seL4_ReplyRecv() combines seL4_Reply() and seL4_Recv(). It exists mostly for efficiency reasons: the common case of replying to a request and waiting for the next can be performed in a single kernel system call instead of two. The transition from the reply to the receive phase is also atomic.
seL4_Call() combines seL4_Send() and seL4_Recv() with some important differ- ences. The call blocks the sending thread until its message is delivered and a reply message is received.