Re: [PATCH net v5] net/smc: fix lgr/lnk lifetime vs diag reader race

[Sidraya Jayagond <sidraya@linux.ibm.com>]

On 23/09/26 11:47 am, Mahanta Jambigi wrote:
> The diag dump walks the socket hash table under a read_lock and dereferences
> conn->lgr and conn->lnk. Two terminal teardown paths drop those references via
> smc_conn_free() while the socket is still hashed:
> 
>   - smc_conn_kill() -> smc_close_active_abort() -> smc_conn_free()
>   - smc_close_passive_work() -> smc_conn_free()
> 
> This allows the diag reader to dereference a freed lgr or lnk.
> 
> Fix it by unhashing the socket before smc_conn_free() is called at the two
> terminal teardown sites in smc_close.c.
> 
> Additionally, smc_conn_abort() calls smc_conn_free() during early handshake
> aborts while the socket remains hashed in SMC_INIT state. The
> smc_listen_out_err() path leaves the socket hashed in SMC_CLOSED state after
> smc_conn_abort() returns. Guard the conn/lgr/lnk inspection blocks in
> __smc_diag_dump() by skipping them when the socket is in SMC_INIT or SMC_CLOSED
> state.
> 
> Fixes: f16a7dd5cf27 ("smc: netlink interface for SMC sockets")
> Fixes: 9dbe086c69b8 ("net/smc: fix invalid link access in dumping SMC-R connections")
> Signed-off-by: Mahanta Jambigi <mjambigi@linux.ibm.com>
> ---
> Changes in v5:
> - guard conn/lgr/lnk blocks in __smc_diag_dump() with an
>   (SMC_INIT || SMC_CLOSED) state check placed after the
>   SMC_DIAG_FALLBACK nla_put; SMC_INIT can coexist with fallback
>   (smc_sendmsg() MSG_FASTOPEN path), so the guard must not suppress
>   the fallback reason
> 
> Changes in v4:
> - dropped smc_conn_unhash() wrapper, conn->unhashed flag, and all
>   changes to af_smc.c, smc.h, smc_core.c and smc_core.h;
>   smc_unhash_sk() is already idempotent via sk_hashed(), so direct
>   calls at the two teardown sites in smc_close.c are sufficient
> - dropped the __smc_release() hunk: it needs no change since the
>   subsequent unhash there is already a safe no-op
> - fixed premature-unhash issue present in v3: smc_conn_free() must not
>   unhash because smc_conn_abort() calls it before
>   smc_switch_to_fallback() in both smc_listen_decline() and
>   smc_connect_rdma() error paths; unhashing there would make live
>   fallback sockets invisible to smcss
> - likewise, the ISM/RDMA retry loops (smc_find_ism_v2_device_serv(),
>   smc_find_rdma_v2_device_serv()) call smc_conn_abort() on a failed
>   attempt and then smc_conn_create() on the next device; unhashing in
>   smc_conn_free() would permanently hide the established connection from
>   smc_diag since smc_conn_create() does not re-hash the socket
> 
> Changes in v3:
> - redesigned as a single patch; dropped the lgr_lnk_lock spinlock
>   approach and the 2-patch split
> - fix is now at the socket hash layer: introduce smc_conn_unhash() with
>   a per-connection unhashed flag; smc_conn_free() unhashes before dropping
>   lgr/lnk refs, so any socket visible to the diag reader under the hash
>   read_lock has valid conn->lgr and conn->lnk pointers
> - __smc_release() updated to call smc_conn_unhash() for non-fallback
>   sockets so the flag is honoured when smc_conn_free() already ran first
> - smc_diag.c needs no changes; the hash read_lock invariant is
>   sufficient without any per-connection lock in the dump path
> - dropped the clcsock/mutex_trylock fix as that will be addressed separately
> 
> Changes in v2:
> - this is v2 of the 2-patch series; the earlier submission was mislabelled
>   [PATCH v3] but was in fact the first version sent to the list
> - split into a 2-patch series; patch 1/2 adds per-connection lgr_lnk_lock
>   infrastructure to smc_core, patch 2/2 fixes the diag dump path using it
> - dropped lock_sock()/release_sock() from __smc_diag_dump(); v1 held the
>   socket lock across all lgr/lnk dereferences, requiring the hash read_lock
>   to be dropped and re-acquired around each socket
> - dropped the restart-from-head loop in smc_diag_dump_proto(); the new
>   design does not drop the hash read_lock mid-walk so the hlist truncation
>   concern no longer applies
> - dropped refcount_inc_not_zero() socket pinning from the dump loop for the
>   same reason: the hash read_lock is now held for the full walk
> - added per-connection lgr_lnk_lock spinlock to struct smc_connection;
>   conn->lgr and conn->lnk are NULLed under this lock in smc_conn_free()
>   before borrowed references are released, establishing the invariant: a
>   non-NULL conn->lgr seen under lgr_lnk_lock guarantees the lgr is alive
> - added lgr_lnk_lock to smc_switch_link_and_count() to protect the conn->lnk
>   pointer swap from concurrent diag readers
> - replaced mutex_lock() on clcsock_release_lock in smc_diag_msg_common_fill()
>   with mutex_trylock(); mutex_lock() was valid in v1 because the hash
>   spinlock had been dropped, but the new design holds the hash read_lock
>   throughout so only a non-sleeping trylock is safe; a failed trylock leaves
>   address fields zeroed, which is acceptable for a monitoring tool
> - all conn->lgr and conn->lnk accesses in __smc_diag_dump() use a
>   snapshot-then-use pattern: fields are copied into local stack variables
>   under lgr_lnk_lock and nla_put() is called after releasing the lock,
>   avoiding any sleeping operation under the spinlock
> 
>  net/smc/smc_close.c | 10 ++++++++++
>  net/smc/smc_diag.c  |  4 ++++
>  2 files changed, 14 insertions(+)
> 
> diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c
> index bb0313ef5f7c..3940098c651d 100644
> --- a/net/smc/smc_close.c
> +++ b/net/smc/smc_close.c
> @@ -130,6 +130,11 @@ void smc_close_active_abort(struct smc_sock *smc)
>  	struct sock *sk = &smc->sk;
>  	bool release_clcsock = false;
> 
> +	/*
> +	 * Unhash before any branch calls smc_conn_free() and drops lgr/lnk refs;
> +	 * the diag reader must not see a hashed socket with freed lgr/lnk.
> +	 */
> +	sk->sk_prot->unhash(sk);
>  	if (sk->sk_state != SMC_INIT && smc->clcsock && smc->clcsock->sk) {
>  		sk->sk_err = ECONNABORTED;
>  		if (smc->clcsock && smc->clcsock->sk)
> @@ -433,6 +438,11 @@ static void smc_close_passive_work(struct work_struct *work)
>  		sk->sk_state_change(sk);
>  		if ((sk->sk_state == SMC_CLOSED) &&
>  		    (sock_flag(sk, SOCK_DEAD) || !sk->sk_socket)) {
> +			/*
> +			 * Unhash before smc_conn_free() drops lgr/lnk refs so the diag
> +			 * reader cannot see a hashed socket with freed lgr/lnk.
> +			 */
> +			sk->sk_prot->unhash(sk);
>  			smc_conn_free(conn);
>  			if (smc->clcsock)
>  				release_clcsock = true;
> diff --git a/net/smc/smc_diag.c b/net/smc/smc_diag.c
> index bf0beaa23bdb..fb615d3af058 100644
> --- a/net/smc/smc_diag.c
> +++ b/net/smc/smc_diag.c
> @@ -103,5 +103,8 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb,
>  	if (nla_put(skb, SMC_DIAG_FALLBACK, sizeof(fallback), &fallback) < 0)
>  		goto errout;
> 
> +	if (sk->sk_state == SMC_INIT || sk->sk_state == SMC_CLOSED)
> +		goto out;
> +
>  	if ((req->diag_ext & (1 << (SMC_DIAG_CONNINFO - 1))) &&
>  	    smc->conn.alert_token_local) {
> @@ -185,5 +188,6 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb,
>  			goto errout;
>  	}
> 
> +out:
>  	nlmsg_end(skb, nlh);
>  	return 0;

Reviewed-by: Sidraya Jayagond <sidraya@linux.ibm.com>