net: sched: introduce block mechanism to handle netif_keep_dst calls

Couple of classifiers call netif_keep_dst directly on q->dev. That is
not possible to do directly for shared blocke where multiple qdiscs are
owning the block. So introduce a infrastructure to keep track of the
block owners in list and use this list to implement block variant of
netif_keep_dst.

Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/pkt_cls.h

@@ -39,6 +39,7 @@ bool tcf_queue_work(struct work_struct *work);
 struct tcf_chain *tcf_chain_get(struct tcf_block *block, u32 chain_index,
 				bool create);
 void tcf_chain_put(struct tcf_chain *chain);
+void tcf_block_netif_keep_dst(struct tcf_block *block);
 int tcf_block_get(struct tcf_block **p_block,
 		  struct tcf_proto __rcu **p_filter_chain, struct Qdisc *q,
 		  struct netlink_ext_ack *extack);

include/net/sch_generic.h

@@ -289,6 +289,8 @@ struct tcf_block {
 	struct net *net;
 	struct Qdisc *q;
 	struct list_head cb_list;
+	struct list_head owner_list;
+	bool keep_dst;
 };
 
 static inline void qdisc_cb_private_validate(const struct sk_buff *skb, int sz)

net/sched/cls_api.c

@@ -375,6 +375,7 @@ static struct tcf_block *tcf_block_create(struct net *net, struct Qdisc *q,
 	}
 	INIT_LIST_HEAD(&block->chain_list);
 	INIT_LIST_HEAD(&block->cb_list);
+	INIT_LIST_HEAD(&block->owner_list);
 
 	/* Create chain 0 by default, it has to be always present. */
 	chain = tcf_chain_create(block, 0);
@@ -406,6 +407,65 @@ static struct tcf_chain *tcf_block_chain_zero(struct tcf_block *block)
 	return list_first_entry(&block->chain_list, struct tcf_chain, list);
 }
 
+struct tcf_block_owner_item {
+	struct list_head list;
+	struct Qdisc *q;
+	enum tcf_block_binder_type binder_type;
+};
+
+static void
+tcf_block_owner_netif_keep_dst(struct tcf_block *block,
+			       struct Qdisc *q,
+			       enum tcf_block_binder_type binder_type)
+{
+	if (block->keep_dst &&
+	    binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS &&
+	    binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS)
+		netif_keep_dst(qdisc_dev(q));
+}
+
+void tcf_block_netif_keep_dst(struct tcf_block *block)
+{
+	struct tcf_block_owner_item *item;
+
+	block->keep_dst = true;
+	list_for_each_entry(item, &block->owner_list, list)
+		tcf_block_owner_netif_keep_dst(block, item->q,
+					       item->binder_type);
+}
+EXPORT_SYMBOL(tcf_block_netif_keep_dst);
+
+static int tcf_block_owner_add(struct tcf_block *block,
+			       struct Qdisc *q,
+			       enum tcf_block_binder_type binder_type)
+{
+	struct tcf_block_owner_item *item;
+
+	item = kmalloc(sizeof(*item), GFP_KERNEL);
+	if (!item)
+		return -ENOMEM;
+	item->q = q;
+	item->binder_type = binder_type;
+	list_add(&item->list, &block->owner_list);
+	return 0;
+}
+
+static void tcf_block_owner_del(struct tcf_block *block,
+				struct Qdisc *q,
+				enum tcf_block_binder_type binder_type)
+{
+	struct tcf_block_owner_item *item;
+
+	list_for_each_entry(item, &block->owner_list, list) {
+		if (item->q == q && item->binder_type == binder_type) {
+			list_del(&item->list);
+			kfree(item);
+			return;
+		}
+	}
+	WARN_ON(1);
+}
+
 int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q,
 		      struct tcf_block_ext_info *ei,
 		      struct netlink_ext_ack *extack)
@@ -435,6 +495,12 @@ int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q,
 		}
 	}
 
+	err = tcf_block_owner_add(block, q, ei->binder_type);
+	if (err)
+		goto err_block_owner_add;
+
+	tcf_block_owner_netif_keep_dst(block, q, ei->binder_type);
+
 	err = tcf_chain_head_change_cb_add(tcf_block_chain_zero(block),
 					   ei, extack);
 	if (err)
@@ -444,6 +510,8 @@ int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q,
 	return 0;
 
 err_chain_head_change_cb_add:
+	tcf_block_owner_del(block, q, ei->binder_type);
+err_block_owner_add:
 	if (created) {
 		if (tcf_block_shared(block))
 			tcf_block_remove(block, net);
@@ -489,6 +557,7 @@ void tcf_block_put_ext(struct tcf_block *block, struct Qdisc *q,
 	if (!block)
 		return;
 	tcf_chain_head_change_cb_del(tcf_block_chain_zero(block), ei);
+	tcf_block_owner_del(block, q, ei->binder_type);
 
 	if (--block->refcnt == 0) {
 		if (tcf_block_shared(block))

net/sched/cls_bpf.c

@@ -392,8 +392,8 @@ static int cls_bpf_prog_from_efd(struct nlattr **tb, struct cls_bpf_prog *prog,
 	prog->bpf_name = name;
 	prog->filter = fp;
 
-	if (fp->dst_needed && !(tp->q->flags & TCQ_F_INGRESS))
-		netif_keep_dst(qdisc_dev(tp->q));
+	if (fp->dst_needed)
+		tcf_block_netif_keep_dst(tp->chain->block);
 
 	return 0;
 }

net/sched/cls_flow.c

@@ -526,7 +526,7 @@ static int flow_change(struct net *net, struct sk_buff *in_skb,
 
 	timer_setup(&fnew->perturb_timer, flow_perturbation, TIMER_DEFERRABLE);
 
-	netif_keep_dst(qdisc_dev(tp->q));
+	tcf_block_netif_keep_dst(tp->chain->block);
 
 	if (tb[TCA_FLOW_KEYS]) {
 		fnew->keymask = keymask;

net/sched/cls_route.c

@@ -527,7 +527,7 @@ static int route4_change(struct net *net, struct sk_buff *in_skb,
 		if (f->handle < f1->handle)
 			break;
 
-	netif_keep_dst(qdisc_dev(tp->q));
+	tcf_block_netif_keep_dst(tp->chain->block);
 	rcu_assign_pointer(f->next, f1);
 	rcu_assign_pointer(*fp, f);