rxrpc: Fix missing unlock in rxrpc_do_sendmsg()

[ Upstream commit 4feb2c4 ]

One of the error paths in rxrpc_do_sendmsg() doesn't unlock the call mutex
before returning.  Fix it to do this.

Note that this still doesn't get rid of the checker warning:

   ../net/rxrpc/sendmsg.c:617:5: warning: context imbalance in 'rxrpc_do_sendmsg' - wrong count at exit

I think the interplay between the socket lock and the call's user_mutex may
be too complicated for checker to analyse, especially as
rxrpc_new_client_call_for_sendmsg(), which it calls, returns with the
call's user_mutex if successful but unconditionally drops the socket lock.

Fixes: e754eba ("rxrpc: Provide a cmsg to specify the amount of Tx data for a call")
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/rxrpc/sendmsg.c

@@ -716,7 +716,7 @@ int rxrpc_do_sendmsg(struct rxrpc_sock *rx, struct msghdr *msg, size_t len)
 			if (call->tx_total_len != -1 ||
 			    call->tx_pending ||
 			    call->tx_top != 0)
-				goto error_put;
+				goto out_put_unlock;
 			call->tx_total_len = p.call.tx_total_len;
 		}
 	}