Added test for cancellation, fixed retry bug.

On retry we were pushing back to the queue.  The problem with this is that
we could wind up pushing back many copies of the message if no reader was
present.  The new code ensures at most one retry is outstanding.

1 files changed, 17 insertions, 8 deletions

diff --git a/src/protocol/reqrep/req.c b/src/protocol/reqrep/req.c
index abd6e25c..d4098526 100644
--- a/src/protocol/reqrep/req.c
+++ b/src/protocol/reqrep/req.c
@@ -32,6 +32,7 @@ struct nni_req_sock {
 	int		raw;
 	int		closing;
 	nni_msg *	reqmsg;
+	nni_msg *	retrymsg;
 	uint32_t	nextid;         // next id
 	uint8_t		reqid[4];       // outstanding request ID (big endian)
 };
@@ -69,6 +70,7 @@ nni_req_init(void **reqp, nni_sock *sock)
 	req->retry = NNI_SECOND * 60;
 	req->sock = sock;
 	req->reqmsg = NULL;
+	req->retrymsg = NULL;
 	req->raw = 0;
 	req->resend = NNI_TIME_ZERO;
 
@@ -106,6 +108,9 @@ nni_req_fini(void *arg)
 	if (req->reqmsg != NULL) {
 		nni_msg_free(req->reqmsg);
 	}
+	if (req->retrymsg != NULL) {
+		nni_msg_free(req->retrymsg);
+	}
 	NNI_FREE_STRUCT(req);
 }
 
@@ -167,9 +172,16 @@ nni_req_pipe_send(void *arg)
 	int rv;
 
 	for (;;) {
-		rv = nni_msgq_get_sig(uwq, &msg, &rp->sigclose);
-		if (rv != 0) {
-			break;
+		nni_mtx_lock(&req->mx);
+		if ((msg = req->retrymsg) != NULL) {
+			req->retrymsg = NULL;
+		}
+		nni_mtx_unlock(&req->mx);
+		if (msg == NULL) {
+			rv = nni_msgq_get_sig(uwq, &msg, &rp->sigclose);
+			if (rv != 0) {
+				break;
+			}
 		}
 		rv = nni_pipe_send(pipe, msg);
 		if (rv != 0) {
@@ -294,12 +306,9 @@ nni_req_resender(void *arg)
 		}
 		rv = nni_cv_until(&req->cv, req->resend);
 		if ((rv == NNG_ETIMEDOUT) && (req->reqmsg != NULL)) {
-			nni_msg *dup;
 			// XXX: check for final timeout on this?
-			if (nni_msg_dup(&dup, req->reqmsg) != 0) {
-				if (nni_msgq_putback(req->uwq, dup) != 0) {
-					nni_msg_free(dup);
-				}
+			if (req->retrymsg == NULL) {
+				nni_msg_dup(&req->retrymsg, req->reqmsg);
 			}
 			req->resend = nni_clock() + req->retry;
 		}