On Wed, Mar 29, 2023 at 09:44:09PM +0300, Dan Carpenter wrote:
> On Wed, Mar 29, 2023 at 09:14:01PM +0300, Leon Romanovsky wrote:
> > @@ -1287,11 +1279,8 @@ static struct ib_mr *rxe_alloc_mr(struct ib_pd *ibpd, enum ib_mr_type mr_type,
> > }
> >
> > mr = kzalloc(sizeof(*mr), GFP_KERNEL);
> > - if (!mr) {
> > - err = -ENOMEM;
> > - rxe_dbg_mr(mr, "no memory for mr");
> > - goto err_out;
> > - }
> > + if (!mr)
> > + return ERR_PTR(-ENOMEM);
> >
> > err = rxe_add_to_pool(&rxe->mr_pool, mr);
> > if (err) {
> ^^^^^^^^^^
> If the rxe_add_to_pool() fails then it calls:
>
> rxe_dbg_mr(mr, "unable to create mr, err = %d", err);
> ^^
>
> "rm" is zeroed mem and not useful at this point. Possibly in a later
> patch though.
I will delete that line when will apply the patch.
Thanks
>
> regards,
> dan carpenter
>
