| Age | Commit message (Collapse) | Author |
|---|
|
Kui-Feng Lee says:
====================
Remove expired routes with a separated list of routes.
FIB6 GC walks trees of fib6_tables to remove expired routes. Walking a tree
can be expensive if the number of routes in a table is big, even if most of
them are permanent. Checking routes in a separated list of routes having
expiration will avoid this potential issue.
Background
==========
The size of a Linux IPv6 routing table can become a big problem if not
managed appropriately. Now, Linux has a garbage collector to remove
expired routes periodically. However, this may lead to a situation in
which the routing path is blocked for a long period due to an
excessive number of routes.
For example, years ago, there is a commit c7bb4b89033b ("ipv6: tcp:
drop silly ICMPv6 packet too big messages"). The root cause is that
malicious ICMPv6 packets were sent back for every small packet sent to
them. These packets add routes with an expiration time that prompts
the GC to periodically check all routes in the tables, including
permanent ones.
Why Route Expires
=================
Users can add IPv6 routes with an expiration time manually. However,
the Neighbor Discovery protocol may also generate routes that can
expire. For example, Router Advertisement (RA) messages may create a
default route with an expiration time. [RFC 4861] For IPv4, it is not
possible to set an expiration time for a route, and there is no RA, so
there is no need to worry about such issues.
Create Routes with Expires
==========================
You can create routes with expires with the command.
For example,
ip -6 route add 2001:b000:591::3 via fe80::5054:ff:fe12:3457 \
dev enp0s3 expires 30
The route that has been generated will be deleted automatically in 30
seconds.
GC of FIB6
==========
The function called fib6_run_gc() is responsible for performing
garbage collection (GC) for the Linux IPv6 stack. It checks for the
expiration of every route by traversing the trees of routing
tables. The time taken to traverse a routing table increases with its
size. Holding the routing table lock during traversal is particularly
undesirable. Therefore, it is preferable to keep the lock for the
shortest possible duration.
Solution
========
The cause of the issue is keeping the routing table locked during the
traversal of large trees. To solve this problem, we can create a separate
list of routes that have expiration. This will prevent GC from checking
permanent routes.
Result
======
We conducted a test to measure the execution times of fib6_gc_timer_cb()
and observed that it enhances the GC of FIB6. During the test, we added
permanent routes with the following numbers: 1000, 3000, 6000, and
9000. Additionally, we added a route with an expiration time.
Here are the average execution times for the kernel without the patch.
- 120020 ns with 1000 permanent routes
- 308920 ns with 3000 ...
- 581470 ns with 6000 ...
- 855310 ns with 9000 ...
The kernel with the patch consistently takes around 14000 ns to execute,
regardless of the number of permanent routes that are installed.
Major changes from v7:
- Fix warings raised by the patchwork.
Major changes from v6:
- Remove unnecessary check of tb6 in fib6_clean_expires_locked().
- Use ib6_clean_expires_locked() instead in fib6_purge_rt().
Major changes from v5:
- Change the order of adding new routes to the GC list and starting
GC timer.
- Remove time measurements from the test case.
- Stop forcing GC flush.
Major changes from v4:
- Detect existence of 'strace' in the test case.
Major changes from v3:
- Fix the type of arg according to feedback.
- Add 1k temporary routes and 5K permanent routes in the test case.
Measure time spending on GC with strace.
Major changes from v2:
- Remove unnecessary and incorrect sysctl restoring in the test case.
Major changes from v1:
- Moved gc_link to avoid creating a hole in fib6_info.
- Moved fib6_set_expires*() and fib6_clean_expires*() to the header
file and inlined. And removed duplicated lines.
- Added a test case.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Add 1000 IPv6 routes with expiration time (w/ and w/o additional 5000
permanet routes in the background.) Wait for a few seconds to make sure
they are removed correctly.
The expected output of the test looks like the following example.
> Fib6 garbage collection test
> TEST: ipv6 route garbage collection [ OK ]
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
FIB6 GC walks trees of fib6_tables to remove expired routes. Walking a tree
can be expensive if the number of routes in a table is big, even if most of
them are permanent. Checking routes in a separated list of routes having
expiration will avoid this potential issue.
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
On some I219 devices, ethernet cable plugging detection only works once
from PCI D3 state. Subsequent cable plugging does set PME bit correctly,
but device still doesn't get woken up.
Since I219 connects to the root complex directly, it relies on platform
firmware (ACPI) to wake it up. In this case, the GPE from _PRW only
works for first cable plugging but fails to notify the driver for
subsequent plugging events.
The issue was originally found on CNP, but the same issue can be found
on ADL too. So workaround the issue by continuing use PME poll after
first ACPI wake. As PME poll is always used, the runtime suspend
restriction for CNP can also be removed.
Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Acked-by: Sasha Neftin <sasha.neftin@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Now there are two indicators of socket memory pressure sit inside
struct mem_cgroup, socket_pressure and tcpmem_pressure, indicating
memory reclaim pressure in memcg->memory and ->tcpmem respectively.
When in legacy mode (cgroupv1), the socket memory is charged into
->tcpmem which is independent of ->memory, so socket_pressure has
nothing to do with socket's pressure at all. Things could be worse
by taking socket_pressure into consideration in legacy mode, as a
pressure in ->memory can lead to premature reclamation/throttling
in socket.
While for the default mode (cgroupv2), the socket memory is charged
into ->memory, and ->tcpmem/->tcpmem_pressure are simply not used.
So {socket,tcpmem}_pressure are only used in default/legacy mode
respectively for indicating socket memory pressure. This patch fixes
the pieces of code that make mixed use of both.
Fixes: 8e8ae645249b ("mm: memcontrol: hook up vmpressure to socket pressure")
Signed-off-by: Abel Wu <wuyun.abel@bytedance.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Take over maintainership of the nfp driver from Simon as he
is moving away from Corigine.
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Acked-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
This patch adds MQPRIO Qdisc offload in full 'channel' mode which allows
not only setting up pri:tc mapping, but also configuring TX shapers on
external port FIFOs. The K3 CPSW MQPRIO Qdisc offload is expected to work
with VLAN/priority tagged packets. Non-tagged packets have to be mapped
only to TC0.
- TX traffic classes must be rated starting from TC that has highest
priority and with no gaps
- Traffic classes are used starting from 0, that has highest priority
- min_rate defines Committed Information Rate (guaranteed)
- max_rate defines Excess Information Rate (non guaranteed) and offloaded
as (max_rate[i] - tcX_min_rate[i])
- VLAN/priority tagged packets mapped to TC0 will exit switch with VLAN tag
priority 0
The configuration example:
ethtool -L eth1 tx 5
ethtool --set-priv-flags eth1 p0-rx-ptype-rrobin off
tc qdisc add dev eth1 parent root handle 100: mqprio num_tc 3 \
map 0 0 1 2 0 0 0 0 0 0 0 0 0 0 0 0 \
queues 1@0 1@1 1@2 hw 1 mode channel \
shaper bw_rlimit min_rate 0 100mbit 200mbit max_rate 0 101mbit 202mbit
tc qdisc replace dev eth2 handle 100: parent root mqprio num_tc 1 \
map 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 queues 1@0 hw 1
ip link add link eth1 name eth1.100 type vlan id 100
ip link set eth1.100 type vlan egress 0:0 1:1 2:2 3:3 4:4 5:5 6:6 7:7
In the above example two ports share the same TX CPPI queue 0 for low
priority traffic. 3 traffic classes are defined for eth1 and mapped to:
TC0 - low priority, TX CPPI queue 0 -> ext Port 1 fifo0, no rate limit
TC1 - prio 2, TX CPPI queue 1 -> ext Port 1 fifo1, CIR=100Mbit/s, EIR=1Mbit/s
TC2 - prio 3, TX CPPI queue 2 -> ext Port 1 fifo2, CIR=200Mbit/s, EIR=2Mbit/s
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: Roger Quadros <rogerq@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Eric Dumazet says:
====================
inet: socket lock and data-races avoidance
In this series, I converted 20 bits in "struct inet_sock" and made
them truly atomic.
This allows to implement many IP_ socket options in a lockless
fashion (no need to acquire socket lock), and fixes data-races
that were showing up in various KCSAN reports.
I also took care of IP_TTL/IP_MINTTL, but left few other options
for another series.
v4: Rebased after recent mptcp changes.
Added Reviewed-by: tags from Simon (thanks !)
v3: fixed patch 7, feedback from build bot about ipvs set_mcast_loop()
v2: addressed a feedback from a build bot in patch 9 by removing
unused issk variable in mptcp_setsockopt_sol_ip_set_transparent()
Added Acked-by: tags from Soheil (thanks !)
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
inet->min_ttl is already read with READ_ONCE().
Implementing IP_MINTTL socket option set/read
without holding the socket lock is easy.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
ip_select_ttl() is racy, because it reads inet->uc_ttl
without proper locking.
Add READ_ONCE()/WRITE_ONCE() annotations while
allowing IP_TTL socket option to be set/read without
holding the socket lock.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Make room in struct inet_sock by removing this bit field,
using one available bit in inet_flags instead.
Also move local_port_range to fill the resulting hole,
saving 8 bytes on 64bit arches.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_BIND_ADDRESS_NO_PORT socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_NODEFRAG socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We move single bit fields to inet->inet_flags to avoid races.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_TRANSPARENT socket option can now be set/read
without locking the socket.
v2: removed unused issk variable in mptcp_setsockopt_sol_ip_set_transparent()
v4: rebased after commit 3f326a821b99 ("mptcp: change the mpc check helper to return a sk")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_MULTICAST_ALL socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_MULTICAST_LOOP socket option can now be set/read
without locking the socket.
v3: fix build bot error reported in ipvs set_mcast_loop()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_HDRINCL socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_FREEBIND socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_RECVERR_RFC4884 socket option can now be set/read
without locking the socket.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
IP_RECVERR socket option can now be set/get without locking the socket.
This patch potentially avoid data-races around inet->recverr.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Now we have inet->inet_flags, we can set following options
without having to hold the socket lock:
IP_PKTINFO, IP_RECVTTL, IP_RECVTOS, IP_RECVOPTS, IP_RETOPTS,
IP_PASSSEC, IP_RECVORIGDSTADDR, IP_RECVFRAGSIZE.
ip_sock_set_pktinfo() no longer hold the socket lock.
Similarly we can get the following options whithout holding
the socket lock:
IP_PKTINFO, IP_RECVTTL, IP_RECVTOS, IP_RECVOPTS, IP_RETOPTS,
IP_PASSSEC, IP_RECVORIGDSTADDR, IP_CHECKSUM, IP_RECVFRAGSIZE.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Various inet fields are currently racy.
do_ip_setsockopt() and do_ip_getsockopt() are mostly holding
the socket lock, but some (fast) paths do not.
Use a new inet->inet_flags to hold atomic bits in the series.
Remove inet->cmsg_flags, and use instead 9 bits from inet_flags.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Ruan Jinjie says:
====================
net: Remove redundant of_match_ptr() macro
Since these net drivers depend on CONFIG_OF, there is
no need to wrap the macro of_match_ptr() here.
Changes in v3:
- Collect responses from v1 and v2.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The driver depends on CONFIG_OF, it is not necessary to use
of_match_ptr() here.
Signed-off-by: Ruan Jinjie <ruanjinjie@huawei.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Use the module_misc_device macro to simplify the code, which is the
same as declaring with module_init() and module_exit().
Signed-off-by: Li Zetao <lizetao1@huawei.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Jijie Shao says:
====================
hns3: refactor registers information for ethtool -d
refactor registers information for ethtool -d
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
In the original RPU query command, the status register values of
multiple RPU tunnels are accumulated by default, which is unreasonable.
This patch Fix it by querying the specified tunnel ID.
The tunnel number of the device can be obtained from firmware
during initialization.
Fixes: ddb54554fa51 ("net: hns3: add DFX registers information for ethtool -d")
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The dump register function is being refactored.
The third step in refactoring is to support tlv info in regs data for
HNS3 PF driver.
Currently, if we use "ethtool -d" to dump regs value,
the output is as follows:
offset1: 00 01 02 03 04 05 ...
offset2：10 11 12 13 14 15 ...
......
We can't get the value of a register directly.
This patch deletes the original separator information and
add tag_len_value information in regs data.
ethtool can parse register data in key-value format by -d command.
a patch will be added to the ethtool to parse regs data
in the following format:
reg1 : value2
reg2 : value2
......
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The dump register function is being refactored.
The second step in refactoring is to support tlv info in regs data for
HNS3 PF driver.
Currently, if we use "ethtool -d" to dump regs value,
the output is as follows:
offset1: 00 01 02 03 04 05 ...
offset2：10 11 12 13 14 15 ...
......
We can't get the value of a register directly.
This patch deletes the original separator information and
add tag_len_value information in regs data.
ethtool can parse register data in key-value format by -d command.
a patch will be added to the ethtool to parse regs data
in the following format:
reg1 : value2
reg2 : value2
......
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The dump register function is being refactored.
The first step in refactoring is put the dump regs function
into a separate file.
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Wei Fang says:
====================
net: fec: add XDP_TX feature support
This patch set is to support the XDP_TX feature of FEC driver, the first
patch is add initial XDP_TX support, and the second patch improves the
performance of XDP_TX by not using xdp_convert_buff_to_frame(). Please
refer to the commit message of each patch for more details.
====================
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
As suggested by Jesper and Alexander, we can avoid converting xdp_buff
to xdp_frame in case of XDP_TX to save a bunch of CPU cycles, so that
we can further improve the XDP_TX performance.
Before this patch on i.MX8MP-EVK board, the performance shows as follows.
root@imx8mpevk:~# ./xdp2 eth0
proto 17: 353918 pkt/s
proto 17: 352923 pkt/s
proto 17: 353900 pkt/s
proto 17: 352672 pkt/s
proto 17: 353912 pkt/s
proto 17: 354219 pkt/s
After applying this patch, the performance is improved.
root@imx8mpevk:~# ./xdp2 eth0
proto 17: 369261 pkt/s
proto 17: 369267 pkt/s
proto 17: 369206 pkt/s
proto 17: 369214 pkt/s
proto 17: 369126 pkt/s
proto 17: 369272 pkt/s
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Suggested-by: Jesper Dangaard Brouer <hawk@kernel.org>
Reviewed-by: Jesper Dangaard Brouer <hawk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The XDP_TX feature is not supported before, and all the frames
which are deemed to do XDP_TX action actually do the XDP_DROP
action. So this patch adds the XDP_TX support to FEC driver.
I tested the performance of XDP_TX in XDP_DRV mode and XDP_SKB
mode respectively on i.MX8MP-EVK platform, and as suggested by
Jesper, I also tested the performance of XDP_REDIRECT on the
same platform. And the test steps and results are as follows.
XDP_TX test:
Step 1: One board is used as generator and connects to switch,and
the FEC port of DUT also connects to the switch. Both boards with
flow control off. Then the generator runs the
pktgen_sample03_burst_single_flow.sh script to generate and send
burst traffic to DUT. Note that the size of packet was set to 64
bytes and the procotol of packet was UDP in my test scenario. In
addition, the SMAC of the packet need to be different from the MAC
of the generator, because the xdp2 program will swap the DMAC and
SMAC of the packet and send it back to the generator. If the SMAC
of the generated packet is the MAC of the generator, the generator
will receive the returned traffic which increase the CPU loading
and significantly degrade the transmit speed of the generator, and
finally it affects the test of XDP_TX performance.
Step 2: The DUT runs the xdp2 program to transmit received UDP
packets back out on the same port where they were received.
root@imx8mpevk:~# ./xdp2 eth0
proto 17: 353918 pkt/s
proto 17: 352923 pkt/s
proto 17: 353900 pkt/s
proto 17: 352672 pkt/s
proto 17: 353912 pkt/s
proto 17: 354219 pkt/s
root@imx8mpevk:~# ./xdp2 -S eth0
proto 17: 160604 pkt/s
proto 17: 160708 pkt/s
proto 17: 160564 pkt/s
proto 17: 160684 pkt/s
proto 17: 160640 pkt/s
proto 17: 160720 pkt/s
The above results show that the XDP_TX performance of XDP_DRV mode
is much better than XDP_SKB mode, more than twice that of XDP_SKB
mode, which is in line with our expectation.
XDP_REDIRECT test:
Step1: Both the generator and the FEC port of the DUT connet to the
switch port. All the ports with flow control off, then the generator
runs the pktgen script to generate and send burst traffic to DUT.
Note that the size of packet was set to 64 bytes and the procotol of
packet was UDP in my test scenario.
Step2: The DUT runs the xdp_redirect program to redirect the traffic
from the FEC port to the FEC port itself.
root@imx8mpevk:~# ./xdp_redirect eth0 eth0
Redirecting from eth0 (ifindex 2; driver fec) to eth0
(ifindex 2; driver fec)
Summary 232,302 rx/s 0 err,drop/s 232,344 xmit/s
Summary 234,579 rx/s 0 err,drop/s 234,577 xmit/s
Summary 235,548 rx/s 0 err,drop/s 235,549 xmit/s
Summary 234,704 rx/s 0 err,drop/s 234,703 xmit/s
Summary 235,504 rx/s 0 err,drop/s 235,504 xmit/s
Summary 235,223 rx/s 0 err,drop/s 235,224 xmit/s
Summary 234,509 rx/s 0 err,drop/s 234,507 xmit/s
Summary 235,481 rx/s 0 err,drop/s 235,482 xmit/s
Summary 234,684 rx/s 0 err,drop/s 234,683 xmit/s
Summary 235,520 rx/s 0 err,drop/s 235,520 xmit/s
Summary 235,461 rx/s 0 err,drop/s 235,461 xmit/s
Summary 234,627 rx/s 0 err,drop/s 234,627 xmit/s
Summary 235,611 rx/s 0 err,drop/s 235,611 xmit/s
Packets received : 3,053,753
Average packets/s : 234,904
Packets transmitted : 3,053,792
Average transmit/s : 234,907
Compared the performance of XDP_TX with XDP_REDIRECT, XDP_TX is also
much better than XDP_REDIRECT. It's also in line with our expectation.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Suggested-by: Jesper Dangaard Brouer <hawk@kernel.org>
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When run command "ip netns delete client", device link1_1 has been
deleted. So, it is no need to delete link1_1 again. Remove it.
Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-updates-2023-08-14
1) Handle PTP out of order CQEs issue
2) Check FW status before determining reset successful
3) Expose maximum supported SFs via devlink resource
4) MISC cleanups
* tag 'mlx5-updates-2023-08-14' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
net/mlx5: Don't query MAX caps twice
net/mlx5: Remove unused MAX HCA capabilities
net/mlx5: Remove unused CAPs
net/mlx5: Fix error message in mlx5_sf_dev_state_change_handler()
net/mlx5: Remove redundant check of mlx5_vhca_event_supported()
net/mlx5: Use mlx5_sf_start_function_id() helper instead of directly calling MLX5_CAP_GEN()
net/mlx5: Remove redundant SF supported check from mlx5_sf_hw_table_init()
net/mlx5: Use auxiliary_device_uninit() instead of device_put()
net/mlx5: E-switch, Add checking for flow rule destinations
net/mlx5: Check with FW that sync reset completed successfully
net/mlx5: Expose max possible SFs via devlink resource
net/mlx5e: Add recovery flow for tx devlink health reporter for unhealthy PTP SQ
net/mlx5e: Make tx_port_ts logic resilient to out-of-order CQEs
net/mlx5: Consolidate devlink documentation in devlink/mlx5.rst
====================
Link: https://lore.kernel.org/r/20230814214144.159464-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Jakub Kicinski says:
====================
net: warn about attempts to register negative ifindex
Follow up to the recently posted fix for OvS lacking input
validation:
https://lore.kernel.org/all/20230814203840.2908710-1-kuba@kernel.org/
Warn about negative ifindex more explicitly and misc YNL updates.
====================
Link: https://lore.kernel.org/r/20230814205627.2914583-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
When developing specs its useful to know which attr space
YNL was trying to find an attribute in on key error.
Instead of printing:
KeyError: 0
add info about the space:
Exception: Space 'vport' has no attribute with value '0'
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://lore.kernel.org/r/20230814205627.2914583-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Add NEW to the spec, it was useful testing the fix for OvS
input validation.
Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://lore.kernel.org/r/20230814205627.2914583-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Since the xarray changes we mix returning valid ifindex and negative
errno in a single int returned from dev_index_reserve(). This depends
on the fact that ifindexes can't be negative. Otherwise we may insert
into the xarray and return a very large negative value. This in turn
may break ERR_PTR().
OvS is susceptible to this problem and lacking validation (fix posted
separately for net).
Reject negative ifindex explicitly. Add a warning because the input
validation is better handled by the caller.
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/20230814205627.2914583-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Mario reports that loading r8152 on his system leads to a:
netif_napi_add_weight() called with weight 256
warning getting printed. We don't have any solid data
on why such high budget was chosen, and it may cause
stalls in processing other softirqs and rt threads.
So try to switch back to the default (64) weight.
If this slows down someone's system we should investigate
which part of stopping starting the NAPI poll in this
driver are expensive.
Reported-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://lore.kernel.org/all/0bfd445a-81f7-f702-08b0-bd5a72095e49@amd.com/
Acked-by: Hayes Wang <hayeswang@realtek.com>
Link: https://lore.kernel.org/r/20230814153521.2697982-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Commit bdfe2da6aefd ("e1000e: cosmetic move of function prototypes to the new mac.h")
declared but never implemented them.
Signed-off-by: Yue Haibing <yuehaibing@huawei.com>
Reviewed-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://lore.kernel.org/r/20230814135821.4808-1-yuehaibing@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Newer versions of clang warn about this variable being assigned but
never used:
drivers/net/ethernet/qlogic/qed/qed_vf.c:63:67: error: parameter 'resp_size' set but not used [-Werror,-Wunused-but-set-parameter]
There is no indication in the git history on how this was ever
meant to be used, so just remove the entire calculation and argument
passing for it to avoid the warning.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Link: https://lore.kernel.org/r/20230814074512.1067715-1-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Implement netdev trigger and primitive bliking offloading as well as
simple set_brigthness function for both PHY LEDs of the in-SoC PHYs
found in MT7981 and MT7988.
For MT7988, read boottrap register and apply LED polarities accordingly
to get uniform behavior from all LEDs on MT7988.
This requires syscon phandle 'mediatek,pio' present in parenting MDIO bus
which should point to the syscon holding the boottrap register.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/dc324d48c00cd7350f3a506eaa785324cae97372.1691977904.git.daniel@makrotopia.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Ido Schimmel says:
====================
nexthop: Various cleanups
Benefit from recent bug fixes and simplify the nexthop dump code.
No regressions in existing tests:
# ./fib_nexthops.sh
[...]
Tests passed: 234
Tests failed: 0
====================
Link: https://lore.kernel.org/r/20230813164856.2379822-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The nexthop and nexthop bucket dump callbacks previously returned a
positive return code even when the dump was complete, prompting the core
netlink code to invoke the callback again, until returning zero.
Zero was only returned by these callbacks when no information was filled
in the provided skb, which was achieved by incrementing the dump
sentinel at the end of the dump beyond the ID of the last nexthop.
This is no longer necessary as when the dump is complete these callbacks
return zero.
Remove the unnecessary increment.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20230813164856.2379822-3-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
