| Age | Commit message (Collapse) | Author |
|---|
|
Currently "net/sctp/checksum.h" including "net/sctp/sctp.h" is
included in quite some places in netfilter and openswitch and
net/sched. It's not necessary to include "net/sctp/sctp.h" if
a module does not have dependence on SCTP, "linux/sctp.h" is
the right one to include.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Reviewed-by: Saeed Mahameed <saeed@kernel.org>
Link: https://lore.kernel.org/r/ca7ea96d62a26732f0491153c3979dc1c0d8d34a.1668526793.git.lucien.xin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Currently the driver is able to create leaf nodes for the devlink-rate,
but is unable to set parent for them. This wasn't as issue before the
possibility to export hierarchy from the driver. After adding the export
feature, in order for the driver to supply correct hierarchy, it's
necessary for it to be able to supply a parent name to
devl_rate_leaf_create().
Introduce a new parameter 'parent_name' in devl_rate_leaf_create().
Signed-off-by: Michal Wilczynski <michal.wilczynski@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Intel 100G card internal firmware hierarchy for Hierarchicial QoS is very
rigid and can't be easily removed. This requires an ability to export
default hierarchy to allow user to modify it. Currently the driver is
only able to create the 'leaf' nodes, which usually represent the vport.
This is not enough for HQoS implemented in Intel hardware.
Introduce new function devl_rate_node_create() that allows for creation
of the devlink-rate nodes from the driver.
Signed-off-by: Michal Wilczynski <michal.wilczynski@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
To fully utilize offload capabilities of Intel 100G card QoS capabilities
new attribute 'tx_weight' needs to be introduced. This attribute allows
for usage of Weighted Fair Queuing arbitration scheme among siblings.
This arbitration scheme can be used simultaneously with the strict
priority.
Introduce new attribute in devlink-rate that will allow for configuration
of Weighted Fair Queueing. New attribute is optional.
Signed-off-by: Michal Wilczynski <michal.wilczynski@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
To fully utilize offload capabilities of Intel 100G card QoS capabilities
new attribute 'tx_priority' needs to be introduced. This attribute allows
for usage of strict priority arbiter among siblings. This arbitration
scheme attempts to schedule nodes based on their priority as long as the
nodes remain within their bandwidth limit.
Introduce new attribute in devlink-rate that will allow for configuration
of strict priority. New attribute is optional.
Signed-off-by: Michal Wilczynski <michal.wilczynski@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The DSA tagging protocol driver macros are in the public include/net/dsa.h
probably because that's also where the DSA_TAG_PROTO_*_VALUE macros are
(MODULE_ALIAS_DSA_TAG_DRIVER hinges on those macro definitions).
But there is no reason to expose these helpers to <net/dsa.h>. That
header is shared between switch drivers (drivers/net/dsa/), tagging
protocol drivers (net/dsa/tag_*.c), the DSA core (net/dsa/ sans tag_*.c),
and the rest of the world (DSA master drivers, network stack, etc).
Too much exposure.
On the other hand, net/dsa/dsa_priv.h is included only by the DSA core
and by DSA tagging protocol drivers (or IOW, "friend" modules). Also a
bit too much exposure - I've contemplated creating a new header which is
only included by tagging protocol drivers, but completely separating a
new dsa_tag_proto.h from dsa_priv.h is not immediately trivial - for
example dsa_slave_to_port() is used both from the fast path and from the
control path.
So for now, move these definitions to dsa_priv.h which at least hides
them from the world.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Michael Walle <michael@walle.cc>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
include/linux/bpf.h
1f6e04a1c7b8 ("bpf: Fix offset calculation error in __copy_map_value and zero_map_value")
aa3496accc41 ("bpf: Refactor kptr_off_tab into btf_record")
f71b2f64177a ("bpf: Refactor map->off_arr handling")
https://lore.kernel.org/all/20221114095000.67a73239@canb.auug.org.au/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
kernel test robot reported warnings when build bonding module with
make W=1 O=build_dir ARCH=x86_64 SHELL=/bin/bash drivers/net/bonding/:
from ../drivers/net/bonding/bond_main.c:35:
In function ‘fortify_memcpy_chk’,
inlined from ‘iph_to_flow_copy_v4addrs’ at ../include/net/ip.h:566:2,
inlined from ‘bond_flow_ip’ at ../drivers/net/bonding/bond_main.c:3984:3:
../include/linux/fortify-string.h:413:25: warning: call to ‘__read_overflow2_field’ declared with attribute warning: detected read beyond size of f
ield (2nd parameter); maybe use struct_group()? [-Wattribute-warning]
413 | __read_overflow2_field(q_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In function ‘fortify_memcpy_chk’,
inlined from ‘iph_to_flow_copy_v6addrs’ at ../include/net/ipv6.h:900:2,
inlined from ‘bond_flow_ip’ at ../drivers/net/bonding/bond_main.c:3994:3:
../include/linux/fortify-string.h:413:25: warning: call to ‘__read_overflow2_field’ declared with attribute warning: detected read beyond size of f
ield (2nd parameter); maybe use struct_group()? [-Wattribute-warning]
413 | __read_overflow2_field(q_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This is because we try to copy the whole ip/ip6 address to the flow_key,
while we only point the to ip/ip6 saddr. Note that since these are UAPI
headers, __struct_group() is used to avoid the compiler warnings.
Reported-by: kernel test robot <lkp@intel.com>
Fixes: c3f8324188fa ("net: Add full IPv6 addresses to flow_keys")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Link: https://lore.kernel.org/r/20221115142400.1204786-1-liuhangbin@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
sk->sk_user_data has multiple users, which are not compatible with each
other. Writers must synchronize by grabbing the sk->sk_callback_lock.
l2tp currently fails to grab the lock when modifying the underlying tunnel
socket fields. Fix it by adding appropriate locking.
We err on the side of safety and grab the sk_callback_lock also inside the
sk_destruct callback overridden by l2tp, even though there should be no
refs allowing access to the sock at the time when sk_destruct gets called.
v4:
- serialize write to sk_user_data in l2tp sk_destruct
v3:
- switch from sock lock to sk_callback_lock
- document write-protection for sk_user_data
v2:
- update Fixes to point to origin of the bug
- use real names in Reported/Tested-by tags
Cc: Tom Parkin <tparkin@katalix.com>
Fixes: 3557baabf280 ("[L2TP]: PPP over L2TP driver core")
Reported-by: Haowei Yan <g1042620637@gmail.com>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Long standing KCSAN issues are caused by data-race around
some dev->stats changes.
Most performance critical paths already use per-cpu
variables, or per-queue ones.
It is reasonable (and more correct) to use atomic operations
for the slow paths.
This patch adds an union for each field of net_device_stats,
so that we can convert paths that are not yet protected
by a spinlock or a mutex.
netdev_stats_to_stats64() no longer has an #if BITS_PER_LONG==64
Note that the memcpy() we were using on 64bit arches
had no provision to avoid load-tearing,
while atomic_long_read() is providing the needed protection
at no cost.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
The maximum hash table size is 64K due to the nature of the protocol. [0]
It's smaller than TCP, and fewer sockets can cause a performance drop.
On an EC2 c5.24xlarge instance (192 GiB memory), after running iperf3 in
different netns, creating 32Mi sockets without data transfer in the root
netns causes regression for the iperf3's connection.
uhash_entries sockets length Gbps
64K 1 1 5.69
1Mi 16 5.27
2Mi 32 4.90
4Mi 64 4.09
8Mi 128 2.96
16Mi 256 2.06
32Mi 512 1.12
The per-netns hash table breaks the lengthy lists into shorter ones. It is
useful on a multi-tenant system with thousands of netns. With smaller hash
tables, we can look up sockets faster, isolate noisy neighbours, and reduce
lock contention.
The max size of the per-netns table is 64K as well. This is because the
possible hash range by udp_hashfn() always fits in 64K within the same
netns and we cannot make full use of the whole buckets larger than 64K.
/* 0 < num < 64K -> X < hash < X + 64K */
(num + net_hash_mix(net)) & mask;
Also, the min size is 128. We use a bitmap to search for an available
port in udp_lib_get_port(). To keep the bitmap on the stack and not
fire the CONFIG_FRAME_WARN error at build time, we round up the table
size to 128.
The sysctl usage is the same with TCP:
$ dmesg | cut -d ' ' -f 6- | grep "UDP hash"
UDP hash table entries: 65536 (order: 9, 2097152 bytes, vmalloc)
# sysctl net.ipv4.udp_hash_entries
net.ipv4.udp_hash_entries = 65536 # can be changed by uhash_entries
# sysctl net.ipv4.udp_child_hash_entries
net.ipv4.udp_child_hash_entries = 0 # disabled by default
# ip netns add test1
# ip netns exec test1 sysctl net.ipv4.udp_hash_entries
net.ipv4.udp_hash_entries = -65536 # share the global table
# sysctl -w net.ipv4.udp_child_hash_entries=100
net.ipv4.udp_child_hash_entries = 100
# ip netns add test2
# ip netns exec test2 sysctl net.ipv4.udp_hash_entries
net.ipv4.udp_hash_entries = 128 # own a per-netns table with 2^n buckets
We could optimise the hash table lookup/iteration further by removing
the netns comparison for the per-netns one in the future. Also, we
could optimise the sparse udp_hslot layout by putting it in udp_table.
[0]: https://lore.kernel.org/netdev/4ACC2815.7010101@gmail.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We will soon introduce an optional per-netns hash table
for UDP.
This means we cannot use the global sk->sk_prot->h.udp_table
to fetch a UDP hash table.
Instead, set NULL to sk->sk_prot->h.udp_table for UDP and get
a proper table from net->ipv4.udp_table.
Note that we still need sk->sk_prot->h.udp_table for UDP LITE.
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
When built with Control Flow Integrity, function prototypes between
caller and function declaration must match. These mismatches are visible
at compile time with the new -Wcast-function-type-strict in Clang[1].
Fix a total of 73 warnings like these:
drivers/net/wireless/intersil/orinoco/wext.c:1379:27: warning: cast from 'int (*)(struct net_device *, struct iw_request_info *, struct iw_param *, char *)' to 'iw_handler' (aka 'int (*)(struct net_device *, struct iw_request_info *, union iwreq_data *, char *)') converts to incompatible function type [-Wcast-function-type-strict]
IW_HANDLER(SIOCGIWPOWER, (iw_handler)orinoco_ioctl_getpower),
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../net/wireless/wext-compat.c:1607:33: warning: cast from 'int (*)(struct net_device *, struct iw_request_info *, struct iw_point *, char *)' to 'iw_handler' (aka 'int (*)(struct net_device *, struct iw_request_info *, union iwreq_data *, char *)') converts to incompatible function type [-Wcast-function-type-strict]
[IW_IOCTL_IDX(SIOCSIWGENIE)] = (iw_handler) cfg80211_wext_siwgenie,
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../drivers/net/wireless/intersil/orinoco/wext.c:1390:27: error: incompatible function pointer types initializing 'const iw_handler' (aka 'int (*const)(struct net_device *, struct iw_request_info *, union iwreq_data *, char *)') with an expression of type 'int (struct net_device *, struct iw_request_info *, struct iw_param *, char *)' [-Wincompatible-function-pointer-types]
IW_HANDLER(SIOCGIWRETRY, cfg80211_wext_giwretry),
^~~~~~~~~~~~~~~~~~~~~~
The cfg80211 Wireless Extension handler callbacks (iw_handler) use a
union for the data argument. Actually use the union and perform explicit
member selection in the function body instead of having a function
prototype mismatch. There are no resulting binary differences
before/after changes.
These changes were made partly manually and partly with the help of
Coccinelle.
Link: https://github.com/KSPP/linux/issues/234
Link: https://reviews.llvm.org/D134831 [1]
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/a68822bf8dd587988131bb6a295280cb4293f05d.1667934775.git.gustavoars@kernel.org
|
|
As of now, no DSA driver uses a custom link mode validation procedure
anymore. So remove this DSA operation and let phylink determine what is
supported based on config->mac_capabilities (if provided by the driver).
Leave a comment why we left the code that we did, and that there is more
work to do.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Pablo Neira Ayuso says:
====================
Netfilter updates for net-next
1) Fix sparse warning in the new nft_inner expression, reported
by Jakub Kicinski.
2) Incorrect vlan header check in nft_inner, from Peng Wu.
3) Two patches to pass reset boolean to expression dump operation,
in preparation for allowing to reset stateful expressions in rules.
This adds a new NFT_MSG_GETRULE_RESET command. From Phil Sutter.
4) Inconsistent indentation in nft_fib, from Jiapeng Chong.
5) Speed up siphash calculation in conntrack, from Florian Westphal.
* git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next:
netfilter: conntrack: use siphash_4u64
netfilter: rpfilter/fib: clean up some inconsistent indenting
netfilter: nf_tables: Introduce NFT_MSG_GETRULE_RESET
netfilter: nf_tables: Extend nft_expr_ops::dump callback parameters
netfilter: nft_inner: fix return value check in nft_inner_parse_l2l3()
netfilter: nft_payload: use __be16 to store gre version
====================
Link: https://lore.kernel.org/r/20221115095922.139954-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Analogous to NFT_MSG_GETOBJ_RESET, but for rules: Reset stateful
expressions like counters or quotas. The latter two are the only
consumers, adjust their 'dump' callbacks to respect the parameter
introduced earlier.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
Add a 'reset' flag just like with nft_object_ops::dump. This will be
useful to reset "anonymous stateful objects", e.g. simple rule counters.
No functional change intended.
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
|
|
This adds a new flow_rule_match_arp function that allows drivers
to be able to dissect ARP frames.
Signed-off-by: Steen Hegelund <steen.hegelund@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
gdma_obj_handle_t is no more than redefinition of basic
u64 type. Remove such obfuscation.
Link: https://lore.kernel.org/r/3c1e821279e6a165d058655d2343722d6650e776.1668160486.git.leonro@nvidia.com
Acked-by: Long Li <longli@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
Add a RDMA VF driver for Microsoft Azure Network Adapter (MANA).
Co-developed-by: Ajay Sharma <sharmaajay@microsoft.com>
Signed-off-by: Ajay Sharma <sharmaajay@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-13-git-send-email-longli@linuxonhyperv.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Long Li says:
====================
Introduce Microsoft Azure Network Adapter (MANA) RDMA driver [netdev prep]
The first 11 patches which modify the MANA Ethernet driver to support
RDMA driver.
* 'mana-shared-6.2' of https://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
net: mana: Define data structures for protection domain and memory registration
net: mana: Define data structures for allocating doorbell page from GDMA
net: mana: Define and process GDMA response code GDMA_STATUS_MORE_ENTRIES
net: mana: Define max values for SGL entries
net: mana: Move header files to a common location
net: mana: Record port number in netdev
net: mana: Export Work Queue functions for use by RDMA driver
net: mana: Set the DMA device max segment size
net: mana: Handle vport sharing between devices
net: mana: Record the physical address for doorbell page region
net: mana: Add support for auxiliary device
====================
Link: https://lore.kernel.org/all/1667502990-2559-1-git-send-email-longli@linuxonhyperv.com/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
The MANA hardware support protection domain and memory registration for use
in RDMA environment. Add those definitions and expose them for use by the
RDMA driver.
Signed-off-by: Ajay Sharma <sharmaajay@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-12-git-send-email-longli@linuxonhyperv.com
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Acked-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
The RDMA device needs to allocate doorbell pages for each user context.
Define the GDMA data structures for use by the RDMA driver.
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-11-git-send-email-longli@linuxonhyperv.com
Acked-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
When doing memory registration, the PF may respond with
GDMA_STATUS_MORE_ENTRIES to indicate a follow request is needed. This is
not an error and should be processed as expected.
Signed-off-by: Ajay Sharma <sharmaajay@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-10-git-send-email-longli@linuxonhyperv.com
Acked-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
The number of maximum SGl entries should be computed from the maximum
WQE size for the intended queue type and the corresponding OOB data
size. This guarantees the hardware queue can successfully queue requests
up to the queue depth exposed to the upper layer.
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-9-git-send-email-longli@linuxonhyperv.com
Acked-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
In preparation to add MANA RDMA driver, move all the required header files
to a common location for use by both Ethernet and RDMA drivers.
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Long Li <longli@microsoft.com>
Link: https://lore.kernel.org/r/1667502990-2559-8-git-send-email-longli@linuxonhyperv.com
Acked-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
|
|
Add packet traps for 802.1X operation. The "eapol" control trap is used
to trap EAPOL packets and is required for the correct operation of the
control plane. The "locked_port" drop trap can be enabled to gain
visibility into packets that were dropped by the device due to the
locked bridge port check.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
When the bridge is offloaded to hardware, FDB entries are learned and
aged-out by the hardware. Some device drivers synchronize the hardware
and software FDBs by generating switchdev events towards the bridge.
When a port is locked, the hardware must not learn autonomously, as
otherwise any host will blindly gain authorization. Instead, the
hardware should generate events regarding hosts that are trying to gain
authorization and their MAC addresses should be notified by the device
driver as locked FDB entries towards the bridge driver.
Allow device drivers to notify the bridge driver about such entries by
extending the 'switchdev_notifier_fdb_info' structure with the 'locked'
bit. The bit can only be set by device drivers and not by the bridge
driver.
Prevent a locked entry from being installed if MAB is not enabled on the
bridge port.
If an entry already exists in the bridge driver, reject the locked entry
if the current entry does not have the "locked" flag set or if it points
to a different port. The same semantics are implemented in the software
data path.
Signed-off-by: Hans J. Schultz <netdev@kapio-technology.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs
rxrpc changes
David Howells says:
====================
rxrpc: Increasing SACK size and moving away from softirq, part 1
AF_RXRPC has some issues that need addressing:
(1) The SACK table has a maximum capacity of 255, but for modern networks
that isn't sufficient. This is hard to increase in the upstream code
because of the way the application thread is coupled to the softirq
and retransmission side through a ring buffer. Adjustments to the rx
protocol allows a capacity of up to 8192, and having a ring
sufficiently large to accommodate that would use an excessive amount
of memory as this is per-call.
(2) Processing ACKs in softirq mode causes the ACKs get conflated, with
only the most recent being considered. Whilst this has the upside
that the retransmission algorithm only needs to deal with the most
recent ACK, it causes DATA transmission for a call to be very bursty
because DATA packets cannot be transmitted in softirq mode. Rather
transmission must be delegated to either the application thread or a
workqueue, so there tend to be sudden bursts of traffic for any
particular call due to scheduling delays.
(3) All crypto in a single call is done in series; however, each DATA
packet is individually encrypted so encryption and decryption of large
calls could be parallelised if spare CPU resources are available.
This is the first of a number of sets of patches that try and address them.
The overall aims of these changes include:
(1) To get rid of the TxRx ring and instead pass the packets round in
queues (eg. sk_buff_head). On the Tx side, each ACK packet comes with
a SACK table that can be parsed as-is, so there's no particular need
to maintain our own; we just have to refer to the ACK.
On the Rx side, we do need to maintain a SACK table with one bit per
entry - but only if packets go missing - and we don't want to have to
perform a complex transformation to get the information into an ACK
packet.
(2) To try and move almost all processing of received packets out of the
softirq handler and into a high-priority kernel I/O thread. Only the
transferral of packets would be left there. I would still use the
encap_rcv hook to receive packets as there's a noticeable performance
drop from letting the UDP socket put the packets into its own queue
and then getting them out of there.
(3) To make the I/O thread also do all the transmission. The app thread
would be responsible for packaging the data into packets and then
buffering them for the I/O thread to transmit. This would make it
easier for the app thread to run ahead of the I/O thread, and would
mean the I/O thread is less likely to have to wait around for a new
packet to come available for transmission.
(4) To logically partition the socket/UAPI/KAPI side of things from the
I/O side of things. The local endpoint, connection, peer and call
objects would belong to the I/O side. The socket side would not then
touch the private internals of calls and suchlike and would not change
their states. It would only look at the send queue, receive queue and
a way to pass a message to cause an abort.
(5) To remove as much locking, synchronisation, barriering and atomic ops
as possible from the I/O side. Exclusion would be achieved by
limiting modification of state to the I/O thread only. Locks would
still need to be used in communication with the UDP socket and the
AF_RXRPC socket API.
(6) To provide crypto offload kernel threads that, when there's slack in
the system, can see packets that need crypting and provide
parallelisation in dealing with them.
(7) To remove the use of system timers. Since each timer would then send
a poke to the I/O thread, which would then deal with it when it had
the opportunity, there seems no point in using system timers if,
instead, a list of timeouts can be sensibly consulted. An I/O thread
only then needs to schedule with a timeout when it is idle.
(8) To use zero-copy sendmsg to send packets. This would make use of the
I/O thread being the sole transmitter on the socket to manage the
dead-reckoning sequencing of the completion notifications. There is a
problem with zero-copy, though: the UDP socket doesn't handle running
out of option memory very gracefully.
With regard to this first patchset, the changes made include:
(1) Some fixes, including a fallback for proc_create_net_single_write(),
setting ack.bufferSize to 0 in ACK packets and a fix for rxrpc
congestion management, which shouldn't be saving the cwnd value
between calls.
(2) Improvements in rxrpc tracepoints, including splitting the timer
tracepoint into a set-timer and a timer-expired trace.
(3) Addition of a new proc file to display some stats.
(4) Some code cleanups, including removing some unused bits and
unnecessary header inclusions.
(5) A change to the recently added UDP encap_err_rcv hook so that it has
the same signature as {ip,ipv6}_icmp_error(), and then just have rxrpc
point its UDP socket's hook directly at those.
(6) Definition of a new struct, rxrpc_txbuf, that is used to hold
transmissible packets of DATA and ACK type in a single 2KiB block
rather than using an sk_buff. This allows the buffer to be on a
number of queues simultaneously more easily, and also guarantees that
the entire block is in a single unit for zerocopy purposes and that
the data payload is aligned for in-place crypto purposes.
(7) ACK txbufs are allocated at proposal and queued for later transmission
rather than being stored in a single place in the rxrpc_call struct,
which means only a single ACK can be pending transmission at a time.
The queue is then drained at various points. This allows the ACK
generation code to be simplified.
(8) The Rx ring buffer is removed. When a jumbo packet is received (which
comprises a number of ordinary DATA packets glued together), it used
to be pointed to by the ring multiple times, with an annotation in a
side ring indicating which subpacket was in that slot - but this is no
longer possible. Instead, the packet is cloned once for each
subpacket, barring the last, and the range of data is set in the skb
private area. This makes it easier for the subpackets in a jumbo
packet to be decrypted in parallel.
(9) The Tx ring buffer is removed. The side annotation ring that held the
SACK information is also removed. Instead, in the event of packet
loss, the SACK data attached an ACK packet is parsed.
(10) Allocate an skcipher request when needed in the rxkad security class
rather than caching one in the rxrpc_call struct. This deals with a
race between externally-driven call disconnection getting rid of the
skcipher request and sendmsg/recvmsg trying to use it because they
haven't seen the completion yet. This is also needed to support
parallelisation as the skcipher request cannot be used by two or more
threads simultaneously.
(11) Call udp_sendmsg() and udpv6_sendmsg() directly rather than going
through kernel_sendmsg() so that we can provide our own iterator
(zerocopy explicitly doesn't work with a KVEC iterator). This also
lets us avoid the overhead of the security hook.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Change the udp encap_err_rcv signature to match ip_icmp_error() and
ipv6_icmp_error() so that those can be used from the called function and
export them.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
cc: netdev@vger.kernel.org
|
|
This patch is to add helper support in act_ct for OVS actions=ct(alg=xxx)
offloading, which is corresponding to Commit cae3a2627520 ("openvswitch:
Allow attaching helpers to ct action") in OVS kernel part.
The difference is when adding TC actions family and proto cannot be got
from the filter/match, other than helper name in tb[TCA_CT_HELPER_NAME],
we also need to send the family in tb[TCA_CT_HELPER_FAMILY] and the
proto in tb[TCA_CT_HELPER_PROTO] to kernel.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Move ovs_ct_add_helper from openvswitch to nf_conntrack_helper and
rename as nf_ct_add_helper, so that it can be used in TC act_ct in
the next patch.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Move ovs_ct_helper from openvswitch to nf_conntrack_helper and rename
as nf_ct_helper so that it can be used in TC act_ct in the next patch.
Note that it also adds the checks for the family and proto, as in TC
act_ct, the packets with correct family and proto are not guaranteed.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Let families to hook in the new split ops.
They are more flexible and should not be much larger than
full ops. Each split op is 40B while full op is 48B.
Devlink for example has 54 dos and 19 dumps, 2 of the dumps
do not have a do -> 56 full commands = 2688B.
Split ops would have taken 2920B, so 9% more space while
allowing individual per/post doit and per-type policies.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
We currently have two forms of operations - small ops and "full" ops
(or just ops). The former does not have pointers for some of the less
commonly used features (namely dump start/done and policy).
The "full" ops, however, still don't contain all the necessary
information. In particular the policy is per command ID, while
do and dump often accept different attributes. It's also not
possible to define different pre_doit and post_doit callbacks
for different commands within the family.
At the same time a lot of commands do not support dumping and
therefore all the dump-related information is wasted space.
Create a new command representation which can hold info about
a do implementation or a dump implementation, but not both at
the same time.
Use this new representation on the command execution path
(genl_family_rcv_msg) as we either run a do or a dump and
don't have to create a "full" op there.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Move the private fields down to form a "private section".
Use the kdoc "private:" label comment thing to hide them
from the main kdoc comment.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Johannes Berg <johannes@sipsolutions.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
|
Expose devlink port handle related to netdev over rtnetlink. Introduce a
new nested IFLA attribute to carry the info. Call into devlink code to
fill-up the nest with existing devlink attributes that are used over
devlink netlink.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
without RTNL held
To avoid a need to take RTNL mutex in port_fill() function, benefit from
the introduce infrastructure that tracks netdevice notifier events.
Store the ifindex and ifname upon register and change name events.
Remove the rtnl_held bool propagated down to port_fill() function as it
is no longer needed.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Since devlink_port_type_eth_set() should no longer be called by any
driver with netdev pointer as it should rather use
SET_NETDEV_DEVLINK_PORT, remove the netdev arg. Add a warn to
type_clear()
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Instead of storing type_dev as a void pointer, convert it to union and
use it to store either struct net_device or struct ib_device pointer.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
No conflicts.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Add new apptrust extension attributes to the 8021Qaz APP managed object.
Two new attributes, DCB_ATTR_DCB_APP_TRUST_TABLE and
DCB_ATTR_DCB_APP_TRUST, has been added. Trusted selectors are passed in
the nested attribute DCB_ATTR_DCB_APP_TRUST, in order of precedence.
The new attributes are meant to allow drivers, whose hw supports the
notion of trust, to be able to set whether a particular app selector is
trusted - and in which order.
Signed-off-by: Daniel Machon <daniel.machon@microchip.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
|
|
Both bond_alb_xmit() and bond_tlb_xmit() produce a valid warning with
gcc-13:
drivers/net/bonding/bond_alb.c:1409:13: error: conflicting types for 'bond_tlb_xmit' due to enum/integer mismatch; have 'netdev_tx_t(struct sk_buff *, struct net_device *)' ...
include/net/bond_alb.h:160:5: note: previous declaration of 'bond_tlb_xmit' with type 'int(struct sk_buff *, struct net_device *)'
drivers/net/bonding/bond_alb.c:1523:13: error: conflicting types for 'bond_alb_xmit' due to enum/integer mismatch; have 'netdev_tx_t(struct sk_buff *, struct net_device *)' ...
include/net/bond_alb.h:159:5: note: previous declaration of 'bond_alb_xmit' with type 'int(struct sk_buff *, struct net_device *)'
I.e. the return type of the declaration is int, while the definitions
spell netdev_tx_t. Synchronize both of them to the latter.
Cc: Martin Liska <mliska@suse.cz>
Cc: Jay Vosburgh <j.vosburgh@gmail.com>
Cc: Veaceslav Falico <vfalico@gmail.com>
Cc: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: Jiri Slaby (SUSE) <jirislaby@kernel.org>
Link: https://lore.kernel.org/r/20221031114409.10417-1-jirislaby@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Jakub reported that the addition of the "network_byte_order"
member in struct nla_policy increases size of 32bit platforms.
Instead of scraping the bit from elsewhere Johannes suggested
to add explicit NLA_BE types instead, so do this here.
NLA_POLICY_MAX_BE() macro is removed again, there is no need
for it: NLA_POLICY_MAX(NLA_BE.., ..) will do the right thing.
NLA_BE64 can be added later.
Fixes: 08724ef69907 ("netlink: introduce NLA_POLICY_MAX_BE")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Suggested-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Link: https://lore.kernel.org/r/20221031123407.9158-1-fw@strlen.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
IPv4 reassembly unit can decide to drop frags based on
/proc/sys/net/ipv4/ipfrag_max_dist sysctl.
Add a specific drop reason to track this specific
and weird case.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
Used to track skbs freed after a timeout happened
in a reassmbly unit.
Passing a @reason argument to inet_frag_rbtree_purge()
allows to use correct consumed status for frags
that have been successfully re-assembled.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This is used to track when a duplicate segment received by various
reassembly units is dropped.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This will allow to simply use in the future:
kfree_skb_reason(skb, reason);
Instead of repeating sequences like:
if (dropped)
kfree_skb_reason(skb, reason);
else
consume_skb(skb);
For instance, following patch in the series is adding
@reason to skb_release_data() and skb_release_all(),
so that we can propagate a meaningful @reason whenever
consume_skb()/kfree_skb() have to take care of a potential frag_list.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This patch use the new helper unregister_netdevice_many_notify() for
rtnl_delete_link(), so that the kernel could reply unicast when userspace
set NLM_F_ECHO flag to request the new created interface info.
At the same time, the parameters of rtnl_delete_link() need to be updated
since we need nlmsghdr and portid info.
Suggested-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
|
This patch pass netlink message header and portid to rtnl_configure_link()
All the functions in this call chain need to add the parameters so we can
use them in the last call rtnl_notify(), and notify the userspace about
the new link info if NLM_F_ECHO flag is set.
- rtnl_configure_link()
- __dev_notify_flags()
- rtmsg_ifinfo()
- rtmsg_ifinfo_event()
- rtmsg_ifinfo_build_skb()
- rtmsg_ifinfo_send()
- rtnl_notify()
Also move __dev_notify_flags() declaration to net/core/dev.h, as Jakub
suggested.
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
