linux-toradex.git/include/linux/mlx5/driver.h, branch v4.9.16

net/mlx5: PCI error recovery health care simulation

2016-10-29T16:00:39+00:00

In case that the kernel PCI error handlers are not called, we will
trigger our own recovery flow.

The health work will give priority to the kernel pci error handlers to
recover the PCI by waiting for a small period, if the pci error handlers
are not triggered the manual recovery flow will be executed.

We don't save pci state in case of manual recovery because it will ruin the
pci configuration space and we will lose dma sync.

Fixes: 89d44f0a6c73 ('net/mlx5_core: Add pci error handlers to mlx5_core driver')
Signed-off-by: Mohamad Haj Yahia 
Signed-off-by: Saeed Mahameed 
Signed-off-by: David S. Miller

net/mlx5: Fix race between PCI error handlers and health work

2016-10-29T16:00:39+00:00

Currently there is a race between the health care work and the kernel
pci error handlers because both of them detect the error, the first one
to be called will do the error handling.
There is a chance that health care will disable the pci after resuming
pci slot.
Also create a separate WQ because now we will have two types of health
works, one for the error detection and one for the recovery.

Fixes: 89d44f0a6c73 ('net/mlx5_core: Add pci error handlers to mlx5_core driver')
Signed-off-by: Mohamad Haj Yahia 
Signed-off-by: Saeed Mahameed 
Signed-off-by: David S. Miller

{net, ib}/mlx5: Make cache line size determination at runtime.

2016-10-29T16:00:39+00:00

ARM 64B cache line systems have L1_CACHE_BYTES set to 128.
cache_line_size() will return the correct size.

Fixes: cf50b5efa2fe('net/mlx5_core/ib: New device capabilities
handling.')
Signed-off-by: Daniel Jurgens 

Signed-off-by: Saeed Mahameed 
Signed-off-by: David S. Miller

net/mlx5: Introduce attach/detach to interface API

2016-09-11T04:21:49+00:00

Add attach/detach callbacks to interface API.
This is crucial for implementing seamless reset flow which releases the
hardware and it's resources upon detach while keeping software
structures and state (e.g netdev) then reset and reallocate the hardware
needed resources upon attach.

Signed-off-by: Mohamad Haj Yahia 
Signed-off-by: Saeed Mahameed 
Signed-off-by: David S. Miller

net/mlx5: SRIOV core code refactoring

2016-09-11T04:21:49+00:00

Simplify the code and makes it look modular and symmetric.
Split sriov enable/disable to two levels: device level and pci level.
When user enable/disable sriov (via sriov_configure driver callback) we
will enable/disable both device and pci sriov.
When driver load/unload we will enable/disable (on demand) only device
sriov while keeping the PCI sriov enabled for next driver load.
On internal/pci error, VFs will be kept enabled on PCI and the reset
is done only in device level.

Signed-off-by: Mohamad Haj Yahia 
Signed-off-by: Saeed Mahameed 
Signed-off-by: David S. Miller

net/mlx5: Vport LAG creation support

2016-08-18T15:49:57+00:00

Add interfaces for issuing CREATE_VPORT_LAG and
DESTROY_VPORT_LAG commands.

Used for receiving PF1's eth traffic on PF0's
root ft.

Signed-off-by: Aviv Heller 
Signed-off-by: Saeed Mahameed 
Signed-off-by: Leon Romanovsky

net/mlx5: Get RoCE netdev

2016-08-18T15:49:54+00:00

Used by IB driver for determining the IB bond
device's netdev, when LAG is active.

Returns PF0's netdev if mode is not active-backup,
or the PF netdev of the active slave when mode is
active-backup.

Signed-off-by: Aviv Heller 
Signed-off-by: Saeed Mahameed 
Signed-off-by: Leon Romanovsky

net/mlx5: Implement RoCE LAG feature

2016-08-18T15:49:54+00:00

Available on dual port cards only, this feature keeps
track, using netdev LAG events, of the bonding
and link status of each port's PF netdev.

When both of the card's PF netdevs are enslaved to the
same bond/team master, and only them, LAG state
is active.

During LAG, only one IB device is present for both ports.

In addition to the above, this commit includes FW commands
used for managing the LAG, new facilities for adding and removing
a single device by interface, and port remap functionality according to
bond events.

Please note that this feature is currently used only for mimicking
Ethernet bonding for RoCE - netdevs functionality is not altered,
and their bonding continues to be managed solely by bond/team driver.

Signed-off-by: Aviv Heller 
Signed-off-by: Saeed Mahameed 
Signed-off-by: Leon Romanovsky

net/mlx5: Modify RQ bitmask from mlx5 ifc

2016-08-18T15:49:08+00:00

Use mlx5 ifc MODIFY_BITMASK_VSD in mlx5e_modify_rq_vsd and expose counter
set capability bit in hca caps structure.

Signed-off-by: Alex Vesker 
Signed-off-by: Saeed Mahameed 
Signed-off-by: Leon Romanovsky

net/mlx5: Unify and improve command interface

2016-08-17T14:45:58+00:00

Now as all commands use mlx5 ifc interface, instead of doing two calls
for executing a command we embed command status checking into
mlx5_cmd_exec to simplify the interface.

Also we do here some cleanup for redundant software structures
(inbox/outbox) and functions and improved command failure output.

Signed-off-by: Saeed Mahameed 
Signed-off-by: Leon Romanovsky