Re: [PATCH rdma-next v1 4/4] RDMA/core: Convert RWQ table logic to ib_core allocation scheme

[Leon Romanovsky <leon@kernel.org>]

On Tue, Jun 30, 2020 at 02:31:16PM +0300, Yishai Hadas wrote:
> On 6/30/2020 1:18 PM, Leon Romanovsky wrote:
> > From: Leon Romanovsky <leonro@mellanox.com>
> >
> > Move struct ib_rwq_ind_table allocation to ib_core.
> >
> > Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
> > ---
> >   drivers/infiniband/core/device.c           |  1 +
> >   drivers/infiniband/core/uverbs_cmd.c       | 37 ++++++++++++-------
> >   drivers/infiniband/core/uverbs_std_types.c | 16 +++++---
> >   drivers/infiniband/core/verbs.c            | 23 ------------
> >   drivers/infiniband/hw/mlx4/main.c          |  4 +-
> >   drivers/infiniband/hw/mlx4/mlx4_ib.h       | 12 +++---
> >   drivers/infiniband/hw/mlx4/qp.c            | 40 ++++++--------------
> >   drivers/infiniband/hw/mlx5/main.c          |  3 ++
> >   drivers/infiniband/hw/mlx5/mlx5_ib.h       |  8 ++--
> >   drivers/infiniband/hw/mlx5/qp.c            | 43 +++++++++-------------
> >   include/rdma/ib_verbs.h                    | 11 +++---
> >   11 files changed, 85 insertions(+), 113 deletions(-)

<...>

> > +
> > +	if (rwq_ind_tbl->device->ops.destroy_rwq_ind_table)
> > +		rwq_ind_tbl->device->ops.destroy_rwq_ind_table(rwq_ind_tbl);
>
>
> We had here two notes from previous review that need to be settled, ref
> count decrement before object destruction (high priority) and considering
> the existance of both alloc/destroy functions from driver point of view from
> symetic reasons. (low priority).
>
> Let's get Jason's feedback here please.

I'm confident that Jason will give his feedback like he always does
while accepting/declining patches. It goes without saying.

From my point of view, there is nothing to change.

Thanks