Message-Id: <20230830053517.166611-1-john.fastabend@gmail.com>
Date: Tue, 29 Aug 2023 22:35:17 -0700
From: John Fastabend <john.fastabend@...il.com>
To: bpf@...r.kernel.org
Cc: john.fastabend@...il.com,
	jakub@...udflare.com,
	netdev@...r.kernel.org
Subject: [PATCH bpf] bpf: sockmap, fix preempt_rt splat when using raw_spin_lock_t

Sockmap and sockhash maps are a collection of psocks that are
objects representing a socket plus a set of metadata needed
to manage the BPF programs associated with the socket. These
maps use the stab->lock to protect from concurrent operations
on the maps, e.g. trying to insert to objects into the array
at the same time in the same slot. Additionally, a sockhash map
has a bucket lock to protect iteration and insert/delete into
the hash entry.

Each psock has a psock->link which is a linked list of all the
maps that a psock is attached to. This allows a psock (socket)
to be included in multiple sockmap and sockhash maps. This
linked list is protected the psock->link_lock.

They _must_ be nested correctly to avoid deadlock,

  lock(stab->lock)
    : do BPF map operations and psock insert/delete
    lock(psock->link_lock)
       : add map to psock linked list of maps
    unlock(psock->link_lock)
  unlock(stab->lock)

For non PREEMPT_RT kernels both raw_spin_lock_t and spin_lock_t
are guaranteed to not sleep. But, with PREEMPT_RT kernels the
spin_lock_t variants may sleep. In the current code we have
many patterns like this,

   rcu_critical_section:
      raw_spin_lock(stab->lock)
         spin_lock(psock->link_lock) <- may sleep ouch
         spin_unlock(psock->link_lock)
      raw_spin_unlock(stab->lock)
   rcu_critical_section

Nesting spin_lock() inside a raw_spin_lock() violates locking
rules for PREEMPT_RT kernels. And additionally we do alloc(GFP_ATOMICS)
inside the stab->lock, but those might sleep on PREEMPT_RT kernels.
The result is splats like this,

./test_progs -t sockmap_basic
[   33.344330] bpf_testmod: loading out-of-tree module taints kernel.
[   33.441933]
[   33.442089] =============================
[   33.442421] [ BUG: Invalid wait context ]
[   33.442763] 6.5.0-rc5-01731-gec0ded2e0282 #4958 Tainted: G           O
[   33.443320] -----------------------------
[   33.443624] test_progs/2073 is trying to lock:
[   33.443960] ffff888102a1c290 (&psock->link_lock){....}-{3:3}, at: sock_map_update_common+0x2c2/0x3d0
[   33.444636] other info that might help us debug this:
[   33.444991] context-{5:5}
[   33.445183] 3 locks held by test_progs/2073:
[   33.445498]  #0: ffff88811a208d30 (sk_lock-AF_INET){+.+.}-{0:0}, at: sock_map_update_elem_sys+0xff/0x330
[   33.446159]  #1: ffffffff842539e0 (rcu_read_lock){....}-{1:3}, at: sock_map_update_elem_sys+0xf5/0x330
[   33.446809]  #2: ffff88810d687240 (&stab->lock){+...}-{2:2}, at: sock_map_update_common+0x177/0x3d0
[   33.447445] stack backtrace:
[   33.447655] CPU: 10 PID

To fix observe we can't readily remove the allocations (for that
we would need to use/create something similar to bpf_map_alloc). So
convert raw_spin_lock_t to spin_lock_t. We note that sock_map_update
that would trigger the allocate and potential sleep is only allowed
through sys_bpf ops and via sock_ops which precludes hw interrupts
and low level atomic sections in RT preempt kernel. On non RT
preempt kernel there are no changes here and spin locks sections
and alloc(GFP_ATOMIC) are still not sleepable.

Signed-off-by: John Fastabend <john.fastabend@...il.com>
---
 net/core/sock_map.c | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 8f07fea39d9e..cb11750b1df5 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -18,7 +18,7 @@ struct bpf_stab {
 	struct bpf_map map;
 	struct sock **sks;
 	struct sk_psock_progs progs;
-	raw_spinlock_t lock;
+	spinlock_t lock;
 };
 
 #define SOCK_CREATE_FLAG_MASK				\
@@ -44,7 +44,7 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr)
 		return ERR_PTR(-ENOMEM);
 
 	bpf_map_init_from_attr(&stab->map, attr);
-	raw_spin_lock_init(&stab->lock);
+	spin_lock_init(&stab->lock);
 
 	stab->sks = bpf_map_area_alloc((u64) stab->map.max_entries *
 				       sizeof(struct sock *),
@@ -411,7 +411,7 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
 	struct sock *sk;
 	int err = 0;
 
-	raw_spin_lock_bh(&stab->lock);
+	spin_lock_bh(&stab->lock);
 	sk = *psk;
 	if (!sk_test || sk_test == sk)
 		sk = xchg(psk, NULL);
@@ -421,7 +421,7 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
 	else
 		err = -EINVAL;
 
-	raw_spin_unlock_bh(&stab->lock);
+	spin_unlock_bh(&stab->lock);
 	return err;
 }
 
@@ -487,7 +487,7 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
 	psock = sk_psock(sk);
 	WARN_ON_ONCE(!psock);
 
-	raw_spin_lock_bh(&stab->lock);
+	spin_lock_bh(&stab->lock);
 	osk = stab->sks[idx];
 	if (osk && flags == BPF_NOEXIST) {
 		ret = -EEXIST;
@@ -501,10 +501,10 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
 	stab->sks[idx] = sk;
 	if (osk)
 		sock_map_unref(osk, &stab->sks[idx]);
-	raw_spin_unlock_bh(&stab->lock);
+	spin_unlock_bh(&stab->lock);
 	return 0;
 out_unlock:
-	raw_spin_unlock_bh(&stab->lock);
+	spin_unlock_bh(&stab->lock);
 	if (psock)
 		sk_psock_put(sk, psock);
 out_free:
@@ -835,7 +835,7 @@ struct bpf_shtab_elem {
 
 struct bpf_shtab_bucket {
 	struct hlist_head head;
-	raw_spinlock_t lock;
+	spinlock_t lock;
 };
 
 struct bpf_shtab {
@@ -910,7 +910,7 @@ static void sock_hash_delete_from_link(struct bpf_map *map, struct sock *sk,
 	 * is okay since it's going away only after RCU grace period.
 	 * However, we need to check whether it's still present.
 	 */
-	raw_spin_lock_bh(&bucket->lock);
+	spin_lock_bh(&bucket->lock);
 	elem_probe = sock_hash_lookup_elem_raw(&bucket->head, elem->hash,
 					       elem->key, map->key_size);
 	if (elem_probe && elem_probe == elem) {
@@ -918,7 +918,7 @@ static void sock_hash_delete_from_link(struct bpf_map *map, struct sock *sk,
 		sock_map_unref(elem->sk, elem);
 		sock_hash_free_elem(htab, elem);
 	}
-	raw_spin_unlock_bh(&bucket->lock);
+	spin_unlock_bh(&bucket->lock);
 }
 
 static long sock_hash_delete_elem(struct bpf_map *map, void *key)
@@ -932,7 +932,7 @@ static long sock_hash_delete_elem(struct bpf_map *map, void *key)
 	hash = sock_hash_bucket_hash(key, key_size);
 	bucket = sock_hash_select_bucket(htab, hash);
 
-	raw_spin_lock_bh(&bucket->lock);
+	spin_lock_bh(&bucket->lock);
 	elem = sock_hash_lookup_elem_raw(&bucket->head, hash, key, key_size);
 	if (elem) {
 		hlist_del_rcu(&elem->node);
@@ -940,7 +940,7 @@ static long sock_hash_delete_elem(struct bpf_map *map, void *key)
 		sock_hash_free_elem(htab, elem);
 		ret = 0;
 	}
-	raw_spin_unlock_bh(&bucket->lock);
+	spin_unlock_bh(&bucket->lock);
 	return ret;
 }
 
@@ -1000,7 +1000,7 @@ static int sock_hash_update_common(struct bpf_map *map, void *key,
 	hash = sock_hash_bucket_hash(key, key_size);
 	bucket = sock_hash_select_bucket(htab, hash);
 
-	raw_spin_lock_bh(&bucket->lock);
+	spin_lock_bh(&bucket->lock);
 	elem = sock_hash_lookup_elem_raw(&bucket->head, hash, key, key_size);
 	if (elem && flags == BPF_NOEXIST) {
 		ret = -EEXIST;
@@ -1026,10 +1026,10 @@ static int sock_hash_update_common(struct bpf_map *map, void *key,
 		sock_map_unref(elem->sk, elem);
 		sock_hash_free_elem(htab, elem);
 	}
-	raw_spin_unlock_bh(&bucket->lock);
+	spin_unlock_bh(&bucket->lock);
 	return 0;
 out_unlock:
-	raw_spin_unlock_bh(&bucket->lock);
+	spin_unlock_bh(&bucket->lock);
 	sk_psock_put(sk, psock);
 out_free:
 	sk_psock_free_link(link);
@@ -1115,7 +1115,7 @@ static struct bpf_map *sock_hash_alloc(union bpf_attr *attr)
 
 	for (i = 0; i < htab->buckets_num; i++) {
 		INIT_HLIST_HEAD(&htab->buckets[i].head);
-		raw_spin_lock_init(&htab->buckets[i].lock);
+		spin_lock_init(&htab->buckets[i].lock);
 	}
 
 	return &htab->map;
@@ -1147,11 +1147,11 @@ static void sock_hash_free(struct bpf_map *map)
 		 * exists, psock exists and holds a ref to socket. That
 		 * lets us to grab a socket ref too.
 		 */
-		raw_spin_lock_bh(&bucket->lock);
+		spin_lock_bh(&bucket->lock);
 		hlist_for_each_entry(elem, &bucket->head, node)
 			sock_hold(elem->sk);
 		hlist_move_list(&bucket->head, &unlink_list);
-		raw_spin_unlock_bh(&bucket->lock);
+		spin_unlock_bh(&bucket->lock);
 
 		/* Process removed entries out of atomic context to
 		 * block for socket lock before deleting the psock's
-- 
2.33.0

Powered by blists - more mailing lists