Currently the mentioned helper can end-up freeing the socket wmem without waking-up any processes waiting for more write memory.
If the partially orphaned skb is attached to an UDP (or raw) socket, the lack of wake-up can hang the user-space. Even for TCP sockets not calling the sk destructor could have bad effects on TSQ. Address the issue using skb_orphan to release the sk wmem before setting the new sock_efree destructor. Additionally bundle the whole ownership update in a new helper, so that later other potential users could avoid duplicate code. v1 -> v2: - use skb_orphan() instead of sort of open coding it (Eric) - provide an helper for the ownership change (Eric) Fixes: f6ba8d33cfbb ("netem: fix skb_orphan_partial()") Suggested-by: Eric Dumazet <eduma...@google.com> Signed-off-by: Paolo Abeni <pab...@redhat.com> --- include/net/sock.h | 9 +++++++++ net/core/sock.c | 12 +++--------- 2 files changed, 12 insertions(+), 9 deletions(-) diff --git a/include/net/sock.h b/include/net/sock.h index 636810ddcd9ba..4a1c3f53630b3 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2221,6 +2221,15 @@ static inline void skb_set_owner_r(struct sk_buff *skb, struct sock *sk) sk_mem_charge(sk, skb->truesize); } +static inline void skb_set_owner_sk_safe(struct sk_buff *skb, struct sock *sk) +{ + if (sk && refcount_inc_not_zero(&sk->sk_refcnt)) { + skb_orphan(skb); + skb->destructor = sock_efree; + skb->sk = sk; + } +} + void sk_reset_timer(struct sock *sk, struct timer_list *timer, unsigned long expires); diff --git a/net/core/sock.c b/net/core/sock.c index 0ed98f20448a2..7092e3b3076ee 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2132,16 +2132,10 @@ void skb_orphan_partial(struct sk_buff *skb) if (skb_is_tcp_pure_ack(skb)) return; - if (can_skb_orphan_partial(skb)) { - struct sock *sk = skb->sk; - - if (refcount_inc_not_zero(&sk->sk_refcnt)) { - WARN_ON(refcount_sub_and_test(skb->truesize, &sk->sk_wmem_alloc)); - skb->destructor = sock_efree; - } - } else { + if (can_skb_orphan_partial(skb)) + skb_set_owner_sk_safe(skb, skb->sk); + else skb_orphan(skb); - } } EXPORT_SYMBOL(skb_orphan_partial); -- 2.26.2