linux.git/drivers/staging/fsl-dpaa2, branch v5.10.258

staging: dpaa2-switch: add a dpaa2_switch prefix to all functions in ethsw.c

2020-10-10T10:33:41+00:00

Some static functions in the dpaa2-switch driver don't have a distinct
prefix and this is becoming an inconvenience when looking at, for
example, a perf top output and trying to determine easily which entries
are dpaa2 switch related. Ammend this by adding the prefix to all the
functions.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20201009153000.14550-3-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-switch: add a dpaa2_switch_ prefix to all functions in ethsw-ethtool.c

2020-10-10T10:33:41+00:00

Some static functions in the dpaa2-switch driver don't have a distinct
prefix and this is becoming an inconvenience when looking at, for
example, a perf top output and trying to determine easily which entries
are dpaa2 switch related. Ammend this by adding the prefix to all the
functions.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20201009153000.14550-2-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: check if there is space for a new VLAN

2020-07-21T14:53:34+00:00

Avoid getting into a WARNING as below by checking, while in the prepare
state of the transactional operation, if there is space for a new VLAN.
If we reached the maximum number, return an appropriate error.

[ 6503.657564] eth3: Commit of object (id=1) failed.
[ 6503.657588] WARNING: CPU: 2 PID: 17144 at net/switchdev/switchdev.c:277 switchdev_port_obj_add_now+0xcc/0x110
...
[ 6503.657628] x1 : 70887ce26695c500 x0 : 0000000000000000
[ 6503.657630] Call trace:
[ 6503.657633]  switchdev_port_obj_add_now+0xcc/0x110
[ 6503.657635]  switchdev_port_obj_add+0x40/0xc0
[ 6503.657638]  br_switchdev_port_vlan_add+0x50/0x78
[ 6503.657640]  __vlan_add+0x2dc/0x758
[ 6503.657642]  nbp_vlan_add+0xc0/0x180
[ 6503.657644]  br_vlan_info.isra.0+0x68/0x128
[ 6503.657646]  br_process_vlan_info+0x224/0x2f8
[ 6503.657647]  br_afspec+0x158/0x188
[ 6503.657649]  br_setlink+0x1a4/0x290

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-7-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: read the port state from firmware

2020-07-21T14:53:34+00:00

Rely on the port state seen by the firmware since it will also be the
one erroring out when trying to setup anything major when the port is
up.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-6-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: destroy workqueue after deregistering the notifiers

2020-07-21T14:53:34+00:00

We should destroy the switch workqueue only after deregistering the
switchdev notifiers. Without this fix, we could end up with switchdev
notifications on a draining workqueue and also with a lock up since the
netdevice reference count is increased (in port_switchdev_event) and not
decreased ever (since the workqueue did not run).

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-5-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: setup the STP state for all installed VLANs

2020-07-21T14:53:34+00:00

Setup the STP state for all VLANs installed on the port. This is also
avoiding the error situation when the DEFAULT_VLAN_ID is not installed
on the port (thus the firmware complains that it cannot setup the
required STP state).

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-4-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: don't allow interfaces from different DPSWs to be bridged

2020-07-21T14:53:34+00:00

Error out when the user tries to bridge two switch interfaces that are
from different DPSW instances. This is not supported by the hardware and
we should reflect this into what the user is aware of.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-3-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: verify the nofifier block

2020-07-21T14:53:34+00:00

Since now we have a notifier block for each DPSW instance probed, we
have to also check that the netdev is indeed connected to the notifier
received. Without this, we end up with the same switchdev callback being
executed multiple times (because it would be received by all notifier
blocks, not just the one intended to).
Also, move the function higher in the source file because it will be
used in later patches from multiple places.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200721091919.20394-2-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: fix switch/case fallthrough warning

2020-07-21T14:52:47+00:00

Fix the fallthrough warning that is reported by checkpatch.

Signed-off-by: Marian Posteuca 
Link: https://lore.kernel.org/r/20200720064205.10323-1-posteuca@mutex.one
Signed-off-by: Greg Kroah-Hartman

staging: dpaa2-ethsw: setup MAC address of switch netdevices

2020-07-15T14:22:36+00:00

At probe time, retrieve the MAC addresses of the switch ports using a
firmware call and use that to setup the switch interface net_device
instead of relying entirely on the user to configure a MAC address on
the interface. In case a switch interface is not connected to a MAC,
thus the dpsw_if_get_port_mac_addr() will return all zeroes, generate a
random MAC address and use that.

This new functionality is dependent on a firmware call which is
available only on newer versions, so depending on the running DPSW
object version skip this step.

Signed-off-by: Ioana Ciornei 
Link: https://lore.kernel.org/r/20200714133431.17532-7-ioana.ciornei@nxp.com
Signed-off-by: Greg Kroah-Hartman