Commit graph

1043032 commits

Author SHA1 Message Date
Johannes Berg
a5b983c607 mac80211: mesh: clean up rx_bcn_presp API
We currently pass the entire elements to the rx_bcn_presp()
method, but only need mesh_config. Additionally, we use the
length of the elements to calculate back the entire frame's
length, but that's confusing - just pass the length of the
frame instead.

Link: https://lore.kernel.org/r/20210920154009.a18ed3d2da6c.I1824b773a0fbae4453e1433c184678ca14e8df45@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-09-23 16:26:33 +02:00
Johannes Berg
01f84f0ed3 mac80211: reduce stack usage in debugfs
We put a few large buffers on the stack here, but it's easy to
just allocate them on the heap, so do that.

Link: https://lore.kernel.org/r/20210920154009.1387f44e7382.Ife043c169e6a44edace516fea9f8311a5ca4282a@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-09-23 16:26:14 +02:00
Len Baker
40f231e75a nl80211: prefer struct_size over open coded arithmetic
As noted in the "Deprecated Interfaces, Language Features, Attributes,
and Conventions" documentation [1], size calculations (especially
multiplication) should not be performed in memory allocator (or similar)
function arguments due to the risk of them overflowing. This could lead
to values wrapping around and a smaller allocation being made than the
caller was expecting. Using those allocations could lead to linear
overflows of heap memory and other misbehaviors.

So, use the struct_size() helper to do the arithmetic instead of the
argument "size + count * size" in the kzalloc() functions.

Also, take the opportunity to refactor the memcpy() call to use the
flex_array_size() helper.

[1] https://www.kernel.org/doc/html/latest/process/deprecated.html#open-coded-arithmetic-in-allocator-arguments

Signed-off-by: Len Baker <len.baker@gmx.com>
Link: https://lore.kernel.org/r/20210919114040.41522-1-len.baker@gmx.com
[remove unnecessary variable]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-09-23 13:38:11 +02:00
Lorenzo Bianconi
68ba1131d4 mac80211: check hostapd configuration parsing twt requests
Check twt_responder in ieee80211_process_rx_twt_action routine
in order to take into account the case where twt has been disabled
in hostapd configuration.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Link: https://lore.kernel.org/r/516057fe4ca73ad257e8c2762e25f4b7872957fc.1630051438.git.lorenzo@kernel.org
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-09-23 13:37:09 +02:00
Johannes Berg
3df15f3451 cfg80211: honour V=1 in certificate code generation
When we generate the code for built-in certificates, honour
the V=1 build option to print out the script doing it.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Link: https://lore.kernel.org/r/20210827131112.dc5492458d55.Idefe4ce8f9681a5ad576d3c6e57c7bff142244de@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2021-09-23 13:37:02 +02:00
David S. Miller
428168f995 Merge branch 'mlxsw-trap-adjacency'
Ido Schimmel says:

====================
mlxsw: Alter trap adjacency entry allocation scheme

In commit 0c3cbbf96d ("mlxsw: Add specific trap for packets routed via
invalid nexthops"), mlxsw started allocating a new adjacency entry
during driver initialization, to trap packets routed via invalid
nexthops.

This behavior was later altered in commit 983db6198f ("mlxsw:
spectrum_router: Allocate discard adjacency entry when needed") to only
allocate the entry upon the first route that requires it. The motivation
for the change is explained in the commit message.

The problem with the current behavior is that the entry shows up as a
"leak" in a new BPF resource monitoring tool [1]. This is caused by the
asymmetry of the allocation/free scheme. While the entry is allocated
upon the first route that requires it, it is only freed during
de-initialization of the driver.

Instead, this patchset tracks the number of active nexthop groups and
allocates the adjacency entry upon the creation of the first group. The
entry is freed when the number of active groups reaches zero.

Patch #1 adds the new entry.

Patch #2 converts mlxsw to start using the new entry and removes the old
one.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:35:02 +01:00
Ido Schimmel
e3a3aae74d mlxsw: spectrum_router: Start using new trap adjacency entry
Start using the trap adjacency entry that was added in the previous
patch and remove the existing one which is no longer needed.

Note that the name of the old entry was inaccurate as the entry did not
discard packets, but trapped them.

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:35:02 +01:00
Ido Schimmel
4bdf80bcb7 mlxsw: spectrum_router: Add trap adjacency entry upon first nexthop group
In commit 0c3cbbf96d ("mlxsw: Add specific trap for packets routed via
invalid nexthops"), mlxsw started allocating a new adjacency entry
during driver initialization, to trap packets routed via invalid
nexthops.

This behavior was later altered in commit 983db6198f ("mlxsw:
spectrum_router: Allocate discard adjacency entry when needed") to only
allocate the entry upon the first route that requires it. The motivation
for the change is explained in the commit message.

The problem with the current behavior is that the entry shows up as a
"leak" in a new BPF resource monitoring tool [1]. This is caused by the
asymmetry of the allocation/free scheme. While the entry is allocated
upon the first route that requires it, it is only freed during
de-initialization of the driver.

Instead, track the number of active nexthop groups and allocate the
adjacency entry upon the creation of the first group. Free it when the
number of active groups reaches zero.

The next patch will convert mlxsw to start using the new entry and
remove the old one.

[1] https://github.com/Mellanox/mlxsw/tree/master/Debugging/libbpf-tools/resmon

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:35:01 +01:00
M Chetan Kumar
8bea96efa7 net: wwan: iosm: fw flashing and cd improvements
1> Function comments moved to .c file.
2> Use literals in return to improve readability.
3> Do error handling check instead of success check.
4> Redundant ret assignment removed.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:23:33 +01:00
Li RongQing
a5df6333f1 skbuff: pass the result of data ksize to __build_skb_around
Avoid to call ksize again in __build_skb_around by passing
the result of data ksize to __build_skb_around

nginx stress test shows this change can reduce ksize cpu usage,
and give a little performance boost

Signed-off-by: Li RongQing <lirongqing@baidu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:20:01 +01:00
Leon Romanovsky
db4278c55f devlink: Make devlink_register to be void
devlink_register() can't fail and always returns success, but all drivers
are obligated to check returned status anyway. This adds a lot of boilerplate
code to handle impossible flow.

Make devlink_register() void and simplify the drivers that use that
API call.

Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Acked-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Vladimir Oltean <olteanv@gmail.com> # dsa
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-22 14:15:12 +01:00
Mianhan Liu
07b855628c net/ipv4/sysctl_net_ipv4.c: remove superfluous header files from sysctl_net_ipv4.c
sysctl_net_ipv4.c hasn't use any macro or function declared in igmp.h,
inetdevice.h, mm.h, module.h, nsproxy.h, swap.h, inet_frag.h, route.h
and snmp.h. Thus, these files can be removed from sysctl_net_ipv4.c
safely without affecting the compilation of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 11:02:28 +01:00
David S. Miller
6a38075363 Merge branch 'iddq-sr-mode'
Florian Fainelli says:

====================
net: phy: broadcom: IDDQ-SR mode

This patch series adds support for the IDDQ with soft recovery mode
which allows power savings of roughly 150mW compared to a simple
BMCR.PDOWN power off (called standby power down in Broadcom datasheets).

In order to leverage these modes we add a new PHY driver flags for
drivers to opt-in for that behavior, the PHY driver is modified to do
the appropriate programming and the PHYs on which this was tested get
updated to have an appropriate suspend/resume set of functions.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:36 +01:00
Florian Fainelli
4972ce7201 net: dsa: bcm_sf2: Request APD, DLL disable and IDDQ-SR
When interfacing with a Broadcom PHY, request the auto-power down, DLL
disable and IDDQ-SR modes to be enabled.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:35 +01:00
Florian Fainelli
c3a4c69360 net: bcmgenet: Request APD, DLL disable and IDDQ-SR
When interfacing with a Broadcom PHY, request the auto-power down, DLL
disable and IDDQ-SR modes to be enabled.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:35 +01:00
Florian Fainelli
72e78d22e1 net: phy: broadcom: Utilize appropriate suspend for BCM54810/11
Since we enable APD and DLL/RXC/TXC disable we need to use
bcm54xx_suspend() in order not to do a read/modify/write of the BMCR
register which is incompatible with the desired settings.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:35 +01:00
Florian Fainelli
38b6a90730 net: phy: broadcom: Wire suspend/resume for BCM50610 and BCM50610M
These two Ethernet PHYs support IDDQ-SR therefore wire-up the suspend
and resume callbacks to point to bcm54xx_suspend() and bcm54xx_resume().

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:35 +01:00
Florian Fainelli
d6da08ed14 net: phy: broadcom: Add IDDQ-SR mode
Add support for putting the PHY into IDDQ Soft Recovery mode by setting
the TOP_MISC register bits accordingly. This requires us to implement a
custom bcm54xx_suspend() routine which diverges from genphy_suspend() in
order to configure the PHY to enter IDDQ with software recovery as well
as avoid doing a read/modify/write on the BMCR register.

Doing a read/modify/write on the BMCR register means that the
auto-negotation bit may remain which interferes with the ability to put
the PHY into IDDQ-SR mode. We do software reset upon suspend in order to
put the PHY back into its state prior to suspend as recommended by the
datasheet.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:58:35 +01:00
Mianhan Liu
c595b120eb net/ipv4/syncookies.c: remove superfluous header files from syncookies.c
syncookies.c hasn't use any macro or function declared in slab.h and random.h,
Thus, these files can be removed from syncookies.c safely without
affecting the compilation of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:48:47 +01:00
Mianhan Liu
bea714581a net/ipv4/udp_tunnel_core.c: remove superfluous header files from udp_tunnel_core.c
udp_tunnel_core.c hasn't use any macro or function declared in udp.h, types.h,
and net_namespace.h. Thus, these files can be removed from udp_tunnel_core.c
safely without affecting the compilation of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-21 10:17:20 +01:00
Mianhan Liu
85c698863c net/ipv4/tcp_minisocks.c: remove superfluous header files from tcp_minisocks.c
tcp_minisocks.c hasn't use any macro or function declared in mm.h, module.h,
slab.h, sysctl.h, workqueue.h, static_key.h and inet_common.h. Thus, these
files can be removed from tcp_minisocks.c safely without affecting the
compilation of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 13:09:06 +01:00
Mianhan Liu
222a31408a net/ipv4/tcp_fastopen.c: remove superfluous header files from tcp_fastopen.c
tcp_fastopen.c hasn't use any macro or function declared in crypto.h, err.h,
init.h, list.h, rculist.h and inetpeer.h. Thus, these files can be removed
from tcp_fastopen.c safely without affecting the compilation of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 13:09:06 +01:00
Mianhan Liu
ffa66f15e4 net/ipv4/route.c: remove superfluous header files from route.c
route.c hasn't use any macro or function declared in uaccess.h, types.h,
string.h, sockios.h, times.h, protocol.h, arp.h and l3mdev.h. Thus, these
files can be removed from route.c safely without affecting the compilation
of the net module.

Signed-off-by: Mianhan Liu <liumh1@shanghaitech.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 13:09:06 +01:00
Aleksander Jan Bajkowski
998ac35801 net: lantiq: add support for jumbo frames
Add support for jumbo frames. Full support for jumbo frames requires
changes in the DSA switch driver (lantiq_gswip.c).

Tested on BT Hone Hub 5A.

Signed-off-by: Aleksander Jan Bajkowski <olek2@wp.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:07:52 +01:00
David S. Miller
13f356f5dc Merge branch 'wwan-iosm-fw-flashing'
M Chetan Kumar says:

====================
net: wwan: iosm: fw flashing & cd collection

This patch series brings-in support for M.2 7560 Device firmware flashing &
coredump collection using devlink.
- Driver Registers with Devlink framework.
- Register devlink params callback for configuring device params
  required in flashing or coredump flow.
- Implements devlink ops flash_update callback that programs modem
  firmware.
- Creates region & snapshot required for device coredump log collection.

On early detection of device in boot rom stage. Driver registers with
Devlink framework and establish transport channel for PSI (Primary Signed
Image) injection. Once PSI is injected to device, the device execution
stage details are read to determine whether device is in flash or
exception mode. The collected information is reported to devlink user
space application & based on this informationi, application proceeds with
either modem firmware flashing or coredump collection.

Refer to iosm devlink documentation for details on Devlink Params, flash
update and coredump collection command usage.

Note: Patches are interdependent. Need to apply complete patch series for
compilation.

Changes since v1:
  * Break down single patch into multiple patches.
  * IOSM Driver devlink documentation.
  * Fixes NULL parameter deference in ipc_devlink_flash_update() reported
    by smatch static checker.
  * Fixes memory leak in ipc_devlink_create_region().
  * Use kmemdup instead of kzalloc and memcpy in ipc_flash_boot_psi().
  * Fixes linux-net build error.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:04:01 +01:00
M Chetan Kumar
607d574aba net: wwan: iosm: fw flashing & cd collection infrastructure changes
IOSM Makefile & WWAN Kconfig changes to support fw flashing & cd
collection module compliation.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
M Chetan Kumar
64302024bc net: wwan: iosm: devlink fw flashing & cd collection documentation
Documents devlink params, fw update & cd collection commands
and its usage.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
M Chetan Kumar
8d9be06341 net: wwan: iosm: transport layer support for fw flashing/cd
Implements transport layer protocol for fw flashing/coredump
collection.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
M Chetan Kumar
09e7b002ff net: wwan: iosm: coredump collection support
Implements protocol for coredump collection.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
M Chetan Kumar
b557347455 net: wwan: iosm: fw flashing support
Implements protocol for fw flashing and PSI injection for
coredump collection.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
M Chetan Kumar
4dcd183fbd net: wwan: iosm: devlink registration
Register with devlink framework and implment callbacks required
for fw flashing and coredump collection.

Signed-off-by: M Chetan Kumar <m.chetan.kumar@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 10:03:37 +01:00
Ansuel Smith
d44fd8604a net: phy: at803x: fix spacing and improve name for 83xx phy
Fix spacing and improve name for 83xx phy following other phy in the
same driver.

Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 09:51:21 +01:00
Ansuel Smith
15b9df4ece net: phy: at803x: add resume/suspend function to qca83xx phy
Add resume/suspend function to qca83xx internal phy.
We can't use the at803x generic function as the documentation lacks of
any support for WoL regs.

Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 09:51:21 +01:00
Ansuel Smith
b4df02b562 net: phy: at803x: add support for qca 8327 A variant internal phy
For qca8327 internal phy there are 2 different switch variant with 2
different phy id. Add this missing variant so the internal phy can be
correctly identified and fixed.

Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 09:51:21 +01:00
Tony Lu
a520794b06 virtio_net: introduce TX timeout watchdog
This implements ndo_tx_timeout handler and put this into stats. When
there is something wrong to send out packets, we could notice tx timeout
events and total timeout counter.

We have suffered send timeout issues due to the backends hung. With this,
we can find the details, and collect the counters by monitor systems.

Signed-off-by: Tony Lu <tony.ly@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 09:42:51 +01:00
Hariprasad Kelam
14e94f9445 octeontx2-af: verify CQ context updates
As per HW errata AQ modification to CQ could be discarded on heavy
traffic. This patch implements workaround for the same after each
CQ write by AQ check whether the requested fields (except those
which HW can update eg: avg_level) are properly updated or not.

If CQ context is not updated then perform AQ write again.

Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 14:05:50 +01:00
Jakub Kicinski
f7116fb460 net: sched: move and reuse mq_change_real_num_tx()
The code for handling active queue changes is identical
between mq and mqprio, reuse it.

Suggested-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 13:26:01 +01:00
Russell King (Oracle)
cbcca2e396 net: phylink: don't call netif_carrier_off() with NULL netdev
Dan Carpenter points out that we have a code path that permits a NULL
netdev pointer to be passed to netif_carrier_off(), which will cause
a kernel oops. In any case, we need to set pl->old_link_state to false
to have the desired effect when there is no netdev present.

Fixes: f97493657c ("net: phylink: add suspend/resume support")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 13:23:47 +01:00
Yajun Deng
4fc2998983 net: rtnetlink: convert rcu_assign_pointer to RCU_INIT_POINTER
It no need barrier when assigning a NULL value to an RCU protected
pointer. So use RCU_INIT_POINTER() instead for more fast.

Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 12:56:02 +01:00
Xuan Zhuo
9ce4e3d6d8 virtio_net: use netdev_warn_once to output warn when without enough queues
This warning is output when virtnet does not have enough queues, but it
only needs to be printed once to inform the user of this situation. It
is not necessary to print it every time. If the user loads xdp
frequently, this log appears too much.

Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 12:46:17 +01:00
wangzhitong
db9c8e2b1e NET: IPV4: fix error "do not initialise globals to 0"
this patch fixes below Errors reported by checkpatch
    ERROR: do not initialise globals to 0
    +int cipso_v4_rbm_optfmt = 0;

Signed-off-by: wangzhitong <wangzhitong@uniontech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 12:43:56 +01:00
Yajun Deng
aed0826b0c net: net_namespace: Fix undefined member in key_remove_domain()
The key_domain member in struct net only exists if we define CONFIG_KEYS.
So we should add the define when we used key_domain.

Fixes: 9b24261051 ("keys: Network namespace domain tag")
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 12:43:04 +01:00
Russell King
9eb7b5e7cb net: dpaa2-mac: add support for more ethtool 10G link modes
Phylink documentation says:
  Note that the PHY may be able to transform from one connection
  technology to another, so, eg, don't clear 1000BaseX just
  because the MAC is unable to BaseX mode. This is more about
  clearing unsupported speeds and duplex settings. The port modes
  should not be cleared; phylink_set_port_modes() will help with this.

So add the missing 10G modes.

Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Acked-by: Marek Behún <kabel@kernel.org>
Acked-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-19 12:11:40 +01:00
David S. Miller
983e59a27b Merge branch 'mptcp-next'
Mat Martineau says:

====================
mptcp: Add SOL_MPTCP getsockopt support

Here's the first new MPTCP feature for the v5.16 cycle, and I'll defer
to Florian's helpful description of the series implementing some new
MPTCP socket options:

========

This adds the MPTCP_INFO, MPTCP_TCPINFO and MPTCP_SUBFLOW_ADDRS
mptcp getsockopt optnames.

MPTCP_INFO exposes the mptcp_info struct as an alternative to the
existing netlink diag interface.

MPTCP_TCPINFO exposes the tcp_info struct.
Unlike SOL_TCP/TCP_INFO, this returns one struct for each active
subflow.

MPTCP_SUBFLOW_ADDRS allows userspace to discover the ip addresses/ports
used by the local and remote endpoints, one for each active tcp subflow.

MPTCP_TCPINFO and MPTCP_SUBFLOW_ADDRS share the same meta-header that
needs to be pre-filled by userspace with the size of the data structures
it expects.  This is done to allow extension of the involved structs
later on, without breaking backwards compatibility.

The meta-structure can also be used to discover the required space
to obtain all information, as kernel will fill in the number of
active subflows even if there is not enough room for the requested info
itself.

More information is available in the individual patches.
Last patch adds test cases for the three optnames.

========
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:01 +01:00
Florian Westphal
ce9979129a selftests: mptcp: add mptcp getsockopt test cases
Add a test program that retrieves the three info types:
1. mptcp meta information
2. tcp info for subflow
3. subflow endpoint addresses

For all three rudimentary checks are added.

1. Meta information checks that the logical mptcp
   sequence numbers advance as expected, based on the bytes read
   (init seq + bytes_received/sent) and the connection state
   (after close, we should exect 1 extra byte due to FIN).

2. TCP info checks the number of bytes sent/received vs.
   sums of read/write syscall return values.

3. Subflow endpoint addresses are checked vs. getsockname/getpeername
   result.

Tests for forward compatibility (0-initialisation of output-only
fields in mptcp_subflow_data structure) are added as well.

Co-developed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:01 +01:00
Florian Westphal
c11c5906bc mptcp: add MPTCP_SUBFLOW_ADDRS getsockopt support
This retrieves the address pairs of all subflows currently
active for a given mptcp connection.

It re-uses the same meta-header as for MPTCP_TCPINFO.

A new structure is provided to hold the subflow
address data:

struct mptcp_subflow_addrs {
	union {
		__kernel_sa_family_t sa_family;
		struct sockaddr sa_local;
		struct sockaddr_in sin_local;
		struct sockaddr_in6 sin6_local;
		struct sockaddr_storage ss_local;
	};
	union {
		struct sockaddr sa_remote;
		struct sockaddr_in sin_remote;
		struct sockaddr_in6 sin6_remote;
		struct sockaddr_storage ss_remote;
	};
};

Usage of the new getsockopt is very similar to
MPTCP_TCPINFO one.

Userspace allocates a
'struct mptcp_subflow_data', followed by one or
more 'struct mptcp_subflow_addrs', then inits the
mptcp_subflow_data structure as follows:

struct mptcp_subflow_addrs *sf_addr;
struct mptcp_subflow_data *addr;
socklen_t olen = sizeof(*addr) + (8 * sizeof(*sf_addr));

addr = malloc(olen);
addr->size_subflow_data = sizeof(*addr);
addr->num_subflows = 0;
addr->size_kernel = 0;
addr->size_user = sizeof(struct mptcp_subflow_addrs);

sf_addr = (struct mptcp_subflow_addrs *)(addr + 1);

and then retrieves the endpoint addresses via:
ret = getsockopt(fd, SOL_MPTCP, MPTCP_SUBFLOW_ADDRS,
		 addr, &olen);

If the call succeeds, kernel will have added up to 8
endpoint addresses after the 'mptcp_subflow_data' header.

Userspace needs to re-check 'olen' value to detect how
many bytes have been filled in by the kernel.

Userspace can check addr->num_subflows to discover when
there were more subflows that available data space.

Co-developed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:01 +01:00
Florian Westphal
06f15cee36 mptcp: add MPTCP_TCPINFO getsockopt support
Allow users to retrieve TCP_INFO data of all subflows.

Users need to pre-initialize a meta header that has to be
prepended to the data buffer that will be filled with the tcp info data.

The meta header looks like this:

struct mptcp_subflow_data {
 __u32 size_subflow_data;/* size of this structure in userspace */
 __u32 num_subflows;	/* must be 0, set by kernel */
 __u32 size_kernel;	/* must be 0, set by kernel */
 __u32 size_user;	/* size of one element in data[] */
} __attribute__((aligned(8)));

size_subflow_data has to be set to 'sizeof(struct mptcp_subflow_data)'.
This allows to extend mptcp_subflow_data structure later on without
breaking backwards compatibility.

If the structure is extended later on, kernel knows where the
userspace-provided meta header ends, even if userspace uses an older
(smaller) version of the structure.

num_subflows must be set to 0. If the getsockopt request succeeds (return
value is 0), it will be updated to contain the number of active subflows
for the given logical connection.

size_kernel must be set to 0. If the getsockopt request is successful,
it will contain the size of the 'struct tcp_info' as known by the kernel.
This is informational only.

size_user must be set to 'sizeof(struct tcp_info)'.

This allows the kernel to only fill in the space reserved/expected by
userspace.

Example:

struct my_tcp_info {
  struct mptcp_subflow_data d;
  struct tcp_info ti[2];
};
struct my_tcp_info ti;
socklen_t olen;

memset(&ti, 0, sizeof(ti));

ti.d.size_subflow_data = sizeof(struct mptcp_subflow_data);
ti.d.size_user = sizeof(struct tcp_info);
olen = sizeof(ti);

ret = getsockopt(fd, SOL_MPTCP, MPTCP_TCPINFO, &ti, &olen);
if (ret < 0)
	die_perror("getsockopt MPTCP_TCPINFO");

mptcp_subflow_data.num_subflows is populated with the number of
subflows that exist on the kernel side for the logical mptcp connection.

This allows userspace to re-try with a larger tcp_info array if the number
of subflows was larger than the available space in the ti[] array.

olen has to be set to the number of bytes that userspace has allocated to
receive the kernel data.  It will be updated to contain the real number
bytes that have been copied to by the kernel.

In the above example, if the number if subflows was 1, olen is equal to
'sizeof(struct mptcp_subflow_data) + sizeof(struct tcp_info).
For 2 or more subflows olen is equal to 'sizeof(struct my_tcp_info)'.

If there was more data that could not be copied due to lack of space
in the option buffer, userspace can detect this by checking
mptcp_subflow_data->num_subflows.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:01 +01:00
Florian Westphal
55c42fa7fa mptcp: add MPTCP_INFO getsockopt
Its not compatible with multipath-tcp.org kernel one.

1. The out-of-tree implementation defines a different 'struct mptcp_info',
   with embedded __user addresses for additional data such as
   endpoint addresses.

2. Mat Martineau points out that embedded __user addresses doesn't work
with BPF_CGROUP_RUN_PROG_GETSOCKOPT() which assumes that copying in
optsize bytes from optval provides all data that got copied to userspace.

This provides mptcp_info data for the given mptcp socket.

Userspace sets optlen to the size of the structure it expects.
The kernel updates it to contain the number of bytes that it copied.

This allows to append more information to the structure later.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:01 +01:00
Florian Westphal
61bc6e82f9 mptcp: add new mptcp_fill_diag helper
Will be re-used from getsockopt path.
Since diag can be a module, we can't export the helper from diag, it
needs to be moved to core.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:20:00 +01:00
David S. Miller
95dca2d578 Merge branch 'macb-MII-on-RGMII'
Claudiu Beznea says:

====================
net: macb: add support for MII on RGMII interface

This series adds support for MII mode on RGMII interface (patches 3/4,
4/4). Along with this the series also contains minor cleanups (patches 1/3,
2/3) on macb.h.

Changes in v2:
- added patch 4/4 to enable MII on RGMII support for SAMA7G5 MAC IPs
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-18 14:14:39 +01:00