From: Leon Romanovsky <[email protected]>
Over the years, the main.c file grew above all imagination and was >8K
LOC of the code. This caused to a huge burden while I started to work on
ib_flow allocation patches.
This series implements long standing "internal" wish to move flow logic
from the main to separate file.
Based on
https://lore.kernel.org/linux-rdma/[email protected]
Thanks
Leon Romanovsky (6):
RDMA/mlx5: Limit the scope of mlx5_ib_enable_driver function
RDMA/mlx5: Separate restrack callbacks initialization from main.c
RDMA/mlx5: Separate counters from main.c
RDMA/mlx5: Separate flow steering logic from main.c
RDMA/mlx5: Cleanup DEVX initialization flow
RDMA/mlx5: Delete one-time used functions
drivers/infiniband/hw/mlx5/Makefile | 3 +-
drivers/infiniband/hw/mlx5/cmd.c | 12 -
drivers/infiniband/hw/mlx5/cmd.h | 1 -
drivers/infiniband/hw/mlx5/counters.c | 709 +++++
drivers/infiniband/hw/mlx5/counters.h | 17 +
drivers/infiniband/hw/mlx5/devx.c | 102 +-
drivers/infiniband/hw/mlx5/devx.h | 45 +
drivers/infiniband/hw/mlx5/flow.c | 765 -----
drivers/infiniband/hw/mlx5/fs.c | 2514 +++++++++++++++
drivers/infiniband/hw/mlx5/fs.h | 29 +
drivers/infiniband/hw/mlx5/main.c | 4112 +++++--------------------
drivers/infiniband/hw/mlx5/mlx5_ib.h | 76 +-
drivers/infiniband/hw/mlx5/qp.c | 1 +
drivers/infiniband/hw/mlx5/qp.h | 1 +
drivers/infiniband/hw/mlx5/restrack.c | 29 +-
drivers/infiniband/hw/mlx5/restrack.h | 13 +
16 files changed, 4184 insertions(+), 4245 deletions(-)
create mode 100644 drivers/infiniband/hw/mlx5/counters.c
create mode 100644 drivers/infiniband/hw/mlx5/counters.h
create mode 100644 drivers/infiniband/hw/mlx5/devx.h
delete mode 100644 drivers/infiniband/hw/mlx5/flow.c
create mode 100644 drivers/infiniband/hw/mlx5/fs.c
create mode 100644 drivers/infiniband/hw/mlx5/fs.h
create mode 100644 drivers/infiniband/hw/mlx5/restrack.h
--
2.26.2
On Thu, Jul 02, 2020 at 11:18:03AM +0300, Leon Romanovsky wrote:
> From: Leon Romanovsky <[email protected]>
>
> Over the years, the main.c file grew above all imagination and was >8K
> LOC of the code. This caused to a huge burden while I started to work on
> ib_flow allocation patches.
>
> This series implements long standing "internal" wish to move flow logic
> from the main to separate file.
>
> Based on
> https://lore.kernel.org/linux-rdma/[email protected]
Isn't this the series you said to drop? Can this be applied
independently?
Jason
On Thu, Jul 02, 2020 at 10:08:09AM -0300, Jason Gunthorpe wrote:
> On Thu, Jul 02, 2020 at 11:18:03AM +0300, Leon Romanovsky wrote:
> > From: Leon Romanovsky <[email protected]>
> >
> > Over the years, the main.c file grew above all imagination and was >8K
> > LOC of the code. This caused to a huge burden while I started to work on
> > ib_flow allocation patches.
> >
> > This series implements long standing "internal" wish to move flow logic
> > from the main to separate file.
> >
> > Based on
> > https://lore.kernel.org/linux-rdma/[email protected]
>
> Isn't this the series you said to drop? Can this be applied
> independently?
I asked to drop one patch in question, the one that revealed issue with
reference counting and convoluted error unwind flow.
https://lore.kernel.org/lkml/20200630145926.GA4837@unreal
It probably can be applied independently, but I didn't try.
Thanks
>
> Jason
On Thu, Jul 02, 2020 at 11:18:03AM +0300, Leon Romanovsky wrote:
> From: Leon Romanovsky <[email protected]>
>
> Over the years, the main.c file grew above all imagination and was >8K
> LOC of the code. This caused to a huge burden while I started to work on
> ib_flow allocation patches.
>
> This series implements long standing "internal" wish to move flow logic
> from the main to separate file.
>
> Based on
> https://lore.kernel.org/linux-rdma/[email protected]
>
> Thanks
>
> Leon Romanovsky (6):
> RDMA/mlx5: Limit the scope of mlx5_ib_enable_driver function
> RDMA/mlx5: Separate restrack callbacks initialization from main.c
> RDMA/mlx5: Separate counters from main.c
> RDMA/mlx5: Separate flow steering logic from main.c
> RDMA/mlx5: Cleanup DEVX initialization flow
> RDMA/mlx5: Delete one-time used functions
Applied to for-next, thanks
Jason