The semantics for bpf_sk_assign are as follows:

    sk = some_lookup_func()
    bpf_sk_assign(skb, sk)
    bpf_sk_release(sk)

That is, the sk is not consumed by bpf_sk_assign. The function
therefore needs to make sure that sk lives long enough to be
consumed from __inet_lookup_skb. The path through the stack for a
TCPv4 packet is roughly:

  netif_receive_skb_core: takes RCU read lock
    __netif_receive_skb_core:
      sch_handle_ingress:
        tcf_classify:
          bpf_sk_assign()
      deliver_ptype_list_skb:
        deliver_skb:
          ip_packet_type->func == ip_rcv:
            ip_rcv_core:
            ip_rcv_finish_core:
              dst_input:
                ip_local_deliver:
                  ip_local_deliver_finish:
                    ip_protocol_deliver_rcu:
                      tcp_v4_rcv:
                        __inet_lookup_skb:
                          skb_steal_sock

The existing helper takes advantage of the fact that everything
happens in the same RCU critical section: for sockets with
SOCK_RCU_FREE set bpf_sk_assign never takes a reference.
skb_steal_sock then checks SOCK_RCU_FREE again and does sock_put
if necessary.

This approach assumes that SOCK_RCU_FREE is never set on a sk
between bpf_sk_assign and skb_steal_sock, but this invariant is
violated by unhashed UDP sockets. A new UDP socket is created
in TCP_CLOSE state but without SOCK_RCU_FREE set. That flag is only
added in udp_lib_get_port() which happens when a socket is bound.

When bpf_sk_assign was added it wasn't possible to access unhashed
UDP sockets from BPF, so this wasn't a problem. This changed
in commit 0c48eefa ("sock_map: Lift socket state restriction
for datagram sockets"), but the helper wasn't adjusted accordingly.
The following sequence of events will therefore lead to a refcount
leak:

1. Add socket(AF_INET, SOCK_DGRAM) to a sockmap.
2. Pull socket out of sockmap and bpf_sk_assign it. Since
   SOCK_RCU_FREE is not set we increment the refcount.
3. bind() or connect() the socket, setting SOCK_RCU_FREE.
4. skb_steal_sock will now set refcounted = false due to
   SOCK_RCU_FREE.
5. tcp_v4_rcv() skips sock_put().

Fix the problem by rejecting unhashed sockets in bpf_sk_assign().
This matches the behaviour of __inet_lookup_skb which is ultimately
the goal of bpf_sk_assign().

Fixes: cf7fbe66 ("bpf: Add socket assign support")
Cc: Joe Stringer <joe@cilium.io>
Signed-off-by: Lorenz Bauer <lmb@isovalent.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/20230720-so-reuseport-v6-2-7021b683cdae@isovalent.comSigned-off-by: Martin KaFai Lau <martin.lau@kernel.org>

Contrary to TCP, UDP reuseport groups can contain TCP_ESTABLISHED
sockets. To support these properly we remember whether a group has
a connected socket and skip the fast reuseport early-return. In
effect we continue scoring all reuseport sockets and then choose the
one with the highest score.

The current code fails to re-calculate the score for the result of
lookup_reuseport. According to Kuniyuki Iwashima:

    1) SO_INCOMING_CPU is set
       -> selected sk might have +1 score

    2) BPF prog returns ESTABLISHED and/or SO_INCOMING_CPU sk
       -> selected sk will have more than 8

  Using the old score could trigger more lookups depending on the
  order that sockets are created.

    sk -> sk (SO_INCOMING_CPU) -> sk (ESTABLISHED)
    |     |
    `-> select the next SO_INCOMING_CPU sk
          |
          `-> select itself (We should save this lookup)

Fixes: efc6b6f6 ("udp: Improve load balancing for SO_REUSEPORT.")
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Lorenz Bauer <lmb@isovalent.com>
Link: https://lore.kernel.org/r/20230720-so-reuseport-v6-1-7021b683cdae@isovalent.comSigned-off-by: Martin KaFai Lau <martin.lau@kernel.org>

Switch from corporate email address to linux.dev address.
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20230725054100.1013421-1-yonghong.song@linux.devSigned-off-by: Martin KaFai Lau <martin.lau@kernel.org>

On qdisc destruction, the ingress_destroy() needs to update the correct
entry, that is, tcx_entry_update must NULL the dev->tcx_ingress pointer.
Therefore, fix the typo.

Fixes: e420bed0 ("bpf: Add fd-based tcx multi-prog infra with link support")
Reported-by: syzbot+bdcf141f362ef83335cf@syzkaller.appspotmail.com
Reported-by: syzbot+b202b7208664142954fa@syzkaller.appspotmail.com
Reported-by: syzbot+14736e249bce46091c18@syzkaller.appspotmail.com
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: syzbot+bdcf141f362ef83335cf@syzkaller.appspotmail.com
Tested-by: syzbot+b202b7208664142954fa@syzkaller.appspotmail.com
Tested-by: syzbot+14736e249bce46091c18@syzkaller.appspotmail.com
Tested-by: Petr Machata <petrm@nvidia.com>
Link: https://lore.kernel.org/r/20230721233330.5678-1-daniel@iogearbox.netSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Shannon Nelson says:

====================
ionic: add FLR support

Add support for handing and recovering from a PCI FLR event.
This patchset first moves some code around to make it usable
from multiple paths, then adds the PCI error handler callbacks
for reset_prepare and reset_done.

Example test:
    echo 1 > /sys/bus/pci/devices/0000:2a:00.0/reset

v4:
 - don't remove ionic_dev_teardown() in ionic_probe() in patch 2/4
 - remove clear_bit() change from patch 3/4

v3:
Link: https://lore.kernel.org/netdev/20230717170001.30539-1-shannon.nelson@amd.com/
 - removed first patch, it is already merged into net

v2:
Link: https://lore.kernel.org/netdev/20230713192936.45152-1-shannon.nelson@amd.com/
 - removed redundant pci_save/restore_state() calls
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for the PCI reset handlers in order to manage an FLR event.
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull out some code from ionic_lif_handle_fw_up() that can be
used in the coming FLR recovery patch.
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull out some chunks of code from ionic_probe() that will
be common in rebuild paths.
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Pull out a chunk of code from ionic_remove() that will
be common in teardown paths.
Signed-off-by: Shannon Nelson <shannon.nelson@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Samin Guo says:

====================
Add motorcomm phy pad-driver-strength-cfg support

The motorcomm phy (YT8531) supports the ability to adjust the drive
strength of the rx_clk/rx_data, and the default strength may not be
suitable for all boards. So add configurable options to better match
the boards.(e.g. StarFive VisionFive 2)

The first patch adds a description of dt-bingding, and the second patch adds
YT8531's parsing and settings for pad-driver-strength-cfg.

Changes since v4:
Patch 1:
- Removed register-related DS(3b) values and added vol descriptions (by Andrew Lunn)
- Dropped the type and added '-microamp' suffix. (by Rob Herring)
Patch 2:
- Return -EINVAL if the value in DT but it is invalid (by Andrew Lunn)

Changes since v3:
Patch 1:
- Used current values instead of register values
- Added units and numerical descriptions of driver-strength
Patch 2:
- Added a lookup table to listing the valid values in the schema (by Andrew Lunn)

Changes since v2:
Patch 2:
- Readjusted the order of YT8531_RGMII_xxx to below YTPHY_PAD_DRIVE_STRENGTH_REG (by Frank Sae)
- Reversed Christmas tree, sort these longest first, shortest last (by Andrew Lunn)
- Rebased on tag v6.4

Changes since v1:
Patch 1:
- Renamed "rx-xxx-driver-strength" to "motorcomm,rx-xxx-driver-strength" (by Frank Sae)
Patch 2:
- Added default values for rxc/rxd driver strength (by Frank Sea/Andrew Lunn)
- Added range checking when val is in DT (by Frank Sea/Andrew Lunn)

Previous versions:
v1 - https://patchwork.kernel.org/project/netdevbpf/cover/20230426063541.15378-1-samin.guo@starfivetech.com
v2 - https://patchwork.kernel.org/project/netdevbpf/cover/20230505090558.2355-1-samin.guo@starfivetech.com
v3 - https://patchwork.kernel.org/project/netdevbpf/cover/20230526090502.29835-1-samin.guo@starfivetech.com
v4 - https://patchwork.kernel.org/project/netdevbpf/cover/20230714101406.17686-1-samin.guo@starfivetech.com
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The motorcomm phy (YT8531) supports the ability to adjust the drive
strength of the rx_clk/rx_data, and the default strength may not be
suitable for all boards. So add configurable options to better match
the boards.(e.g. StarFive VisionFive 2)

When we configure the drive strength, we need to read the current
LDO voltage value to ensure that it is a legal value at that LDO
voltage.
Reviewed-by: Hal Feng <hal.feng@starfivetech.com>
Signed-off-by: Samin Guo <samin.guo@starfivetech.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

The motorcomm phy (YT8531) supports the ability to adjust the drive
strength of the rx_clk/rx_data.

The YT8531 RGMII LDO voltage supports 1.8V/3.3V, and the
LDO voltage can be configured with hardware pull-up resistors to match
the SOC voltage (usually 1.8V). The software can read the registers
0xA001 obtain the current LDO voltage value.
Reviewed-by: Hal Feng <hal.feng@starfivetech.com>
Signed-off-by: Samin Guo <samin.guo@starfivetech.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

IPv6 inet sockets are supposed to have a "struct ipv6_pinfo"
field at the end of their definition, so that inet6_sk_generic()
can derive from socket size the offset of the "struct ipv6_pinfo".

This is very fragile, and prevents adding bigger alignment
in sockets, because inet6_sk_generic() does not work
if the compiler adds padding after the ipv6_pinfo component.

We are currently working on a patch series to reorganize
TCP structures for better data locality and found issues
similar to the one fixed in commit f5d54767
("tcp: fix tcp_inet6_sk() for 32bit kernels")

Alternative would be to force an alignment on "struct ipv6_pinfo",
greater or equal to __alignof__(any ipv6 sock) to ensure there is
no padding. This does not look great.

v2: fix typo in mptcp_proto_v6_init() (Paolo)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Chao Wu <wwchao@google.com>
Cc: Wei Wang <weiwan@google.com>
Cc: Coco Li <lixiaoyan@google.com>
Cc: YiFei Zhu <zhuyifei@google.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This change adds a new sysctl accept_ra_min_rtr_lft to specify the
minimum acceptable router lifetime in an RA. If the received RA router
lifetime is less than the configured value (and not 0), the RA is
ignored.
This is useful for mobile devices, whose battery life can be impacted
by networks that configure RAs with a short lifetime. On such networks,
the device should never gain IPv6 provisioning and should attempt to
drop RAs via hardware offload, if available.
Signed-off-by: Patrick Rohr <prohr@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

`strncpy` is deprecated for use on NUL-terminated destination strings [1].

Even call sites utilizing length-bounded destination buffers should
switch over to using `strtomem` or `strtomem_pad`. In this case,
however, the compiler is unable to determine the size of the `data`
buffer which renders `strtomem` unusable. Due to this, `strscpy`
should be used.

It should be noted that most call sites already zero-initialize the
destination buffer. However, I've opted to use `strscpy_pad` to maintain
the same exact behavior that `strncpy` produced (zero-padded tail up to
`len`).

Also see [3].

[1]: www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
[2]: elixir.bootlin.com/linux/v6.3/source/net/ethtool/ioctl.c#L1944
[3]: manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html

Link: https://github.com/KSPP/linux/issues/90Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Justin Stitt <justinstitt@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Anjali Kulkarni says:

====================
Process connector bug fixes & enhancements

Oracle DB is trying to solve a performance overhead problem it has been
facing for the past 10 years and using this patch series, we can fix this
issue.

Oracle DB runs on a large scale with 100000s of short lived processes,
starting up and exiting quickly. A process monitoring DB daemon which
tracks and cleans up after processes that have died without a proper exit
needs notifications only when a process died with a non-zero exit code
(which should be rare).

Due to the pmon architecture, which is distributed, each process is
independent and has minimal interaction with pmon. Hence fd based
solutions to track a process's spawning and exit cannot be used. Pmon
needs to detect the abnormal death of a process so it can cleanup after.
Currently it resorts to checking /proc every few seconds. Other methods
we tried like using system call to reduce the above overhead were not
accepted upstream.

With this change, we add event based filtering to proc connector module
so that DB can only listen to the events it is interested in. A new
event type PROC_EVENT_NONZERO_EXIT is added, which is only sent by kernel
to a listening application when any process exiting has a non-zero exit
status.

This change will give Oracle DB substantial performance savings - it takes
50ms to scan about 8K PIDs in /proc, about 500ms for 100K PIDs. DB does
this check every 3 secs, so over an hour we save 10secs for 100K PIDs.

With this, a client can register to listen for only exit or fork or a mix or
all of the events. This greatly enhances performance - currently, we
need to listen to all events, and there are 9 different types of events.
For eg. handling 3 types of events - 8K-forks + 8K-exits + 8K-execs takes
200ms, whereas handling 2 types - 8K-forks + 8K-exits takes about 150ms,
and handling just one type - 8K exits takes about 70ms.

Measuring the time using pidfds for monitoring 8K process exits took 4
times longer - 200ms, as compared to 70ms using only exit notifications
of proc connector. Hence, we cannot use pidfd for our use case.

This kind of a new event could also be useful to other applications like
Google's lmkd daemon, which needs a killed process's exit notification.

This patch series is organized as follows -

Patch 1 : Needed for patch 3 to work.
Patch 2 : Needed for patch 3 to work.
Patch 3 : Fixes some bugs in proc connector, details in the patch.
Patch 4 : Adds event based filtering for performance enhancements.
Patch 5 : Allow non-root users access to proc connector events.
Patch 6 : Selftest code for proc connector.

v9->v10 changes:
- Rebased to net-next, re-compiled and re-tested.

v8->v9 changes:
- Added sha1 ("title") of reversed patch as suggested by Eric Dumazet.

v7->v8 changes:
- Fixed an issue pointed by Liam Howlett in v7.

v6->v7 changes:
- Incorporated Liam Howlett's comments on v6
- Incorporated Kalesh Anakkur Purayil's comments

v5->v6 changes:
- Incorporated Liam Howlett's comments
- Removed FILTER define from proc_filter.c and added a "-f" run-time
  option to run new filter code.
- Made proc_filter.c a selftest in tools/testing/selftests/connector

v4->v5 changes:
- Change the cover letter
- Fix a small issue in proc_filter.c

v3->v4 changes:
- Fix comments by Jakub Kicinski to incorporate root access changes
  within bind call of connector

v2->v3 changes:
- Fix comments by Jakub Kicinski to separate netlink (patch 2) (after
  layering) from connector fixes (patch 3).
- Minor fixes suggested by Jakub.
- Add new multicast group level permissions check at netlink layer.
  Split this into netlink & connector layers (patches 6 & 7)

v1->v2 changes:
- Fix comments by Jakub Kicinski to keep layering within netlink and
  update kdocs.
- Move non-root users access patch last in series so remaining patches
  can go in first.

v->v1 changes:
- Changed commit log in patch 4 as suggested by Christian Brauner
- Changed patch 4 to make more fine grained access to non-root users
- Fixed warning in cn_proc.c,
Reported-by: kernel test robot <lkp@intel.com>
- Fixed some existing warnings in cn_proc.c
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Run as ./proc_filter -f to run new filter code. Run without "-f" to run
usual proc connector code without the new filtering code.
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There were a couple of reasons for not allowing non-root users access
initially  - one is there was some point no proper receive buffer
management in place for netlink multicast. But that should be long
fixed. See link below for more context.

Second is that some of the messages may contain data that is root only. But
this should be handled with a finer granularity, which is being done at the
protocol layer.  The only problematic protocols are nf_queue and the
firewall netlink. Hence, this restriction for non-root access was relaxed
for NETLINK_ROUTE initially:
https://lore.kernel.org/all/20020612013101.A22399@wotan.suse.de/

This restriction has also been removed for following protocols:
NETLINK_KOBJECT_UEVENT, NETLINK_AUDIT, NETLINK_SOCK_DIAG,
NETLINK_GENERIC, NETLINK_SELINUX.

Since process connector messages are not sensitive (process fork, exit
notifications etc.), and anyone can read /proc data, we can allow non-root
access here. However, since process event notification is not the only
consumer of NETLINK_CONNECTOR, we can make this change even more
fine grained than the protocol level, by checking for multicast group
within the protocol.

Allow non-root access for NETLINK_CONNECTOR via NL_CFG_F_NONROOT_RECV
but add new bind function cn_bind(), which allows non-root access only
for CN_IDX_PROC multicast group.
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds the capability to filter messages sent by the proc
connector on the event type supplied in the message from the client
to the connector. The client can register to listen for an event type
given in struct proc_input.

This event based filteting will greatly enhance performance - handling
8K exits takes about 70ms, whereas 8K-forks + 8K-exits takes about 150ms
& handling 8K-forks + 8K-exits + 8K-execs takes 200ms. There are currently
9 different types of events, and we need to listen to all of them. Also,
measuring the time using pidfds for monitoring 8K process exits took
much longer - 200ms, as compared to 70ms using only exit notifications of
proc connector.

We also add a new event type - PROC_EVENT_NONZERO_EXIT, which is
only sent by kernel to a listening application when any process exiting,
has a non-zero exit status. This will help the clients like Oracle DB,
where a monitoring process wants notfications for non-zero process exits
so it can cleanup after them.

This kind of a new event could also be useful to other applications like
Google's lmkd daemon, which needs a killed process's exit notification.

The patch takes care that existing clients using old mechanism of not
sending the event type work without any changes.

cn_filter function checks to see if the event type being notified via
proc connector matches the event type requested by client, before
sending(matches) or dropping(does not match) a packet.
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current proc connector code has the foll. bugs - if there are more
than one listeners for the proc connector messages, and one of them
deregisters for listening using PROC_CN_MCAST_IGNORE, they will still get
all proc connector messages, as long as there is another listener.

Another issue is if one client calls PROC_CN_MCAST_LISTEN, and another one
calls PROC_CN_MCAST_IGNORE, then both will end up not getting any messages.

This patch adds filtering and drops packet if client has sent
PROC_CN_MCAST_IGNORE. This data is stored in the client socket's
sk_user_data. In addition, we only increment or decrement
proc_event_num_listeners once per client. This fixes the above issues.

cn_release is the release function added for NETLINK_CONNECTOR. It uses
the newly added netlink_release function added to netlink_sock. It will
free sk_user_data.
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

A new function netlink_release is added in netlink_sock to store the
protocol's release function. This is called when the socket is deleted.
This can be supplied by the protocol via the release function in
netlink_kernel_cfg. This is being added for the NETLINK_CONNECTOR
protocol, so it can free it's data when socket is deleted.
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

To use filtering at the connector & cn_proc layers, we need to enable
filtering in the netlink layer. This reverses the patch which removed
netlink filtering - commit ID for that patch:
549017aa (netlink: remove netlink_broadcast_filtered).
Signed-off-by: Anjali Kulkarni <anjali.k.kulkarni@oracle.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jakub Kicinski says:

====================
net: page_pool: remove page_pool_release_page()

page_pool_return_page() is a historic artefact from before
recycling of pages attached to skbs was supported. Theoretical
uses for it may be thought up but in practice all existing
users can be converted to use skb_mark_for_recycle() instead.

This code was previously posted as part of the memory provider RFC.
https://lore.kernel.org/all/20230707183935.997267-1-kuba@kernel.org/
====================

Link: https://lore.kernel.org/r/20230720010409.1967072-1-kuba@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Now that page_pool_release_page() is not exported we can
merge it with page_pool_return_page(). I believe that
the "Do not replace this with page_pool_return_page()"
comment was there in case page_pool_return_page() was
not inlined, to avoid two function calls.
Acked-by: Jesper Dangaard Brouer <hawk@kernel.org>
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Link: https://lore.kernel.org/r/20230720010409.1967072-5-kuba@kernel.orgReviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

There seems to be no user calling page_pool_release_page()
for legit reasons, all the users simply haven't been converted
to skb-based recycling, yet. Previous changes converted them.
Update the docs, and unexport the function.

Link: https://lore.kernel.org/r/20230720010409.1967072-4-kuba@kernel.orgReviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

stmmac removes pages from the page pool after attaching them
to skbs. Use page recycling instead.

skb heads are always copied, and pages are always from page
pool in this driver. We could as well mark all allocated skbs
for recycling.
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Link: https://lore.kernel.org/r/20230720010409.1967072-3-kuba@kernel.orgReviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

tsnep builds an skb with napi_build_skb() and then calls
page_pool_release_page() for the page in which that skb's
head sits. Use recycling instead, recycling of heads works
just fine.
Reviewed-by: Yunsheng Lin <linyunsheng@huawei.com>
Link: https://lore.kernel.org/r/20230720010409.1967072-2-kuba@kernel.orgReviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Currently, if cmd in the split ops array is of lower value than the
previous one, genl_validate_ops() continues to do the checks as if
the values are equal. This may result in non-obvious WARN_ON() hit in
these check.

Instead, check the incorrect ordering explicitly and put a WARN_ON()
in case it is broken.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Link: https://lore.kernel.org/r/20230720111354.562242-1-jiri@resnulli.usSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Linus seems to like the MAINTAINERS file sorted, see
c192ac73 ("MAINTAINERS 2: Electric Boogaloo").

Since this is currently not the case, restore the sort order.

Fixes: 3abf3d15 ("MAINTAINERS: ASP 2.0 Ethernet driver maintainers")
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Acked-by: Justin Chen <justin.chen@broadcom.com>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20230720151107.679668-1-mkl@pengutronix.deSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Hariprasad Kelam says:

====================
octeontx2-pf: support Round Robin scheduling

octeontx2 and CN10K silicons support Round Robin scheduling. When multiple
traffic flows reach transmit level with the same priority, with Round Robin
scheduling traffic flow with the highest quantum value is picked. With this
support, the user can add multiple classes with the same priority and
different quantum in htb offload.

This series of patches adds support for the same.

Patch1: implement transmit schedular allocation algorithm as preparation
        for support round robin scheduling.

Patch2: Allow quantum parameter in HTB offload mode.

Patch3: extends octeontx2 htb offload support for Round Robin scheduling

Patch4: extend QOS documentation for Round Robin scheduling

Hariprasad Kelam (1):
  docs: octeontx2: extend documentation for Round Robin scheduling

Naveen Mamindlapalli (3):
  octeontx2-pf: implement transmit schedular allocation algorithm
  sch_htb: Allow HTB quantum parameter in offload mode
  octeontx2-pf: htb offload support for Round Robin scheduling
---
v4 * update classid values in documentation.

v3 * 1. update QOS documentation for round robin scheduling
     2. added out of bound checks for quantum parameter

v2 * change data type of otx2_index_used to reduce size of structure
     otx2_qos_cfg
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add example tc-htb commands for Round robin scheduling
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When multiple traffic flows reach Transmit level with the same
priority, with Round robin scheduling traffic flow with the highest
quantum value is picked. With this support, the user can add multiple
classes with the same priority and different quantum. This patch
does necessary changes to support the same.
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The current implementation of HTB offload returns the EINVAL error for
quantum parameter. This patch removes the error returning checks for
'quantum' parameter and populates its value to tc_htb_qopt_offload
structure such that driver can use the same.

Add quantum parameter check in mlx5 driver, as mlx5 devices are not capable
of supporting the quantum parameter when htb offload is used. Report error
if quantum parameter is set to a non-default value.
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

unlike strict priority, where number of classes are limited to max
8, there is no restriction on the number of dwrr child nodes unless
the count increases the max number of child nodes supported.

Hardware expects strict priority transmit schedular indexes mapped
to their priority. This patch adds defines transmit schedular allocation
algorithm such that the above requirement is honored.
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Petr Machata says:

====================
mlxsw: Permit enslavement to netdevices with uppers

The mlxsw driver currently makes the assumption that the user applies
configuration in a bottom-up manner. Thus netdevices need to be added to
the bridge before IP addresses are configured on that bridge or SVI added
on top of it. Enslaving a netdevice to another netdevice that already has
uppers is in fact forbidden by mlxsw for this reason. Despite this safety,
it is rather easy to get into situations where the offloaded configuration
is just plain wrong.

As an example, take a front panel port, configure an IP address: it gets a
RIF. Now enslave the port to the bridge, and the RIF is gone. Remove the
port from the bridge again, but the RIF never comes back. There is a number
of similar situations, where changing the configuration there and back
utterly breaks the offload.

Similarly, detaching a front panel port from a configured topology means
unoffloading of this whole topology -- VLAN uppers, next hops, etc.
Attaching the port back is then not permitted at all. If it were, it would
not result in a working configuration, because much of mlxsw is written to
react to changes in immediate configuration. There is nothing that would go
visit netdevices in the attached-to topology and offload existing routes
and VLAN memberships, for example.

In this patchset, introduce a number of replays to be invoked so that this
sort of post-hoc offload is supported. Then remove the vetoes that
disallowed enslavement of front panel ports to other netdevices with
uppers.

The patchset progresses as follows:

- In patch #1, fix an issue in the bridge driver. To my knowledge, the
  issue could not have resulted in a buggy behavior previously, and thus is
  packaged with this patchset instead of being sent separately to net.

- In patch #2, add a new helper to the switchdev code.

- In patch #3, drop mlxsw selftests that will not be relevant after this
  patchset anymore.

- Patches #4, #5, #6, #7 and #8 prepare the codebase for smoother
  introduction of the rest of the code.

- Patches #9, #10, #11, #12, #13 and #14 replay various aspects of upper
  configuration when a front panel port is introduced into a topology.
  Individual patches take care of bridge and LAG RIF memberships, switchdev
  replay, nexthop and neighbors replay, and MACVLAN offload.

- Patches #15 and #16 introduce RIFs for newly-relevant netdevices when a
  front panel port is enslaved (in which case all uppers are newly
  relevant), or, respectively, deslaved (in which case the newly-relevant
  netdevice is the one being deslaved).

- Up until this point, the introduced scaffolding was not really used,
  because mlxsw still forbids enslavement of mlxsw netdevices to uppers
  with uppers. In patch #17, this condition is finally relaxed.

A sizable selftest suite is available to test all this new code. That will
be sent in a separate patchset.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Enslaving of front panel ports (and their uppers) to netdevices that
already have uppers is currently forbidden. In the previous patches, a
number of replays have been added. Those ensure that various bits of state,
such as next hops or switchdev objects, are offloaded when they become
relevant due to a mlxsw lower being introduced into the topology.

However the act of actually, for example, enslaving a front-panel port to
a bridge with uppers, has been vetoed so far. In this patch, remove the
vetoes and permit the operation.

mlxsw currently validates creation of "interesting" uppers. Thus creating
VLAN netdevices on top of 802.1ad bridges is forbidden if the bridge has an
mlxsw lower, but permitted in general. This validation code never gets run
when a port is introduced as a lower of an existing netdevice structure.

Thus when enslaving an mlxsw netdevice to netdevices with uppers, invoke
the PRECHANGEUPPER event handler for each netdevice above the one that the
front panel port is being enslaved to. This way the tower of netdevices
above the attachment point is validated.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When a netdevice is removed from a bridge or a LAG, and it has an IP
address, it should join the router and gain a RIF. Do that by replaying
address addition event on the netdevice.

When handling deslavement of LAG or its upper from a bridge device, the
replay should be done after all the lowers of the LAG have left the bridge.
Thus these scenarios are handled by passing replay_deslavement of false,
and by invoking, after the lowers have been processed, a new helper,
mlxsw_sp_netdevice_post_lag_event(), which does the per-LAG / -upper
handling, and in particular invokes the replay.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Enslaving of front panel ports (and their uppers) to netdevices that
already have uppers is currently forbidden. When this is permitted, any
uppers with IP addresses need to have the NETDEV_UP inetaddr event
replayed, so that any RIFs are created.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As neighbours are created, mlxsw is involved through the netevent
notifications. When at the time there is no RIF for a given neighbour, the
notification is not acted upon. When the RIF is later created, these
outstanding neighbours are left unoffloaded and cause traffic to go through
the SW datapath.

In order to fix this issue, as a RIF is created, walk the ARP and ND tables
and find neighbours for the netdevice that represents the RIF. Then
schedule neighbour work for them, allowing them to be offloaded.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

If IP address is added to a MACVLAN netdevice, the effect is of configuring
VRRP on the RIF for the netdevice linked to the MACVLAN. Because the
MACVLAN offload is tied to existence of a RIF at the linked netdevice,
adding a MACVLAN is currently not allowed until a RIF is present.

If this requirement stays, it will never be possible to attach a first port
into a topology that involves a MACVLAN. Thus topologies would need to be
built in a certain order, which is impractical.

Additionally, IP address removal, which leads to disappearance of the RIF
that the MACVLAN depends on, cannot be vetoed. Thus even as things stand
now it is possible to get to a state where a MACVLAN netdevice exists
without a RIF, despite having mlxsw lowers. And once the MACVLAN is
un-offloaded due to RIF getting destroyed, recreating the RIF does not
bring it back.

In this patch, accept that MACVLAN can be created out of order and support
that use case.

One option would seem to be to simply recognize MACVLAN netdevices as
"interesting", and let the existing replay mechanisms take care of the
offload. However, that does not address the necessity to reoffload MACVLAN
once a RIF is created.

Thus add a new replay hook, symmetrical to mlxsw_sp_rif_macvlan_flush(),
called mlxsw_sp_rif_macvlan_replay(), which instead of unwinding the
existing offloads, applies the configuration as if the netdevice were
created just now.

Additionally, remove all vetoes and warning messages that checked for
presence of a RIF at the linked device.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Danielle Ratson <danieller@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>