1 files changed, 115 insertions, 119 deletions

diff --git a/tcp_splice.c b/tcp_splice.c
index 483e45d..a7c04ca 100644
--- a/tcp_splice.c
+++ b/tcp_splice.c
@@ -28,7 +28,7 @@
  * - FIN_SENT_0:		FIN (write shutdown) sent to accepted socket
  * - FIN_SENT_1:		FIN (write shutdown) sent to target socket
  *
- * #syscalls:pasta pipe2|pipe fcntl arm:fcntl64 ppc64:fcntl64
+ * #syscalls:pasta pipe2|pipe fcntl arm:fcntl64 ppc64:fcntl64|fcntl i686:fcntl64
  */
 
 #include <sched.h>
@@ -44,7 +44,6 @@
 #include <net/ethernet.h>
 #include <netinet/in.h>
 #include <netinet/tcp.h>
-#include <sys/epoll.h>
 #include <sys/types.h>
 #include <sys/socket.h>
 
@@ -56,6 +55,7 @@
 #include "siphash.h"
 #include "inany.h"
 #include "flow.h"
+#include "epoll_ctl.h"
 
 #include "flow_table.h"
 
@@ -95,7 +95,7 @@ static int tcp_conn_sock_ns(const struct ctx *c, sa_family_t af);
  * conn_at_sidx() - Get spliced TCP connection specific flow at given sidx
  * @sidx:	Flow and side to retrieve
  *
- * Return: Spliced TCP connection at @sidx, or NULL of @sidx is invalid.
+ * Return: spliced TCP connection at @sidx, or NULL of @sidx is invalid.
  *         Asserts if the flow at @sidx is not FLOW_TCP_SPLICE.
  */
 static struct tcp_splice_conn *conn_at_sidx(flow_sidx_t sidx)
@@ -114,68 +114,62 @@ static struct tcp_splice_conn *conn_at_sidx(flow_sidx_t sidx)
  * @events:	Connection event flags
  * @ev:		Events to fill in, 0 is accepted socket, 1 is connecting socket
  */
-static void tcp_splice_conn_epoll_events(uint16_t events,
-					 struct epoll_event ev[])
+static uint32_t tcp_splice_conn_epoll_events(uint16_t events, unsigned sidei)
 {
-	unsigned sidei;
-
-	flow_foreach_sidei(sidei)
-		ev[sidei].events = 0;
+	uint32_t e = 0;
 
 	if (events & SPLICE_ESTABLISHED) {
-		flow_foreach_sidei(sidei) {
-			if (!(events & FIN_SENT(!sidei)))
-				ev[sidei].events = EPOLLIN | EPOLLRDHUP;
-		}
-	} else if (events & SPLICE_CONNECT) {
-		ev[1].events = EPOLLOUT;
+		if (!(events & FIN_SENT(!sidei)))
+			e = EPOLLIN | EPOLLRDHUP;
+	} else if (sidei == 1 && events & SPLICE_CONNECT) {
+		e = EPOLLOUT;
 	}
 
-	flow_foreach_sidei(sidei)
-		ev[sidei].events |= (events & OUT_WAIT(sidei)) ? EPOLLOUT : 0;
+	if (events & OUT_WAIT(sidei))
+		e |= EPOLLOUT;
+	if (events & OUT_WAIT(!sidei))
+		e &= ~EPOLLIN;
+
+	return e;
 }
 
 /**
  * tcp_splice_epoll_ctl() - Add/modify/delete epoll state from connection events
- * @c:		Execution context
  * @conn:	Connection pointer
  *
  * Return: 0 on success, negative error code on failure (not on deletion)
  */
-static int tcp_splice_epoll_ctl(const struct ctx *c,
-				struct tcp_splice_conn *conn)
+static int tcp_splice_epoll_ctl(struct tcp_splice_conn *conn)
 {
-	int m = conn->in_epoll ? EPOLL_CTL_MOD : EPOLL_CTL_ADD;
-	const union epoll_ref ref[SIDES] = {
-		{ .type = EPOLL_TYPE_TCP_SPLICE, .fd = conn->s[0],
-		  .flowside = FLOW_SIDX(conn, 0) },
-		{ .type = EPOLL_TYPE_TCP_SPLICE, .fd = conn->s[1],
-		  .flowside = FLOW_SIDX(conn, 1) }
-	};
-	struct epoll_event ev[SIDES] = { { .data.u64 = ref[0].u64 },
-					 { .data.u64 = ref[1].u64 } };
-
-	tcp_splice_conn_epoll_events(conn->events, ev);
-
-	if (epoll_ctl(c->epollfd, m, conn->s[0], &ev[0]) ||
-	    epoll_ctl(c->epollfd, m, conn->s[1], &ev[1])) {
+	uint32_t events[2];
+	int m;
+
+	if (flow_in_epoll(&conn->f)) {
+		m = EPOLL_CTL_MOD;
+	} else {
+		flow_epollid_set(&conn->f, EPOLLFD_ID_DEFAULT);
+		m = EPOLL_CTL_ADD;
+	}
+
+	events[0] = tcp_splice_conn_epoll_events(conn->events, 0);
+	events[1] = tcp_splice_conn_epoll_events(conn->events, 1);
+
+	if (flow_epoll_set(&conn->f, m, events[0], conn->s[0], 0) ||
+	    flow_epoll_set(&conn->f, m, events[1], conn->s[1], 1)) {
 		int ret = -errno;
-		flow_err(conn, "ERROR on epoll_ctl(): %s", strerror(errno));
+		flow_perror(conn, "ERROR on epoll_ctl()");
 		return ret;
 	}
 
-	conn->in_epoll = true;
-
 	return 0;
 }
 
 /**
  * conn_flag_do() - Set/unset given flag, log, update epoll on CLOSING flag
- * @c:		Execution context
  * @conn:	Connection pointer
  * @flag:	Flag to set, or ~flag to unset
  */
-static void conn_flag_do(const struct ctx *c, struct tcp_splice_conn *conn,
+static void conn_flag_do(struct tcp_splice_conn *conn,
 			 unsigned long flag)
 {
 	if (flag & (flag - 1)) {
@@ -200,25 +194,23 @@ static void conn_flag_do(const struct ctx *c, struct tcp_splice_conn *conn,
 	}
 
 	if (flag == CLOSING) {
-		epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->s[0], NULL);
-		epoll_ctl(c->epollfd, EPOLL_CTL_DEL, conn->s[1], NULL);
+		epoll_del(flow_epollfd(&conn->f), conn->s[0]);
+		epoll_del(flow_epollfd(&conn->f), conn->s[1]);
 	}
 }
 
-#define conn_flag(c, conn, flag)					\
+#define conn_flag(conn, flag)					\
 	do {								\
 		flow_trace(conn, "flag at %s:%i", __func__, __LINE__);	\
-		conn_flag_do(c, conn, flag);				\
+		conn_flag_do(conn, flag);				\
 	} while (0)
 
 /**
  * conn_event_do() - Set and log connection events, update epoll state
- * @c:		Execution context
  * @conn:	Connection pointer
  * @event:	Connection event
  */
-static void conn_event_do(const struct ctx *c, struct tcp_splice_conn *conn,
-			  unsigned long event)
+static void conn_event_do(struct tcp_splice_conn *conn, unsigned long event)
 {
 	if (event & (event - 1)) {
 		int flag_index = fls(~event);
@@ -240,14 +232,14 @@ static void conn_event_do(const struct ctx *c, struct tcp_splice_conn *conn,
 			flow_dbg(conn, "%s", tcp_splice_event_str[flag_index]);
 	}
 
-	if (tcp_splice_epoll_ctl(c, conn))
-		conn_flag(c, conn, CLOSING);
+	if (tcp_splice_epoll_ctl(conn))
+		conn_flag(conn, CLOSING);
 }
 
-#define conn_event(c, conn, event)					\
+#define conn_event(conn, event)					\
 	do {								\
 		flow_trace(conn, "event at %s:%i",__func__, __LINE__);	\
-		conn_event_do(c, conn, event);				\
+		conn_event_do(conn, event);				\
 	} while (0)
 
 
@@ -313,14 +305,14 @@ static int tcp_splice_connect_finish(const struct ctx *c,
 
 		if (conn->pipe[sidei][0] < 0) {
 			if (pipe2(conn->pipe[sidei], O_NONBLOCK | O_CLOEXEC)) {
-				flow_err(conn, "cannot create %d->%d pipe: %s",
-					 sidei, !sidei, strerror(errno));
-				conn_flag(c, conn, CLOSING);
+				flow_perror(conn, "cannot create %d->%d pipe",
+					    sidei, !sidei);
+				conn_flag(conn, CLOSING);
 				return -EIO;
 			}
 
 			if (fcntl(conn->pipe[sidei][0], F_SETPIPE_SZ,
-				  c->tcp.pipe_size)) {
+				  c->tcp.pipe_size) != (int)c->tcp.pipe_size) {
 				flow_trace(conn,
 					   "cannot set %d->%d pipe size to %zu",
 					   sidei, !sidei, c->tcp.pipe_size);
@@ -329,7 +321,7 @@ static int tcp_splice_connect_finish(const struct ctx *c,
 	}
 
 	if (!(conn->events & SPLICE_ESTABLISHED))
-		conn_event(c, conn, SPLICE_ESTABLISHED);
+		conn_event(conn, SPLICE_ESTABLISHED);
 
 	return 0;
 }
@@ -347,10 +339,10 @@ static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn)
 	sa_family_t af = inany_v4(&tgt->eaddr) ? AF_INET : AF_INET6;
 	uint8_t tgtpif = conn->f.pif[TGTSIDE];
 	union sockaddr_inany sa;
-	socklen_t sl;
+	int one = 1;
 
 	if (tgtpif == PIF_HOST)
-		conn->s[1] = tcp_conn_sock(c, af);
+		conn->s[1] = tcp_conn_sock(af);
 	else if (tgtpif == PIF_SPLICE)
 		conn->s[1] = tcp_conn_sock_ns(c, af);
 	else
@@ -359,24 +351,33 @@ static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn)
 	if (conn->s[1] < 0)
 		return -1;
 
-	if (setsockopt(conn->s[1], SOL_TCP, TCP_QUICKACK,
-		       &((int){ 1 }), sizeof(int))) {
+	if (setsockopt(conn->s[1], SOL_TCP, TCP_QUICKACK, &one, sizeof(one))) {
 		flow_trace(conn, "failed to set TCP_QUICKACK on socket %i",
 			   conn->s[1]);
 	}
 
-	pif_sockaddr(c, &sa, &sl, tgtpif, &tgt->eaddr, tgt->eport);
+	if (setsockopt(conn->s[0], SOL_TCP, TCP_NODELAY, &one, sizeof(one))) {
+		flow_trace(conn, "failed to set TCP_NODELAY on socket %i",
+			   conn->s[0]);
+	}
+
+	if (setsockopt(conn->s[1], SOL_TCP, TCP_NODELAY, &one, sizeof(one))) {
+		flow_trace(conn, "failed to set TCP_NODELAY on socket %i",
+			   conn->s[1]);
+	}
+
+	pif_sockaddr(c, &sa, tgtpif, &tgt->eaddr, tgt->eport);
+
+	conn_event(conn, SPLICE_CONNECT);
 
-	if (connect(conn->s[1], &sa.sa, sl)) {
+	if (connect(conn->s[1], &sa.sa, socklen_inany(&sa))) {
 		if (errno != EINPROGRESS) {
 			flow_trace(conn, "Couldn't connect socket for splice: %s",
-				   strerror(errno));
+				   strerror_(errno));
 			return -errno;
 		}
-
-		conn_event(c, conn, SPLICE_CONNECT);
 	} else {
-		conn_event(c, conn, SPLICE_ESTABLISHED);
+		conn_event(conn, SPLICE_ESTABLISHED);
 		return tcp_splice_connect_finish(c, conn);
 	}
 
@@ -388,7 +389,7 @@ static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn)
  * @c:		Execution context
  * @af:		Address family (AF_INET or AF_INET6)
  *
- * Return: Socket fd in the namespace on success, -errno on failure
+ * Return: socket fd in the namespace on success, -errno on failure
  */
 static int tcp_conn_sock_ns(const struct ctx *c, sa_family_t af)
 {
@@ -436,7 +437,7 @@ void tcp_splice_conn_from_sock(const struct ctx *c, union flow *flow, int s0)
 		flow_trace(conn, "failed to set TCP_QUICKACK on %i", s0);
 
 	if (tcp_splice_connect(c, conn))
-		conn_flag(c, conn, CLOSING);
+		conn_flag(conn, CLOSING);
 
 	FLOW_ACTIVATE(conn);
 }
@@ -468,11 +469,10 @@ void tcp_splice_sock_handler(struct ctx *c, union epoll_ref ref,
 
 		rc = getsockopt(ref.fd, SOL_SOCKET, SO_ERROR, &err, &sl);
 		if (rc)
-			flow_err(conn, "Error retrieving SO_ERROR: %s",
-				 strerror(errno));
+			flow_perror(conn, "Error retrieving SO_ERROR");
 		else
 			flow_trace(conn, "Error event on socket: %s",
-				   strerror(err));
+				   strerror_(err));
 
 		goto close;
 	}
@@ -486,14 +486,14 @@ void tcp_splice_sock_handler(struct ctx *c, union epoll_ref ref,
 
 	if (events & EPOLLOUT) {
 		fromsidei = !evsidei;
-		conn_event(c, conn, ~OUT_WAIT(evsidei));
+		conn_event(conn, ~OUT_WAIT(evsidei));
 	} else {
 		fromsidei = evsidei;
 	}
 
 	if (events & EPOLLRDHUP)
 		/* For side 0 this is fake, but implied */
-		conn_event(c, conn, FIN_RCVD(evsidei));
+		conn_event(conn, FIN_RCVD(evsidei));
 
 swap:
 	eof = 0;
@@ -503,49 +503,53 @@ swap:
 	lowat_act_flag = RCVLOWAT_ACT(fromsidei);
 
 	while (1) {
-		ssize_t readlen, to_write = 0, written;
+		ssize_t readlen, written, pending;
 		int more = 0;
 
 retry:
-		readlen = splice(conn->s[fromsidei], NULL,
-				 conn->pipe[fromsidei][1], NULL,
-				 c->tcp.pipe_size,
-				 SPLICE_F_MOVE | SPLICE_F_NONBLOCK);
-		flow_trace(conn, "%zi from read-side call", readlen);
-		if (readlen < 0) {
-			if (errno == EINTR)
-				goto retry;
+		do
+			readlen = splice(conn->s[fromsidei], NULL,
+					 conn->pipe[fromsidei][1], NULL,
+					 c->tcp.pipe_size,
+					 SPLICE_F_MOVE | SPLICE_F_NONBLOCK);
+		while (readlen < 0 && errno == EINTR);
+
+		if (readlen < 0 && errno != EAGAIN)
+			goto close;
 
-			if (errno != EAGAIN)
-				goto close;
+		flow_trace(conn, "%zi from read-side call", readlen);
 
-			to_write = c->tcp.pipe_size;
-		} else if (!readlen) {
+		if (!readlen) {
 			eof = 1;
-			to_write = c->tcp.pipe_size;
-		} else {
+		} else if (readlen > 0) {
 			never_read = 0;
-			to_write += readlen;
+
 			if (readlen >= (long)c->tcp.pipe_size * 90 / 100)
 				more = SPLICE_F_MORE;
 
 			if (conn->flags & lowat_set_flag)
-				conn_flag(c, conn, lowat_act_flag);
+				conn_flag(conn, lowat_act_flag);
 		}
 
-eintr:
-		written = splice(conn->pipe[fromsidei][0], NULL,
-				 conn->s[!fromsidei], NULL, to_write,
-				 SPLICE_F_MOVE | more | SPLICE_F_NONBLOCK);
+		do
+			written = splice(conn->pipe[fromsidei][0], NULL,
+					 conn->s[!fromsidei], NULL,
+					 c->tcp.pipe_size,
+					 SPLICE_F_MOVE | more | SPLICE_F_NONBLOCK);
+		while (written < 0 && errno == EINTR);
+
+		if (written < 0 && errno != EAGAIN)
+			goto close;
+
 		flow_trace(conn, "%zi from write-side call (passed %zi)",
-			   written, to_write);
+			   written, c->tcp.pipe_size);
 
 		/* Most common case: skip updating counters. */
 		if (readlen > 0 && readlen == written) {
 			if (readlen >= (long)c->tcp.pipe_size * 10 / 100)
 				continue;
 
-			if (conn->flags & lowat_set_flag &&
+			if (!(conn->flags & lowat_set_flag) &&
 			    readlen > (long)c->tcp.pipe_size / 10) {
 				int lowat = c->tcp.pipe_size / 4;
 
@@ -554,10 +558,10 @@ eintr:
 					       &lowat, sizeof(lowat))) {
 					flow_trace(conn,
 						   "Setting SO_RCVLOWAT %i: %s",
-						   lowat, strerror(errno));
+						   lowat, strerror_(errno));
 				} else {
-					conn_flag(c, conn, lowat_set_flag);
-					conn_flag(c, conn, lowat_act_flag);
+					conn_flag(conn, lowat_set_flag);
+					conn_flag(conn, lowat_act_flag);
 				}
 			}
 
@@ -568,26 +572,19 @@ eintr:
 		conn->written[fromsidei] += written > 0 ? written : 0;
 
 		if (written < 0) {
-			if (errno == EINTR)
-				goto eintr;
-
-			if (errno != EAGAIN)
-				goto close;
-
 			if (conn->read[fromsidei] == conn->written[fromsidei])
 				break;
 
-			conn_event(c, conn, OUT_WAIT(!fromsidei));
+			conn_event(conn, OUT_WAIT(!fromsidei));
 			break;
 		}
 
 		if (never_read && written == (long)(c->tcp.pipe_size))
 			goto retry;
 
-		if (!never_read && written < to_write) {
-			to_write -= written;
+		pending = conn->read[fromsidei] - conn->written[fromsidei];
+		if (!never_read && written > 0 && written < pending)
 			goto retry;
-		}
 
 		if (eof)
 			break;
@@ -600,7 +597,7 @@ eintr:
 			if ((conn->events & FIN_RCVD(sidei)) &&
 			    !(conn->events & FIN_SENT(!sidei))) {
 				shutdown(conn->s[!sidei], SHUT_WR);
-				conn_event(c, conn, FIN_SENT(!sidei));
+				conn_event(conn, FIN_SENT(!sidei));
 			}
 		}
 	}
@@ -621,7 +618,7 @@ eintr:
 	return;
 
 close:
-	conn_flag(c, conn, CLOSING);
+	conn_flag(conn, CLOSING);
 }
 
 /**
@@ -676,7 +673,7 @@ static void tcp_splice_pipe_refill(const struct ctx *c)
 			continue;
 
 		if (fcntl(splice_pipe_pool[i][0], F_SETPIPE_SZ,
-			  c->tcp.pipe_size)) {
+			  c->tcp.pipe_size) != (int)c->tcp.pipe_size) {
 			trace("TCP (spliced): cannot set pool pipe size to %zu",
 			      c->tcp.pipe_size);
 		}
@@ -697,16 +694,16 @@ static int tcp_sock_refill_ns(void *arg)
 	ns_enter(c);
 
 	if (c->ifi4) {
-		int rc = tcp_sock_refill_pool(c, ns_sock_pool4, AF_INET);
+		int rc = tcp_sock_refill_pool(ns_sock_pool4, AF_INET);
 		if (rc < 0)
 			warn("TCP: Error refilling IPv4 ns socket pool: %s",
-			     strerror(-rc));
+			     strerror_(-rc));
 	}
 	if (c->ifi6) {
-		int rc = tcp_sock_refill_pool(c, ns_sock_pool6, AF_INET6);
+		int rc = tcp_sock_refill_pool(ns_sock_pool6, AF_INET6);
 		if (rc < 0)
 			warn("TCP: Error refilling IPv6 ns socket pool: %s",
-			     strerror(-rc));
+			     strerror_(-rc));
 	}
 
 	return 0;
@@ -741,10 +738,9 @@ void tcp_splice_init(struct ctx *c)
 
 /**
  * tcp_splice_timer() - Timer for spliced connections
- * @c:		Execution context
  * @conn:	Connection to handle
  */
-void tcp_splice_timer(const struct ctx *c, struct tcp_splice_conn *conn)
+void tcp_splice_timer(struct tcp_splice_conn *conn)
 {
 	unsigned sidei;
 
@@ -758,10 +754,10 @@ void tcp_splice_timer(const struct ctx *c, struct tcp_splice_conn *conn)
 				flow_trace(conn, "can't set SO_RCVLOWAT on %d",
 					   conn->s[sidei]);
 			}
-			conn_flag(c, conn, ~RCVLOWAT_SET(sidei));
+			conn_flag(conn, ~RCVLOWAT_SET(sidei));
 		}
 	}
 
 	flow_foreach_sidei(sidei)
-		conn_flag(c, conn, ~RCVLOWAT_ACT(sidei));
+		conn_flag(conn, ~RCVLOWAT_ACT(sidei));
 }