Merge tag 'net-6.17-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Including fixes from netfilter, wireless and Bluetooth.
We're reverting the removal of a Sundance driver, a user has appeared.
This makes the PR rather large in terms of LoC.
There's a conspicuous absence of real, user-reported 6.17 issues.
Slightly worried that the summer distracted people from testing.
Previous releases - regressions:
- ax25: properly unshare skbs in ax25_kiss_rcv()
Previous releases - always broken:
- phylink: disable autoneg for interfaces that have no inband, fix
regression on pcs-lynx (NXP LS1088)
- vxlan: fix null-deref when using nexthop objects
- batman-adv: fix OOB read/write in network-coding decode
- icmp: icmp_ndo_send: fix reversing address translation for replies
- tcp: fix socket ref leak in TCP-AO failure handling for IPv6
- mctp:
- mctp_fraq_queue should take ownership of passed skb
- usb: initialise mac header in RX path, avoid WARN
- wifi: mac80211: do not permit 40 MHz EHT operation on 5/6 GHz,
respect device limitations
- wifi: wilc1000: avoid buffer overflow in WID string configuration
- wifi: mt76:
- fix regressions from mt7996 MLO support rework
- fix offchannel handling issues on mt7996
- fix multiple wcid linked list corruption issues
- mt7921: don't disconnect when AP requests switch to a channel
which requires radar detection
- mt7925u: use connac3 tx aggr check in tx complete
- wifi: intel:
- improve validation of ACPI DSM data
- cfg: restore some 1000 series configs
- wifi: ath:
- ath11k: a fix for GTK rekeying
- ath12k: a missed WiFi7 capability (multi-link EMLSR)
- eth: intel:
- ice: fix races in "low latency" firmware interface for Tx timestamps
- idpf: set mac type when adding and removing MAC filters
- i40e: remove racy read access to some debugfs files
Misc:
- Revert "eth: remove the DLink/Sundance (ST201) driver"
Merge tag 'slab-for-6.17-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab
Pull slab fixes from Vlastimil Babka:
- Stable fix to make slub_debug code not access invalid pointers in the
process of reporting issues (Li Qiong)
- Stable fix to make object tracking pass gfp flags to stackdepot to
avoid deadlock in contexts that can't even wake up kswapd due to e.g.
timers debugging enabled (yangshiguang)
* tag 'slab-for-6.17-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab:
mm: slub: avoid wake up kswapd in set_track_prepare
mm/slub: avoid accessing metadata when pointer is invalid in object_err()
phy: mscc: Stop taking ts_lock for tx_queue and use its own lock
When transmitting a PTP frame which is timestamp using 2 step, the
following warning appears if CONFIG_PROVE_LOCKING is enabled:
=============================
[ BUG: Invalid wait context ] 6.17.0-rc1-00326-ge6160462704e #427 Not tainted
-----------------------------
ptp4l/119 is trying to lock: c2a44ed4 (&vsc8531->ts_lock){+.+.}-{3:3}, at: vsc85xx_txtstamp+0x50/0xac
other info that might help us debug this:
context-{4:4}
4 locks held by ptp4l/119:
#0: c145f068 (rcu_read_lock_bh){....}-{1:2}, at: __dev_queue_xmit+0x58/0x1440
#1: c29df974 (dev->qdisc_tx_busylock ?: &qdisc_tx_busylock){+...}-{2:2}, at: __dev_queue_xmit+0x5c4/0x1440
#2: c2aaaad0 (_xmit_ETHER#2){+.-.}-{2:2}, at: sch_direct_xmit+0x108/0x350
#3: c2aac170 (&lan966x->tx_lock){+.-.}-{2:2}, at: lan966x_port_xmit+0xd0/0x350
stack backtrace:
CPU: 0 UID: 0 PID: 119 Comm: ptp4l Not tainted 6.17.0-rc1-00326-ge6160462704e #427 NONE
Hardware name: Generic DT based system
Call trace:
unwind_backtrace from show_stack+0x10/0x14
show_stack from dump_stack_lvl+0x7c/0xac
dump_stack_lvl from __lock_acquire+0x8e8/0x29dc
__lock_acquire from lock_acquire+0x108/0x38c
lock_acquire from __mutex_lock+0xb0/0xe78
__mutex_lock from mutex_lock_nested+0x1c/0x24
mutex_lock_nested from vsc85xx_txtstamp+0x50/0xac
vsc85xx_txtstamp from lan966x_fdma_xmit+0xd8/0x3a8
lan966x_fdma_xmit from lan966x_port_xmit+0x1bc/0x350
lan966x_port_xmit from dev_hard_start_xmit+0xc8/0x2c0
dev_hard_start_xmit from sch_direct_xmit+0x8c/0x350
sch_direct_xmit from __dev_queue_xmit+0x680/0x1440
__dev_queue_xmit from packet_sendmsg+0xfa4/0x1568
packet_sendmsg from __sys_sendto+0x110/0x19c
__sys_sendto from sys_send+0x18/0x20
sys_send from ret_fast_syscall+0x0/0x1c
Exception stack(0xf0b05fa8 to 0xf0b05ff0)
5fa0: 000000010000000e0000000e0004b47a0000003a00000000
5fc0: 000000010000000e00000000000001210004af58000448740000000000000000
5fe0: 00000001bee9d42000025a10b6e75c7c
So, instead of using the ts_lock for tx_queue, use the spinlock that
skb_buff_head has.
selftest: net: Fix weird setsockopt() in bind_bhash.c.
bind_bhash.c passes (SO_REUSEADDR | SO_REUSEPORT) to setsockopt().
In the asm-generic definition, the value happens to match with the
bare SO_REUSEPORT, (2 | 15) == 15, but not on some arch.
arch/alpha/include/uapi/asm/socket.h:18:#define SO_REUSEADDR 0x0004
arch/alpha/include/uapi/asm/socket.h:24:#define SO_REUSEPORT 0x0200
arch/mips/include/uapi/asm/socket.h:24:#define SO_REUSEADDR 0x0004 /* Allow reuse of local addresses. */
arch/mips/include/uapi/asm/socket.h:33:#define SO_REUSEPORT 0x0200 /* Allow local address and port reuse. */
arch/parisc/include/uapi/asm/socket.h:12:#define SO_REUSEADDR 0x0004
arch/parisc/include/uapi/asm/socket.h:18:#define SO_REUSEPORT 0x0200
arch/sparc/include/uapi/asm/socket.h:13:#define SO_REUSEADDR 0x0004
arch/sparc/include/uapi/asm/socket.h:20:#define SO_REUSEPORT 0x0200
include/uapi/asm-generic/socket.h:12:#define SO_REUSEADDR 2
include/uapi/asm-generic/socket.h:27:#define SO_REUSEPORT 15
Let's pass SO_REUSEPORT only.
Fixes: c35ecb95c448 ("selftests/net: Add test for timing a bind request to a port with a populated bhash entry") Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20250903222938.2601522-1-kuniyu@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Wed, 3 Sep 2025 21:20:54 +0000 (14:20 -0700)]
MAINTAINERS: add Sabrina to TLS maintainers
Sabrina has been very helpful reviewing TLS patches, fixing bugs,
and, I believe, the last one to implement any major feature in
the TLS code base (rekeying). Add her as a maintainer.
Qianfeng Rong [Wed, 3 Sep 2025 12:34:03 +0000 (20:34 +0800)]
net: dsa: dsa_loop: use int type to store negative error codes
Change the 'ret' variable in dsa_loop_init() from unsigned int to int, as
it needs to store either negative error codes or zero returned by
mdio_driver_register().
Storing the negative error codes in unsigned type, doesn't cause an issue
at runtime but can be confusing. Additionally, assigning negative error
codes to unsigned type may trigger a GCC warning when the -Wsign-conversion
flag is enabled.
Qingfang Deng [Wed, 3 Sep 2025 10:07:26 +0000 (18:07 +0800)]
ppp: fix memory leak in pad_compress_skb
If alloc_skb() fails in pad_compress_skb(), it returns NULL without
releasing the old skb. The caller does:
skb = pad_compress_skb(ppp, skb);
if (!skb)
goto drop;
drop:
kfree_skb(skb);
When pad_compress_skb() returns NULL, the reference to the old skb is
lost and kfree_skb(skb) ends up doing nothing, leading to a memory leak.
Align pad_compress_skb() semantics with realloc(): only free the old
skb if allocation and compression succeed. At the call site, use the
new_skb variable so the original skb is not lost when pad_compress_skb()
fails.
Fixes: b3f9b92a6ec1 ("[PPP]: add PPP MPPE encryption module") Signed-off-by: Qingfang Deng <dqfext@gmail.com> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Yue Haibing <yuehaibing@huawei.com> Link: https://patch.msgid.link/20250903100726.269839-1-dqfext@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add proper error checking for dmaengine_desc_get_metadata_ptr() which
can return an error pointer and lead to potential crashes or undefined
behaviour if the pointer retrieval fails.
Properly handle the error by unmapping DMA buffer, freeing the skb and
returning early to prevent further processing with invalid data.
Jakub Kicinski [Thu, 4 Sep 2025 13:59:27 +0000 (06:59 -0700)]
Merge tag 'nf-25-09-04' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Florian Westphal says:
====================
netfilter: updates for net
1) Fix a silly bug in conntrack selftest, busyloop may get optimized to
for (;;), reported by Yi Chen.
2) Introduce new NFTA_DEVICE_PREFIX attribute in nftables netlink api,
re-using old NFTA_DEVICE_NAME led to confusion with different
kernel/userspace versions. This refines the wildcard interface
support added in 6.16 release. From Phil Sutter.
* tag 'nf-25-09-04' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: nf_tables: Introduce NFTA_DEVICE_PREFIX
selftests: netfilter: fix udpclash tool hang
====================
====================
eth: fbnic: support queue API and zero-copy Rx
Add support for queue API to fbnic, enable zero-copy Rx.
Patch 10 is likely of most interest as it adds a new core helper
(and touches mlx5). The rest of the patches are fbnic-specific
(and relatively boring).
Patches 1-3 reshuffle the Rx init/allocation path to better
align structures and functions which operate on them. Notably
patch 1 moves the page pool pointer to the queue struct (from NAPI).
Patch 4 converts the driver to use netmem_ref. The driver has
separate and explicit buffer queue for scatter / payloads, so only
references to those are converted.
Next 5 patches are more boring code shifts.
Patch 11 adds unreadable memory support to page pool allocation.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:14 +0000 (14:12 -0700)]
eth: fbnic: support queue ops / zero-copy Rx
Support queue ops. fbnic doesn't shut down the entire device
just to restart a single queue.
./tools/testing/selftests/drivers/net/hw/iou-zcrx.py
TAP version 13
1..3
ok 1 iou-zcrx.test_zcrx
ok 2 iou-zcrx.test_zcrx_oneshot
ok 3 iou-zcrx.test_zcrx_rss
# Totals: pass:3 fail:0 xfail:0 xpass:0 skip:0 error:0
Jakub Kicinski [Mon, 1 Sep 2025 21:12:13 +0000 (14:12 -0700)]
eth: fbnic: don't pass NAPI into pp alloc
Queue API may ask us to allocate page pools when the device
is down, to validate that we ingested a memory provider binding.
Don't require NAPI to be passed to fbnic_alloc_qt_page_pools(),
to make calling fbnic_alloc_qt_page_pools() without NAPI possible.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:12 +0000 (14:12 -0700)]
eth: fbnic: defer page pool recycling activation to queue start
We need to be more careful about when direct page pool recycling
is enabled in preparation for queue ops support. Don't set the
NAPI pointer, call page_pool_enable_direct_recycling() from
the function that activates the queue (once the config can
no longer fail).
Jakub Kicinski [Mon, 1 Sep 2025 21:12:11 +0000 (14:12 -0700)]
eth: fbnic: allocate unreadable page pool for the payloads
Allow allocating a page pool with unreadable memory for the payload
ring (sub1). We need to provide the queue ID so that the memory provider
can match the PP. Use the appropriate page pool DMA sync helper.
For unreadable mem the direction has to be FROM_DEVICE. The default
is BIDIR for XDP, but obviously unreadable mem is not compatible
with XDP in the first place, so that's fine. While at it remove
the define for page pool flags.
The rxq_idx is passed to fbnic_alloc_rx_qt_resources() explicitly
to make it easy to allocate page pools without NAPI (see the patch
after the next).
Jakub Kicinski [Mon, 1 Sep 2025 21:12:10 +0000 (14:12 -0700)]
net: add helper to pre-check if PP for an Rx queue will be unreadable
mlx5 pokes into the rxq state to check if the queue has a memory
provider, and therefore whether it may produce unreadable mem.
Add a helper for doing this in the page pool API. fbnic will want
a similar thing (tho, for a slightly different reason).
Jakub Kicinski [Mon, 1 Sep 2025 21:12:08 +0000 (14:12 -0700)]
eth: fbnic: split fbnic_enable()
Factor out handling a single nv from fbnic_enable() to make
it reusable for queue ops. Use a __ prefix for the factored
out code. The real fbnic_nv_enable() which will include
fbnic_wrfl() will be added with the qops, to avoid unused
function warnings.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:06 +0000 (14:12 -0700)]
eth: fbnic: split fbnic_disable()
Factor out handling a single nv from fbnic_disable() to make
it reusable for queue ops. Use a __ prefix for the factored
out code. The real fbnic_nv_disable() which will include
fbnic_wrfl() will be added with the qops, to avoid unused
function warnings.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:05 +0000 (14:12 -0700)]
eth: fbnic: request ops lock
We'll add queue ops soon so. queue ops will opt the driver into
extra locking. Request this locking explicitly already to make
future patches smaller and easier to review.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:04 +0000 (14:12 -0700)]
eth: fbnic: use netmem_ref where applicable
Use netmem_ref instead of struct page pointer in prep for
unreadable memory. fbnic has separate free buffer submission
queues for headers and for data. Refactor the helper which
returns page pointer for a submission buffer to take the
high level queue container, create a separate handler
for header and payload rings. This ties the "upcast" from
netmem to system page to use of sub0 which we know has
system pages.
Jakub Kicinski [Mon, 1 Sep 2025 21:12:02 +0000 (14:12 -0700)]
eth: fbnic: move xdp_rxq_info_reg() to resource alloc
Move rxq_info and mem model registration from fbnic_alloc_napi_vector()
and fbnic_alloc_nv_resources() to fbnic_alloc_rx_qt_resources().
The rxq_info is now registered later in the process, but that
should not cause any issues.
rxq_info lives in the fbnic_q_triad (qt) struct so qt init is a more
natural place. Encapsulating the logic in the qt functions will also
allow simplifying the cleanup in the NAPI related alloc functions
in the next commit.
Rx does not have a dedicated fbnic_free_rx_qt_resources(),
but we can use xdp_rxq_info_is_reg() to tell whether given
rxq_info was in use (effectively - if it's a qt for an Rx queue).
Having to pass nv into fbnic_alloc_rx_qt_resources() is not
great in terms of layering, but that's temporary, pp will
move soon..
Jakub Kicinski [Mon, 1 Sep 2025 21:12:01 +0000 (14:12 -0700)]
eth: fbnic: move page pool pointer from NAPI to the ring struct
In preparation for memory providers we need a closer association
between queues and page pools. We used to have a page pool at the
NAPI level to serve all associated queues but with MP the queues
under a NAPI may no longer be created equal.
The "ring" structure in fbnic is a descriptor ring. We have separate
"rings" for payload and header pages ("to device"), as well as a ring
for completions ("from device"). Technically we only need the page
pool pointers in the "to device" rings, so adding the pointer to
the ring struct is a bit wasteful. But it makes passing the structures
around much easier.
For now both "to device" rings store a pointer to the same
page pool. Using more than one queue per NAPI is extremely rare
so don't bother trying to share a single page pool between queues.
ipv6: sit: Add ipip6_tunnel_dst_find() for cleanup
Extract the dst lookup logic from ipip6_tunnel_xmit() into new helper
ipip6_tunnel_dst_find() to reduce code duplication and enhance readability.
No functional change intended.
On a x86_64, with allmodconfig object size is also reduced:
Wang Liang [Mon, 1 Sep 2025 06:35:37 +0000 (14:35 +0800)]
net: atm: fix memory leak in atm_register_sysfs when device_register fail
When device_register() return error in atm_register_sysfs(), which can be
triggered by kzalloc fail in device_private_init() or other reasons,
kmemleak reports the following memory leaks:
The current R-Car S4 rswitch driver only supports port based fowarding.
This patch set adds HW offloading for L2 switching/bridgeing. The driver
hooks into switchdev.
1. Rename the base driver file to keep the driver name (rswitch.ko)
2. Add setting of default MAC ageing time in hardware.
3. Add the L2 driver extension in a separate file. The HW offloading
is automatically configured when a port is added to the bridge device.
Usage example:
ip link add name br0 type bridge
ip link set dev tsn0 master br0
ip link set dev tsn1 master br0
ip link set dev br0 up
ip link set dev tsn0 up
ip link set dev tsn1 up
Layer 2 traffic is now fowarded by HW from port TSN0 to port TSN1.
4. Provides the functionality to set the MAC table ageing time in the
Rswitch.
Usage example:
ip link change dev br0 type bridge ageing 100
Michael Dege [Mon, 1 Sep 2025 04:58:07 +0000 (06:58 +0200)]
net: renesas: rswitch: add offloading for L2 switching
Add hardware offloading for L2 switching on R-Car S4.
On S4 brdev is limited to one per-device (not per port). Reasoning
is that hw L2 forwarding support lacks any sort of source port based
filtering, which makes it unusable to offload more than one bridge
device. Either you allow hardware to forward destination MAC to a
port, or you have to send it to CPU. You can't make it forward only
if src and dst ports are in the same brdev.
Michael Dege [Mon, 1 Sep 2025 04:58:05 +0000 (06:58 +0200)]
net: renesas: rswitch: rename rswitch.c to rswitch_main.c
Adding new functionality to the driver. Therefore splitting into multiple
c files to keep them manageable. New functionality will be added to
separate files.
Reviewed-by: Andrew Lunn <andrew@lunn.ch> Reviewed-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> Reviewed-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Michael Dege <michael.dege@renesas.com> Link: https://patch.msgid.link/20250901-add_l2_switching-v5-1-5f13e46860d5@renesas.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This new attribute is supposed to be used instead of NFTA_DEVICE_NAME
for simple wildcard interface specs. It holds a NUL-terminated string
representing an interface name prefix to match on.
While kernel code to distinguish full names from prefixes in
NFTA_DEVICE_NAME is simpler than this solution, reusing the existing
attribute with different semantics leads to confusion between different
versions of kernel and user space though:
* With old kernels, wildcards submitted by user space are accepted yet
silently treated as regular names.
* With old user space, wildcards submitted by kernel may cause crashes
since libnftnl expects NUL-termination when there is none.
Using a distinct attribute type sanitizes these situations as the
receiving part detects and rejects the unexpected attribute nested in
*_HOOK_DEVS attributes.
Fixes: 6d07a289504a ("netfilter: nf_tables: Support wildcard netdev hook specs") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Florian Westphal <fw@strlen.de>
Florian Westphal [Wed, 27 Aug 2025 17:17:32 +0000 (19:17 +0200)]
selftests: netfilter: fix udpclash tool hang
Yi Chen reports that 'udpclash' loops forever depending on compiler
(and optimization level used); while (x == 1) gets optimized into
for (;;). Add volatile qualifier to avoid that.
While at it, also run it under timeout(1) and fix the resize script
to not ignore the timeout passed as second parameter to insert_flood.
Reported-by: Yi Chen <yiche@redhat.com> Suggested-by: Yi Chen <yiche@redhat.com> Fixes: 78a588363587 ("selftests: netfilter: add conntrack clash resolution test case") Signed-off-by: Florian Westphal <fw@strlen.de>
====================
net: phy: micrel: Add PTP support for lan8842
The PTP block in lan8842 is the same as lan8814 so reuse all these
functions. The first patch of the series just does cosmetic changes such
that lan8842 can reuse the function lan8814_ptp_probe. There should not be
any functional changes here. While the second patch adds the PTP support
to lan8842.
====================
It has the same PTP IP block as lan8814, only the number of GPIOs is
different, all the other functionality is the same. So reuse the same
functions as lan8814 for lan8842.
There is a revision of lan8842 called lan8832 which doesn't have the PTP
IP block. So make sure in that case the PTP is not initialized.
net: phy: micrel: Introduce function __lan8814_ptp_probe_once
Introduce the function __lan8814_ptp_probe_once as this function will be
used also by lan8842 driver which has a different number of GPIOs
compared to lan8814. This change doesn't have any functional
changes.
Juraj Šarinay [Tue, 2 Sep 2025 11:36:28 +0000 (13:36 +0200)]
net: nfc: nci: Increase NCI_DATA_TIMEOUT to 3000 ms
An exchange with a NFC target must complete within NCI_DATA_TIMEOUT.
A delay of 700 ms is not sufficient for cryptographic operations on smart
cards. CardOS 6.0 may need up to 1.3 seconds to perform 256-bit ECDH
or 3072-bit RSA. To prevent brute-force attacks, passports and similar
documents introduce even longer delays into access control protocols
(BAC/PACE).
The timeout should be higher, but not too much. The expiration allows
us to detect that a NFC target has disappeared.
mctp: return -ENOPROTOOPT for unknown getsockopt options
In mctp_getsockopt(), unrecognized options currently return -EINVAL.
In contrast, mctp_setsockopt() returns -ENOPROTOOPT for unknown
options.
Update mctp_getsockopt() to also return -ENOPROTOOPT for unknown
options. This aligns the behavior of getsockopt() and setsockopt(),
and matches the standard kernel socket API convention for handling
unsupported options.
net/smc: Remove validation of reserved bits in CLC Decline message
Currently SMC code is validating the reserved bits while parsing the incoming
CLC decline message & when this validation fails, its treated as a protocol
error. As a result, the SMC connection is terminated instead of falling back to
TCP. As per RFC7609[1] specs we shouldn't be validating the reserved bits that
is part of CLC message. This patch fixes this issue.
Dan Carpenter [Tue, 2 Sep 2025 06:36:08 +0000 (09:36 +0300)]
ipv4: Fix NULL vs error pointer check in inet_blackhole_dev_init()
The inetdev_init() function never returns NULL. Check for error
pointers instead.
Fixes: 22600596b675 ("ipv4: give an IPv4 dev to blackhole_netdev") Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Eric Dumazet <edumazet@google.com> Link: https://patch.msgid.link/aLaQWL9NguWmeM1i@stanley.mountain Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: thunder_bgx: decrement cleanup index before use
All paths in probe that call goto defer do so before assigning phydev
and thus it makes sense to cleanup the prior index. It also fixes a bug
where index 0 does not get cleaned up.
Fixes: b7d3e3d3d21a ("net: thunderx: Don't leak phy device references on -EPROBE_DEFER condition.") Signed-off-by: Rosen Penev <rosenp@gmail.com> Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20250901213314.48599-1-rosenp@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
net: stmmac: allow generation of flexible PPS relative to MAC time
When doing some testing on stm32mp2x platforms(MACv5), I noticed that
the command previously used with a MACv4 for genering a PPS signal:
echo "0 0 0 1 1" > /sys/class/ptp/ptp0/period
did not work.
This is because the arguments passed through this command must contain
the start time at which the PPS should be generated, relative to the
MAC system time. For some reason, a time set in the past seems to work
with a MACv4.
Because passing such an argument is tedious, consider that any time
set in the past is an offset regarding the MAC system time. This way,
this does not impact existing scripts and the past time use case is
handled. Edit: But maybe that's not important and we can just change
the default behavior to this.
Example to generate a flexible PPS signal that has a 1s period 3s
relative to when the command was entered:
drivers: net: stmmac: handle start time set in the past for flexible PPS
In case the time arguments used for flexible PPS signal generation are in
the past, consider the arguments to be a time offset relative to the MAC
system time.
This way, past time use case is handled and it avoids the tedious work
of passing an absolute time value for the flexible PPS signal generation
while not breaking existing scripts that may rely on this behavior.
Russell King (Oracle) [Mon, 1 Sep 2025 11:52:56 +0000 (12:52 +0100)]
net: phylink: move PHY interrupt request to non-fail path
The blamed commit added code which could return an error after we
requested the PHY interrupt. When we return an error, the caller
will call phy_detach() which fails to free the interrupt.
Rearrange the code such that failing operations happen before the
interrupt is requested, thereby allowing phy_detach() to be used.
Note that replacing phy_detach() with phy_disconnect() in these
paths could lead to freeing an interrupt which was never requested.
Fixes: 1942b1c6f687 ("net: phylink: make configuring clock-stop dependent on MAC support") Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Link: https://patch.msgid.link/E1ut35k-00000001UEl-0iq6@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Joy Zou [Mon, 1 Sep 2025 10:36:32 +0000 (18:36 +0800)]
net: stmmac: imx: add i.MX91 support
Add i.MX91 specific settings for EQoS.
Reviewed-by: Alexander Stein <alexander.stein@ew.tq-group.com> Reviewed-by: Frank Li <Frank.Li@nxp.com> Signed-off-by: Joy Zou <joy.zou@nxp.com> Link: https://patch.msgid.link/20250901103632.3409896-7-joy.zou@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For ice:
Jake adds checks for initialization of Tx timestamp tracking structure
to prevent NULL pointer dereferences.
For idpf:
Josh moves freeing of auxiliary device id to prevent use-after-free issue.
Emil sets, expected, MAC type value when sending virtchnl add/delete MAC
commands.
For i40e:
Jake removes read debugfs access as 'netdev_ops' has the possibility to
overflow.
Zhen Ni adds handling for when MAC list is empty.
For ixgbe:
Alok Tiwari corrects bitmap being used for link speeds.
For e1000e:
Vitaly adds check to ensure overflow does not occur in
e1000_set_eeprom().
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
e1000e: fix heap overflow in e1000_set_eeprom
ixgbe: fix incorrect map used in eee linkmode
i40e: Fix potential invalid access when MAC list is empty
i40e: remove read access to debugfs files
idpf: set mac type when adding and removing MAC filters
idpf: fix UAF in RDMA core aux dev deinitialization
ice: fix NULL access of tx->in_use in ice_ll_ts_intr
ice: fix NULL access of tx->in_use in ice_ptp_ts_irq
====================
Eric Dumazet [Tue, 2 Sep 2025 18:36:03 +0000 (18:36 +0000)]
net: lockless sock_i_ino()
Followup of commit c51da3f7a161 ("net: remove sock_i_uid()")
A recent syzbot report was the trigger for this change.
Over the years, we had many problems caused by the
read_lock[_bh](&sk->sk_callback_lock) in sock_i_uid().
We could fix smc_diag_dump_proto() or make a more radical move:
Instead of waiting for new syzbot reports, cache the socket
inode number in sk->sk_ino, so that we no longer
need to acquire sk->sk_callback_lock in sock_i_ino().
This makes socket dumps faster (one less cache line miss,
and two atomic ops avoided).
Prior art:
commit 25a9c8a4431c ("netlink: Add __sock_i_ino() for __netlink_diag_dump().")
commit 4f9bf2a2f5aa ("tcp: Don't acquire inet_listen_hashbucket::lock with disabled BH.")
commit efc3dbc37412 ("rds: Make rds_sock_lock BH rather than IRQ safe.")
Fixes: d2d6422f8bd1 ("x86: Allow to enable PREEMPT_RT.") Reported-by: syzbot+50603c05bbdf4dfdaffa@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/68b73804.050a0220.3db4df.01d8.GAE@google.com/T/#u Signed-off-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com> Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Link: https://patch.msgid.link/20250902183603.740428-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Wed, 3 Sep 2025 23:06:44 +0000 (16:06 -0700)]
Merge tag 'nf-next-25-09-02' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next
Florian Westphal says:
====================
netfilter: updates for net-next
1) prefer vmalloc_array in ebtables, from Qianfeng Rong.
2) Use csum_replace4 instead of open-coding it, from Christophe Leroy.
3+4) Get rid of GFP_ATOMIC in transaction object allocations, those
cause silly failures with large sets under memory pressure, from
myself.
5) Remove test for AVX cpu feature in nftables pipapo set type,
testing for AVX2 feature is sufficient.
6) Unexport a few function in nf_reject infra: no external callers.
7) Extend payload offset to u16, this was restricted to values <=255
so far, from Fernando Fernandez Mancera.
* tag 'nf-next-25-09-02' of https://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next:
netfilter: nft_payload: extend offset to 65535 bytes
netfilter: nf_reject: remove unneeded exports
netfilter: nft_set_pipapo: remove redundant test for avx feature bit
netfilter: nf_tables: all transaction allocations can now sleep
netfilter: nf_tables: allow iter callbacks to sleep
netfilter: nft_payload: Use csum_replace4() instead of opencoding
netfilter: ebtables: Use vmalloc_array() to improve code
====================
STMMAC on SoCFPGA uses exactly one interrupt in in-kernel DTS and common
snps,dwmac.yaml binding is flexible, so define precise constraint for
this device.
The blamed commit introduced the concept of split attribute
counting, and later allocating an array to hold them, however
TypeArrayNest wasn't updated to use the new counting variable.
Abbreviated example from tools/net/ynl/generated/nl80211-user.c:
nl80211_if_combination_attributes_parse(...):
unsigned int n_limits = 0;
[...]
ynl_attr_for_each(attr, nlh, yarg->ys->family->hdr_len)
if (type == NL80211_IFACE_COMB_LIMITS)
ynl_attr_for_each_nested(attr2, attr)
dst->_count.limits++;
if (n_limits) {
dst->_count.limits = n_limits;
/* allocate and parse attributes */
}
In the above example n_limits is guaranteed to always be 0,
hence the conditional is unsatisfiable and is optimized out.
This patch changes the attribute counting to use n_limits++ in the
attribute counting loop in the above example.
tools: ynl-gen: use macro for binary min-len check
This patch changes the generated min-len check for binary
attributes to use the NLA_POLICY_MIN_LEN() macro, thereby the
generated code supports strict policy validation.
With this change TypeBinary will always generate a NLA_BINARY
attribute policy.
This doesn't change any currently generated code, as it isn't
used in any specs currently used for generating code.
While updating the binary min-len implementation, I noticed that
the only user, should AFAICT be using exact-len instead.
In net/ipv4/fou_core.c FOU_ATTR_LOCAL_V6 and FOU_ATTR_PEER_V6
are only used for singular IPv6 addresses, and there are AFAICT
no known implementations trying to send more, it therefore
appears safe to change it to an exact-len policy.
This patch therefore changes the local-v6/peer-v6 attributes to
use an exact-len check, instead of a min-len check.
Christoph Paasch [Tue, 2 Sep 2025 21:11:36 +0000 (23:11 +0200)]
mptcp: record subflows in RPS table
Accelerated Receive Flow Steering (aRFS) relies on sockets recording
their RX flow hash into the rps_sock_flow_table so that incoming packets
are steered to the CPU where the application runs.
With MPTCP, the application interacts with the parent MPTCP socket while
data is carried over per-subflow TCP sockets. Without recording these
subflows, aRFS cannot steer interrupts and RX processing for the flows
to the desired CPU.
Record all subflows in the RPS table by calling sock_rps_record_flow()
for each subflow at the start of mptcp_sendmsg(), mptcp_recvmsg() and
mptcp_stream_accept(), by using the new helper
mptcp_rps_record_subflows().
It does not by itself improve throughput, but ensures that IRQ and RX
processing are directed to the right CPU, which is a
prerequisite for effective aRFS.
Christoph Paasch [Tue, 2 Sep 2025 21:11:35 +0000 (23:11 +0200)]
net: Add rfs_needed() helper
Add a helper to check if RFS is needed or not. Allows to make the code a
bit cleaner and the next patch to have MPTCP use this helper to decide
whether or not to iterate over the subflows.
tun_flow_update() was calling sock_rps_record_flow_hash() regardless of
the state of rfs_needed. This was not really a bug as sock_flow_table
simply ends up being NULL and thus everything will be fine.
This commit here thus also implicitly makes tun_flow_update() respect
the state of rfs_needed.
Gang Yan [Tue, 2 Sep 2025 21:11:34 +0000 (23:11 +0200)]
selftests: mptcp: add checks for fallback counters
Recently, some mib counters about fallback has been added, this patch
provides a method to check the expected behavior of these mib counters
during the test execution.
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/571 Signed-off-by: Gang Yan <yangang@kylinos.cn> Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20250902-net-next-mptcp-misc-feat-6-18-v2-2-fa02bb3188b1@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Wed, 3 Sep 2025 21:56:14 +0000 (14:56 -0700)]
Merge tag 'wireless-2025-09-03' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless
Johannes Berg says:
====================
Just a few updates:
- a set of buffer overflow fixes
- ath11k: a fix for GTK rekeying
- ath12k: a missed WiFi7 capability
* tag 'wireless-2025-09-03' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: wilc1000: avoid buffer overflow in WID string configuration
wifi: cfg80211: sme: cap SSID length in __cfg80211_connect_result()
wifi: libertas: cap SSID len in lbs_associate()
wifi: cw1200: cap SSID length in cw1200_do_join()
wifi: ath11k: fix group data packet drops during rekey
wifi: ath12k: Set EMLSR support flag in MLO flags for EML-capable stations
====================
Merge tag 'soc-fixes-6.17-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc
Pull SoC fixes from Arnd Bergmann:
"These are mainly devicetree fixes for the rockchip and nxp platforms
on arm64, addressing mistakes in the board and soc specific
descriptions.
In particular the newly added Rock 5T board required multiple bugfixes
for PCIe and USB, while on the i.MX platform there are a number of
regulator related fixes. The only other platforms with devicetree
fixes are at91 with a fixup for SD/MMC and a change to enable all the
available UARTS on the Axiado reference board.
Also on the at91 platform, a Kconfig change addresses a regression
that stopped the DMA engine from working in 6.17-rc.
Three drivers each have a simple bugfix, stopping incorrect behavior
in op-tee firmware, the tee subsystem and the qualcomm mdt_loader.
Two trivial MAINTAINERS file changes are needed to make sure that
patches reach the correct maintainer, but don't change the actual
responsibilities"
* tag 'soc-fixes-6.17-2' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: (27 commits)
ARM: dts: microchip: sama7d65: Force SDMMC Legacy mode
ARM: at91: select ARCH_MICROCHIP
arm64: dts: rockchip: fix second M.2 slot on ROCK 5T
arm64: dts: rockchip: fix USB on RADXA ROCK 5T
MAINTAINERS: exclude defconfig from ARM64 PORT
arm64: dts: axiado: Add missing UART aliases
MAINTAINERS: Update Nobuhiro Iwamatsu's email address
arm64: dts: rockchip: Add vcc-supply to SPI flash on Pinephone Pro
arm64: dts: rockchip: fix es8388 address on rk3588s-roc-pc
arm64: dts: rockchip: Fix Bluetooth interrupts flag on Neardi LBA3368
arm64: dts: rockchip: correct network description on Sige5
arm64: dts: rockchip: Minor whitespace cleanup
ARM: dts: rockchip: Minor whitespace cleanup
arm64: dts: rockchip: Add supplies for eMMC on rk3588-orangepi-5
arm64: dts: rockchip: Fix the headphone detection on the orangepi 5 plus
arm64: dts: imx95: Fix JPEG encoder node assigned clock
arm64: dts: imx95-19x19-evk: correct the phy setting for flexcan1/2
arm64: dts: imx8mp: Fix missing microSD slot vqmmc on Data Modul i.MX8M Plus eDM SBC
arm64: dts: imx8mp: Fix missing microSD slot vqmmc on DH electronics i.MX8M Plus DHCOM
arm64: dts: imx8mp-tqma8mpql: remove virtual 3.3V regulator
...
Merge tag 'at91-fixes-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/at91/linux into arm/fixes
Microchip AT91 fixes for v6.17
This update includes:
- adaptation to the SDHCI capabilities on sama7d65 curiosity board DT as
SDHCI quirks are not in place yet. SD/MMC don't work without these
- addition of one Kconfig symbol that is already used in DMA tree for
6.17. XDMA cannot be selected if not present.
* tag 'at91-fixes-6.17' of git://git.kernel.org/pub/scm/linux/kernel/git/at91/linux:
ARM: dts: microchip: sama7d65: Force SDMMC Legacy mode
ARM: at91: select ARCH_MICROCHIP
Merge tag 'bitmap-for-6.17-rc5' of https://github.com/norov/linux
Pull bitmap fix from Yury Norov:
"Fix sched_numa_find_nth_cpu() if mask offline
sched_numa_find_nth_cpu() uses a bsearch to look for the 'closest' CPU
in sched_domains_numa_masks and given cpus mask. However they might
not intersect if all CPUs in the cpus mask are offline.
bsearch will return NULL in that case, bail out instead of
dereferencing a bogus pointer"
* tag 'bitmap-for-6.17-rc5' of https://github.com/norov/linux:
sched: Fix sched_numa_find_nth_cpu() if mask offline
Robert Marko [Wed, 13 Aug 2025 17:44:38 +0000 (19:44 +0200)]
ARM: at91: select ARCH_MICROCHIP
Like with the ARM64 Microchip platforms, lets add a generic ARCH_MICROCHIP
symbol and select it so that drivers that are reused for multiple product
generation or lines, can just depend on it instead of adding each SoC
symbol as their dependencies.
Signed-off-by: Robert Marko <robert.marko@sartura.hr> Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com> Acked-by: Daniel Machon <daniel.machon@microchip.com> Link: https://lore.kernel.org/r/20250813174720.540015-3-robert.marko@sartura.hr Signed-off-by: Claudiu Beznea <claudiu.beznea@tuxon.dev> Signed-off-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Christian Loehle [Wed, 3 Sep 2025 15:48:32 +0000 (16:48 +0100)]
sched: Fix sched_numa_find_nth_cpu() if mask offline
sched_numa_find_nth_cpu() uses a bsearch to look for the 'closest'
CPU in sched_domains_numa_masks and given cpus mask. However they
might not intersect if all CPUs in the cpus mask are offline. bsearch
will return NULL in that case, bail out instead of dereferencing a
bogus pointer.
The previous behaviour lead to this bug when using maxcpus=4 on an
rk3399 (LLLLbb) (i.e. booting with all big CPUs offline):
Merge tag 'v6.17-rockchip-dtsfixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip into arm/fixes
The newly added Rock 5T board needed slightly bigger fixes to make the
PCIe and USB actually work, because the PCIe does share its lanes between
two ports and the usb needs to toggle a gpio to supply power.
The other interesting fix is the headphone detection on the Orange Pi 5+.
The rest are some added supplies to make the boot log less scary and a
number of styling fixes.
* tag 'v6.17-rockchip-dtsfixes1' of git://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip:
arm64: dts: rockchip: fix second M.2 slot on ROCK 5T
arm64: dts: rockchip: fix USB on RADXA ROCK 5T
arm64: dts: rockchip: Add vcc-supply to SPI flash on Pinephone Pro
arm64: dts: rockchip: fix es8388 address on rk3588s-roc-pc
arm64: dts: rockchip: Fix Bluetooth interrupts flag on Neardi LBA3368
arm64: dts: rockchip: correct network description on Sige5
arm64: dts: rockchip: Minor whitespace cleanup
ARM: dts: rockchip: Minor whitespace cleanup
arm64: dts: rockchip: Add supplies for eMMC on rk3588-orangepi-5
arm64: dts: rockchip: Fix the headphone detection on the orangepi 5 plus
arm64: dts: rockchip: Add vcc-supply to SPI flash on rk3399-pinebook-pro
arm64: dts: rockchip: mark eeprom as read-only for Radxa E52C
Johannes Berg [Wed, 3 Sep 2025 07:40:04 +0000 (09:40 +0200)]
Merge tag 'ath-current-20250902' of git://git.kernel.org/pub/scm/linux/kernel/git/ath/ath
Jeff Johnson says:
==================
ath.git update for v6.17-rc5
Fix a long-standing issue with ath11k dropping group data packets
during GTK rekey, and fix an omission in the ath12k multi-link EMLSR
support introduced in v6.16.
==================
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Ajay.Kathat@microchip.com [Fri, 29 Aug 2025 22:58:43 +0000 (22:58 +0000)]
wifi: wilc1000: avoid buffer overflow in WID string configuration
Fix the following copy overflow warning identified by Smatch checker.
drivers/net/wireless/microchip/wilc1000/wlan_cfg.c:184 wilc_wlan_parse_response_frame()
error: '__memcpy()' 'cfg->s[i]->str' copy overflow (512 vs 65537)
This patch introduces size check before accessing the memory buffer.
The checks are base on the WID type of received data from the firmware.
For WID string configuration, the size limit is determined by individual
element size in 'struct wilc_cfg_str_vals' that is maintained in 'len' field
of 'struct wilc_cfg_str'.
Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Closes: https://lore.kernel.org/linux-wireless/aLFbr9Yu9j_TQTey@stanley.mountain Suggested-by: Dan Carpenter <dan.carpenter@linaro.org> Signed-off-by: Ajay Singh <ajay.kathat@microchip.com> Link: https://patch.msgid.link/20250829225829.5423-1-ajay.kathat@microchip.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
====================
net: dsa: lantiq_gswip: prepare for supporting MaxLinear GSW1xx
Continue to prepare for supporting the newer standalone MaxLinear GSW1xx
switch family by extending the existing lantiq_gswip driver to allow it
to support MII interfaces and MDIO bus of the GSW1xx.
This series has been preceded by an RFC series which covers everything
needed to support the MaxLinear GSW1xx family of switches. Andrew Lunn
had suggested to split it into a couple of smaller series and start
with the changes which don't yet make actual functional changes or
support new features.
Everything has been compile and runtime tested on AVM Fritz!Box 7490
(GSWIP version 2.1, VR9 v1.2)
Daniel Golle [Sat, 30 Aug 2025 02:34:48 +0000 (03:34 +0100)]
net: dsa: lantiq_gswip: move MDIO bus registration to .setup()
Instead of registering the switch MDIO bus in the probe() function, move
the call to gswip_mdio() into the .setup() DSA switch op, so it can be
reused independently of the probe() function.
Daniel Golle [Sat, 30 Aug 2025 02:34:23 +0000 (03:34 +0100)]
net: dsa: lantiq_gswip: support standard MDIO node name
Instead of matching against the child node's compatible string also
support locating the node of the device tree node of the MDIO bus
in the standard way by referencing the node name ("mdio").
Daniel Golle [Sat, 30 Aug 2025 02:34:13 +0000 (03:34 +0100)]
net: dsa: lantiq_gswip: support offset of MII registers
The MaxLinear GSW1xx family got a single (R)(G)MII port at index 5 but
the registers MII_PCDU and MII_CFG are those of port 0.
Allow applying an offset for the port index to access those registers.
Daniel Golle [Sat, 30 Aug 2025 02:33:48 +0000 (03:33 +0100)]
net: dsa: lantiq_gswip: ignore SerDes modes in phylink_mac_config()
We can safely ignore SerDes interface modes 1000Base-X, 2500Base-X and
SGMII in phylink_mac_config() as they are being taken care of by the PCS
and the SGMII port anyway doesn't have MII_CFG and MII_PCDU registers
and hence gswip_phylink_mac_config() is already a no-op apart from
outputing a misleading error message.
Return early in case of SerDes interface modes to avoid printing that
error message.
Daniel Golle [Sat, 30 Aug 2025 02:33:00 +0000 (03:33 +0100)]
net: dsa: lantiq_gswip: support model-specific mac_select_pcs()
Call mac_select_pcs() function if provided in struct gswip_hwinfo.
The MaxLinear GSW1xx series got one port wired to a SerDes PCS and
PHY which can do 1000Base-X, 2500Base-X and SGMII. Support for the
SerDes port will be provided using phylink_pcs, so provide a
convenient way for mac_select_pcs() to differ based on the hardware
model.
Daniel Golle [Sat, 30 Aug 2025 02:32:42 +0000 (03:32 +0100)]
net: dsa: lantiq_gswip: move to dedicated folder
Move the lantiq_gswip driver to its own folder and update
MAINTAINERS file accordingly.
This is done ahead of extending the driver to support the MaxLinear
GSW1xx series of standalone switch ICs, which includes adding a bunch
of files.
ipv6: Add sanity checks on ipv6_devconf.rpl_seg_enabled
In ipv6_rpl_srh_rcv() we use min(net->ipv6.devconf_all->rpl_seg_enabled,
idev->cnf.rpl_seg_enabled) is intended to return 0 when either value is
zero, but if one of the values is negative it will in fact return non-zero.
====================
vxlan: Fix NPDs when using nexthop objects
With FDB nexthop groups, VXLAN FDB entries do not necessarily point to
a remote destination but rather to an FDB nexthop group. This means that
first_remote_{rcu,rtnl}() can return NULL and a few places in the driver
were not ready for that, resulting in NULL pointer dereferences.
Patches #1-#2 fix these NPDs.
Note that vxlan_fdb_find_uc() still dereferences the remote returned by
first_remote_rcu() without checking that it is not NULL, but this
function is only invoked by a single driver which vetoes the creation of
FDB nexthop groups. I will patch this in net-next to make the code less
fragile.
Patch #3 adds a selftests which exercises these code paths and tests
basic Tx functionality with FDB nexthop groups. I verified that the test
crashes the kernel without the first two patches.
====================
vxlan: Fix NPD in {arp,neigh}_reduce() when using nexthop objects
When the "proxy" option is enabled on a VXLAN device, the device will
suppress ARP requests and IPv6 Neighbor Solicitation messages if it is
able to reply on behalf of the remote host. That is, if a matching and
valid neighbor entry is configured on the VXLAN device whose MAC address
is not behind the "any" remote (0.0.0.0 / ::).
The code currently assumes that the FDB entry for the neighbor's MAC
address points to a valid remote destination, but this is incorrect if
the entry is associated with an FDB nexthop group. This can result in a
NPD [1][3] which can be reproduced using [2][4].
Fix by checking that the remote destination exists before dereferencing
it.
Fixes: 1274e1cc4226 ("vxlan: ecmp support for mac fdb entries") Reviewed-by: Petr Machata <petrm@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org> Link: https://patch.msgid.link/20250901065035.159644-3-idosch@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
vxlan: Fix NPD when refreshing an FDB entry with a nexthop object
VXLAN FDB entries can point to either a remote destination or an FDB
nexthop group. The latter is usually used in EVPN deployments where
learning is disabled.
However, when learning is enabled, an incoming packet might try to
refresh an FDB entry that points to an FDB nexthop group and therefore
does not have a remote. Such packets should be dropped, but they are
only dropped after dereferencing the non-existent remote, resulting in a
NPD [1] which can be reproduced using [2].
Fix by dropping such packets earlier. Remove the misleading comment from
first_remote_rcu().
ip address add 192.0.2.1/32 dev lo
ip address add 192.0.2.2/32 dev lo
ip nexthop add id 1 via 192.0.2.3 fdb
ip nexthop add id 10 group 1 fdb
ip link add name vx0 up type vxlan id 10010 local 192.0.2.1 dstport 12345 localbypass
ip link add name vx1 up type vxlan id 10020 local 192.0.2.2 dstport 54321 learning
bridge fdb add 00:11:22:33:44:55 dev vx0 self static dst 192.0.2.2 port 54321 vni 10020
bridge fdb add 00:aa:bb:cc:dd:ee dev vx1 self static nhid 10
mausezahn vx0 -a 00:aa:bb:cc:dd:ee -b 00:11:22:33:44:55 -c 1 -q
Fixes: 1274e1cc4226 ("vxlan: ecmp support for mac fdb entries") Reported-by: Marlin Cremers <mcremers@cloudbear.nl> Reviewed-by: Petr Machata <petrm@nvidia.com> Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org> Link: https://patch.msgid.link/20250901065035.159644-2-idosch@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
projects / users / hch / misc.git / log