On Tue, Jul 24, 2018 at 10:09:27AM +0300, Leon Romanovsky wrote:
> diff --git a/include/rdma/ib_cache.h b/include/rdma/ib_cache.h
> index 1108d4220276..511ff6be503a 100644
> +++ b/include/rdma/ib_cache.h
> @@ -128,8 +128,23 @@ int ib_get_cached_port_state(struct ib_device *device,
> enum ib_port_state *port_active);
>
> bool rdma_is_zero_gid(const union ib_gid *gid);
> -const struct ib_gid_attr *rdma_get_gid_attr(struct ib_device *device,
> - u8 port_num, int index);
> +
> +const struct ib_gid_attr *_rdma_get_gid_attr(struct ib_device *device,
> + u8 port_num, int index,
> + struct net *net);
> +static inline const struct ib_gid_attr *
> +rdma_get_gid_attr(struct ib_device *device, u8 port_num, int index)
> +{
> + return _rdma_get_gid_attr(device, port_num, index, NULL);
> +}
I don't think supporting NULL makes much sense here. ie the ability
for NULL to search all namespaces is just ugly/potential security hole
upon misuse.
It should be init_net and this wrapper shouldn't exist - all callers
of the API are broken for namespaces and should be fixed. Prefer to
push the init_net out to them.
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
