mptcp: deliver ssk errors to msk

Currently all errors received on msk subflows are ignored.
We need to catch at least the errors on connect() and
on fallback sockets.

Use a custom sk_error_report callback at subflow level,
and do the real action under the msk socket lock - via
the usual sock_owned_by_user()/release_callback() schema.

Fixes: 6e628cd3a8 ("mptcp: use mptcp release_cb for delayed tasks")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Paolo Abeni 2021-02-11 15:30:37 -08:00 committed by David S. Miller
parent 4c0d2e96ba
commit 15cc104533
3 changed files with 54 additions and 0 deletions

View File

@ -2959,6 +2959,8 @@ static void mptcp_release_cb(struct sock *sk)
mptcp_push_pending(sk, 0);
spin_lock_bh(&sk->sk_lock.slock);
}
if (test_and_clear_bit(MPTCP_ERROR_REPORT, &mptcp_sk(sk)->flags))
__mptcp_error_report(sk);
/* clear any wmem reservation and errors */
__mptcp_update_wmem(sk);
@ -3355,6 +3357,11 @@ static __poll_t mptcp_poll(struct file *file, struct socket *sock,
if (sk->sk_shutdown & RCV_SHUTDOWN)
mask |= EPOLLIN | EPOLLRDNORM | EPOLLRDHUP;
/* This barrier is coupled with smp_wmb() in tcp_reset() */
smp_rmb();
if (sk->sk_err)
mask |= EPOLLERR;
return mask;
}

View File

@ -95,6 +95,7 @@
#define MPTCP_WORK_CLOSE_SUBFLOW 5
#define MPTCP_PUSH_PENDING 6
#define MPTCP_CLEAN_UNA 7
#define MPTCP_ERROR_REPORT 8
static inline bool before64(__u64 seq1, __u64 seq2)
{
@ -414,6 +415,7 @@ struct mptcp_subflow_context {
void (*tcp_data_ready)(struct sock *sk);
void (*tcp_state_change)(struct sock *sk);
void (*tcp_write_space)(struct sock *sk);
void (*tcp_error_report)(struct sock *sk);
struct rcu_head rcu;
};
@ -478,6 +480,7 @@ static inline void mptcp_subflow_tcp_fallback(struct sock *sk,
sk->sk_data_ready = ctx->tcp_data_ready;
sk->sk_state_change = ctx->tcp_state_change;
sk->sk_write_space = ctx->tcp_write_space;
sk->sk_error_report = ctx->tcp_error_report;
inet_csk(sk)->icsk_af_ops = ctx->icsk_af_ops;
}
@ -505,6 +508,7 @@ bool mptcp_finish_join(struct sock *sk);
bool mptcp_schedule_work(struct sock *sk);
void __mptcp_check_push(struct sock *sk, struct sock *ssk);
void __mptcp_data_acked(struct sock *sk);
void __mptcp_error_report(struct sock *sk);
void mptcp_subflow_eof(struct sock *sk);
bool mptcp_update_rcv_data_fin(struct mptcp_sock *msk, u64 data_fin_seq, bool use_64bit);
void __mptcp_flush_join_list(struct mptcp_sock *msk);

View File

@ -1043,6 +1043,46 @@ static void subflow_write_space(struct sock *ssk)
/* we take action in __mptcp_clean_una() */
}
void __mptcp_error_report(struct sock *sk)
{
struct mptcp_subflow_context *subflow;
struct mptcp_sock *msk = mptcp_sk(sk);
mptcp_for_each_subflow(msk, subflow) {
struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
int err = sock_error(ssk);
if (!err)
continue;
/* only propagate errors on fallen-back sockets or
* on MPC connect
*/
if (sk->sk_state != TCP_SYN_SENT && !__mptcp_check_fallback(msk))
continue;
inet_sk_state_store(sk, inet_sk_state_load(ssk));
sk->sk_err = -err;
/* This barrier is coupled with smp_rmb() in mptcp_poll() */
smp_wmb();
sk->sk_error_report(sk);
break;
}
}
static void subflow_error_report(struct sock *ssk)
{
struct sock *sk = mptcp_subflow_ctx(ssk)->conn;
mptcp_data_lock(sk);
if (!sock_owned_by_user(sk))
__mptcp_error_report(sk);
else
set_bit(MPTCP_ERROR_REPORT, &mptcp_sk(sk)->flags);
mptcp_data_unlock(sk);
}
static struct inet_connection_sock_af_ops *
subflow_default_af_ops(struct sock *sk)
{
@ -1352,9 +1392,11 @@ static int subflow_ulp_init(struct sock *sk)
ctx->tcp_data_ready = sk->sk_data_ready;
ctx->tcp_state_change = sk->sk_state_change;
ctx->tcp_write_space = sk->sk_write_space;
ctx->tcp_error_report = sk->sk_error_report;
sk->sk_data_ready = subflow_data_ready;
sk->sk_write_space = subflow_write_space;
sk->sk_state_change = subflow_state_change;
sk->sk_error_report = subflow_error_report;
out:
return err;
}
@ -1407,6 +1449,7 @@ static void subflow_ulp_clone(const struct request_sock *req,
new_ctx->tcp_data_ready = old_ctx->tcp_data_ready;
new_ctx->tcp_state_change = old_ctx->tcp_state_change;
new_ctx->tcp_write_space = old_ctx->tcp_write_space;
new_ctx->tcp_error_report = old_ctx->tcp_error_report;
new_ctx->rel_write_seq = 1;
new_ctx->tcp_sock = newsk;