Add document for IFLA_BRPORT enum so we can use it in
Documentation/networking/bridge.rst.
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Add document for IFLA_BR enum so we can use it in
Documentation/networking/bridge.rst.
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The current bridge kernel doc is too old. It only pointed to the
linuxfoundation wiki page which lacks of the new features.
Here let's start the new bridge document and put all the bridge info
so new developers and users could catch up the last bridge status soon.
In this patch, Convert the doc to rst format. Add bridge brief introduction,
FAQ and contact info.
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Rohan G Thomas says:
====================
net: stmmac: EST implementation
This patchset extends EST interrupt handling support to DWXGMAC IP
followed by refactoring of EST implementation. Added a separate
module for EST and moved all EST related functions to the new module.
Also added support for EST cycle-time-extension.
====================
Link: https://lore.kernel.org/r/20231201055252.1302-1-rohan.g.thomas@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Refactor EST implementation by moving common code for DWMAC4 and
DWXGMAC IPs into a separate EST module. EST implementation for DWMAC4
and DWXGMAC differs only for CSR base address, PTOV field offset
width, and PTOV clock multiplier value.
Thanks, Serge Semin and Jakub Kicinski for the suggestions on
refactoring EST implementation into a separate EST module.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20231201055252.1302-3-rohan.g.thomas@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Enabled the following EST related interrupts:
1) Constant Gate Control Error (CGCE)
2) Head-of-Line Blocking due to Scheduling (HLBS)
3) Head-of-Line Blocking due to Frame Size (HLBF)
4) Base Time Register error (BTRE)
5) Switch to S/W owned list Complete (SWLC)
Also, add EST errors into the ethtool statistic.
The commit e49aa315cb ("net: stmmac: EST interrupts handling and
error reporting") and commit 9f29895919 ("net: stmmac: Add EST
errors into ethtool statistic") add EST interrupts handling and error
reporting support to DWMAC4 core. This patch enables the same support
for XGMAC.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20231201055252.1302-2-rohan.g.thomas@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Amritha Nambiar says:
====================
Introduce queue and NAPI support in netdev-genl (Was: Introduce NAPI queues support)
Add the capability to export the following via netdev-genl interface:
- queue information supported by the device
- NAPI information supported by the device
Introduce support for associating queue and NAPI instance.
Extend the netdev_genl generic netlink family for netdev
with queue and NAPI data.
The queue parameters exposed are:
- queue index
- queue type
- ifindex
- NAPI id associated with the queue
Additional rx and tx queue parameters can be exposed in follow up
patches by stashing them in netdev queue structures. XDP queue type
can also be supported in future.
The NAPI fields exposed are:
- NAPI id
- NAPI device ifindex
- Interrupt number associated with the NAPI instance
- PID for the NAPI thread
This series only supports 'get' ability for retrieving
certain queue and NAPI attributes. The 'set' ability for
configuring queue and associated NAPI instance via netdev-genl
will be submitted as a separate patch series.
Previous discussion at:
https://lore.kernel.org/netdev/c8476530638a5f4381d64db0e024ed49c2db3b02.camel@gmail.com/T/#m00999652a8b4731fbdb7bf698d2e3666c65a60e7
$ ./cli.py --spec netdev.yaml --do queue-get --json='{"ifindex": 12, "id": 0, "type": 0}'
{'id': 0, 'ifindex': 12, 'napi-id': 593, 'type': 'rx'}
$ ./cli.py --spec netdev.yaml --do queue-get --json='{"ifindex": 12, "id": 0, "type": 1}'
{'id': 0, 'ifindex': 12, 'napi-id': 593, 'type': 'tx'}
$ ./cli.py --spec netdev.yaml --dump queue-get --json='{"ifindex": 12}'
[{'id': 0, 'ifindex': 12, 'napi-id': 593, 'type': 'rx'},
{'id': 1, 'ifindex': 12, 'napi-id': 594, 'type': 'rx'},
{'id': 2, 'ifindex': 12, 'napi-id': 595, 'type': 'rx'},
{'id': 3, 'ifindex': 12, 'napi-id': 596, 'type': 'rx'},
{'id': 0, 'ifindex': 12, 'napi-id': 593, 'type': 'tx'},
{'id': 1, 'ifindex': 12, 'napi-id': 594, 'type': 'tx'},
{'id': 2, 'ifindex': 12, 'napi-id': 595, 'type': 'tx'},
{'id': 3, 'ifindex': 12, 'napi-id': 596, 'type': 'tx'}]
$ ./cli.py --spec netdev.yaml --do napi-get --json='{"id": 593}'
{'id': 593, 'ifindex': 12, 'irq': 291, 'pid': 3727}
$ ./cli.py --spec netdev.yaml --dump napi-get --json='{"ifindex": 12}'
[{'id': 596, 'ifindex': 12, 'irq': 294, 'pid': 3724},
{'id': 595, 'ifindex': 12, 'irq': 293, 'pid': 3725},
{'id': 594, 'ifindex': 12, 'irq': 292, 'pid': 3726},
{'id': 593, 'ifindex': 12, 'irq': 291, 'pid': 3727}]
====================
Link: https://lore.kernel.org/r/170147307026.5260.9300080745237900261.stgit@anambiarhost.jf.intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Michael Chan says:
====================
bnxt_en: Support new 5760X P7 devices
This series completes the basic support for the new 5760X P7 devices
with new PCI IDs added in the last patch.
Thie first patch fixes a backing store issue introduced in the last
patchset last week. The 2nd patch is the new firmware interface
required to support the new chips. The next few patches are doorbell
changes, refactoring, and new hardware interface structures. New
changes to support packet reception including TPA are added in patch 10.
The next 4 patches are ethernet link related changes to support the
new chip.
====================
Link: https://lore.kernel.org/r/20231201223924.26955-1-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Refactor bnxt_tpa_start() by adding bnxt_tpa_metadata() to gather the
metadata from the TPA_START completion. This makes it easier to
support the new P7 chip which has a modified TPA_START completion
structure with different metadata formats. We also add vlan_valid
and cfa_code_valid fields to the bnxt_tpa_info structure so that the
VLAN and VF rep logic can be common for all chips. The VLAN metadata
is now collected in bnxt_tpa_start() only when it is valid and the
vlan_valid field will be set. bnxt_tpa_end() can now use common VLAN
logic for all chips.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/20231201223924.26955-10-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The new chip family passes the Toggle bits to the driver in the NQE
notification. The driver now stores this value and sends it back to
hardware when it re-arms the RX and TX CQs. Together with the earlier
patch that guarantees the driver will only re-arm the CQ at the end of
NAPI polling if it has seen a new NQE, this method allows the hardware
to detect any dropped doorbells.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/20231201223924.26955-6-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Repurpose the BNXT_FLAG_CHIP_SR2 flag by renaming it to
BNXT_FLAG_CHIP_P7 since the SR2 chip never went to production. The SR2
statictics structure is also renamed for the P7 chip. Define the basic
P7 doorbell bits (Epoch. Toggle, etc) and implement the Epoch bit
logic. The next higher bit beyond the legal doorbell mask is the
Epoch bit used for doorbells on P7 chips. This bit is used by the
chip to detect dropped doorbells.
The 57608 chip ID belonging to the P7 family is also defined. Note
that the PCI ID is not added until the last patch in the series.
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/20231201223924.26955-4-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The current code determines the last backing store valid type during
bnxt_hwrm_func_backing_store_qcaps_v2(). In effect, the last type
is determined based on what firmware advertises. The more correct
way is to determine it based on what the driver is configuring. The
driver may not configure all the backing store types advertised by
firmware.
Move the logic to determine the last type to bnxt_backing_store_cfg_v2().
We need to pass the legacy enable flags to the function in case only
the legacy types are being configured.
Fixes: 236e237f8f ("bnxt_en: Add support for HWRM_FUNC_BACKING_STORE_CFG_V2 firmware calls")
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/20231201223924.26955-2-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Walk the hashinfo->bhash2 table so that inet_diag can dump TCP sockets
that are bound but haven't yet called connect() or listen().
The code is inspired by the ->lhash2 loop. However there's no manual
test of the source port, since this kind of filtering is already
handled by inet_diag_bc_sk(). Also, a maximum of 16 sockets are dumped
at a time, to avoid running with bh disabled for too long.
There's no TCP state for bound but otherwise inactive sockets. Such
sockets normally map to TCP_CLOSE. However, "ss -l", which is supposed
to only dump listening sockets, actually requests the kernel to dump
sockets in either the TCP_LISTEN or TCP_CLOSE states. To avoid dumping
bound-only sockets with "ss -l", we therefore need to define a new
pseudo-state (TCP_BOUND_INACTIVE) that user space will be able to set
explicitly.
With an IPv4, an IPv6 and an IPv6-only socket, bound respectively to
40000, 64000, 60000, an updated version of iproute2 could work as
follow:
$ ss -t state bound-inactive
Recv-Q Send-Q Local Address:Port Peer Address:Port Process
0 0 0.0.0.0:40000 0.0.0.0:*
0 0 [::]:60000 [::]:*
0 0 *:64000 *:*
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/b3a84ae61e19c06806eea9c602b3b66e8f0cfc81.1701362867.git.gnault@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Heiko Stuebner says:
====================
net: phy: micrel: additional clock handling
Some Micrel phys define a specific rmii-ref clock (added in 2014) while
the generic phy binding specifies an unnamed clock for ethernet phys.
This allows Micrel phys to use both, so as to keep the phys not using
the named rmii-ref clock to conform to the generic binding while allowing
them to enable a supplying clock, when the phy is not supplied by a
dedicated oscillator.
====================
Link: https://lore.kernel.org/r/20231201150131.326766-1-heiko@sntech.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The generic ethernet-phy binding allows describing an external clock since
commit 350b7a258f ("dt-bindings: net: phy: Document support for external PHY clk")
for cases where the phy is not supplied by an oscillator but instead
by a clock from the host system.
And the old named "rmii-ref" clock from 2014 is only specified for phys
of the KSZ8021, KSZ8031, KSZ8081, KSZ8091 types.
So allow retrieving and enabling the optional generic clock on phys that
do not provide a rmii-ref clock.
Signed-off-by: Heiko Stuebner <heiko.stuebner@cherry.de>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20231201150131.326766-3-heiko@sntech.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Suman Ghosh says:
====================
octeontx2: Multicast/mirror offload changes
This patchset includes changes to support TC multicast/mirror offload.
Patch #1: Adds changes to support new mailbox to offload multicast/mirror
offload.
Patch #2: Adds TC related changes which uses the newly added mailboxes to
offload multicast/mirror rules.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch extends TC flower offload support for mirroring ingress
traffic to a different PF/VF. Below is an example command,
'tc filter add dev eth1 ingress protocol ip flower src_ip <ip-addr>
skip_sw action mirred ingress mirror dev eth2'
Signed-off-by: Suman Ghosh <sumang@marvell.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
A new mailbox is added to support offloading of multicast/mirror
functionality. The mailbox also supports dynamic updation of the
multicast/mirror list.
Signed-off-by: Suman Ghosh <sumang@marvell.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Coco Li says:
====================
Analyze and Reorganize core Networking Structs to optimize cacheline consumption
Currently, variable-heavy structs in the networking stack is organized
chronologically, logically and sometimes by cacheline access.
This patch series attempts to reorganize the core networking stack
variables to minimize cacheline consumption during the phase of data
transfer. Specifically, we looked at the TCP/IP stack and the fast
path definition in TCP.
For documentation purposes, we also added new files for each core data
structure we considered, although not all ended up being modified due
to the amount of existing cacheline they span in the fast path. In
the documentation, we recorded all variables we identified on the
fast path and the reasons. We also hope that in the future when
variables are added/modified, the document can be referred to and
updated accordingly to reflect the latest variable organization.
Tested:
Our tests were run with neper tcp_rr using tcp traffic. The tests have $cpu
number of threads and variable number of flows (see below).
Tests were run on 6.5-rc1
Efficiency is computed as cpu seconds / throughput (one tcp_rr round trip).
The following result shows efficiency delta before and after the patch
series is applied.
On AMD platforms with 100Gb/s NIC and 256Mb L3 cache:
IPv4
Flows with patches clean kernel Percent reduction
30k 0.0001736538065 0.0002741191042 -36.65%
20k 0.0001583661752 0.0002712559158 -41.62%
10k 0.0001639148817 0.0002951800751 -44.47%
5k 0.0001859683866 0.0003320642536 -44.00%
1k 0.0002035190546 0.0003152056382 -35.43%
IPv6
Flows with patches clean kernel Percent reduction
30k 0.000202535503 0.0003275329163 -38.16%
20k 0.0002020654777 0.0003411304786 -40.77%
10k 0.0002122427035 0.0003803674705 -44.20%
5k 0.0002348776729 0.0004030403953 -41.72%
1k 0.0002237384583 0.0002813646157 -20.48%
On Intel platforms with 200Gb/s NIC and 105Mb L3 cache:
IPv6
Flows with patches clean kernel Percent reduction
30k 0.0006296537873 0.0006370427753 -1.16%
20k 0.0003451029365 0.0003628016076 -4.88%
10k 0.0003187646958 0.0003346835645 -4.76%
5k 0.0002954676348 0.000311807592 -5.24%
1k 0.0001909169342 0.0001848069709 3.31%
v8 changes:
1. Update net_device_read_txrx cache group maximum
2. Update MAINTAINERS for documentations
3. Skip __cache_group variables in scripts/kernel-doc
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Reorganize fast path variables on tx-txrx-rx order.
Fastpath cacheline ends after sysctl_tcp_rmem.
There are only read-only variables here. (write is on the control path
and not considered in this case)
Below data generated with pahole on x86 architecture.
Fast path variables span cache lines before change: 4
Fast path variables span cache lines after change: 2
Suggested-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Wei Wang <weiwan@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Coco Li <lixiaoyan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set up build time warnings to safeguard against future header changes of
organized structs.
Warning includes:
1) whether all variables are still in the same cache group
2) whether all the cache groups have the sum of the members size (in the
maximum condition, including all members defined in configs)
The __cache_group* variables are ignored in kernel-doc check in the
various header files they appear in to enforce the cache groups.
Suggested-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Coco Li <lixiaoyan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Analyzed a few structs in the networking stack by looking at variables
within them that are used in the TCP/IP fast path.
Fast path is defined as TCP path where data is transferred from sender to
receiver unidirectionally. It doesn't include phases other than
TCP_ESTABLISHED, nor does it look at error paths.
We hope to re-organizing variables that span many cachelines whose fast
path variables are also spread out, and this document can help future
developers keep networking fast path cachelines small.
Optimized_cacheline field is computed as
(Fastpath_Bytes/L3_cacheline_size_x86), and not the actual organized
results (see patches to come for these).
Investigation is done on 6.5
Name Struct_Cachelines Cur_fastpath_cache Fastpath_Bytes Optimized_cacheline
tcp_sock 42 (2664 Bytes) 12 396 8
net_device 39 (2240 bytes) 12 234 4
inet_sock 15 (960 bytes) 14 922 14
Inet_connection_sock 22 (1368 bytes) 18 1166 18
Netns_ipv4 (sysctls) 12 (768 bytes) 4 77 2
linux_mib 16 (1060) 6 104 2
Note how there isn't much improvement space for inet_sock and
Inet_connection_sock because sk and icsk_inet respectively takes up so
much of the struct that rest of the variables become a small portion of
the struct size.
So, we decided to reorganize tcp_sock, net_device, netns_ipv4
Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Coco Li <lixiaoyan@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When breaking out the Gen4 gPTP support to its own module the dependency
on the PTP_1588_CLOCK framework was left as optional and only stated for
the driver using the module. This leads to issues when doing
COMPILE_TEST of RENESAS_GEN4_PTP separately and PTP_1588_CLOCK is built
as a module and the other as a built-in. Add an explicit depend on
PTP_1588_CLOCK.
While at it remove the optional support for PTP_1588_CLOCK from
RENESAS_ETHER_SWITCH as the driver unconditionally calls the Gen4 gPTP
module and thus also requires the PTP_1588_CLOCK framework.
Reported-by: Arnd Bergmann <arnd@arndb.de>
Fixes: 8c1c66235e ("net: ethernet: renesas: rcar_gen4_ptp: Break out to module")
Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20231129111142.3322667-1-niklas.soderlund+renesas@ragnatech.se
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
