linux.git/drivers/net/ethernet, branch v6.9

net: hns3: fix kernel crash when devlink reload during initialization

2024-05-09T08:47:32+00:00

The devlink reload process will access the hardware resources,
but the register operation is done before the hardware is initialized.
So, processing the devlink reload during initialization may lead to kernel
crash.

This patch fixes this by registering the devlink after
hardware initialization.

Fixes: cd6242991d2e ("net: hns3: add support for registering devlink for VF")
Fixes: 93305b77ffcb ("net: hns3: fix kernel crash when devlink reload during pf initialization")
Signed-off-by: Yonglong Liu 
Signed-off-by: Jijie Shao 
Signed-off-by: Paolo Abeni

net: hns3: fix port vlan filter not disabled issue

2024-05-09T08:47:32+00:00

According to hardware limitation, for device support modify
VLAN filter state but not support bypass port VLAN filter,
it should always disable the port VLAN filter. but the driver
enables port VLAN filter when initializing, if there is no
VLAN(except VLAN 0) id added, the driver will disable it
in service task. In most time, it works fine. But there is
a time window before the service task shceduled and net device
being registered. So if user adds VLAN at this time, the driver
will not update the VLAN filter state,  and the port VLAN filter
remains enabled.

To fix the problem, if support modify VLAN filter state but not
support bypass port VLAN filter, set the port vlan filter to "off".

Fixes: 184cd221a863 ("net: hns3: disable port VLAN filter when support function level VLAN filter control")
Fixes: 2ba306627f59 ("net: hns3: add support for modify VLAN filter state")
Signed-off-by: Yonglong Liu 
Signed-off-by: Jijie Shao 
Reviewed-by: Simon Horman 
Signed-off-by: Paolo Abeni

net: hns3: use appropriate barrier function after setting a bit value

2024-05-09T08:47:32+00:00

There is a memory barrier in followed case. When set the port down,
hclgevf_set_timmer will set DOWN in state. Meanwhile, the service task has
different behaviour based on whether the state is DOWN. Thus, to make sure
service task see DOWN, use smp_mb__after_atomic after calling set_bit().

          CPU0                        CPU1
========================== ===================================
hclgevf_set_timer_task()    hclgevf_periodic_service_task()
  set_bit(DOWN,state)         test_bit(DOWN,state)

pf also has this issue.

Fixes: ff200099d271 ("net: hns3: remove unnecessary work in hclgevf_main")
Fixes: 1c6dfe6fc6f7 ("net: hns3: remove mailbox and reset work in hclge_main")
Signed-off-by: Peiyang Wang 
Signed-off-by: Jijie Shao 
Reviewed-by: Simon Horman 
Signed-off-by: Paolo Abeni

net: hns3: release PTP resources if pf initialization failed

2024-05-09T08:47:32+00:00

During the PF initialization process, hclge_update_port_info may return an
error code for some reason. At this point,  the ptp initialization has been
completed. To void memory leaks, the resources that are applied by ptp
should be released. Therefore, when hclge_update_port_info returns an error
code, hclge_ptp_uninit is called to release the corresponding resources.

Fixes: eaf83ae59e18 ("net: hns3: add querying fec ability from firmware")
Signed-off-by: Peiyang Wang 
Signed-off-by: Jijie Shao 
Reviewed-by: Hariprasad Kelam 
Signed-off-by: Paolo Abeni

net: hns3: change type of numa_node_mask as nodemask_t

2024-05-09T08:47:32+00:00

It provides nodemask_t to describe the numa node mask in kernel. To
improve transportability, change the type of numa_node_mask as nodemask_t.

Fixes: 38caee9d3ee8 ("net: hns3: Add support of the HNAE3 framework")
Signed-off-by: Peiyang Wang 
Signed-off-by: Jijie Shao 
Reviewed-by: Simon Horman 
Signed-off-by: Paolo Abeni

net: hns3: direct return when receive a unknown mailbox message

2024-05-09T08:47:31+00:00

Currently, the driver didn't return when receive a unknown
mailbox message, and continue checking whether need to
generate a response. It's unnecessary and may be incorrect.

Fixes: bb5790b71bad ("net: hns3: refactor mailbox response scheme between PF and VF")
Signed-off-by: Jian Shen 
Signed-off-by: Jijie Shao 
Reviewed-by: Simon Horman 
Signed-off-by: Paolo Abeni

net: hns3: using user configure after hardware reset

2024-05-09T08:47:31+00:00

When a reset occurring, it's supposed to recover user's configuration.
Currently, the port info(speed, duplex and autoneg) is stored in hclge_mac
and will be scheduled updated. Consider the case that reset was happened
consecutively. During the first reset, the port info is configured with
a temporary value cause the PHY is reset and looking for best link config.
Second reset start and use pervious configuration which is not the user's.
The specific process is as follows:

+------+               +----+                +----+
| USER |               | PF |                | HW |
+---+--+               +-+--+                +-+--+
    |  ethtool --reset   |                     |
    +------------------->|    reset command    |
    |  ethtool --reset   +-------------------->|
    +------------------->|                     +---+
    |                    +---+                 |   |
    |                    |   |reset currently  |   | HW RESET
    |                    |   |and wait to do   |   |
    |                    |<--+                 |   |
    |                    | send pervious cfg   |<--+
    |                    | (1000M FULL AN_ON)  |
    |                    +-------------------->|
    |                    | read cfg(time task) |
    |                    | (10M HALF AN_OFF)   +---+
    |                    |<--------------------+   | cfg take effect
    |                    |    reset command    |<--+
    |                    +-------------------->|
    |                    |                     +---+
    |                    | send pervious cfg   |   | HW RESET
    |                    | (10M HALF AN_OFF)   |<--+
    |                    +-------------------->|
    |                    | read cfg(time task) |
    |                    |  (10M HALF AN_OFF)  +---+
    |                    |<--------------------+   | cfg take effect
    |                    |                     |   |
    |                    | read cfg(time task) |<--+
    |                    |  (10M HALF AN_OFF)  |
    |                    |<--------------------+
    |                    |                     |
    v                    v                     v

To avoid aboved situation, this patch introduced req_speed, req_duplex,
req_autoneg to store user's configuration and it only be used after
hardware reset and to recover user's configuration

Fixes: f5f2b3e4dcc0 ("net: hns3: add support for imp-controlled PHYs")
Signed-off-by: Peiyang Wang 
Signed-off-by: Jijie Shao 
Reviewed-by: Przemek Kitszel 
Reviewed-by: Simon Horman 
Signed-off-by: Paolo Abeni

net: ks8851: Queue RX packets in IRQ handler instead of disabling BHs

2024-05-03T22:15:03+00:00

Currently the driver uses local_bh_disable()/local_bh_enable() in its
IRQ handler to avoid triggering net_rx_action() softirq on exit from
netif_rx(). The net_rx_action() could trigger this driver .start_xmit
callback, which is protected by the same lock as the IRQ handler, so
calling the .start_xmit from netif_rx() from the IRQ handler critical
section protected by the lock could lead to an attempt to claim the
already claimed lock, and a hang.

The local_bh_disable()/local_bh_enable() approach works only in case
the IRQ handler is protected by a spinlock, but does not work if the
IRQ handler is protected by mutex, i.e. this works for KS8851 with
Parallel bus interface, but not for KS8851 with SPI bus interface.

Remove the BH manipulation and instead of calling netif_rx() inside
the IRQ handler code protected by the lock, queue all the received
SKBs in the IRQ handler into a queue first, and once the IRQ handler
exits the critical section protected by the lock, dequeue all the
queued SKBs and push them all into netif_rx(). At this point, it is
safe to trigger the net_rx_action() softirq, since the netif_rx()
call is outside of the lock that protects the IRQ handler.

Fixes: be0384bf599c ("net: ks8851: Handle softirqs at the end of IRQ thread to fix hang")
Tested-by: Ronald Wahl  # KS8851 SPI
Signed-off-by: Marek Vasut 
Reviewed-by: Eric Dumazet 
Link: https://lore.kernel.org/r/20240502183436.117117-1-marex@denx.de
Signed-off-by: Jakub Kicinski

e1000e: change usleep_range to udelay in PHY mdic access

2024-05-01T01:40:59+00:00

This is a partial revert of commit 6dbdd4de0362 ("e1000e: Workaround
for sporadic MDI error on Meteor Lake systems"). The referenced commit
used usleep_range inside the PHY access routines, which are sometimes
called from an atomic context. This can lead to a kernel panic in some
scenarios, such as cable disconnection and reconnection on vPro systems.

Solve this by changing the usleep_range calls back to udelay.

Fixes: 6dbdd4de0362 ("e1000e: Workaround for sporadic MDI error on Meteor Lake systems")
Cc: stable@vger.kernel.org
Reported-by: Jérôme Carretero 
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218740
Closes: https://lore.kernel.org/lkml/a7eb665c74b5efb5140e6979759ed243072cb24a.camel@zougloub.eu/
Co-developed-by: Sasha Neftin 
Signed-off-by: Sasha Neftin 
Signed-off-by: Vitaly Lifshits 
Tested-by: Dima Ruinskiy 
Signed-off-by: Tony Nguyen 
Reviewed-by: Simon Horman 
Link: https://lore.kernel.org/r/20240429171040.1152516-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski

cxgb4: Properly lock TX queue for the selftest.

2024-05-01T01:30:56+00:00

The selftest for the driver sends a dummy packet and checks if the
packet will be received properly as it should be. The regular TX path
and the selftest can use the same network queue so locking is required
and was missing in the selftest path. This was addressed in the commit
cited below.
Unfortunately locking the TX queue requires BH to be disabled which is
not the case in selftest path which is invoked in process context.
Lockdep should be complaining about this.

Use __netif_tx_lock_bh() for TX queue locking.

Fixes: c650e04898072 ("cxgb4: Fix race between loopback and normal Tx path")
Reported-by: "John B. Wyatt IV" 
Closes: https://lore.kernel.org/all/Zic0ot5aGgR-V4Ks@thinkpad2021/
Signed-off-by: Sebastian Andrzej Siewior 
Link: https://lore.kernel.org/r/20240429091147.YWAaal4v@linutronix.de
Signed-off-by: Jakub Kicinski