Home » Mailing list archives » linux-netdev » 2009 » March » 2

[PATCH net-next 0/5] sctp: Bug fixes

Previous thread: [net-next PATCH 0/3] Add Device ID 8000. by Ron Mercer on Monday, March 2, 2009 - 11:07 am. (5 messages)

Next thread: [net-next PATCH v2] igb: remove skb_orphan calls by Jeff Kirsher on Monday, March 2, 2009 - 4:44 pm. (2 messages)
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 0/5] sctp: Bug fixes
Date: Monday, March 2, 2009 - 12:46 pm

Hi David

This is a set of bug fixes and cleanups for SCTP for the net-next tree.

Please apply.

Thanks
-vlad
--


[ message continues ]
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 1/5] sctp: Add some missing types for debug message
Date: Monday, March 2, 2009 - 12:46 pm

From: Wei Yongjun <yjwei@cn.fujitsu.com>

This patch add the type name "AUTH" and primitive type name
"PRIMITIVE_ASCONF" for debug message.

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
---
 net/sctp/debug.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/net/sctp/debug.c b/net/sctp/debug.c
index 67715f4..7ff548a 100644
--- a/net/sctp/debug.c
+++ b/net/sctp/debug.c
@@ -86,6 +86,9 @@ const char *sctp_cname(const sctp_subtype_t cid)
 	case SCTP_CID_FWD_TSN:
 		return "FWD_TSN";
 
+	case SCTP_CID_AUTH:
+		return "AUTH";
+
 	default:
 		break;
 	}
@@ -135,6 +138,7 @@ static const char *sctp_primitive_tbl[SCTP_NUM_PRIMITIVE_TYPES] = {
 	"PRIMITIVE_ABORT",
 	"PRIMITIVE_SEND",
 	"PRIMITIVE_REQUESTHEARTBEAT",
+	"PRIMITIVE_ASCONF",
 };
 
 /* Lookup primitive debug name. */
-- 
1.5.4.3

--


[ message continues ]
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 2/5] sctp: remove dup code in net/sctp/socket.c
Date: Monday, March 2, 2009 - 12:46 pm

From: Wei Yongjun <yjwei@cn.fujitsu.com>

Remove dup check of "if (optlen < sizeof(int))".

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
---
 net/sctp/socket.c |    6 ------
 1 files changed, 0 insertions(+), 6 deletions(-)

diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index dea864f..4bc558c 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -3069,9 +3069,6 @@ static int sctp_setsockopt_maxburst(struct sock *sk,
 	int val;
 	int assoc_id = 0;
 
-	if (optlen < sizeof(int))
-		return -EINVAL;
-
 	if (optlen == sizeof(int)) {
 		printk(KERN_WARNING
 		   "SCTP: Use of int in max_burst socket option deprecated\n");
@@ -5283,9 +5280,6 @@ static int sctp_getsockopt_maxburst(struct sock *sk, int len,
 	struct sctp_sock *sp;
 	struct sctp_association *asoc;
 
-	if (len < sizeof(int))
-		return -EINVAL;
-
 	if (len == sizeof(int)) {
 		printk(KERN_WARNING
 		   "SCTP: Use of int in max_burst socket option deprecated\n");
-- 
1.5.4.3

--


[ message continues ]
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 3/5] sctp: fix the length check in sctp_getsockopt_maxburst()
Date: Monday, March 2, 2009 - 12:46 pm

From: Wei Yongjun <yjwei@cn.fujitsu.com>

The code in sctp_getsockopt_maxburst() doesn't allow len to be larger
then struct sctp_assoc_value, which is a common case where app writers
just pass down the sizeof(buf) or something similar.

This patch fix the problem.

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
---
 net/sctp/socket.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 4bc558c..bbd3cd2 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -5286,7 +5286,8 @@ static int sctp_getsockopt_maxburst(struct sock *sk, int len,
 		printk(KERN_WARNING
 		   "SCTP: Use struct sctp_assoc_value instead\n");
 		params.assoc_id = 0;
-	} else if (len == sizeof (struct sctp_assoc_value)) {
+	} else if (len >= sizeof(struct sctp_assoc_value)) {
+		len = sizeof(struct sctp_assoc_value);
 		if (copy_from_user(&params, optval, len))
 			return -EFAULT;
 	} else
-- 
1.5.4.3

--


[ message continues ]
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 4/5] sctp: use time_before or time_after for comparing jiffies
Date: Monday, March 2, 2009 - 12:46 pm

From: Wei Yongjun <yjwei@cn.fujitsu.com>

The functions time_before or time_after are more robust
for comparing jiffies against other values.

Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
---
 net/sctp/outqueue.c  |    3 ++-
 net/sctp/transport.c |    7 ++++---
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/net/sctp/outqueue.c b/net/sctp/outqueue.c
index bc411c8..a367d15 100644
--- a/net/sctp/outqueue.c
+++ b/net/sctp/outqueue.c
@@ -428,7 +428,8 @@ void sctp_retransmit_mark(struct sctp_outq *q,
 			 * retransmitting due to T3 timeout.
 			 */
 			if (reason == SCTP_RTXR_T3_RTX &&
-			    (jiffies - chunk->sent_at) < transport->last_rto)
+			    time_before(jiffies, chunk->sent_at +
+						 transport->last_rto))
 				continue;
 
 			/* RFC 2960 6.2.1 Processing a Received SACK
diff --git a/net/sctp/transport.c b/net/sctp/transport.c
index 5c29b14..e5dde45 100644
--- a/net/sctp/transport.c
+++ b/net/sctp/transport.c
@@ -543,8 +543,8 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		 * congestion indications more than once every window of
 		 * data (or more loosely more than once every round-trip time).
 		 */
-		if ((jiffies - transport->last_time_ecne_reduced) >
-		    transport->rtt) {
+		if (time_after(jiffies, transport->last_time_ecne_reduced +
+					transport->rtt)) {
 			transport->ssthresh = max(transport->cwnd/2,
 						  4*transport->asoc->pathmtu);
 			transport->cwnd = transport->ssthresh;
@@ -561,7 +561,8 @@ void sctp_transport_lower_cwnd(struct sctp_transport *transport,
 		 * to be done every RTO interval, we do it every hearbeat
 		 * interval.
 		 */
-		if ((jiffies - transport->last_time_used) > transport->rto)
+		if (time_after(jiffies, transport->last_time_used +
+					transport->rto))
 			transport->cwnd = max(transport->cwnd/2,
 						 4*transport->asoc->pathmtu);
 		break;
-- 
1.5.4.3

--


[ message continues ]
[view original message]
From: Vlad Yasevich
Subject: [PATCH net-next 5/5] sctp: Fix broken RTO-doubling for data retransmits
Date: Monday, March 2, 2009 - 12:46 pm

Commit faee47cdbfe8d74a1573c2f81ea6dbb08d735be6
(sctp: Fix the RTO-doubling on idle-link heartbeats)
broke the RTO doubling for data retransmits.  If the
heartbeat was sent before the data T3-rtx time, the
the RTO will not double upon the T3-rtx expiration.
Distingish between the operations by passing an argument
to the function.

Additionally, Wei Youngjun pointed out that our treatment
of requested HEARTBEATS and timer HEARTBEATS is the same
wrt resetting congestion window.  That needs to be separated,
since user requested HEARTBEATS should not treat the link
as idle.

Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
---
 include/net/sctp/command.h |    3 ++-
 net/sctp/sm_sideeffect.c   |   32 +++++++++++++-------------------
 net/sctp/sm_statefuns.c    |    6 ++++--
 3 files changed, 19 insertions(+), 22 deletions(-)

diff --git a/include/net/sctp/command.h b/include/net/sctp/command.h
index 88988ab..3b96680 100644
--- a/include/net/sctp/command.h
+++ b/include/net/sctp/command.h
@@ -77,7 +77,8 @@ typedef enum {
 	SCTP_CMD_HB_TIMERS_START,    /* Start the heartbeat timers. */
 	SCTP_CMD_HB_TIMER_UPDATE,    /* Update a heartbeat timers.  */
 	SCTP_CMD_HB_TIMERS_STOP,     /* Stop the heartbeat timers.  */
-	SCTP_CMD_TRANSPORT_RESET,    /* Reset the status of a transport. */
+	SCTP_CMD_TRANSPORT_HB_SENT,  /* Reset the status of a transport. */
+	SCTP_CMD_TRANSPORT_IDLE,     /* Do manipulations on idle transport */
 	SCTP_CMD_TRANSPORT_ON,       /* Mark the transport as active. */
 	SCTP_CMD_REPORT_ERROR,   /* Pass this error back out of the sm. */
 	SCTP_CMD_REPORT_BAD_TAG, /* Verification tags didn't match. */
diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 0146cfb..5385150 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -434,7 +434,8 @@ sctp_timer_event_t *sctp_timer_events[SCTP_NUM_TIMEOUT_TYPES] = {
  *
  */
 static void sctp_do_8_2_transport_strike(struct sctp_association *asoc,
-					 struct sctp_transport ...
[ message continues ]
[view original message]
From: David Miller
Subject: Re: [PATCH net-next 0/5] sctp: Bug fixes
Date: Monday, March 2, 2009 - 11:56 pm

From: Vlad Yasevich <vladislav.yasevich@hp.com>

All applied, thanks Vlad.
--


[ message continues ]
Previous thread: [net-next PATCH 0/3] Add Device ID 8000. by Ron Mercer on Monday, March 2, 2009 - 11:07 am. (5 messages)

Next thread: [net-next PATCH v2] igb: remove skb_orphan calls by Jeff Kirsher on Monday, March 2, 2009 - 4:44 pm. (2 messages)