net: ravb: Use common defines for time stamping control
Instead of translating to/from driver specific flags for packet time
stamp control use the common flags directly. This simplifies the driver
as the translating code can be removed while at the same time making it
clear the flags are not flags written to hardware registers.
The change from a device specific bit-field track variable to the common
enum datatypes forces us to touch the ravb_rx_rcar_hwstamp() in a non
trivial way. To make this cleaner and easier to understand expand the
nested conditions.
Prepare for moving away from device specific bit-fields to track how to
do hardware Rx timestamping to using net common enums by breaking out
the timestamping to a helper function. This is done to create cleaner
code and prepare for easier changes improving the hardware timestapming.
The driver specific flags to control packet time stamps have all been
replaced by values from enum hwtstamp_tx_types and enum
hwtstamp_rx_filters. Remove the driver specific flags as there are no
more users.
net: rtsn: Use common defines for time stamping control
Instead of translating to/from driver specific flags for packet time
stamp control use the common flags directly. This simplifies the driver
as the translating code can be removed while at the same time making it
clear the flags are not flags written to hardware registers.
One thing to note is that the bit-wise and check in rtsn_rx() of
RCAR_GEN4_RXTSTAMP_TYPE_V2_L2_EVENT is replaced with a not set check of
HWTSTAMP_FILTER_NONE. This is okay as the bit of device specific event
replaced was set for all modes except HWTSTAMP_FILTER_NONE.
net: rswitch: Use common defines for time stamping control
Instead of translating to/from driver specific flags for packet time
stamp control use the common flags directly. This simplifies the driver
as the translating code can be removed while at the same time making it
clear the flags are not flags written to hardware registers.
One thing to note is that the bit-wise and check in rswitch_rx() of
RCAR_GEN4_RXTSTAMP_TYPE_V2_L2_EVENT is replaced with a not set check of
HWTSTAMP_FILTER_NONE. This is okay as the bit of device specific event
replaced was set for all modes except HWTSTAMP_FILTER_NONE.
The struct rcar_gen4_ptp_private provides two fields for convenience of
its users, tstamp_tx_ctrl and tstamp_rx_ctrl. These fields are not used
by the rcar_gen4_ptp driver itself but only by the drivers using it.
Upcoming work will enable the RAVB driver currently only supporting gPTP
on pre-Gen4 SoCs to use the Gen4 implementation as well. To facilitate
this the convenience of having these fields in struct
rcar_gen4_ptp_private becomes a problem as the RAVB driver already have
it's own driver specific fields for the same thing.
Move the fields from struct rcar_gen4_ptp_private to each driver using
the Gen4 gPTP clocks own private data structures. There is no functional
change.
The files rcar_gen4_ptp.{c,h} implements an abstraction of the gPTP
support implemented together with different other IP blocks. The first
device added which supported this was RSWITCH on R-Car S4.
While doing so the RSWITCH R-Car S4 specific offset was added to the
generic Gen4 gPTP header file. Move it to the RSWITCH driver to make it
clear it only applies to this driver.
Daniel Borkmann [Fri, 31 Oct 2025 21:20:59 +0000 (22:20 +0100)]
netkit: Document fast vs slowpath members via macros
Instead of a comment, just use two cachline groups to document the intent
for members often accessed in fast or slow path.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Co-developed-by: David Wei <dw@davidwei.uk> Signed-off-by: David Wei <dw@davidwei.uk> Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org> Link: https://patch.msgid.link/20251031212103.310683-11-daniel@iogearbox.net Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Daniel Borkmann [Fri, 31 Oct 2025 21:20:55 +0000 (22:20 +0100)]
xsk: Move NETDEV_XDP_ACT_ZC into generic header
Move NETDEV_XDP_ACT_ZC into xdp_sock_drv.h header such that external code
can reuse it, and rename it into more generic NETDEV_XDP_ACT_XSK.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Co-developed-by: David Wei <dw@davidwei.uk> Signed-off-by: David Wei <dw@davidwei.uk> Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org> Link: https://patch.msgid.link/20251031212103.310683-7-daniel@iogearbox.net Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Breno Leitao [Wed, 5 Nov 2025 18:01:12 +0000 (10:01 -0800)]
tg3: extract GRXRINGS from .get_rxnfc
Commit 84eaf4359c36 ("net: ethtool: add get_rx_ring_count callback to
optimize RX ring queries") added specific support for GRXRINGS callback,
simplifying .get_rxnfc.
Remove the handling of GRXRINGS in .get_rxnfc() by moving it to the new
.get_rx_ring_count().
Given that tg3_get_rxnfc() only handles ETHTOOL_GRXRINGS, then this
function becomes useless now, and it is removed.
This also fixes the behavior for devices without MSIX support.
Previously, the function would return -EOPNOTSUPP, but now it correctly
returns 1.
The functionality remains the same: return the current queue count
if the device is running, otherwise return the minimum of online
CPUs and TG3_RSS_MAX_NUM_QS.
====================
net: dsa: lantiq_gswip: Add support for MaxLinear GSW1xx switch family
This patch series extends the existing lantiq_gswip DSA driver to
support the MaxLinear GSW1xx family of dedicated Ethernet switch ICs.
These switches are based on the same IP as the Lantiq/Intel GSWIP found
in VR9 and xRX MIPS router SoCs which are currently supported by the
lantiq_gswip driver, but they are dedicated ICs connected via MDIO
rather than built-in components of a SoC accessible via memory-mapped
I/O.
The series includes several improvements and refactoring to implement
support for GSW1xx switch ICs by reusing the existing lantiq_gswip
driver.
The GSW1xx family includes several variants:
- GSW120: 4 ports, 2 PHYs, RGMII & SGMII/2500Base-X
- GSW125: 4 ports, 2 PHYs, RGMII & SGMII/2500Base-X, industrial temperature
- GSW140: 6 ports, 4 PHYs, RGMII & SGMII/2500Base-X
- GSW141: 6 ports, 4 PHYs, RGMII & SGMII
- GSW145: 6 ports, 4 PHYs, RGMII & SGMII/2500Base-X, industrial temperature
Key features implemented:
- MDIO-based register access using regmap
- Support for SGMII/1000Base-X/2500Base-X SerDes interfaces
- Configurable RGMII delays via device tree properties
- Configurable RMII clock direction
- Energy Efficient Ethernet (EEE) support
- enabling/disabling learning
====================
Daniel Golle [Mon, 3 Nov 2025 12:20:28 +0000 (12:20 +0000)]
net: dsa: add driver for MaxLinear GSW1xx switch family
Add driver for the MaxLinear GSW1xx family of Ethernet switch ICs which
are based on the same IP as the Lantiq/Intel GSWIP found in the Lantiq VR9
and Intel GRX MIPS router SoCs. The main difference is that instead of
using memory-mapped I/O to communicate with the host CPU these ICs are
connected via MDIO (or SPI, which isn't supported by this driver).
Implement the regmap API to access the switch registers over MDIO to allow
reusing lantiq_gswip_common for all core functionality.
The GSW1xx also comes with a SerDes port capable of 1000Base-X, SGMII and
2500Base-X, which can either be used to connect an external PHY or SFP
cage, or as the CPU port. Support for the SerDes interface is implemented
in this driver using the phylink_pcs interface.
Daniel Golle [Mon, 3 Nov 2025 12:20:20 +0000 (12:20 +0000)]
net: dsa: add tagging driver for MaxLinear GSW1xx switch family
Add support for a new DSA tagging protocol driver for the MaxLinear
GSW1xx switch family. The GSW1xx switches use a proprietary 8-byte
special tag inserted between the source MAC address and the EtherType
field to indicate the source and destination ports for frames
traversing the CPU port.
Implement the tag handling logic to insert the special tag on transmit
and parse it on receive.
Daniel Golle [Mon, 3 Nov 2025 12:20:12 +0000 (12:20 +0000)]
dt-bindings: net: dsa: lantiq,gswip: add support for MaxLinear GSW1xx switches
Extend the Lantiq GSWIP device tree binding to also cover MaxLinear
GSW1xx switches which are based on the same hardware IP but connected
via MDIO instead of being memory-mapped.
Add compatible strings for MaxLinear GSW120, GSW125, GSW140, GSW141,
and GSW145 switches and adjust the schema to handle the different
connection methods with conditional properties.
Add MaxLinear GSW125 example showing MDIO-connected configuration.
Daniel Golle [Mon, 3 Nov 2025 12:20:05 +0000 (12:20 +0000)]
net: dsa: lantiq_gswip: allow adjusting MII delays
Currently the MII clk vs. data delay is configured based on the PHY
interface mode.
In addition to that add support for setting up MII delays using the
standard Device Tree properties 'tx-internal-delay-ps' and
'rx-internal-delay-ps', using the values determined by the PHY interface
mode as default to maintain backward compatibility with legacy device
trees.
Daniel Golle [Mon, 3 Nov 2025 12:19:58 +0000 (12:19 +0000)]
dt-bindings: net: dsa: lantiq,gswip: add support for MII delay properties
Add support for standard tx-internal-delay-ps and rx-internal-delay-ps
properties on port nodes to allow fine-tuning of RGMII clock delays.
The GSWIP switch hardware supports delay values in 500 picosecond
increments from 0 to 3500 picoseconds, with a post-reset default of 2000
picoseconds for both TX and RX delays. The driver currently sets the
delay to 0 in case the PHY is setup to carry out the delay by the
corresponding interface modes ("rgmii-id", "rgmii-rxid", "rgmii-txid").
This corresponds to the driver changes that allow adjusting MII delays
using Device Tree properties instead of relying solely on the PHY
interface mode.
Daniel Golle [Mon, 3 Nov 2025 12:19:47 +0000 (12:19 +0000)]
net: dsa: lantiq_gswip: add vendor property to setup MII refclk output
Read boolean Device Tree property "maxlinear,rmii-refclk-out" and switch
the RMII reference clock to be a clock output rather than an input if it
is set.
Add support for the maxlinear,rmii-refclk-out boolean property on port
nodes to configure the RMII reference clock to be an output rather than
an input.
This property is only applicable for ports in RMII mode and allows the
switch to provide the reference clock for RMII-connected PHYs instead
of requiring an external clock source.
This corresponds to the driver changes that read this Device Tree
property to configure the RMII clock direction.
Daniel Golle [Mon, 3 Nov 2025 12:19:25 +0000 (12:19 +0000)]
net: dsa: lantiq_gswip: define and use GSWIP_TABLE_MAC_BRIDGE_VAL1_VALID
When adding FDB entries to the MAC bridge table on GSWIP 2.2 or later it
is needed to set an (undocumented) bit to mark the entry as valid. If this
bit isn't set for entries in the MAC bridge table, then those entries won't
be considered as valid MAC addresses.
Daniel Golle [Mon, 3 Nov 2025 12:19:10 +0000 (12:19 +0000)]
net: dsa: lantiq_gswip: set link parameters also for CPU port
On standalone switch ICs the link parameters of the CPU port need to
be setup just like user ports. The destinction in the driver to not
carry out link parameter setup for the CPU port does make sense for
in-SoC switches on which the CPU port is internally connected to the
SoC's Ethernet MAC.
Set link parameters also for the CPU port unless it is an internal
interface. Note that the internal TP PHYs anyway cannot be used as
CPU ports, hence it doesn't matter that they are now also covered by
that condition.
Daniel Golle [Mon, 3 Nov 2025 12:18:49 +0000 (12:18 +0000)]
net: dsa: lantiq_gswip: support enable/disable learning
Switch API 2.2 or later supports enabling or disabling learning on each
port. Implement support for BR_LEARNING bridge flag and announce support
for BR_LEARNING on GSWIP 2.2 or later.
Daniel Golle [Mon, 3 Nov 2025 12:18:41 +0000 (12:18 +0000)]
net: dsa: lantiq_gswip: split into common and MMIO parts
Move all parts specific for the MMIO/SoC driver into a module of its own
to prepare for supporting MDIO-connected switch ICs.
Modify gswip_probe() functions by splitting it into a common function
gswip_probe_common() which covers allocating, initializing and registering
the DSA switch, while keeping transport-specific regmap initialization as
well as PHY firmware loading in the new MMIO/SoC-specific gswip_probe()
function.
Cross-merge networking fixes after downstream PR (net-6.18-rc5).
Conflicts:
drivers/net/wireless/ath/ath12k/mac.c 9222582ec524 ("Revert "wifi: ath12k: Fix missing station power save configuration"") 6917e268c433 ("wifi: ath12k: Defer vdev bring-up until CSA finalize to avoid stale beacon")
https://lore.kernel.org/11cece9f7e36c12efd732baa5718239b1bf8c950.camel@sipsolutions.net
Adjacent changes:
drivers/net/ethernet/intel/Kconfig b1d16f7c0063 ("libie: depend on DEBUG_FS when building LIBIE_FWLOG") 93f53db9f9dc ("ice: switch to Page Pool")
Linus Torvalds [Thu, 6 Nov 2025 16:52:30 +0000 (08:52 -0800)]
Merge tag 'net-6.18-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
Including fixes from bluetooth and wireless.
Current release - new code bugs:
- ptp: expose raw cycles only for clocks with free-running counter
- bonding: fix null-deref in actor_port_prio setting
- mdio: ERR_PTR-check regmap pointer returned by
device_node_to_regmap()
- eth: libie: depend on DEBUG_FS when building LIBIE_FWLOG
Previous releases - regressions:
- virtio_net: fix perf regression due to bad alignment of
virtio_net_hdr_v1_hash
- Revert "wifi: ath10k: avoid unnecessary wait for service ready
message" caused regressions for QCA988x and QCA9984
- Revert "wifi: ath12k: Fix missing station power save configuration"
caused regressions for WCN7850
- eth: bnxt_en: shutdown FW DMA in bnxt_shutdown(), fix memory
corruptions after kexec
Previous releases - always broken:
- virtio-net: fix received packet length check for big packets
- sctp: fix races in socket diag handling
- wifi: add an hrtimer-based delayed work item to avoid low
granularity of timers set relatively far in the future, and use it
where it matters (e.g. when performing AP-scheduled channel switch)
- eth: mlx5e:
- correctly propagate error in case of module EEPROM read failure
- fix HW-GRO on systems with PAGE_SIZE == 64kB
- dsa: b53: fixes for tagging, link configuration / RMII, FDB,
multicast
- phy: lan8842: implement latest errata"
* tag 'net-6.18-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (63 commits)
selftests/vsock: avoid false-positives when checking dmesg
net: bridge: fix MST static key usage
net: bridge: fix use-after-free due to MST port state bypass
lan966x: Fix sleeping in atomic context
bonding: fix NULL pointer dereference in actor_port_prio setting
net: dsa: microchip: Fix reserved multicast address table programming
net: wan: framer: pef2256: Switch to devm_mfd_add_devices()
net: libwx: fix device bus LAN ID
net/mlx5e: SHAMPO, Fix header formulas for higher MTUs and 64K pages
net/mlx5e: SHAMPO, Fix skb size check for 64K pages
net/mlx5e: SHAMPO, Fix header mapping for 64K pages
net: ti: icssg-prueth: Fix fdb hash size configuration
net/mlx5e: Fix return value in case of module EEPROM read error
net: gro_cells: Reduce lock scope in gro_cell_poll
libie: depend on DEBUG_FS when building LIBIE_FWLOG
wifi: mac80211_hwsim: Limit destroy_on_close radio removal to netgroup
netpoll: Fix deadlock in memory allocation under spinlock
net: ethernet: ti: netcp: Standardize knav_dma_open_channel to return NULL on error
virtio-net: fix received length check in big packets
bnxt_en: Fix warning in bnxt_dl_reload_down()
...
Bobby Eshleman [Wed, 5 Nov 2025 15:59:19 +0000 (07:59 -0800)]
selftests/vsock: avoid false-positives when checking dmesg
Sometimes VMs will have some intermittent dmesg warnings that are
unrelated to vsock. Change the dmesg parsing to filter on strings
containing 'vsock' to avoid false positive failures that are unrelated
to vsock. The downside is that it is possible for some vsock related
warnings to not contain the substring 'vsock', so those will be missed.
Fixes: a4a65c6fe08b ("selftests/vsock: add initial vmtest.sh for vsock") Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: Bobby Eshleman <bobbyeshleman@meta.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Link: https://patch.msgid.link/20251105-vsock-vmtest-dmesg-fix-v2-1-1a042a14892c@meta.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 6 Nov 2025 15:32:19 +0000 (07:32 -0800)]
Merge branch 'net-bridge-fix-two-mst-bugs'
Nikolay Aleksandrov says:
====================
net: bridge: fix two MST bugs
Patch 01 fixes a race condition that exists between expired fdb deletion
and port deletion when MST is enabled. Learning can happen after the
port's state has been changed to disabled which could lead to that
port's memory being used after it's been freed. The issue was reported
by syzbot, more information in patch 01. Patch 02 fixes an issue with
MST's static key which Ido spotted, we can have multiple bridges with MST
and a single bridge can erroneously disable it for all.
====================
As Ido pointed out, the static key usage in MST is buggy and should use
inc/dec instead of enable/disable because we can have multiple bridges
with MST enabled which means a single bridge can disable MST for all.
Use static_branch_inc/dec to avoid that. When destroying a bridge decrement
the key if MST was enabled.
Fixes: ec7328b59176 ("net: bridge: mst: Multiple Spanning Tree (MST) mode") Reported-by: Ido Schimmel <idosch@nvidia.com> Closes: https://lore.kernel.org/netdev/20251104120313.1306566-1-razor@blackwall.org/T/#m6888d87658f94ed1725433940f4f4ebb00b5a68b Signed-off-by: Nikolay Aleksandrov <razor@blackwall.org> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20251105111919.1499702-3-razor@blackwall.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: bridge: fix use-after-free due to MST port state bypass
syzbot reported[1] a use-after-free when deleting an expired fdb. It is
due to a race condition between learning still happening and a port being
deleted, after all its fdbs have been flushed. The port's state has been
toggled to disabled so no learning should happen at that time, but if we
have MST enabled, it will bypass the port's state, that together with VLAN
filtering disabled can lead to fdb learning at a time when it shouldn't
happen while the port is being deleted. VLAN filtering must be disabled
because we flush the port VLANs when it's being deleted which will stop
learning. This fix adds a check for the port's vlan group which is
initialized to NULL when the port is getting deleted, that avoids the port
state bypass. When MST is enabled there would be a minimal new overhead
in the fast-path because the port's vlan group pointer is cache-hot.
Horatiu Vultur [Wed, 5 Nov 2025 07:49:55 +0000 (08:49 +0100)]
lan966x: Fix sleeping in atomic context
The following warning was seen when we try to connect using ssh to the device.
BUG: sleeping function called from invalid context at kernel/locking/mutex.c:575
in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 104, name: dropbear
preempt_count: 1, expected: 0
INFO: lockdep is turned off.
CPU: 0 UID: 0 PID: 104 Comm: dropbear Tainted: G W 6.18.0-rc2-00399-g6f1ab1b109b9-dirty #530 NONE
Tainted: [W]=WARN
Hardware name: Generic DT based system
Call trace:
unwind_backtrace from show_stack+0x10/0x14
show_stack from dump_stack_lvl+0x7c/0xac
dump_stack_lvl from __might_resched+0x16c/0x2b0
__might_resched from __mutex_lock+0x64/0xd34
__mutex_lock from mutex_lock_nested+0x1c/0x24
mutex_lock_nested from lan966x_stats_get+0x5c/0x558
lan966x_stats_get from dev_get_stats+0x40/0x43c
dev_get_stats from dev_seq_printf_stats+0x3c/0x184
dev_seq_printf_stats from dev_seq_show+0x10/0x30
dev_seq_show from seq_read_iter+0x350/0x4ec
seq_read_iter from seq_read+0xfc/0x194
seq_read from proc_reg_read+0xac/0x100
proc_reg_read from vfs_read+0xb0/0x2b0
vfs_read from ksys_read+0x6c/0xec
ksys_read from ret_fast_syscall+0x0/0x1c
Exception stack(0xf0b11fa8 to 0xf0b11ff0)
1fa0: 000000010000100000000008be9048d80000100000000001
1fc0: 00000001000010000000000800000003be9059200000001e0000000000000001
1fe0: 0005404cbe9048c000018684b6ec2cd8
It seems that we are using a mutex in a atomic context which is wrong.
Change the mutex with a spinlock.
Fixes: 12c2d0a5b8e2 ("net: lan966x: add ethtool configuration and statistics") Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20251105074955.1766792-1-horatiu.vultur@microchip.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Hangbin Liu [Wed, 5 Nov 2025 07:26:20 +0000 (07:26 +0000)]
bonding: fix NULL pointer dereference in actor_port_prio setting
Liang reported an issue where setting a slave’s actor_port_prio to
predefined values such as 0, 255, or 65535 would cause a system crash.
The problem occurs because in bond_opt_parse(), when the provided value
matches a predefined table entry, the function returns that table entry,
which does not contain slave information. Later, in
bond_option_actor_port_prio_set(), calling bond_slave_get_rtnl() leads
to a NULL pointer dereference.
Since actor_port_prio is defined as a u16 and initialized to the default
value of 255 in ad_initialize_port(), there is no need for the
bond_actor_port_prio_tbl. Using the BOND_OPTFLAG_RAWVAL flag is sufficient.
Fixes: 6b6dc81ee7e8 ("bonding: add support for per-port LACP actor priority") Reported-by: Liang Li <liali@redhat.com> Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> Link: https://patch.msgid.link/20251105072620.164841-1-liuhangbin@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
KSZ9477/KSZ9897 and LAN937X families of switches use a reserved multicast
address table for some specific forwarding with some multicast addresses,
like the one used in STP. The hardware assumes the host port is the last
port in KSZ9897 family and port 5 in LAN937X family. Most of the time
this assumption is correct but not in other cases like KSZ9477.
Originally the function just setups the first entry, but the others still
need update, especially for one common multicast address that is used by
PTP operation.
LAN937x also uses different register bits when accessing the reserved
table.
Fixes: 457c182af597 ("net: dsa: microchip: generic access to ksz9477 static and reserved table") Signed-off-by: Tristram Ha <tristram.ha@microchip.com> Tested-by: Łukasz Majewski <lukma@nabladev.com> Link: https://patch.msgid.link/20251105033741.6455-1-Tristram.Ha@microchip.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
idpf: add support for IDPF PCI programming interface
At present IDPF supports only 0x1452 and 0x145C as PF and VF device IDs
on our current generation hardware. Future hardware exposes a new set of
device IDs for each generation. To avoid adding a new device ID for each
generation and to make the driver forward and backward compatible,
make use of the IDPF PCI programming interface to load the driver.
Write and read the VF_ARQBAL mailbox register to find if the current
device is a PF or a VF.
PCI SIG allocated a new programming interface for the IDPF compliant
ethernet network controller devices. It can be found at:
https://members.pcisig.com/wg/PCI-SIG/document/20113
with the document titled as 'PCI Code and ID Assignment Revision 1.16'
or any latest revisions.
Tested this patch by doing a simple driver load/unload on Intel IPU E2000
hardware which supports 0x1452 and 0x145C device IDs and new hardware
which supports the IDPF PCI programming interface.
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: Pavan Kumar Linga <pavan.kumar.linga@intel.com> Signed-off-by: Madhu Chittim <madhu.chittim@intel.com> Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Tested-by: Marek Landowski <marek.landowski@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Link: https://patch.msgid.link/20251103224631.595527-1-anthony.l.nguyen@intel.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Aswin Karuvally [Mon, 3 Nov 2025 10:16:52 +0000 (11:16 +0100)]
s390/ctcm: Use info level for handshake UC_RCRESET
CTC adapter throws CTC_EVENT_UC_RCRESET (Unit check remote reset event)
during initial handshake, if the peer is not ready yet. This causes the
ctcm driver to re-attempt the handshake.
As it is normal for the event to occur during initialization, use info
instead of warn level in kernel log and NOTICE instead of ERROR level
in s390 debug feature. Also reword the log message for clarity.
====================
amd-xgbe: introduce support for ethtool selftests
This patch series introduces support for ethtool selftests, which helps
in finding the misconfiguration of HW. Makes use of network selftest
packet creation infrastructure.
Raju Rangoju [Fri, 31 Oct 2025 11:15:55 +0000 (16:45 +0530)]
amd-xgbe: add ethtool phy loopback selftest
Add support for PHY loopback testing via ethtool self-test.
The test uses phy_loopback() which enables PHY-level loopback
through the PHY driver's set_loopback callback if provided,
else uses the genphy_loopback().
Raju Rangoju [Fri, 31 Oct 2025 11:15:53 +0000 (16:45 +0530)]
amd-xgbe: introduce support ethtool selftest
Add support for ethtool selftest for MAC loopback. This includes the
sanity check and helps in finding the misconfiguration of HW. Uses the
existing selftest infrastructure to create test packets.
Raju Rangoju [Fri, 31 Oct 2025 11:18:11 +0000 (16:48 +0530)]
net: selftests: export packet creation helpers for driver use
Export the network selftest packet creation infrastructure to allow
network drivers to reuse the existing selftest framework instead of
duplicating packet creation code.
eswin,hsp-sp-csr attribute is one phandle with multiple arguments,
so the syntax should be in the form of:
items:
- items:
- description: ...
- description: ...
- description: ...
- description: ...
To align with the description of the 'eswin-sp-csr'
attribute in the mmc,usb modules, the description
of the 'eswin,hsp-sp-csr' attribute has been modified.
Fixes: 888bd0eca93c ("dt-bindings: ethernet: eswin: Document for EIC7700 SoC") Reported-by: Rob Herring (Arm) <robh@kernel.org> Closes: https://lore.kernel.org/all/176096011380.22917.1988679321096076522.robh@kernel.org/ Signed-off-by: Shangjuan Wei <weishangjuan@eswincomputing.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://patch.msgid.link/20251104073305.299-1-weishangjuan@eswincomputing.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
net: stmmac: socfpga: Add Agilex5 platform support and enhancements
This patch series adds support for the Agilex5 EMAC platform to the
dwmac-socfpga driver.
The series includes:
- Platform configuration for Agilex5 EMAC
- Enabling Time-Based Scheduling (TBS) for Tx queues 6 and 7
- Enabling TCP Segmentation Offload(TSO)
- Adding hardware-supported cross timestamping using the SMTG IP,
allowing precise synchronization between MAC and system time via
PTP_SYS_OFFSET_PRECISE.
====================
Cross timestamping is supported on Agilex5 platform with Synchronized
Multidrop Timestamp Gathering(SMTG) IP. The hardware cross-timestamp
result is made available the applications through the ioctl call
PTP_SYS_OFFSET_PRECISE, which inturn calls stmmac_getcrosststamp().
Device time is stored in the MAC Auxiliary register. The 64-bit System
time (ARM_ARCH_COUNTER) is stored in SMTG IP. SMTG IP is an MDIO device
with 0xC - 0xF MDIO register space holds 64-bit system time.
This commit is similar to following commit for Intel platforms:
Commit 341f67e424e5 ("net: stmmac: Add hardware supported cross-timestamp")
Agilex5 HPS EMAC uses the dwxgmac-3.10a IP, unlike previous socfpga
platforms which use dwmac1000 IP. Due to differences in platform
configuration, Agilex5 requires a distinct setup.
Introduce a setup_plat_dat() callback in socfpga_dwmac_ops to handle
platform-specific setup. This callback is invoked before
stmmac_dvr_probe() to ensure the platform data is correctly
configured. Also, implemented separate setup_plat_dat() callback for
current socfpga platforms and Agilex5.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@altera.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Tested-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20251101-agilex5_ext-v2-1-a6b51b4dca4d@altera.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 6 Nov 2025 02:09:23 +0000 (18:09 -0800)]
Merge tag 'wireless-next-2025-11-05' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next
Johannes Berg says:
====================
More changes from drivers are coming in, notably:
- ath10k: factory test support
- ath11k: TX power insertion support
- ath12k: BSS color change support
- iwlwifi: new sniffer API support
* tag 'wireless-next-2025-11-05' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next: (63 commits)
wifi: ath10k: use = {} to initialize bmi_target_info instead of memset
wifi: ath10k: use = {} to initialize pm_qos_request instead of memset
wifi: ath12k: unassign arvif on scan vdev create failure
wifi: ath12k: enforce vdev limit in ath12k_mac_vdev_create()
wifi: ath12k: Set EHT fixed rates for associated STAs
wifi: ath12k: add EHT rates to ath12k_mac_op_set_bitrate_mask()
wifi: ath12k: Add EHT fixed GI/LTF
wifi: ath12k: Add EHT MCS/NSS rates to Peer Assoc
wifi: ath12k: add EHT rate handling to existing set rate functions
wifi: ath12k: generalize GI and LTF fixed rate functions
wifi: ath12k: fix error handling in creating hardware group
wifi: ath12k: fix reusing m3 memory
wifi: ath12k: fix potential memory leak in ath12k_wow_arp_ns_offload()
wifi: iwlwifi: mld: add null check for kzalloc() in iwl_mld_send_proto_offload()
wifi: iwlwifi: mld: check for NULL pointer after kmalloc
wifi: iwlwifi: cfg: fix a few device names
wifi: iwlwifi: mld: Move EMLSR prints to IWL_DL_EHT
wifi: iwlwifi: disable EHT if the device doesn't allow it
wifi: iwlwifi: bump core version for BZ/SC/DR
wifi: iwlwifi: mld: use FW_CHECK on bad ROC notification
...
====================
Jakub Kicinski [Thu, 6 Nov 2025 02:04:54 +0000 (18:04 -0800)]
Merge tag 'wireless-2025-11-05' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless
Johannes Berg says:
====================
Just two small fixes:
- ath12k: revert a change that caused performance regressions
- hwsim: don't ignore netns on netlink socket matching
* tag 'wireless-2025-11-05' of https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: mac80211_hwsim: Limit destroy_on_close radio removal to netgroup
Revert "wifi: ath12k: Fix missing station power save configuration"
====================
Haotian Zhang [Wed, 5 Nov 2025 03:47:16 +0000 (11:47 +0800)]
net: wan: framer: pef2256: Switch to devm_mfd_add_devices()
The driver calls mfd_add_devices() but fails to call mfd_remove_devices()
in error paths after successful MFD device registration and in the remove
function. This leads to resource leaks where MFD child devices are not
properly unregistered.
Replace mfd_add_devices with devm_mfd_add_devices to automatically
manage the device resources.
Fixes: c96e976d9a05 ("net: wan: framer: Add support for the Lantiq PEF2256 framer") Suggested-by: Herve Codina <herve.codina@bootlin.com> Signed-off-by: Haotian Zhang <vulab@iscas.ac.cn> Acked-by: Herve Codina <herve.codina@bootlin.com> Link: https://patch.msgid.link/20251105034716.662-1-vulab@iscas.ac.cn Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Dan Carpenter [Fri, 31 Oct 2025 13:05:01 +0000 (16:05 +0300)]
net: dsa: microchip: Fix a link check in ksz9477_pcs_read()
The BMSR_LSTATUS define is 0x4 but the "p->phydev.link" variable
is a 1 bit bitfield in a u32. Since 4 doesn't fit in 0-1 range
it means that ".link" is always set to false. Add a !! to fix
this.
[Jakub: According to Maxime the phydev struct isn't really
used and we should consider removing it completely. So not
treating this as a fix.]
Jiawen Wu [Tue, 4 Nov 2025 06:23:21 +0000 (14:23 +0800)]
net: libwx: fix device bus LAN ID
The device bus LAN ID was obtained from PCI_FUNC(), but when a PF
port is passthrough to a virtual machine, the function number may not
match the actual port index on the device. This could cause the driver
to perform operations such as LAN reset on the wrong port.
Fix this by reading the LAN ID from port status register.
Dragos Tatulea [Tue, 4 Nov 2025 06:48:35 +0000 (08:48 +0200)]
net/mlx5e: SHAMPO, Fix header formulas for higher MTUs and 64K pages
The MLX5E_SHAMPO_WQ_HEADER_PER_PAGE and
MLX5E_SHAMPO_LOG_MAX_HEADER_ENTRY_SIZE macros are used directly in
several places under the assumption that there will always be more
headers per WQE than headers per page. However, this assumption doesn't
hold for 64K page sizes and higher MTUs (> 4K). This can be first
observed during header page allocation: ksm_entries will become 0 during
alignment to MLX5E_SHAMPO_WQ_HEADER_PER_PAGE.
This patch introduces 2 additional members to the mlx5e_shampo_hd struct
which are meant to be used instead of the macrose mentioned above.
When the number of headers per WQE goes below
MLX5E_SHAMPO_WQ_HEADER_PER_PAGE, clamp the number of headers per
page and expand the header size accordingly so that the headers
for one WQE cover a full page.
All the formulas are adapted to use these two new members.
Fixes: 945ca432bfd0 ("net/mlx5e: SHAMPO, Drop info array") Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/1762238915-1027590-4-git-send-email-tariqt@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Dragos Tatulea [Tue, 4 Nov 2025 06:48:34 +0000 (08:48 +0200)]
net/mlx5e: SHAMPO, Fix skb size check for 64K pages
mlx5e_hw_gro_skb_has_enough_space() uses a formula to check if there is
enough space in the skb frags to store more data. This formula is
incorrect for 64K page sizes and it triggers early GRO session
termination because the first fragment will blow up beyond
GRO_LEGACY_MAX_SIZE.
This patch adds a special case for page sizes >= GRO_LEGACY_MAX_SIZE
(64K) which uses the skb->len instead. Within this context,
the check is safe from fragment overflow because the hardware
will continuously fill the data up to the reservation size of 64K
and the driver will coalesce all data from the same page to the same
fragment. This means that the data will span one fragment or at most
two for such a large page size.
It is expected that the if statement will be optimized out as the
check is done with constants.
Dragos Tatulea [Tue, 4 Nov 2025 06:48:33 +0000 (08:48 +0200)]
net/mlx5e: SHAMPO, Fix header mapping for 64K pages
HW-GRO is broken on mlx5 for 64K page sizes. The patch in the fixes tag
didn't take into account larger page sizes when doing an align down
of max_ksm_entries. For 64K page size, max_ksm_entries is 0 which will skip
mapping header pages via WQE UMR. This breaks header-data split
and will result in the following syndrome:
Furthermore, the function that fills in WQE UMRs for the headers
(mlx5e_build_shampo_hd_umr()) only supports mapping page sizes that
fit in a single UMR WQE.
This patch goes back to the old non-aligned max_ksm_entries value and it
changes mlx5e_build_shampo_hd_umr() to support mapping a large page over
multiple UMR WQEs.
This means that mlx5e_build_shampo_hd_umr() can now leave a page only
partially mapped. The caller, mlx5e_alloc_rx_hd_mpwqe(), ensures that
there are enough UMR WQEs to cover complete pages by working on
ksm_entries that are multiples of MLX5E_SHAMPO_WQ_HEADER_PER_PAGE.
Fixes: 8a0ee54027b1 ("net/mlx5e: SHAMPO, Simplify UMR allocation for headers") Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/1762238915-1027590-2-git-send-email-tariqt@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The ICSSG driver does the initial FDB configuration which
includes setting the control registers. Other run time
management like learning is managed by the PRU's. The default
FDB hash size used by the firmware is 512 slots, which is
currently missing in the current driver. Update the driver
FDB config to include FDB hash size as well.
Please refer trm [1] 6.4.14.12.17 section on how the FDB config
register gets configured. From the table 6-1404, there is a reset
field for FDB_HAS_SIZE which is 4, meaning 1024 slots. Currently
the driver is not updating this reset value from 4(1024 slots) to
3(512 slots). This patch fixes this by updating the reset value
to 512 slots.
[1]: https://www.ti.com/lit/pdf/spruim2 Fixes: abd5576b9c57f ("net: ti: icssg-prueth: Add support for ICSSG switch firmware") Signed-off-by: Meghana Malladi <m-malladi@ti.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20251104104415.3110537-1-m-malladi@ti.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Gal Pressman [Tue, 4 Nov 2025 14:15:36 +0000 (16:15 +0200)]
net/mlx5e: Fix return value in case of module EEPROM read error
mlx5e_get_module_eeprom_by_page() has weird error handling.
First, it is treating -EINVAL as a special case, but it is unclear why.
Second, it tries to fail "gracefully" by returning the number of bytes
read even in case of an error. This results in wrongly returning
success (0 return value) if the error occurs before any bytes were
read.
Simplify the error handling by returning an error when such occurs. This
also aligns with the error handling we have in mlx5e_get_module_eeprom()
for the old API.
This fixes the following case where the query fails, but userspace
ethtool wrongly treats it as success and dumps an output:
net: gro_cells: Reduce lock scope in gro_cell_poll
One GRO-cell device's NAPI callback can nest into the GRO-cell of
another device if the underlying device is also using GRO-cell.
This is the case for IPsec over vxlan.
These two GRO-cells are separate devices. From lockdep's point of view
it is the same because each device is sharing the same lock class and so
it reports a possible deadlock assuming one device is nesting into
itself.
Hold the bh_lock only while accessing gro_cell::napi_skbs in
gro_cell_poll(). This reduces the locking scope and avoids acquiring the
same lock class multiple times.
Fixes: 25718fdcbdd2 ("net: gro_cells: Use nested-BH locking for gro_cell") Reported-by: Gal Pressman <gal@nvidia.com> Closes: https://lore.kernel.org/all/66664116-edb8-48dc-ad72-d5223696dd19@nvidia.com/ Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Link: https://patch.msgid.link/20251104153435.ty88xDQt@linutronix.de Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tim Hostetler [Tue, 4 Nov 2025 22:59:15 +0000 (14:59 -0800)]
ptp: Return -EINVAL on ptp_clock_register if required ops are NULL
ptp_clock should never be registered unless it stubs one of gettimex64()
or gettime64() and settime64(). WARN_ON_ONCE and error out if either set
of function pointers is null.
For consistency, n_alarm validation is also folded into the
WARN_ON_ONCE.
Suggested-by: Kuniyuki Iwashima <kuniyu@google.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@google.com> Reviewed-by: Harshitha Ramamurthy <hramamurthy@google.com> Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev> Signed-off-by: Tim Hostetler <thostet@google.com> Acked-by: Richard Cochran <richardcochran@gmail.com> Link: https://patch.msgid.link/20251104225915.2040080-1-thostet@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
libie: depend on DEBUG_FS when building LIBIE_FWLOG
LIBIE_FWLOG is unusable without DEBUG_FS. Mark it in Kconfig.
Fix build error on ixgbe when DEBUG_FS is not set. To not add another
layer of #if IS_ENABLED(LIBIE_FWLOG) in ixgbe fwlog code define debugfs
dentry even when DEBUG_FS isn't enabled. In this case the dummy
functions of LIBIE_FWLOG will be used, so not initialized dentry isn't a
problem.
Fixes: 641585bc978e ("ixgbe: fwlog support for e610") Reported-by: Guenter Roeck <linux@roeck-us.net> Closes: https://lore.kernel.org/lkml/f594c621-f9e1-49f2-af31-23fbcb176058@roeck-us.net/ Signed-off-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Tested-by: Rinitha S <sx.rinitha@intel.com> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Link: https://patch.msgid.link/20251104172333.752445-1-anthony.l.nguyen@intel.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Wed, 5 Nov 2025 19:15:36 +0000 (11:15 -0800)]
Merge tag 'rust-fixes-6.18' of git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux
Pull rust fixes from Miguel Ojeda:
- Fix/workaround a couple Rust 1.91.0 build issues when sanitizers are
enabled due to extra checking performed by the compiler and an
upstream issue already fixed for Rust 1.93.0
- Fix future Rust 1.93.0 builds by supporting the stabilized name for
the 'no-jump-tables' flag
- Fix a couple private/broken intra-doc links uncovered by the future
move of pin-init to 'syn'
* tag 'rust-fixes-6.18' of git://git.kernel.org/pub/scm/linux/kernel/git/ojeda/linux:
rust: kbuild: support `-Cjump-tables=n` for Rust 1.93.0
rust: kbuild: workaround `rustdoc` doctests modifier bug
rust: kbuild: treat `build_error` and `rustdoc` as kernel objects
rust: condvar: fix broken intra-doc link
rust: devres: fix private intra-doc link
Johannes Berg [Wed, 5 Nov 2025 15:28:53 +0000 (16:28 +0100)]
Merge tag 'ath-next-20251103' of git://git.kernel.org/pub/scm/linux/kernel/git/ath/ath into wireless-next
Jeff Johnson says:
==================
ath.git patches for v6.19
Highlights for some specific drivers include:
ath10k:
Add support for Factory Test TLV commands
ath11k:
Add support for Tx Power insertion
ath12k:
Add support for BSS color change
And of course there is the usual set of cleanups and bug fixes across
the entire family of "ath" drivers.
We do expect to have one more pull request before the v6.19 merge
window to pull in the refactored ath12k driver from the ath12k-ng
branch.
==================
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Martin Willi [Mon, 3 Nov 2025 08:24:36 +0000 (09:24 +0100)]
wifi: mac80211_hwsim: Limit destroy_on_close radio removal to netgroup
hwsim radios marked destroy_on_close are removed when the Netlink socket
that created them is closed. As the portid is not unique across network
namespaces, closing a socket in one namespace may remove radios in another
if it has the destroy_on_close flag set.
Instead of matching the network namespace, match the netgroup of the radio
to limit radio removal to those that have been created by the closing
Netlink socket. The netgroup of a radio identifies the network namespace
it was created in, and matching on it removes a destroy_on_close radio
even if it has been moved to another namespace.
Linus Torvalds [Wed, 5 Nov 2025 09:56:15 +0000 (18:56 +0900)]
Merge tag 'media/v6.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
- honour privacy led with pdx86/int3472
- fix invalid file access on cx18 and ivtv
- forbid remove_bufs when legacy fileio is active on videbuf2
- add an heuristic to find stream entity on uvcvideo
* tag 'media/v6.18-2' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: videobuf2: forbid remove_bufs when legacy fileio is active
media: uvcvideo: Use heuristic to find stream entity
media: v4l2-subdev / pdx86: int3472: Use "privacy" as con_id for the privacy LED
media: ivtv: Fix invalid access to file *
media: cx18: Fix invalid access to file *
Breno Leitao [Mon, 3 Nov 2025 16:38:17 +0000 (08:38 -0800)]
netpoll: Fix deadlock in memory allocation under spinlock
Fix a AA deadlock in refill_skbs() where memory allocation while holding
skb_pool->lock can trigger a recursive lock acquisition attempt.
The deadlock scenario occurs when the system is under severe memory
pressure:
1. refill_skbs() acquires skb_pool->lock (spinlock)
2. alloc_skb() is called while holding the lock
3. Memory allocator fails and calls slab_out_of_memory()
4. This triggers printk() for the OOM warning
5. The console output path calls netpoll_send_udp()
6. netpoll_send_udp() attempts to acquire the same skb_pool->lock
7. Deadlock: the lock is already held by the same CPU
This bug was exposed by commit 248f6571fd4c51 ("netpoll: Optimize skb
refilling on critical path") which removed refill_skbs() from the
critical path (where nested printk was being deferred), letting nested
printk being called from inside refill_skbs()
Refactor refill_skbs() to never allocate memory while holding
the spinlock.
Another possible solution to fix this problem is protecting the
refill_skbs() from nested printks, basically calling
printk_deferred_{enter,exit}() in refill_skbs(), then, any nested
pr_warn() would be deferred.
I prefer this approach, given I _think_ it might be a good idea to move
the alloc_skb() from GFP_ATOMIC to GFP_KERNEL in the future, so, having
the alloc_skb() outside of the lock will be necessary step.
There is a possible TOCTOU issue when checking for the pool length, and
queueing the new allocated skb, but, this is not an issue, given that
an extra SKB in the pool is harmless and it will be eventually used.
Nishanth Menon [Mon, 3 Nov 2025 16:28:11 +0000 (10:28 -0600)]
net: ethernet: ti: netcp: Standardize knav_dma_open_channel to return NULL on error
Make knav_dma_open_channel consistently return NULL on error instead
of ERR_PTR. Currently the header include/linux/soc/ti/knav_dma.h
returns NULL when the driver is disabled, but the driver
implementation does not even return NULL or ERR_PTR on failure,
causing inconsistency in the users. This results in a crash in
netcp_free_navigator_resources as followed (trimmed):
Unhandled fault: alignment exception (0x221) at 0xfffffff2
[fffffff2] *pgd=80000800207003, *pmd=82ffda003, *pte=00000000
Internal error: : 221 [#1] SMP ARM
Modules linked in:
CPU: 0 UID: 0 PID: 1 Comm: swapper/0 Not tainted 6.17.0-rc7 #1 NONE
Hardware name: Keystone
PC is at knav_dma_close_channel+0x30/0x19c
LR is at netcp_free_navigator_resources+0x2c/0x28c
[... TRIM...]
Call trace:
knav_dma_close_channel from netcp_free_navigator_resources+0x2c/0x28c
netcp_free_navigator_resources from netcp_ndo_open+0x430/0x46c
netcp_ndo_open from __dev_open+0x114/0x29c
__dev_open from __dev_change_flags+0x190/0x208
__dev_change_flags from netif_change_flags+0x1c/0x58
netif_change_flags from dev_change_flags+0x38/0xa0
dev_change_flags from ip_auto_config+0x2c4/0x11f0
ip_auto_config from do_one_initcall+0x58/0x200
do_one_initcall from kernel_init_freeable+0x1cc/0x238
kernel_init_freeable from kernel_init+0x1c/0x12c
kernel_init from ret_from_fork+0x14/0x38
[... TRIM...]
Standardize the error handling by making the function return NULL on
all error conditions. The API is used in just the netcp_core.c so the
impact is limited.
Note, this change, in effect reverts commit 5b6cb43b4d62 ("net:
ethernet: ti: netcp_core: return error while dma channel open issue"),
but provides a less error prone implementation.
Suggested-by: Simon Horman <horms@kernel.org> Suggested-by: Jacob Keller <jacob.e.keller@intel.com> Signed-off-by: Nishanth Menon <nm@ti.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20251103162811.3730055-1-nm@ti.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The historically fixed-size struct sockaddr is part of UAPI and embedded
in many existing structures. The kernel uses struct sockaddr extensively
within the kernel to represent arbitrarily sized sockaddr structures,
which caused problems with the compiler's ability to determine object
sizes correctly. The "temporary" solution was to make sockaddr explicitly
use a flexible array, but this causes problems for embedding struct
sockaddr in structures, where once again the compiler has to guess about
the size of such objects, and causes thousands of warnings under the
coming -Wflex-array-member-not-at-end warning.
Switching to sockaddr_storage internally everywhere wastes a lot of memory,
so we are left with needing two changes:
- introduction of an explicitly arbitrarily sized sockaddr struct
- switch struct sockaddr back to being fixed size
Doing the latter step requires all "arbitrarily sized" uses of struct
sockaddr to be replaced with the new struct from the first step.
So, introduce the new struct and do enough conversions that we can
switch sockaddr back to a fixed-size sa_data.
====================
Kees Cook [Tue, 4 Nov 2025 00:26:16 +0000 (16:26 -0800)]
net: Convert struct sockaddr to fixed-size "sa_data[14]"
Revert struct sockaddr from flexible array to fixed 14-byte "sa_data",
to solve over 36,000 -Wflex-array-member-not-at-end warnings, since
struct sockaddr is embedded within many network structs.
With socket/proto sockaddr-based internal APIs switched to use struct
sockaddr_unsized, there should be no more uses of struct sockaddr that
depend on reading beyond the end of struct sockaddr::sa_data that might
trigger bounds checking.
Comparing an x86_64 "allyesconfig" vmlinux build before and after this
patch showed no new "ud1" instructions from CONFIG_UBSAN_BOUNDS nor any
new "field-spanning" memcpy CONFIG_FORTIFY_SOURCE instrumentations.
Kees Cook [Tue, 4 Nov 2025 00:26:15 +0000 (16:26 -0800)]
bpf: Convert bpf_sock_addr_kern "uaddr" to sockaddr_unsized
Change struct bpf_sock_addr_kern to use sockaddr_unsized for the "uaddr"
field instead of sockaddr. This improves type safety in the BPF cgroup
socket address filtering code.
The casting in __cgroup_bpf_run_filter_sock_addr() is updated to match the
new type, removing an unnecessary cast in the initialization and updating
the conditional assignment to use the appropriate sockaddr_unsized cast.
Additionally rename the "unspec" variable to "storage" to better align
with its usage.
Kees Cook [Tue, 4 Nov 2025 00:26:14 +0000 (16:26 -0800)]
bpf: Convert cgroup sockaddr filters to use sockaddr_unsized consistently
Update BPF cgroup sockaddr filtering infrastructure to use sockaddr_unsized
consistently throughout the call chain, removing redundant explicit casts
from callers.
Kees Cook [Tue, 4 Nov 2025 00:26:13 +0000 (16:26 -0800)]
net: Convert proto callbacks from sockaddr to sockaddr_unsized
Convert struct proto pre_connect(), connect(), bind(), and bind_add()
callback function prototypes from struct sockaddr to struct sockaddr_unsized.
This does not change per-implementation use of sockaddr for passing around
an arbitrarily sized sockaddr struct. Those will be addressed in future
patches.
Additionally removes the no longer referenced struct sockaddr from
include/net/inet_common.h.
Kees Cook [Tue, 4 Nov 2025 00:26:11 +0000 (16:26 -0800)]
net: Convert proto_ops connect() callbacks to use sockaddr_unsized
Update all struct proto_ops connect() callback function prototypes from
"struct sockaddr *" to "struct sockaddr_unsized *" to avoid lying to the
compiler about object sizes. Calls into struct proto handlers gain casts
that will be removed in the struct proto conversion patch.
Kees Cook [Tue, 4 Nov 2025 00:26:10 +0000 (16:26 -0800)]
net: Convert proto_ops bind() callbacks to use sockaddr_unsized
Update all struct proto_ops bind() callback function prototypes from
"struct sockaddr *" to "struct sockaddr_unsized *" to avoid lying to the
compiler about object sizes. Calls into struct proto handlers gain casts
that will be removed in the struct proto conversion patch.
Kees Cook [Tue, 4 Nov 2025 00:26:09 +0000 (16:26 -0800)]
net: Add struct sockaddr_unsized for sockaddr of unknown length
Add flexible sockaddr structure to support addresses longer than the
traditional 14-byte struct sockaddr::sa_data limitation without
requiring the full 128-byte sa_data of struct sockaddr_storage. This
allows the network APIs to pass around a pointer to an object that
isn't lying to the compiler about how big it is, but must be accompanied
by its actual size as an additional parameter.
It's possible we may way to migrate to including the size with the
struct in the future, e.g.:
virtio-net: fix received length check in big packets
Since commit 4959aebba8c0 ("virtio-net: use mtu size as buffer length
for big packets"), when guest gso is off, the allocated size for big
packets is not MAX_SKB_FRAGS * PAGE_SIZE anymore but depends on
negotiated MTU. The number of allocated frags for big packets is stored
in vi->big_packets_num_skbfrags.
Because the host announced buffer length can be malicious (e.g. the host
vhost_net driver's get_rx_bufs is modified to announce incorrect
length), we need a check in virtio_net receive path. Currently, the
check is not adapted to the new change which can lead to NULL page
pointer dereference in the below while loop when receiving length that
is larger than the allocated one.
This commit fixes the received length check corresponding to the new
change.
Fixes: 4959aebba8c0 ("virtio-net: use mtu size as buffer length for big packets") Cc: stable@vger.kernel.org Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com> Tested-by: Lei Yang <leiyang@redhat.com> Link: https://patch.msgid.link/20251030144438.7582-1-minhquangbui99@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
net: phy: remove fixed_phy_add and first its users
fixed_phy_add() has a number of problems/disadvantages:
- It uses phy address 0 w/o checking whether a fixed phy with this
address exists already.
- A subsequent call to fixed_phy_register() would also use phy address 0,
because fixed_phy_add() doesn't mark it as used.
- fixed_phy_add() is used from platform code, therefore requires that
fixed phy code is built-in.
fixed_phy_add() has only two users
- coldfire/5272, using fec
- bcm47xx, using b44
So migrate fec and b44 to use fixed_phy_register_100fd(), afterwards
remove usage of fixed_phy_add() from the two platforms, and eventually
remove fixed_phy_add().
====================
Heiner Kallweit [Thu, 30 Oct 2025 21:46:32 +0000 (22:46 +0100)]
net: phy: fixed_phy: remove fixed_phy_add
fixed_phy_add() has a number of problems/disadvantages:
- It uses phy address 0 w/o checking whether a fixed phy with this
address exists already.
- A subsequent call to fixed_phy_register() would also use phy address 0,
because fixed_phy_add() doesn't mark it as used.
- fixed_phy_add() is used from platform code, therefore requires that
fixed_phy code is built-in.
Now that for the only two users (coldfire/5272 and bcm47xx) fixed_phy
creation has been moved to the respective ethernet driver (fec, b44),
we can remove fixed_phy_add().
Heiner Kallweit [Thu, 30 Oct 2025 21:44:35 +0000 (22:44 +0100)]
net: b44: register a fixed phy using fixed_phy_register_100fd if needed
In case of bcm47xx a fixed phy is used, which so far is created
by platform code, using fixed_phy_add(). This function has a number of
problems, therefore create a potentially needed fixed phy here, using
fixed_phy_register_100fd.
Due to lack of hardware, this is compile-tested only.
Heiner Kallweit [Thu, 30 Oct 2025 21:42:30 +0000 (22:42 +0100)]
net: fec: register a fixed phy using fixed_phy_register_100fd if needed
In case of coldfire/5272 a fixed phy is used, which so far is created
by platform code, using fixed_phy_add(). This function has a number of
problems, therefore create a potentially needed fixed phy here, using
fixed_phy_register_100fd.
Note 1: This includes a small functional change, as coldfire/5272
created a fixed phy in half-duplex mode. Likely this was by mistake,
because the fec MAC is 100FD-capable, and connection is to a switch.
Note 2: Usage of phy_find_next() makes use of the fact that dev_id can
only be 0 or 1.
Due to lack of hardware, this is compile-tested only.
Altera TSE cleanup to make sure everything is properly intialized
before registering the netdev.
When Altera TSE was converted to phylink, the PCS and phylink creation
were added after register_netdev(), which is wrong as this may race
with .ndo_open() once the netdev is registered.
This series makes so that we register the netdev once all resources are
cleanly initialised, that includes PCS and phylink creation as well as a
few other operations such as reading the IP version.
No errors were found in the wild, so this series doesn't target net, but
given that we fix some racy-ness, a point could be made to send that to
net.
This series doesn't introduce functional changes, however the internal
mii_bus for PCS configuration is renamed.
net: altera-tse: Init PCS and phylink before registering netdev
register_netdev() must be done only once all resources are ready, as
they may be used in .ndo_open() immediately upon registration.
Move the lynx PCS and phylink initialisation before registerng the
netdevice. We also remove the call to netif_carrier_off(), as phylink
takes care of that.
net: altera-tse: Don't use netdev name for the PCS mdio bus
The PCS mdio bus must be created before registering the net_device. To
do that, we musn't depend on the netdev name to create the mdio bus
name. Let's use the device's name instead.
Note that this changes the bus name in /sys/bus/mdiobus
net: altera-tse: Warn on bad revision at probe time
Instead of reading the core revision at probe time, and print a warning
for an unexecpected version at .ndo_open() time, let's print that
warning directly in .probe().
This allows getting rid of the "revision" private field, and also
prevent a potential race between reading the revision in .probe() after
netdev registration, and accessing that revision in .ndo_open().
By printing the warning after register_netdev(), we are sure that we
have a netdev name, and that we try to print the revision after having
read it from the internal registers.
Heiner Kallweit [Mon, 3 Nov 2025 22:26:49 +0000 (23:26 +0100)]
net: phy: make phy_device members pause and asym_pause bitfield bits
We can reduce the size of struct phy_device a little by switching
the type of members pause and asym_pause from int to a single bit.
As C99 is supported now, we can use type bool for the bitfield members,
what provides us with the benefit of the usual implicit bool conversions.
====================
Add driver for 1Gbe network chips from MUCSE
This patch series adds support for MUCSE RNPGBE 1Gbps PCIe Ethernet controllers
(N500/N210 series), including build infrastructure, hardware initialization,
mailbox (MBX) communication with firmware, and basic netdev registration
(Can show mac witch is got from firmware, and tx/rx will be added later).
Series breakdown (5 patches):
01/05: net: ethernet/mucse: Add build support for rnpgbe
- Kconfig/Makefile for MUCSE vendor, basic PCI probe (no netdev)
02/05: net: ethernet/mucse: Add N500/N210 chip support
- netdev allocation, BAR mapping
03/05: net: ethernet/mucse: Add basic MBX ops for PF-FW communication
- base read/write, write with poll ack, poll and read data
04/05: net: ethernet/mucse: Add FW commands (sync, reset, MAC query)
- FW sync retry logic, MAC address retrieval, reset hw with
base mbx ops in patch4
05/05: net: ethernet/mucse: Complete netdev registration
- HW reset, MAC setup, netdev_ops registration
====================
Dong Yibo [Sat, 1 Nov 2025 01:38:49 +0000 (09:38 +0800)]
net: rnpgbe: Add register_netdev
Complete the network device (netdev) registration flow for Mucse Gbe
Ethernet chips, including:
1. Hardware state initialization:
- Send powerup notification to firmware (via echo_fw_status)
- Sync with firmware
- Reset hardware
2. MAC address handling:
- Retrieve permanent MAC from firmware (via mucse_mbx_get_macaddr)
- Fallback to random valid MAC (eth_random_addr) if not valid mac
from Fw
Dong Yibo [Sat, 1 Nov 2025 01:38:48 +0000 (09:38 +0800)]
net: rnpgbe: Add basic mbx_fw support
Add fundamental firmware (FW) communication operations via PF-FW
mailbox, including:
- FW sync (via HW info query with retries)
- HW reset (post FW command to reset hardware)
- MAC address retrieval (request FW for port-specific MAC)
- Power management (powerup/powerdown notification to FW)
Vadim Fedorenko [Mon, 3 Nov 2025 17:29:02 +0000 (17:29 +0000)]
ti: netcp: convert to ndo_hwtstamp callbacks
Convert TI NetCP driver to use ndo_hwtstamp_get()/ndo_hwtstamp_set()
callbacks. The logic is slightly changed, because I believe the original
logic was not really correct. Config reading part is using the very
first module to get the configuration instead of iterating over all of
them and keep the last one as the configuration is supposed to be identical
for all modules. HW timestamp config set path is now trying to configure
all modules, but in case of error from one module it adds extack
message. This way the configuration will be as synchronized as possible.
There are only 2 modules using netcp core infrastructure, and both use
the very same function to configure HW timestamping, so no actual
difference in behavior is expected.
====================
convert drivers to use ndo_hwtstamp callbacks part 3 [part]
This patchset converts the rest of ethernet drivers to use ndo callbacks
instead ioctl to configure and report time stamping. The drivers in part
3 originally implemented only SIOCSHWTSTAMP command, but converted to
also provide configuration back to users.
====================
projects / thirdparty / kernel / linux.git / log
