iscsi_tcp: do not bind sockets that already have extra callbacks
PATCH 1/2 iscsi_tcp: do not bind sockets that already have extra callbacks
This attempts to avoid a situation where a misbehaving iscsi daemon passes a socket for a different iSCSI connection to BIND_CONN - which would result in infinite recursion and stack overflow. This will also prevent passing *other* sockets which had sk_user_data overridden, but that wouldn't have been safe anyways - since we throw away that pointer anyways. This does not cover all hypothetical scenarios where we pass bad sockets to BIND_CONN.
This also papers over a different bug - we allow a daemon to call BIND_CONN twice for the same connection - which would result in, at the least, failing to uninitialize/teardown the previous socket, which will be addressed separately.
--- drivers/scsi/iscsi_tcp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/iscsi_tcp.c +1 -1
@@ -725,7 +725,7 @@ iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session, } err = -EINVAL ; -if (!sk_is_tcp (sock->sk)) +if (!sk_is_tcp (sock->sk) || sock->sk->sk_user_data) goto free_socket; err =iscsi_conn_bind (cls_session, cls_conn, is_leading);
Signature · 1 line
2.44.0.291.gc1ea87d7ee-goog
PATCH 2/2 iscsi_tcp: disallow binding the same connection twice
iscsi_sw_tcp_conn_bind does not check or cleanup previously bound sockets, nor should we allow binding the same connection twice.
--- drivers/scsi/iscsi_tcp.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/scsi/iscsi_tcp.c +3 -0
@@ -716,6 +716,9 @@ iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session,struct socket *sock;int err; +if (tcp_sw_conn->sock) +return -EINVAL ; +/* lookup for existing socket */ sock =sockfd_lookup ((int )transport_eph, &err);if (!sock) {
Signature · 1 line
2.44.0.291.gc1ea87d7ee-goog
Mike Christie
maintainerRe: [PATCH 2/2] iscsi_tcp: disallow binding the same connection twice
Quoted 3 lines · Khazhismel Kumykov
iscsi_sw_tcp_conn_binddoes not check or cleanup previously bound sockets, nor should we allow binding the same connection twice.
This looks like a problem for all the iscsi drivers.
I think you could:
1. Add a check for ISCSI_CONN_FLAG_BOUND in iscsi_conn_bind.
2. Have iscsi_sw_tcp_conn_stop do:
/* stop xmit side */
- iscsi_suspend_tx(conn);
+ iscsi_conn_unbind(cls_conn, true);
to clear the flag when we clean up the conn for relogin.
3. Fix up the other iscsi drivers so they call:
iscsi_conn_unbind(cls_conn, true);
in their failure paths so when they fail they clear ISCSI_CONN_FLAG_BOUND and iscsi_conn_bind can be called on the retry.
Quoted 19 lines
Signed-off-by Khazhismel Kumykov <[email protected]>--- drivers/scsi/iscsi_tcp.c | 3 +++ 1 file changed, 3 insertions(+)drivers/scsi/iscsi_tcp.c +3
-0@@ -716,6 +716,9 @@ iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session,struct socket *sock;int err; +if (tcp_sw_conn->sock) +return -EINVAL ; +/* lookup for existing socket */ sock =sockfd_lookup ((int )transport_eph, &err);if (!sock) {
Re: [PATCH 2/2] iscsi_tcp: disallow binding the same connection twice
Quoted 17 lines · Mike Christie
Quoted 3 lines · Khazhismel Kumykov
iscsi_sw_tcp_conn_binddoes not check or cleanup previously bound sockets, nor should we allow binding the same connection twice.This looks like a problem for all the iscsi drivers.
I think you could:
1. Add a check for
ISCSI_CONN_FLAG_BOUNDiniscsi_conn_bind.
2. Haveiscsi_sw_tcp_conn_stopdo:/* stop xmit side */(quoted diff) +1
-1- iscsi_suspend_tx(conn); + iscsi_conn_unbind(cls_conn, true);
Quoted 5 lines
to clear the flag when we clean up the conn for relogin.
3. Fix up the other iscsi drivers so they call:
iscsi_conn_unbind(cls_conn, true);
I took a look, and it seems like the other drivers in-tree all use iscsi_conn_unbind already, and iscsi_tcp is the odd one out - only supporting stop_conn, and not setting ep_disconnect. Just swapping the iscsi_suspend_tx call for iscsi_conn_unbind works, and makes sense to me - but makes me wonder why we were using suspend_tx directly in the first place.
The only other side effects I see are we set the session to failed momentarily if we stop the conn while still logging in (iscsi_tcp will quickly change this to _TERMINATE or _IN_RECOVERY once we hit the stop_conn), which seems OK.
Quoted 26 lines
in their failure paths so when they fail they clear
ISCSI_CONN_FLAG_BOUNDandiscsi_conn_bindcan be called on the retry.Quoted 19 lines
Signed-off-by Khazhismel Kumykov <[email protected]>--- drivers/scsi/iscsi_tcp.c | 3 +++ 1 file changed, 3 insertions(+)drivers/scsi/iscsi_tcp.c +3
-0@@ -716,6 +716,9 @@ iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session,struct socket *sock;int err; +if (tcp_sw_conn->sock) +return -EINVAL ; +/* lookup for existing socket */ sock =sockfd_lookup ((int )transport_eph, &err);if (!sock) {
Quoted 23 lines · Khazhismel Kumykov
This attempts to avoid a situation where a misbehaving iscsi daemon passes a socket for a different iSCSI connection to
BIND_CONN- which would result in infinite recursion and stack overflow. This will also prevent passing *other* sockets which hadsk_user_dataoverridden, but that wouldn't have been safe anyways - since we throw away that pointer anyways. This does not cover all hypothetical scenarios where we pass bad sockets toBIND_CONN.This also papers over a different bug - we allow a daemon to call
BIND_CONNtwice for the same connection - which would result in, at the least, failing to uninitialize/teardown the previous socket, which will be addressed separately.Signed-off-by Khazhismel Kumykov <[email protected]>--- drivers/scsi/iscsi_tcp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/scsi/iscsi_tcp.c +0
-0
drivers/scsi/iscsi_tcp.c iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session, · line 731
Quoted 8 lines
drivers/scsi/iscsi_tcp.c +1
-1@@ -725,7 +725,7 @@ iscsi_sw_tcp_conn_bind(struct iscsi_cls_session *cls_session, } err = -EINVAL ; -if (!sk_is_tcp (sock->sk)) +if (!sk_is_tcp (sock->sk) || sock->sk->sk_user_data) goto free_socket; err =iscsi_conn_bind (cls_session, cls_conn, is_leading);