Re: [PATCH] RDMA/rxe: Remove the member 'type' of struct rxe_mr

[Leon Romanovsky <leon@kernel.org>]

On Mon, Oct 24, 2022 at 09:07:11AM -0500, Bob Pearson wrote:
> On 10/24/22 06:51, Leon Romanovsky wrote:
> > On Fri, Oct 21, 2022 at 01:45:17PM +0000, yangx.jy@fujitsu.com wrote:
> >> The member 'type' is included in both struct rxe_mr and struct ib_mr
> >> so remove the duplicate one of struct rxe_mr.
> >>
> >> Signed-off-by: Xiao Yang <yangx.jy@fujitsu.com>
> >> ---
> >>  drivers/infiniband/sw/rxe/rxe_mr.c    | 16 ++++++++--------
> >>  drivers/infiniband/sw/rxe/rxe_verbs.h |  1 -
> >>  2 files changed, 8 insertions(+), 9 deletions(-)
> > 
> > <...>
> > 
> >>  	default:
> >>  		pr_warn("%s: mr type (%d) not supported\n",
> >> -			__func__, mr->type);
> >> +			__func__, mr->ibmr.type);
> >>  		return -EFAULT;
> > 
> > <...>
> > 
> >> -	if (unlikely(mr->type != IB_MR_TYPE_MEM_REG)) {
> >> -		pr_warn("%s: mr->type (%d) is wrong type\n", __func__, mr->type);
> >> +	if (unlikely(mr->ibmr.type != IB_MR_TYPE_MEM_REG)) {
> >> +		pr_warn("%s: mr type (%d) is wrong\n", __func__, mr->ibmr.type);
> > 
> > Someone needs to convert pr_*() calls to ibdev_*() prints.
> > 
> > Thanks
> 
> I have a patch laying around that does this. Hoping some of the backlog of rxe patches gets
> accepted so I don't have to backport everything.

The way to move forward with RXE patches is to collect Reviewed-by from
other active RXE contributors.

I'm not taking RXE patches without ROB, that don't apply cleanly and which
don't follow kernel coding style both in code and in commit messages.

Thanks

> 
> Bob