stmmac has had a long history of problems with resuming, illustrated by
reset failure due to the receive clock not running.
Several attempts have been attempted over the years to address this
issue, such as moving phylink_start() (now phylink_resume()) super
early in stmmac_resume() in commit 90702dcd19c0 ("net: stmmac: fix MAC
not working when system resume back with WoL a ctive.") However, this
has the downside that stmmac_mac_link_up() can (and demonstrably is)
called before or during the driver initialisation in another thread.
This can cause issues as packets could begin to be queued, and the
transmit/receive enable bits will be set before any initialisation has
been done.
Another attempt is used by dwmac-socfpga.c in commit 2d871aa07136 ("net:
stmmac: add platform init/exit for Altera's ARM socfpga") which
pre-dates the above commit.
Neither of these two approaches consider the effect of EEE with a PHY
that supports receive clock-stop and has that feature enabled (which
the stmmac driver does enable). If the link is up, then there is the
possibility for the receive path to be in low-power mode, and the PHY
may stop its receive clock.
This series addresses these issues by (each is not necessarily a
separate patch):
1) introducing phylink_prepare_resume(), which can be used by MAC
drivers to ensure that the PHY is resumed prior to doing any
re-initialisation work. This call is added to stmmac_resume().
2) moving phylink_resume() after all re-initialisation has completed,
thereby ensuring that the hardware is ready to be enabled for
packet reception/transmission.
3) with (1) and (2) addressed, the need for socfpga to have a private
work-around is no longer necessary, so it is removed.
4) introducing phylink functions to block/unblock the receive clock-
stop at the PHY. As these require PHY access over the MDIO bus,
they can sleep, so are not suitable for atomic access.
5) the stmmac hardware requires the receive clock to be running for
reset to complete. Depending on synthesis options, this requirement
may also extend to writing various registers as well, e.g. setting
the MAC address, writing some of the vlan registers, etc. Full
details are in the databook.
We add blocking/unblocking of the PHY receive clock-stop around
parts of the main stmmac driver where we have a context that we
can sleep. These are wrapped with the new phylink functions.
However, depending on synthesis options, there could be other
places where the net core calls the driver with a BH-disabled
context where we can't sleep, and thus can't block the PHY from
disabling its receive clock. These are documented with FIXME
comments.
Given the last paragraph above, I am wondering whether a better
approach would be to ensure that receive clock-stop is always disabled
at the PHY with stmmac. From what I can see, implementations do not
document to this level of detail, which makes it difficult to tell
which registers require the receive clock to be running to behave
correctly.
This patch series has been tested on the Tegra194 Jetson Xavier NX
board kindly donated by NVidia, with two additional patches that are
pending in patchwork - the first is required to have EEE's LPI mode
passed through to the MAC on this platform to allow testing under
PHY clock-stop scenarios. The second is a bug fix for PHYLIB and
makes "eee off" functional, but should not affect this series.
All patches on top of net-next commit f749448ce9f1 ("Merge branch
'net-mlx5-hw-steering-cleanups'")
The DesignWare core requires the receive clock to be running during
certain operations. Ensure that we block PHY RXC clock-stop during
these operations.
This is a best-efforts change - not everywhere can be covered by this
because of net's core locking, which means we can't access the MDIO
bus to configure the PHY to disable RXC clock-stop in certain areas.
These are marked with FIXME comments.
net: phylink: add functions to block/unblock rx clock stop
Some MACs require the PHY receive clock to be running to complete setup
actions. This may fail if the PHY has negotiated EEE, the MAC supports
receive clock stop, and the link has entered LPI state. Provide a pair
of APIs that MAC drivers can use to temporarily block the PHY disabling
the receive clock.
As the previous commit addressed DWGMAC resuming with a PHY in
suspended state, there is now no need for socfpga to work around
this. Remove this code.
The Synopsys Designware GMAC core databook requires all clocks to be
active in order to complete software reset, which we perform during
resume.
However, IEEE 802.3 allows a PHY to stop its clocks when placed in
low-power mode, which happens when the system is suspended and WoL
is not enabled.
As an attempt to work around this, commit 36d18b5664ef ("net: stmmac:
start phylink instance before stmmac_hw_setup()") started phylink
early, but this has the side effect that the mac_link_up() method may
be called before or during the initialisation of GMAC hardware.
We also have the socfpga glue driver directly calling phy_resume()
also as an attempt to work around this.
In a previous commit, phylink_prepare_resume() has been introduced
to give MAC drivers a way to ensure that the PHY is resumed prior to
their initialisation of their MAC hardware. This commit adds the call,
and moves the phylink_resume() call back to where it should be before
the aforementioned commit.
When the system is suspended, the PHY may be placed in low-power mode
by setting the BMCR 0.11 Power down bit. IEEE 802.3 states that the
behaviour of the PHY in this state is implementation specific, and
the PHY is not required to meet the RX_CLK and TX_CLK requirements.
Essentially, this means that a PHY may stop the clocks that it is
generating while in power down state.
However, MACs exist which require the clocks from the PHY to be running
in order to properly resume. phylink_prepare_resume() provides them
with a way to clear the Power down bit early.
Note, however, that IEEE 802.3 gives PHYs up to 500ms grace before the
transmit and receive clocks meet the requirements after clearing the
power down bit.
Add a resume preparation function, which will ensure that the receive
clock from the PHY is appropriately configured while resuming.
Jakub Kicinski [Tue, 25 Mar 2025 14:55:13 +0000 (07:55 -0700)]
Merge branch 'sfc-devlink-flash-for-x4'
Edward Cree says:
====================
sfc: devlink flash for X4
Updates to support devlink flash on X4 NICs.
Patch #2 is needed for NVRAM_PARTITION_TYPE_AUTO, and patch #1 is
needed because the latest MCDI headers from firmware no longer
include MDIO read/write commands.
Edward Cree [Thu, 20 Mar 2025 17:57:12 +0000 (17:57 +0000)]
sfc: support X4 devlink flash
Unlike X2 and EF100, we do not attempt to parse the firmware file to
find an image within it; we simply hand the entire file to the MC,
which is responsible for understanding any container formats we might
use and validating that the firmware file is applicable to this NIC.
Edward Cree [Thu, 20 Mar 2025 17:57:10 +0000 (17:57 +0000)]
sfc: rip out MDIO support
Unlike Siena, no EF10 board ever had an external PHY, and consequently
MDIO handling isn't even built into the firmware. Since Siena has
been split out into its own driver, the MDIO code can be deleted from
the sfc driver.
Eric Dumazet [Thu, 20 Mar 2025 10:14:34 +0000 (10:14 +0000)]
net: reorganize IP MIB values (II)
Commit 14a196807482 ("net: reorganize IP MIB values") changed
MIB values to group hot fields together.
Since then 5 new fields have been added without caring about
data locality.
This patch moves IPSTATS_MIB_OUTPKTS, IPSTATS_MIB_NOECTPKTS,
IPSTATS_MIB_ECT1PKTS, IPSTATS_MIB_ECT0PKTS, IPSTATS_MIB_CEPKTS
to the hot portion of per-cpu data.
====================
nexthop: Convert RTM_{NEW,DEL}NEXTHOP to per-netns RTNL.
Patch 1 - 5 move some validation for RTM_NEWNEXTHOP so that it can be
called without RTNL.
Patch 6 & 7 converts RTM_NEWNEXTHOP and RTM_DELNEXTHOP to per-netns RTNL.
Note that RTM_GETNEXTHOP and RTM_GETNEXTHOPBUCKET are not touched in
this series.
rtm_get_nexthop() can be easily converted to RCU, but rtm_dump_nexthop()
needs more work due to the left-to-right rbtree walk, which looks prone
to node deletion and tree rotation without a retry mechanism.
Jianbo Liu [Wed, 19 Mar 2025 14:03:03 +0000 (16:03 +0200)]
net/mlx5e: TC, Don't offload CT commit if it's the last action
For CT action with commit argument, it's usually followed by the
forward action, either to the output netdev or next chain. The default
behavior for software is to drop by setting action attribute to
TC_ACT_SHOT instead of TC_ACT_PIPE if it's the last action. But driver
can't handle it, so block the offload for such case.
Paul Blakey [Wed, 19 Mar 2025 14:03:02 +0000 (16:03 +0200)]
net/mlx5e: CT: Filter legacy rules that are unrelated to nic
In nic mode CT setup where we do hairpin between the two
nics, both nics register to the same flow table (per zone),
and try to offload all rules on it.
Instead, filter the rules that originated from the relevant nic
(so only one side is offloaded for each nic).
Shay Drory [Wed, 19 Mar 2025 14:03:01 +0000 (16:03 +0200)]
net/mlx5: Update pfnum retrieval for devlink port attributes
Align mlx5 driver usage of 'pfnum' with the documentation clarification
introduced in commit bb70b0d48d8e ("devlink: Improve the port attributes
description").
Amir Tzin [Wed, 19 Mar 2025 14:03:00 +0000 (16:03 +0200)]
net/mlx5: fw reset, check bridge accessibility at earlier stage
Currently, mlx5_is_reset_now_capable() checks whether the pci bridge is
accessible only on bridge hot plug capability check. If the pci bridge
is not accessible, reset now will fail regardless of bridge hotplug
capability. Move this check to function mlx5_is_reset_now_capable()
which, in such case, aborts the reset and does so in the request phase
instead of the reset now phase.
Signed-off-by: Aya Levin <ayal@nvidia.com> Signed-off-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Amir Tzin <amirtz@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Link: https://patch.msgid.link/1742392983-153050-3-git-send-email-tariqt@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Mark Bloch [Wed, 19 Mar 2025 14:02:59 +0000 (16:02 +0200)]
net/mlx5: Lag, use port selection tables when available
As queue affinity is being deprecated and will no longer be supported
in the future, Always check for the presence of the port selection
namespace. When available, leverage it to distribute traffic
across the physical ports via steering, ensuring compatibility with
future NICs.
Tariq Toukan [Wed, 19 Mar 2025 13:42:26 +0000 (15:42 +0200)]
net/mlx5e: TX, Utilize WQ fragments edge for multi-packet WQEs
For simplicity reasons, the driver avoids crossing work queue fragment
boundaries within the same TX WQE (Work-Queue Element). Until today, as
the number of packets in a TX MPWQE (Multi-Packet WQE) descriptor is not
known in advance, the driver pre-prepared contiguous memory for the
largest possible WQE. For this, when getting too close to the fragment
edge, having no room for the largest WQE possible, the driver was
filling the fragment remainder with NOP descriptors, aligning the next
descriptor to the beginning of the next fragment.
Generating and handling these NOPs wastes resources, like: CPU cycles,
work-queue entries fetched to the device, and PCI bandwidth.
In this patch, we replace this NOPs filling mechanism in the TX MPWQE
flow. Instead, we utilize the remaining entries of the fragment with a
TX MPWQE. If this room turns out to be too small, we simply open an
additional descriptor starting at the beginning of the next fragment.
Performance benchmark:
uperf test, single server against 3 clients.
TCP multi-stream, bidir, traffic profile "2x350B read, 1400B write".
Bottleneck is in inbound PCI bandwidth (device POV).
Jing Su [Wed, 19 Mar 2025 08:57:51 +0000 (16:57 +0800)]
dql: Fix dql->limit value when reset.
Executing dql_reset after setting a non-zero value for limit_min can
lead to an unreasonable situation where dql->limit is less than
dql->limit_min.
For instance, after setting
/sys/class/net/eth*/queues/tx-0/byte_queue_limits/limit_min,
an ifconfig down/up operation might cause the ethernet driver to call
netdev_tx_reset_queue, which in turn invokes dql_reset.
In this case, dql->limit is reset to 0 while dql->limit_min remains
non-zero value, which is unexpected. The limit should always be
greater than or equal to limit_min.
Dmitry Safonov [Wed, 19 Mar 2025 03:13:38 +0000 (03:13 +0000)]
selftests/net: Print the testing side in unsigned-md5
As both client and server print the same test name on failure or pass,
add "[server]" so that it's more obvious from a log which side printed
"ok" or "not ok".
Dmitry Safonov [Wed, 19 Mar 2025 03:13:37 +0000 (03:13 +0000)]
selftests/net: Add mixed select()+polling mode to TCP-AO tests
Currently, tcp_ao tests have two timeouts: TEST_RETRANSMIT_SEC and
TEST_TIMEOUT_SEC [by default 1 and 5 seconds]. The first one,
TEST_RETRANSMIT_SEC is used for operations that are expected to succeed
in order for a test to pass. It is usually not consumed and exists only
to avoid indefinite test run if the operation didn't complete.
The second one, TEST_RETRANSMIT_SEC exists for the tests that checking
operations, that are expected to fail/timeout. It is shorter as it is
fully consumed, with an expectation that if operation didn't succeed
during that period, it will timeout. And the related test that expects
the timeout is passing. The actual operation failure is then
cross-verified by other means like counters checks.
The issue with TEST_RETRANSMIT_SEC timeout is that 1 second is the exact
initial TCP timeout. So, in case the initial segment gets lost (quite
unlikely on local veth interface between two net namespaces, yet happens
in slow VMs), the retransmission never happens and as a result, the test
is not actually testing the functionality. Which in the end fails
counters checks.
As I want tcp_ao selftests to be fast and finishing in a reasonable
amount of time on manual run, I didn't consider increasing
TEST_RETRANSMIT_SEC.
Rather, initially, BPF_SOCK_OPS_TIMEOUT_INIT looked promising as a lever
to make the initial TCP timeout shorter. But as it's not a socket bpf
attached thing, but sock_ops (attaches to cgroups), the selftests would
have to use libbpf, which I wanted to avoid if not absolutely required.
Instead, use a mixed select() and counters polling mode with the longer
TEST_TIMEOUT_SEC timeout to detect running-away failed tests. It
actually not only allows losing segments and succeeding after
the previous TEST_RETRANSMIT_SEC timeout was consumed, but makes
the tests expecting timeout/failure pass faster.
The only test case taking longer (TEST_TIMEOUT_SEC) now is connect-deny
"wrong snd id", which checks for no key on SYN-ACK for which there is no
counter in the kernel (see tcp_make_synack()). Yet it can be speed up
by poking skpair from the trace event (see trace_tcp_ao_synack_no_key).
Fixes: ed9d09b309b1 ("selftests/net: Add a test for TCP-AO keys matching") Reported-by: Jakub Kicinski <kuba@kernel.org> Closes: https://lore.kernel.org/netdev/20241205070656.6ef344d7@kernel.org/ Signed-off-by: Dmitry Safonov <0x7f454c46@gmail.com> Link: https://patch.msgid.link/20250319-tcp-ao-selftests-polling-v2-4-da48040153d1@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Dmitry Safonov [Wed, 19 Mar 2025 03:13:36 +0000 (03:13 +0000)]
selftests/net: Fetch and check TCP-MD5 counters
There are related TCP-MD5 <=> TCP and TCP-MD5 <=> TCP-AO tests
that can benefit from checking the related counters, not only from
validating operations timeouts.
It also prepares the code for introduction of mixed select()+poll mode,
see the follow-up patches.
Dmitry Safonov [Wed, 19 Mar 2025 03:13:35 +0000 (03:13 +0000)]
selftests/net: Provide tcp-ao counters comparison helper
Rename __test_tcp_ao_counters_cmp() into test_assert_counters_ao() and
test_tcp_ao_key_counters_cmp() into test_assert_counters_key() as they
are asserts, rather than just compare functions.
Provide test_cmp_counters() helper, that's going to be used to compare
ao_info and netns counters as a stop condition for polling the sockets.
For the history, I think the initial format was to emphasize the absence
of flags as well as their presence (!R meant no RST flag). But looking
again, it's just unreadable and hard to understand.
Make it the standard/expected one.
====================
net: xdp: Add missing metadata support for some xdp drvs
Introduce missing metadata support for some xdp drivers setting metadata
size building the skb from xdp_buff.
Please note most of the drivers are just compile tested.
Lorenzo Bianconi [Tue, 18 Mar 2025 11:46:10 +0000 (12:46 +0100)]
net: mana: Add metadata support for xdp mode
Set metadata size building the skb from xdp_buff in mana driver.
mana driver sets xdp headroom to XDP_PACKET_HEADROOM so the headroom is
large enough to contain xdp_frame and xdp metadata.
Please note this patch is just compiled tested.
Lorenzo Bianconi [Tue, 18 Mar 2025 11:46:09 +0000 (12:46 +0100)]
net: ethernet: mediatek: Add metadata support for xdp mode
Set metadata size building the skb from xdp_buff in mediatek driver.
mtk_eth_soc driver sets xdp headroom to XDP_PACKET_HEADROOM so the
headroom is large enough to contain xdp_frame and xdp metadata.
Please note this patch is just compiled tested.
Simon Horman [Tue, 18 Mar 2025 13:53:34 +0000 (13:53 +0000)]
net: tulip: avoid unused variable warning
There is an effort to achieve W=1 kernel builds without warnings.
As part of that effort Helge Deller highlighted the following warnings
in the tulip driver when compiling with W=1 and CONFIG_TULIP_MWI=n:
.../tulip_core.c: In function ‘tulip_init_one’:
.../tulip_core.c:1309:22: warning: variable ‘force_csr0’ set but not used
This patch addresses that problem using IS_ENABLED(). This approach has
the added benefit of reducing conditionally compiled code. And thus
increasing compile coverage. E.g. for allmodconfig builds which enable
CONFIG_TULIP_MWI.
net/unix/*.c include many unnecessary header files (rtnetlink.h,
netdevice.h, etc).
Let's clean them up.
af_unix.c:
+uapi/linux/sockios.h : Only exist under include/uapi
+uapi/linux/termios.h : Only exist under include/uapi
-linux/freezer.h : No longer use freezable_schedule_timeout()
-linux/in.h : No ipv4_is_XXX() etc
-linux/module.h : No longer support CONFIG_UNIX=m
-linux/netdevice.h : No dev used
-linux/rtnetlink.h : Not part of rtnetlink API
-linux/signal.h : signal_pending() is defined in sched/signal.h
-linux/stat.h : No struct stat used
-net/checksum.h : CHECKSUM_UNNECESSARY is defined in skbuff.h
diag.c:
+linux/dcache.h : struct dentry in sk_diag_dump_vfs()
+linux/user_namespace.h : struct user_namespace in sk_diag_dump_uid()
+uapi/linux/unix_diag.h : Only exist under include/uapi/
-linux/file.h : No fget() etc
-linux/kernel.h : No cond_resched() etc
-linux/netdevice.h : No dev used
-linux/proc_fs.h : No procfs provided
-linux/string.h : No memcpy(), kmemdup(), etc
Jason Xing [Mon, 17 Mar 2025 12:03:14 +0000 (20:03 +0800)]
tcp: support TCP_DELACK_MAX_US for set/getsockopt use
Support adjusting/reading delayed ack max for socket level by using
set/getsockopt().
This option aligns with TCP_BPF_DELACK_MAX usage. Considering that bpf
option was implemented before this patch, so we need to use a standalone
new option for pure tcp set/getsockopt() use.
Add WRITE_ONCE/READ_ONCE() to prevent data-race if setsockopt()
happens to write one value to icsk_delack_max while icsk_delack_max is
being read.
Jason Xing [Mon, 17 Mar 2025 12:03:13 +0000 (20:03 +0800)]
tcp: support TCP_RTO_MIN_US for set/getsockopt use
Support adjusting/reading RTO MIN for socket level by using set/getsockopt().
This new option has the same effect as TCP_BPF_RTO_MIN, which means it
doesn't affect RTAX_RTO_MIN usage (by using ip route...). Considering that
bpf option was implemented before this patch, so we need to use a standalone
new option for pure tcp set/getsockopt() use.
When the socket is created, its icsk_rto_min is set to the default
value that is controlled by sysctl_tcp_rto_min_us. Then if application
calls setsockopt() with TCP_RTO_MIN_US flag to pass a valid value, then
icsk_rto_min will be overridden in jiffies unit.
This patch adds WRITE_ONCE/READ_ONCE to avoid data-race around
icsk_rto_min.
====================
mlxsw: Add VXLAN to the same hardware domain as physical bridge ports
Amit Cohen writes:
Packets which are trapped to CPU for forwarding in software data path
are handled according to driver marking of skb->offload_{,l3}_fwd_mark.
Packets which are marked as L2-forwarded in hardware, will not be flooded
by the bridge to bridge ports which are in the same hardware domain as the
ingress port.
Currently, mlxsw does not add VXLAN bridge ports to the same hardware
domain as physical bridge ports despite the fact that the device is able
to forward packets to and from VXLAN tunnels in hardware. In some
scenarios this can result in remote VTEPs receiving duplicate packets.
To solve such packets duplication, add VXLAN bridge ports to the same
hardware domain as other bridge ports.
One complication is ARP suppression which requires the local VTEP to avoid
flooding ARP packets to remote VTEPs if the local VTEP is able to reply on
behalf of remote hosts. This is currently implemented by having the device
flood ARP packets in hardware and trapping them during VXLAN encapsulation,
but marking them with skb->offload_fwd_mark=1 so that the bridge will not
re-flood them to physical bridge ports.
The above scheme will break when VXLAN bridge ports are added to the same
hardware domain as physical bridge ports as ARP packets that cannot be
suppressed by the bridge will not be able to egress the VXLAN bridge ports
due to hardware domain filtering. This is solved by trapping ARP packets
when they enter the device and not marking them as being forwarded in
hardware.
Patch set overview:
Patch #1 sets hardware to trap ARP packets at layer 2
Patches #2-#4 are preparations for setting hardwarwe domain of VXLAN
Patch #5 sets hardware domain of VXLAN
Patch #6 extends VXLAN flood test to verify that this set solves the
packets duplication
====================
Amit Cohen [Mon, 17 Mar 2025 17:37:30 +0000 (18:37 +0100)]
mlxsw: Add VXLAN bridge ports to same hardware domain as physical bridge ports
When hardware floods packets to bridge ports, but flooding to VXLAN bridge
port fails during encapsulation to one of the remote VTEPs, the packets are
trapped to CPU. In such case, the packets are marked with
skb->offload_fwd_mark, which means that packet was L2-forwarded in
hardware. Software data path repeats flooding, but packets which are
marked with skb->offload_fwd_mark will not be flooded by the bridge to
bridge ports which are in the same hardware domain as the ingress port.
Currently, mlxsw does not add VXLAN bridge ports to the same hardware
domain as physical bridge ports despite the fact that the device is able
to forward packets to and from VXLAN tunnels in hardware. In some scenarios
(as mentioned above) this can result in remote VTEPs receiving duplicate
packets. The packets are first flooded by hardware and after an
encapsulation failure, they are flooded again to all remote VTEPs by
software.
Solve this by adding VXLAN bridge ports to the same hardware domain as
physical bridge ports, so then nbp_switchdev_allowed_egress() will return
false also for VXLAN, and packets will not be sent twice from VXLAN device.
switchdev_bridge_port_offload() should get vxlan_dev not as const, so
some changes are required. Call switchdev API from
mlxsw_sp_bridge_vxlan_{join,leave}() which handle offload configurations.
Reported-by: Vladimir Oltean <olteanv@gmail.com> Closes: https://lore.kernel.org/all/20250210152246.4ajumdchwhvbarik@skbuf/ Reported-by: Vladyslav Mykhaliuk <vmykhaliuk@nvidia.com> Signed-off-by: Amit Cohen <amcohen@nvidia.com> Reviewed-by: Petr Machata <petrm@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Petr Machata <petrm@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/7279056843140fae3a72c2d204c7886b79d03899.1742224300.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Next patch will call __mlxsw_sp_bridge_vxlan_leave() from
mlxsw_sp_bridge_vxlan_join() as part of error flow, move the function to
be able to call the second one.
Amit Cohen [Mon, 17 Mar 2025 17:37:28 +0000 (18:37 +0100)]
mlxsw: spectrum_switchdev: Add an internal API for VXLAN leave
There is asymmetry in how the VXLAN join and leave functions are used.
The join function (mlxsw_sp_bridge_vxlan_join()) is only called in
response to netdev events (e.g., VXLAN device joining a bridge), but the
leave function is also called in response to switchdev events (e.g.,
VLAN configuration on top of the VXLAN device) in order to invalidate
VNI to FID mappings.
This asymmetry will cause problems when the functions will be later
extended to mark VXLAN bridge ports as offloaded or not.
Therefore, create an internal function (__mlxsw_sp_bridge_vxlan_leave())
that is used to invalidate VNI to FID mappings and call it from
mlxsw_sp_bridge_vxlan_leave() which will only be invoked in response to
netdev events, like mlxsw_sp_bridge_vxlan_join().
Amit Cohen [Mon, 17 Mar 2025 17:37:27 +0000 (18:37 +0100)]
mlxsw: spectrum: Call mlxsw_sp_bridge_vxlan_{join, leave}() for VLAN-aware bridge
mlxsw_sp_bridge_vxlan_{join,leave}() are not called when a VXLAN device
joins or leaves a VLAN-aware bridge. As mentioned in the comment - when the
bridge is VLAN-aware, the VNI of the VXLAN device needs to be mapped to a
VLAN, but at this point no VLANs are configured on the VxLAN device. This
means that we can call the APIs, but there is no point to do that, as they
do not configure anything in such cases.
Next patch will extend mlxsw_sp_bridge_vxlan_{join,leave}() to set hardware
domain for VXLAN, this should be done also when a VXLAN device joins or
leaves a VLAN-aware bridge. Call the APIs, which for now do not do anything
in these flows.
Align the call to mlxsw_sp_bridge_vxlan_leave() to be called like
mlxsw_sp_bridge_vxlan_join(), only in case that the VXLAN device is up,
so move the check to be done before calling
mlxsw_sp_bridge_vxlan_{join,leave}(). This does not change the existing
behavior, as there is a similar check inside mlxsw_sp_bridge_vxlan_leave().
Amit Cohen [Mon, 17 Mar 2025 17:37:26 +0000 (18:37 +0100)]
mlxsw: Trap ARP packets at layer 2 instead of layer 3
Next patch will set the same hardware domain for all bridge ports,
including VXLAN, to prevent packets from being forwarded by software when
they were already forwarded by hardware.
ARP packets are not flooded by hardware to VXLAN, so software should handle
such flooding. When hardware domain of VXLAN device will be changed, ARP
packets which are trapped and marked with offload_fwd_mark will not be
flooded to VXLAN also in software, which will break VXLAN traffic.
To prevent such breaking, trap ARP packets at layer 2 and don't mark them
as L2-forwarded in hardware, then flooding ARP packets will be done only
in software, and VXLAN will send ARP packets.
Remove NVE_ENCAP_ARP which is no longer needed, as now ARP packets are
trapped when they enter the device.
Paolo Abeni [Thu, 20 Mar 2025 18:22:38 +0000 (19:22 +0100)]
net: introduce per netns packet chains
Currently network taps unbound to any interface are linked in the
global ptype_all list, affecting the performance in all the network
namespaces.
Add per netns ptypes chains, so that in the mentioned case only
the netns owning the packet socket(s) is affected.
While at that drop the global ptype_all list: no in kernel user
registers a tap on "any" type without specifying either the target
device or the target namespace (and IMHO doing that would not make
any sense).
Note that this adds a conditional in the fast path (to check for
per netns ptype_specific list) and increases the dataset size by
a cacheline (owing the per netns lists).
Simon Horman [Thu, 20 Mar 2025 11:19:20 +0000 (11:19 +0000)]
tty: caif: removed unused function debugfs_tx()
Remove debugfs_tx() which was added when the caif driver was added in
commit 9b27105b4a44 ("net-caif-driver: add CAIF serial driver (ldisc)")
but it has never been used.
Sudeep Holla [Wed, 19 Mar 2025 13:52:09 +0000 (13:52 +0000)]
net: phy: fixed_phy: transition to the faux device interface
The net fixed phy driver does not require the creation of a platform
device. Originally, this approach was chosen for simplicity when the
driver was first implemented.
With the introduction of the lightweight faux device interface, we now
have a more appropriate alternative. Migrate the device to utilize the
faux bus, given that the platform device it previously created was not
a real one anyway. This will get rid of the fake platform device.
Mark Zhang [Wed, 19 Mar 2025 19:23:18 +0000 (21:23 +0200)]
net/mlx5e: Use right API to free bitmap memory
Use bitmap_free() to free memory allocated with bitmap_zalloc_node().
This fixes memtrack error:
mtl rsc inconsistency: memtrack_free: .../drivers/net/ethernet/mellanox/mlx5/core/en_main.c::466: kfree for unknown address=0xFFFF0000CA3619E8, device=0x0
Signed-off-by: Mark Zhang <markzhang@nvidia.com> Reviewed-by: Maher Sanalla <msanalla@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Link: https://patch.msgid.link/1742412199-159596-3-git-send-email-tariqt@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alexander Duyck [Wed, 19 Mar 2025 17:46:25 +0000 (10:46 -0700)]
net: phylink: Remove unused function pointer from phylink structure
From what I can tell the .get_fixed_state pointer in the phylink structure
hasn't been used since commit 5c05c1dbb177 ("net: phylink, dsa: eliminate
phylink_fixed_state_cb()") . Since I can't find any users for it we might
as well just drop the pointer.
net: stmmac: Call xpcs_config_eee_mult_fact() only when xpcs is present
Some dwmac variants such as dwmac_socfpga don't use xpcs but lynx_pcs.
Don't call xpcs_config_eee_mult_fact() in this case, as this causes a
crash at init :
Unable to handle kernel NULL pointer dereference at virtual address 00000039 when write
[...]
Call trace:
xpcs_config_eee_mult_fact from stmmac_pcs_setup+0x40/0x10c
stmmac_pcs_setup from stmmac_dvr_probe+0xc0c/0x1244
stmmac_dvr_probe from socfpga_dwmac_probe+0x130/0x1bc
socfpga_dwmac_probe from platform_probe+0x5c/0xb0
Fixes: 060fb27060e8 ("net: stmmac: call xpcs_config_eee_mult_fact()") Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Link: https://patch.msgid.link/20250321103502.1303539-1-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Gal Pressman [Tue, 18 Mar 2025 11:24:26 +0000 (13:24 +0200)]
selftests: drv-net: rss_ctx: Don't assume indirection table is present
The test_rss_context_dump() test assumes the indirection table is always
supported, which is not true for all drivers, e.g., virtio_net when
VIRTIO_NET_F_RSS is disabled.
====================
r8169: enable more devices ASPM support
This series of patches will enable more devices ASPM support.
It also fix a RTL8126 cannot enter L1 substate issue when ASPM is
enabled.
====================
ChunHao Lin [Tue, 18 Mar 2025 08:37:21 +0000 (16:37 +0800)]
r8169: disable RTL8126 ZRX-DC timeout
Disable it due to it dose not meet ZRX-DC specification. If it is enabled,
device will exit L1 substate every 100ms. Disable it for saving more power
in L1 substate.
Peter Seiderer [Mon, 17 Mar 2025 09:04:00 +0000 (10:04 +0100)]
net: pktgen: add strict buffer parsing index check
Add strict buffer parsing index check to avoid the following Smatch
warning:
net/core/pktgen.c:877 get_imix_entries()
warn: check that incremented offset 'i' is capped
Checking the buffer index i after every get_user/i++ step and returning
with error code immediately avoids the current indirect (but correct)
error handling.
Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Closes: https://lore.kernel.org/netdev/36cf3ee2-38b1-47e5-a42a-363efeb0ace3@stanley.mountain/ Signed-off-by: Peter Seiderer <ps.report@gmx.net> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20250317090401.1240704-1-ps.report@gmx.net Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Lukas Bulwahn [Mon, 17 Mar 2025 09:27:17 +0000 (10:27 +0100)]
MAINTAINERS: adjust the file entry in INTEL PMC CORE DRIVER
Commit 7e2f7e25f6ff ("arch: x86: add IPC mailbox accessor function and add
SoC register access") adds a new file entry referring to the non-existent
file linux/platform_data/x86/intel_pmc_ipc.h in section INTEL PMC CORE
DRIVER rather than referring to the file
include/linux/platform_data/x86/intel_pmc_ipc.h added with this commit.
Note that it was missing 'include' in the beginning.
Dimitri Fedrau [Mon, 17 Mar 2025 07:48:34 +0000 (08:48 +0100)]
net: phy: dp83822: fix transmit amplitude if CONFIG_OF_MDIO not defined
When CONFIG_OF_MDIO is not defined the index for selecting the transmit
amplitude voltage for 100BASE-TX is set to 0, but it should be -1, if there
is no need to modify the transmit amplitude voltage. Move initialization of
the index from dp83822_of_init to dp8382x_probe.
Fix copy-paste error in the code comment for Interface Mode definitions.
The code refers to Internal TX delay, not Internal RX delay. It was likely
copied from the line above this one.
Realtek RTL8211F has a "PHY-mode" EEE support which interferes with an
IEEE 802.3 compliant implementation. This mode defaults to enabled, and
results in the MAC receive path not seeing the link transition to LPI
state.
net: phy: fix genphy_c45_eee_is_active() for disabled EEE
Commit 809265fe96fe ("net: phy: c45: remove local advertisement
parameter from genphy_c45_eee_is_active") stopped reading the local
advertisement from the PHY earlier in this development cycle, which
broke "ethtool --set-eee ethX eee off".
When ethtool is used to set EEE off, genphy_c45_eee_is_active()
indicates that EEE was active if the link partner reported an
advertisement, which causes phylib to set phydev->enable_tx_lpi on
link up, despite our local advertisement in hardware being empty.
However, phydev->advertising_eee is preserved while EEE is turned off,
which leads to genphy_c45_eee_is_active() incorrectly reporting that
EEE is active.
Fix it by checking phydev->eee_cfg.eee_enabled, and if clear,
immediately indicate that EEE is not active.
====================
mlx5e: Support recovery counter in reset
This series by Yael adds a recovery counter in ethtool, for any recovery
type during port reset cycle.
Series starts with some cleanup and refactoring patches.
New counter is added and exposed to ethtool stats in patch #4.
====================
Yael Chemla [Sun, 16 Mar 2025 08:14:36 +0000 (10:14 +0200)]
net/mlx5e: Expose port reset cycle recovery counter via ethtool
Display recovery event of PPCNT recovery counters group. Counts (per
link) the number of total successful recovery events of any recovery
types during port reset cycle.
Yael Chemla [Sun, 16 Mar 2025 08:14:33 +0000 (10:14 +0200)]
net/mlx5e: Ensure each counter group uses its PCAM bit
The code was incorrectly relying on PCAM bit of ppcnt_statistical_group
for accessing per_lane_error_counters.
If ppcnt_statistical_group PCAM bit was not set, we would not read
per_lane_error_counters, even when its PCAM bit is set.
Given the existing device capabilities, it seems to cause no harm, so
this change primarily serves as cleanup.
Signed-off-by: Yael Chemla <ychemla@nvidia.com> Reviewed-by: Cosmin Ratiu <cratiu@nvidia.com> Signed-off-by: Tariq Toukan <tariqt@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Link: https://patch.msgid.link/1742112876-2890-2-git-send-email-tariqt@nvidia.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When adding a socket option support in MPTCP, both the get and set parts
are supposed to be implemented.
IP(V6)_FREEBIND and IP(V6)_TRANSPARENT support for the setsockopt part
has been added a while ago, but it looks like the get part got
forgotten. It should have been present as a way to verify a setting has
been set as expected, and not to act differently from TCP or any other
socket types.
Everything was in place to expose it, just the last step was missing.
Only new code is added to cover these specific getsockopt(), that seems
safe.
When adding a socket option support in MPTCP, both the get and set parts
are supposed to be implemented.
IPV6_V6ONLY support for the setsockopt part has been added a while ago,
but it looks like the get part got forgotten. It should have been
present as a way to verify a setting has been set as expected, and not
to act differently from TCP or any other socket types.
Not supporting this getsockopt(IPV6_V6ONLY) blocks some apps which want
to check the default value, before doing extra actions. On Linux, the
default value is 0, but this can be changed with the net.ipv6.bindv6only
sysctl knob. On Windows, it is set to 1 by default. So supporting the
get part, like for all other socket options, is important.
Everything was in place to expose it, just the last step was missing.
Only new code is added to cover this specific getsockopt(), that seems
safe.
Fixes: c9b95a135987 ("mptcp: support IPV6_V6ONLY setsockopt") Cc: stable@vger.kernel.org Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/550 Reviewed-by: Mat Martineau <martineau@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20250314-net-mptcp-fix-data-stream-corr-sockopt-v1-2-122dbb249db3@kernel.org Signed-off-by: Paolo Abeni <pabeni@redhat.com>
====================
netconsole: Add support for userdata release
I am submitting a series of patches that introduce a new feature for the
netconsole subsystem, specifically the addition of the 'release' field
to the sysdata structure. This feature allows the kernel release/version
to be appended to the userdata dictionary in every message sent,
enhancing the information available for debugging and monitoring
purposes.
This complements the already supported release prepend feature, which
was added some time ago. The release prepend appends the release
information at the message header, which is not ideal for two reasons:
1) It is difficult to determine if a message includes this information,
making it hard and resource-intensive to parse.
2) When a message is fragmented, the release information is appended to
every message fragment, consuming valuable space in the packet.
The "release prepend" feature was created before the concept of userdata
and sysdata. Now that this format has proven successful, we are
implementing the release feature as part of this enhanced structure.
This patch series aims to improve the netconsole subsystem by providing
a more efficient and user-friendly way to include kernel release
information in messages. I believe these changes will significantly aid
in system analysis and troubleshooting.
Suggested-by: Manu Bretelle <chantr4@gmail.com> Signed-off-by: Breno Leitao <leitao@debian.org>
====================
projects / thirdparty / kernel / linux.git / log
