bpf: Retire the struct_ops map kvalue->refcnt.

We have replaced kvalue-refcnt with synchronize_rcu() to wait for an
RCU grace period.

Maintenance of kvalue->refcnt was a complicated task, as we had to
simultaneously keep track of two reference counts: one for the
reference count of bpf_map. When the kvalue->refcnt reaches zero, we
also have to reduce the reference count on bpf_map - yet these steps
are not performed in an atomic manner and require us to be vigilant
when managing them. By eliminating kvalue->refcnt, we can make our
maintenance more straightforward as the refcount of bpf_map is now
solely managed!

To prevent the trampoline image of a struct_ops from being released
while it is still in use, we wait for an RCU grace period. The
setsockopt(TCP_CONGESTION, "...") command allows you to change your
socket's congestion control algorithm and can result in releasing the
old struct_ops implementation. It is fine. However, this function is
exposed through bpf_setsockopt(), it may be accessed by BPF programs
as well. To ensure that the trampoline image belonging to struct_op
can be safely called while its method is in use, the trampoline
safeguarde the BPF program with rcu_read_lock(). Doing so prevents any
destruction of the associated images before returning from a
trampoline and requires us to wait for an RCU grace period.

Signed-off-by: Kui-Feng Lee <kuifeng@meta.com>
Link: https://lore.kernel.org/r/20230323032405.3735486-2-kuifeng@meta.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>

include/linux/bpf.h

@@ -1945,6 +1945,7 @@ struct bpf_map *bpf_map_get_with_uref(u32 ufd);
 struct bpf_map *__bpf_map_get(struct fd f);
 void bpf_map_inc(struct bpf_map *map);
 void bpf_map_inc_with_uref(struct bpf_map *map);
+struct bpf_map *__bpf_map_inc_not_zero(struct bpf_map *map, bool uref);
 struct bpf_map * __must_check bpf_map_inc_not_zero(struct bpf_map *map);
 void bpf_map_put_with_uref(struct bpf_map *map);
 void bpf_map_put(struct bpf_map *map);

kernel/bpf/bpf_struct_ops.c

@@ -11,6 +11,7 @@
 #include <linux/refcount.h>
 #include <linux/mutex.h>
 #include <linux/btf_ids.h>
+#include <linux/rcupdate_wait.h>
 
 enum bpf_struct_ops_state {
 	BPF_STRUCT_OPS_STATE_INIT,
@@ -249,6 +250,7 @@ int bpf_struct_ops_map_sys_lookup_elem(struct bpf_map *map, void *key,
 	struct bpf_struct_ops_map *st_map = (struct bpf_struct_ops_map *)map;
 	struct bpf_struct_ops_value *uvalue, *kvalue;
 	enum bpf_struct_ops_state state;
+	s64 refcnt;
 
 	if (unlikely(*(u32 *)key != 0))
 		return -ENOENT;
@@ -267,7 +269,14 @@ int bpf_struct_ops_map_sys_lookup_elem(struct bpf_map *map, void *key,
 	uvalue = value;
 	memcpy(uvalue, st_map->uvalue, map->value_size);
 	uvalue->state = state;
-	refcount_set(&uvalue->refcnt, refcount_read(&kvalue->refcnt));
+
+	/* This value offers the user space a general estimate of how
+	 * many sockets are still utilizing this struct_ops for TCP
+	 * congestion control. The number might not be exact, but it
+	 * should sufficiently meet our present goals.
+	 */
+	refcnt = atomic64_read(&map->refcnt) - atomic64_read(&map->usercnt);
+	refcount_set(&uvalue->refcnt, max_t(s64, refcnt, 0));
 
 	return 0;
 }
@@ -491,7 +500,6 @@ static long bpf_struct_ops_map_update_elem(struct bpf_map *map, void *key,
 		*(unsigned long *)(udata + moff) = prog->aux->id;
 	}
 
-	refcount_set(&kvalue->refcnt, 1);
 	bpf_map_inc(map);
 
 	set_memory_rox((long)st_map->image, 1);
@@ -536,7 +544,6 @@ static long bpf_struct_ops_map_delete_elem(struct bpf_map *map, void *key)
 	switch (prev_state) {
 	case BPF_STRUCT_OPS_STATE_INUSE:
 		st_map->st_ops->unreg(&st_map->kvalue.data);
-		if (refcount_dec_and_test(&st_map->kvalue.refcnt))
 		bpf_map_put(map);
 		return 0;
 	case BPF_STRUCT_OPS_STATE_TOBEFREE:
@@ -570,7 +577,7 @@ static void bpf_struct_ops_map_seq_show_elem(struct bpf_map *map, void *key,
 	kfree(value);
 }
 
-static void bpf_struct_ops_map_free(struct bpf_map *map)
+static void __bpf_struct_ops_map_free(struct bpf_map *map)
 {
 	struct bpf_struct_ops_map *st_map = (struct bpf_struct_ops_map *)map;
 
@@ -582,6 +589,28 @@ static void bpf_struct_ops_map_free(struct bpf_map *map)
 	bpf_map_area_free(st_map);
 }
 
+static void bpf_struct_ops_map_free(struct bpf_map *map)
+{
+	/* The struct_ops's function may switch to another struct_ops.
+	 *
+	 * For example, bpf_tcp_cc_x->init() may switch to
+	 * another tcp_cc_y by calling
+	 * setsockopt(TCP_CONGESTION, "tcp_cc_y").
+	 * During the switch,  bpf_struct_ops_put(tcp_cc_x) is called
+	 * and its refcount may reach 0 which then free its
+	 * trampoline image while tcp_cc_x is still running.
+	 *
+	 * A vanilla rcu gp is to wait for all bpf-tcp-cc prog
+	 * to finish. bpf-tcp-cc prog is non sleepable.
+	 * A rcu_tasks gp is to wait for the last few insn
+	 * in the tramopline image to finish before releasing
+	 * the trampoline image.
+	 */
+	synchronize_rcu_mult(call_rcu, call_rcu_tasks);
+
+	__bpf_struct_ops_map_free(map);
+}
+
 static int bpf_struct_ops_map_alloc_check(union bpf_attr *attr)
 {
 	if (attr->key_size != sizeof(unsigned int) || attr->max_entries != 1 ||
@@ -630,7 +659,7 @@ static struct bpf_map *bpf_struct_ops_map_alloc(union bpf_attr *attr)
 				   NUMA_NO_NODE);
 	st_map->image = bpf_jit_alloc_exec(PAGE_SIZE);
 	if (!st_map->uvalue || !st_map->links || !st_map->image) {
-		bpf_struct_ops_map_free(map);
+		__bpf_struct_ops_map_free(map);
 		return ERR_PTR(-ENOMEM);
 	}
 
@@ -676,41 +705,23 @@ const struct bpf_map_ops bpf_struct_ops_map_ops = {
 bool bpf_struct_ops_get(const void *kdata)
 {
 	struct bpf_struct_ops_value *kvalue;
+	struct bpf_struct_ops_map *st_map;
+	struct bpf_map *map;
 
 	kvalue = container_of(kdata, struct bpf_struct_ops_value, data);
+	st_map = container_of(kvalue, struct bpf_struct_ops_map, kvalue);
 
-	return refcount_inc_not_zero(&kvalue->refcnt);
-}
-
-static void bpf_struct_ops_put_rcu(struct rcu_head *head)
-{
-	struct bpf_struct_ops_map *st_map;
-
-	st_map = container_of(head, struct bpf_struct_ops_map, rcu);
-	bpf_map_put(&st_map->map);
+	map = __bpf_map_inc_not_zero(&st_map->map, false);
+	return !IS_ERR(map);
 }
 
 void bpf_struct_ops_put(const void *kdata)
 {
 	struct bpf_struct_ops_value *kvalue;
-
-	kvalue = container_of(kdata, struct bpf_struct_ops_value, data);
-	if (refcount_dec_and_test(&kvalue->refcnt)) {
 	struct bpf_struct_ops_map *st_map;
 
-		st_map = container_of(kvalue, struct bpf_struct_ops_map,
-				      kvalue);
-		/* The struct_ops's function may switch to another struct_ops.
-		 *
-		 * For example, bpf_tcp_cc_x->init() may switch to
-		 * another tcp_cc_y by calling
-		 * setsockopt(TCP_CONGESTION, "tcp_cc_y").
-		 * During the switch,  bpf_struct_ops_put(tcp_cc_x) is called
-		 * and its map->refcnt may reach 0 which then free its
-		 * trampoline image while tcp_cc_x is still running.
-		 *
-		 * Thus, a rcu grace period is needed here.
-		 */
-		call_rcu(&st_map->rcu, bpf_struct_ops_put_rcu);
-	}
+	kvalue = container_of(kdata, struct bpf_struct_ops_value, data);
+	st_map = container_of(kvalue, struct bpf_struct_ops_map, kvalue);
+
+	bpf_map_put(&st_map->map);
 }

kernel/bpf/syscall.c

@@ -1303,8 +1303,10 @@ struct bpf_map *bpf_map_get_with_uref(u32 ufd)
 	return map;
 }
 
-/* map_idr_lock should have been held */
-static struct bpf_map *__bpf_map_inc_not_zero(struct bpf_map *map, bool uref)
+/* map_idr_lock should have been held or the map should have been
+ * protected by rcu read lock.
+ */
+struct bpf_map *__bpf_map_inc_not_zero(struct bpf_map *map, bool uref)
 {
 	int refold;