Commit graph

1,412,004 commits

Author SHA1 Message Date
Po-Hao Huang
18dab90f56 wifi: rtw89: debug: add FW log component for MLO
This allows showing MLO related logs when FW debug mode is on.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-10-pkshih@realtek.com
2025-05-10 08:58:08 +08:00
Po-Hao Huang
0c400c0a68 wifi: rtw89: debug: add MLD table dump
Add definition for MLD table dump, this is for debug purpose only.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-9-pkshih@realtek.com
2025-05-10 08:57:01 +08:00
Po-Hao Huang
23a5c37ffb wifi: rtw89: debug: extend dbgfs for MLO
Extend dbgfs vif/sta info to show current designated link.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-8-pkshih@realtek.com
2025-05-10 08:56:49 +08:00
Po-Hao Huang
e264a4d1c7 wifi: rtw89: add MLO track for MLSR switch decision
Add MLSR switch mechanism based on tracking RSSI. Switch to a 2 GHz link
(if any) when average RSSI is lower than threshold -53. And, switch out
from a 2 GHz link when average RSSI is larger than threshold -38.

The sequence of MLSR switch handling is like the following.
1. initialize target link and configure to PS mode
2. configure current designated link to PS mode
3. configure target link to non-PS mode
4. deinitialize currently active links except target link

The above flow also implies that target link becomes new designated link.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-7-pkshih@realtek.com
2025-05-10 08:55:27 +08:00
Zong-Zhe Yang
c3dba0653b wifi: rtw89: add handling of mlo_link_cfg H2C command and C2H event
The mlo_link_cfg H2C command is used to tell FW to enter/leave PS mode
on a given link. And, need to wait for status of C2H event to ensure if
FW deals with it successfully.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-6-pkshih@realtek.com
2025-05-10 08:54:05 +08:00
Zong-Zhe Yang
5b6247de57 wifi: rtw89: chan: re-calculate MLO DBCC mode during setting channel
Wi-Fi 7 chips have dual HW bands. After impending MLO support, they
can work with HW-[0] / HW-[1] / HW-[0,1] according to active links.
So, during setting channel, also re-calculate the MLO DBCC mode flag.
Then, leaf chip functions of setting channel can configure with HWs
based on current case.

Besides, tweak the initial and idle MLO DBCC mode of Wi-Fi 7 chips to
MLO_1_PLUS_1_1RF to work with dual HW bands. And, after disconnecting,
due to no active links, MLO DBCC mode will re-calculate to idle case,
i.e. MLO_1_PLUS_1_1RF.

Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-5-pkshih@realtek.com
2025-05-10 08:53:51 +08:00
Po-Hao Huang
a8ba4acab7 wifi: rtw89: send nullfunc based on the given link
The nullfunc sender function is link specific. Use core_tx_write_link
with sw_mld flag to TX the nullfunc via the given link.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-4-pkshih@realtek.com
2025-05-10 08:53:40 +08:00
Po-Hao Huang
829bd3599a wifi: rtw89: allow driver to do specific band TX for MLO
For data packets that can be sent on any band, fill in main mac_id
and let HW decide. For packets that we wish to transmit on specific
band, fill in sw_mld field so HW would only send it on that band.
Main mac_id is the corresponding mac_id on band 0.

Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-3-pkshih@realtek.com
2025-05-10 08:52:14 +08:00
Zong-Zhe Yang
0f34fbd274 wifi: rtw89: extract link part from core tx write function
Extract core_tx_write_link from core_tx_write to accept given links as
parameters. To make the follow-up changes of TX description more clear,
tweak SW functions to split things ahead.

(don't change logic at all)

Signed-off-by: Zong-Zhe Yang <kevin_yang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250505072440.45113-2-pkshih@realtek.com
2025-05-10 08:52:02 +08:00
Yuuki NAGAO
b7f0cc647e wifi: rtw88: rtw8822bu VID/PID for BUFFALO WI-U2-866DM
Add VID/PID 0411/03d1 for recently released
BUFFALO WI-U2-866DM USB WiFi adapter.

Signed-off-by: Yuuki NAGAO <wf.yn386@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/20250503003227.6673-1-wf.yn386@gmail.com
2025-05-10 08:49:04 +08:00
Bitterblue Smith
2c17afde9f wifi: rtw88: Handle RTL8723D(S) with blank efuse
Some users have RTL8723DS chips with nearly blank efuse. Currently these
chips cannot connect when using rtw88, but they do work using the old
out-of-tree driver.

Use reasonable default values for TX power, antenna configuration, and
crystal cap if the chip's efuse is missing these things.

RTL8723D can use the same default values as RTL8703B, so simply move
the code from rtl8703b_read_efuse() to the shared function
__rtl8723x_read_efuse().

Link: https://github.com/lwfinger/rtw88/issues/157
Signed-off-by: Bitterblue Smith <rtl8821cerfe2@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/5734afe7-0870-40b2-acd4-5657a02d7c56@gmail.com
2025-05-10 08:45:13 +08:00
Bitterblue Smith
0ffa1ba81b wifi: rtw88: Fix RX aggregation settings for RTL8723DS
Use the same RX aggregation size and timeout used by the out-of-tree
RTL8723DS driver. Also set mystery bit 31 of REG_RXDMA_AGG_PG_TH. This
improves the RX speed from ~44 Mbps to ~67 Mbps.

Signed-off-by: Bitterblue Smith <rtl8821cerfe2@gmail.com>
Acked-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Link: https://patch.msgid.link/4c79fdc1-54bc-4986-9931-bb3ceb418b97@gmail.com
2025-05-10 08:43:49 +08:00
Jakub Kicinski
0b28182c73 Merge branch 'refactoring-designware-vlan-code'
Boon Khai Ng says:

====================
Refactoring designware VLAN code.

Refactoring designware VLAN code and introducing support for
hardware-accelerated VLAN stripping for dwxgmac2 IP,
the current patch set consists of two key changes:

1) Refactoring VLAN Functions:
The first change involves moving common VLAN-related functions
of the DesignWare Ethernet MAC into a dedicated file, stmmac_vlan.c.
This refactoring aims to improve code organization and maintainability
by centralizing VLAN handling logic.

2) Renaming all the functions, symbols and macro into more
generic name and consolidate the same function pointer into one.

3) Enabling VLAN for 10G Ethernet MAC IP:
The second change enables VLAN support specifically
for the 10G Ethernet MAC IP. This enhancement leverages
the hardware capabilities of the to perform VLAN stripping,

Changes from previous submmited patches.
v5:
a) Divided the refactor patch in to two patches,
first patch is to move the code into the separate file
and second patch to update the symbol name

b) get the dwmac4 vlan function up to date and port to
stmmac_vlan.c

c) remove the inline function in function pointer and
use only static function defeination.

d) Remove the outer parenthese that is not needed
on the 1 line return statement.

v4:
a) Updated the commit message to explain the descriptors
behaviour on different hardware.

b) Updated the perfect_match variable with the correct
byte order.

Link: https://lore.kernel.org/20250421162930.10237-1-boon.khai.ng@altera.com

v3: https://lore.kernel.org/20250408081354.25881-1-boon.khai.ng@altera.com
v2: https://lore.kernel.org/BL3PR11MB5748AC693D9D61FB56DB7313C1F32@BL3PR11MB5748.namprd11.prod.outlook.com
v1: https://lore.kernel.org/DM8PR11MB5751E5388AEFCFB80BCB483FC13FA@DM8PR11MB5751.namprd11.prod.outlook.com
====================

Link: https://patch.msgid.link/20250507063812.34000-1-boon.khai.ng@altera.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:29:45 -07:00
Boon Khai Ng
534df0c172 net: stmmac: dwxgmac2: Add support for HW-accelerated VLAN stripping
This patch adds support for MAC level VLAN tag stripping for the
dwxgmac2 IP. This feature can be configured through ethtool.
If the rx-vlan-offload is off, then the VLAN tag will be stripped
by the driver. If the rx-vlan-offload is on then the VLAN tag
will be stripped by the MAC.

Signed-off-by: Boon Khai Ng <boon.khai.ng@altera.com>
Reviewed-by: Matthew Gerlach <matthew.gerlach@altera.com>
Link: https://patch.msgid.link/20250507063812.34000-4-boon.khai.ng@altera.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:29:43 -07:00
Boon Khai Ng
f3acaf7364 net: stmmac: stmmac_vlan: rename VLAN functions and symbol to generic symbol.
With the VLAN handling code decoupled from dwmac4 and dwxgmac2 and
consolidated in stmmac_vlan.c, functions and symbols are renamed to
use a generic prefix. This change improves code clarity and
maintainability by reflecting the shared nature of the VLAN logic,
facilitating future enhancements or reuse without being tied to
specific MAC implementations.

No functional changes are introduced in this patch.

Note: The dwxgmac2_update_vlan_hash function is not combined due
to minor differences in setting the VTFE bit. A separate fix patch
will be submitted to align its behavior with the dwmac4 driver.

Signed-off-by: Boon Khai Ng <boon.khai.ng@altera.com>
Link: https://patch.msgid.link/20250507063812.34000-3-boon.khai.ng@altera.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:29:43 -07:00
Boon Khai Ng
1d2c7a5fee net: stmmac: Refactor VLAN implementation
Refactor VLAN implementation by moving common code for DWMAC4 and
DWXGMAC IPs into a separate VLAN module. VLAN implementation for
DWMAC4 and DWXGMAC differs only for CSR base address, the descriptor
for the VLAN ID and VLAN VALID bit field.

The descriptor format for VLAN is not moved to the common code due
to hardware-specific differences between DWMAC4 and DWXGMAC.

For the DWMAC4 IP, the Receive Normal Descriptor 0 (RDES0) is
formatted as follows:
    31                                                0
      ------------------------ -----------------------
RDES0| Inner VLAN TAG [31:16] | Outer VLAN TAG [15:0] |
      ------------------------ -----------------------

For the DWXGMAC IP, the RDES0 format varies based on the
Tunneled Frame bit (TNP):

a) For Non-Tunneled Frame (TNP=0)

    31                                                0
      ------------------------ -----------------------
RDES0| Inner VLAN TAG [31:16] | Outer VLAN TAG [15:0] |
      ------------------------ -----------------------

b) For Tunneled Frame (TNP=1)

     31                   8 7                3 2      0
      --------------------- ------------------ -------
RDES0| VNID/VSID           | Reserved         | OL2L3 |
      --------------------- ------------------ ------

The logic for handling tunneled frames is not yet implemented
in the dwxgmac2_wrback_get_rx_vlan_tci() function. Therefore,
it is prudent to maintain separate functions within their
respective descriptor driver files
(dwxgmac2_descs.c and dwmac4_descs.c).

Signed-off-by: Boon Khai Ng <boon.khai.ng@altera.com>
Link: https://patch.msgid.link/20250507063812.34000-2-boon.khai.ng@altera.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:29:43 -07:00
Jakub Kicinski
4d64321c4f Here is a batman-adv bugfix:
- fix duplicate MAC address check, by Matthias Schiffer
 -----BEGIN PGP SIGNATURE-----
 
 iQJKBAABCgA0FiEE1ilQI7G+y+fdhnrfoSvjmEKSnqEFAmgdxEYWHHN3QHNpbW9u
 d3VuZGVybGljaC5kZQAKCRChK+OYQpKeoSG7D/9UpiHwuCb3FCT9hZVNbRX6XChV
 DSv8lYedVflwYgIZCqcKlg4/svygmnNNYy4XOJeLw7kjeegGmdQsol0TouZGkqTB
 pNxKnporVTxwA4FDSaRwVxwF1UOlelAzcMRxAk5tpIkGPl1wBFLx7dFzWWuU7wvT
 c0mtkJlV0kag0OZJEZqY5fYmQNHmYde9rsTpdP44ie7SPqGHpby6MYjqfGl0ztxy
 70XCMgp9ByRsCnL45UQYua17gYavKWalCsAYCikQc/+nZGkgymAaIFv5elFrou9N
 Dbw+k9X2ipw3v2atAUP40Bw6dvdIkTRaEpAyO5oDGM1//kqwAfUQazS+IWFYE50v
 eN2KRGSNJH/J2RHY0ODRomcFDotHykR8aESbkePJHc5ZFgM9E6vAzy+BPrFJaROX
 nRO/XmWXeVwyl9CrxoEtA0UusIEr3OQBMi2p5cV7vbcPHSPzQRuA2iKah0ZFsmB0
 pGyaeEh60lO2r84lYsO04bA9q8eyZkNHUxd3pZnCpy4iAzEmgKaD7RAIygwHtRim
 UezzP8BO7m6dlox16/eJ2Bev2EmA3m3w/rkyFU2jrE9xv8tTkjIYs04zTifs8+F7
 hAU+2WM/P0z+lCVEZzTaguEjFK4UrmnmSZxF1B4mh5AVQL5q4R5np4ncTiBdGJCl
 uS34escHTXWW4gHttQ==
 =kqur
 -----END PGP SIGNATURE-----

Merge tag 'batadv-net-pullrequest-20250509' of git://git.open-mesh.org/linux-merge

Simon Wunderlich says:

====================
Here is a batman-adv bugfix:

 - fix duplicate MAC address check, by Matthias Schiffer

* tag 'batadv-net-pullrequest-20250509' of git://git.open-mesh.org/linux-merge:
  batman-adv: fix duplicate MAC address check
====================

Link: https://patch.msgid.link/20250509090240.107796-1-sw@simonwunderlich.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:09:39 -07:00
Jakub Kicinski
6a63b01567 This cleanup patchset includes the following patches:
- bump version strings, by Simon Wunderlich
 
  - constify and move broadcast addr definition, Matthias Schiffer
 
  - remove start/stop queue function for mesh-iface, by Antonio Quartulli
 
  - switch to crc32 header for crc32c, by Sven Eckelmann
 
  - drop unused net_namespace.h include, by Sven Eckelmann
 -----BEGIN PGP SIGNATURE-----
 
 iQJKBAABCgA0FiEE1ilQI7G+y+fdhnrfoSvjmEKSnqEFAmgdxeMWHHN3QHNpbW9u
 d3VuZGVybGljaC5kZQAKCRChK+OYQpKeoSR2EADDS+gfYJHfGK+9R2N+GFVf+/nu
 avyo9HPi2ZdQEgzeuSDZFQb96yzQNOcW3iTkM+nSAJMANHV7I9n28ZKIufqtXNaR
 E3dKpi592ZjwedaEYp1JFHp8L4bGxEWQ68ie8zqdlyT4QXryUyVVZG061jKF2Pjl
 vns3KAL6kzFv8SlweR+JuoYudQTjn/P0m1iuPylpzf2cCD8yUaY6GsI11D4XDgxQ
 IJDkNpmVulBEJ6oYqUfA9fqQffnnxiAw4/IaA4NOrzIGvNGHKfDfnWEMTqNB/YU7
 xZ7/zD6iKefxd8OQkiFnA1NuNw+dVuii/HOeH3GcWMQ9blApeAHCUZ+o9xp6uGP0
 8f6AQ3TU/STe9/whDCgHvjMHED5V2hpYSVvtWn0w3L9nqhxrZq5eY0Cek8k7vt+G
 tRj1tyzrEcpoPQMb/eL8mS179PgCV72KKe8gwSrFdgKzLaxWVcqnByPVUjSzhHoN
 4yOWY+KUhSH9CUTes0/dMwag2EorK1B/cUXm1forHk2xNosw/FvXD1hYXJB1Mcx5
 tZvN9BalJcag4aAhpx2CCDYI0XMZrAXDB2duS/lrgSbLdbvWN3WDGAr2aI3GXpWS
 97ezGLIMWn8l9oE2BJr1s6a4/2sXJ/HSYteUGkRxIQbaitVUsN9rueuQvRzSvUvW
 i8Y5mZZaX5EUW/2gfw==
 =T+6l
 -----END PGP SIGNATURE-----

Merge tag 'batadv-next-pullrequest-20250509' of git://git.open-mesh.org/linux-merge

Simon Wunderlich says:

====================
This cleanup patchset includes the following patches:

 - bump version strings, by Simon Wunderlich

 - constify and move broadcast addr definition, Matthias Schiffer

 - remove start/stop queue function for mesh-iface, by Antonio Quartulli

 - switch to crc32 header for crc32c, by Sven Eckelmann

 - drop unused net_namespace.h include, by Sven Eckelmann

* tag 'batadv-next-pullrequest-20250509' of git://git.open-mesh.org/linux-merge:
  batman-adv: Drop unused net_namespace.h include
  batman-adv: Switch to crc32 header for crc32c
  batman-adv: no need to start/stop queue on mesh-iface
  batman-adv: constify and move broadcast addr definition
  batman-adv: Start new development cycle
====================

Link: https://patch.msgid.link/20250509091041.108416-1-sw@simonwunderlich.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 17:09:38 -07:00
Linus Torvalds
1a33418a69 two smb3 client fixes
-----BEGIN PGP SIGNATURE-----
 
 iQGzBAABCgAdFiEE6fsu8pdIjtWE/DpLiiy9cAdyT1EFAmgebFUACgkQiiy9cAdy
 T1HI+gv/Q1R0t6HlpL3fgMPw9RZIOcndvIulYI1CyNQ/xdhqDcPBGl3qt3vm6247
 OM7o/easyr6TXXY5rdO/zne2ekORRLKBKFeoH/1v4jQu3XBoc/O2S0jaM9rmNuvm
 gDgVU+rzkfIUWLYQz1r8XxQ7umz+rZYEX3E9K96mlQWlykSdepx8wtLDE5kZoZZb
 fo5sHw1iEuWXLualM0fuNbqScdWKQ/Wg//yjopefwsldLY/JcoOfA/1HLXOgI451
 a3AcL7QmZzQG6wa/aq1kUmU10fL9DxbDhNmCPyiGTlCIcyithECrMaR4IRq6/FlU
 E1Zh1v7K+zCH48DSrb9SlVaCoqkZmkztvadTZBqp9NqnjOUlbpX0mHK2I0hFxcMN
 nrYFg5HPKYdo0pbLSobgIGQPph6UIrAjSlKiTfTdpEGjyyStRjtftmhhky5vKHKS
 ZXkQ9XNdS6Cwx1DT9JDYQrUwt2xH44kDx+OYsxYYYaw9yLkClJBDfQZWr4zeMnWi
 0E0mx1cz
 =m9eD
 -----END PGP SIGNATURE-----

Merge tag '6.15-rc5-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6

Pull smb client fixes from Steve French:

 - Fix dentry leak which can cause umount crash

 - Add warning for parse contexts error on compounded operation

* tag '6.15-rc5-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
  smb: client: Avoid race in open_cached_dir with lease breaks
  smb3 client: warn when parse contexts returns error on compounded operation
2025-05-09 16:45:21 -07:00
Paul Cacheux
e41b5af451 tracing: add missing trace_probe_log_clear for eprobes
Make sure trace_probe_log_clear is called in the tracing
eprobe code path, matching the trace_probe_log_init call.

Link: https://lore.kernel.org/all/20250504-fix-trace-probe-log-race-v3-1-9e99fec7eddc@gmail.com/

Signed-off-by: Paul Cacheux <paulcacheux@gmail.com>
Acked-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
2025-05-10 08:44:50 +09:00
Vladimir Oltean
3c9ff6eb2d net: mvpp2: convert to ndo_hwtstamp_get() and ndo_hwtstamp_set()
New timestamping API was introduced in commit 66f7223039 ("net: add
NDOs for configuring hardware timestamping") from kernel v6.6. It is
time to convert the mvpp2 driver to the new API, so that the
ndo_eth_ioctl() path can be removed completely.

Note that on the !port->hwtstamp condition, the old code used to fall
through in mvpp2_ioctl(), and return either -ENOTSUPP if !port->phylink,
or -EOPNOTSUPP, in phylink_mii_ioctl(). Keep the test for port->hwtstamp
in the newly introduced net_device_ops, but consolidate the error code
to just -EOPNOTSUPP. The other one is documented as NFS-specific, it's
best to avoid it anyway.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250508144630.1979215-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:43:41 -07:00
Vladimir Oltean
17c6c5a09d net: gianfar: convert to ndo_hwtstamp_get() and ndo_hwtstamp_set()
New timestamping API was introduced in commit 66f7223039 ("net: add
NDOs for configuring hardware timestamping") from kernel v6.6. It is
time to convert the gianfar driver to the new API, so that the
ndo_eth_ioctl() path can be removed completely.

Don't propagate the unnecessary "config.flags = 0;" assignment to
gfar_hwtstamp_get(), because dev_get_hwtstamp() provides a zero
initialized struct kernel_hwtstamp_config.

After removing timestamping logic from gfar_ioctl(), the rest is
equivalent to phy_do_ioctl_running(), so provide that directly as our
ndo_eth_ioctl() implementation.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> # LS1021A
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250508143659.1944220-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:42:29 -07:00
Vladimir Oltean
d27c6e8975 net: dpaa2-eth: add ndo_hwtstamp_get() implementation
Allow SIOCGHWTSTAMP to retrieve the current timestamping settings on
DPNIs. This reflects what has been done in dpaa2_eth_hwtstamp_set().

Tested with hwstamp_ctl -i endpmac17. Before the change, it prints
"Device driver does not have support for non-destructive SIOCGHWTSTAMP."
After the change, it retrieves the previously set configuration.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> # LX2160A
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250508134102.1747075-2-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:41:51 -07:00
Vladimir Oltean
b6e79c5da8 net: dpaa2-eth: convert to ndo_hwtstamp_set()
New timestamping API was introduced in commit 66f7223039 ("net: add
NDOs for configuring hardware timestamping") from kernel v6.6. It is
time to convert the DPAA2 Ethernet driver to the new API, so that the
ndo_eth_ioctl() path can be removed completely.

This driver only responds to SIOCSHWTSTAMP (not SIOCGHWTSTAMP) so
convert just that.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> # LX2160A
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250508134102.1747075-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:41:51 -07:00
Jakub Kicinski
4b3f6fb8d0 Merge branch 'dpaa_eth-conversion-to-ndo_hwtstamp_get-and-ndo_hwtstamp_set'
Vladimir Oltean says:

====================
dpaa_eth conversion to ndo_hwtstamp_get() and ndo_hwtstamp_set()

This is part of the effort to finalize the conversion of drivers to the
dedicated hardware timestamping API.

In the case of the DPAA1 Ethernet driver, a bit more care is needed,
because dpaa_ioctl() looks a bit strange. It handles the "set" IOCTL but
not the "get", and even the phylink_mii_ioctl() portion could do with
some cleanup.
====================

Link: https://patch.msgid.link/20250508124753.1492742-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:40:24 -07:00
Vladimir Oltean
c2d0b7da61 net: dpaa_eth: simplify dpaa_ioctl()
phylink_mii_ioctl() handles multiple ioctls in addition to just
SIOCGMIIREG: SIOCGMIIPHY, SIOCSMIIREG. Don't filter these out.

Also, phylink can handle the case where net_dev->phydev is NULL (like
optical SFP module, fixed-link). Be like other drivers and let phylink
do so without any driver-side call filtering.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Acked-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Link: https://patch.msgid.link/20250508124753.1492742-4-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:40:22 -07:00
Vladimir Oltean
7bf230556b net: dpaa_eth: add ndo_hwtstamp_get() implementation
Allow SIOCGHWTSTAMP to retrieve the current timestamping settings
on DPAA1 interfaces. This reflects what has been done in
dpaa_hwtstamp_set().

Tested with hwstamp_ctl -i fm1-mac9.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Acked-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Link: https://patch.msgid.link/20250508124753.1492742-3-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:40:22 -07:00
Vladimir Oltean
b45bf3f84e net: dpaa_eth: convert to ndo_hwtstamp_set()
New timestamping API was introduced in commit 66f7223039 ("net: add
NDOs for configuring hardware timestamping") from kernel v6.6. It is
time to convert the DPAA1 driver to the new API, so that the
ndo_eth_ioctl() path can be removed completely.

This driver only responds to SIOCSHWTSTAMP (not SIOCGHWTSTAMP) so
convert just that.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Acked-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Link: https://patch.msgid.link/20250508124753.1492742-2-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:40:22 -07:00
Vladimir Oltean
6c14058edf net: dsa: convert to ndo_hwtstamp_get() and ndo_hwtstamp_set()
New timestamping API was introduced in commit 66f7223039 ("net: add
NDOs for configuring hardware timestamping") from kernel v6.6. It is
time to convert DSA to the new API, so that the ndo_eth_ioctl() path can
be removed completely.

Move the ds->ops->port_hwtstamp_get() and ds->ops->port_hwtstamp_set()
calls from dsa_user_ioctl() to dsa_user_hwtstamp_get() and
dsa_user_hwtstamp_set().

Due to the fact that the underlying ifreq type changes to
kernel_hwtstamp_config, the drivers and the Ocelot switchdev front-end,
all hooked up directly or indirectly, must also be converted all at once.

The conversion also updates the comment from dsa_port_supports_hwtstamp(),
which is no longer true because kernel_hwtstamp_config is kernel memory
and does not need copy_to_user(). I've deliberated whether it is
necessary to also update "err != -EOPNOTSUPP" to a more general "!err",
but all drivers now either return 0 or -EOPNOTSUPP.

The existing logic from the ocelot_ioctl() function, to avoid
configuring timestamping if the PHY supports the operation, is obsoleted
by more advanced core logic in dev_set_hwtstamp_phylib().

This is only a partial preparation for proper PHY timestamping support.
None of these switch driver currently sets up PTP traps for PHY
timestamping, so setting dev->see_all_hwtstamp_requests is not yet
necessary and the conversion is relatively trivial.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com> # felix, sja1105, mv88e6xxx
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://patch.msgid.link/20250508095236.887789-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:34:09 -07:00
Donald Hunter
0df6932485 tools: ynl: handle broken pipe gracefully in CLI
When sending YNL CLI output into a pipe, closing the pipe causes a
BrokenPipeError. E.g. running the following and quitting less:

./tools/net/ynl/pyynl/cli.py --family rt-link --dump getlink | less
Traceback (most recent call last):
  File "/home/donaldh/net-next/./tools/net/ynl/pyynl/cli.py", line 160, in <module>
    main()
    ~~~~^^
  File "/home/donaldh/net-next/./tools/net/ynl/pyynl/cli.py", line 142, in main
    output(reply)
    ~~~~~~^^^^^^^
  File "/home/donaldh/net-next/./tools/net/ynl/pyynl/cli.py", line 97, in output
    pprint.PrettyPrinter().pprint(msg)
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^
[...]
BrokenPipeError: [Errno 32] Broken pipe

Consolidate the try block for ops and notifications, and gracefully
handle the BrokenPipeError by adding an exception handler to the
consolidated try block.

Signed-off-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250508112102.63539-1-donald.hunter@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:28:23 -07:00
Breno Leitao
9dda18a32b tracing: fprobe: Fix RCU warning message in list traversal
When CONFIG_PROVE_RCU_LIST is enabled, fprobe triggers the following
warning:

    WARNING: suspicious RCU usage
    kernel/trace/fprobe.c:457 RCU-list traversed in non-reader section!!

    other info that might help us debug this:
	#1: ffffffff863c4e08 (fprobe_mutex){+.+.}-{4:4}, at: fprobe_module_callback+0x7b/0x8c0

    Call Trace:
	fprobe_module_callback
	notifier_call_chain
	blocking_notifier_call_chain

This warning occurs because fprobe_remove_node_in_module() traverses an
RCU list using RCU primitives without holding an RCU read lock. However,
the function is only called from fprobe_module_callback(), which holds
the fprobe_mutex lock that provides sufficient protection for safely
traversing the list.

Fix the warning by specifying the locking design to the
CONFIG_PROVE_RCU_LIST mechanism. Add the lockdep_is_held() argument to
hlist_for_each_entry_rcu() to inform the RCU checker that fprobe_mutex
provides the required protection.

Link: https://lore.kernel.org/all/20250410-fprobe-v1-1-068ef5f41436@debian.org/

Fixes: a3dc2983ca ("tracing: fprobe: Cleanup fprobe hash when module unloading")
Signed-off-by: Breno Leitao <leitao@debian.org>
Tested-by: Antonio Quartulli <antonio@mandelbit.com>
Tested-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
2025-05-10 08:28:02 +09:00
Gal Pressman
1b2900db01 ethtool: Block setting of symmetric RSS when non-symmetric rx-flow-hash is requested
Symmetric RSS hash requires that:
* No other fields besides IP src/dst and/or L4 src/dst are set
* If src is set, dst must also be set

This restriction was only enforced when RXNFC was configured after
symmetric hash was enabled. In the opposite order of operations (RXNFC
then symmetric enablement) the check was not performed.

Perform the sanity check on set_rxfh as well, by iterating over all flow
types hash fields and making sure they are all symmetric.

Introduce a function that returns whether a flow type is hashable (not
spec only) and needs to be iterated over. To make sure that no one
forgets to update the list of hashable flow types when adding new flow
types, a static assert is added to draw the developer's attention.

The conversion of uapi #defines to enum is not ideal, but as Jakub
mentioned [1], we have precedent for that.

[1] https://lore.kernel.org/netdev/20250324073509.6571ade3@kernel.org/

Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250508103034.885536-1-gal@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:24:28 -07:00
Andrew Jeffery
e4f349bd6e net: mctp: Ensure keys maintain only one ref to corresponding dev
mctp_flow_prepare_output() is called in mctp_route_output(), which
places outbound packets onto a given interface. The packet may represent
a message fragment, in which case we provoke an unbalanced reference
count to the underlying device. This causes trouble if we ever attempt
to remove the interface:

    [   48.702195] usb 1-1: USB disconnect, device number 2
    [   58.883056] unregister_netdevice: waiting for mctpusb0 to become free. Usage count = 2
    [   69.022548] unregister_netdevice: waiting for mctpusb0 to become free. Usage count = 2
    [   79.172568] unregister_netdevice: waiting for mctpusb0 to become free. Usage count = 2
    ...

Predicate the invocation of mctp_dev_set_key() in
mctp_flow_prepare_output() on not already having associated the device
with the key. It's not yet realistic to uphold the property that the key
maintains only one device reference earlier in the transmission sequence
as the route (and therefore the device) may not be known at the time the
key is associated with the socket.

Fixes: 67737c4572 ("mctp: Pass flow data & flow release events to drivers")
Acked-by: Jeremy Kerr <jk@codeconstruct.com.au>
Signed-off-by: Andrew Jeffery <andrew@codeconstruct.com.au>
Link: https://patch.msgid.link/20250508-mctp-dev-refcount-v1-1-d4f965c67bb5@codeconstruct.com.au
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:22:53 -07:00
Jason Xing
179542a987 net: thunder: make tx software timestamp independent
skb_tx_timestamp() is used for tx software timestamp enabled by
SOF_TIMESTAMPING_TX_SOFTWARE while SKBTX_HW_TSTAMP is used for
SOF_TIMESTAMPING_TX_HARDWARE. As it clearly shows they are different
timestamps in two dimensions, it's not appropriate to group these two
together in the if-statement.

This patch completes three things:
1. make the software one standalone. Users are able to set both
timestamps together with SOF_TIMESTAMPING_OPT_TX_SWHW flag.
2. make the software one generated after the hardware timestamp logic to
avoid generating sw and hw timestamps at one time without
SOF_TIMESTAMPING_OPT_TX_SWHW being set.
3. move the software timestamp call as close to the door bell.

Signed-off-by: Jason Xing <kernelxing@tencent.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/20250508034433.14408-1-kerneljasonxing@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 16:21:25 -07:00
Lad Prabhakar
a92b9efaf2 dt-bindings: gpu: mali-bifrost: Add compatible for RZ/V2N SoC
Add a compatible string for the Renesas RZ/V2N SoC variants that include a
Mali-G31 GPU. These variants share the same restrictions on interrupts,
clocks, and power domains as the RZ/G2L SoC, so extend the existing schema
validation accordingly.

Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Acked-by: Conor Dooley <conor.dooley@microchip.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20250502162540.165962-1-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
2025-05-09 17:26:40 -05:00
Dan Carpenter
493e6cb63a of: unittest: Unlock on error in unittest_data_add()
The of_overlay_mutex_unlock() was accidentally deleted if "of_root" is
NULL.  Change this to a goto unlock.

Fixes: d1eabd218e ("of: unittest: treat missing of_root as error instead of fixing up")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Link: https://lore.kernel.org/r/aBHZ1DvXiBcZkWmk@stanley.mountain
Signed-off-by: Rob Herring (Arm) <robh@kernel.org>
2025-05-09 17:26:40 -05:00
Dmitry Baryshkov
3022ae40e9 dt-bindings: soc: qcom: qcom,rpm: add missing clock/-names properties
Qualcomm MSM8660 and IPQ0864 platforms use additional clock for the RPM
device. Document it in the schema.

Fixes: aa0c4b8150 ("mfd: devicetree: bindings: Add Qualcomm RPM DT binding")
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Rob Herring (Arm) <robh@kernel.org>
Link: https://lore.kernel.org/r/20250318-fix-nexus-4-v2-2-bcedd1406790@oss.qualcomm.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Dmitry Baryshkov
dc4ec78031 dt-bindings: soc: qcom,rpm: add missing clock-controller node
On Qualcomm platforms in addition to regulators the RPM also provides
clocks via the child clock-controller node. Describe it properly in the
schema.

Fixes: 872f91b5ea ("clk: qcom: Add support for RPM Clocks")
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250318-fix-nexus-4-v2-1-bcedd1406790@oss.qualcomm.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Unnathi Chalicheemala
acda397ad5 soc: qcom: smem: Update max processor count
Update max processor count to reflect the number of co-processors on
upcoming SoC.

Signed-off-by: Unnathi Chalicheemala <unnathi.chalicheemala@oss.qualcomm.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250424-smem_count-v1-1-931cf68f71a8@oss.qualcomm.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Danila Tikhonov
db3de3ff26 firmware: qcom: tzmem: disable sm7150 platform
The SHM bridge makes the SM7150 devices reset while probing the RMTFS
(in qcom_scm_assign_mem()). Blacklist the SHM Bridge on corresponding
platforms using SoC-level compat string.

Signed-off-by: Danila Tikhonov <danila@jiaxyga.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250422213137.80366-13-danila@jiaxyga.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Jens Reidel
7dc837b556 soc: qcom: pd-mapper: Add support for SM7150
SM7150 protection domains are the same as SC7180, with the subtle
difference that SM7150 has a CDSP.

Signed-off-by: Jens Reidel <adrian@mainlining.org>
Signed-off-by: Danila Tikhonov <danila@jiaxyga.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20250422213137.80366-11-danila@jiaxyga.com
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Johan Hovold
5090ac9191 soc: qcom: pmic_glink_altmode: fix spurious DP hotplug events
The PMIC GLINK driver is currently generating DisplayPort hotplug
notifications whenever something is connected to (or disconnected from)
a port regardless of the type of notification sent by the firmware.

These notifications are forwarded to user space by the DRM subsystem as
connector "change" uevents:

    KERNEL[1556.223776] change   /devices/platform/soc@0/ae00000.display-subsystem/ae01000.display-controller/drm/card0 (drm)
    ACTION=change
    DEVPATH=/devices/platform/soc@0/ae00000.display-subsystem/ae01000.display-controller/drm/card0
    SUBSYSTEM=drm
    HOTPLUG=1
    CONNECTOR=36
    DEVNAME=/dev/dri/card0
    DEVTYPE=drm_minor
    SEQNUM=4176
    MAJOR=226
    MINOR=0

On the Lenovo ThinkPad X13s and T14s, the PMIC GLINK firmware sends two
identical notifications with orientation information when connecting a
charger, each generating a bogus DRM hotplug event. On the X13s, two
such notification are also sent every 90 seconds while a charger remains
connected, which again are forwarded to user space:

    port = 1, svid = ff00, mode = 255, hpd_state = 0
    payload = 01 00 00 00 00 00 00 ff 00 00 00 00 00 00 00 00

Note that the firmware only sends on of these when connecting an
ethernet adapter.

Fix the spurious hotplug events by only forwarding hotplug notifications
for the Type-C DisplayPort service id. This also reduces the number of
uevents from four to two when an actual DisplayPort altmode device is
connected:

    port = 0, svid = ff01, mode = 2, hpd_state = 0
    payload = 00 01 02 00 f2 0c 01 ff 03 00 00 00 00 00 00 00
    port = 0, svid = ff01, mode = 2, hpd_state = 1
    payload = 00 01 02 00 f2 0c 01 ff 43 00 00 00 00 00 00 00

Fixes: 080b4e2485 ("soc: qcom: pmic_glink: Introduce altmode support")
Cc: stable@vger.kernel.org	# 6.3
Cc: Bjorn Andersson <andersson@kernel.org>
Reported-by: Clayton Craft <clayton@craftyguy.net>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Acked-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Tested-by: Clayton Craft <clayton@craftyguy.net>
Link: https://lore.kernel.org/r/20250324132448.6134-1-johan+linaro@kernel.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Barnabás Czémán
421777a02b soc: qcom: smp2p: Fix fallback to qcom,ipc parse
mbox_request_channel() returning value was changed in case of error.
It uses returning value of of_parse_phandle_with_args().
It is returning with -ENOENT instead of -ENODEV when no mboxes property
exists.

Fixes: 24fdd5074b ("mailbox: use error ret code of of_parse_phandle_with_args()")
Signed-off-by: Barnabás Czémán <barnabas.czeman@mainlining.org>
Reviewed-by: Stephan Gerhold <stephan.gerhold@linaro.org>
Tested-by: Stephan Gerhold <stephan.gerhold@linaro.org> # msm8939
Link: https://lore.kernel.org/r/20250421-fix-qcom-smd-v1-2-574d071d3f27@mainlining.org
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2025-05-09 17:12:55 -05:00
Al Viro
d1ddc6f1d9 fix IS_MNT_PROPAGATING uses
propagate_mnt() does not attach anything to mounts created during
propagate_mnt() itself.  What's more, anything on ->mnt_slave_list
of such new mount must also be new, so we don't need to even look
there.

When move_mount() had been introduced, we've got an additional
class of mounts to skip - if we are moving from anon namespace,
we do not want to propagate to mounts we are moving (i.e. all
mounts in that anon namespace).

Unfortunately, the part about "everything on their ->mnt_slave_list
will also be ignorable" is not true - if we have propagation graph
	A -> B -> C
and do OPEN_TREE_CLONE open_tree() of B, we get
	A -> [B <-> B'] -> C
as propagation graph, where B' is a clone of B in our detached tree.
Making B private will result in
	A -> B' -> C
C still gets propagation from A, as it would after making B private
if we hadn't done that open_tree(), but now the propagation goes
through B'.  Trying to move_mount() our detached tree on subdirectory
in A should have
	* moved B' on that subdirectory in A
	* skipped the corresponding subdirectory in B' itself
	* copied B' on the corresponding subdirectory in C.
As it is, the logics in propagation_next() and friends ends up
skipping propagation into C, since it doesn't consider anything
downstream of B'.

IOW, walking the propagation graph should only skip the ->mnt_slave_list
of new mounts; the only places where the check for "in that one
anon namespace" are applicable are propagate_one() (where we should
treat that as the same kind of thing as "mountpoint we are looking
at is not visible in the mount we are looking at") and
propagation_would_overmount().  The latter is better dealt with
in the caller (can_move_mount_beneath()); on the first call of
propagation_would_overmount() the test is always false, on the
second it is always true in "move from anon namespace" case and
always false in "move within our namespace" one, so it's easier
to just use check_mnt() before bothering with the second call and
be done with that.

Fixes: 064fe6e233 ("mount: handle mount propagation for detached mount trees")
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2025-05-09 18:06:27 -04:00
Al Viro
267fc3a06a do_move_mount(): don't leak MNTNS_PROPAGATING on failures
as it is, a failed move_mount(2) from anon namespace breaks
all further propagation into that namespace, including normal
mounts in non-anon namespaces that would otherwise propagate
there.

Fixes: 064fe6e233 ("mount: handle mount propagation for detached mount trees")
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2025-05-09 18:06:10 -04:00
Al Viro
65781e19dc do_umount(): add missing barrier before refcount checks in sync case
do_umount() analogue of the race fixed in 119e1ef80e "fix
__legitimize_mnt()/mntput() race".  Here we want to make sure that
if __legitimize_mnt() doesn't notice our lock_mount_hash(), we will
notice their refcount increment.  Harder to hit than mntput_no_expire()
one, fortunately, and consequences are milder (sync umount acting
like umount -l on a rare race with RCU pathwalk hitting at just the
wrong time instead of use-after-free galore mntput_no_expire()
counterpart used to be hit).  Still a bug...

Fixes: 48a066e72d ("RCU'd vfsmounts")
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2025-05-09 18:05:55 -04:00
Cosmin Ratiu
97c4e094a4 tests/ncdevmem: Fix double-free of queue array
netdev_bind_rx takes ownership of the queue array passed as parameter
and frees it, so a queue array buffer cannot be reused across multiple
netdev_bind_rx calls.

This commit fixes that by always passing in a newly created queue array
to all netdev_bind_rx calls in ncdevmem.

Fixes: 85585b4bc8 ("selftests: add ncdevmem, netcat for devmem TCP")
Signed-off-by: Cosmin Ratiu <cratiu@nvidia.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Reviewed-by: Joe Damato <jdamato@fastly.com>
Reviewed-by: Mina Almasry <almasrymina@google.com>
Link: https://patch.msgid.link/20250508084434.1933069-1-cratiu@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 15:05:07 -07:00
Matt Johnston
f11cf946c0 net: mctp: Don't access ifa_index when missing
In mctp_dump_addrinfo, ifa_index can be used to filter interfaces, but
only when the struct ifaddrmsg is provided. Otherwise it will be
comparing to uninitialised memory - reproducible in the syzkaller case from
dhcpd, or busybox "ip addr show".

The kernel MCTP implementation has always filtered by ifa_index, so
existing userspace programs expecting to dump MCTP addresses must
already be passing a valid ifa_index value (either 0 or a real index).

BUG: KMSAN: uninit-value in mctp_dump_addrinfo+0x208/0xac0 net/mctp/device.c:128
 mctp_dump_addrinfo+0x208/0xac0 net/mctp/device.c:128
 rtnl_dump_all+0x3ec/0x5b0 net/core/rtnetlink.c:4380
 rtnl_dumpit+0xd5/0x2f0 net/core/rtnetlink.c:6824
 netlink_dump+0x97b/0x1690 net/netlink/af_netlink.c:2309

Fixes: 583be982d9 ("mctp: Add device handling and netlink interface")
Reported-by: syzbot+e76d52dadc089b9d197f@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/68135815.050a0220.3a872c.000e.GAE@google.com/
Reported-by: syzbot+1065a199625a388fce60@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/681357d6.050a0220.14dd7d.000d.GAE@google.com/
Signed-off-by: Matt Johnston <matt@codeconstruct.com.au>
Link: https://patch.msgid.link/20250508-mctp-addr-dump-v2-1-c8a53fd2dd66@codeconstruct.com.au
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 15:03:53 -07:00
Hangbin Liu
45375814eb tools/net/ynl: ethtool: fix crash when Hardware Clock info is missing
Fix a crash in the ethtool YNL implementation when Hardware Clock information
is not present in the response. This ensures graceful handling of devices or
drivers that do not provide this optional field. e.g.

  Traceback (most recent call last):
    File "/net/tools/net/ynl/pyynl/./ethtool.py", line 438, in <module>
      main()
      ~~~~^^
    File "/net/tools/net/ynl/pyynl/./ethtool.py", line 341, in main
      print(f'PTP Hardware Clock: {tsinfo["phc-index"]}')
                                   ~~~~~~^^^^^^^^^^^^^
  KeyError: 'phc-index'

Fixes: f3d07b02b2 ("tools: ynl: ethtool testing tool")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://patch.msgid.link/20250508035414.82974-1-liuhangbin@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2025-05-09 15:03:07 -07:00
Al Viro
250cf36930 __legitimize_mnt(): check for MNT_SYNC_UMOUNT should be under mount_lock
... or we risk stealing final mntput from sync umount - raising mnt_count
after umount(2) has verified that victim is not busy, but before it
has set MNT_SYNC_UMOUNT; in that case __legitimize_mnt() doesn't see
that it's safe to quietly undo mnt_count increment and leaves dropping
the reference to caller, where it'll be a full-blown mntput().

Check under mount_lock is needed; leaving the current one done before
taking that makes no sense - it's nowhere near common enough to bother
with.

Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2025-05-09 18:02:12 -04:00