Hello Maor-
> On Dec 19, 2021, at 4:51 AM, Maor Gottlieb <maorg@xxxxxxxxxx> wrote:
>
>
> On 12/18/2021 11:25 PM, Chuck Lever III wrote:
>> NFS/RDMA with an NFS client using mlx5-based hardware triggers a
>> system deadlock (no error messages) on the client. I bisected to
>> f0ae4afe3d35 ("RDMA/mlx5: Fix releasing unallocated memory in
>> dereg MR flow").
>>
>> --
>> Chuck Lever
>
> Hi Chuck,
>
> I found some bug in the cited commit. Can you please test if the below patch fixes this deadlock ?
>
> diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
> index 157d862fb864..3cb4e34fe199 100644
> --- a/drivers/infiniband/hw/mlx5/mr.c
> +++ b/drivers/infiniband/hw/mlx5/mr.c
> @@ -1978,7 +1978,7 @@ int mlx5_ib_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata)
> return rc;
> }
>
> - if (mr->umem) {
> + if (udata && mr->umem) {
> bool is_odp = is_odp_mr(mr);
>
> if (!is_odp)
After applying this one-liner, I am not able to reproduce the
NFS client system deadlock.
Tested-by: Chuck Lever <chuck.lever@xxxxxxxxxx>
--
Chuck Lever
