user/sven/linux.git/include/linux/mlx5, branch v4.9.51

net/mlx5: Cancel delayed recovery work when unloading the driver

2017-07-21T05:42:18Z

commit 2a0165a034ac024b60cca49c61e46f4afa2e4d98 upstream. Draining the health workqueue will ignore future health works including the one that report hardware failure and thus we can't enter error state Instead cancel the recovery flow and make sure only recovery flow won't be scheduled. Fixes: 5e44fca50470 ('net/mlx5: Only cancel recovery work when cleaning up device') Signed-off-by: Mohamad Haj Yahia Signed-off-by: Moshe Shemesh Signed-off-by: Saeed Mahameed Signed-off-by: Greg Kroah-Hartman

net/mlx5: Avoid using pending command interface slots

2017-06-07T10:07:45Z

[ Upstream commit 73dd3a4839c1d27c36d4dcc92e1ff44225ecbeb7 ] Currently when firmware command gets stuck or it takes long time to complete, the driver command will get timeout and the command slot is freed and can be used for new commands, and if the firmware receive new command on the old busy slot its behavior is unexpected and this could be harmful. To fix this when the driver command gets timeout we return failure, but we don't free the command slot and we wait for the firmware to explicitly respond to that command. Once all the entries are busy we will stop processing new firmware commands. Fixes: 9cba4ebcf374 ('net/mlx5: Fix potential deadlock in command mode change') Signed-off-by: Mohamad Haj Yahia Cc: kernel-team@fb.com Signed-off-by: Saeed Mahameed Signed-off-by: Greg Kroah-Hartman

net/mlx5: PCI error recovery health care simulation

2016-10-29T16:00:39Z

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

2016-10-29T16:00:39Z

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.

2016-10-29T16:00:39Z

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: Add MLX5_ARRAY_SET64 to fix BUILD_BUG_ON

2016-10-13T14:13:24Z

I am hitting this in mlx5: drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c: In function reclaim_pages_cmd.clone.0: drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c:346: error: call to __compiletime_assert_346 declared with attribute error: BUILD_BUG_ON failed: __mlx5_bit_off(manage_pages_out, pas[i]) % 64 drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c: In function give_pages: drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c:291: error: call to __compiletime_assert_291 declared with attribute error: BUILD_BUG_ON failed: __mlx5_bit_off(manage_pages_in, pas[i]) % 64 Problem is that this is doing a BUILD_BUG_ON on a non-constant expression because of trying to take offset of pas[i] in the structure. Fix is to create MLX5_ARRAY_SET64 that takes an additional argument that is the field index to separate between BUILD_BUG_ON on the array constant field and the indexed field to assign the value to. There are two callers of MLX5_SET64 that are trying to get a variable offset, change those to call MLX5_ARRAY_SET64 passing 'pas' and 'i' as the arguments to use in the offset check and the indexed value assignment. Fixes: a533ed5e179cd ("net/mlx5: Pages management commands via mlx5 ifc") Signed-off-by: Tom Herbert Signed-off-by: Saeed Mahameed Signed-off-by: David S. Miller

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net

2016-09-12T22:52:44Z

Conflicts: drivers/net/ethernet/mediatek/mtk_eth_soc.c drivers/net/ethernet/qlogic/qed/qed_dcbx.c drivers/net/phy/Kconfig All conflicts were cases of overlapping commits. Signed-off-by: David S. Miller

net/mlx5: Introduce attach/detach to interface API

2016-09-11T04:21:49Z

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

2016-09-11T04:21:49Z

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/mlx5e: Move an_disable_cap bit to a new position

2016-09-08T23:15:28Z

Previous an_disable_cap position bit31 is deprecated to be use in driver with newer firmware. New firmware will advertise the same capability in bit29. Old capability didn't allow setting more than one protocol for a specific speed when autoneg is off, while newer firmware will allow this and it is indicated in the new capability location. Signed-off-by: Bodong Wang Signed-off-by: Saeed Mahameed Signed-off-by: David S. Miller