On Sun, 2018-09-16 at 20:43 +0300, Leon Romanovsky wrote:
> From: Leon Romanovsky <leonro@xxxxxxxxxxxx>
>
> From Jason,
>
> There is too much duplicated code here, centralized it all into the core
> code and provide a much better implementation.
>
> This comes about because Linus remarked that he did not like the manipulation
> of vma_ops by the drivers, this new code changes that approach and corrects
> some locking problems.
>
> Thanks
>
> Jason Gunthorpe (5):
> RDMA/ucontext: Add a core API for mmaping driver IO memory
> RDMA/mlx4: Use rdma_user_mmap_io
> RDMA/mlx5: Use rdma_user_mmap_io
> RDMA/hns: Use rdma_user_mmap_io
> RDMA/ucontext: Get rid of the old disassociate flow
>
> drivers/infiniband/core/rdma_core.c | 55 ++-----
> drivers/infiniband/core/rdma_core.h | 1 +
> drivers/infiniband/core/uverbs.h | 3 +
> drivers/infiniband/core/uverbs_main.c | 223 ++++++++++++++++++++++++++++
> drivers/infiniband/hw/hns/hns_roce_device.h | 8 -
> drivers/infiniband/hw/hns/hns_roce_main.c | 101 +++----------
> drivers/infiniband/hw/mlx4/main.c | 142 +++---------------
> drivers/infiniband/hw/mlx4/mlx4_ib.h | 5 -
> drivers/infiniband/hw/mlx5/main.c | 122 +--------------
> drivers/infiniband/hw/mlx5/mlx5_ib.h | 10 --
> include/rdma/ib_verbs.h | 22 +++
> 11 files changed, 314 insertions(+), 378 deletions(-)
>
> --
> 2.14.4
>
Series looks good. I made two very minor edits (added a comma in a
commit message for grammatical correctness, and you left out the word
zero where you meant to write "zero page" in a comment block in patch
#1). Applied to for-next.
--
Doug Ledford <dledford@xxxxxxxxxx>
GPG KeyID: B826A3330E572FDD
Key fingerprint = AE6B 1BDA 122B 23B4 265B 1274 B826 A333 0E57 2FDD
Attachment:
signature.asc
Description: This is a digitally signed message part
