The FMan driver has support for 2 MACs: mEMAC (newer, present on
Layerscape and PowerPC T series) and dTSEC/TGEC (older, present on
PowerPC P series). I only have handy access to the mEMAC, and this adds
support for MAC counters for those platforms.
MAC counters are necessary for any kind of low-level debugging, and
currently there is no mechanism to dump them.
Vladimir Oltean [Sat, 22 Nov 2025 11:55:23 +0000 (13:55 +0200)]
net: dpaa: fman_memac: complete phylink support with 2500base-x
The DPAA phylink conversion in the following commits partially developed
code for handling the 2500base-x host interface mode (called "2.5G
SGMII" in LS1043A/LS1046A reference manuals).
In principle, having phy-interface-mode = "2500base-x" and a pcsphy-handle
(unnamed or with pcs-handle-names = "sgmii") in the MAC device tree node
results in PHY_INTERFACE_MODE_2500BASEX being set in phylink_config ::
supported_interfaces, but this isn't sufficient.
Because memac_select_pcs() returns no PCS for PHY_INTERFACE_MODE_2500BASEX,
the Lynx PCS code never engages. There's a chance the PCS driver doesn't
have any configuration to change for 2500base-x fixed-link (based on
bootloader pre-initialization), but there's an even higher chance that
this is not the case, and the PCS remains misconfigured.
More importantly, memac_if_mode() does not handle
PHY_INTERFACE_MODE_2500BASEX, and it should be telling the mEMAC to
configure itself in GMII mode (which is upclocked by the PCS). Currently
it prints a WARN_ON() and returns zero, aka IF_MODE_10G (incorrect).
The additional case statement in memac_prepare() for calling
phy_set_mode_ext() does not make any difference, because there is no
generic PHY driver for the Lynx 10G SerDes from LS1043A/LS1046A. But we
add it nonetheless, for consistency.
Regarding the question "did 2500base-x ever work with the FMan mEMAC
mainline code prior to the phylink conversion?" - the answer is more
nuanced.
For context, the previous phylib-based implementation was unable to
describe the fixed-link speed as 2500, because the software PHY
implementation is limited to 1G. However, improperly describing the link
as an sgmii fixed-link with speed = <1000> would have resulted in a
functional 2.5G speed, because there is no other difference than the
SerDes lane clock net frequency (3.125 GHz for 2500base-x) - all the
other higher-level settings are the same, and the SerDes lane frequency
is currently handled by the RCW.
But this hack cannot be extended towards a phylib PHY such as Aquantia
operating in OCSGMII, because the latter requires phy-mode = "2500base-x",
which the mEMAC driver did not support prior to the phylink conversion.
So I do not really consider this a regression, just completing support
for a missing feature.
The FMan mEMAC driver sets phylink's "default_an_inband" property to
true, making it as if the device tree node had the managed =
"in-band-status" property anyway. This default made sense for SGMII,
where it was added to avoid regressions, but for 2500base-x we learned
only recently how to enable in-band autoneg:
https://lore.kernel.org/netdev/20251122113433.141930-1-vladimir.oltean@nxp.com/
so the driver needs to opt out of this default in-band enabled
behaviour, and only enable in-band based on the device tree property.
Vladimir Oltean [Sat, 22 Nov 2025 11:04:27 +0000 (13:04 +0200)]
net: phy: dp83867: implement configurability for SGMII in-band auto-negotiation
Implement the inband_caps() and config_inband() PHY driver methods, to
allow working with PCS devices that do not support or want in-band to be
used.
There is a complication due to existing logic from commit c76acfb7e19d
("net: phy: dp83867: retrigger SGMII AN when link change") which might
re-enable what dp83867_config_inband() has disabled. So we need to
modify dp83867_link_change_notify() to use phy_modify_changed() when
temporarily disabling in-band autoneg. If the return code is 0, it means
the original in-band was disabled and we need to keep it disabled.
If the return code is 1, the original was enabled and we need to
re-enable it. If negative, there was an error, which was silent before,
and remains silent now.
dp83867_config_inband() and dp83867_link_change_notify() are serialized
by the phydev->lock.
Hangbin Liu [Mon, 24 Nov 2025 02:20:55 +0000 (02:20 +0000)]
tools: ynl: add YNL test framework
Add a test framework for YAML Netlink (YNL) tools, covering both CLI and
ethtool functionality. The framework includes:
1) cli: family listing, netdev, ethtool, rt-* families, and nlctrl
operations
2) ethtool: device info, statistics, ring/coalesce/pause parameters, and
feature gettings
The current YNL syntax is a bit obscure, and end users may not always know
how to use it. This test framework provides usage examples and also serves
as a regression test to catch potential breakages caused by future changes.
Reviewed-by: Donald Hunter <donald.hunter@gmail.com> Acked-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> Link: https://patch.msgid.link/20251124022055.33389-1-liuhangbin@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Hangbin Liu [Tue, 25 Nov 2025 11:20:48 +0000 (11:20 +0000)]
netlink: specs: add big-endian byte-order for u32 IPv4 addresses
The fix commit converted several IPv4 address attributes from binary
to u32, but forgot to specify byte-order: big-endian. Without this,
YNL tools display IPv4 addresses incorrectly due to host-endian
interpretation.
Add the missing byte-order: big-endian to all affected u32 IPv4
address fields to ensure correct parsing and display.
Fixes: 1064d521d177 ("netlink: specs: support ipv4-or-v6 for dual-stack fields") Reported-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> Reviewed-by: Asbjørn Sloth Tønnesen <ast@fiberby.net> Link: https://patch.msgid.link/20251125112048.37631-1-liuhangbin@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
net: intel: migrate to .get_rx_ring_count() ethtool callback
This series migrates Intel network drivers to use the new .get_rx_ring_count()
ethtool callback introduced in commit 84eaf4359c36 ("net: ethtool: add
get_rx_ring_count callback to optimize RX ring queries").
The new callback simplifies the .get_rxnfc() implementation by removing
ETHTOOL_GRXRINGS handling and moving it to a dedicated callback. This provides
a cleaner separation of concerns and aligns these drivers with the modern
ethtool API.
The series updates the following Intel drivers:
- idpf
- igb
- igc
- ixgbevf
- fm10k
====================
Breno Leitao [Tue, 25 Nov 2025 10:19:51 +0000 (02:19 -0800)]
fm10k: 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().
This simplifies the RX ring count retrieval and aligns fm10k with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:50 +0000 (02:19 -0800)]
ixgbevf: 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().
This simplifies the RX ring count retrieval and aligns ixgbevf with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:49 +0000 (02:19 -0800)]
igc: 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().
This simplifies the RX ring count retrieval and aligns igc with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:48 +0000 (02:19 -0800)]
igb: 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().
This simplifies the RX ring count retrieval and aligns igb with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:47 +0000 (02:19 -0800)]
idpf: 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().
This simplifies the RX ring count retrieval and aligns idpf with the new
ethtool API for querying RX ring parameters.
I was not totally convinced I needed to have the lock, but, I decided to
be on the safe side and get the exact same behaviour it was before.
Breno Leitao [Tue, 25 Nov 2025 10:19:46 +0000 (02:19 -0800)]
ice: 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().
This simplifies the RX ring count retrieval and aligns ice with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:45 +0000 (02:19 -0800)]
iavf: 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().
This simplifies the RX ring count retrieval and aligns iavf with the new
ethtool API for querying RX ring parameters.
Breno Leitao [Tue, 25 Nov 2025 10:19:44 +0000 (02:19 -0800)]
i40e: 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().
This simplifies the RX ring count retrieval and aligns i40e with the new
ethtool API for querying RX ring parameters.
====================
Unify platform suspend/resume routines for PCI DWMAC glue
There are currently three PCI-based DWMAC glue drivers in tree,
stmmac_pci.c, dwmac-intel.c, and dwmac-loongson.c. Both stmmac_pci.c and
dwmac-intel.c implements the same and duplicated platform suspend/resume
routines.
This series introduces a new PCI helper library, stmmac_libpci.c,
providing a pair of helpers, stmmac_pci_plat_{suspend,resume}, and
replaces the driver-specific implementation with the helpers to reduce
code duplication. The helper will also simplify the Motorcomm DWMAC glue
driver which I'm working on.
The glue driver for Intel controllers isn't covered by the series, since
its suspend routine doesn't call pci_disable_device() and thus is a
little different from the new generic helpers.
I only have Loongson hardware on hand, thus the series is only tested on
Loongson 3A5000 machine. I could confirm the controller works after
resume, and WoL works as expected. This shouldn't break stmmac_pci.c,
either, since the new helpers have the exactly same code as the old
driver-specific suspend/resume hooks.
====================
Yao Zi [Mon, 24 Nov 2025 16:04:17 +0000 (16:04 +0000)]
net: stmmac: pci: Use generic PCI suspend/resume routines
Convert STMMAC PCI glue driver to use the generic platform
suspend/resume routines for PCI controllers, instead of implementing its
own one.
Signed-off-by: Yao Zi <ziyao@disroot.org> Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by: Yanteng Si <siyanteng@cqsoftware.com.cn> Link: https://patch.msgid.link/20251124160417.51514-4-ziyao@disroot.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Yao Zi [Mon, 24 Nov 2025 16:04:16 +0000 (16:04 +0000)]
net: stmmac: loongson: Use generic PCI suspend/resume routines
Convert glue driver for Loongson DWMAC controller to use the generic
platform suspend/resume routines for PCI controllers, instead of
implementing its own one.
Signed-off-by: Yao Zi <ziyao@disroot.org> Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Acked-by: Yanteng Si <siyanteng@cqsoftware.com.cn> Link: https://patch.msgid.link/20251124160417.51514-3-ziyao@disroot.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Yao Zi [Mon, 24 Nov 2025 16:04:15 +0000 (16:04 +0000)]
net: stmmac: Add generic suspend/resume helper for PCI-based controllers
Most glue driver for PCI-based DWMAC controllers utilize similar
platform suspend/resume routines. Add a generic implementation to reduce
duplicated code.
Signed-off-by: Yao Zi <ziyao@disroot.org> Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by: Yanteng Si <siyanteng@cqsoftware.com.cn> Link: https://patch.msgid.link/20251124160417.51514-2-ziyao@disroot.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
add hwtstamp_get callback to phy drivers
PHY drivers are able to configure HW time stamping and are not able to
report configuration back to user space. Add callback to report
configuration like it's done for net_device and add implementation to
the drivers.
====================
The driver stores configuration of TX timestamping and can technically
report it. Patch RX timestamp configuration storage to be more precise
on reporting and add callback to actually report it.
This patch refines and strengthens the statistics collection of TX queue
wake/stop events introduced by commit c39add9b2423 ("virtio_net: Add TX
stopped and wake counters").
Previously, the driver only recorded partial wake/stop statistics
for TX queues. Some wake events triggered by 'skb_xmit_done()' or resume
operations were not counted, which made the per-queue metrics incomplete.
Since the tagged commit, ice stopped respecting Rx buffer length
passed from VFs.
At that point, the buffer length was hardcoded in ice, so VFs still
worked up to some point (until, for example, a VF wanted an MTU
larger than its PF).
The next commit 93f53db9f9dc ("ice: switch to Page Pool"), broke
Rx on VFs completely since ice started accounting per-queue buffer
lengths again, but now VF queues always had their length zeroed, as
ice was already ignoring what iavf was passing to it.
Restore the line that initializes the buffer length on VF queues
basing on the virtchnl messages.
Fixes: 3a4f419f7509 ("ice: drop page splitting and recycling") Reported-by: Jakub Slepecki <jakub.slepecki@intel.com> Reviewed-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com> Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Tested-by: Jakub Slepecki <jakub.slepecki@intel.com> Link: https://patch.msgid.link/20251124170735.3077425-1-aleksander.lobakin@intel.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-Wflex-array-member-not-at-end was introduced in GCC-14, and we are
getting ready to enable it, globally.
Use the `DEFINE_RAW_FLEX()` helper for on-stack definitions of
a flexible structure where the size of the flexible-array member
is known at compile-time, and refactor the rest of the code,
accordingly.
So, with these changes, fix the following warning:
drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_io.c:163:36: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
This patch adds a new CLI argument for overriding the default
function prefix, as used for naming the doit/dumpit functions
in the generated kernel code.
When not specified the default "$(FAMILY)-nl" is used.
This can also be specified persistently in generated files:
/* YNL-ARG --function-prefix wg */
In the above example it causes the following changes:
wireguard_nl_get_device_dumpit() -> wg_get_device_dumpit()
wireguard_nl_get_device_doit() -> wg_get_device_doit()
The variable name fn_prefix, was chosen as it relates to op_prefix
which is used to prefix the UAPI commands enum entries.
Jakub Kicinski [Wed, 26 Nov 2025 03:17:23 +0000 (19:17 -0800)]
Merge branch 'ptp-ocp-a-fix-and-refactoring'
Andy Shevchenko says:
====================
ptp: ocp: A fix and refactoring
Here is the fix for incorrect use of %ptT with the associated
refactoring and additional cleanups.
Note, %ptS, which is introduced in another series that is already
applied to PRINTK tree, doesn't fit here, that's why this fix
is separated from that series.
====================
Andy Shevchenko [Mon, 24 Nov 2025 08:45:45 +0000 (09:45 +0100)]
ptp: ocp: Refactor signal_show() and fix %ptT misuse
Refactor signal_show() to avoid sequential calls to sysfs_emit*()
and use the same pattern to get the index of a signal as it's done
in signal_store().
While at it, fix wrong use of %ptT against struct timespec64.
It's kinda lucky that it worked just because the first member
there 64-bit and it's of time64_t type. Now with %ptS it may
be used correctly.
Michal Luczaj [Sun, 23 Nov 2025 21:43:59 +0000 (22:43 +0100)]
vsock/test: Extend transport change null-ptr-deref test
syzkaller reported a lockdep lock order inversion warning[1] due to
commit 687aa0c5581b ("vsock: Fix transport_* TOCTOU"). This was fixed in
commit f7c877e75352 ("vsock: fix lock inversion in
vsock_assign_transport()").
Redo syzkaller's repro by piggybacking on a somewhat related test
implemented in commit 3a764d93385c ("vsock/test: Add test for null ptr
deref when transport changes").
Heiner Kallweit [Mon, 24 Nov 2025 07:37:53 +0000 (08:37 +0100)]
r8169: improve MAC EEE handling
Let phydev->enable_tx_lpi control whether MAC enables TX LPI, instead of
enabling it unconditionally. This way TX LPI is disabled if e.g. link
partner doesn't support EEE. This helps to avoid potential issues like
link flaps.
Daniel Golle [Sat, 22 Nov 2025 13:33:47 +0000 (13:33 +0000)]
net: phy: mxl-gpy: add support for MxL86252 and MxL86282
Add PHY driver support for Maxlinear MxL86252 and MxL86282 switches.
The PHYs built-into those switches are just like any other GPY 2.5G PHYs
with the exception of the temperature sensor data being encoded in a
different way.
====================
net: enetc: add port MDIO support for both i.MX94 and i.MX95
The NETC IP has one external master MDIO interface (eMDIO) for managing
external PHYs, all ENETC ports share this eMDIO. The EMDIO function and
the ENETC port MDIO are the virtual ports of this eMDIO, ENETC can use
these virtual ports to access their PHYs. The difference is that EMDIO
function is a 'global port', it can access all the PHYs on the eMDIO, so
it provides a means for different software modules to share a single set
of MDIO signals to access their PHYs.
The ENETC port MDIO can only access its own external PHY. Furthermore,
its PHY address must be set to its corresponding LaBCR register in IERB
module, which is is a 64 KB size page containing registers that are used
for pre-boot initialization for all NETC PCIe functions. And this IERB
is owned by the host OS and it will be locked after the initialization,
so it cannot be configured at running time any more. The port MDIO can
only work properly when the PHY address accessed by it matches the value
of its corresponding LaBCR[MDIO_PHYAD_PRTAD]. Otherwise, the MDIO access
by the port MDIO will not take effect.
Note that the same PHY is either controlled by port MDIO or by the EMDIO
function. The netc-blk-ctrl driver will only set the PHY address in the
LaBCR register corresponding to the ENETC when the ENETC node contains
an mdio child node, and the ENETC driver will only create the port MDIO
bus then. An example in DTS is as follows, the EMDIO function will not\
access this PHY.
If users want to use EMDIO funtion to manage the PHY, they only need to
place the PHY node in the emdio node. The same PHY must not be placed
simultaneously within the ENETC node. An example in DTS to use EMDIO
is as below.
In the host OS, when there are multiple ENETCs, they can all access their
PHYs using their own port MDIO, or they can all access their PHYs using
the EMDIO function, or they can partially use port MDIO and partially use
the EMDIO function.
Another typical use case of port MDIO is the Jailhouse usage. An ENETC is
assigned to a guest OS. The EMDIO function will be unavailable in the
guest OS because EMDIO is controlled by the host OS. Therefore, the ENETC
can use its port MDIO to manage its external PHY in this situation. In
this use case, the host OS's root dtb will disable the ENETC node, so the
host OS's ENETC driver will not probe the ENETC and its PHY.
In addition, this series also adds the internal MDIO bus support, each
ENETC has an internal MDIO interface for managing on-die PHY (PCS) if it
has PCS layer.
====================
Wei Fang [Wed, 19 Nov 2025 10:25:57 +0000 (18:25 +0800)]
net: enetc: update the base address of port MDIO registers for ENETC v4
Each ENETC has a set of external MDIO registers to access its external
PHY based on its port EMDIO bus, these registers are used for MDIO bus
access, such as setting the PHY address, PHY register address and value,
read or write operations, C22 or C45 format, etc. The base address of
this set of registers has been modified in ENETC v4 and is different
from that in ENETC v1. So the base address needs to be updated so that
ENETC v4 can use port MDIO to manage its own external PHY.
Additionally, if ENETC has the PCS layer, it also has a set of internal
MDIO registers for managing its on-die PHY (PCS/Serdes). The base address
of this set of registers is also different from that of ENETC v1, so the
base address also needs to be updated so that ENETC v4 can support the
management of on-die PHY through the internal MDIO bus.
Wei Fang [Wed, 19 Nov 2025 10:25:56 +0000 (18:25 +0800)]
net: enetc: set external PHY address in IERB for i.MX94 ENETC
NETC IP has only one external master MDIO interface (eMDIO) for managing
the external PHYs. ENETC can use the interfaces provided by the EMDIO
function or its port MDIO to access and manage its external PHY. Both
the EMDIO function and the port MDIO are all virtual ports of the eMDIO.
The difference is that the EMDIO function is a 'global port', it can
access all the PHYs on the eMDIO, but port MDIO can only access its own
PHY. To ensure that ENETC can only access its own PHY through port MDIO,
LaBCR[MDIO_PHYAD_PRTAD] needs to be set, which represents the address of
the external PHY connected to ENETC. If the accessed PHY address is not
consistent with LaBCR[MDIO_PHYAD_PRTAD], then the MDIO access initiated
by port MDIO will be invalid.
Wei Fang [Wed, 19 Nov 2025 10:25:55 +0000 (18:25 +0800)]
net: enetc: set the external PHY address in IERB for port MDIO usage
The ENETC supports managing its own external PHY through its port MDIO
functionality. To use this function, the PHY address needs be set in the
corresponding LaBCR register in the Integrated Endpoint Register Block
(IERB), which is used for pre-boot initialization of NETC PCIe functions.
The port MDIO can only work properly when the PHY address accessed by the
port MDIO matches the corresponding LaBCR[MDIO_PHYAD_PRTAD] value.
Because the ENETC driver only registers the MDIO bus (port MDIO bus) when
it detects an MDIO child node in its node, similarly, the netc-blk-ctrl
driver only resolves the PHY address and sets it in the corresponding
LaBCR when it detects an MDIO child node in the ENETC node.
Co-developed-by: Aziz Sellami <aziz.sellami@nxp.com> Signed-off-by: Aziz Sellami <aziz.sellami@nxp.com> Signed-off-by: Wei Fang <wei.fang@nxp.com> Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com> Link: https://patch.msgid.link/20251119102557.1041881-2-wei.fang@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
Improvements over DSA conduit ethtool ops
DSA interceps 'ethtool -S eth0', where eth0 is the host port of the
switch (called 'conduit'). It does this because otherwise there is no
way to report port counters for the CPU port, which is a MAC like any
other of that switch, except Linux exposes no net_device for it, thus no
ethtool hook.
Having understood all downsides of this debugging interface, when we
need it we needed, so the proposed changes here are to make it more
useful by dumping more counters in it: not just the switch CPU port,
but all other switch ports in the tree which lack a net_device. Not
reinventing any wheel, just putting more output in an existing command.
That is patch 3/3. The other 2 are cleanup.
====================
Vladimir Oltean [Sat, 22 Nov 2025 11:23:11 +0000 (13:23 +0200)]
net: dsa: append ethtool counters of all hidden ports to conduit
Currently there is no way to see packet counters on cascade ports, and
no clarity on how the API for that would look like.
Because it's something that is currently needed, just extend the hack
where ethtool -S on the conduit interface dumps CPU port counters, and
also use it to dump counters of cascade ports.
Note that the "pXX_" naming convention changes to "sXX_pYY", to
distinguish between ports having the same index but belonging to
different switches. This has a slight chance of causing regressions to
existing tooling:
- grepping for "p04_counter_name" still works, but might return more
than one string now
- grepping for " p04_counter_name" no longer works
Vladimir Oltean [Sat, 22 Nov 2025 11:23:09 +0000 (13:23 +0200)]
net: dsa: cpu_dp->orig_ethtool_ops might be NULL
In theory this would have been seen by now, but it seems that all
drivers used as DSA conduit interfaces thus far have had ethtool_ops
set, and it's hard to even find modern Ethernet drivers (and not VF
ones) which don't use ethtool.
Here is the unfiltered list of drivers which register any sort of
net_device but don't set its ethtool_ops pointer. I don't think any of
them 'risks' being used as a DSA conduit, maybe except for moxart,
rnpbge and icssm, I'm not sure.
Alan Maguire [Fri, 21 Nov 2025 18:12:31 +0000 (18:12 +0000)]
cxgb4: Rename sched_class to avoid type clash
drivers/net/ethernet/chelsio/cxgb4/sched.h declares a sched_class
struct which has a type name clash with struct sched_class
in kernel/sched/sched.h (a type used in a field in task_struct).
When cxgb4 is a builtin we end up with both sched_class types,
and as a result of this we wind up with DWARF (and derived from
that BTF) with a duplicate incorrect task_struct representation.
When cxgb4 is built-in this type clash can cause kernel builds to
fail as resolve_btfids will fail when confused which task_struct
to use. See [1] for more details.
As such, renaming sched_class to ch_sched_class (in line with
other structs like ch_sched_flowc) makes sense.
Eric Dumazet [Fri, 21 Nov 2025 08:32:55 +0000 (08:32 +0000)]
net_sched: add qdisc_dequeue_drop() helper
Some qdisc like cake, codel, fq_codel might drop packets
in their dequeue() method.
This is currently problematic because dequeue() runs with
the qdisc spinlock held. Freeing skbs can be extremely expensive.
Add qdisc_dequeue_drop() method and a new TCQ_F_DEQUEUE_DROPS
so that these qdiscs can opt-in to defer the skb frees
after the socket spinlock is released.
TCQ_F_DEQUEUE_DROPS is an attempt to not penalize other qdiscs
with an extra cache line miss.
Eric Dumazet [Fri, 21 Nov 2025 08:32:49 +0000 (08:32 +0000)]
net_sched: add Qdisc_read_mostly and Qdisc_write groups
It is possible to reorg Qdisc to avoid always dirtying 2 cache lines in
fast path by reducing this to a single dirtied cache line.
In current layout, we change only four/six fields in the first cache line:
- q.spinlock
- q.qlen
- bstats.bytes
- bstats.packets
- some Qdisc also change q.next/q.prev
In the second cache line we change in the fast path:
- running
- state
- qstats.backlog
Add "aspeed,ast2700-mdio" compatible to the binding schema with a fallback
to "aspeed,ast2600-mdio".
Although the MDIO controller on AST2700 is functionally the same as the
one on AST2600, it's good practice to add a SoC-specific compatible for
new silicon. This allows future driver updates to handle any 2700-specific
integration issues without requiring devicetree changes or complex
runtime detection logic.
For now, the driver continues to bind via the existing
"aspeed,ast2600-mdio" compatible, so no driver changes are needed.
====================
mptcp: memcg accounting for passive sockets & backlog processing
This series is split in two: the 4 first patches are linked to memcg
accounting for passive sockets, and the rest introduce the backlog
processing. They are sent together, because the first one appeared to be
needed to get the second one fully working.
The second part includes RX path improvement built around backlog
processing. The main goals are improving the RX performances _and_
increase the long term maintainability.
- Patches 1-3: preparation work to ease the introduction of the next
patch.
- Patch 4: fix memcg accounting for passive sockets. Note that this is a
(non-urgent) fix, but it depends on material that is currently only in
net-next, e.g. commit 4a997d49d92a ("tcp: Save lock_sock() for memcg
in inet_csk_accept().").
- Patches 5-6: preparation of the stack for backlog processing, removing
assumptions that will not hold true any more after the backlog
introduction.
- Patches 7,8,10,11,12 are more cleanups that will make the backlog
patch a little less huge.
- Patch 9: somewhat an unrelated cleanup, included here not to forget
about it.
- Patches 13-14: The real work is done by them. Patch 13 introduces the
helpers needed to manipulate the msk-level backlog, and the data
struct itself, without any actual functional change. Patch 14 finally
uses the backlog for RX skb processing. Note that MPTCP can't use the
sk_backlog, as the MPTCP release callback can also release and
re-acquire the msk-level spinlock and core backlog processing works
under the assumption that such event is not possible.
A relevant point is memory accounts for skbs in the backlog. It's
somewhat "original" due to MPTCP constraints. Such skbs use space from
the incoming subflow receive buffer, do not use explicitly any forward
allocated memory, as we can't update the msk fwd mem while enqueuing,
nor we want to acquire again the ssk socket lock while processing the
skbs. Instead the msk borrows memory from the subflow and reserve it
for the backlog, see patch 5 and 14 for the gory details.
====================
Paolo Abeni [Fri, 21 Nov 2025 17:02:13 +0000 (18:02 +0100)]
mptcp: leverage the backlog for RX packet processing
When the msk socket is owned or the msk receive buffer is full,
move the incoming skbs in a msk level backlog list. This avoid
traversing the joined subflows and acquiring the subflow level
socket lock at reception time, improving the RX performances.
When processing the backlog, use the fwd alloc memory borrowed from
the incoming subflow. skbs exceeding the msk receive space are
not dropped; instead they are kept into the backlog until the receive
buffer is freed. Dropping packets already acked at the TCP level is
explicitly discouraged by the RFC and would corrupt the data stream
for fallback sockets.
Special care is needed to avoid adding skbs to the backlog of a closed
msk and to avoid leaving dangling references into the backlog
at subflow closing time.
Paolo Abeni [Fri, 21 Nov 2025 17:02:12 +0000 (18:02 +0100)]
mptcp: introduce mptcp-level backlog
We are soon using it for incoming data processing.
MPTCP can't leverage the sk_backlog, as the latter is processed
before the release callback, and such callback for MPTCP releases
and re-acquire the socket spinlock, breaking the sk_backlog processing
assumption.
Add a skb backlog list inside the mptcp sock struct, and implement
basic helper to transfer packet to and purge such list.
Packets in the backlog are memory accounted and still use the incoming
subflow receive memory, to allow back-pressure. The backlog size is
implicitly bounded to the sum of subflows rcvbuf.
When a subflow is closed, references from the backlog to such sock
are removed.
No packet is currently added to the backlog, so no functional changes
intended here.
Paolo Abeni [Fri, 21 Nov 2025 17:02:11 +0000 (18:02 +0100)]
mptcp: borrow forward memory from subflow
In the MPTCP receive path, we release the subflow allocated fwd
memory just to allocate it again shortly after for the msk.
That could increases the failures chances, especially when we will
add backlog processing, with other actions could consume the just
released memory before the msk socket has a chance to do the
rcv allocation.
Replace the skb_orphan() call with an open-coded variant that
explicitly borrows, the fwd memory from the subflow socket instead
of releasing it.
The borrowed memory does not have PAGE_SIZE granularity; rounding to
the page size will make the fwd allocated memory higher than what is
strictly required and could make the incoming subflow fwd mem
consistently negative. Instead, keep track of the accumulated frag and
borrow the full page at subflow close time.
This allow removing the last drop in the TCP to MPTCP transition and
the associated, now unused, MIB.
Paolo Abeni [Fri, 21 Nov 2025 17:02:10 +0000 (18:02 +0100)]
mptcp: handle first subflow closing consistently
Currently, as soon as the PM closes a subflow, the msk stops accepting
data from it, even if the TCP socket could be still formally open in the
incoming direction, with the notable exception of the first subflow.
The root cause of such behavior is that code currently piggy back two
separate semantic on the subflow->disposable bit: the subflow context
must be released and that the subflow must stop accepting incoming
data.
The first subflow is never disposed, so it also never stop accepting
incoming data. Use a separate bit to mark the latter status and set such
bit in __mptcp_close_ssk() for all subflows.
Beyond making per subflow behaviour more consistent this will also
simplify the next patch.
Paolo Abeni [Fri, 21 Nov 2025 17:02:07 +0000 (18:02 +0100)]
mptcp: do not miss early first subflow close event notification
The MPTCP protocol is not currently emitting the NL event when the first
subflow is closed before msk accept() time.
By replacing the in use close helper is such scenario, implicitly introduce
the missing notification. Note that in such scenario we want to be sure
that mptcp_close_ssk() will not trigger any PM work, move the msk state
change update earlier, so that the previous patch will offer such
guarantee.
Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Geliang Tang <geliang@kernel.org> Tested-by: Geliang Tang <geliang@kernel.org> Reviewed-by: Mat Martineau <martineau@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20251121-net-next-mptcp-memcg-backlog-imp-v1-8-1f34b6c1e0b1@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni [Fri, 21 Nov 2025 17:02:06 +0000 (18:02 +0100)]
mptcp: ensure the kernel PM does not take action too late
The PM hooks can currently take place when the msk is already shutting
down. Subflow creation will fail, thanks to the existing check at join
time, but we can entirely avoid starting the to be failed operations.
Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Geliang Tang <geliang@kernel.org> Tested-by: Geliang Tang <geliang@kernel.org> Reviewed-by: Mat Martineau <martineau@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20251121-net-next-mptcp-memcg-backlog-imp-v1-7-1f34b6c1e0b1@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni [Fri, 21 Nov 2025 17:02:05 +0000 (18:02 +0100)]
mptcp: cleanup fallback dummy mapping generation
MPTCP currently access ack_seq outside the msk socket log scope to
generate the dummy mapping for fallback socket. Soon we are going
to introduce backlog usage and even for fallback socket the ack_seq
value will be significantly off outside of the msk socket lock scope.
Avoid relying on ack_seq for dummy mapping generation, using instead
the subflow sequence number. Note that in case of disconnect() and
(re)connect() we must ensure that any previous state is re-set.
Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Geliang Tang <geliang@kernel.org> Tested-by: Geliang Tang <geliang@kernel.org> Reviewed-by: Mat Martineau <martineau@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20251121-net-next-mptcp-memcg-backlog-imp-v1-6-1f34b6c1e0b1@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni [Fri, 21 Nov 2025 17:02:04 +0000 (18:02 +0100)]
mptcp: cleanup fallback data fin reception
MPTCP currently generate a dummy data_fin for fallback socket
when the fallback subflow has completed data reception using
the current ack_seq.
We are going to introduce backlog usage for the msk soon, even
for fallback sockets: the ack_seq value will not match the most recent
sequence number seen by the fallback subflow socket, as it will ignore
data_seq sitting in the backlog.
Instead use the last map sequence number to set the data_fin,
as fallback (dummy) map sequences are always in sequence.
Reviewed-by: Geliang Tang <geliang@kernel.org> Tested-by: Geliang Tang <geliang@kernel.org> Signed-off-by: Paolo Abeni <pabeni@redhat.com> Reviewed-by: Mat Martineau <martineau@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20251121-net-next-mptcp-memcg-backlog-imp-v1-5-1f34b6c1e0b1@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni [Fri, 21 Nov 2025 17:02:03 +0000 (18:02 +0100)]
mptcp: fix memcg accounting for passive sockets
The passive sockets never got proper memcg accounting: the msk
socket is associated with the memcg at accept time, but the
passive subflows never got it right.
At accept time, traverse the subflows list and associate each of them
with the msk memcg, and try to do the same at join completion time, if
the msk has been already accepted.
Paolo Abeni [Fri, 21 Nov 2025 17:02:00 +0000 (18:02 +0100)]
net: factor-out _sk_charge() helper
Move out of __inet_accept() the code dealing charging newly
accepted socket to memcg. MPTCP will soon use it to on a per
subflow basis, in different contexts.
ipvlan_core.c:56: warning: incorrect type in argument 1
(different base types) expected unsigned int [usertype] a
got restricted __be32 const [usertype] s_addr
Breno Leitao [Fri, 21 Nov 2025 17:02:36 +0000 (09:02 -0800)]
net: mvpp2: 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() for the mvpp2 driver.
This simplifies the RX ring count retrieval and aligns mvpp2 with the new
ethtool API for querying RX ring parameters, while keeping the other
rxnfc handlers (GRXCLSRLCNT, GRXCLSRULE, GRXCLSRLALL) intact.
Breno Leitao [Fri, 21 Nov 2025 17:02:35 +0000 (09:02 -0800)]
net: mvneta: convert to use .get_rx_ring_count
Convert the mvneta driver to use the new .get_rx_ring_count ethtool
operation instead of implementing .get_rxnfc solely for handling
ETHTOOL_GRXRINGS command. This simplifies the code by removing the
switch statement and replacing it with a direct return of the queue
count.
The new callback provides the same functionality in a more direct way,
following the ongoing ethtool API modernization.
Breno Leitao [Fri, 21 Nov 2025 09:59:23 +0000 (01:59 -0800)]
net: hyperv: convert to use .get_rx_ring_count
Convert the hyperv netvsc driver to use the new .get_rx_ring_count
ethtool operation instead of implementing .get_rxnfc solely for handling
ETHTOOL_GRXRINGS command. This simplifies the code by replacing the
switch statement with a direct return of the queue count.
The new callback provides the same functionality in a more direct way,
following the ongoing ethtool API modernization.
projects / thirdparty / kernel / linux.git / log
