linux-stable.git/drivers/net, branch v6.6.7

r8169: fix rtl8125b PAUSE frames blasting when suspended

2023-12-13T17:45:24+00:00

commit 4b0768b6556af56ee9b7cf4e68452a2b6289ae45 upstream.

When FIFO reaches near full state, device will issue pause frame.
If pause slot is enabled(set to 1), in this time, device will issue
pause frame only once. But if pause slot is disabled(set to 0), device
will keep sending pause frames until FIFO reaches near empty state.

When pause slot is disabled, if there is no one to handle receive
packets, device FIFO will reach near full state and keep sending
pause frames. That will impact entire local area network.

This issue can be reproduced in Chromebox (not Chromebook) in
developer mode running a test image (and v5.10 kernel):
1) ping -f $CHROMEBOX (from workstation on same local network)
2) run "powerd_dbus_suspend" from command line on the $CHROMEBOX
3) ping $ROUTER (wait until ping fails from workstation)

Takes about ~20-30 seconds after step 2 for the local network to
stop working.

Fix this issue by enabling pause slot to only send pause frame once
when FIFO reaches near full state.

Fixes: f1bce4ad2f1c ("r8169: add support for RTL8125")
Reported-by: Grant Grundler 
Tested-by: Grant Grundler 
Cc: stable@vger.kernel.org
Signed-off-by: ChunHao Lin 
Reviewed-by: Jacob Keller 
Reviewed-by: Heiner Kallweit 
Link: https://lore.kernel.org/r/20231129155350.5843-1-hau@realtek.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Greg Kroah-Hartman

nfp: flower: fix for take a mutex lock in soft irq context and rcu lock

2023-12-13T17:45:23+00:00

commit 0ad722bd9ee3a9bdfca9613148645e4c9b7f26cf upstream.

The neighbour event callback call the function nfp_tun_write_neigh,
this function will take a mutex lock and it is in soft irq context,
change the work queue to process the neighbour event.

Move the nfp_tun_write_neigh function out of range rcu_read_lock/unlock()
in function nfp_tunnel_request_route_v4 and nfp_tunnel_request_route_v6.

Fixes: abc210952af7 ("nfp: flower: tunnel neigh support bond offload")
CC: stable@vger.kernel.org # 6.2+
Signed-off-by: Hui Zhou 
Signed-off-by: Louis Peens 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net: dsa: microchip: provide a list of valid protocols for xmit handler

2023-12-13T17:45:11+00:00

[ Upstream commit 1499b89289bf272fd83cb296c82fb5519d0fe93f ]

Provide a list of valid protocols for which the driver will provide
it's deferred xmit handler.

When using DSA_TAG_PROTO_KSZ8795 protocol, it does not provide a
"connect" method, therefor ksz_connect() is not allocating ksz_tagger_data.

This avoids the following null pointer dereference:
 ksz_connect_tag_protocol from dsa_register_switch+0x9ac/0xee0
 dsa_register_switch from ksz_switch_register+0x65c/0x828
 ksz_switch_register from ksz_spi_probe+0x11c/0x168
 ksz_spi_probe from spi_probe+0x84/0xa8
 spi_probe from really_probe+0xc8/0x2d8

Fixes: ab32f56a4100 ("net: dsa: microchip: ptp: add packet transmission timestamping")
Signed-off-by: Sean Nyekjaer 
Reviewed-by: Florian Fainelli 
Reviewed-by: Vladimir Oltean 
Link: https://lore.kernel.org/r/20231206071655.1626479-1-sean@geanix.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

net: dsa: mv88e6xxx: Restore USXGMII support for 6393X

2023-12-13T17:45:10+00:00

[ Upstream commit 0c7ed1f9197aecada33a08b022e484a97bf584ba ]

In 4a56212774ac, USXGMII support was added for 6393X, but this was
lost in the PCS conversion (the blamed commit), most likely because
these efforts where more or less done in parallel.

Restore this feature by porting Michal's patch to fit the new
implementation.

Reviewed-by: Florian Fainelli 
Tested-by: Michal Smulski 
Reviewed-by: Vladimir Oltean 
Fixes: e5b732a275f5 ("net: dsa: mv88e6xxx: convert 88e639x to phylink_pcs")
Signed-off-by: Tobias Waldekranz 
Link: https://lore.kernel.org/r/20231205221359.3926018-1-tobias@waldekranz.com
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

octeontx2-af: Update Tx link register range

2023-12-13T17:45:09+00:00

[ Upstream commit 7336fc196748f82646b630d5a2e9d283e200b988 ]

On new silicons the TX channels for transmit level has increased.
This patch fixes the respective register offset range to
configure the newly added channels.

Fixes: b279bbb3314e ("octeontx2-af: NIX Tx scheduler queue config support")
Signed-off-by: Rahul Bhansali 
Signed-off-by: Geetha sowjanya 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

octeontx2-af: Add missing mcs flr handler call

2023-12-13T17:45:09+00:00

[ Upstream commit d431abd0a9aa27be379fb5f8304062071b0f5a7e ]

If mcs resources are attached to PF/VF. These resources need
to be freed on FLR. This patch add missing mcs flr call on PF FLR.

Fixes: bd69476e86fc ("octeontx2-af: cn10k: mcs: Install a default TCAM for normal traffic")
Signed-off-by: Geetha sowjanya 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

octeontx2-af: Fix mcs stats register address

2023-12-13T17:45:09+00:00

[ Upstream commit 3ba98a8c6f8ceb4e01a78f973d8d9017020bbd57 ]

This patch adds the miss mcs stats register
for mcs supported platforms.

Fixes: 9312150af8da ("octeontx2-af: cn10k: mcs: Support for stats collection")
Signed-off-by: Geetha sowjanya 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

octeontx2-af: Fix mcs sa cam entries size

2023-12-13T17:45:08+00:00

[ Upstream commit 9723b2cca1f0e980c53156b52ea73b93966b3c8a ]

On latest silicon versions SA cam entries increased to 256.
This patch fixes the datatype of sa_entries in mcs_hw_info
struct to u16 to hold 256 entries.

Fixes: 080bbd19c9dd ("octeontx2-af: cn10k: mcs: Add mailboxes for port related operations")
Signed-off-by: Geetha sowjanya 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

octeontx2-af: Adjust Tx credits when MCS external bypass is disabled

2023-12-13T17:45:08+00:00

[ Upstream commit dca6fa8644b89f54345e55501b1419316ba5cb29 ]

When MCS external bypass is disabled, MCS returns additional
2 credits(32B) for every packet Tx'ed on LMAC. To account for
these extra credits, NIX_AF_TX_LINKX_NORM_CREDIT.CC_MCS_CNT
needs to be configured as otherwise NIX Tx credits would overflow
and will never be returned to idle state credit count
causing issues with credit control and MTU change.

This patch fixes the same by configuring CC_MCS_CNT at probe
time for MCS enabled SoC's

Fixes: bd69476e86fc ("octeontx2-af: cn10k: mcs: Install a default TCAM for normal traffic")
Signed-off-by: Nithin Dabilpuram 
Signed-off-by: Geetha sowjanya 
Signed-off-by: Sunil Goutham 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin

net: hns: fix fake link up on xge port

2023-12-13T17:45:08+00:00

[ Upstream commit f708aba40f9c1eeb9c7e93ed4863b5f85b09b288 ]

If a xge port just connect with an optical module and no fiber,
it may have a fake link up because there may be interference on
the hardware. This patch adds an anti-shake to avoid the problem.
And the time of anti-shake is base on tests.

Fixes: b917078c1c10 ("net: hns: Add ACPI support to check SFP present")
Signed-off-by: Yonglong Liu 
Signed-off-by: Jijie Shao 
Reviewed-by: Wojciech Drewek 
Signed-off-by: Paolo Abeni 
Signed-off-by: Sasha Levin