The 16-bit pause watermark register is little endian as
described in the ASIX 4.1.0 out-of-tree driver. Correct the
register byte sequence but also swap the configuration values
used in the code in order to keep the current behaviour.
The endianness is relevant for 16-bit writes to the register.
Signed-off-by: Birger Koblitz <mail@birger-koblitz.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Jianhui Xu <neuromoments@gmail.com>
Link: https://patch.msgid.link/20260916-ax88179a-v12-3-60c04c9924a2@birger-koblitz.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The 'priv' variable is defined before an #ifdef block but used inside
of it:
drivers/net/dsa/motorcomm/chip.c: In function 'yt921x_dsa_teardown':
drivers/net/dsa/motorcomm/chip.c:4229:29: error: unused variable 'priv' [-Werror=unused-variable]
Move it one line down.
Fixes: 9af9a81682 ("net: dsa: motorcomm: Add LED support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://patch.msgid.link/20260915200148.3523045-1-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ip6gre_exit_rtnl_net() iterates ip6gre devices and variants whose
link_net is dying and queues them for destruction.
The devices may reside in different netns.
Let's use unregister_netdevice_queue_net() to support per-netns
device unregistration.
Even after ip6gre_exit_rtnl_net() queues a cross-netns ip6gre
device, ip6gre_changelink_common(), ip6gre_dellink(), and
ip6gre_tunnel_siocdevprivate() could be called concurrently for
it (once RTNL is removed). In such a case, __rtnl_net_unlock()
will perform the unregistration.
Also, ip6gre_tunnel_siocdevprivate() needs to check check_net(t->net),
otherwise it could create a new dev in dying netns after
ip6gre_exit_rtnl_net().
In the example below, we can see the fallback tunnel device
(ip6gre0) and the cross-netns device (ip6gre1) are unregistered
by different processes:
# bpftrace -e '#include <linux/netdevice.h>
kprobe:ip6gre_tunnel_uninit {
$dev = (struct net_device *)arg0;
printf("PID: %d | DEV: %s%s\n", pid, $dev->name, kstack());
}
kprobe:ip6gre_exit_rtnl_net {
printf("PID: %d%s\n", pid, kstack());
}' &
# ip netns add ns1
# ip netns add ns2
# ip -n ns1 link add name ip6gre1 link-netns ns2 \
type ip6gre local 2001:9292::1 remote 2001:9292::2
# ip netns del ns2
PID: 12
ip6gre_exit_rtnl_net+5
ops_undo_list+702
cleanup_net+1122
...
PID: 12 | DEV: ip6gre0 <------ fallback device (itn->fb_tunnel_dev).
ip6gre_tunnel_uninit+5
unregister_netdevice_many_notify+7137
unregister_netdevice_many_net+1050
__rtnl_net_unlock+37
ops_undo_list+754
cleanup_net+1122
...
PID: 10 | DEV: ip6gre1
ip6gre_tunnel_uninit+5
unregister_netdevice_many_notify+7137
unregister_netdevice_many_net+1050
rtnl_net_work_func+136
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-7-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
struct ip6_tnl.net is the netns where encapsulated packets
flow into.
struct ip6_tnl is linked to ip6gre_net.tunnels[][] of netns.
During netns dismantle or module unload, ip6gre_exit_rtnl_net()
iterates the list and queues devices for destruction regardless
of the devices' netns.
Thus, once RTNL is removed, the list can be modified concurrently
from different netns due to device removal.
Let's protect it with per-netns mutex.
Note that dev_siocdevprivate() calls netdev_lock_ops() but
it must be NOP for tunnel devices to avoid AB-BA deadlock.
DEBUG_NET_WARN_ON_ONCE() is added to annotate the locking
explicitly.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-6-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
With the previous patch, itn->fb_tunnel_dev can be removed
via ->dellink().
However, ioctl(SIOCDELTUNNEL) still uses unregister_netdevice(),
which requires ip6gre_tunnel_unlink() in ip6gre_tunnel_uninit().
Let's use __ip6gre_dellink() everywhere to remove ip6gre (and
variant) devices and remove ip6gre_tunnel_unlink() and
ip6gre_tunnel_unlink_md() in ip6gre_tunnel_uninit().
ip6erspan_tunnel_uninit() is also changed accordingly.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-5-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ip6gre_dellink() ignores itn->fb_tunnel_dev, so the per-netns
fallback tunnel device cannot be removed by userspace.
This also makes default_device_exit_batch() impossible to remove
the device since it calls ->dellink().
So, ip6gre_exit_rtnl_net() has to iterate devices in the dying
netns and call unregister_netdevice_queue() directly, and
ip6gre_tunnel_unlink() must be called at ->ndo_uninit().
For per-netns device unregistration, we want to check if the
device is being removed in ->dellink(), for which we must call
ip6gre_tunnel_unlink() before unregister_netdevice_queue().
As a prep, let's set itn->fb_tunnel_dev to NULL and remove
remove for_each_netdev_safe() in ip6gre_exit_rtnl_net().
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-4-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
struct ip6_tnl has a pointer to the next one.
Single linked list requires O(n) iteration for device removal
and IPv4 tunnel already uses hlist for its hash table.
Let's convert ip6gre_net.tunnels[][] to hlist.
Once we convert other users (ip6_tunnel.c, sit.c), we can remove
the *next pointer from struct ip6_tnl.
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-3-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ip6gre_init_net() links the fallback device ip6gre0 to
ign->tunnels_wc[0] after publishing it by register_netdev().
While ip6_gre module is being loaded, there is a small race window.
If ioctl(SIOCADDTUNNEL) is called before ip6gre0 is linked to
the hash table, a new device could be linked to ign->tunnels_wc[0].
Then, the fallback device overwrites ign->tunnels_wc[0], and the new
device will be no longer found by ip6gre_tunnel_lookup(), resulting
in no RX.
Let's link the fallback device before register_netdev().
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Link: https://patch.msgid.link/20260916230353.367014-2-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tariq Toukan says:
====================
net/mlx5: more HWS misc enhancements
This series by Yevgeny brings more misc improvements to the mlx5 HWS
driver.
====================
Link: https://patch.msgid.link/20260915111912.3926585-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Remove the unneeded WRITE_ONCE, for two reasons:
a) This write is erroneous - it just writes to a local variable
and doesn't have any effect.
b) This code originated from another place (copy-paste bug),
where there was actually a CQ polling right after this line,
and the WRITE_ONCE was writing to another place.
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Erez Shitrit <erezsh@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/20260915111912.3926585-7-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Handle a case in mlx5hws_send_stes_fw where
draining a queue fails with timeout.
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Erez Shitrit <erezsh@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/20260915111912.3926585-6-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
HWS has a timeout mechanism for usual polling for completion,
but it was missing such a mechanism for draining the send queues.
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Erez Shitrit <erezsh@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://patch.msgid.link/20260915111912.3926585-5-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Walleij says:
====================
net: dsa: microchip: Add support for KSZ8995XA/KS8995XA
This series breaks with the dated attempt to polish the old
KS8995 driver, and instead implement support for the KS8995XA
in the KSZ driver, and after that delete the old KS8995 driver.
The hardware clearly has the same ancestry, the KSZ8995XA is
just a rebrand of the much older Micrel KX8995XA switch.
The old driver identified the shared 0x95/0x0 ID as KS8995MA, but
this ID cannot distinguish MA from XA. No KS8995MA users are known,
and KS8995E has no documented SPI management interface. Neither
variant has been verified with the replacement driver.
Add new compatibles, add special code paths for the KSZ8995XA,
including support for its two-level DCB priority model, and add a
new tagger for the special front tag found in the KSZ8995XA.
The patches, including the per-port PHY reset polling, were tested
with the Actiontec MI424WR rev D (which has the KS8995XA) and
OpenWrt as userspace.
====================
Link: https://patch.msgid.link/20260916-ks8995-to-ksz8-v15-0-8825ff5fd4be@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The Microchip KSZ driver now handles the KS(Z)8995XA. The old
driver identified the shared 0x95/0x0 ID as KS8995MA, but this ID
cannot distinguish MA from XA. No KS8995MA users are known, and
KS8995E has no documented SPI management interface. Neither variant
has been verified with the replacement driver.
The two in-tree device trees using the "micrel,ks8995" compatible:
arch/arm/boot/dts/intel/ixp/intel-ixp42x-linksys-wrv54g.dts
arch/arm/boot/dts/intel/ixp/intel-ixp42x-actiontec-mi424wr.dtsi
both use KS8995XA hardware and continue to be supported by the
replacement driver without regressions.
Delete the old driver, but leave a stub behind in Kconfig so new
users will be automatically phased over to the new symbol when
migrating their configs.
The old driver supports platform data (board file) instantiation,
but nothing in the kernel tree makes use of this legacy mechanism so
it is fine to delete.
Reviewed-by: Nicolai Buchwitz <nb@tipi-net.de>
Signed-off-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260916-ks8995-to-ksz8-v15-4-8825ff5fd4be@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This adds support for the Microchip KSZ8995XA also known as the
Micrel KS8995XA switch to the KSZ driver.
The KSZ8995XA shares family ID 0x95 and chip-ID nibble 0x0 with
the KS8995MA, so hardware detection cannot distinguish them. The
legacy "micrel,ks8995" compatible is therefore treated as KSZ8995XA.
No KS8995MA users are known, and KS8995E has no documented SPI
management interface. Neither MA nor E has been verified with this
driver.
The helper macros are named ksz_is_ksz8995xa() to make it
possible to add E and MA support in the future.
Wait for the self-clearing per-port PHY reset bits before continuing
with switch setup. Poll each port for up to 600 ms and allow an
additional millisecond for the PHYs to settle.
The KSZ8995XA has eight global TOS priority registers with one
bit per DSCP code point, providing two internal priority levels.
The registers are ordered in the opposite direction from the
other KSZ8 switches. Describe the two priorities in the chip
data, then augment the DCB helpers for this.
Practical testing also shows that its packet-size control bits use
the same semantics as the other KSZ8 switches. The polarity in the
manual appears to be wrong and makes the switch hang on big packets.
The legacy KS8995 driver also matches the Micrel compatibles for
KSZ8864 and KSZ8795. Add these to the KSZ SPI driver here and use
an exact tristate dependency to make the drivers mutually exclusive.
This transfers ownership atomically, avoiding module probe races and
gaps in driver coverage between patches.
Reviewed-by: Nicolai Buchwitz <nb@tipi-net.de>
Signed-off-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260916-ks8995-to-ksz8-v15-3-8825ff5fd4be@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The KS8995 100Mbit switch can do proper DSA per-port tagging
with the proper set-up. This adds the code to handle ingress
and egress KS8995 tags.
The tagger is used by the KSZ8995XA support added later in this
series.
The tag is a modified 0x8100 ethertype tag where a bit in the
last nibble is set for each target port.
Always insert a distinct outer switch tag. User ports remove this
field on egress, so reusing an existing 802.1Q header would consume
the sender VLAN tag instead of preserving it.
Push any hardware-accelerated VLAN tag into the packet first. This
keeps the switch tag at offset 12 while retaining 802.1Q, 802.1ad and
stacked VLAN headers as inner payload, and avoids relying on a valid
SKB MAC header in direct transmit paths.
Rate-limit receive untagging errors so repeated failures cannot flood
the kernel log.
Reviewed-by: Nicolai Buchwitz <nb@tipi-net.de>
Signed-off-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260916-ks8995-to-ksz8-v15-2-8825ff5fd4be@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The KSZ8995XA is just like the KSZ8795 and KSZ8864 a Micrel
product. It was renamed from KS8995XA to KSZ8995XA at some point,
but it has the same properties as the KS8995XA.
Be careful to use the full product name in this new compatible:
there is also KSZ8995MA and KSZ8995E which are not compatible
with the KS8995XA.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Reviewed-by: Nicolai Buchwitz <nb@tipi-net.de>
Signed-off-by: Linus Walleij <linusw@kernel.org>
Link: https://patch.msgid.link/20260916-ks8995-to-ksz8-v15-1-8825ff5fd4be@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
inet_unhash() sets inet_csk(sk)->unhashed_state only when the
socket is hashed because tcp_set_state(sk, TCP_CLOSE) could be
called multiple times, e.g. tcp_abort() calls it directly and
tcp_done_with_error().
However, inet_twsk_hashdance_schedule() also unhashes a socket
when replacing it with twsk, allowing the socket to bypass
checks for inet_csk(sk)->unhashed_state.
Let's update inet_csk(sk)->unhashed_state there as well.
Fixes: 8cc3aef0cb ("tcp: Do not allow buggy transitions between ehash and lhash2.")
Reported-by: Daniel Zahka <daniel.zahka@gmail.com>
Closes: https://lore.kernel.org/netdev/DLHLRA8GVI5B.2Q1IRQG5BVJNZ@gmail.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@google.com>
Reviewed-by: Daniel Zahka <daniel.zahka@gmail.com>
Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260917191554.1600494-1-kuniyu@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The phy_may_wakeup() documentation refers to
device_set_wakeup_enabled(), but the function is named
device_set_wakeup_enable(). Correct the reference.
Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://patch.msgid.link/20260917063051.52727-1-kmehltretter@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
PHY fixup registration is used from platform code in init phase only.
Let's move the PHY fixup code from the modular part of phylib to the
always built-in part of phylib. This allows to annotate the fixup
registration as __init. No caller uses the return code of PHY fixup
registration, therefore change related functions to return void.
phy_needs_fixup() and phy_scan_fixups() wouldn't have to be moved to
the built-in part of phylib. But doing so allows to fully factor out
legacy fixup support into its own source code file, and make struct
phy_fixup and phy_fixup_list strictly private to phy_fixup.c.
phy_scan_fixups() is used after init phase only, then phy_fixup_list
is read-only. So we don't need the mutex when accessing the list.
Also when registering PHY fixups the mutex isn't needed, because
fixup registration is done sequentially from platform init code.
Actually there is only one platform with more than one fixup.
In addition this change lays the foundation for enabling modular
phylib on platforms where the fixup just sets a flag.
Last but not least reflect the change in the documentation,
and add a hint that fixup's are legacy functionality.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://patch.msgid.link/2d722b84-a263-400b-ac91-bf37125eb85b@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Since commit d4349ba987 ("netfilter: allow nfnetlink built-in only"),
NETFILTER_NETLINK was changed to a bool. Use 'y' instead of 'm'.
Otherwise the build emits:
.config:5688:warning: symbol value 'm' invalid for NETFILTER_NETLINK
Signed-off-by: Hangbin Liu <liuhangbin@kylinos.cn>
Reviewed-by: Fernando Fernandez Mancera <fmancera@suse.de>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Link: https://patch.msgid.link/20260917-selftest_config-v1-1-bf25ccbfb18e@kylinos.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ionic_probe() allocates VF tracking data and maps a statistics buffer
for each already enabled VF with ionic_vf_alloc().
The probe failure path releases these resources with
ionic_vf_dealloc(). However, after a successful probe, ionic_remove()
tears down the LIF and device without releasing the VF resources.
This leaves the VF tracking allocation and statistics DMA mappings
behind when the PF driver is removed.
Call ionic_vf_dealloc() before deinitializing the LIF so the VF
statistics addresses can be cleared and the associated DMA mappings
and tracking data released.
This issue was found by manual code inspection.
Cc: stable+noautosel@kernel.org # untested fix to a driver init path
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Link: https://patch.msgid.link/20260916053117.2906446-1-lgs201920130244@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
hsr: lockless hsr_fill_info()
hsr_fill_info() currently relies on RTNL being held, because
hsr_port_get_hsr() walks hsr->ports with hsr_for_each_port_rtnl().
- Patch 1 annotates the data-race around hsr->sequence_nr, which
is incremented from the transmit paths under hsr->seqnr_lock but
read without it.
- Patch 2 converts hsr_fill_info() to RCU, using a local RCU
variant of hsr_port_get_hsr().
hsr->ports is already an RCU protected list: ports are added with
list_add_tail_rcu(), removed with list_del_rcu() and freed with
kfree_rcu(), and the data path already iterates over it with
hsr_for_each_port().
This is part of a larger effort to remove the RTNL dependency from
rtnl_link_ops->fill_info().
====================
Link: https://patch.msgid.link/20260916132822.1791040-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
hsr_fill_info() used hsr_port_get_hsr(), which walks hsr->ports
with hsr_for_each_port_rtnl(), forcing "ip link show" dumps to
hold RTNL.
hsr->ports is an RCU protected list: ports are added with
list_add_tail_rcu(), removed with list_del_rcu() and freed with
kfree_rcu(), and the data path already iterates over it with
hsr_for_each_port().
Add a local RCU variant of hsr_port_get_hsr() and use it from
hsr_fill_info() under rcu_read_lock().
The remaining fields (sup_multicast_addr, prot_version) are set
once at device creation. hsr->sequence_nr is now read with
READ_ONCE(), since the transmit paths increment it under
hsr->seqnr_lock only.
Also add missing const qualifiers.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260916132822.1791040-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
hsr->sequence_nr is incremented from the transmit paths under
hsr->seqnr_lock, but hsr_fill_info() reads it without holding
this lock.
Use WRITE_ONCE() in the two writers, so that a matching
READ_ONCE() can be used in hsr_fill_info().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260916132822.1791040-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
vlan: lockless vlan_fill_info()
vlan_fill_info() currently relies on RTNL being held, because it
uses rcu_dereference_rtnl() to walk vlan->egress_priority_map[].
While auditing the fields exposed there, it appears that
vlan->flags, vlan->ingress_priority_map[] and vlan->nr_*_mappings
are written under RTNL but already read from lockless contexts
(RX, TX, /proc/net/vlan/<dev> and vlan_get_size()).
- Patch 1 adds the missing READ_ONCE()/WRITE_ONCE() annotations.
- Patch 2 converts vlan_fill_info() to RCU.
vlan->egress_priority_map[] needed no work: it is already updated
with rcu_assign_pointer() and freed with kfree_rcu().
This is part of a larger effort to remove the RTNL dependency from
rtnl_link_ops->fill_info().
====================
Link: https://patch.msgid.link/20260916132532.1785657-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
vlan_fill_info() used rcu_dereference_rtnl() to walk the egress
priority map, forcing "ip link show" dumps to hold RTNL.
The egress priority map is already fully RCU compliant:
entries are published with rcu_assign_pointer() and freed with
kfree_rcu() from vlan_dev_set_egress_priority() and
vlan_dev_free_egress_priority().
Wrap the QoS mapping serialization in rcu_read_lock() /
rcu_read_unlock() and use rcu_dereference().
Now that all the involved fields are annotated, use READ_ONCE()
for vlan->flags, vlan->nr_ingress_mappings, vlan->nr_egress_mappings
and vlan->ingress_priority_map[].
Also add missing const qualifiers in vlan_get_size() and
vlan_fill_info().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260916132532.1785657-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
vlan->flags, vlan->ingress_priority_map[] and vlan->nr_*_mappings
are written under RTNL, but read from lockless contexts:
- the RX path reads vlan->flags (VLAN_FLAG_REORDER_HDR) and
the ingress priority map,
- the TX path reads vlan->flags,
- /proc/net/vlan/<dev> reads vlan->flags and the ingress priority
map,
- vlan_get_size() reads vlan->nr_*_mappings,
- vlan_fill_info() will soon run without RTNL.
Add the missing READ_ONCE()/WRITE_ONCE() annotations.
vlan_dev_change_flags() is also made a bit more readable by using
a local new_flags variable, avoiding the repeated re-reads of
vlan->flags after it has been published.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20260916132532.1785657-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
dpaa2_ptp_probe() gets a referenced device node with
of_find_compatible_node() and stores it in dev->of_node.
The probe failure paths release this reference with of_node_put().
However, after a successful probe, dpaa2_ptp_remove() tears down the
PTP and MC resources without dropping the device node reference.
Release dev->of_node during driver removal and clear the pointer after
dropping the reference.
This issue was found by manual code inspection.
Cc: stable+noautosel@kernel.org # untested fix to a driver init path
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Link: https://patch.msgid.link/20260916030015.2818545-1-lgs201920130244@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Breno Leitao says:
====================
net: a sockopt_t quirk for the options that write past optlen
This series continues the migration of our protocols to sockopt_t, as
described in [1].
There are some protocols that use optlen as the header size, and the real
buffer size comes from a field inside the header. This means a bug, given
that optlen should be the full buffer size, but there are indications [2]
that we have programs that use the bad behaviour above, and we want to
avoid breaking them (or, honestly, avoid being cursed by Linus).
That said, create a quirk helper that preserves the same behaviour, even
using sockopt_t. The way to do it is simple:
1) Only do it for userspace callers (ubuf), otherwise a bug here will
corrupt the kernel instead of a simple SIGSEGV.
2) Expand optval mid-air based on the header field.
IP_MSFILTER is the first user, and the smallest one: a single caller, no
compat variant, and a reply written front to back. MCAST_MSFILTER on ipv4
and ipv6 comes next, and TCP_AO_GET_KEYS has the same shape. Do this
quirk on IP_MSFILTER to make sure the dynamic is ok, so, we can expand
it later.
None of this is meant to change what userspace sees.
====================
Link: https://patch.msgid.link/20260914-getsockopt_phase6-v2-0-e48befc9602e@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
IP_MSFILTER reads its reply through ip_mc_msfget(), reached from
do_ip_getsockopt() and from nowhere else. Convert it, and build the
sockopt_t at the call site for as long as the caller still carries a
sockptr_t pair.
optlen here only has to cover the header, and the real reply size comes
from the imsf_numsrc field inside it. This is nasty, but userspace
relies on it, so sockopt_expand_out() preserves the same mechanism: it
grows optval only for a user address, and assumes the caller left room
for the size its own header asked for.
The *optlen store moves out of ip_mc_msfget() and into the call site,
guarded by !err so the -EINVAL, -ENODEV and -EADDRNOTAVAIL returns still
leave the caller's optlen word untouched.
The source list also moves from copy_to_sockptr_offset() to a sequential
copy_to_iter(). IP_MSFILTER_SIZE(0) and offsetof(struct ip_msfilter,
imsf_slist_flex) are both 16, so the bytes land where they did.
Signed-off-by: Breno Leitao <leitao@debian.org>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260914-getsockopt_phase6-v2-2-e48befc9602e@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some getsockopt options size their reply from a count the caller left in
optval rather than from optlen, and so write past the optlen the caller
declared. Userspace relies on that, so the sockopt_t conversion has to
keep doing it.
IP_MSFILTER is the first one to convert: its reply covers the
imsf_numsrc sources the caller asked for, while optlen only has to cover
the fixed header.
Add sockopt_expand_out() to grow opt->iter_out mid-air, so the quirk
sits in one place instead of each protocol assuming it implicitly. It is
a no-op unless the reply outruns optlen. Growing re-anchors the iterator
at the head of optval, so it has to be called before anything is written
through iter_out.
Only a user buffer can be longer than optlen says. A kernel-backed
optval keeps the bounded iterator, and a callback that asks to grow one
gets a WARN_ON_ONCE() and -EINVAL.
Nothing in tree can trip that WARN: the only kernel-backed path into
do_ip_getsockopt() is sol_ip_sockopt(), which takes IP_TOS and
IP_TRANSPARENT only. It is an assert for the in-kernel callers BPF and
io_uring gain once the conversion is done, so they declare an optlen
covering the whole buffer instead of repeating the userspace mistake.
The conversion in the next patch calls sockptr_to_sockopt() from
net/ipv4/, so drop its static and declare it in net.h.
Signed-off-by: Breno Leitao <leitao@debian.org>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20260914-getsockopt_phase6-v2-1-e48befc9602e@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Eric Dumazet says:
====================
net: rmnet: lockless rmnet_fill_info()
rmnet_fill_info() currently relies on RTNL being held, because it
uses rmnet_get_port_rtnl() to reach the rmnet_port attached to the
underlying real device.
While auditing the fields exposed there, it appears that both
priv->mux_id and port->data_format are written under RTNL but read
from the data path without any lock. Note that rmnet_newlink() can
update port->data_format of an already active port, and that
rmnet_changelink() can change both fields while traffic is flowing.
- Patch 1 annotates the data-races around port->data_format. It also
samples the field only once per packet and passes the value down,
so that both ends of a packet transformation agree on the format
(TX used to size the MAP headroom and set MAP_NEXT_HEADER_FLAG
from two separate reads), and prevents rmnet_changelink() from
publishing an intermediate value to the data path.
- Patch 2 annotates the data-races around priv->mux_id and
ep->mux_id. It also publishes ep->mux_id before the endpoint is
inserted into its new bucket, so that a concurrent lookup walking
that bucket can no longer miss it and drop the packet.
- Patch 3 converts rmnet_fill_info() to RCU. rmnet_get_port_rcu()
was only called from the data path and thus used
rcu_dereference_bh(); its lockdep condition is relaxed so that it
can also be called from process context.
The lockless lookup is safe because rmnet_get_port_rcu() checks
real_dev->rx_handler before returning rx_handler_data, while
rmnet_unregister_real_device() clears rx_handler and waits for a
grace period (in netdev_rx_handler_unregister()) before freeing the
port.
This is part of a larger effort to remove the RTNL dependency from
rtnl_link_ops->fill_info().
====================
Link: https://patch.msgid.link/20260916132231.1780091-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
rmnet_fill_info() used rmnet_get_port_rtnl(), and thus required
RTNL.
Now that priv->mux_id and port->data_format are properly
annotated, rmnet_fill_info() can simply fetch the port under
rcu_read_lock().
rmnet_get_port_rcu() was only used from the data path and thus
used rcu_dereference_bh(). Relax its lockdep condition so that it
can also be called from process context under a plain
rcu_read_lock(), and use it from rmnet_fill_info().
Note that rmnet_get_port_rcu() checks real_dev->rx_handler before
returning rx_handler_data: this is what makes the lockless lookup
safe against rmnet_unregister_real_device(), which clears
rx_handler, waits for a grace period in
netdev_rx_handler_unregister(), and only then frees the port.
While at it, add missing const qualifiers.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Sean Tranchetti <sean.tranchetti@oss.qualcomm.com>
Reviewed-by: Subash Abhinov Kasiviswanathan <subash.a.kasiviswanathan@oss.qualcomm.com>
Link: https://patch.msgid.link/20260916132231.1780091-4-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
priv->mux_id is written under RTNL from rmnet_vnd_newlink() and
rmnet_changelink(), but rmnet_egress_handler() reads it from the
transmit path without any lock.
Likewise ep->mux_id is written under RTNL from rmnet_changelink(),
while rmnet_get_endpoint() reads it from the receive path under
RCU only.
Add the missing READ_ONCE()/WRITE_ONCE() annotations.
While at it, publish ep->mux_id before inserting the endpoint in
its new bucket in rmnet_changelink(). Otherwise a concurrent
rmnet_get_endpoint() walking the new bucket could find the
endpoint still carrying its old mux_id, and drop the packet.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Sean Tranchetti <sean.tranchetti@oss.qualcomm.com>
Reviewed-by: Subash Abhinov Kasiviswanathan <subash.a.kasiviswanathan@oss.qualcomm.com>
Link: https://patch.msgid.link/20260916132231.1780091-3-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
port->data_format is written under RTNL from rmnet_newlink() and
rmnet_changelink(), but it is read from the data path without any
lock, both from RX (rmnet_rx_handler() context) and TX
(rmnet_egress_handler() context).
Note that rmnet_newlink() can be called for a real_dev which is
already hooked to rmnet: the rx_handler is live and traffic can
already be flowing when port->data_format is overwritten.
Add the missing READ_ONCE()/WRITE_ONCE() annotations.
While at it, sample port->data_format only once per packet and pass
the value down, so that all the decisions taken for a given packet
are based on a single consistent value.
Otherwise the two ends of a packet transformation could disagree.
On TX, rmnet_map_egress_handler() sized the headroom from one read
while rmnet_map_add_map_header() decided on MAP_NEXT_HEADER_FLAG
from another one, so a concurrent rmnet_changelink() could produce
a MAP header announcing a v5 csum header that was neither reserved
nor written. On RX, rmnet_map_validate_packet_len() and
__rmnet_map_ingress_handler() could likewise disagree on the
expected layout, and rmnet_map_send_ack() could trim a dl csum
trailer that the ingress path never accounted for.
rmnet_map_add_map_header(), rmnet_map_command(),
rmnet_map_deaggregate(), rmnet_map_send_ack(),
rmnet_map_validate_packet_len() and __rmnet_map_ingress_handler()
therefore get the value from their caller instead of re-reading it.
rmnet_changelink() now computes the new value in a local variable
and publishes it with a single store, instead of letting the data
path observe the intermediate (old_data_format & ~flags->mask)
value.
rmnet_vnd_headroom() is left alone, all its callers hold RTNL.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Sean Tranchetti <sean.tranchetti@oss.qualcomm.com>
Reviewed-by: Subash Abhinov Kasiviswanathan <subash.a.kasiviswanathan@oss.qualcomm.com>
Link: https://patch.msgid.link/20260916132231.1780091-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
fs_enet_probe() allocates the platform information structure with
kzalloc_obj() and stores it in fep->fpi.
The probe failure paths release this allocation with kfree(). However,
after a successful probe, fs_enet_remove() tears down the network
device without freeing fep->fpi, leaving the platform information
structure allocated after driver removal.
Free fep->fpi before releasing the network device.
This issue was found by manual code inspection.
Cc: stable+noautosel@kernel.org # untested fix to a driver init path
Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20260916031414.2875022-1-lgs201920130244@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Arria10 SoCFPGA DWMAC nodes use reset-names = "stmmaceth",
"stmmaceth-ocp", which matches what dwmac-socfpga looks up for the
OCP reset. The generic snps,dwmac binding only allowed "ahb" as the
second reset name, so Arria10 DTS failed schema validation.
Allow "stmmaceth-ocp" as an alternative second reset name and drop
the related TODO from the Altera SoCFPGA DWMAC binding.
Signed-off-by: Nazim Amirul <muhammad.nazim.amirul.nazle.asmade@altera.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://patch.msgid.link/20260909080559.2028-1-muhammad.nazim.amirul.nazle.asmade@altera.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The `strcpy()` function is deprecated and moving towards code-tree
elimination. Replacing it with `strscpy()` fixes potential buffer
overflow vectors by ensuring safe NULL-termination based on the
destination buffer size limit [1][2].
To make the interface safer and more robust for future callers, refactor
`tipc_bearer_get_name()` to accept a destination buffer length parameter.
Replace `strcpy()` with `strscpy()` and pass through any potential
`-E2BIG` truncation error code up to the caller. Update the existing
caller in `net/tipc/monitor.c` to pass its array size.
Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strcpy [1]
Link: https://github.com/KSPP/linux/issues/88 [2]
Signed-off-by: Ajith Parappalliyalil Venugopal <ajithpv.linux@gmail.com>
Reviewed-by: Tung Nguyen <tung.quang.nguyen@est.tech>
Link: https://patch.msgid.link/20260915145641.2434403-1-ajithpv.linux@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
port-index is the only way rate-set can address a leaf (port) rate object,
but it was never listed in the request, so the generated policy has no
entry for it. The op declares .maxattr = DEVLINK_ATTR_PARENT_DEV, so the
attribute still reaches info->attrs[], validated against a zeroed slot -
NLA_UNSPEC, length 0 - which GENL_DONT_VALIDATE_STRICT accepts at any
length.
devlink_port_get_from_attrs() then runs nla_get_u32() on it. Handed a
zero-length port-index the kernel reads the four bytes past the payload,
which are the next attribute's header, and acts on the port index those
spell out. Since we're reading a linear skb the OOB read is still
within the same memory allocation, it's just garbage. We also do not
echo the garbage back to the user so it's not an info leak either.
Hence not treating this is a real bug fix.
rate-new is left alone on purpose. It creates rate nodes, resolved by
name through devlink_rate_node_get_from_attrs(), and never looks at
port-index.
Link: https://patch.msgid.link/20260915161341.1053476-10-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We are missing the pad attribute in a number of subsets.
Worse, dl-attr-stats which is a separate attr space doesn't
define pad, but reuses the value from the root attr set
(DEVLINK_ATTR_PAD). This would break parsing stats on
an arch without CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS.
Link: https://patch.msgid.link/20260915161341.1053476-9-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
devlink_nl_rate_parent_fill() emits DEVLINK_ATTR_PARENT_DEV via
devlink_nl_put_nested_handle(), which adds DEVLINK_ATTR_NETNS_ID
whenever the parent rate node lives on a devlink instance in another
netns. Now that we defined dl-nested-devlink subset, which unlike
dl-parent-dev includes netns-id, let's switch to it. The rate APIs
input genuinely does not accept the netns-id, so we need to subset
them to stick with dl-parent-dev.
Fixes: b5f90fd458 ("devlink: Add parent dev to devlink API")
Link: https://patch.msgid.link/20260915161341.1053476-8-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Three attributes are emitted repeatedly inside their nest:
- devlink_nl_region_read_fill() calls
devlink_nl_cmd_region_read_chunk_fill() once per 256 byte unit
- devlink_nl_region_snapshots_id_put() emits one
DEVLINK_ATTR_REGION_SNAPSHOT sub-nest per entry on the region's
snapshot list
- devlink_nl_linecard_fill() emits one DEVLINK_ATTR_LINECARD_TYPE per
linecard->types_count entry
In all three cases we're missing the multi-attr properties, and the
attributes themselves are missing from the attr list used by C code gen.
Link: https://patch.msgid.link/20260915161341.1053476-7-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
devlink has a one huge root attribute set for the whole family,
we haven't taken the time to properly define the sub-sets for
each command. Do it for port-set so that we can drop the hand
written policy used by devlink_port_function_set().
We need this subsetting because within the DEVLINK_ATTR_PORT_FUNCTION
nest DEVLINK_PORT_FN_ATTR_OPSTATE and DEVLINK_PORT_FN_ATTR_DEVLINK
are output-only so we have to filter them out of the input set.
Link: https://patch.msgid.link/20260915161341.1053476-6-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
dl-port-function stops at caps, but the nest also carries
DEVLINK_PORT_FN_ATTR_DEVLINK (5) and DEVLINK_PORT_FN_ATTR_MAX_IO_EQS (6)
both put by devlink_nl_port_function_attrs_put() on every port-get
do and dump. YNL raises
Space 'dl-port-function' has no attribute with value '6'
for any port reporting max_io_eqs or a nested devlink handle,
i.e. for mlx5 SFs and VFs.
Commit 5af3e3876d ("devlink: Support setting max_io_eqs") added the
uAPI value and the hand written policy but never touched the spec.
Add the missing attributes, subsequent commit reworks the code
to use the YNL-generated policy.
Link: https://patch.msgid.link/20260915161341.1053476-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
