When an afs network namespace is torn down, it cancels and waits for the
work item that keeps the preallocated rxrpc call/conn/peer queue charged
before disabling incoming (i.e. listen 0), but there's a small window in
which it can be requeued by an incoming call wending through the I/O
thread.
Fix this by cancelling the charger work item again after reducing the
listen backlog to zero.
Fixes: 47694fbc9d ("afs: Fix netns teardown to cancel the preallocation charger")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Link: https://sashiko.dev/#/patchset/20260609140911.838677-1-dhowells%40redhat.com
cc: Li Daming <d4n.for.sec@gmail.com>
cc: Ren Wei <n05ec@lzu.edu.cn>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Jeffrey Altman <jaltman@auristor.com>
cc: Simon Horman <horms@kernel.org>
cc: linux-afs@lists.infradead.org
cc: stable@kernel.org
Link: https://patch.msgid.link/20260624163819.3017002-5-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
rxrpc_input_ackall() accepts ACKALL packets without checking whether the
call is in a state that can legitimately have outstanding transmit buffers.
A forged ACKALL can therefore reach a new service call in
RXRPC_CALL_SERVER_RECV_REQUEST before any reply packets have been queued.
In that state call->tx_top is zero and call->tx_queue is NULL, so
rxrpc_rotate_tx_window() dereferences a NULL txqueue and triggers a
null-pointer dereference.
Fix the handling of ACKALL packets by the following means:
(1) Add two new call states: RXRPC_CALL_CLIENT_PRE_SEND which indicates
that the client call is connected, but nothing has been transmitted as
yet; and RXRPC_CALL_CLIENT_AWAIT_ACK, which indicates that everything
has been transmitted at least once, but we're now waiting for the
stuff remaining in the Tx buffer to be ACK'd (retransmissions may
still happen).
The RXRPC_CALL_CLIENT_PRE_SEND state is set when the call is assigned
a channel and transitions to RXRPC_CALL_CLIENT_SEND_REQUEST when the
first packet is transmitted.
RXRPC_CALL_CLIENT_AWAIT_REPLY is then narrowed in scope to indicate
that all Tx packets have been ACK'd and we're now waiting for the
reply to be received.
(2) As per Wyatt Feng's original patch[1], the ACKALL handler then checks
that the call state is one in which there might be stuff in the Tx
buffer to ACK, but now this includes AWAIT_ACK rather than
AWAIT_REPLY. ACKALL packets are ignored if received in the wrong
state.
Note that unlike Wyatt Feng's patch, it's no longer necessary to check
to see if the Tx buffer exists as this the state set now covers this.
(3) Make the ACKALL handler use call->tx_transmitted rather than
call->tx_top as the former is explicitly the highest packet seq number
transmitted, whereas the latter has a looser definition.
Thanks to Jeffrey Altman for a description of the history of the ACKALL
packet[1].
Fixes: b341a0263b ("rxrpc: Implement progressive transmission queue struct")
Reported-by: Yuan Tan <yuantan098@gmail.com>
Reported-by: Yifan Wu <yifanwucs@gmail.com>
Reported-by: Juefei Pu <tomapufckgml@gmail.com>
Reported-by: Zhengchuan Liang <zcliangcn@gmail.com>
Reported-by: Xin Liu <bird@lzu.edu.cn>
Signed-off-by: Wyatt Feng <bronzed_45_vested@icloud.com>
Co-developed-by: David Howells <dhowells@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Ren Wei <n05ec@lzu.edu.cn>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20260616155749.2125907-2-dhowells@redhat.com/ [1]
Link: https://lore.kernel.org/r/c0fd4fec-1576-4070-b31e-a37d5506f5ed@auristor.com/ [2]
Reviewed-by: Jeffrey Altman <jaltman@auristor.com>
Link: https://patch.msgid.link/20260624163819.3017002-2-dhowells@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jijie Shao says:
====================
net: hns3: fix configuration deadlocks and refactor link setup
This patch series addresses a sequence of link configuration deadlocks
and parameter contamination issues in the hns3 network driver, which
typically occur during hardware resets or driver initialization under
specific user-configured scenarios.
The bugs root from asynchronous discrepancies between the MAC state
machine and cached user requests during sudden hardware resets, leading
to invalid parameter combos or frozen registers.
====================
Link: https://patch.msgid.link/20260624141319.271439-1-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix a link loss issue during driver initialization on optical ports
connected to forced-mode (non-autoneg) remote switches.
Previously, during driver probe or initialization, hclge_configure()
blindly hardcoded hdev->hw.mac.req_autoneg to AUTONEG_ENABLE for all
media types. While this is necessary for copper (BASE-T) ports to
establish a link, many high-speed optical (fiber) ports in data
centers are connected to switches running in forced mode (fixed speed,
autoneg disabled). Forcing autoneg on these optical ports during
initialization causes a permanent link failure since the remote end
refuses to respond to autoneg pulses.
Fix this by implementing media-type differentiated initialization in
hclge_init_ae_dev(). Copper ports continue to default to
AUTONEG_ENABLE, while optical ports strictly inherit the preset
autoneg status pre-configured by the firmware (hdev->hw.mac.autoneg),
preserving native compatibility with forced-mode network environments.
Fixes: 05eb60e964 ("net: hns3: using user configure after hardware reset")
Signed-off-by: Shuaisong Yang <yangshuaisong@h-partners.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://patch.msgid.link/20260624141319.271439-5-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix a critical race condition deadlock where the network interface
remains permanently Link Down after a hardware reset under specific
ethtool sequences.
This issue exclusively manifests in firmware-controlled PHY topologies
where the driver relies on the IMP firmware to arbitrate link parameters.
Standard devices driven by the kernel's native PHY_LIB are unaffected.
The deadlock occurs via the following path:
1. User disables autoneg and forces an unmatched speed, forcing link
down: `ethtool -s ethx autoneg off speed 10 duplex full`
2. User re-enables autoneg: `ethtool -s ethx autoneg on`. The netdev
stack passes cmd->base.speed as SPEED_UNKNOWN (0xffffffff).
3. Driver saves req_autoneg=1, but before the interface can link up,
a hardware reset is triggered.
4. During reset recovery, MAC init reads the un-synchronized runtime
state mac.autoneg (which is still 0/OFF), misinterprets it as
forced mode, and pushes the cached SPEED_UNKNOWN into the hardware
registers, causing the MAC firmware state machine to freeze.
Meanwhile, PHY init reads req_autoneg=1 and enables PHY autoneg.
Since the MAC is frozen with 0xffffffff and PHY is running autoneg,
they mismatch permanently.
Fix this by:
1. Intercepting SPEED_UNKNOWN/DUPLEX_UNKNOWN in
hclge_set_phy_link_ksettings() and hclge_cfg_mac_speed_dup_h() to
prevent it from corrupting the driver's cached valid configuration.
2. Save req_autoneg in hclge_set_autoneg().
3. Aligning the state judgment in hclge_set_autoneg_speed_dup() to use
req_autoneg instead of the un-synchronized runtime mac.autoneg,
ensuring both MAC and PHY consistently enter the autoneg branch to
eliminate configuration discrepancies during reset recovery.
Fixes: 05eb60e964 ("net: hns3: using user configure after hardware reset")
Signed-off-by: Shuaisong Yang <yangshuaisong@h-partners.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://patch.msgid.link/20260624141319.271439-4-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Extract the MAC autoneg and speed/duplex/lane configuration logic out
of hclge_mac_init() and encapsulate it into a new dedicated helper
function hclge_set_autoneg_speed_dup().
In the init path (hclge_init_ae_dev), this helper is now called after
hclge_update_port_info() so that firmware-reported autoneg values are
already populated before applying the link configuration.
Introduce a separate req_lane_num field in struct hclge_mac to isolate
the user-requested lane count from mac.lane_num, which firmware may
overwrite via hclge_get_sfp_info() with stale values from a prior link
lifecycle (e.g., lane_num=4 from 100G). During probe, req_lane_num is
initialized to 0, which instructs firmware to auto-select the correct
lane count for the current speed, rather than reusing the firmware-
reported mac.lane_num that may be inconsistent with the target speed.
This prevents probe failures from mismatched (speed, lane_num) pairs.
In the reset path (hclge_reset_ae_dev), it runs immediately after
hclge_mac_init(), using the previously cached req_* values to restore
the link without re-querying firmware.
Signed-off-by: Shuaisong Yang <yangshuaisong@h-partners.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://patch.msgid.link/20260624141319.271439-3-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Refactor hns3_set_link_ksettings() and hclge_set_phy_link_ksettings()
to unify the configuration path for copper ports.
Previously, netdevs with a native kernel phy attached bypassed the main
MAC parameter caching logic and returned early via
phy_ethtool_ksettings_set(). This prevented the driver from updating
hdev->hw.mac.req_xxx variables for kernel PHY setups, leaving them
out-of-sync during reset recovery.
Clean this up by routing all copper port configurations through
ops->set_phy_link_ksettings(), and perform driver-level or kernel-level
PHY arbitration inside hclge_set_phy_link_ksettings() via
hnae3_dev_phy_imp_supported(). This ensures that the user's intended link
profiles (req_speed, req_duplex, req_autoneg) are uniformly recorded
across all copper and fiber deployment topologies, laying the groundwork
for stable reset recovery.
For copper ports where neither IMP firmware nor a kernel PHY is available
(e.g. PHY_INEXISTENT), hclge_set_phy_link_ksettings() returns -ENODEV.
In hns3_set_link_ksettings(), this is caught so the configuration falls
through to the existing MAC-level path (check_ksettings_param ->
cfg_mac_speed_dup_h), preserving compatibility with PHY-less copper
deployments.
Signed-off-by: Shuaisong Yang <yangshuaisong@h-partners.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://patch.msgid.link/20260624141319.271439-2-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Before the commit 7553911210 ("net: mana: Allocate MSI-X vectors
dynamically"), all the MANA IRQs were assigned statically and together
during early driver load.
After this commit, the IRQ allocation for MANA was done in two phases.
HWC IRQ allocated earlier and then, queue IRQs dynamically added at a
later point. By this time, the IRQ weights on vCPUs can become imbalanced
and if IRQ count is greater than the vCPU count the topology aware IRQ
distribution logic in MANA can cause multiple MANA IRQs to land on the
same vCPUs, while other sibling vCPUs have none (case 1).
On SMP enabled, low-vCPU systems, this becomes a bigger problem as the
softIRQ handling overhead of two IRQs on the same vCPUs becomes much more
than their overheads if they were spread across sibling vCPUs.
In such cases when many parallel TCP connections are tested, the
throughput drops significantly.
Fix the affinity assignment logic, in cases where the IRQ count is greater
than the vCPU count and when IRQs are added dynamically, by utilizing all
the vCPUs irrespective of their NUMA/core bindings (case 2).
The results of setting the affinity and hint to NULL were also studied,
and we observed that, with this logic if there are pre-existing IRQs
allocated on the VM (apart from MANA), during MANA IRQs allocation, it
leads to clustering of the MANA queue IRQs again (case 3).
=======================================================
Case 1: without this patch
=======================================================
4 vcpu(2 cores), 5 MANA IRQs (1 HWC + 4 Queue)
TYPE effective vCPU aff
=======================================================
IRQ0: HWC 0
IRQ1: mana_q1 0
IRQ2: mana_q2 2
IRQ3: mana_q3 0
IRQ4: mana_q4 3
%soft on each vCPU(mpstat -P ALL 1) on receiver
vCPU 0 1 2 3
=======================================================
pass 1: 38.85 0.03 24.89 24.65
pass 2: 39.15 0.03 24.57 25.28
pass 3: 40.36 0.03 23.20 23.17
=======================================================
Case 2: with this patch
=======================================================
4 vcpu(2 cores), 5 MANA IRQs (1 HWC + 4 Queue)
TYPE effective vCPU aff
=======================================================
IRQ0: HWC 0
IRQ1: mana_q1 0
IRQ2: mana_q2 1
IRQ3: mana_q3 2
IRQ4: mana_q4 3
%soft on each vCPU(mpstat -P ALL 1) on receiver
vCPU 0 1 2 3
=======================================================
pass 1: 15.42 15.85 14.99 14.51
pass 2: 15.53 15.94 15.81 15.93
pass 3: 16.41 16.35 16.40 16.36
=======================================================
Case 3: with affinity set to NULL
=======================================================
4 vCPU(2 cores), 5 MANA IRQs (1 HWC + 4 Queue)
TYPE effective vCPU aff
=======================================================
IRQ0: HWC 0
IRQ1: mana_q1 2
IRQ2: mana_q2 3
IRQ3: mana_q3 2
IRQ4: mana_q4 3
=======================================================
Throughput Impact(in Gbps, same env)
=======================================================
TCP conn with patch w/o patch aff NULL
20480 15.65 7.73 5.25
10240 15.63 8.93 5.77
8192 15.64 9.69 7.16
6144 15.64 13.16 9.33
4096 15.69 15.75 13.50
2048 15.69 15.83 13.61
1024 15.71 15.28 13.60
Fixes: 7553911210 ("net: mana: Allocate MSI-X vectors dynamically")
Cc: stable@vger.kernel.org
Co-developed-by: Erni Sri Satya Vennela <ernis@linux.microsoft.com>
Signed-off-by: Erni Sri Satya Vennela <ernis@linux.microsoft.com>
Signed-off-by: Shradha Gupta <shradhagupta@linux.microsoft.com>
Reviewed-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Yury Norov <ynorov@nvidia.com>
Link: https://patch.msgid.link/20260624072138.1632849-1-shradhagupta@linux.microsoft.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
splice_short grows its pipe with (MAX_FRAGS + 1) * 0x1000 so it can
queue one short vmsplice() buffer for each fragment before draining the
pipe. That assumes 4K pipe buffers.
On 64K-page kernels the request is rounded to 262144 bytes, which
provides only four pipe buffers. The fifth one-byte vmsplice() blocks in
pipe_wait_writable and the test times out before it reaches the TLS path.
Request enough bytes for the same number of pipe buffers using the
runtime page size, and assert that the kernel granted at least that much.
If an unprivileged run cannot raise the pipe above the system
pipe-max-size limit, skip the test because it cannot exercise the
intended path.
Fixes: 3667e9b442 ("selftests: tls: add test for short splice due to full skmsg")
Signed-off-by: Nirmoy Das <nirmoyd@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260624134416.3235403-1-nirmoyd@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
sparx5_register_notifier_blocks() registers the switchdev blocking
notifier before allocating the ordered workqueue. If the workqueue
allocation fails, the error path unregisters the switchdev and netdevice
notifiers, but leaves the blocking notifier registered.
Add a separate error label for the workqueue allocation failure path and
unregister the switchdev blocking notifier there.
Fixes: d6fce51419 ("net: sparx5: add switching support")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260623115714.2192074-1-haoxiang_li2024@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
tipc: syzbot related fixes
First patch fixes a recent syzbot report.
Second patch is inspired by numerous syzbot soft lockup
reports with RTNL pressure.
====================
Link: https://patch.msgid.link/20260623173030.2925059-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Blamed commit introduced a busy-wait loop in tipc_exit_net()
to wait for pending UDP bearer cleanup works to complete:
while (atomic_read(&tn->wq_count))
cond_resched();
This loop can busy-wait for a long time if cond_resched() is a NOP. This
typically happens if the netns exit is executed by a high priority task,
or under kernels configured without preemption (CONFIG_PREEMPT_NONE). In
such cases, it wastes CPU cycles and can lead to soft lockups.
Fix this by replacing the busy loop with wait_var_event(), allowing the
thread to sleep properly until the work queue count reaches zero.
Accordingly, update cleanup_bearer() to use atomic_dec_and_test() and
wake_up_var() to wake up the waiter when the count drops to zero.
This uses the global wait queue hash table, avoiding the need to bloat
struct tipc_net with a wait_queue_head_t. The atomic_dec_and_test()
provides the necessary memory barrier to ensure the wakeup is not missed.
Fixes: 04c26faa51 ("tipc: wait and exit until all work queues are done")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jon Maloy <jmaloy@redhat.com>
Cc: tipc-discussion@lists.sourceforge.net
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260623173030.2925059-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
TIPC UDP media bearer teardown calls dst_cache_destroy() on its
replicast caches before calling synchronize_net() to wait for
concurrent RCU readers (transmitters) to finish:
static void cleanup_bearer(struct work_struct *work)
{
...
list_for_each_entry_safe(rcast, tmp, &ub->rcast.list, list) {
dst_cache_destroy(&rcast->dst_cache);
list_del_rcu(&rcast->list);
kfree_rcu(rcast, rcu);
}
...
dst_cache_destroy(&ub->rcast.dst_cache);
udp_tunnel_sock_release(ub->sk);
synchronize_net();
...
}
This is highly buggy because dst_cache_destroy() immediately frees the
per-CPU cache memory (free_percpu()) and releases the cached dst
entries without any synchronization.
If a concurrent transmitter (e.g., tipc_udp_xmit()) is running on another
CPU under RCU protection, it can call dst_cache_get() concurrently,
leading to:
1. Use-After-Free on the per-CPU cache pointer itself (crash).
2. "rcuref - imbalanced put()" warning if it attempts to release a
dst that was concurrently released by dst_cache_destroy().
Furthermore, calling kfree(ub) immediately after synchronize_net() without
closing the socket first (or waiting after closing it) leaves a window
where a concurrent receiver (tipc_udp_recv()) could start after
synchronize_net(), access ub, and suffer a UAF when kfree(ub) runs.
To fix this, we must defer dst_cache_destroy() and kfree(ub) until after
we have ensured that no more readers can see the bearer/socket and all
existing readers have finished:
1. Defer rcast entry destruction (both dst_cache_destroy() and kfree())
to an RCU callback using call_rcu_hurry().
Using call_rcu_hurry() ensures the dst entries are released quickly.
2. Release the bearer socket using udp_tunnel_sock_release() (stops
new receive readers).
3. Call synchronize_net() to wait for all outstanding RCU readers
(both transmit and receive) to finish.
4. Now that it is safe, call dst_cache_destroy() on the main bearer
cache, and free ub.
Note: 3) and 4) can be changed later in net-next to also use
call_rcu_hurry() and get rid of the synchronize_net() latency.
Fixes: e9c1a79321 ("tipc: add dst_cache support for udp media")
Reported-by: syzbot+e14bc5d4942756023b77@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6a396a66.52ae72c2.136ac7.0003.GAE@google.com/T/#u
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jon Maloy <jmaloy@redhat.com>
Cc: tipc-discussion@lists.sourceforge.net
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Link: https://patch.msgid.link/20260623173030.2925059-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
nix_setup_bpids() allocates bp->bpids with rvu_alloc_bitmap(), which uses
a plain kcalloc(). If any of the following devm_kcalloc() allocations for
the BPID mapping arrays fails, the function returns without freeing the
bitmap. Free the BPID bitmap before returning from those error paths.
Fixes: d6212d2e41 ("octeontx2-af: Create BPIDs free pool")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260623114316.2182271-1-haoxiang_li2024@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Emails to the maintainer of Qualcomm PPE Ethernet driver (Luo Jie
<quic_luoj@quicinc.com>) bounce permanently (full mailbox), because the
"quicinc.com" addresses were deprecated for public work. All Qualcomm
contributors are aware of that and were asked to fix their addresses.
Driver is not supported - in terms of how netdev understands supported
commitment - if maintainer does not care to receive the patches for its
code, so demote it to "maintained" to reflect true status.
Fix all occurences of Luo Jie email address to preferred and working
domain.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Acked-by: Luo Jie <jie.luo@oss.qualcomm.com>
Link: https://patch.msgid.link/20260623073307.36483-2-krzysztof.kozlowski@oss.qualcomm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For i.MX94 series, all the standalone ENETCs do not support SR-IOV, so
pf->caps.num_vsi is zero. This leads to a divide-by-zero in
enetc4_default_rings_allocation() when distributing rings among PF and
VFs.
Division by zero is undefined behavior in C. On ARM64, the UDIV/SDIV
instructions silently return zero rather than raising an exception, so
the issue does not cause a visible crash. However, relying on this
behavior is incorrect and poses a cross-platform compatibility risk.
Add an explicit check for num_vsi == 0 and return early after the PF's
rings have been configured.
Fixes: 2d673b0e2f ("net: enetc: add standalone ENETC support for i.MX94")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Link: https://patch.msgid.link/20260624072726.1238903-1-wei.fang@oss.nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Fix the Micrel PHY in the example which shouldn't have the
fallback "ethernet-phy-ieee802.3-c22" compatible:
Documentation/devicetree/bindings/net/renesas,ether.example.dtb: ethernet-phy@1 \
(ethernet-phy-id0022.1537): compatible: ['ethernet-phy-id0022.1537', 'ethernet-phy-ieee802.3-c22'] is too long
from schema $id: http://devicetree.org/schemas/net/micrel.yaml
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Acked-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Fixes: 37a2fce090 ("dt-bindings: sh_eth convert bindings to json-schema")
Link: https://patch.msgid.link/20260624150250.131966-2-robh@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ct_limit_set() is documented as being called with ovs_mutex held. It
walks the ct limit hlist with hlist_for_each_entry_rcu(), but the
iterator does not currently pass the OVS lockdep condition used
elsewhere for RCU-protected OVS objects.
Pass lockdep_ovsl_is_held() to the iterator. This matches the function's
existing caller contract and lets CONFIG_PROVE_RCU_LIST distinguish the
ovs_mutex-protected update path from the RCU read-side ct_limit_get()
path.
This was found by our static analysis tool and then manually reviewed
against the current tree. In the reviewed CONFIG_PROVE_RCU_LIST triage
run, the writer-side ct limit update produced the expected "RCU-list
traversed in non-reader section!!" warning while ovs_mutex was held,
with the stack matching ct_limit_set() and ovs_ct_limit_set_zone_limit().
The change is limited to documenting the existing protection contract.
This is a lockdep annotation cleanup. It does not change the conntrack
limit list update or release behavior.
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Reviewed-by: Eelco Chaudron <echaudro@redhat.com>
Link: https://patch.msgid.link/20260624150149.3510541-1-runyu.xiao@seu.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Yue Sun reported a use-after-free and debugobjects warning in
udp_tunnel_nic_device_sync_work() during concurrent device operations.
The workqueue core clears the internal pending bit before invoking the
worker. At that point, a concurrent thread can queue the work again.
When the already running worker eventually clears the work_pending flag
to 0, it mistakenly clears the flag for the newly queued instance.
udp_tunnel_nic_unregister() then observes work_pending as 0 and frees
the structure while the second work item is still active in the queue,
leading to UAF.
Fix this by returning early in udp_tunnel_nic_device_sync() if
work_pending is already set, preventing redundant work queueing.
Fixes: cc4e3835ef ("udp_tunnel: add central NIC RX port offload infrastructure")
Reported-by: Yue Sun <samsun1006219@gmail.com>
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260625065938.654652-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=BXU0
-----END PGP SIGNATURE-----
Merge tag 'nf-26-06-23' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net:
1) Add a workaround to avoid a possible crash if nf_nat and nft_chain_nat are
compiled built-in and nf_nat fails to register, allowing nft_chain_nat to
access the incorrect pernetns area. This is crash specific of all built-in
compilation. From Matias Krause.
2) Revisit conncount GC optimization for confirmed conntracks, skip GC round
if IPS_ASSURED is set on. This is addressing an issue for corner case
use case scenario involving locally generated traffic. No crash, just a
functionality fix. From Fernando F. Mancera.
3) Validate iph->ihl in flowtable IPIP tunnel support, from Lorenzo Bianconi.
This a sanity check to bounces back malformed IPIP packets to classic
forwarding path.
4) Kdoc fixes for x_tables.h, from Randy Dunlap.
5) Use info->options so nft_synproxy_tcp_options() stays on the same local
snapshot, otherwise eval path can observe inconsistent mix of mss and
timestamps. From Runyu Xiao.
6) Add conntrack_sctp_collision.sh to cover for SCTP INIT collisions.
From Yi Chen.
7) Do not allow NFPROTO_UNSPEC targets if family is NFPROTO_BRIDGE in
nft_compat. This allows to use non-sense targets such as xt_nat leading
to crash. From Florian Westphal.
8) Add a selftest queueing from bridge family. From Florian Westphal.
9) Do not allow to reset a conntrack helper via ctnetlink. This feature
antedates the creation of the conntrack-tools, and it is not used
I don't have a usecase for it, I prefer to remove than fixing it.
10) Add deprecation warning for IPv4 only conntrack helpers for PPTP
and IRC. From Florian Westphal.
11) Store the master tuple in the expectation object and use it,
otherwise SLAB_TYPESAFE_RCU rules allow to display incorrect
master tuple information through ctnetlink.
12) Run expectation eviction when inserting an expectation with no
helper, this is a fix for the nft_ct custom expectation support.
13) Fix nft_ct custom expectation timeouts, userspace provides a
timeout in milliseconds but kernel assumes this comes in seconds.
From Florian Westphal.
14) Cap maximum number of expectations per class to 255 expectations
per master conntrack at helper registration. This is a fix to
restrict the maximum number of expectations per master conntrack
which can be a issue for the new lazy GC expectation approach.
* tag 'nf-26-06-23' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: nf_conntrack_helper: cap maximum number of expectation at helper registration
netfilter: nft_ct: expectation timeouts are passed in milliseconds
netfilter: nf_conntrack_expect: run expectation eviction with no helper
netfilter: nf_conntrack_expect: store master_tuple in expectation
netfilter: conntrack: add deprecation warnings for irc and pptp trackers
netfilter: ctnetlink: do not allow to reset helper on existing conntrack
selftests: nft_queue.sh: add a bridge queue test
netfilter: nft_compat: ebtables emulation must reject non-bridge targets
selftests: netfilter: conntrack_sctp_collision.sh: Introduce SCTP INIT collision test
netfilter: nft_synproxy: stop bypassing the priv->info snapshot
netfilter: x_tables.h: fix all kernel-doc warnings
netfilter: flowtable: Validate iph->ihl in nf_flow_ip4_tunnel_proto()
netfilter: nf_conncount: prevent connlimit drops for early confirmed ct
netfilter: nf_nat: avoid invalid nat_net pointer use on failed nf_nat_init()
====================
Link: https://patch.msgid.link/20260623221548.701545-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2026-06-22 (ice, i40e, e1000e)
For ice:
Dawid changes call to release control VSI during reset to prevent
leaking it.
Lukasz fixes flow control error check to check value rather than treat
is as bitmap values.
Paul makes link related errors non-fatal to probe to allow for recovery
in certain NVM update situations.
Marcin moves netif_keep_dst() to only be called once when entering
switchdev mode.
ZhaoJinming adds a cleanup path for ice_dpll_init_info() to prevent
memory leaks on error path.
For i40e:
Mohamed Khalfella corrects argument passed in macro to match the
one provided to the macro.
For e1000e:
Dima resolves power state issues by adjusting value of PLL clock gate
and re-enabling K1; a quirk table is added to keep it off for known bad
systems.
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
e1000e: Reconfigure PLL clock gate timeout and re-enable K1 on Meteor Lake
i40e: Fix i40e_debug() to use struct i40e_hw argument
ice: dpll: fix memory leak in ice_dpll_init_info error paths
ice: dpll: set pointers to NULL after kfree in ice_dpll_deinit_info
ice: call netif_keep_dst() once when entering switchdev mode
ice: fix ice_init_link() error return preventing probe
ice: fix AQ error code comparison in ice_set_pauseparam()
ice: fix FDIR CTRL VSI resource leak in ice_reset_all_vfs()
====================
Link: https://patch.msgid.link/20260622220059.2471844-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The current irq definition of the wake irq and the lpi irq
is wrong, replace them with the right number and name.
Fixes: 30f0ba420e ("net: stmmac: Add glue layer for Spacemit K3 SoC")
Signed-off-by: Inochi Amaoto <inochiama@gmail.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Link: https://patch.msgid.link/20260623074637.503864-3-inochiama@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The current MII interface register definition from the vendor is wrong,
use the right number for the macro. Also, correct the interface mask
in spacemit_set_phy_intf_sel() so it can update the register with the
right number
Fixes: 30f0ba420e ("net: stmmac: Add glue layer for Spacemit K3 SoC")
Signed-off-by: Inochi Amaoto <inochiama@gmail.com>
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Link: https://patch.msgid.link/20260623074637.503864-2-inochiama@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
mac->phy_node is acquired via of_parse_phandle() in spl2sw_probe() and
stored in the mac private data, transferring ownership of the
device_node reference to mac. On driver removal, spl2sw_phy_remove()
disconnects the PHY but never drops that reference, so each
probe-then-remove cycle leaks one of_node refcount per port permanently.
Drop the reference after phy_disconnect(). While at it, remove the
redundant inner "if (ndev)" check; comm->ndev[i] was just verified
non-NULL on the line above.
Compile-tested only; no SP7021 hardware available.
Fixes: fd3040b939 ("net: ethernet: Add driver for Sunplus SP7021")
Signed-off-by: Shitalkumar Gandhi <shitalkumar.gandhi@cambiumnetworks.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/f3bdd4c91f3e2269b4e256075f9dc70808b1b8e9.1782195965.git.shitalkumar.gandhi@cambiumnetworks.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
drm_ras includes drm/drm_ras.h, which is a relatively new header not yet
shipped in most distro kernel-header packages. Without the explicit
entry, the build might fail with a message like this:
drm_ras-user.c:19:10: error: ‘DRM_RAS_CMD_CLEAR_ERROR_COUNTER’ \
undeclared here (not in a function); did you mean \
‘DRM_RAS_CMD_GET_ERROR_COUNTER’
Signed-off-by: Thorsten Leemhuis <linux@leemhuis.info>
Link: https://patch.msgid.link/20260623070818.2161810-1-linux@leemhuis.info
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
gem_init_one() calls gem_remove_one() when register_netdev() fails.
gem_remove_one() unregisters and frees resources owned by the net_device,
including the DMA block, MMIO mapping, PCI regions, and the net_device
itself. gem_init_one() then falls through to its own cleanup labels and
frees the same resources again.
Keep the register_netdev() error path in gem_init_one(): clear drvdata so
PM/remove paths do not see a half-registered device, remove the NAPI
instance added during probe, and let the existing cleanup labels release
the resources once.
The issue was found by a local static-analysis checker for probe error
paths. The reported path was manually inspected before sending this fix.
Compile-tested with CONFIG_SUNGEM=y. Runtime testing was not performed
because no sungem hardware is available.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Ruoyu Wang <ruoyuw560@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260623025759.3468566-1-ruoyuw560@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In tcp_ao_delete_key(), the del_async path skips the current_key
and rnext_key validity checks present in the synchronous path,
assuming these pointers are always NULL on LISTEN sockets. However,
if a key was added with set_current=1/set_rnext=1 while the socket
was in CLOSE state, current_key and rnext_key will be non-NULL
after listen() transitions the socket to LISTEN.
When such a key is deleted with del_async=1, hlist_del_rcu() and
call_rcu() free the key without clearing the dangling pointers.
After the RCU grace period, getsockopt(TCP_AO_INFO) dereferences
current_key->sndid and rnext_key->rcvid from freed slab memory.
Clear current_key and rnext_key in the del_async path when they
reference the key being deleted.
Fixes: d6732b95b6 ("net/tcp: Allow asynchronous delete for TCP-AO keys (MKTs)")
Signed-off-by: HanQuan <eilaimemedsnaimel@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260623015208.1191687-1-eilaimemedsnaimel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use $(AR) to allow build system to override the archiver tool (e.g.,
when cross-compiling for a different architecture) by setting the AR
environment variable.
GNU Make defaults AR to ar, so this change will not break existing build
environments that do not explicitly set AR.
Fixes: 07c3cc51a0 ("tools: net: package libynl for use in selftests")
Fixes: 86878f14d7 ("tools: ynl: user space helpers")
Signed-off-by: Greg Thelen <gthelen@google.com>
Link: https://patch.msgid.link/20260622161659.145047-1-gthelen@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Configurations with mlx5 built-in but macsec=m fail to link:
x86_64-linux-ld: drivers/infiniband/hw/mlx5/macsec.o: in function `mlx5r_add_gid_macsec_operations':
macsec.c:(.text+0x77d): undefined reference to `macsec_netdev_is_offloaded'
x86_64-linux-ld: drivers/infiniband/hw/mlx5/macsec.o: in function `mlx5r_del_gid_macsec_operations':
macsec.c:(.text+0xe81): undefined reference to `macsec_netdev_is_offloaded'
Fix the dependency so this configuration cannot happen.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Link: https://patch.msgid.link/20260622124229.2444502-1-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
kalmia_rx_fixup() computes usb_packet_length = skb->len - (2 *
KALMIA_HEADER_LENGTH) as a u16, guarded only by a pre-loop check that
skb->len is at least KALMIA_HEADER_LENGTH, which is 6. A device can
deliver a short bulk-IN frame with skb->len in the 6 to 11 range, or
leave a short trailing remainder on a later loop iteration. Either case
underflows usb_packet_length to about 65530.
That bypasses the usb_packet_length < ether_packet_length truncation path.
The device-supplied ether_packet_length, a le16 up to 65535 read from
header_start[2], then drives a memcmp() and the following skb_trim() and
skb_pull() past the end of the rx buffer. The rx buffer is hard_mtu * 10,
which is 14000 bytes. That is an out of bounds read.
Require both the start and end framing headers to be present before
subtracting them, on every loop iteration.
Fixes: d40261236e ("net/usb: Add Samsung Kalmia driver for Samsung GT-B3730")
Cc: stable@vger.kernel.org
Signed-off-by: Maoyi Xie <maoyixie.tju@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://patch.msgid.link/178211531778.2216480.12637613349790980750@maoyixie.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Even with both paths gated on gs->gro_hint, geneve_gro_complete()
re-derives the inner dispatch type and length from the packet and the
current gs->gro_hint, independently of geneve_gro_receive(). The two can
disagree if gs->gro_hint flips under a concurrent geneve_quiesce()/
geneve_unquiesce() (sk_user_data is NULL across a synchronize_net()), or if
the re-read option bytes differ from the ones receive parsed.
geneve_gro_receive() already records the inner network header position in
NAPI_GRO_CB()->inner_network_offset. Have geneve_gro_complete() compute the
offset it is about to dispatch at, adding ETH_HLEN in the ETH_P_TEB case
where eth_gro_complete() steps over the inner MAC header, and bail out if
it lands past inner_network_offset.
Use a lower bound rather than exact equality: between gh_len and the inner
L3 header, geneve_gro_receive() may also have pulled an inner VLAN tag
(vlan_gro_receive() advances the recorded offset past it), which only moves
inner_network_offset further out. A valid frame therefore always satisfies
inner_nh <= inner_network_offset, while a gh_len inflated by a hint
gro_receive() did not honour dispatches past the validated inner header,
i.e. the out-of-bounds completion. Only the latter is rejected.
Fixes: fd0dd79657 ("geneve: use GRO hint option in the RX path")
Suggested-by: Paolo Abeni <pabeni@redhat.com>
Co-developed-by: Weiming Shi <bestswngs@gmail.com>
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/20260618032622.484720-2-xmei5@asu.edu
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
geneve_gro_receive() reads the GRO hint through geneve_sk_gro_hint_off(),
which honours it only when the socket enabled IFLA_GENEVE_GRO_HINT
(gs->gro_hint). geneve_gro_complete() instead calls the low-level
geneve_opt_gro_hint_off() and acts on the hint unconditionally.
On a tunnel without the hint, receive aggregates the frames as plain
ETH_P_TEB while complete still honours an attacker-supplied hint option: it
inflates gh_len by gro_hint->nested_hdr_len (u8) and redirects the dispatch
type, so the inner gro_complete handler runs at nhoff + gh_len, an offset
receive never pulled nor validated, reading out of bounds of the skb head:
BUG: KASAN: slab-out-of-bounds in ipv6_gro_complete (net/ipv6/ip6_offload.c:196)
Read of size 1 at addr ffff88800fe91980 by task exploit/153
ipv6_gro_complete (net/ipv6/ip6_offload.c:196)
geneve_gro_complete (drivers/net/geneve.c:965)
udp_gro_complete (net/ipv4/udp_offload.c:940)
inet_gro_complete (net/ipv4/af_inet.c:1621)
__gro_flush (net/core/gro.c:306)
Gate the complete path on gs->gro_hint too via geneve_sk_gro_hint_off(), so
both paths agree. Tunnels that enable the hint are unaffected.
Fixes: fd0dd79657 ("geneve: use GRO hint option in the RX path")
Reported-by: Weiming Shi <bestswngs@gmail.com>
Reported-by: Kyle Zeng <kylebot@openai.com>
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Link: https://patch.msgid.link/20260618032622.484720-1-xmei5@asu.edu
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On Marvell MPIC platforms (Armada 370/XP/38x), mvneta uses a percpu
IRQ disable/enable scheme for NAPI: the ISR (mvneta_percpu_isr) calls
disable_percpu_irq() to mask the MPIC per-CPU interrupt and schedules
NAPI poll, which calls enable_percpu_irq() on completion to unmask.
If suspend occurs while NAPI poll is pending (between
disable_percpu_irq in the ISR and enable_percpu_irq in poll
completion), the interrupt is never re-enabled:
1. mvneta_percpu_isr: disable_percpu_irq() + napi_schedule()
=> MPIC masked, percpu_enabled cpumask bit cleared
2. NAPI poll does not complete before suspend proceeds
(on PREEMPT_RT this is highly likely since softirqs run in
ksoftirqd which gets frozen; on non-RT it can happen when
softirq processing is deferred to ksoftirqd)
3. mvneta_stop_dev => napi_disable(): cancels the pending poll
without executing the completion path
4. suspend_device_irqs => IRQCHIP_MASK_ON_SUSPEND: masks MPIC
(already masked, but records IRQS_SUSPENDED)
5. Resume: mpic_resume checks irq_percpu_is_enabled() => false
(bit was cleared in step 1) => skips unmask
6. mvneta_start_dev only restores device-level INTR_NEW_MASK,
does not touch the MPIC per-CPU mask
Result: MPIC per-CPU interrupt stays masked permanently. The NIC
generates interrupts (INTR_NEW_CAUSE != 0) but the CPU never
receives them, causing complete loss of network connectivity.
Fix by calling on_each_cpu(mvneta_percpu_enable) in the resume path
to unconditionally unmask the MPIC per-CPU interrupt regardless of
pre-suspend state.
Fixes: 12bb03b436 ("net: mvneta: Handle per-cpu interrupts")
Signed-off-by: Yun Zhou <yun.zhou@windriver.com>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Link: https://patch.msgid.link/20260622074350.1666290-1-yun.zhou@windriver.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
CGX per-lmac debugfs seq readers obtained struct rvu via
pci_get_drvdata(pci_get_device(..., PCI_DEVID_OCTEONTX2_RVU_AF, ...)),
which leaks a PCI device reference on every read. Store rvu and the CGX
handle in debugfs inode private data when creating stats, mac_filter,
and fwdata files (one context per CGX), and use debugfs aux numbers for
fwdata so lmac_id matches the other CGX debugfs entries.
Fixes: f967488d09 ("octeontx2-af: Add per CGX port level NIX Rx/Tx counters")
Fixes: dbc52debf9 ("octeontx2-af: Debugfs support for DMAC filters")
Fixes: 49f02e6877 ("Octeontx2-af: Debugfs support for firmware data")
Cc: Linu Cherian <lcherian@marvell.com>
Reported-by: Yuho Choi <dbgh9129@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com>
Link: https://patch.msgid.link/20260622034229.2254145-1-rkannoth@marvell.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
NIX maximum number of LFs can be set via devlink command
but that can be done before assigning any LFs to a PF/VF.
The condition used to check whether any LFs are assigned is
incorrect. This patch fixes that condition.
Fixes: dd78428786 ("octeontx2-af: Add new devlink param to configure maximum usable NIX block LFs")
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Link: https://patch.msgid.link/1782082853-6941-1-git-send-email-sbhatta@marvell.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
t7xx_cldma_late_init() creates md_ctrl->gpd_dmapool before
initializing the TX and RX rings. If any ring initialization
fails, the error path frees the already initialized rings but
leaves the DMA pool allocated.
Destroy md_ctrl->gpd_dmapool on the late-init failure path
to avoid leaking the DMA pool.
Fixes: 39d439047f ("net: wwan: t7xx: Add control DMA interface")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
Reviewed-by: Loic Poulain <loic.poulain@oss.qualcomm.com>
Link: https://patch.msgid.link/20260621031714.3605022-1-haoxiang_li2024@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When multiple netdevs share a QDMA TX ring and one device is stopped,
netdev_tx_reset_subqueue() zeroes that device's BQL counters while its
pending skbs remain in the shared HW TX ring. When NAPI later completes
those skbs via netdev_tx_completed_queue(), the already-zeroed
dql->num_queued counter underflows.
Fix the issue:
- Remove netdev_tx_reset_subqueue() from airoha_dev_stop() so pending
skbs are completed naturally by NAPI with proper BQL accounting.
- Rework airoha_qdma_tx_cleanup() to disable TX DMA, flush BQL
counters, DMA-unmap and free all pending skbs while skb->dev
references are still valid. Use a per-queue flushing flag checked
under q->lock in airoha_dev_xmit() to prevent races between teardown
and transmit. Call airoha_qdma_stop_napi() before
airoha_qdma_tx_cleanup() at the call sites.
- Move DMA engine start into probe. Split DMA teardown so TX DMA is
disabled in airoha_qdma_tx_cleanup() and RX DMA in
airoha_qdma_cleanup().
- Remove qdma->users counter since DMA lifetime is now tied to
probe/cleanup rather than per-netdev open/stop.
Fixes: a9c2ca61fe ("net: airoha: Support multiple net_devices for a single FE GDM port")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260620-airoha-bql-fixes-v3-1-76b95374e63e@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On RTL8127A connected to a link partner that advertises 10000baseT
speed cannot be changed to anything other than 10000baseT as 10GbE
is always advertised regardless of any setting. Fix this by
clearing MDIO_AN_10GBT_CTRL_ADV10G bit in rtl822x_config_aneg()'s
call to phy_modify_mmd_changed().
Fixes: 83d9623161 ("net: phy: realtek: add RTL8127-internal PHY")
Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com>
Signed-off-by: Jan Klos <honza.klos@gmail.com>
Link: https://patch.msgid.link/20260620011956.37181-1-honza.klos@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
subbank_srch_order[i] is the physical subbank at search-order slot i,
so each subbank's arr_idx must be i (its slot), not
subbank_srch_order[sb->idx]. The old logic mis-keyed xa_sb_free
and broke allocation traversal order.
Populate arr_idx and xa_sb_free in a single pass over the search
order after subbank structs are initialized.
Fixes: 7ac9d4c407 ("octeontx2-af: npc: cn20k: add subbank search order control")
Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260619095100.1864440-1-rkannoth@marvell.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit d7709812e1 ("net: mana: hardening: Validate adapter_mtu from
MANA_QUERY_DEV_CONFIG") rejected any adapter_mtu value smaller than
ETH_MIN_MTU + ETH_HLEN, including 0, returning -EPROTO and failing
mana_probe().
Some older PF firmware versions still in the field report
adapter_mtu as 0 in the MANA_QUERY_DEV_CONFIG response. With the
hardening check in place, the MANA VF driver now fails to load on
those hosts, breaking networking entirely for guests.
MANA hardware always supports the standard Ethernet MTU. Treat a
reported adapter_mtu of 0 as "the PF did not advertise a value" and
fall back to ETH_FRAME_LEN, the same value used for the pre-V2
message version path. Only jumbo frames remain unavailable until
the PF reports a valid MTU.
Other small-but-nonzero bogus values are still rejected, preserving
the original protection against the unsigned-subtraction wrap that
would otherwise let ndev->max_mtu underflow to a huge value.
Fixes: d7709812e1 ("net: mana: hardening: Validate adapter_mtu from MANA_QUERY_DEV_CONFIG")
Signed-off-by: Erni Sri Satya Vennela <ernis@linux.microsoft.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260619055348.467224-1-ernis@linux.microsoft.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Upon an MDIO CRC error mxl862xx_crc_err_work_fn() walks the DSA ports
and closes the CPU port conduits:
dsa_switch_for_each_cpu_port(dp, priv->ds)
dev_close(dp->conduit);
mxl862xx_remove() unregisters the switch before cancelling this work:
set_bit(MXL862XX_FLAG_WORK_STOPPED, &priv->flags);
cancel_delayed_work_sync(&priv->stats_work);
dsa_unregister_switch(ds);
mxl862xx_host_shutdown(priv);
dsa_unregister_switch() frees the dsa_port objects. If a CRC error
schedules the work during teardown it can run after the ports have been
freed and dereference freed memory.
Guard the port walk with MXL862XX_FLAG_WORK_STOPPED, which is already set
before dsa_unregister_switch(). DSA tears the ports down under
rtnl_lock(), so checking the flag under rtnl_lock() means the work either
runs before teardown and sees valid ports, or runs afterwards, observes
the flag and skips the walk. This mirrors the host_flood_work handler,
which skips torn-down ports under rtnl_lock().
Link: https://sashiko.dev/#/patchset/cover.1780968180.git.daniel%40makrotopia.org?part=2
Fixes: a319d0c8c8 ("net: dsa: mxl862xx: add CRC for MDIO communication")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Link: https://patch.msgid.link/5e55169926c02f2b914e5ada529d7453b943cda4.1781702256.git.daniel@makrotopia.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The MXL862XX_API_* macros pass the address of a stack-allocated, __packed
firmware-ABI struct to mxl862xx_api_wrap() as a void *. The struct has an
alignment of 1, so the compiler is free to place it at an odd address.
mxl862xx_api_wrap() reinterprets that buffer as a __le16 * and accesses it
with data[i], for which the compiler assumes the natural 2-byte alignment
of __le16 and emits aligned 16-bit loads/stores (e.g. lhu/sh on MIPS).
When the buffer lands on an odd address these fault on architectures that
do not support unaligned access, such as MIPS32.
-Waddress-of-packed-member does not catch this: the packed origin is
laundered through the void * parameter, so the cast inside api_wrap looks
alignment-safe to the compiler and no warning is emitted.
Use get_unaligned_le16()/put_unaligned_le16() for the three 16-bit word
accesses. The byte accesses (*(u8 *)&data[i], crc16()) are already safe
and are left unchanged.
Link: https://sashiko.dev/#/patchset/cover.1781319534.git.daniel%40makrotopia.org?part=4
Fixes: 23794bec1c ("net: dsa: add basic initial driver for MxL862xx switches")
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Link: https://patch.msgid.link/599327521db465a534d277de53ab9b6cac01928b.1781702256.git.daniel@makrotopia.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
led_classdev_register_ext() only reads init_data.devicename - it never
stores the pointer. However, the caller allocated devicename with
kasprintf() but never freed it, leaking the string memory.
Fix it with a stack buffer to avoid dynamic buffers completely.
Fixes: 32d6170054 ("net: dsa: realtek: add LED drivers for rtl8366rb")
Signed-off-by: David Yang <mmyangfl@gmail.com>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260618140200.1888707-1-mmyangfl@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ixp4xx_hss_probe() allocates two HDLC netdevs. The first one is stored
in ndev, initialized, and registered with register_hdlc_device(). The
second one is stored in port->netdev and later used by the remove path
for unregister_hdlc_device() and free_netdev().
This means that the registered netdev is not the same object that is
unregistered and freed on remove. It also leaks the first allocation if
the second alloc_hdlcdev() call fails, and the first allocation is not
checked before ndev is used.
Older code allocated the HDLC netdev only once and stored the same object
in both the local variable and port->netdev. The buggy conversion split
this into two alloc_hdlcdev() calls. A later rename changed the local
variable name to ndev, but the underlying mismatch remained.
Fix this by allocating the HDLC netdev only once and assigning the same
object to port->netdev.
Fixes: 99ebe65eb9 ("net: ixp4xx_hss: move out assignment in if condition")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
Reviewed-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260622043015.643637-1-haoxiang_li2024@163.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
airoha_tc_htb_alloc_leaf_queue() assigns queue IDs based on the channel
index (opt->qid = AIROHA_NUM_TX_RING + channel), but updates
real_num_tx_queues with a simple increment (num_tx_queues + 1). When QoS
channels are allocated sparsely (e.g., channels 0 and 3 without 1 and
2), the returned qid can exceed real_num_tx_queues, causing out-of-bounds
accesses in the networking stack.
For example, allocating channel 0 then channel 3 results in
real_num_tx_queues = 34 but qid = 35, which is out of range [0, 34).
Fix this by computing real_num_tx_queues based on the highest active
channel index rather than using a simple counter, in both the allocation
and deletion paths.
Fixes: ef1ca92713 ("net: airoha: Add sched HTB offload support")
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260619-airoha-qos-fixes-v2-2-5c43485038f9@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
airoha_tc_htb_alloc_leaf_queue() computes the HTB QoS channel index
as opt->classid % AIROHA_NUM_QOS_CHANNELS and stores it in qos_sq_bmap.
However, airoha_tc_remove_htb_queue() clears the HTB configuration
using queue + 1 as the channel index, causing an off-by-one error.
Use queue directly as the QoS channel index to match the allocation
logic.
Fixes: ef1ca92713 ("net: airoha: Add sched HTB offload support")
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Link: https://patch.msgid.link/20260619-airoha-qos-fixes-v2-1-5c43485038f9@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
