linux-stable.git/drivers/net/dsa, branch v5.9.2

net: dsa: seville: the packet buffer is 2 megabits, not megabytes

2020-10-29T09:11:59+00:00

[ Upstream commit a15a6afb3bf9388eb83a4b876d3453f305fba909 ]

The VSC9953 Seville switch has 2 megabits of buffer split into 4360
words of 60 bytes each. 2048 * 1024 is 2 megabytes instead of 2 megabits.
2 megabits is (2048 / 8) * 1024 = 256 * 1024.

Signed-off-by: Maxim Kochetkov 
Reviewed-by: Vladimir Oltean 
Fixes: a63ed92d217f ("net: dsa: seville: fix buffer size of the queue system")
Link: https://lore.kernel.org/r/20201019050625.21533-1-fido_max@inbox.ru
Signed-off-by: Jakub Kicinski 
Signed-off-by: Sasha Levin

net: dsa: rtl8366rb: Support all 4096 VLANs

2020-10-29T09:11:25+00:00

[ Upstream commit a7920efdd86d8a0d74402dbc80ead03b023294ba ]

There is an off-by-one error in rtl8366rb_is_vlan_valid()
making VLANs 0..4094 valid while it should be 1..4095.
Fix it.

Fixes: d8652956cf37 ("net: dsa: realtek-smi: Add Realtek SMI driver")
Signed-off-by: Linus Walleij 
Reviewed-by: Florian Fainelli 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

net: dsa: rtl8366: Skip PVID setting if not requested

2020-10-29T09:11:24+00:00

[ Upstream commit 3dfe8dde093a07e82fa472c0f8c29a7f6a2006a5 ]

We go to lengths to determine whether the PVID should be set
for this port or not, and then fail to take it into account.
Fix this oversight.

Fixes: d8652956cf37 ("net: dsa: realtek-smi: Add Realtek SMI driver")
Signed-off-by: Linus Walleij 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

net: dsa: rtl8366: Refactor VLAN/PVID init

2020-10-29T09:11:24+00:00

[ Upstream commit 7e1301ed1881447d2a25f9c6423738c33cbca133 ]

The VLANs and PVIDs on the RTL8366 utilizes a "member
configuration" (MC) which is largely unexplained in the
code.

This set-up requires a special ordering: rtl8366_set_pvid()
must be called first, followed by rtl8366_set_vlan(),
else the MC will not be properly allocated. Relax this
by factoring out the code obtaining an MC and reuse
the helper in both rtl8366_set_pvid() and
rtl8366_set_vlan() so we remove this strict ordering
requirement.

In the process, add some better comments and debug prints
so people who read the code understand what is going on.

Reviewed-by: Florian Fainelli 
Signed-off-by: Linus Walleij 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

net: dsa: rtl8366: Check validity of passed VLANs

2020-10-29T09:11:23+00:00

[ Upstream commit 6641a2c42b0a307b7638d10e5d4b90debc61389d ]

The rtl8366_set_vlan() and rtl8366_set_pvid() get invalid
VLANs tossed at it, especially VLAN0, something the hardware
and driver cannot handle. Check validity and bail out like
we do in the other callbacks.

Reviewed-by: Florian Fainelli 
Signed-off-by: Linus Walleij 
Signed-off-by: David S. Miller 
Signed-off-by: Sasha Levin

net: dsa: microchip: fix race condition

2020-10-29T09:10:50+00:00

[ Upstream commit 8098bd69bc4e925070313b1b95d03510f4f24738 ]

Between queuing the delayed work and finishing the setup of the dsa
ports, the process may sleep in request_module() (via
phy_device_create()) and the queued work may be executed prior to the
switch net devices being registered. In ksz_mib_read_work(), a NULL
dereference will happen within netof_carrier_ok(dp->slave).

Not queuing the delayed work in ksz_init_mib_timer() makes things even
worse because the work will now be queued for immediate execution
(instead of 2000 ms) in ksz_mac_link_down() via
dsa_port_link_register_of().

Call tree:
ksz9477_i2c_probe()
\--ksz9477_switch_register()
   \--ksz_switch_register()
      +--dsa_register_switch()
      |  \--dsa_switch_probe()
      |     \--dsa_tree_setup()
      |        \--dsa_tree_setup_switches()
      |           +--dsa_switch_setup()
      |           |  +--ksz9477_setup()
      |           |  |  \--ksz_init_mib_timer()
      |           |  |     |--/* Start the timer 2 seconds later. */
      |           |  |     \--schedule_delayed_work(&dev->mib_read, msecs_to_jiffies(2000));
      |           |  \--__mdiobus_register()
      |           |     \--mdiobus_scan()
      |           |        \--get_phy_device()
      |           |           +--get_phy_id()
      |           |           \--phy_device_create()
      |           |              |--/* sleeping, ksz_mib_read_work() can be called meanwhile */
      |           |              \--request_module()
      |           |
      |           \--dsa_port_setup()
      |              +--/* Called for non-CPU ports */
      |              +--dsa_slave_create()
      |              |  +--/* Too late, ksz_mib_read_work() may be called beforehand */
      |              |  \--port->slave = ...
      |             ...
      |              +--Called for CPU port */
      |              \--dsa_port_link_register_of()
      |                 \--ksz_mac_link_down()
      |                    +--/* mib_read must be initialized here */
      |                    +--/* work is already scheduled, so it will be executed after 2000 ms */
      |                    \--schedule_delayed_work(&dev->mib_read, 0);
      \-- /* here port->slave is setup properly, scheduling the delayed work should be safe */

Solution:
1. Do not queue (only initialize) delayed work in ksz_init_mib_timer().
2. Only queue delayed work in ksz_mac_link_down() if init is completed.
3. Queue work once in ksz_switch_register(), after dsa_register_switch()
has completed.

Fixes: 7c6ff470aa86 ("net: dsa: microchip: add MIB counter reading support")
Signed-off-by: Christian Eggers 
Reviewed-by: Florian Fainelli 
Reviewed-by: Vladimir Oltean 
Signed-off-by: Jakub Kicinski 
Signed-off-by: Greg Kroah-Hartman

net: mscc: ocelot: warn when encoding an out-of-bounds watermark value

2020-10-06T13:05:47+00:00

There is an upper bound to the value that a watermark may hold. That
upper bound is not immediately obvious during configuration, and it
might be possible to have accidental truncation.

Actually this has happened already, add a warning to prevent it from
happening again.

Signed-off-by: Vladimir Oltean 
Signed-off-by: David S. Miller

net: dsa: seville: fix VCAP IS2 action width

2020-09-29T20:24:17+00:00

Since the actions are packed together in the action RAM, an incorrect
action width means that no action except the first one would behave
correctly.

The tc-flower offload has probably not been tested on this hardware
since its introduction.

Signed-off-by: Vladimir Oltean 
Signed-off-by: David S. Miller

net: dsa: felix: fix incorrect action offsets for VCAP IS2

2020-09-29T20:24:17+00:00

The port mask width was larger than the actual number of ports, and
therefore, all fields following this one were also shifted by the number
of excess bits. But the driver doesn't use the REW_OP, SMAC_REPLACE_ENA
or ACL_ID bits from the action vector, so the bug was inconsequential.

Signed-off-by: Vladimir Oltean 
Signed-off-by: David S. Miller

net: dsa: microchip: really look for phy-mode in port nodes

2020-09-25T03:08:48+00:00

The previous implementation failed to account for the "ports" node. The
actual port nodes are not child nodes of the switch node, but a "ports"
node sits in between.

Fixes: edecfa98f602 ("net: dsa: microchip: look for phy-mode in port nodes")
Signed-off-by: Helmut Grohne 
Signed-off-by: David S. Miller