unregister_vlan_dev() for refcount
imbalance if toggling filtering during runtime:
$ ip link add bond0 type bond mode 0
$ ip link add link bond0 name vlan0 type vlan id 0 protocol 802.1q
$ ethtool -K bond0 rx-vlan-filter off
$ ifconfig bond0 up
$ ethtool -K bond0 rx-vlan-filter on
$ ifconfig bond0 down
$ ip link del vlan0
Root cause is as below:
step1: add vlan0 for real_dev, such as bond, team.
register_vlan_dev
vlan_vid_add(real_dev,htons(ETH_P_8021Q),0) //refcnt=1
step2: disable vlan filter feature and enable real_dev
step3: change filter from 0 to 1
vlan_device_event
vlan_filter_push_vids
ndo_vlan_rx_add_vid //No refcnt added to real_dev vlan0
step4: real_dev down
vlan_device_event
vlan_vid_del(dev, htons(ETH_P_8021Q), 0); //refcnt=0
vlan_info_rcu_free //free vlan0
step5: delete vlan0
unregister_vlan_dev
BUG_ON(!vlan_info); //vlan_info is null
Fix both problems by noting in the VLAN info whether VLAN 0 was
automatically added upon NETDEV_UP and based on that decide whether it
should be deleted upon NETDEV_DOWN, regardless of the state of the
"rx-vlan-filter" feature.
[1]
unreferenced object 0xffff8880068e3100 (size 256):
comm "ip", pid 384, jiffies 4296130254
hex dump (first 32 bytes):
00 20 30 0d 80 88 ff ff 00 00 00 00 00 00 00 00 . 0.............
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
backtrace (crc 81ce31fa):
__kmalloc_cache_noprof+0x2b5/0x340
vlan_vid_add+0x434/0x940
vlan_device_event.cold+0x75/0xa8
notifier_call_chain+0xca/0x150
__dev_notify_flags+0xe3/0x250
rtnl_configure_link+0x193/0x260
rtnl_newlink_create+0x383/0x8e0
__rtnl_newlink+0x22c/0xa40
rtnl_newlink+0x627/0xb00
rtnetlink_rcv_msg+0x6fb/0xb70
netlink_rcv_skb+0x11f/0x350
netlink_unicast+0x426/0x710
netlink_sendmsg+0x75a/0xc20
__sock_sendmsg+0xc1/0x150
____sys_sendmsg+0x5aa/0x7b0
___sys_sendmsg+0xfc/0x180
[2]
kernel BUG at net/8021q/vlan.c:99!
Oops: invalid opcode: 0000 [#1] SMP KASAN PTI
CPU: 0 UID: 0 PID: 382 Comm: ip Not tainted 6.16.0-rc3 #61 PREEMPT(voluntary)
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:unregister_vlan_dev (net/8021q/vlan.c:99 (discriminator 1))
RSP: 0018:ffff88810badf310 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff88810da84000 RCX: ffffffffb47ceb9a
RDX: dffffc0000000000 RSI: 0000000000000008 RDI: ffff88810e8b43c8
RBP: 0000000000000000 R08: 0000000000000000 R09: fffffbfff6cefe80
R10: ffffffffb677f407 R11: ffff88810badf3c0 R12: ffff88810e8b4000
R13: 0000000000000000 R14: ffff88810642a5c0 R15: 000000000000017e
FS: 00007f1ff68c20c0(0000) GS:ffff888163a24000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f1ff5dad240 CR3: 0000000107e56000 CR4: 00000000000006f0
Call Trace:
<TASK>
rtnl_dellink (net/core/rtnetlink.c:3511 net/core/rtnetlink.c:3553)
rtnetlink_rcv_msg (net/core/rtnetlink.c:6945)
netlink_rcv_skb (net/netlink/af_netlink.c:2535)
netlink_unicast (net/netlink/af_netlink.c:1314 net/netlink/af_netlink.c:1339)
netlink_sendmsg (net/netlink/af_netlink.c:1883)
____sys_sendmsg (net/socket.c:712 net/socket.c:727 net/socket.c:2566)
___sys_sendmsg (net/socket.c:2622)
__sys_sendmsg (net/socket.c:2652)
do_syscall_64 (arch/x86/entry/syscall_64.c:63 arch/x86/entry/syscall_64.c:94)
Fixes: ad1afb0039 ("vlan_dev: VLAN 0 should be treated as "no vlan tag" (802.1p packet)")
Reported-by: syzbot+a8b046e462915c65b10b@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=a8b046e462915c65b10b
Suggested-by: Ido Schimmel <idosch@idosch.org>
Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250716034504.2285203-2-dongchenchen2@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Antonio Quartulli says:
====================
This bugfix batch includes the following changes:
* properly propagate sk mark to skb->mark field
* reject unexpected incoming netlink attributes
* reset GSO state when moving skb from transport to tunnel layer
* tag 'ovpn-net-20250716' of https://github.com/OpenVPN/ovpn-net-next:
ovpn: reset GSO metadata after decapsulation
ovpn: reject unexpected netlink attributes
ovpn: propagate socket mark to skb in UDP
====================
Link: https://patch.msgid.link/20250716115443.16763-1-antonio@openvpn.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After recent changes in net-next TCP compacts skbs much more
aggressively. This unearthed a bug in TLS where we may try
to operate on an old skb when checking if all skbs in the
queue have matching decrypt state and geometry.
BUG: KASAN: slab-use-after-free in tls_strp_check_rcv+0x898/0x9a0 [tls]
(net/tls/tls_strp.c:436 net/tls/tls_strp.c:530 net/tls/tls_strp.c:544)
Read of size 4 at addr ffff888013085750 by task tls/13529
CPU: 2 UID: 0 PID: 13529 Comm: tls Not tainted 6.16.0-rc5-virtme
Call Trace:
kasan_report+0xca/0x100
tls_strp_check_rcv+0x898/0x9a0 [tls]
tls_rx_rec_wait+0x2c9/0x8d0 [tls]
tls_sw_recvmsg+0x40f/0x1aa0 [tls]
inet_recvmsg+0x1c3/0x1f0
Always reload the queue, fast path is to have the record in the queue
when we wake, anyway (IOW the path going down "if !strp->stm.full_len").
Fixes: 0d87bbd39d ("tls: strp: make sure the TCP skbs do not have overlapping data")
Link: https://patch.msgid.link/20250716143850.1520292-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The deadlock appears in a stack trace like:
virtnet_probe()
rtnl_lock()
virtio_config_changed_work()
netdev_notify_peers()
rtnl_lock()
It happens if the VMM sends a VIRTIO_NET_S_ANNOUNCE request while the
virtio-net driver is still probing.
The config_work in probe() will get scheduled until virtnet_open() enables
the config change notification via virtio_config_driver_enable().
Fixes: df28de7b00 ("virtio-net: synchronize operstate with admin state on up/down")
Signed-off-by: Zigit Zo <zuozhijie@bytedance.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Link: https://patch.msgid.link/20250716115717.1472430-1-zuozhijie@bytedance.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A new warning in clang [1] points out a place in pep_sock_accept() where
dst is uninitialized then passed as a const pointer to pep_find_pipe():
net/phonet/pep.c:829:37: error: variable 'dst' is uninitialized when passed as a const pointer argument here [-Werror,-Wuninitialized-const-pointer]
829 | newsk = pep_find_pipe(&pn->hlist, &dst, pipe_handle);
| ^~~:
Move the call to pn_skb_get_dst_sockaddr(), which initializes dst, to
before the call to pep_find_pipe(), so that dst is consistently used
initialized throughout the function.
Cc: stable@vger.kernel.org
Fixes: f7ae8d59f6 ("Phonet: allocate sock from accept syscall rather than soft IRQ")
Link: 00dacf8c22 [1]
Closes: https://github.com/ClangBuiltLinux/linux/issues/2101
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://patch.msgid.link/20250715-net-phonet-fix-uninit-const-pointer-v1-1-8efd1bd188b3@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Configuration request only configure the incoming direction of the peer
initiating the request, so using the MTU is the other direction shall
not be used, that said the spec allows the peer responding to adjust:
Bluetooth Core 6.1, Vol 3, Part A, Section 4.5
'Each configuration parameter value (if any is present) in an
L2CAP_CONFIGURATION_RSP packet reflects an ‘adjustment’ to a
configuration parameter value that has been sent (or, in case of
default values, implied) in the corresponding
L2CAP_CONFIGURATION_REQ packet.'
That said adjusting the MTU in the response shall be limited to ERTM
channels only as for older modes the remote stack may not be able to
detect the adjustment causing it to silently drop packets.
Link: https://github.com/bluez/bluez/issues/1422
Link: https://gitlab.archlinux.org/archlinux/packaging/packages/linux/-/issues/149
Link: https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/4793
Fixes: 042bb9603c ("Bluetooth: L2CAP: Fix L2CAP MTU negotiation")
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Johannes Berg says:
====================
Couple of fixes:
- ath12k performance regression from -rc1
- cfg80211 counted_by() removal for scan request
as it doesn't match usage and keeps complaining
- iwlwifi crash with certain older devices
- iwlwifi missing an error path unlock
- iwlwifi compatibility with certain BIOS updates
* tag 'wireless-2025-07-17' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: iwlwifi: Fix botched indexing conversion
wifi: cfg80211: remove scan request n_channels counted_by
wifi: ath12k: Fix packets received in WBM error ring with REO LUT enabled
wifi: iwlwifi: mask reserved bits in chan_state_active_bitmap
wifi: iwlwifi: pcie: fix locking on invalid TOP reset
====================
Link: https://patch.msgid.link/20250717091831.18787-5-johannes@sipsolutions.net
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following batch contains Netfilter fixes for net:
1) Three patches to enhance conntrack selftests for resize and clash
resolution, from Florian Westphal.
2) Expand nft_concat_range.sh selftest to improve coverage from error
path, from Florian Westphal.
3) Hide clash bit to userspace from netlink dumps until there is a
good reason to expose, from Florian Westphal.
4) Revert notification for device registration/unregistration for
nftables basechains and flowtables, we decided to go for a better
way to handle this through the nfnetlink_hook infrastructure which
will come via nf-next, patch from Phil Sutter.
5) Fix crash in conntrack due to race related to SLAB_TYPESAFE_BY_RCU
that results in removing a recycled object that is not yet in the
hashes. Move IPS_CONFIRM setting after the object is in the hashes.
From Florian Westphal.
netfilter pull request 25-07-17
* tag 'nf-25-07-17' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: nf_conntrack: fix crash due to removal of uninitialised entry
Revert "netfilter: nf_tables: Add notifications for hook changes"
netfilter: nf_tables: hide clash bit from userspace
selftests: netfilter: nft_concat_range.sh: send packets to empty set
selftests: netfilter: conntrack_resize.sh: also use udpclash tool
selftests: netfilter: add conntrack clash resolution test case
selftests: netfilter: conntrack_resize.sh: extend resize test
====================
Link: https://patch.msgid.link/20250717095808.41725-1-pablo@netfilter.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
A crash in conntrack was reported while trying to unlink the conntrack
entry from the hash bucket list:
[exception RIP: __nf_ct_delete_from_lists+172]
[..]
#7 [ff539b5a2b043aa0] nf_ct_delete at ffffffffc124d421 [nf_conntrack]
#8 [ff539b5a2b043ad0] nf_ct_gc_expired at ffffffffc124d999 [nf_conntrack]
#9 [ff539b5a2b043ae0] __nf_conntrack_find_get at ffffffffc124efbc [nf_conntrack]
[..]
The nf_conn struct is marked as allocated from slab but appears to be in
a partially initialised state:
ct hlist pointer is garbage; looks like the ct hash value
(hence crash).
ct->status is equal to IPS_CONFIRMED|IPS_DYING, which is expected
ct->timeout is 30000 (=30s), which is unexpected.
Everything else looks like normal udp conntrack entry. If we ignore
ct->status and pretend its 0, the entry matches those that are newly
allocated but not yet inserted into the hash:
- ct hlist pointers are overloaded and store/cache the raw tuple hash
- ct->timeout matches the relative time expected for a new udp flow
rather than the absolute 'jiffies' value.
If it were not for the presence of IPS_CONFIRMED,
__nf_conntrack_find_get() would have skipped the entry.
Theory is that we did hit following race:
cpu x cpu y cpu z
found entry E found entry E
E is expired <preemption>
nf_ct_delete()
return E to rcu slab
init_conntrack
E is re-inited,
ct->status set to 0
reply tuplehash hnnode.pprev
stores hash value.
cpu y found E right before it was deleted on cpu x.
E is now re-inited on cpu z. cpu y was preempted before
checking for expiry and/or confirm bit.
->refcnt set to 1
E now owned by skb
->timeout set to 30000
If cpu y were to resume now, it would observe E as
expired but would skip E due to missing CONFIRMED bit.
nf_conntrack_confirm gets called
sets: ct->status |= CONFIRMED
This is wrong: E is not yet added
to hashtable.
cpu y resumes, it observes E as expired but CONFIRMED:
<resumes>
nf_ct_expired()
-> yes (ct->timeout is 30s)
confirmed bit set.
cpu y will try to delete E from the hashtable:
nf_ct_delete() -> set DYING bit
__nf_ct_delete_from_lists
Even this scenario doesn't guarantee a crash:
cpu z still holds the table bucket lock(s) so y blocks:
wait for spinlock held by z
CONFIRMED is set but there is no
guarantee ct will be added to hash:
"chaintoolong" or "clash resolution"
logic both skip the insert step.
reply hnnode.pprev still stores the
hash value.
unlocks spinlock
return NF_DROP
<unblocks, then
crashes on hlist_nulls_del_rcu pprev>
In case CPU z does insert the entry into the hashtable, cpu y will unlink
E again right away but no crash occurs.
Without 'cpu y' race, 'garbage' hlist is of no consequence:
ct refcnt remains at 1, eventually skb will be free'd and E gets
destroyed via: nf_conntrack_put -> nf_conntrack_destroy -> nf_ct_destroy.
To resolve this, move the IPS_CONFIRMED assignment after the table
insertion but before the unlock.
Pablo points out that the confirm-bit-store could be reordered to happen
before hlist add resp. the timeout fixup, so switch to set_bit and
before_atomic memory barrier to prevent this.
It doesn't matter if other CPUs can observe a newly inserted entry right
before the CONFIRMED bit was set:
Such event cannot be distinguished from above "E is the old incarnation"
case: the entry will be skipped.
Also change nf_ct_should_gc() to first check the confirmed bit.
The gc sequence is:
1. Check if entry has expired, if not skip to next entry
2. Obtain a reference to the expired entry.
3. Call nf_ct_should_gc() to double-check step 1.
nf_ct_should_gc() is thus called only for entries that already failed an
expiry check. After this patch, once the confirmed bit check passes
ct->timeout has been altered to reflect the absolute 'best before' date
instead of a relative time. Step 3 will therefore not remove the entry.
Without this change to nf_ct_should_gc() we could still get this sequence:
1. Check if entry has expired.
2. Obtain a reference.
3. Call nf_ct_should_gc() to double-check step 1:
4 - entry is still observed as expired
5 - meanwhile, ct->timeout is corrected to absolute value on other CPU
and confirm bit gets set
6 - confirm bit is seen
7 - valid entry is removed again
First do check 6), then 4) so the gc expiry check always picks up either
confirmed bit unset (entry gets skipped) or expiry re-check failure for
re-inited conntrack objects.
This change cannot be backported to releases before 5.19. Without
commit 8a75a2c174 ("netfilter: conntrack: remove unconfirmed list")
|= IPS_CONFIRMED line cannot be moved without further changes.
Cc: Razvan Cojocaru <rzvncj@gmail.com>
Link: https://lore.kernel.org/netfilter-devel/20250627142758.25664-1-fw@strlen.de/
Link: https://lore.kernel.org/netfilter-devel/4239da15-83ff-4ca4-939d-faef283471bb@gmail.com/
Fixes: 1397af5bfd ("netfilter: conntrack: remove the percpu dying list")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
