[PATCH bpf-next] bpf, sockmap: fix deadlocks in the sockhash and sockmap

From: Xin Liu
Date: Thu Apr 06 2023 - 08:26:33 EST


When huang uses sched_switch tracepoint, the tracepoint
does only one thing in the mounted ebpf program, which
deletes the fixed elements in sockhash ([0])

It seems that elements in sockhash are rarely actively
deleted by users or ebpf program. Therefore, we do not
pay much attention to their deletion. Compared with hash
maps, sockhash only provides spin_lock_bh protection.
This causes it to appear to have self-locking behavior
in the interrupt context.

[0]:https://lore.kernel.org/all/CABcoxUayum5oOqFMMqAeWuS8+EzojquSOSyDA3J_2omY=2EeAg@xxxxxxxxxxxxxx/

Reported-by: Hsin-Wei Hung <hsinweih@xxxxxxx>
Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Xin Liu <liuxin350@xxxxxxxxxx>
---
net/core/sock_map.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 7c189c2e2fbf..66305b7bf8b7 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -414,8 +414,9 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
{
struct sock *sk;
int err = 0;
+ unsigned long flags;

- raw_spin_lock_bh(&stab->lock);
+ raw_spin_lock_irqsave(&stab->lock, flags);
sk = *psk;
if (!sk_test || sk_test == sk)
sk = xchg(psk, NULL);
@@ -425,7 +426,7 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
else
err = -EINVAL;

- raw_spin_unlock_bh(&stab->lock);
+ raw_spin_unlock_irqrestore(&stab->lock, flags);
return err;
}

@@ -932,11 +933,12 @@ static long sock_hash_delete_elem(struct bpf_map *map, void *key)
struct bpf_shtab_bucket *bucket;
struct bpf_shtab_elem *elem;
int ret = -ENOENT;
+ unsigned long flags;

hash = sock_hash_bucket_hash(key, key_size);
bucket = sock_hash_select_bucket(htab, hash);

- raw_spin_lock_bh(&bucket->lock);
+ raw_spin_lock_irqsave(&bucket->lock, flags);
elem = sock_hash_lookup_elem_raw(&bucket->head, hash, key, key_size);
if (elem) {
hlist_del_rcu(&elem->node);
@@ -944,7 +946,7 @@ 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);
+ raw_spin_unlock_irqrestore(&bucket->lock, flags);
return ret;
}

--
2.33.0