From afb651821b9e603fc9347b3fb5d0ba9ddfdad355 Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Thu, 27 Jul 2006 01:24:39 +0000
Subject: Just use the normal "remote closed" handler when reading ident stings

--HG--
extra : convert_revision : 9a4e042fd565f46141e81e0c1ab90260303348fe
---
 common-session.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'common-session.c')

diff --git a/common-session.c b/common-session.c
index 4c15391..b8ea6f7 100644
--- a/common-session.c
+++ b/common-session.c
@@ -229,7 +229,7 @@ void session_identification() {
 	/* write our version string, this blocks */
 	if (atomicio(write, ses.sock, LOCAL_IDENT "\r\n",
 				strlen(LOCAL_IDENT "\r\n")) == DROPBEAR_FAILURE) {
-		dropbear_exit("Error writing ident string");
+		ses.remoteclosed();
 	}
 
     /* If they send more than 50 lines, something is wrong */
@@ -250,7 +250,7 @@ void session_identification() {
 
 	if (!done) {
 		TRACE(("err: %s for '%s'\n", strerror(errno), linebuf))
-		dropbear_exit("Failed to get remote version");
+		ses.remoteclosed();
 	} else {
 		/* linebuf is already null terminated */
 		ses.remoteident = m_malloc(len);
-- 
cgit v1.2.3


From dd06653e53c9a6a3734eaf5ce5e9374f559a2bf4 Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Tue, 5 Dec 2006 13:27:59 +0000
Subject: Tidy up behaviour when select() is interrupted. We follow normal
 codepaths, just with no FDs set.

--HG--
branch : channel-fix
extra : convert_revision : d348546b80847bc0d42a7b5208bb31a54f1fdfaf
---
 common-session.c | 28 +++++++++++-----------------
 1 file changed, 11 insertions(+), 17 deletions(-)

(limited to 'common-session.c')

diff --git a/common-session.c b/common-session.c
index b8ea6f7..a16d1f9 100644
--- a/common-session.c
+++ b/common-session.c
@@ -143,27 +143,21 @@ void session_loop(void(*loophandler)()) {
 			dropbear_exit("Terminated by signal");
 		}
 		
-		if (val < 0) {
-			if (errno == EINTR) {
-				/* This must happen even if we've been interrupted, so that
-				 * changed signal-handler vars can take effect etc */
-				if (loophandler) {
-					loophandler();
-				}
-				continue;
-			} else {
-				dropbear_exit("Error in select");
-			}
+		if (val < 0 && errno != EINTR) {
+			dropbear_exit("Error in select");
+		}
+
+		if (val <= 0) {
+			/* If we were interrupted or the select timed out, we still
+			 * want to iterate over channels etc for reading, to handle
+			 * server processes exiting etc. 
+			 * We don't want to read/write FDs. */
+			FD_ZERO(&writefd);
+			FD_ZERO(&readfd);
 		}
 
 		/* check for auth timeout, rekeying required etc */
 		checktimeouts();
-		
-		if (val == 0) {
-			/* timeout */
-			TRACE(("select timeout"))
-			continue;
-		}
 
 		/* process session socket's incoming/outgoing data */
 		if (ses.sock != -1) {
-- 
cgit v1.2.3


From f5ad5c15531a8288222aeafa5b28f30298b3497f Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Fri, 9 Feb 2007 10:43:16 +0000
Subject: Improve behaviour when flushing out after a process has exited.

--HG--
branch : channel-fix
extra : convert_revision : e73ee8f7ae404a9355685c30828a0ad4524031bc
---
 channel.h         |  2 ++
 common-channel.c  | 52 +++++++++++++++++++++++++++++++++++++++++++++-------
 common-session.c  | 19 ++++++++++++++++++-
 session.h         |  3 ++-
 svr-chansession.c | 21 +++++++++++++++++----
 5 files changed, 84 insertions(+), 13 deletions(-)

(limited to 'common-session.c')

diff --git a/channel.h b/channel.h
index fc3cb90..1bfe392 100644
--- a/channel.h
+++ b/channel.h
@@ -84,6 +84,8 @@ struct Channel {
 					   for this channel (and are awaiting a confirmation
 					   or failure). */
 
+	int flushing;
+
 	const struct ChanType* type;
 
 };
diff --git a/common-channel.c b/common-channel.c
index 19b807f..d87ea7d 100644
--- a/common-channel.c
+++ b/common-channel.c
@@ -148,6 +148,7 @@ struct Channel* newchannel(unsigned int remotechan,
 	newchan->errfd = FD_CLOSED; /* this isn't always set to start with */
 	newchan->initconn = 0;
 	newchan->await_open = 0;
+	newchan->flushing = 0;
 
 	newchan->writebuf = cbuf_new(RECV_MAXWINDOW);
 	newchan->extrabuf = NULL; /* The user code can set it up */
@@ -203,12 +204,14 @@ void channelio(fd_set *readfds, fd_set *writefds) {
 
 		/* read data and send it over the wire */
 		if (channel->readfd >= 0 && FD_ISSET(channel->readfd, readfds)) {
+			TRACE(("send normal readfd"))
 			send_msg_channel_data(channel, 0);
 		}
 
 		/* read stderr data and send it over the wire */
-		if (ERRFD_IS_READ(channel) &&
-				channel->errfd >= 0 && FD_ISSET(channel->errfd, readfds)) {
+		if (ERRFD_IS_READ(channel) && channel->errfd >= 0 
+			&& FD_ISSET(channel->errfd, readfds)) {
+				TRACE(("send normal errfd"))
 				send_msg_channel_data(channel, 1);
 		}
 
@@ -265,8 +268,14 @@ static void check_close(struct Channel *channel) {
 				cbuf_getused(channel->writebuf),
 				channel->extrabuf ? cbuf_getused(channel->extrabuf) : 0))
 
+	if (!channel->flushing && channel->type->check_close
+		&& channel->type->check_close(channel))
+	{
+		channel->flushing = 1;
+	}
+
 	if (channel->recv_close && !write_pending(channel)) {
-		if (! channel->sent_close) {
+		if (!channel->sent_close) {
 			TRACE(("Sending MSG_CHANNEL_CLOSE in response to same."))
 			send_msg_channel_close(channel);
 		}
@@ -278,6 +287,22 @@ static void check_close(struct Channel *channel) {
 		close_chan_fd(channel, channel->writefd, SHUT_WR);
 	}
 
+	/* Special handling for flushing read data after an exit. We
+	   read regardless of whether the select FD was set,
+	   and if there isn't data available, the channel will get closed. */
+	if (channel->flushing) {
+		TRACE(("might send data, flushing"))
+		if (channel->readfd >= 0 && channel->transwindow > 0) {
+			TRACE(("send data readfd"))
+			send_msg_channel_data(channel, 0);
+		}
+		if (ERRFD_IS_READ(channel) && channel->readfd >= 0 
+			&& channel->transwindow > 0) {
+			TRACE(("send data errfd"))
+			send_msg_channel_data(channel, 1);
+		}
+	}
+
 	/* If we're not going to send any more data, send EOF */
 	if (!channel->sent_eof
 			&& channel->readfd == FD_CLOSED 
@@ -287,15 +312,13 @@ static void check_close(struct Channel *channel) {
 
 	/* And if we can't receive any more data from them either, close up */
 	if (!channel->sent_close
-			&& channel->writefd == FD_CLOSED
 			&& channel->readfd == FD_CLOSED
-			&& channel->errfd == FD_CLOSED) {
+			&& !write_pending(channel)) {
+		TRACE(("sending close, readfd is closed"))
 		send_msg_channel_close(channel);
 	}
-
 }
 
-
 /* Check whether a deferred (EINPROGRESS) connect() was successful, and
  * if so, set up the channel properly. Otherwise, the channel is cleaned up, so
  * it is important that the channel reference isn't used after a call to this
@@ -341,6 +364,9 @@ static void send_msg_channel_close(struct Channel *channel) {
 
 	channel->sent_eof = 1;
 	channel->sent_close = 1;
+	close_chan_fd(channel, channel->readfd, SHUT_RD);
+	close_chan_fd(channel, channel->errfd, SHUT_RDWR);
+	close_chan_fd(channel, channel->writefd, SHUT_WR);
 	TRACE(("leave send_msg_channel_close"))
 }
 
@@ -556,6 +582,7 @@ static void send_msg_channel_data(struct Channel *channel, int isextended) {
 
 	CHECKCLEARTOWRITE();
 
+	TRACE(("enter send_msg_channel_data"))
 	dropbear_assert(!channel->sent_close);
 
 	if (isextended) {
@@ -591,6 +618,9 @@ static void send_msg_channel_data(struct Channel *channel, int isextended) {
 	len = read(fd, buf_getwriteptr(ses.writepayload, maxlen), maxlen);
 	if (len <= 0) {
 		if (len == 0 || errno != EINTR) {
+			/* This will also get hit in the case of EAGAIN. The only
+			time we expect to receive EAGAIN is when we're flushing a FD,
+			in which case it can be treated the same as EOF */
 			close_chan_fd(channel, fd, SHUT_RD);
 		}
 		ses.writepayload->len = ses.writepayload->pos = 0;
@@ -606,6 +636,14 @@ static void send_msg_channel_data(struct Channel *channel, int isextended) {
 	channel->transwindow -= len;
 
 	encrypt_packet();
+	
+	/* If we receive less data than we requested when flushing, we've
+	   reached the equivalent of EOF */
+	if (channel->flushing && len < maxlen)
+	{
+		TRACE(("closing from channel, flushing out."))
+		close_chan_fd(channel, fd, SHUT_RD);
+	}
 	TRACE(("leave send_msg_channel_data"))
 }
 
diff --git a/common-session.c b/common-session.c
index a16d1f9..6e1abf3 100644
--- a/common-session.c
+++ b/common-session.c
@@ -61,6 +61,12 @@ void common_session_init(int sock, char* remotehost) {
 
 	ses.connecttimeout = 0;
 	
+	if (pipe(ses.signal_pipe) < 0) {
+		dropbear_exit("signal pipe failed");
+	}
+	setnonblocking(ses.signal_pipe[0]);
+	setnonblocking(ses.signal_pipe[1]);
+	
 	kexfirstinitialise(); /* initialise the kex state */
 
 	ses.writepayload = buf_new(MAX_TRANS_PAYLOAD_LEN);
@@ -108,7 +114,6 @@ void common_session_init(int sock, char* remotehost) {
 
 	ses.allowprivport = 0;
 
-
 	TRACE(("leave session_init"))
 }
 
@@ -132,6 +137,10 @@ void session_loop(void(*loophandler)()) {
 				FD_SET(ses.sock, &writefd);
 			}
 		}
+		
+		/* We get woken up when signal handlers write to this pipe.
+		   SIGCHLD in svr-chansession is the only one currently. */
+		FD_SET(ses.signal_pipe[0], &readfd);
 
 		/* set up for channels which require reading/writing */
 		if (ses.dataallowed) {
@@ -155,6 +164,14 @@ void session_loop(void(*loophandler)()) {
 			FD_ZERO(&writefd);
 			FD_ZERO(&readfd);
 		}
+		
+		/* We'll just empty out the pipe if required. We don't do
+		any thing with the data, since the pipe's purpose is purely to
+		wake up the select() above. */
+		if (FD_ISSET(ses.signal_pipe[0], &readfd)) {
+			char x;
+			while (read(ses.signal_pipe[0], &x, 1) > 0) {}
+		}
 
 		/* check for auth timeout, rekeying required etc */
 		checktimeouts();
diff --git a/session.h b/session.h
index a4ad45f..78e3e57 100644
--- a/session.h
+++ b/session.h
@@ -123,7 +123,8 @@ struct sshsession {
 
 	unsigned char lastpacket; /* What the last received packet type was */
 	
-
+    int signal_pipe[2]; /* stores endpoints of a self-pipe used for
+						   race-free signal handling */
 
 	/* KEX/encryption related */
 	struct KEXState kexstate;
diff --git a/svr-chansession.c b/svr-chansession.c
index a45e7a6..23cac85 100644
--- a/svr-chansession.c
+++ b/svr-chansession.c
@@ -120,9 +120,21 @@ static void sesssigchild_handler(int UNUSED(dummy)) {
 			/* we use this to determine how pid exited */
 			exit->exitsignal = -1;
 		}
+		
+		/* Make sure that the main select() loop wakes up */
+		while (1) {
+			/* EAGAIN means the pipe's full, so don't need to write anything */
+			/* isserver is just a random byte to write */
+			if (write(ses.signal_pipe[1], &ses.isserver, 1) == 1 || errno == EAGAIN) {
+				break;
+			}
+			if (errno == EINTR) {
+				continue;
+			}
+			dropbear_exit("error writing signal pipe");
+		}
 	}
 
-	
 	sa_chld.sa_handler = sesssigchild_handler;
 	sa_chld.sa_flags = SA_NOCLDSTOP;
 	sigaction(SIGCHLD, &sa_chld, NULL);
@@ -244,16 +256,17 @@ static void closechansess(struct Channel *channel) {
 	unsigned int i;
 	struct logininfo *li;
 
-	chansess = (struct ChanSess*)channel->typedata;
+	TRACE(("enter closechansess"))
 
-	send_exitsignalstatus(channel);
+	chansess = (struct ChanSess*)channel->typedata;
 
-	TRACE(("enter closechansess"))
 	if (chansess == NULL) {
 		TRACE(("leave closechansess: chansess == NULL"))
 		return;
 	}
 
+	send_exitsignalstatus(channel);
+
 	m_free(chansess->cmd);
 	m_free(chansess->term);
 
-- 
cgit v1.2.3


From 2d4d9627a21f11a3e5caa17149e5391e71af256b Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Tue, 24 Jul 2007 15:40:23 +0000
Subject: Rearrange the channel buffer sizes into three neat use-editable
 values in options.h. Increasing RECV_MAX_WINDOW gives big network performance
 increases - even with the present buffers (which haven't changed) it performs
 a lot better. Next step is to make the window size a cmdline option.

--HG--
extra : convert_revision : 24c7cb47fb56cf5b82e3bc0859b45ea83038eab0
---
 channel.h         |  8 --------
 cli-chansession.c |  2 +-
 common-channel.c  | 22 +++++++++++-----------
 common-session.c  |  2 +-
 options.h         | 31 ++++++++++++++++++++++++-------
 packet.c          |  4 ++--
 6 files changed, 39 insertions(+), 30 deletions(-)

(limited to 'common-session.c')

diff --git a/channel.h b/channel.h
index 1bfe392..46f023e 100644
--- a/channel.h
+++ b/channel.h
@@ -45,16 +45,8 @@
 /* Not a real type */
 #define SSH_OPEN_IN_PROGRESS					99
 
-#define MAX_CHANNELS 100 /* simple mem restriction, includes each tcp/x11
-							connection, so can't be _too_ small */
-
 #define CHAN_EXTEND_SIZE 3 /* how many extra slots to add when we need more */
 
-#define RECV_MAXWINDOW 8000 /* tweak */
-#define RECV_WINDOWEXTEND 1000 /* We send a "window extend" every
-								RECV_WINDOWEXTEND bytes */
-#define RECV_MAXPACKET RECV_MAXWINDOW /* tweak */
-
 struct ChanType;
 
 struct Channel {
diff --git a/cli-chansession.c b/cli-chansession.c
index 0f0d07a..e4ec45c 100644
--- a/cli-chansession.c
+++ b/cli-chansession.c
@@ -350,7 +350,7 @@ static int cli_initchansess(struct Channel *channel) {
 	channel->errfd = STDERR_FILENO;
 	setnonblocking(STDERR_FILENO);
 
-	channel->extrabuf = cbuf_new(RECV_MAXWINDOW);
+	channel->extrabuf = cbuf_new(RECV_MAX_WINDOW);
 
 	if (cli_opts.wantpty) {
 		send_chansess_pty_req(channel);
diff --git a/common-channel.c b/common-channel.c
index d77a575..ed6e5a2 100644
--- a/common-channel.c
+++ b/common-channel.c
@@ -150,11 +150,11 @@ struct Channel* newchannel(unsigned int remotechan,
 	newchan->await_open = 0;
 	newchan->flushing = 0;
 
-	newchan->writebuf = cbuf_new(RECV_MAXWINDOW);
+	newchan->writebuf = cbuf_new(RECV_MAX_WINDOW);
 	newchan->extrabuf = NULL; /* The user code can set it up */
-	newchan->recvwindow = RECV_MAXWINDOW;
+	newchan->recvwindow = RECV_MAX_WINDOW;
 	newchan->recvdonelen = 0;
-	newchan->recvmaxpacket = RECV_MAXPACKET;
+	newchan->recvmaxpacket = RECV_MAX_PAYLOAD_LEN;
 
 	ses.channels[i] = newchan;
 	ses.chancount++;
@@ -421,7 +421,7 @@ static void writechannel(struct Channel* channel, int fd, circbuffer *cbuf) {
 		channel->recvdonelen = 0;
 	}
 
-	dropbear_assert(channel->recvwindow <= RECV_MAXWINDOW);
+	dropbear_assert(channel->recvwindow <= RECV_MAX_WINDOW);
 	dropbear_assert(channel->recvwindow <= cbuf_getavail(channel->writebuf));
 	dropbear_assert(channel->extrabuf == NULL ||
 			channel->recvwindow <= cbuf_getavail(channel->extrabuf));
@@ -710,7 +710,7 @@ void common_recv_msg_channel_data(struct Channel *channel, int fd,
 
 	dropbear_assert(channel->recvwindow >= datalen);
 	channel->recvwindow -= datalen;
-	dropbear_assert(channel->recvwindow <= RECV_MAXWINDOW);
+	dropbear_assert(channel->recvwindow <= RECV_MAX_WINDOW);
 
 	TRACE(("leave recv_msg_channel_data"))
 }
@@ -727,10 +727,10 @@ void recv_msg_channel_window_adjust() {
 	
 	incr = buf_getint(ses.payload);
 	TRACE(("received window increment %d", incr))
-	incr = MIN(incr, MAX_TRANS_WIN_INCR);
+	incr = MIN(incr, TRANS_MAX_WIN_INCR);
 	
 	channel->transwindow += incr;
-	channel->transwindow = MIN(channel->transwindow, MAX_TRANS_WINDOW);
+	channel->transwindow = MIN(channel->transwindow, TRANS_MAX_WINDOW);
 
 }
 
@@ -769,9 +769,9 @@ void recv_msg_channel_open() {
 
 	remotechan = buf_getint(ses.payload);
 	transwindow = buf_getint(ses.payload);
-	transwindow = MIN(transwindow, MAX_TRANS_WINDOW);
+	transwindow = MIN(transwindow, TRANS_MAX_WINDOW);
 	transmaxpacket = buf_getint(ses.payload);
-	transmaxpacket = MIN(transmaxpacket, MAX_TRANS_PAYLOAD_LEN);
+	transmaxpacket = MIN(transmaxpacket, TRANS_MAX_PAYLOAD_LEN);
 
 	/* figure what type of packet it is */
 	if (typelen > MAX_NAME_LEN) {
@@ -970,8 +970,8 @@ int send_msg_channel_open_init(int fd, const struct ChanType *type) {
 	buf_putbyte(ses.writepayload, SSH_MSG_CHANNEL_OPEN);
 	buf_putstring(ses.writepayload, type->name, strlen(type->name));
 	buf_putint(ses.writepayload, chan->index);
-	buf_putint(ses.writepayload, RECV_MAXWINDOW);
-	buf_putint(ses.writepayload, RECV_MAXPACKET);
+	buf_putint(ses.writepayload, RECV_MAX_WINDOW);
+	buf_putint(ses.writepayload, RECV_MAX_PAYLOAD_LEN);
 
 	TRACE(("leave send_msg_channel_open_init()"))
 	return DROPBEAR_SUCCESS;
diff --git a/common-session.c b/common-session.c
index 6e1abf3..b5adad2 100644
--- a/common-session.c
+++ b/common-session.c
@@ -69,7 +69,7 @@ void common_session_init(int sock, char* remotehost) {
 	
 	kexfirstinitialise(); /* initialise the kex state */
 
-	ses.writepayload = buf_new(MAX_TRANS_PAYLOAD_LEN);
+	ses.writepayload = buf_new(TRANS_MAX_PAYLOAD_LEN);
 	ses.transseq = 0;
 
 	ses.readbuf = NULL;
diff --git a/options.h b/options.h
index 8fd7971..15d4522 100644
--- a/options.h
+++ b/options.h
@@ -216,6 +216,20 @@ etc) slower (perhaps by 50%). Recommended for most small systems. */
  * shell/sftp session etc. */
 /* #define LOG_COMMANDS */
 
+/* Window size limits. These tend to be a trade-off between memory
+   usage and network performance: */
+/* Size of the network receive window. This amount of memory is allocated
+   as a per-channel receive buffer. Increasing this value can make a
+   significant difference to network performance. */
+#define RECV_MAX_WINDOW 8192
+/* Maximum size of a received SSH data packet - this _MUST_ be >= 32768
+   in order to interoperate with other implementations */
+#define RECV_MAX_PAYLOAD_LEN 32768
+/* Maximum size of a transmitted data packet - this can be any value,
+   though increasing it may not make a significant difference. */
+#define TRANS_MAX_PAYLOAD_LEN 16384
+
+
 /*******************************************************************
  * You shouldn't edit below here unless you know you need to.
  *******************************************************************/
@@ -317,16 +331,19 @@ etc) slower (perhaps by 50%). Recommended for most small systems. */
 #define MAX_PROPOSED_ALGO 20
 
 /* size/count limits */
-
-#define MAX_PACKET_LEN 35000
 #define MIN_PACKET_LEN 16
-#define MAX_PAYLOAD_LEN 32768
 
-#define MAX_TRANS_PAYLOAD_LEN 32768
-#define MAX_TRANS_PACKET_LEN (MAX_TRANS_PAYLOAD_LEN+50)
+#define RECV_MAX_PACKET_LEN (MAX(35000, ((RECV_MAX_PAYLOAD_LEN)+100)))
+
+/* for channel code */
+#define TRANS_MAX_WINDOW 500000000 /* 500MB is sufficient, stopping overflow */
+#define TRANS_MAX_WIN_INCR 500000000 /* overflow prevention */
+
+#define RECV_WINDOWEXTEND (RECV_MAX_WINDOW / 3) /* We send a "window extend" every
+								RECV_WINDOWEXTEND bytes */
 
-#define MAX_TRANS_WINDOW 500000000 /* 500MB is sufficient, stopping overflow */
-#define MAX_TRANS_WIN_INCR 500000000 /* overflow prevention */
+#define MAX_CHANNELS 100 /* simple mem restriction, includes each tcp/x11
+							connection, so can't be _too_ small */
 
 #define MAX_STRING_LEN 1400 /* ~= MAX_PROPOSED_ALGO * MAX_NAME_LEN, also
 							   is the max length for a password etc */
diff --git a/packet.c b/packet.c
index 9e7c67a..273b9bf 100644
--- a/packet.c
+++ b/packet.c
@@ -212,7 +212,7 @@ static void read_packet_init() {
 	buf_setpos(ses.readbuf, blocksize);
 
 	/* check packet length */
-	if ((len > MAX_PACKET_LEN) ||
+	if ((len > RECV_MAX_PACKET_LEN) ||
 		(len < MIN_PACKET_LEN + macsize) ||
 		((len - macsize) % blocksize != 0)) {
 		dropbear_exit("bad packet size %d", len);
@@ -281,7 +281,7 @@ void decrypt_packet() {
 	/* payload length */
 	/* - 4 - 1 is for LEN and PADLEN values */
 	len = ses.decryptreadbuf->len - padlen - 4 - 1;
-	if ((len > MAX_PAYLOAD_LEN) || (len < 1)) {
+	if ((len > RECV_MAX_PAYLOAD_LEN) || (len < 1)) {
 		dropbear_exit("bad packet size");
 	}
 
-- 
cgit v1.2.3


From 57ae0bfedfaa475c4c49621fd4e2b168b5c73d50 Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Fri, 27 Jul 2007 17:13:42 +0000
Subject: Prevent invalid packets being sent during key-exchange, instead queue
 them until afterwards. This could sometimes terminate connections after 8
 hours if (for example) a new TCP forwarded connection was sent at the KEX
 timeout.

--HG--
extra : convert_revision : 48426bd66b8f5ba50045f7ba190d1672745132e2
---
 common-session.c |  9 +++++++-
 packet.c         | 66 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 packet.h         |  1 +
 session.h        | 10 +++++++++
 4 files changed, 84 insertions(+), 2 deletions(-)

(limited to 'common-session.c')

diff --git a/common-session.c b/common-session.c
index b5adad2..9b248cf 100644
--- a/common-session.c
+++ b/common-session.c
@@ -80,9 +80,12 @@ void common_session_init(int sock, char* remotehost) {
 	initqueue(&ses.writequeue);
 
 	ses.requirenext = SSH_MSG_KEXINIT;
-	ses.dataallowed = 0; /* don't send data yet, we'll wait until after kex */
+	ses.dataallowed = 1; /* we can send data until we actually 
+							send the SSH_MSG_KEXINIT */
 	ses.ignorenext = 0;
 	ses.lastpacket = 0;
+	ses.reply_queue_head = NULL;
+	ses.reply_queue_tail = NULL;
 
 	/* set all the algos to none */
 	ses.keys = (struct key_context*)m_malloc(sizeof(struct key_context));
@@ -192,6 +195,10 @@ void session_loop(void(*loophandler)()) {
 				process_packet();
 			}
 		}
+		
+		/* if required, flush out any queued reply packets that
+		were being held up during a KEX */
+		maybe_flush_reply_queue();
 
 		/* process pipes etc for the channels, ses.dataallowed == 0
 		 * during rekeying ) */
diff --git a/packet.c b/packet.c
index 273b9bf..87921f3 100644
--- a/packet.c
+++ b/packet.c
@@ -403,7 +403,60 @@ static buffer* buf_decompress(buffer* buf, unsigned int len) {
 #endif
 
 
+/* returns 1 if the packet is a valid type during kex (see 7.1 of rfc4253) */
+static int packet_is_okay_kex(unsigned char type) {
+	if (type >= SSH_MSG_USERAUTH_REQUEST) {
+		return 0;
+	}
+	if (type == SSH_MSG_SERVICE_REQUEST || type == SSH_MSG_SERVICE_ACCEPT) {
+		return 0;
+	}
+	if (type == SSH_MSG_KEXINIT) {
+		/* XXX should this die horribly if !dataallowed ?? */
+		return 0;
+	}
+	return 1;
+}
 
+static void enqueue_reply_packet() {
+	struct packetlist * new_item = NULL;
+	new_item = m_malloc(sizeof(struct packetlist));
+	new_item->next = NULL;
+	
+	new_item->payload = buf_newcopy(ses.writepayload);
+	buf_setpos(ses.writepayload, 0);
+	buf_setlen(ses.writepayload, 0);
+	
+	if (ses.reply_queue_tail) {
+		ses.reply_queue_tail->next = new_item;
+	} else {
+		ses.reply_queue_head = new_item;
+		ses.reply_queue_tail = new_item;
+	}
+	TRACE(("leave enqueue_reply_packet"))
+}
+
+void maybe_flush_reply_queue() {
+	struct packetlist *tmp_item = NULL, *curr_item = NULL;
+	if (!ses.dataallowed)
+	{
+		TRACE(("maybe_empty_reply_queue - no data allowed"))
+		return;
+	}
+		
+	for (curr_item = ses.reply_queue_head; curr_item; ) {
+		CHECKCLEARTOWRITE();
+		buf_putbytes(ses.writepayload,
+			curr_item->payload->data, curr_item->payload->len);
+			
+		buf_free(curr_item->payload);
+		tmp_item = curr_item;
+		curr_item = curr_item->next;
+		m_free(tmp_item);
+		encrypt_packet();
+	}
+	ses.reply_queue_head = ses.reply_queue_tail = NULL;
+}
 	
 /* encrypt the writepayload, putting into writebuf, ready for write_packet()
  * to put on the wire */
@@ -413,9 +466,20 @@ void encrypt_packet() {
 	unsigned char blocksize, macsize;
 	buffer * writebuf; /* the packet which will go on the wire */
 	buffer * clearwritebuf; /* unencrypted, possibly compressed */
+	unsigned char type;
 	
+	type = ses.writepayload->data[0];
 	TRACE(("enter encrypt_packet()"))
-	TRACE(("encrypt_packet type is %d", ses.writepayload->data[0]))
+	TRACE(("encrypt_packet type is %d", type))
+	
+	if (!ses.dataallowed && !packet_is_okay_kex(type)) {
+		/* During key exchange only particular packets are allowed.
+			Since this type isn't OK we just enqueue it to send 
+			after the KEX, see maybe_flush_reply_queue */
+		enqueue_reply_packet();
+		return;
+	}
+		
 	blocksize = ses.keys->trans_algo_crypt->blocksize;
 	macsize = ses.keys->trans_algo_mac->hashsize;
 
diff --git a/packet.h b/packet.h
index e9768cd..8fadeb3 100644
--- a/packet.h
+++ b/packet.h
@@ -35,6 +35,7 @@ void encrypt_packet();
 
 void process_packet();
 
+void maybe_flush_reply_queue();
 typedef struct PacketType {
 	unsigned char type; /* SSH_MSG_FOO */
 	void (*handler)();
diff --git a/session.h b/session.h
index 559fe29..8a95614 100644
--- a/session.h
+++ b/session.h
@@ -81,6 +81,12 @@ struct key_context {
 
 };
 
+struct packetlist;
+struct packetlist {
+	struct packetlist *next;
+	buffer * payload;
+};
+
 struct sshsession {
 
 	/* Is it a client or server? */
@@ -137,6 +143,10 @@ struct sshsession {
 	buffer* kexhashbuf; /* session hash buffer calculated from various packets*/
 	buffer* transkexinit; /* the kexinit packet we send should be kept so we
 							 can add it to the hash when generating keys */
+							
+	/* a list of queued replies that should be sent after a KEX has
+	   concluded (ie, while dataallowed was unset)*/
+	struct packetlist *reply_queue_head, *reply_queue_tail;
 
 	algo_type*(*buf_match_algo)(buffer*buf, algo_type localalgos[],
 			int *goodguess); /* The function to use to choose which algorithm
-- 
cgit v1.2.3


From 75ec4d6510aba8780effddb53eee4a0296015608 Mon Sep 17 00:00:00 2001
From: Matt Johnston <matt@ucc.asn.au>
Date: Wed, 8 Aug 2007 15:12:06 +0000
Subject: - Add -K keepalive flag for dropbear and dbclient - Try to reduce the
 frequency of select() timeouts - Add a max receive window size of 1MB

--HG--
extra : convert_revision : 9aa22036cb511cddb35fbc0e09ad05acb39b64d1
---
 cli-runopts.c    | 21 ++++++++++++++++++---
 common-kex.c     |  7 +------
 common-session.c | 46 ++++++++++++++++++++++++++++++++++------------
 dbclient.1       |  7 +++++++
 dropbear.8       |  7 +++++++
 includes.h       |  1 +
 kex.h            |  2 +-
 options.h        |  7 ++++---
 packet.c         |  2 ++
 process-packet.c |  2 +-
 runopts.h        |  1 +
 session.h        |  9 +++++++--
 svr-auth.c       |  2 +-
 svr-main.c       |  8 ++------
 svr-runopts.c    | 27 +++++++++++++++++++++------
 svr-session.c    |  8 +-------
 16 files changed, 109 insertions(+), 48 deletions(-)

(limited to 'common-session.c')

diff --git a/cli-runopts.c b/cli-runopts.c
index f38ccd2..68990fa 100644
--- a/cli-runopts.c
+++ b/cli-runopts.c
@@ -63,11 +63,14 @@ static void printhelp() {
 #ifdef ENABLE_CLI_REMOTETCPFWD
 					"-R <listenport:remotehost:remoteport> Remote port forwarding\n"
 #endif
-					"-W <receive_window_buffer> (default %d, larger may be faster)\n"
+					"-W <receive_window_buffer> (default %d, larger may be faster, max 1MB)\n"
+					"-K <keepalive>  (0 is never, default %d)\n"
 #ifdef DEBUG_TRACE
 					"-v    verbose\n"
 #endif
-					,DROPBEAR_VERSION, cli_opts.progname, DEFAULT_RECV_WINDOW);
+					,DROPBEAR_VERSION, cli_opts.progname,
+					DEFAULT_RECV_WINDOW, DEFAULT_KEEPALIVE);
+					
 }
 
 void cli_getopts(int argc, char ** argv) {
@@ -112,6 +115,7 @@ void cli_getopts(int argc, char ** argv) {
 	*/
 	opts.recv_window = DEFAULT_RECV_WINDOW;
 	char* recv_window_arg = NULL;
+	char* keepalive_arg = NULL;
 
 	/* Iterate all the arguments */
 	for (i = 1; i < (unsigned int)argc; i++) {
@@ -207,6 +211,9 @@ void cli_getopts(int argc, char ** argv) {
 				case 'W':
 					next = &recv_window_arg;
 					break;
+				case 'K':
+					next = &keepalive_arg;
+					break;
 #ifdef DEBUG_TRACE
 				case 'v':
 					debug_trace = 1;
@@ -302,11 +309,19 @@ void cli_getopts(int argc, char ** argv) {
 	if (recv_window_arg)
 	{
 		opts.recv_window = atol(recv_window_arg);
-		if (opts.recv_window == 0)
+		if (opts.recv_window == 0 || opts.recv_window > MAX_RECV_WINDOW)
 		{
 			dropbear_exit("Bad recv window '%s'", recv_window_arg);
 		}
 	}
+	if (keepalive_arg) {
+		opts.keepalive_secs = strtoul(keepalive_arg, NULL, 10);
+		if (opts.keepalive_secs == 0 && errno == EINVAL)
+		{
+			dropbear_exit("Bad keepalive '%s'", keepalive_arg);
+		}
+	}
+	
 }
 
 #ifdef ENABLE_CLI_PUBKEY_AUTH
diff --git a/common-kex.c b/common-kex.c
index dd36cd1..e9c655d 100644
--- a/common-kex.c
+++ b/common-kex.c
@@ -188,8 +188,6 @@ void kexfirstinitialise() {
 /* Reset the kex state, ready for a new negotiation */
 static void kexinitialise() {
 
-	struct timeval tv;
-
 	TRACE(("kexinitialise()"))
 
 	/* sent/recv'd MSG_KEXINIT */
@@ -206,10 +204,7 @@ static void kexinitialise() {
 	ses.kexstate.datatrans = 0;
 	ses.kexstate.datarecv = 0;
 
-	if (gettimeofday(&tv, 0) < 0) {
-		dropbear_exit("Error getting time");
-	}
-	ses.kexstate.lastkextime = tv.tv_sec;
+	ses.kexstate.lastkextime = time(NULL);
 
 }
 
diff --git a/common-session.c b/common-session.c
index 9b248cf..79313f2 100644
--- a/common-session.c
+++ b/common-session.c
@@ -34,8 +34,10 @@
 #include "kex.h"
 #include "channel.h"
 #include "atomicio.h"
+#include "runopts.h"
 
 static void checktimeouts();
+static long select_timeout();
 static int ident_readln(int fd, char* buf, int count);
 
 struct sshsession ses; /* GLOBAL */
@@ -59,7 +61,8 @@ void common_session_init(int sock, char* remotehost) {
 	ses.sock = sock;
 	ses.maxfd = sock;
 
-	ses.connecttimeout = 0;
+	ses.connect_time = 0;
+	ses.last_packet_time = 0;
 	
 	if (pipe(ses.signal_pipe) < 0) {
 		dropbear_exit("signal pipe failed");
@@ -129,7 +132,7 @@ void session_loop(void(*loophandler)()) {
 	/* main loop, select()s for all sockets in use */
 	for(;;) {
 
-		timeout.tv_sec = SELECT_TIMEOUT;
+		timeout.tv_sec = select_timeout();
 		timeout.tv_usec = 0;
 		FD_ZERO(&writefd);
 		FD_ZERO(&readfd);
@@ -359,20 +362,22 @@ static int ident_readln(int fd, char* buf, int count) {
 	return pos+1;
 }
 
+void send_msg_ignore() {
+	CHECKCLEARTOWRITE();
+	buf_putbyte(ses.writepayload, SSH_MSG_IGNORE);
+	buf_putstring(ses.writepayload, "", 0);
+	encrypt_packet();
+}
+
 /* Check all timeouts which are required. Currently these are the time for
  * user authentication, and the automatic rekeying. */
 static void checktimeouts() {
 
-	struct timeval tv;
-	long secs;
-
-	if (gettimeofday(&tv, 0) < 0) {
-		dropbear_exit("Error getting time");
-	}
+	time_t now;
 
-	secs = tv.tv_sec;
+	now = time(NULL);
 	
-	if (ses.connecttimeout != 0 && secs > ses.connecttimeout) {
+	if (ses.connect_time != 0 && now - ses.connect_time >= AUTH_TIMEOUT) {
 			dropbear_close("Timeout before auth");
 	}
 
@@ -382,10 +387,27 @@ static void checktimeouts() {
 	}
 
 	if (!ses.kexstate.sentkexinit
-			&& (secs - ses.kexstate.lastkextime >= KEX_REKEY_TIMEOUT
-			|| ses.kexstate.datarecv+ses.kexstate.datatrans >= KEX_REKEY_DATA)){
+			&& (now - ses.kexstate.lastkextime >= KEX_REKEY_TIMEOUT
+			|| ses.kexstate.datarecv+ses.kexstate.datatrans >= KEX_REKEY_DATA)) {
 		TRACE(("rekeying after timeout or max data reached"))
 		send_msg_kexinit();
 	}
+	
+	if (opts.keepalive_secs > 0 
+		&& now - ses.last_packet_time >= opts.keepalive_secs) {
+		send_msg_ignore();
+	}
 }
 
+static long select_timeout() {
+	/* determine the minimum timeout that might be required, so
+	as to avoid waking when unneccessary */
+	long ret = LONG_MAX;
+	if (KEX_REKEY_TIMEOUT > 0)
+		ret = MIN(KEX_REKEY_TIMEOUT, ret);
+	if (AUTH_TIMEOUT > 0)
+		ret = MIN(AUTH_TIMEOUT, ret);
+	if (opts.keepalive_secs > 0)
+		ret = MIN(opts.keepalive_secs, ret);
+	return ret;
+}
diff --git a/dbclient.1 b/dbclient.1
index 3e67c08..e72f249 100644
--- a/dbclient.1
+++ b/dbclient.1
@@ -79,6 +79,13 @@ connection will abort as normal.
 Specify the per-channel receive window buffer size. Increasing this 
 may improve network performance at the expense of memory use. Use -h to see the
 default buffer size.
+.TP
+.B \-K \fItimeout_seconds
+Ensure that traffic is transmitted at a certain interval in seconds. This is
+useful for working around firewalls or routers that drop connections after
+a certain period of inactivity. The trade-off is that a session may be
+closed if there is a temporary lapse of network connectivity. A setting
+if 0 disables keepalives.
 .SH AUTHOR
 Matt Johnston (matt@ucc.asn.au).
 .br
diff --git a/dropbear.8 b/dropbear.8
index 70dfb3e..c9c2e79 100644
--- a/dropbear.8
+++ b/dropbear.8
@@ -87,6 +87,13 @@ Allow remote hosts to connect to forwarded ports.
 Specify the per-channel receive window buffer size. Increasing this 
 may improve network performance at the expense of memory use. Use -h to see the
 default buffer size.
+.TP
+.B \-K \fItimeout_seconds
+Ensure that traffic is transmitted at a certain interval in seconds. This is
+useful for working around firewalls or routers that drop connections after
+a certain period of inactivity. The trade-off is that a session may be
+closed if there is a temporary lapse of network connectivity. A setting
+if 0 disables keepalives.
 .SH AUTHOR
 Matt Johnston (matt@ucc.asn.au).
 .br
diff --git a/includes.h b/includes.h
index 017de66..03cb1cf 100644
--- a/includes.h
+++ b/includes.h
@@ -56,6 +56,7 @@
 #include <ctype.h>
 #include <stdarg.h>
 #include <dirent.h>
+#include <time.h>
 
 #ifdef HAVE_UTMP_H
 #include <utmp.h>
diff --git a/kex.h b/kex.h
index 448ad1b..d3dd187 100644
--- a/kex.h
+++ b/kex.h
@@ -53,7 +53,7 @@ struct KEXState {
 	unsigned donefirstkex : 1; /* Set to 1 after the first kex has completed,
 								  ie the transport layer has been set up */
 
-	long lastkextime; /* time of the last kex */
+	time_t lastkextime; /* time of the last kex */
 	unsigned int datatrans; /* data transmitted since last kex */
 	unsigned int datarecv; /* data received since last kex */
 
diff --git a/options.h b/options.h
index 939206a..722a43a 100644
--- a/options.h
+++ b/options.h
@@ -231,6 +231,9 @@ etc) slower (perhaps by 50%). Recommended for most small systems. */
    though increasing it may not make a significant difference. */
 #define TRANS_MAX_PAYLOAD_LEN 16384
 
+/* Ensure that data is transmitted every KEEPALIVE seconds. This can
+be overridden at runtime with -K. 0 disables keepalives */
+#define DEFAULT_KEEPALIVE 0
 
 /*******************************************************************
  * You shouldn't edit below here unless you know you need to.
@@ -287,9 +290,6 @@ etc) slower (perhaps by 50%). Recommended for most small systems. */
 
 #define _PATH_CP "/bin/cp"
 
-/* Timeouts in seconds */
-#define SELECT_TIMEOUT 20
-
 /* success/failure defines */
 #define DROPBEAR_SUCCESS 0
 #define DROPBEAR_FAILURE -1
@@ -343,6 +343,7 @@ etc) slower (perhaps by 50%). Recommended for most small systems. */
 
 #define RECV_WINDOWEXTEND (opts.recv_window / 3) /* We send a "window extend" every
 								RECV_WINDOWEXTEND bytes */
+#define MAX_RECV_WINDOW (1024*1024) /* 1 MB should be enough */
 
 #define MAX_CHANNELS 100 /* simple mem restriction, includes each tcp/x11
 							connection, so can't be _too_ small */
diff --git a/packet.c b/packet.c
index 87921f3..28c7346 100644
--- a/packet.c
+++ b/packet.c
@@ -71,6 +71,8 @@ void write_packet() {
 			dropbear_exit("error writing");
 		}
 	} 
+	
+	ses.last_packet_time = time(NULL);
 
 	if (written == 0) {
 		ses.remoteclosed();
diff --git a/process-packet.c b/process-packet.c
index ba39d9f..d96c1cb 100644
--- a/process-packet.c
+++ b/process-packet.c
@@ -56,8 +56,8 @@ void process_packet() {
 	switch(type) {
 
 		case SSH_MSG_IGNORE:
+			goto out;
 		case SSH_MSG_DEBUG:
-			TRACE(("received SSH_MSG_IGNORE or SSH_MSG_DEBUG"))
 			goto out;
 
 		case SSH_MSG_UNIMPLEMENTED:
diff --git a/runopts.h b/runopts.h
index e430371..d6e8917 100644
--- a/runopts.h
+++ b/runopts.h
@@ -37,6 +37,7 @@ typedef struct runopts {
 	int listen_fwd_all;
 #endif
 	unsigned int recv_window;
+	time_t keepalive_secs;
 
 } runopts;
 
diff --git a/session.h b/session.h
index 8a95614..e11c959 100644
--- a/session.h
+++ b/session.h
@@ -45,6 +45,7 @@ void common_session_init(int sock, char* remotehost);
 void session_loop(void(*loophandler)());
 void common_session_cleanup();
 void session_identification();
+void send_msg_ignore();
 
 
 /* Server */
@@ -92,8 +93,9 @@ struct sshsession {
 	/* Is it a client or server? */
 	unsigned char isserver;
 
-	long connecttimeout; /* time to disconnect if we have a timeout (for
-							userauth etc), or 0 for no timeout */
+	time_t connect_time; /* time the connection was established
+							(cleared after auth once we're not
+							respecting AUTH_TIMEOUT any more) */
 
 	int sock;
 
@@ -131,6 +133,9 @@ struct sshsession {
 	
     int signal_pipe[2]; /* stores endpoints of a self-pipe used for
 						   race-free signal handling */
+						
+	time_t last_packet_time; /* time of the last packet transmission, for
+							keepalive purposes */
 
 	/* KEX/encryption related */
 	struct KEXState kexstate;
diff --git a/svr-auth.c b/svr-auth.c
index d0eba9b..ea31e79 100644
--- a/svr-auth.c
+++ b/svr-auth.c
@@ -357,7 +357,7 @@ void send_msg_userauth_success() {
 	encrypt_packet();
 
 	ses.authstate.authdone = 1;
-	ses.connecttimeout = 0;
+	ses.connect_time = 0;
 
 
 	if (ses.authstate.pw->pw_uid == 0) {
diff --git a/svr-main.c b/svr-main.c
index 9e0578e..f7ce221 100644
--- a/svr-main.c
+++ b/svr-main.c
@@ -111,7 +111,6 @@ static void main_inetd() {
 #ifdef NON_INETD_MODE
 void main_noinetd() {
 	fd_set fds;
-	struct timeval seltimeout;
 	unsigned int i, j;
 	int val;
 	int maxsock = -1;
@@ -175,9 +174,6 @@ void main_noinetd() {
 
 		FD_ZERO(&fds);
 		
-		seltimeout.tv_sec = 60;
-		seltimeout.tv_usec = 0;
-		
 		/* listening sockets */
 		for (i = 0; i < listensockcount; i++) {
 			FD_SET(listensocks[i], &fds);
@@ -191,7 +187,7 @@ void main_noinetd() {
 			}
 		}
 
-		val = select(maxsock+1, &fds, NULL, NULL, &seltimeout);
+		val = select(maxsock+1, &fds, NULL, NULL, NULL);
 
 		if (exitflag) {
 			unlink(svr_opts.pidfile);
@@ -199,7 +195,7 @@ void main_noinetd() {
 		}
 		
 		if (val == 0) {
-			/* timeout reached */
+			/* timeout reached - shouldn't happen. eh */
 			continue;
 		}
 
diff --git a/svr-runopts.c b/svr-runopts.c
index f78461b..83c75c3 100644
--- a/svr-runopts.c
+++ b/svr-runopts.c
@@ -80,7 +80,8 @@ static void printhelp(const char * progname) {
 #ifdef INETD_MODE
 					"-i		Start for inetd\n"
 #endif
-					"-W <receive_window_buffer> (default %d, larger may be faster)\n"
+					"-W <receive_window_buffer> (default %d, larger may be faster, max 1MB)\n"
+					"-K <keepalive>  (0 is never, default %d)\n"
 #ifdef DEBUG_TRACE
 					"-v		verbose\n"
 #endif
@@ -91,7 +92,8 @@ static void printhelp(const char * progname) {
 #ifdef DROPBEAR_RSA
 					RSA_PRIV_FILENAME,
 #endif
-					DROPBEAR_MAX_PORTS, DROPBEAR_DEFPORT, DROPBEAR_PIDFILE, DEFAULT_RECV_WINDOW);
+					DROPBEAR_MAX_PORTS, DROPBEAR_DEFPORT, DROPBEAR_PIDFILE,
+					DEFAULT_RECV_WINDOW, DEFAULT_KEEPALIVE);
 }
 
 void svr_getopts(int argc, char ** argv) {
@@ -99,6 +101,8 @@ void svr_getopts(int argc, char ** argv) {
 	unsigned int i;
 	char ** next = 0;
 	int nextisport = 0;
+	char* recv_window_arg = NULL;
+	char* keepalive_arg = NULL;
 
 	/* see printhelp() for options */
 	svr_opts.rsakeyfile = NULL;
@@ -130,7 +134,8 @@ void svr_getopts(int argc, char ** argv) {
 	svr_opts.usingsyslog = 1;
 #endif
 	opts.recv_window = DEFAULT_RECV_WINDOW;
-	char* recv_window_arg = NULL;
+	opts.keepalive_secs = DEFAULT_KEEPALIVE;	
+	
 #ifdef ENABLE_SVR_REMOTETCPFWD
 	opts.listen_fwd_all = 0;
 #endif
@@ -210,6 +215,9 @@ void svr_getopts(int argc, char ** argv) {
 				case 'W':
 					next = &recv_window_arg;
 					break;
+				case 'K':
+					next = &keepalive_arg;
+					break;
 #if defined(ENABLE_SVR_PASSWORD_AUTH) || defined(ENABLE_SVR_PAM_AUTH)
 				case 's':
 					svr_opts.noauthpass = 1;
@@ -274,14 +282,21 @@ void svr_getopts(int argc, char ** argv) {
 
 	}
 	
-	if (recv_window_arg)
-	{
+	if (recv_window_arg) {
 		opts.recv_window = atol(recv_window_arg);
-		if (opts.recv_window == 0)
+		if (opts.recv_window == 0 || opts.recv_window > MAX_RECV_WINDOW)
 		{
 			dropbear_exit("Bad recv window '%s'", recv_window_arg);
 		}
 	}
+	
+	if (keepalive_arg) {
+		opts.keepalive_secs = strtoul(keepalive_arg, NULL, 10);
+		if (opts.keepalive_secs == 0 && errno == EINVAL)
+		{
+			dropbear_exit("Bad keepalive '%s'", keepalive_arg);
+		}
+	}
 }
 
 static void addportandaddress(char* spec) {
diff --git a/svr-session.c b/svr-session.c
index 3701597..18fab6b 100644
--- a/svr-session.c
+++ b/svr-session.c
@@ -77,8 +77,6 @@ static const struct ChanType *svr_chantypes[] = {
 void svr_session(int sock, int childpipe, 
 		char* remotehost, char *addrstring) {
 
-	struct timeval timeout;
-
     reseedrandom();
 
 	crypto_init();
@@ -91,11 +89,7 @@ void svr_session(int sock, int childpipe,
 	chaninitialise(svr_chantypes);
 	svr_chansessinitialise();
 
-	if (gettimeofday(&timeout, 0) < 0) {
-		dropbear_exit("Error getting time");
-	}
-
-	ses.connecttimeout = timeout.tv_sec + AUTH_TIMEOUT;
+	ses.connect_time = time(NULL);
 
 	/* set up messages etc */
 	ses.remoteclosed = svr_remoteclosed;
-- 
cgit v1.2.3