* [PATCH 26/32] [TIPC] Fixed memory leak in tipc_link_send() when destination is unreachable
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/link.c | 6 +++++-
1 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/net/tipc/link.c b/net/tipc/link.c
index ba7d3f1..ff40c91 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -1135,9 +1135,13 @@ int tipc_link_send(struct sk_buff *buf,
if (n_ptr) {
tipc_node_lock(n_ptr);
l_ptr = n_ptr->active_links[selector & 1];
- dbg("tipc_link_send: found link %x for dest %x\n", l_ptr, dest);
if (l_ptr) {
+ dbg("tipc_link_send: found link %x for dest %x\n", l_ptr, dest);
res = tipc_link_send_buf(l_ptr, buf);
+ } else {
+ dbg("Attempt to send msg to unreachable node:\n");
+ msg_dbg(buf_msg(buf),">>>");
+ buf_discard(buf);
}
tipc_node_unlock(n_ptr);
} else {
--
1.4.0
^ permalink raw reply related
* [PATCH 23/32] [TIPC] Optimized argument validation done by connect().
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 17 +++++++++++++----
1 files changed, 13 insertions(+), 4 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 6d4d2b0..32d7784 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -455,7 +455,8 @@ static int send_msg(struct kiocb *iocb,
if (unlikely(!dest))
return -EDESTADDRREQ;
- if (unlikely(dest->family != AF_TIPC))
+ if (unlikely((m->msg_namelen < sizeof(*dest)) ||
+ (dest->family != AF_TIPC)))
return -EINVAL;
needs_conn = (sock->state != SS_READY);
@@ -1245,7 +1246,8 @@ static int connect(struct socket *sock,
if (sock->state == SS_READY)
return -EOPNOTSUPP;
- /* MOVE THE REST OF THIS ERROR CHECKING TO send_msg()? */
+ /* Issue Posix-compliant error code if socket is in the wrong state */
+
if (sock->state == SS_LISTENING)
return -EOPNOTSUPP;
if (sock->state == SS_CONNECTING)
@@ -1253,13 +1255,20 @@ static int connect(struct socket *sock,
if (sock->state != SS_UNCONNECTED)
return -EISCONN;
- if ((destlen < sizeof(*dst)) || (dst->family != AF_TIPC) ||
- ((dst->addrtype != TIPC_ADDR_NAME) && (dst->addrtype != TIPC_ADDR_ID)))
+ /*
+ * Reject connection attempt using multicast address
+ *
+ * Note: send_msg() validates the rest of the address fields,
+ * so there's no need to do it here
+ */
+
+ if (dst->addrtype == TIPC_ADDR_MCAST)
return -EINVAL;
/* Send a 'SYN-' to destination */
m.msg_name = dest;
+ m.msg_namelen = destlen;
if ((res = send_msg(NULL, sock, &m, 0)) < 0) {
sock->state = SS_DISCONNECTING;
return res;
--
1.4.0
^ permalink raw reply related
* [PATCH 18/32] [TIPC] Connected send now checks socket state when retrying congested send.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 16 ++++++++--------
1 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 361dc34..9c834fc 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -565,15 +565,15 @@ static int send_packet(struct kiocb *ioc
return -ERESTARTSYS;
}
- if (unlikely(sock->state != SS_CONNECTED)) {
- if (sock->state == SS_DISCONNECTING)
- res = -EPIPE;
- else
- res = -ENOTCONN;
- goto exit;
- }
-
do {
+ if (unlikely(sock->state != SS_CONNECTED)) {
+ if (sock->state == SS_DISCONNECTING)
+ res = -EPIPE;
+ else
+ res = -ENOTCONN;
+ goto exit;
+ }
+
res = tipc_send(tsock->p->ref, m->msg_iovlen, m->msg_iov);
if (likely(res != -ELINKCONG)) {
exit:
--
1.4.0
^ permalink raw reply related
* [PATCH 20/32] [TIPC] Improved performance of error checking during socket creation.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 9 +++------
1 files changed, 3 insertions(+), 6 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 8cefacb..a1f2210 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -169,12 +169,6 @@ static int tipc_create(struct socket *so
struct sock *sk;
u32 ref;
- if ((sock->type != SOCK_STREAM) &&
- (sock->type != SOCK_SEQPACKET) &&
- (sock->type != SOCK_DGRAM) &&
- (sock->type != SOCK_RDM))
- return -EPROTOTYPE;
-
if (unlikely(protocol != 0))
return -EPROTONOSUPPORT;
@@ -199,6 +193,9 @@ static int tipc_create(struct socket *so
sock->ops = &msg_ops;
sock->state = SS_READY;
break;
+ default:
+ tipc_deleteport(ref);
+ return -EPROTOTYPE;
}
sk = sk_alloc(AF_TIPC, GFP_KERNEL, &tipc_proto, 1);
--
1.4.0
^ permalink raw reply related
* [PATCH 21/32] [TIPC] recvmsg() now returns TIPC ancillary data using correct level (SOL_TIPC)
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index a1f2210..abecf2d 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -744,10 +744,10 @@ static int anc_data_recv(struct msghdr *
if (unlikely(err)) {
anc_data[0] = err;
anc_data[1] = msg_data_sz(msg);
- if ((res = put_cmsg(m, SOL_SOCKET, TIPC_ERRINFO, 8, anc_data)))
+ if ((res = put_cmsg(m, SOL_TIPC, TIPC_ERRINFO, 8, anc_data)))
return res;
if (anc_data[1] &&
- (res = put_cmsg(m, SOL_SOCKET, TIPC_RETDATA, anc_data[1],
+ (res = put_cmsg(m, SOL_TIPC, TIPC_RETDATA, anc_data[1],
msg_data(msg))))
return res;
}
@@ -778,7 +778,7 @@ static int anc_data_recv(struct msghdr *
has_name = 0;
}
if (has_name &&
- (res = put_cmsg(m, SOL_SOCKET, TIPC_DESTNAME, 12, anc_data)))
+ (res = put_cmsg(m, SOL_TIPC, TIPC_DESTNAME, 12, anc_data)))
return res;
return 0;
--
1.4.0
^ permalink raw reply related
* [PATCH 24/32] [TIPC] Withdrawing all names from nameless port now returns success, not error
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/port.c | 3 ---
1 files changed, 0 insertions(+), 3 deletions(-)
diff --git a/net/tipc/port.c b/net/tipc/port.c
index 360920b..899e08e 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -1171,8 +1171,6 @@ int tipc_withdraw(u32 ref, unsigned int
p_ptr = tipc_port_lock(ref);
if (!p_ptr)
return -EINVAL;
- if (!p_ptr->publ.published)
- goto exit;
if (!seq) {
list_for_each_entry_safe(publ, tpubl,
&p_ptr->publications, pport_list) {
@@ -1199,7 +1197,6 @@ int tipc_withdraw(u32 ref, unsigned int
}
if (list_empty(&p_ptr->publications))
p_ptr->publ.published = 0;
-exit:
tipc_port_unlock(p_ptr);
return res;
}
--
1.4.0
^ permalink raw reply related
* [PATCH 22/32] [TIPC] Simplify code for returning partial success of stream send request.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index abecf2d..6d4d2b0 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -969,7 +969,7 @@ static int recv_stream(struct kiocb *ioc
restart:
if (unlikely((skb_queue_len(&sock->sk->sk_receive_queue) == 0) &&
(flags & MSG_DONTWAIT))) {
- res = (sz_copied == 0) ? -EWOULDBLOCK : 0;
+ res = -EWOULDBLOCK;
goto exit;
}
@@ -1060,7 +1060,7 @@ restart:
exit:
up(&tsock->sem);
- return res ? res : sz_copied;
+ return sz_copied ? sz_copied : res;
}
/**
--
1.4.0
^ permalink raw reply related
* [PATCH 25/32] [TIPC] Added missing warning for out-of-memory condition
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/port.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/net/tipc/port.c b/net/tipc/port.c
index 899e08e..99846a1 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -1061,6 +1061,7 @@ int tipc_createport(u32 user_ref,
up_ptr = (struct user_port *)kmalloc(sizeof(*up_ptr), GFP_ATOMIC);
if (up_ptr == NULL) {
+ warn("Port creation failed, no memory\n");
return -ENOMEM;
}
ref = tipc_createport_raw(NULL, port_dispatcher, port_wakeup, importance);
--
1.4.0
^ permalink raw reply related
* [PATCH 14/32] [TIPC] Non-operation-affecting corrections to comments & function definitions.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 12 +++++++-----
1 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index eaf4d69..0923213 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -437,7 +437,7 @@ static int dest_name_check(struct sockad
* @iocb: (unused)
* @sock: socket structure
* @m: message to send
- * @total_len: (unused)
+ * @total_len: length of message
*
* Message must have an destination specified explicitly.
* Used for SOCK_RDM and SOCK_DGRAM messages,
@@ -538,7 +538,7 @@ exit:
* @iocb: (unused)
* @sock: socket structure
* @m: message to send
- * @total_len: (unused)
+ * @total_len: length of message
*
* Used for SOCK_SEQPACKET messages and SOCK_STREAM data.
*
@@ -1386,7 +1386,7 @@ exit:
/**
* shutdown - shutdown socket connection
* @sock: socket structure
- * @how: direction to close (always treated as read + write)
+ * @how: direction to close (unused; always treated as read + write)
*
* Terminates connection (if necessary), then purges socket's receive queue.
*
@@ -1469,7 +1469,8 @@ restart:
* Returns 0 on success, errno otherwise
*/
-static int setsockopt(struct socket *sock, int lvl, int opt, char *ov, int ol)
+static int setsockopt(struct socket *sock,
+ int lvl, int opt, char __user *ov, int ol)
{
struct tipc_sock *tsock = tipc_sk(sock->sk);
u32 value;
@@ -1525,7 +1526,8 @@ static int setsockopt(struct socket *soc
* Returns 0 on success, errno otherwise
*/
-static int getsockopt(struct socket *sock, int lvl, int opt, char *ov, int *ol)
+static int getsockopt(struct socket *sock,
+ int lvl, int opt, char __user *ov, int *ol)
{
struct tipc_sock *tsock = tipc_sk(sock->sk);
int len;
--
1.4.0
^ permalink raw reply related
* [PATCH 13/32] [TIPC] Validate entire interface name when locating bearer to enable.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
This fix prevents a bearer from being enabled using the wrong interface.
For example, specifying "eth:eth14" might enable "eth:eth1" by mistake.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/eth_media.c | 5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/net/tipc/eth_media.c b/net/tipc/eth_media.c
index 3ecb100..682da4a 100644
--- a/net/tipc/eth_media.c
+++ b/net/tipc/eth_media.c
@@ -2,7 +2,7 @@
* net/tipc/eth_media.c: Ethernet bearer support for TIPC
*
* Copyright (c) 2001-2006, Ericsson AB
- * Copyright (c) 2005, Wind River Systems
+ * Copyright (c) 2005-2006, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -127,8 +127,7 @@ static int enable_bearer(struct tipc_bea
/* Find device with specified name */
- while (dev && dev->name &&
- (memcmp(dev->name, driver_name, strlen(dev->name)))) {
+ while (dev && dev->name && strncmp(dev->name, driver_name, IFNAMSIZ)) {
dev = dev->next;
}
if (!dev)
--
1.4.0
^ permalink raw reply related
* [PATCH 15/32] [TIPC] Fixed connect() to detect a dest address that is missing or too short.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 0923213..758b2d2 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1244,7 +1244,7 @@ static int connect(struct socket *sock,
if (sock->state != SS_UNCONNECTED)
return -EISCONN;
- if ((dst->family != AF_TIPC) ||
+ if ((destlen < sizeof(*dst)) || (dst->family != AF_TIPC) ||
((dst->addrtype != TIPC_ADDR_NAME) && (dst->addrtype != TIPC_ADDR_ID)))
return -EINVAL;
--
1.4.0
^ permalink raw reply related
* [PATCH 19/32] [TIPC] Stream socket send indicates partial success if data partially sent.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/socket.c | 18 +++++++++++-------
1 files changed, 11 insertions(+), 7 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 9c834fc..8cefacb 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -601,7 +601,8 @@ exit:
*
* Used for SOCK_STREAM data.
*
- * Returns the number of bytes sent on success, or errno otherwise
+ * Returns the number of bytes sent on success (or partial success),
+ * or errno if no data sent
*/
@@ -615,6 +616,7 @@ static int send_stream(struct kiocb *ioc
char __user *curr_start;
int curr_left;
int bytes_to_send;
+ int bytes_sent;
int res;
if (likely(total_len <= TIPC_MAX_USER_MSG_SIZE))
@@ -637,11 +639,11 @@ static int send_stream(struct kiocb *ioc
* of small iovec entries into send_packet().
*/
- my_msg = *m;
- curr_iov = my_msg.msg_iov;
- curr_iovlen = my_msg.msg_iovlen;
+ curr_iov = m->msg_iov;
+ curr_iovlen = m->msg_iovlen;
my_msg.msg_iov = &my_iov;
my_msg.msg_iovlen = 1;
+ bytes_sent = 0;
while (curr_iovlen--) {
curr_start = curr_iov->iov_base;
@@ -652,16 +654,18 @@ static int send_stream(struct kiocb *ioc
? curr_left : TIPC_MAX_USER_MSG_SIZE;
my_iov.iov_base = curr_start;
my_iov.iov_len = bytes_to_send;
- if ((res = send_packet(iocb, sock, &my_msg, 0)) < 0)
- return res;
+ if ((res = send_packet(iocb, sock, &my_msg, 0)) < 0) {
+ return bytes_sent ? bytes_sent : res;
+ }
curr_left -= bytes_to_send;
curr_start += bytes_to_send;
+ bytes_sent += bytes_to_send;
}
curr_iov++;
}
- return total_len;
+ return bytes_sent;
}
/**
--
1.4.0
^ permalink raw reply related
* [PATCH 27/32] [TIPC] Disallow config operations that aren't supported in certain modes.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
This change provides user-friendly feedback when TIPC is unable to perform
certain configuration operations that don't work properly in certain modes.
(In particular, any reconfiguration request that would temporarily take TIPC
from network mode to standalone mode, or from standalone mode to not running
mode, is disallowed.)
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/config.c | 83 ++++++++++++++++++++++++-----------------------------
1 files changed, 38 insertions(+), 45 deletions(-)
diff --git a/net/tipc/config.c b/net/tipc/config.c
index 48b5de2..41c8447 100644
--- a/net/tipc/config.c
+++ b/net/tipc/config.c
@@ -291,13 +291,22 @@ static struct sk_buff *cfg_set_own_addr(
if (!tipc_addr_node_valid(addr))
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (node address)");
- if (tipc_own_addr)
+ if (tipc_mode == TIPC_NET_MODE)
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
" (cannot change node address once assigned)");
+ tipc_own_addr = addr;
+
+ /*
+ * Must release all spinlocks before calling start_net() because
+ * Linux version of TIPC calls eth_media_start() which calls
+ * register_netdevice_notifier() which may block!
+ *
+ * Temporarily releasing the lock should be harmless for non-Linux TIPC,
+ * but Linux version of eth_media_start() should really be reworked
+ * so that it can be called with spinlocks held.
+ */
spin_unlock_bh(&config_lock);
- tipc_core_stop_net();
- tipc_own_addr = addr;
tipc_core_start_net();
spin_lock_bh(&config_lock);
return tipc_cfg_reply_none();
@@ -350,50 +359,21 @@ static struct sk_buff *cfg_set_max_subsc
static struct sk_buff *cfg_set_max_ports(void)
{
- int orig_mode;
u32 value;
if (!TLV_CHECK(req_tlv_area, req_tlv_space, TIPC_TLV_UNSIGNED))
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
value = *(u32 *)TLV_DATA(req_tlv_area);
value = ntohl(value);
+ if (value == tipc_max_ports)
+ return tipc_cfg_reply_none();
if (value != delimit(value, 127, 65535))
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (max ports must be 127-65535)");
-
- if (value == tipc_max_ports)
- return tipc_cfg_reply_none();
-
- if (atomic_read(&tipc_user_count) > 2)
+ if (tipc_mode != TIPC_NOT_RUNNING)
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
- " (cannot change max ports while TIPC users exist)");
-
- spin_unlock_bh(&config_lock);
- orig_mode = tipc_get_mode();
- if (orig_mode == TIPC_NET_MODE)
- tipc_core_stop_net();
- tipc_core_stop();
+ " (cannot change max ports while TIPC is active)");
tipc_max_ports = value;
- tipc_core_start();
- if (orig_mode == TIPC_NET_MODE)
- tipc_core_start_net();
- spin_lock_bh(&config_lock);
- return tipc_cfg_reply_none();
-}
-
-static struct sk_buff *set_net_max(int value, int *parameter)
-{
- int orig_mode;
-
- if (value != *parameter) {
- orig_mode = tipc_get_mode();
- if (orig_mode == TIPC_NET_MODE)
- tipc_core_stop_net();
- *parameter = value;
- if (orig_mode == TIPC_NET_MODE)
- tipc_core_start_net();
- }
-
return tipc_cfg_reply_none();
}
@@ -405,10 +385,16 @@ static struct sk_buff *cfg_set_max_zones
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
value = *(u32 *)TLV_DATA(req_tlv_area);
value = ntohl(value);
+ if (value == tipc_max_zones)
+ return tipc_cfg_reply_none();
if (value != delimit(value, 1, 255))
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (max zones must be 1-255)");
- return set_net_max(value, &tipc_max_zones);
+ if (tipc_mode == TIPC_NET_MODE)
+ return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
+ " (cannot change max zones once TIPC has joined a network)");
+ tipc_max_zones = value;
+ return tipc_cfg_reply_none();
}
static struct sk_buff *cfg_set_max_clusters(void)
@@ -419,8 +405,8 @@ static struct sk_buff *cfg_set_max_clust
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
value = *(u32 *)TLV_DATA(req_tlv_area);
value = ntohl(value);
- if (value != 1)
- return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
+ if (value != delimit(value, 1, 1))
+ return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (max clusters fixed at 1)");
return tipc_cfg_reply_none();
}
@@ -433,10 +419,16 @@ static struct sk_buff *cfg_set_max_nodes
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
value = *(u32 *)TLV_DATA(req_tlv_area);
value = ntohl(value);
+ if (value == tipc_max_nodes)
+ return tipc_cfg_reply_none();
if (value != delimit(value, 8, 2047))
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (max nodes must be 8-2047)");
- return set_net_max(value, &tipc_max_nodes);
+ if (tipc_mode == TIPC_NET_MODE)
+ return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
+ " (cannot change max nodes once TIPC has joined a network)");
+ tipc_max_nodes = value;
+ return tipc_cfg_reply_none();
}
static struct sk_buff *cfg_set_max_slaves(void)
@@ -461,15 +453,16 @@ static struct sk_buff *cfg_set_netid(voi
return tipc_cfg_reply_error_string(TIPC_CFG_TLV_ERROR);
value = *(u32 *)TLV_DATA(req_tlv_area);
value = ntohl(value);
+ if (value == tipc_net_id)
+ return tipc_cfg_reply_none();
if (value != delimit(value, 1, 9999))
return tipc_cfg_reply_error_string(TIPC_CFG_INVALID_VALUE
" (network id must be 1-9999)");
-
- if (tipc_own_addr)
+ if (tipc_mode == TIPC_NET_MODE)
return tipc_cfg_reply_error_string(TIPC_CFG_NOT_SUPPORTED
- " (cannot change network id once part of network)");
-
- return set_net_max(value, &tipc_net_id);
+ " (cannot change network id once TIPC has joined a network)");
+ tipc_net_id = value;
+ return tipc_cfg_reply_none();
}
struct sk_buff *tipc_cfg_do_cmd(u32 orig_node, u16 cmd, const void *request_area,
--
1.4.0
^ permalink raw reply related
* [PATCH 29/32] [TIPC] Enhanced & cleaned up system messages; fixed 2 obscure memory leaks.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/bcast.c | 2 +
net/tipc/bcast.h | 2 +
net/tipc/bearer.c | 70 +++++++++++++++++++++++++++----------------------
net/tipc/cluster.c | 22 +++++++++------
net/tipc/config.c | 2 +
net/tipc/discover.c | 7 +----
net/tipc/link.c | 39 +++++++++++++++------------
net/tipc/name_distr.c | 10 ++++---
net/tipc/name_table.c | 6 ++--
net/tipc/node.c | 68 +++++++++++++++++++++++++-----------------------
net/tipc/port.c | 10 ++++---
net/tipc/subscr.c | 18 ++++++-------
net/tipc/zone.c | 19 ++++++++-----
13 files changed, 149 insertions(+), 126 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 44645f5..1633ef2 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -785,7 +785,7 @@ int tipc_bclink_init(void)
bclink = kmalloc(sizeof(*bclink), GFP_ATOMIC);
if (!bcbearer || !bclink) {
nomem:
- warn("Memory squeeze; Failed to create multicast link\n");
+ warn("Multicast link creation failed, no memory\n");
kfree(bcbearer);
bcbearer = NULL;
kfree(bclink);
diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h
index 0e3be2a..b243d9d 100644
--- a/net/tipc/bcast.h
+++ b/net/tipc/bcast.h
@@ -180,7 +180,7 @@ static inline void tipc_port_list_add(st
if (!item->next) {
item->next = kmalloc(sizeof(*item), GFP_ATOMIC);
if (!item->next) {
- warn("Memory squeeze: multicast destination port list is incomplete\n");
+ warn("Incomplete multicast delivery, no memory\n");
return;
}
item->next->next = NULL;
diff --git a/net/tipc/bearer.c b/net/tipc/bearer.c
index e213a8e..4fa24b5 100644
--- a/net/tipc/bearer.c
+++ b/net/tipc/bearer.c
@@ -112,39 +112,42 @@ int tipc_register_media(u32 media_type,
goto exit;
if (!media_name_valid(name)) {
- warn("Media registration error: illegal name <%s>\n", name);
+ warn("Media <%s> rejected, illegal name\n", name);
goto exit;
}
if (!bcast_addr) {
- warn("Media registration error: no broadcast address supplied\n");
+ warn("Media <%s> rejected, no broadcast address\n", name);
goto exit;
}
if ((bearer_priority < TIPC_MIN_LINK_PRI) &&
(bearer_priority > TIPC_MAX_LINK_PRI)) {
- warn("Media registration error: priority %u\n", bearer_priority);
+ warn("Media <%s> rejected, illegal priority (%u)\n", name,
+ bearer_priority);
goto exit;
}
if ((link_tolerance < TIPC_MIN_LINK_TOL) ||
(link_tolerance > TIPC_MAX_LINK_TOL)) {
- warn("Media registration error: tolerance %u\n", link_tolerance);
+ warn("Media <%s> rejected, illegal tolerance (%u)\n", name,
+ link_tolerance);
goto exit;
}
media_id = media_count++;
if (media_id >= MAX_MEDIA) {
- warn("Attempt to register more than %u media\n", MAX_MEDIA);
+ warn("Media <%s> rejected, media limit reached (%u)\n", name,
+ MAX_MEDIA);
media_count--;
goto exit;
}
for (i = 0; i < media_id; i++) {
if (media_list[i].type_id == media_type) {
- warn("Attempt to register second media with type %u\n",
+ warn("Media <%s> rejected, duplicate type (%u)\n", name,
media_type);
media_count--;
goto exit;
}
if (!strcmp(name, media_list[i].name)) {
- warn("Attempt to re-register media name <%s>\n", name);
+ warn("Media <%s> rejected, duplicate name\n", name);
media_count--;
goto exit;
}
@@ -283,6 +286,9 @@ static struct bearer *bearer_find(const
struct bearer *b_ptr;
u32 i;
+ if (tipc_mode != TIPC_NET_MODE)
+ return NULL;
+
for (i = 0, b_ptr = tipc_bearers; i < MAX_BEARERS; i++, b_ptr++) {
if (b_ptr->active && (!strcmp(b_ptr->publ.name, name)))
return b_ptr;
@@ -475,26 +481,33 @@ int tipc_enable_bearer(const char *name,
u32 i;
int res = -EINVAL;
- if (tipc_mode != TIPC_NET_MODE)
+ if (tipc_mode != TIPC_NET_MODE) {
+ warn("Bearer <%s> rejected, not supported in standalone mode\n",
+ name);
return -ENOPROTOOPT;
-
- if (!bearer_name_validate(name, &b_name) ||
- !tipc_addr_domain_valid(bcast_scope) ||
- !in_scope(bcast_scope, tipc_own_addr))
+ }
+ if (!bearer_name_validate(name, &b_name)) {
+ warn("Bearer <%s> rejected, illegal name\n", name);
return -EINVAL;
-
+ }
+ if (!tipc_addr_domain_valid(bcast_scope) ||
+ !in_scope(bcast_scope, tipc_own_addr)) {
+ warn("Bearer <%s> rejected, illegal broadcast scope\n", name);
+ return -EINVAL;
+ }
if ((priority < TIPC_MIN_LINK_PRI ||
priority > TIPC_MAX_LINK_PRI) &&
- (priority != TIPC_MEDIA_LINK_PRI))
+ (priority != TIPC_MEDIA_LINK_PRI)) {
+ warn("Bearer <%s> rejected, illegal priority\n", name);
return -EINVAL;
+ }
write_lock_bh(&tipc_net_lock);
- if (!tipc_bearers)
- goto failed;
m_ptr = media_find(b_name.media_name);
if (!m_ptr) {
- warn("No media <%s>\n", b_name.media_name);
+ warn("Bearer <%s> rejected, media <%s> not registered\n", name,
+ b_name.media_name);
goto failed;
}
@@ -510,23 +523,24 @@ restart:
continue;
}
if (!strcmp(name, tipc_bearers[i].publ.name)) {
- warn("Bearer <%s> already enabled\n", name);
+ warn("Bearer <%s> rejected, already enabled\n", name);
goto failed;
}
if ((tipc_bearers[i].priority == priority) &&
(++with_this_prio > 2)) {
if (priority-- == 0) {
- warn("Third bearer <%s> with priority %u, unable to lower to %u\n",
- name, priority + 1, priority);
+ warn("Bearer <%s> rejected, duplicate priority\n",
+ name);
goto failed;
}
- warn("Third bearer <%s> with priority %u, lowering to %u\n",
+ warn("Bearer <%s> priority adjustment required %u->%u\n",
name, priority + 1, priority);
goto restart;
}
}
if (bearer_id >= MAX_BEARERS) {
- warn("Attempt to enable more than %d bearers\n", MAX_BEARERS);
+ warn("Bearer <%s> rejected, bearer limit reached (%u)\n",
+ name, MAX_BEARERS);
goto failed;
}
@@ -536,7 +550,7 @@ restart:
strcpy(b_ptr->publ.name, name);
res = m_ptr->enable_bearer(&b_ptr->publ);
if (res) {
- warn("Failed to enable bearer <%s>\n", name);
+ warn("Bearer <%s> rejected, enable failure (%d)\n", name, -res);
goto failed;
}
@@ -573,9 +587,6 @@ int tipc_block_bearer(const char *name)
struct link *l_ptr;
struct link *temp_l_ptr;
- if (tipc_mode != TIPC_NET_MODE)
- return -ENOPROTOOPT;
-
read_lock_bh(&tipc_net_lock);
b_ptr = bearer_find(name);
if (!b_ptr) {
@@ -584,6 +595,7 @@ int tipc_block_bearer(const char *name)
return -EINVAL;
}
+ info("Blocking bearer <%s>\n", name);
spin_lock_bh(&b_ptr->publ.lock);
b_ptr->publ.blocked = 1;
list_for_each_entry_safe(l_ptr, temp_l_ptr, &b_ptr->links, link_list) {
@@ -595,7 +607,6 @@ int tipc_block_bearer(const char *name)
}
spin_unlock_bh(&b_ptr->publ.lock);
read_unlock_bh(&tipc_net_lock);
- info("Blocked bearer <%s>\n", name);
return TIPC_OK;
}
@@ -611,15 +622,13 @@ static int bearer_disable(const char *na
struct link *l_ptr;
struct link *temp_l_ptr;
- if (tipc_mode != TIPC_NET_MODE)
- return -ENOPROTOOPT;
-
b_ptr = bearer_find(name);
if (!b_ptr) {
warn("Attempt to disable unknown bearer <%s>\n", name);
return -EINVAL;
}
+ info("Disabling bearer <%s>\n", name);
tipc_disc_stop_link_req(b_ptr->link_req);
spin_lock_bh(&b_ptr->publ.lock);
b_ptr->link_req = NULL;
@@ -635,7 +644,6 @@ static int bearer_disable(const char *na
tipc_link_delete(l_ptr);
}
spin_unlock_bh(&b_ptr->publ.lock);
- info("Disabled bearer <%s>\n", name);
memset(b_ptr, 0, sizeof(struct bearer));
return TIPC_OK;
}
diff --git a/net/tipc/cluster.c b/net/tipc/cluster.c
index 1aed815..1dcb694 100644
--- a/net/tipc/cluster.c
+++ b/net/tipc/cluster.c
@@ -60,8 +60,10 @@ struct cluster *tipc_cltr_create(u32 add
int alloc;
c_ptr = (struct cluster *)kmalloc(sizeof(*c_ptr), GFP_ATOMIC);
- if (c_ptr == NULL)
+ if (c_ptr == NULL) {
+ warn("Cluster creation failure, no memory\n");
return NULL;
+ }
memset(c_ptr, 0, sizeof(*c_ptr));
c_ptr->addr = tipc_addr(tipc_zone(addr), tipc_cluster(addr), 0);
@@ -70,30 +72,32 @@ struct cluster *tipc_cltr_create(u32 add
else
max_nodes = tipc_max_nodes + 1;
alloc = sizeof(void *) * (max_nodes + 1);
+
c_ptr->nodes = (struct node **)kmalloc(alloc, GFP_ATOMIC);
if (c_ptr->nodes == NULL) {
+ warn("Cluster creation failure, no memory for node area\n");
kfree(c_ptr);
return NULL;
}
- memset(c_ptr->nodes, 0, alloc);
+ memset(c_ptr->nodes, 0, alloc);
+
if (in_own_cluster(addr))
tipc_local_nodes = c_ptr->nodes;
c_ptr->highest_slave = LOWEST_SLAVE - 1;
c_ptr->highest_node = 0;
z_ptr = tipc_zone_find(tipc_zone(addr));
- if (z_ptr == NULL) {
+ if (!z_ptr) {
z_ptr = tipc_zone_create(addr);
}
- if (z_ptr != NULL) {
- tipc_zone_attach_cluster(z_ptr, c_ptr);
- c_ptr->owner = z_ptr;
- }
- else {
+ if (!z_ptr) {
+ kfree(c_ptr->nodes);
kfree(c_ptr);
- c_ptr = NULL;
+ return NULL;
}
+ tipc_zone_attach_cluster(z_ptr, c_ptr);
+ c_ptr->owner = z_ptr;
return c_ptr;
}
diff --git a/net/tipc/config.c b/net/tipc/config.c
index 41c8447..3ec502f 100644
--- a/net/tipc/config.c
+++ b/net/tipc/config.c
@@ -642,7 +642,7 @@ static void cfg_named_msg_event(void *us
if ((size < sizeof(*req_hdr)) ||
(size != TCM_ALIGN(ntohl(req_hdr->tcm_len))) ||
(ntohs(req_hdr->tcm_flags) != TCM_F_REQUEST)) {
- warn("discarded invalid configuration message\n");
+ warn("Invalid configuration message discarded\n");
return;
}
diff --git a/net/tipc/discover.c b/net/tipc/discover.c
index 9260138..ee9b448 100644
--- a/net/tipc/discover.c
+++ b/net/tipc/discover.c
@@ -176,7 +176,6 @@ void tipc_disc_recv_msg(struct sk_buff *
n_ptr = tipc_node_create(orig);
}
if (n_ptr == NULL) {
- warn("Memory squeeze; Failed to create node\n");
return;
}
spin_lock_bh(&n_ptr->lock);
@@ -191,10 +190,8 @@ void tipc_disc_recv_msg(struct sk_buff *
}
addr = &link->media_addr;
if (memcmp(addr, &media_addr, sizeof(*addr))) {
- char addr_string[16];
-
- warn("New bearer address for %s\n",
- addr_string_fill(addr_string, orig));
+ warn("Resetting link <%s>, peer interface address changed\n",
+ link->name);
memcpy(addr, &media_addr, sizeof(*addr));
tipc_link_reset(link);
}
diff --git a/net/tipc/link.c b/net/tipc/link.c
index 2efced5..d7668b8 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -419,7 +419,7 @@ struct link *tipc_link_create(struct bea
l_ptr = (struct link *)kmalloc(sizeof(*l_ptr), GFP_ATOMIC);
if (!l_ptr) {
- warn("Memory squeeze; Failed to create link\n");
+ warn("Link creation failed, no memory\n");
return NULL;
}
memset(l_ptr, 0, sizeof(*l_ptr));
@@ -469,7 +469,7 @@ struct link *tipc_link_create(struct bea
if (!pb) {
kfree(l_ptr);
- warn("Memory squeeze; Failed to create link\n");
+ warn("Link creation failed, no memory for print buffer\n");
return NULL;
}
tipc_printbuf_init(&l_ptr->print_buf, pb, LINK_LOG_BUF_SIZE);
@@ -819,6 +819,8 @@ static void link_state_event(struct link
break;
case RESET_MSG:
dbg_link("RES -> RR\n");
+ info("Resetting link <%s>, requested by peer\n",
+ l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_RESET;
l_ptr->fsm_msg_cnt = 0;
@@ -843,6 +845,8 @@ static void link_state_event(struct link
break;
case RESET_MSG:
dbg_link("RES -> RR\n");
+ info("Resetting link <%s>, requested by peer "
+ "while probing\n", l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_RESET;
l_ptr->fsm_msg_cnt = 0;
@@ -874,6 +878,8 @@ static void link_state_event(struct link
} else { /* Link has failed */
dbg_link("-> RU (%u probes unanswered)\n",
l_ptr->fsm_msg_cnt);
+ warn("Resetting link <%s>, peer not responding\n",
+ l_ptr->name);
tipc_link_reset(l_ptr);
l_ptr->state = RESET_UNKNOWN;
l_ptr->fsm_msg_cnt = 0;
@@ -1049,7 +1055,7 @@ int tipc_link_send_buf(struct link *l_pt
msg_dbg(msg, "TIPC: Congestion, throwing away\n");
buf_discard(buf);
if (imp > CONN_MANAGER) {
- warn("Resetting <%s>, send queue full", l_ptr->name);
+ warn("Resetting link <%s>, send queue full", l_ptr->name);
tipc_link_reset(l_ptr);
}
return dsz;
@@ -2228,7 +2234,7 @@ static void link_recv_proto_msg(struct l
if (msg_linkprio(msg) &&
(msg_linkprio(msg) != l_ptr->priority)) {
- warn("Changing prio <%s>: %u->%u\n",
+ warn("Resetting link <%s>, priority change %u->%u\n",
l_ptr->name, l_ptr->priority, msg_linkprio(msg));
l_ptr->priority = msg_linkprio(msg);
tipc_link_reset(l_ptr); /* Enforce change to take effect */
@@ -2348,7 +2354,8 @@ void tipc_link_changeover(struct link *l
msg_dbg(&tunnel_hdr, "EMPTY>SEND>");
tipc_link_send_buf(tunnel, buf);
} else {
- warn("Memory squeeze; link changeover failed\n");
+ warn("Link changeover error, "
+ "unable to send changeover msg\n");
}
return;
}
@@ -2398,7 +2405,8 @@ void tipc_link_send_duplicate(struct lin
msg_set_size(&tunnel_hdr, length + INT_H_SIZE);
outbuf = buf_acquire(length + INT_H_SIZE);
if (outbuf == NULL) {
- warn("Memory squeeze; buffer duplication failed\n");
+ warn("Link changeover error, "
+ "unable to send duplicate msg\n");
return;
}
memcpy(outbuf->data, (unchar *)&tunnel_hdr, INT_H_SIZE);
@@ -2473,7 +2481,7 @@ static int link_recv_changeover_msg(stru
}
*buf = buf_extract(tunnel_buf,INT_H_SIZE);
if (*buf == NULL) {
- warn("Memory squeeze; failed to extract msg\n");
+ warn("Link changeover error, duplicate msg dropped\n");
goto exit;
}
msg_dbg(tunnel_msg, "TNL<REC<");
@@ -2485,6 +2493,8 @@ static int link_recv_changeover_msg(stru
if (tipc_link_is_up(dest_link)) {
msg_dbg(tunnel_msg, "UP/FIRST/<REC<");
+ info("Resetting link <%s>, changeover initiated by peer\n",
+ dest_link->name);
tipc_link_reset(dest_link);
dest_link->exp_msg_count = msg_count;
if (!msg_count)
@@ -2514,7 +2524,7 @@ static int link_recv_changeover_msg(stru
buf_discard(tunnel_buf);
return 1;
} else {
- warn("Memory squeeze; dropped incoming msg\n");
+ warn("Link changeover error, original msg dropped\n");
}
}
exit:
@@ -2536,13 +2546,8 @@ void tipc_link_recv_bundle(struct sk_buf
while (msgcount--) {
obuf = buf_extract(buf, pos);
if (obuf == NULL) {
- char addr_string[16];
-
- warn("Buffer allocation failure;\n");
- warn(" incoming message(s) from %s lost\n",
- addr_string_fill(addr_string,
- msg_orignode(buf_msg(buf))));
- return;
+ warn("Link unable to unbundle message(s)\n");
+ break;
};
pos += align(msg_size(buf_msg(obuf)));
msg_dbg(buf_msg(obuf), " /");
@@ -2600,7 +2605,7 @@ int tipc_link_send_long_buf(struct link
}
fragm = buf_acquire(fragm_sz + INT_H_SIZE);
if (fragm == NULL) {
- warn("Memory squeeze; failed to fragment msg\n");
+ warn("Link unable to fragment message\n");
dsz = -ENOMEM;
goto exit;
}
@@ -2715,7 +2720,7 @@ int tipc_link_recv_fragment(struct sk_bu
set_fragm_size(pbuf,fragm_sz);
set_expected_frags(pbuf,exp_fragm_cnt - 1);
} else {
- warn("Memory squeeze; got no defragmenting buffer\n");
+ warn("Link unable to reassemble fragmented message\n");
}
buf_discard(fbuf);
return 0;
diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c
index 5718ecb..f0b063b 100644
--- a/net/tipc/name_distr.c
+++ b/net/tipc/name_distr.c
@@ -127,7 +127,7 @@ void tipc_named_publish(struct publicati
buf = named_prepare_buf(PUBLICATION, ITEM_SIZE, 0);
if (!buf) {
- warn("Memory squeeze; failed to distribute publication\n");
+ warn("Publication distribution failure\n");
return;
}
@@ -151,7 +151,7 @@ void tipc_named_withdraw(struct publicat
buf = named_prepare_buf(WITHDRAWAL, ITEM_SIZE, 0);
if (!buf) {
- warn("Memory squeeze; failed to distribute withdrawal\n");
+ warn("Withdrawl distribution failure\n");
return;
}
@@ -184,8 +184,8 @@ void tipc_named_node_up(unsigned long no
left = (rest <= max_item_buf) ? rest : max_item_buf;
rest -= left;
buf = named_prepare_buf(PUBLICATION, left, node);
- if (buf == NULL) {
- warn("Memory Squeeze; could not send publication\n");
+ if (!buf) {
+ warn("Bulk publication distribution failure\n");
goto exit;
}
item = (struct distr_item *)msg_data(buf_msg(buf));
@@ -291,7 +291,7 @@ void tipc_named_recv(struct sk_buff *buf
ntohl(item->ref), ntohl(item->key));
}
} else {
- warn("tipc_named_recv: unknown msg\n");
+ warn("Unrecognized name table message received\n");
}
item++;
}
diff --git a/net/tipc/name_table.c b/net/tipc/name_table.c
index e90dc80..3857130 100644
--- a/net/tipc/name_table.c
+++ b/net/tipc/name_table.c
@@ -120,7 +120,7 @@ static struct publication *publ_create(u
struct publication *publ =
(struct publication *)kmalloc(sizeof(*publ), GFP_ATOMIC);
if (publ == NULL) {
- warn("Memory squeeze; failed to create publication\n");
+ warn("Publication creation failure, no memory\n");
return NULL;
}
@@ -165,7 +165,7 @@ static struct name_seq *tipc_nameseq_cre
struct sub_seq *sseq = tipc_subseq_alloc(1);
if (!nseq || !sseq) {
- warn("Memory squeeze; failed to create name sequence\n");
+ warn("Name sequence creation failed, no memory\n");
kfree(nseq);
kfree(sseq);
return NULL;
@@ -759,7 +759,7 @@ struct publication *tipc_nametbl_publish
struct publication *publ;
if (table.local_publ_count >= tipc_max_publications) {
- warn("Failed publish: max %u local publication\n",
+ warn("Publication failed, local publication limit reached (%u)\n",
tipc_max_publications);
return NULL;
}
diff --git a/net/tipc/node.c b/net/tipc/node.c
index b54462b..5f09754 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -61,34 +61,37 @@ struct node *tipc_node_create(u32 addr)
struct node **curr_node;
n_ptr = kmalloc(sizeof(*n_ptr),GFP_ATOMIC);
- if (n_ptr != NULL) {
- memset(n_ptr, 0, sizeof(*n_ptr));
- n_ptr->addr = addr;
- n_ptr->lock = SPIN_LOCK_UNLOCKED;
- INIT_LIST_HEAD(&n_ptr->nsub);
-
- c_ptr = tipc_cltr_find(addr);
- if (c_ptr == NULL)
- c_ptr = tipc_cltr_create(addr);
- if (c_ptr != NULL) {
- n_ptr->owner = c_ptr;
- tipc_cltr_attach_node(c_ptr, n_ptr);
- n_ptr->last_router = -1;
-
- /* Insert node into ordered list */
- for (curr_node = &tipc_nodes; *curr_node;
- curr_node = &(*curr_node)->next) {
- if (addr < (*curr_node)->addr) {
- n_ptr->next = *curr_node;
- break;
- }
- }
- (*curr_node) = n_ptr;
- } else {
- kfree(n_ptr);
- n_ptr = NULL;
- }
- }
+ if (!n_ptr) {
+ warn("Node creation failed, no memory\n");
+ return NULL;
+ }
+
+ c_ptr = tipc_cltr_find(addr);
+ if (!c_ptr) {
+ c_ptr = tipc_cltr_create(addr);
+ }
+ if (!c_ptr) {
+ kfree(n_ptr);
+ return NULL;
+ }
+
+ memset(n_ptr, 0, sizeof(*n_ptr));
+ n_ptr->addr = addr;
+ n_ptr->lock = SPIN_LOCK_UNLOCKED;
+ INIT_LIST_HEAD(&n_ptr->nsub);
+ n_ptr->owner = c_ptr;
+ tipc_cltr_attach_node(c_ptr, n_ptr);
+ n_ptr->last_router = -1;
+
+ /* Insert node into ordered list */
+ for (curr_node = &tipc_nodes; *curr_node;
+ curr_node = &(*curr_node)->next) {
+ if (addr < (*curr_node)->addr) {
+ n_ptr->next = *curr_node;
+ break;
+ }
+ }
+ (*curr_node) = n_ptr;
return n_ptr;
}
@@ -132,7 +135,7 @@ void tipc_node_link_up(struct node *n_pt
return;
}
if (l_ptr->priority < active[0]->priority) {
- info("Link is standby\n");
+ info("New link <%s> becomes standby\n", l_ptr->name);
return;
}
tipc_link_send_duplicate(active[0], l_ptr);
@@ -140,8 +143,9 @@ void tipc_node_link_up(struct node *n_pt
active[0] = l_ptr;
return;
}
- info("Link <%s> on network plane %c becomes standby\n",
- active[0]->name, active[0]->b_ptr->net_plane);
+ info("Old link <%s> becomes standby\n", active[0]->name);
+ if (active[1] != active[0])
+ info("Old link <%s> becomes standby\n", active[1]->name);
active[0] = active[1] = l_ptr;
}
@@ -248,7 +252,7 @@ struct node *tipc_node_attach_link(struc
n_ptr->link_cnt++;
return n_ptr;
}
- err("Attempt to establish second link on <%s> to <%s> \n",
+ err("Attempt to establish second link on <%s> to %s \n",
l_ptr->b_ptr->publ.name,
addr_string_fill(addr_string, l_ptr->addr));
}
diff --git a/net/tipc/port.c b/net/tipc/port.c
index 3aab67a..47d9740 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -195,7 +195,7 @@ void tipc_port_recv_mcast(struct sk_buff
struct sk_buff *b = skb_clone(buf, GFP_ATOMIC);
if (b == NULL) {
- warn("Buffer allocation failure\n");
+ warn("Unable to deliver multicast message(s)\n");
msg_dbg(msg, "LOST:");
goto exit;
}
@@ -227,14 +227,14 @@ u32 tipc_createport_raw(void *usr_handle
u32 ref;
p_ptr = kmalloc(sizeof(*p_ptr), GFP_ATOMIC);
- if (p_ptr == NULL) {
- warn("Memory squeeze; failed to create port\n");
+ if (!p_ptr) {
+ warn("Port creation failed, no memory\n");
return 0;
}
memset(p_ptr, 0, sizeof(*p_ptr));
ref = tipc_ref_acquire(p_ptr, &p_ptr->publ.lock);
if (!ref) {
- warn("Reference Table Exhausted\n");
+ warn("Port creation failed, reference table exhausted\n");
kfree(p_ptr);
return 0;
}
@@ -1059,7 +1059,7 @@ int tipc_createport(u32 user_ref,
u32 ref;
up_ptr = (struct user_port *)kmalloc(sizeof(*up_ptr), GFP_ATOMIC);
- if (up_ptr == NULL) {
+ if (!up_ptr) {
warn("Port creation failed, no memory\n");
return -ENOMEM;
}
diff --git a/net/tipc/subscr.c b/net/tipc/subscr.c
index c5f026c..fc17187 100644
--- a/net/tipc/subscr.c
+++ b/net/tipc/subscr.c
@@ -266,7 +266,8 @@ static void subscr_subscribe(struct tipc
/* Refuse subscription if global limit exceeded */
if (atomic_read(&topsrv.subscription_count) >= tipc_max_subscriptions) {
- warn("Failed: max %u subscriptions\n", tipc_max_subscriptions);
+ warn("Subscription rejected, subscription limit reached (%u)\n",
+ tipc_max_subscriptions);
subscr_terminate(subscriber);
return;
}
@@ -274,8 +275,8 @@ static void subscr_subscribe(struct tipc
/* Allocate subscription object */
sub = kmalloc(sizeof(*sub), GFP_ATOMIC);
- if (sub == NULL) {
- warn("Memory squeeze; ignoring subscription\n");
+ if (!sub) {
+ warn("Subscription rejected, no memory\n");
subscr_terminate(subscriber);
return;
}
@@ -298,8 +299,7 @@ static void subscr_subscribe(struct tipc
if ((((sub->filter != TIPC_SUB_PORTS)
&& (sub->filter != TIPC_SUB_SERVICE)))
|| (sub->seq.lower > sub->seq.upper)) {
- warn("Rejecting illegal subscription %u,%u,%u\n",
- sub->seq.type, sub->seq.lower, sub->seq.upper);
+ warn("Subscription rejected, illegal request\n");
kfree(sub);
subscr_terminate(subscriber);
return;
@@ -387,7 +387,7 @@ static void subscr_named_msg_event(void
dbg("subscr_named_msg_event: orig = %x own = %x,\n",
orig->node, tipc_own_addr);
if (size && (size != sizeof(struct tipc_subscr))) {
- warn("Received tipc_subscr of invalid size\n");
+ warn("Subscriber rejected, invalid subscription size\n");
return;
}
@@ -395,7 +395,7 @@ static void subscr_named_msg_event(void
subscriber = kmalloc(sizeof(struct subscriber), GFP_ATOMIC);
if (subscriber == NULL) {
- warn("Memory squeeze; ignoring subscriber setup\n");
+ warn("Subscriber rejected, no memory\n");
return;
}
memset(subscriber, 0, sizeof(struct subscriber));
@@ -403,7 +403,7 @@ static void subscr_named_msg_event(void
INIT_LIST_HEAD(&subscriber->subscriber_list);
subscriber->ref = tipc_ref_acquire(subscriber, &subscriber->lock);
if (subscriber->ref == 0) {
- warn("Failed to acquire subscriber reference\n");
+ warn("Subscriber rejected, reference table exhausted\n");
kfree(subscriber);
return;
}
@@ -422,7 +422,7 @@ static void subscr_named_msg_event(void
NULL,
&subscriber->port_ref);
if (subscriber->port_ref == 0) {
- warn("Memory squeeze; failed to create subscription port\n");
+ warn("Subscriber rejected, unable to create port\n");
tipc_ref_discard(subscriber->ref);
kfree(subscriber);
return;
diff --git a/net/tipc/zone.c b/net/tipc/zone.c
index 2803e1b..316c487 100644
--- a/net/tipc/zone.c
+++ b/net/tipc/zone.c
@@ -44,19 +44,24 @@ #include "node.h"
struct _zone *tipc_zone_create(u32 addr)
{
- struct _zone *z_ptr = NULL;
+ struct _zone *z_ptr;
u32 z_num;
- if (!tipc_addr_domain_valid(addr))
+ if (!tipc_addr_domain_valid(addr)) {
+ err("Zone creation failed, invalid domain 0x%x\n", addr);
return NULL;
+ }
z_ptr = (struct _zone *)kmalloc(sizeof(*z_ptr), GFP_ATOMIC);
- if (z_ptr != NULL) {
- memset(z_ptr, 0, sizeof(*z_ptr));
- z_num = tipc_zone(addr);
- z_ptr->addr = tipc_addr(z_num, 0, 0);
- tipc_net.zones[z_num] = z_ptr;
+ if (!z_ptr) {
+ warn("Zone creation failed, insufficient memory\n");
+ return NULL;
}
+
+ memset(z_ptr, 0, sizeof(*z_ptr));
+ z_num = tipc_zone(addr);
+ z_ptr->addr = tipc_addr(z_num, 0, 0);
+ tipc_net.zones[z_num] = z_ptr;
return z_ptr;
}
--
1.4.0
^ permalink raw reply related
* [PATCH 28/32] [TIPC] First phase of assert() cleanup
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
This also contains enhancements to simplify comparisons in name table
publication removal algorithm and to simplify name table sanity checking
when shutting down TIPC.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/link.c | 13 ++-
net/tipc/name_distr.c | 20 +++++
net/tipc/name_table.c | 179 +++++++++++++++++++++++++-----------------------
net/tipc/node.c | 3 -
net/tipc/node_subscr.c | 15 ++--
net/tipc/port.c | 1
net/tipc/ref.c | 31 +++++++-
7 files changed, 153 insertions(+), 109 deletions(-)
diff --git a/net/tipc/link.c b/net/tipc/link.c
index ff40c91..2efced5 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -574,7 +574,6 @@ void tipc_link_wakeup_ports(struct link
break;
list_del_init(&p_ptr->wait_list);
p_ptr->congested_link = NULL;
- assert(p_ptr->wakeup);
spin_lock_bh(p_ptr->publ.lock);
p_ptr->publ.congested = 0;
p_ptr->wakeup(&p_ptr->publ);
@@ -1246,8 +1245,6 @@ int tipc_link_send_sections_fast(struct
int res;
u32 selector = msg_origport(hdr) & 1;
- assert(destaddr != tipc_own_addr);
-
again:
/*
* Try building message using port's max_pkt hint.
@@ -2310,7 +2307,6 @@ void tipc_link_tunnel(struct link *l_ptr
memcpy(buf->data + INT_H_SIZE, (unchar *)msg, length);
dbg("%c->%c:", l_ptr->b_ptr->net_plane, tunnel->b_ptr->net_plane);
msg_dbg(buf_msg(buf), ">SEND>");
- assert(tunnel);
tipc_link_send_buf(tunnel, buf);
}
@@ -2339,10 +2335,10 @@ void tipc_link_changeover(struct link *l
ORIGINAL_MSG, TIPC_OK, INT_H_SIZE, l_ptr->addr);
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
msg_set_msgcnt(&tunnel_hdr, msgcount);
+
if (!l_ptr->first_out) {
struct sk_buff *buf;
- assert(!msgcount);
buf = buf_acquire(INT_H_SIZE);
if (buf) {
memcpy(buf->data, (unchar *)&tunnel_hdr, INT_H_SIZE);
@@ -2356,6 +2352,7 @@ void tipc_link_changeover(struct link *l
}
return;
}
+
while (crs) {
struct tipc_msg *msg = buf_msg(crs);
@@ -2455,11 +2452,15 @@ static int link_recv_changeover_msg(stru
u32 msg_count = msg_msgcnt(tunnel_msg);
dest_link = (*l_ptr)->owner->links[msg_bearer_id(tunnel_msg)];
- assert(dest_link != *l_ptr);
if (!dest_link) {
msg_dbg(tunnel_msg, "NOLINK/<REC<");
goto exit;
}
+ if (dest_link == *l_ptr) {
+ err("Unexpected changeover message on link <%s>\n",
+ (*l_ptr)->name);
+ goto exit;
+ }
dbg("%c<-%c:", dest_link->b_ptr->net_plane,
(*l_ptr)->b_ptr->net_plane);
*l_ptr = dest_link;
diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c
index a3bbc89..5718ecb 100644
--- a/net/tipc/name_distr.c
+++ b/net/tipc/name_distr.c
@@ -174,7 +174,6 @@ void tipc_named_node_up(unsigned long no
u32 rest;
u32 max_item_buf;
- assert(in_own_cluster(node));
read_lock_bh(&tipc_nametbl_lock);
max_item_buf = TIPC_MAX_USER_MSG_SIZE / ITEM_SIZE;
max_item_buf *= ITEM_SIZE;
@@ -221,15 +220,24 @@ exit:
static void node_is_down(struct publication *publ)
{
struct publication *p;
+
write_lock_bh(&tipc_nametbl_lock);
dbg("node_is_down: withdrawing %u, %u, %u\n",
publ->type, publ->lower, publ->upper);
publ->key += 1222345;
p = tipc_nametbl_remove_publ(publ->type, publ->lower,
publ->node, publ->ref, publ->key);
- assert(p == publ);
write_unlock_bh(&tipc_nametbl_lock);
- kfree(publ);
+
+ if (p != publ) {
+ err("Unable to remove publication from failed node\n"
+ "(type=%u, lower=%u, node=0x%x, ref=%u, key=%u)\n",
+ publ->type, publ->lower, publ->node, publ->ref, publ->key);
+ }
+
+ if (p) {
+ kfree(p);
+ }
}
/**
@@ -275,6 +283,12 @@ void tipc_named_recv(struct sk_buff *buf
if (publ) {
tipc_nodesub_unsubscribe(&publ->subscr);
kfree(publ);
+ } else {
+ err("Unable to remove publication by node 0x%x\n"
+ "(type=%u, lower=%u, ref=%u, key=%u)\n",
+ msg_orignode(msg),
+ ntohl(item->type), ntohl(item->lower),
+ ntohl(item->ref), ntohl(item->key));
}
} else {
warn("tipc_named_recv: unknown msg\n");
diff --git a/net/tipc/name_table.c b/net/tipc/name_table.c
index 0511436..e90dc80 100644
--- a/net/tipc/name_table.c
+++ b/net/tipc/name_table.c
@@ -71,7 +71,7 @@ struct sub_seq {
* @sseq: pointer to dynamically-sized array of sub-sequences of this 'type';
* sub-sequences are sorted in ascending order
* @alloc: number of sub-sequences currently in array
- * @first_free: upper bound of highest sub-sequence + 1
+ * @first_free: array index of first unused sub-sequence entry
* @ns_list: links to adjacent name sequences in hash chain
* @subscriptions: list of subscriptions for this 'type'
* @lock: spinlock controlling access to name sequence structure
@@ -175,7 +175,7 @@ static struct name_seq *tipc_nameseq_cre
nseq->lock = SPIN_LOCK_UNLOCKED;
nseq->type = type;
nseq->sseqs = sseq;
- dbg("tipc_nameseq_create() nseq = %x type %u, ssseqs %x, ff: %u\n",
+ dbg("tipc_nameseq_create(): nseq = %p, type %u, ssseqs %p, ff: %u\n",
nseq, type, nseq->sseqs, nseq->first_free);
nseq->alloc = 1;
INIT_HLIST_NODE(&nseq->ns_list);
@@ -253,16 +253,16 @@ static struct publication *tipc_nameseq_
struct sub_seq *sseq;
int created_subseq = 0;
- assert(nseq->first_free <= nseq->alloc);
sseq = nameseq_find_subseq(nseq, lower);
- dbg("nameseq_ins: for seq %x,<%u,%u>, found sseq %x\n",
+ dbg("nameseq_ins: for seq %p, {%u,%u}, found sseq %p\n",
nseq, type, lower, sseq);
if (sseq) {
/* Lower end overlaps existing entry => need an exact match */
if ((sseq->lower != lower) || (sseq->upper != upper)) {
- warn("Overlapping publ <%u,%u,%u>\n", type, lower, upper);
+ warn("Cannot publish {%u,%u,%u}, overlap error\n",
+ type, lower, upper);
return NULL;
}
} else {
@@ -277,7 +277,8 @@ static struct publication *tipc_nameseq_
if ((inspos < nseq->first_free) &&
(upper >= nseq->sseqs[inspos].lower)) {
- warn("Overlapping publ <%u,%u,%u>\n", type, lower, upper);
+ warn("Cannot publish {%u,%u,%u}, overlap error\n",
+ type, lower, upper);
return NULL;
}
@@ -287,7 +288,8 @@ static struct publication *tipc_nameseq_
struct sub_seq *sseqs = tipc_subseq_alloc(nseq->alloc * 2);
if (!sseqs) {
- warn("Memory squeeze; failed to create sub-sequence\n");
+ warn("Cannot publish {%u,%u,%u}, no memory\n",
+ type, lower, upper);
return NULL;
}
dbg("Allocated %u more sseqs\n", nseq->alloc);
@@ -311,7 +313,7 @@ static struct publication *tipc_nameseq_
sseq->upper = upper;
created_subseq = 1;
}
- dbg("inserting (%u %u %u) from %x:%u into sseq %x(%u,%u) of seq %x\n",
+ dbg("inserting {%u,%u,%u} from <0x%x:%u> into sseq %p(%u,%u) of seq %p\n",
type, lower, upper, node, port, sseq,
sseq->lower, sseq->upper, nseq);
@@ -320,7 +322,7 @@ static struct publication *tipc_nameseq_
publ = publ_create(type, lower, upper, scope, node, port, key);
if (!publ)
return NULL;
- dbg("inserting publ %x, node=%x publ->node=%x, subscr->node=%x\n",
+ dbg("inserting publ %p, node=0x%x publ->node=0x%x, subscr->node=%p\n",
publ, node, publ->node, publ->subscr.node);
if (!sseq->zone_list)
@@ -367,45 +369,47 @@ static struct publication *tipc_nameseq_
/**
* tipc_nameseq_remove_publ -
+ *
+ * NOTE: There may be cases where TIPC is asked to remove a publication
+ * that is not in the name table. For example, if another node issues a
+ * publication for a name sequence that overlaps an existing name sequence
+ * the publication will not be recorded, which means the publication won't
+ * be found when the name sequence is later withdrawn by that node.
+ * A failed withdraw request simply returns a failure indication and lets the
+ * caller issue any error or warning messages associated with such a problem.
*/
static struct publication *tipc_nameseq_remove_publ(struct name_seq *nseq, u32 inst,
u32 node, u32 ref, u32 key)
{
struct publication *publ;
+ struct publication *curr;
struct publication *prev;
struct sub_seq *sseq = nameseq_find_subseq(nseq, inst);
struct sub_seq *free;
struct subscription *s, *st;
int removed_subseq = 0;
- assert(nseq);
-
- if (!sseq) {
- int i;
-
- warn("Withdraw unknown <%u,%u>?\n", nseq->type, inst);
- assert(nseq->sseqs);
- dbg("Dumping subseqs %x for %x, alloc = %u,ff=%u\n",
- nseq->sseqs, nseq, nseq->alloc,
- nseq->first_free);
- for (i = 0; i < nseq->first_free; i++) {
- dbg("Subseq %u(%x): lower = %u,upper = %u\n",
- i, &nseq->sseqs[i], nseq->sseqs[i].lower,
- nseq->sseqs[i].upper);
- }
+ if (!sseq)
return NULL;
- }
- dbg("nameseq_remove: seq: %x, sseq %x, <%u,%u> key %u\n",
+
+ dbg("tipc_nameseq_remove_publ: seq: %p, sseq %p, {%u,%u}, key %u\n",
nseq, sseq, nseq->type, inst, key);
+ /* Remove publication from zone scope list */
+
prev = sseq->zone_list;
publ = sseq->zone_list->zone_list_next;
while ((publ->key != key) || (publ->ref != ref) ||
(publ->node && (publ->node != node))) {
prev = publ;
publ = publ->zone_list_next;
- assert(prev != sseq->zone_list);
+ if (prev == sseq->zone_list) {
+
+ /* Prevent endless loop if publication not found */
+
+ return NULL;
+ }
}
if (publ != sseq->zone_list)
prev->zone_list_next = publ->zone_list_next;
@@ -416,14 +420,24 @@ static struct publication *tipc_nameseq_
sseq->zone_list = NULL;
}
+ /* Remove publication from cluster scope list, if present */
+
if (in_own_cluster(node)) {
prev = sseq->cluster_list;
- publ = sseq->cluster_list->cluster_list_next;
- while ((publ->key != key) || (publ->ref != ref) ||
- (publ->node && (publ->node != node))) {
- prev = publ;
- publ = publ->cluster_list_next;
- assert(prev != sseq->cluster_list);
+ curr = sseq->cluster_list->cluster_list_next;
+ while (curr != publ) {
+ prev = curr;
+ curr = curr->cluster_list_next;
+ if (prev == sseq->cluster_list) {
+
+ /* Prevent endless loop for malformed list */
+
+ err("Unable to de-list cluster publication\n"
+ "{%u%u}, node=0x%x, ref=%u, key=%u)\n",
+ publ->type, publ->lower, publ->node,
+ publ->ref, publ->key);
+ goto end_cluster;
+ }
}
if (publ != sseq->cluster_list)
prev->cluster_list_next = publ->cluster_list_next;
@@ -434,15 +448,26 @@ static struct publication *tipc_nameseq_
sseq->cluster_list = NULL;
}
}
+end_cluster:
+
+ /* Remove publication from node scope list, if present */
if (node == tipc_own_addr) {
prev = sseq->node_list;
- publ = sseq->node_list->node_list_next;
- while ((publ->key != key) || (publ->ref != ref) ||
- (publ->node && (publ->node != node))) {
- prev = publ;
- publ = publ->node_list_next;
- assert(prev != sseq->node_list);
+ curr = sseq->node_list->node_list_next;
+ while (curr != publ) {
+ prev = curr;
+ curr = curr->node_list_next;
+ if (prev == sseq->node_list) {
+
+ /* Prevent endless loop for malformed list */
+
+ err("Unable to de-list node publication\n"
+ "{%u%u}, node=0x%x, ref=%u, key=%u)\n",
+ publ->type, publ->lower, publ->node,
+ publ->ref, publ->key);
+ goto end_node;
+ }
}
if (publ != sseq->node_list)
prev->node_list_next = publ->node_list_next;
@@ -453,22 +478,18 @@ static struct publication *tipc_nameseq_
sseq->node_list = NULL;
}
}
- assert(!publ->node || (publ->node == node));
- assert(publ->ref == ref);
- assert(publ->key == key);
+end_node:
- /*
- * Contract subseq list if no more publications:
- */
- if (!sseq->node_list && !sseq->cluster_list && !sseq->zone_list) {
+ /* Contract subseq list if no more publications for that subseq */
+
+ if (!sseq->zone_list) {
free = &nseq->sseqs[nseq->first_free--];
memmove(sseq, sseq + 1, (free - (sseq + 1)) * sizeof (*sseq));
removed_subseq = 1;
}
- /*
- * Any subscriptions waiting ?
- */
+ /* Notify any waiting subscriptions */
+
list_for_each_entry_safe(s, st, &nseq->subscriptions, nameseq_list) {
tipc_subscr_report_overlap(s,
publ->lower,
@@ -478,6 +499,7 @@ static struct publication *tipc_nameseq_
publ->node,
removed_subseq);
}
+
return publ;
}
@@ -530,7 +552,7 @@ static struct name_seq *nametbl_find_seq
seq_head = &table.types[hash(type)];
hlist_for_each_entry(ns, seq_node, seq_head, ns_list) {
if (ns->type == type) {
- dbg("found %x\n", ns);
+ dbg("found %p\n", ns);
return ns;
}
}
@@ -543,22 +565,21 @@ struct publication *tipc_nametbl_insert_
{
struct name_seq *seq = nametbl_find_seq(type);
- dbg("ins_publ: <%u,%x,%x> found %x\n", type, lower, upper, seq);
+ dbg("tipc_nametbl_insert_publ: {%u,%u,%u} found %p\n", type, lower, upper, seq);
if (lower > upper) {
- warn("Failed to publish illegal <%u,%u,%u>\n",
+ warn("Failed to publish illegal {%u,%u,%u}\n",
type, lower, upper);
return NULL;
}
- dbg("Publishing <%u,%u,%u> from %x\n", type, lower, upper, node);
+ dbg("Publishing {%u,%u,%u} from 0x%x\n", type, lower, upper, node);
if (!seq) {
seq = tipc_nameseq_create(type, &table.types[hash(type)]);
- dbg("tipc_nametbl_insert_publ: created %x\n", seq);
+ dbg("tipc_nametbl_insert_publ: created %p\n", seq);
}
if (!seq)
return NULL;
- assert(seq->type == type);
return tipc_nameseq_insert_publ(seq, type, lower, upper,
scope, node, port, key);
}
@@ -572,7 +593,7 @@ struct publication *tipc_nametbl_remove_
if (!seq)
return NULL;
- dbg("Withdrawing <%u,%u> from %x\n", type, lower, node);
+ dbg("Withdrawing {%u,%u} from 0x%x\n", type, lower, node);
publ = tipc_nameseq_remove_publ(seq, lower, node, ref, key);
if (!seq->first_free && list_empty(&seq->subscriptions)) {
@@ -743,7 +764,7 @@ struct publication *tipc_nametbl_publish
return NULL;
}
if ((type < TIPC_RESERVED_TYPES) && !atomic_read(&rsv_publ_ok)) {
- warn("Failed to publish reserved name <%u,%u,%u>\n",
+ warn("Publication failed, reserved name {%u,%u,%u}\n",
type, lower, upper);
return NULL;
}
@@ -767,10 +788,10 @@ int tipc_nametbl_withdraw(u32 type, u32
{
struct publication *publ;
- dbg("tipc_nametbl_withdraw:<%d,%d,%d>\n", type, lower, key);
+ dbg("tipc_nametbl_withdraw: {%u,%u}, key=%u\n", type, lower, key);
write_lock_bh(&tipc_nametbl_lock);
publ = tipc_nametbl_remove_publ(type, lower, tipc_own_addr, ref, key);
- if (publ) {
+ if (likely(publ)) {
table.local_publ_count--;
if (publ->scope != TIPC_NODE_SCOPE)
tipc_named_withdraw(publ);
@@ -780,6 +801,9 @@ int tipc_nametbl_withdraw(u32 type, u32
return 1;
}
write_unlock_bh(&tipc_nametbl_lock);
+ err("Unable to remove local publication\n"
+ "(type=%u, lower=%u, ref=%u, key=%u)\n",
+ type, lower, ref, key);
return 0;
}
@@ -787,8 +811,7 @@ int tipc_nametbl_withdraw(u32 type, u32
* tipc_nametbl_subscribe - add a subscription object to the name table
*/
-void
-tipc_nametbl_subscribe(struct subscription *s)
+void tipc_nametbl_subscribe(struct subscription *s)
{
u32 type = s->seq.type;
struct name_seq *seq;
@@ -800,11 +823,13 @@ tipc_nametbl_subscribe(struct subscripti
}
if (seq){
spin_lock_bh(&seq->lock);
- dbg("tipc_nametbl_subscribe:found %x for <%u,%u,%u>\n",
+ dbg("tipc_nametbl_subscribe:found %p for {%u,%u,%u}\n",
seq, type, s->seq.lower, s->seq.upper);
- assert(seq->type == type);
tipc_nameseq_subscribe(seq, s);
spin_unlock_bh(&seq->lock);
+ } else {
+ warn("Failed to create subscription for {%u,%u,%u}\n",
+ s->seq.type, s->seq.lower, s->seq.upper);
}
write_unlock_bh(&tipc_nametbl_lock);
}
@@ -813,8 +838,7 @@ tipc_nametbl_subscribe(struct subscripti
* tipc_nametbl_unsubscribe - remove a subscription object from name table
*/
-void
-tipc_nametbl_unsubscribe(struct subscription *s)
+void tipc_nametbl_unsubscribe(struct subscription *s)
{
struct name_seq *seq;
@@ -1049,35 +1073,20 @@ int tipc_nametbl_init(void)
void tipc_nametbl_stop(void)
{
- struct hlist_head *seq_head;
- struct hlist_node *seq_node;
- struct hlist_node *tmp;
- struct name_seq *seq;
u32 i;
if (!table.types)
return;
+ /* Verify name table is empty, then release it */
+
write_lock_bh(&tipc_nametbl_lock);
for (i = 0; i < tipc_nametbl_size; i++) {
- seq_head = &table.types[i];
- hlist_for_each_entry_safe(seq, seq_node, tmp, seq_head, ns_list) {
- struct sub_seq *sseq = seq->sseqs;
-
- for (; sseq != &seq->sseqs[seq->first_free]; sseq++) {
- struct publication *publ = sseq->zone_list;
- assert(publ);
- do {
- struct publication *next =
- publ->zone_list_next;
- kfree(publ);
- publ = next;
- }
- while (publ != sseq->zone_list);
- }
- }
+ if (!hlist_empty(&table.types[i]))
+ err("tipc_nametbl_stop(): hash chain %u is non-null\n", i);
}
kfree(table.types);
table.types = NULL;
write_unlock_bh(&tipc_nametbl_lock);
}
+
diff --git a/net/tipc/node.c b/net/tipc/node.c
index 0d5db06..b54462b 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -234,7 +234,6 @@ struct node *tipc_node_attach_link(struc
u32 bearer_id = l_ptr->b_ptr->identity;
char addr_string[16];
- assert(bearer_id < MAX_BEARERS);
if (n_ptr->link_cnt >= 2) {
char addr_string[16];
@@ -314,7 +313,7 @@ static void node_established_contact(str
struct cluster *c_ptr;
dbg("node_established_contact:-> %x\n", n_ptr->addr);
- if (!tipc_node_has_active_routes(n_ptr)) {
+ if (!tipc_node_has_active_routes(n_ptr) && in_own_cluster(n_ptr->addr)) {
tipc_k_signal((Handler)tipc_named_node_up, n_ptr->addr);
}
diff --git a/net/tipc/node_subscr.c b/net/tipc/node_subscr.c
index cff4068..cc3fff3 100644
--- a/net/tipc/node_subscr.c
+++ b/net/tipc/node_subscr.c
@@ -47,18 +47,19 @@ #include "addr.h"
void tipc_nodesub_subscribe(struct node_subscr *node_sub, u32 addr,
void *usr_handle, net_ev_handler handle_down)
{
- node_sub->node = NULL;
- if (addr == tipc_own_addr)
+ if (addr == tipc_own_addr) {
+ node_sub->node = NULL;
return;
- if (!tipc_addr_node_valid(addr)) {
- warn("node_subscr with illegal %x\n", addr);
+ }
+
+ node_sub->node = tipc_node_find(addr);
+ if (!node_sub->node) {
+ warn("Node subscription rejected, unknown node 0x%x\n", addr);
return;
}
-
node_sub->handle_node_down = handle_down;
node_sub->usr_handle = usr_handle;
- node_sub->node = tipc_node_find(addr);
- assert(node_sub->node);
+
tipc_node_lock(node_sub->node);
list_add_tail(&node_sub->nodesub_list, &node_sub->node->nsub);
tipc_node_unlock(node_sub->node);
diff --git a/net/tipc/port.c b/net/tipc/port.c
index 99846a1..3aab67a 100644
--- a/net/tipc/port.c
+++ b/net/tipc/port.c
@@ -168,7 +168,6 @@ void tipc_port_recv_mcast(struct sk_buff
struct port_list *item = dp;
int cnt = 0;
- assert(buf);
msg = buf_msg(buf);
/* Create destination port list, if one wasn't supplied */
diff --git a/net/tipc/ref.c b/net/tipc/ref.c
index 33bbf50..d2f0cce 100644
--- a/net/tipc/ref.c
+++ b/net/tipc/ref.c
@@ -127,7 +127,14 @@ u32 tipc_ref_acquire(void *object, spinl
u32 next_plus_upper;
u32 reference = 0;
- assert(tipc_ref_table.entries && object);
+ if (!object) {
+ err("Attempt to acquire reference to non-existent object\n");
+ return 0;
+ }
+ if (!tipc_ref_table.entries) {
+ err("Reference table not found during acquisition attempt\n");
+ return 0;
+ }
write_lock_bh(&ref_table_lock);
if (tipc_ref_table.first_free) {
@@ -162,15 +169,28 @@ void tipc_ref_discard(u32 ref)
u32 index;
u32 index_mask;
- assert(tipc_ref_table.entries);
- assert(ref != 0);
+ if (!ref) {
+ err("Attempt to discard reference 0\n");
+ return;
+ }
+ if (!tipc_ref_table.entries) {
+ err("Reference table not found during discard attempt\n");
+ return;
+ }
write_lock_bh(&ref_table_lock);
index_mask = tipc_ref_table.index_mask;
index = ref & index_mask;
entry = &(tipc_ref_table.entries[index]);
- assert(entry->object != 0);
- assert(entry->data.reference == ref);
+
+ if (!entry->object) {
+ err("Attempt to discard reference to non-existent object\n");
+ goto exit;
+ }
+ if (entry->data.reference != ref) {
+ err("Attempt to discard non-existent reference\n");
+ goto exit;
+ }
/* mark entry as unused */
entry->object = NULL;
@@ -184,6 +204,7 @@ void tipc_ref_discard(u32 ref)
/* increment upper bits of entry to invalidate subsequent references */
entry->data.next_plus_upper = (ref & ~index_mask) + (index_mask + 1);
+exit:
write_unlock_bh(&ref_table_lock);
}
--
1.4.0
^ permalink raw reply related
* [PATCH 32/32] [TIPC] Fix incorrect correction to discovery timer frequency computation.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/discover.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/net/tipc/discover.c b/net/tipc/discover.c
index ee9b448..2b84412 100644
--- a/net/tipc/discover.c
+++ b/net/tipc/discover.c
@@ -2,7 +2,7 @@
* net/tipc/discover.c
*
* Copyright (c) 2003-2006, Ericsson AB
- * Copyright (c) 2005, Wind River Systems
+ * Copyright (c) 2005-2006, Wind River Systems
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -267,8 +267,8 @@ static void disc_timeout(struct link_req
/* leave timer interval "as is" if already at a "normal" rate */
} else {
req->timer_intv *= 2;
- if (req->timer_intv > TIPC_LINK_REQ_SLOW)
- req->timer_intv = TIPC_LINK_REQ_SLOW;
+ if (req->timer_intv > TIPC_LINK_REQ_FAST)
+ req->timer_intv = TIPC_LINK_REQ_FAST;
if ((req->timer_intv == TIPC_LINK_REQ_FAST) &&
(req->bearer->nodes.count))
req->timer_intv = TIPC_LINK_REQ_SLOW;
--
1.4.0
^ permalink raw reply related
* [PATCH 30/32] [TIPC] Fixed link switchover bugs
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
Incorporates several related fixes:
- switchover now occurs when switching from an active link to a standby link
- failure of a standby link no longer initiates switchover
- links now display correct # of received packtes following reactivation
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/link.c | 30 ++++++++++++++++++++++++------
net/tipc/node.c | 7 +++++--
net/tipc/node.h | 2 ++
3 files changed, 31 insertions(+), 8 deletions(-)
diff --git a/net/tipc/link.c b/net/tipc/link.c
index d7668b8..d646580 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -690,6 +690,7 @@ void tipc_link_reset(struct link *l_ptr)
struct sk_buff *buf;
u32 prev_state = l_ptr->state;
u32 checkpoint = l_ptr->next_in_no;
+ int was_active_link = tipc_link_is_active(l_ptr);
msg_set_session(l_ptr->pmsg, msg_session(l_ptr->pmsg) + 1);
@@ -711,7 +712,7 @@ #if 0
tipc_printf(TIPC_CONS, "\nReset link <%s>\n", l_ptr->name);
dbg_link_dump();
#endif
- if (tipc_node_has_active_links(l_ptr->owner) &&
+ if (was_active_link && tipc_node_has_active_links(l_ptr->owner) &&
l_ptr->owner->permit_changeover) {
l_ptr->reset_checkpoint = checkpoint;
l_ptr->exp_msg_count = START_CHANGEOVER;
@@ -754,7 +755,7 @@ #endif
static void link_activate(struct link *l_ptr)
{
- l_ptr->next_in_no = 1;
+ l_ptr->next_in_no = l_ptr->stats.recv_info = 1;
tipc_node_link_up(l_ptr->owner, l_ptr);
tipc_bearer_add_dest(l_ptr->b_ptr, l_ptr->addr);
link_send_event(tipc_cfg_link_event, l_ptr, 1);
@@ -2303,12 +2304,18 @@ void tipc_link_tunnel(struct link *l_ptr
u32 length = msg_size(msg);
tunnel = l_ptr->owner->active_links[selector & 1];
- if (!tipc_link_is_up(tunnel))
+ if (!tipc_link_is_up(tunnel)) {
+ warn("Link changeover error, "
+ "tunnel link no longer available\n");
return;
+ }
msg_set_size(tunnel_hdr, length + INT_H_SIZE);
buf = buf_acquire(length + INT_H_SIZE);
- if (!buf)
+ if (!buf) {
+ warn("Link changeover error, "
+ "unable to send tunnel msg\n");
return;
+ }
memcpy(buf->data, (unchar *)tunnel_hdr, INT_H_SIZE);
memcpy(buf->data + INT_H_SIZE, (unchar *)msg, length);
dbg("%c->%c:", l_ptr->b_ptr->net_plane, tunnel->b_ptr->net_plane);
@@ -2328,19 +2335,23 @@ void tipc_link_changeover(struct link *l
u32 msgcount = l_ptr->out_queue_size;
struct sk_buff *crs = l_ptr->first_out;
struct link *tunnel = l_ptr->owner->active_links[0];
- int split_bundles = tipc_node_has_redundant_links(l_ptr->owner);
struct tipc_msg tunnel_hdr;
+ int split_bundles;
if (!tunnel)
return;
- if (!l_ptr->owner->permit_changeover)
+ if (!l_ptr->owner->permit_changeover) {
+ warn("Link changeover error, "
+ "peer did not permit changeover\n");
return;
+ }
msg_init(&tunnel_hdr, CHANGEOVER_PROTOCOL,
ORIGINAL_MSG, TIPC_OK, INT_H_SIZE, l_ptr->addr);
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
msg_set_msgcnt(&tunnel_hdr, msgcount);
+ dbg("Link changeover requires %u tunnel messages\n", msgcount);
if (!l_ptr->first_out) {
struct sk_buff *buf;
@@ -2360,6 +2371,9 @@ void tipc_link_changeover(struct link *l
return;
}
+ split_bundles = (l_ptr->owner->active_links[0] !=
+ l_ptr->owner->active_links[1]);
+
while (crs) {
struct tipc_msg *msg = buf_msg(crs);
@@ -2497,11 +2511,13 @@ static int link_recv_changeover_msg(stru
dest_link->name);
tipc_link_reset(dest_link);
dest_link->exp_msg_count = msg_count;
+ dbg("Expecting %u tunnelled messages\n", msg_count);
if (!msg_count)
goto exit;
} else if (dest_link->exp_msg_count == START_CHANGEOVER) {
msg_dbg(tunnel_msg, "BLK/FIRST/<REC<");
dest_link->exp_msg_count = msg_count;
+ dbg("Expecting %u tunnelled messages\n", msg_count);
if (!msg_count)
goto exit;
}
@@ -2509,6 +2525,8 @@ static int link_recv_changeover_msg(stru
/* Receive original message */
if (dest_link->exp_msg_count == 0) {
+ warn("Link switchover error, "
+ "got too many tunnelled messages\n");
msg_dbg(tunnel_msg, "OVERDUE/DROP/<REC<");
dbg_print_link(dest_link, "LINK:");
goto exit;
diff --git a/net/tipc/node.c b/net/tipc/node.c
index 5f09754..ce9678e 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -125,6 +125,8 @@ void tipc_node_link_up(struct node *n_pt
{
struct link **active = &n_ptr->active_links[0];
+ n_ptr->working_links++;
+
info("Established link <%s> on network plane %c\n",
l_ptr->name, l_ptr->b_ptr->net_plane);
@@ -185,6 +187,8 @@ void tipc_node_link_down(struct node *n_
{
struct link **active;
+ n_ptr->working_links--;
+
if (!tipc_link_is_active(l_ptr)) {
info("Lost standby link <%s> on network plane %c\n",
l_ptr->name, l_ptr->b_ptr->net_plane);
@@ -214,8 +218,7 @@ int tipc_node_has_active_links(struct no
int tipc_node_has_redundant_links(struct node *n_ptr)
{
- return (tipc_node_has_active_links(n_ptr) &&
- (n_ptr->active_links[0] != n_ptr->active_links[1]));
+ return (n_ptr->working_links > 1);
}
static int tipc_node_has_active_routes(struct node *n_ptr)
diff --git a/net/tipc/node.h b/net/tipc/node.h
index 781126e..a07cc79 100644
--- a/net/tipc/node.h
+++ b/net/tipc/node.h
@@ -51,6 +51,7 @@ #include "bearer.h"
* @nsub: list of "node down" subscriptions monitoring node
* @active_links: pointers to active links to node
* @links: pointers to all links to node
+ * @working_links: number of working links to node (both active and standby)
* @link_cnt: number of links to node
* @permit_changeover: non-zero if node has redundant links to this system
* @routers: bitmap (used for multicluster communication)
@@ -76,6 +77,7 @@ struct node {
struct link *active_links[2];
struct link *links[MAX_BEARERS];
int link_cnt;
+ int working_links;
int permit_changeover;
u32 routers[512/32];
int last_router;
--
1.4.0
^ permalink raw reply related
* [PATCH 31/32] [TIPC] Get rid of dynamically allocated arrays in broadcast code.
From: Per Liden @ 2006-06-22 13:59 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Allan Stephens
In-Reply-To: <Pine.LNX.4.64.0606221233570.15853@ulinpc219.uab.ericsson.se>
From: Allan Stephens <allan.stephens@windriver.com>
This change improves an earlier change which replaced the large local
variable arrays used during broadcasting with dynamically allocated arrays.
The temporary arrays are now incoprorated into the multicast link data
structure.
Signed-off-by: Allan Stephens <allan.stephens@windriver.com>
Signed-off-by: Per Liden <per.liden@ericsson.com>
---
net/tipc/bcast.c | 41 +++++++++++++++++------------------------
1 files changed, 17 insertions(+), 24 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 1633ef2..5412804 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -81,7 +81,14 @@ struct bcbearer_pair {
* @bearer: (non-standard) broadcast bearer structure
* @media: (non-standard) broadcast media structure
* @bpairs: array of bearer pairs
- * @bpairs_temp: array of bearer pairs used during creation of "bpairs"
+ * @bpairs_temp: temporary array of bearer pairs used by tipc_bcbearer_sort()
+ * @remains: temporary node map used by tipc_bcbearer_send()
+ * @remains_new: temporary node map used tipc_bcbearer_send()
+ *
+ * Note: The fields labelled "temporary" are incorporated into the bearer
+ * to avoid consuming potentially limited stack space through the use of
+ * large local variables within multicast routines. Concurrent access is
+ * prevented through use of the spinlock "bc_lock".
*/
struct bcbearer {
@@ -89,6 +96,8 @@ struct bcbearer {
struct media media;
struct bcbearer_pair bpairs[MAX_BEARERS];
struct bcbearer_pair bpairs_temp[TIPC_MAX_LINK_PRI + 1];
+ struct node_map remains;
+ struct node_map remains_new;
};
/**
@@ -551,12 +560,8 @@ static int tipc_bcbearer_send(struct sk_
{
static int send_count = 0;
- struct node_map *remains;
- struct node_map *remains_new;
- struct node_map *remains_tmp;
int bp_index;
int swap_time;
- int err;
/* Prepare buffer for broadcasting (if first time trying to send it) */
@@ -577,9 +582,7 @@ static int tipc_bcbearer_send(struct sk_
/* Send buffer over bearers until all targets reached */
- remains = kmalloc(sizeof(struct node_map), GFP_ATOMIC);
- remains_new = kmalloc(sizeof(struct node_map), GFP_ATOMIC);
- *remains = tipc_cltr_bcast_nodes;
+ bcbearer->remains = tipc_cltr_bcast_nodes;
for (bp_index = 0; bp_index < MAX_BEARERS; bp_index++) {
struct bearer *p = bcbearer->bpairs[bp_index].primary;
@@ -588,8 +591,8 @@ static int tipc_bcbearer_send(struct sk_
if (!p)
break; /* no more bearers to try */
- tipc_nmap_diff(remains, &p->nodes, remains_new);
- if (remains_new->count == remains->count)
+ tipc_nmap_diff(&bcbearer->remains, &p->nodes, &bcbearer->remains_new);
+ if (bcbearer->remains_new.count == bcbearer->remains.count)
continue; /* bearer pair doesn't add anything */
if (!p->publ.blocked &&
@@ -607,27 +610,17 @@ swap:
bcbearer->bpairs[bp_index].primary = s;
bcbearer->bpairs[bp_index].secondary = p;
update:
- if (remains_new->count == 0) {
- err = TIPC_OK;
- goto out;
- }
+ if (bcbearer->remains_new.count == 0)
+ return TIPC_OK;
- /* swap map */
- remains_tmp = remains;
- remains = remains_new;
- remains_new = remains_tmp;
+ bcbearer->remains = bcbearer->remains_new;
}
/* Unable to reach all targets */
bcbearer->bearer.publ.blocked = 1;
bcl->stats.bearer_congs++;
- err = ~TIPC_OK;
-
- out:
- kfree(remains_new);
- kfree(remains);
- return err;
+ return ~TIPC_OK;
}
/**
--
1.4.0
^ permalink raw reply related
* Re: [0/5] GSO: Generic Segmentation Offload
From: YOSHIFUJI Hideaki / 吉藤英明 @ 2006-06-22 14:28 UTC (permalink / raw)
To: herbert; +Cc: davem, netdev, yoshfuji
In-Reply-To: <20060622081211.GA22505@gondor.apana.org.au>
Hello.
Yes, I genrally like this idea.
In article <20060622081211.GA22505@gondor.apana.org.au> (at Thu, 22 Jun 2006 18:12:11 +1000), Herbert Xu <herbert@gondor.apana.org.au> says:
> GSO like TSO is only effective if the MTU is significantly less than the
> maximum value of 64K. So only the case where the MTU was set to 1500 is
> of interest. There we can see that the throughput improved by 17.5%
> (3061.05Mb/s => 3598.17Mb/s). The actual saving in transmission cost is
> in fact a lot more than that as the majority of the time here is spent on
> the RX side which still has to deal with 1500-byte packets.
Can you measure some with other sizes,
e.g. 4kByte, 8kByte, 9000Byte?
--yoshfuji
^ permalink raw reply
* IPSec + large packets being corrupted
From: Chris Audley @ 2006-06-22 14:32 UTC (permalink / raw)
To: netdev
I've been using the 2.6 kernel ipsec system for some time and have always
had to work around issues with large packets not traversing the VPN by
setting the LAN interface MTU size to something like 1400.
Because I always thought this was a hack and not a proper fix, I've spent
a few days trying to work out exactly why large packets aren't traversing
the VPN and have found something which may well be the cause. I really
don't know the kernel networking code that well so I was hoping that
someone can either verify that what I've found is really an issue, or
whether I'm doing something wrong.
This has been seen in the field with P4/e100+e1000 systems running 2.6.12
and in testing on Geode/dp8381x systems running 2.6.17, all using IPv4.
VPN is Racoon based, using x509 certs and ESP/AH (3DES/SHA1).
This is my understanding of how large packets get corrupted:
Large packet (eg. 1600 byte ping) received by VPN server A.
Packet encrypted and fragmented then sent from Server A to Server B.
Packet received by network subsytem on B and frag_list created
ah_input() strips the AH header -- frag sizes are not changed!
esp_input() decrypts data
ip_fragment() uses existing frag_list sizes from before the AH
header being stripped, and sends too much data (16 bytes extra). This
breaks the checksum and packets get dropped by destination host.
By setting the MTU on the local interface, this breaks one of the
checks for using the pre-existing frag list in ip_fragment() (MTU is now
smaller than the largest frag size), so the packet fragments are
re-generated from scratch and the large packet gets through.
If I disable the valid frag_list check in ip_fragment(), again large
packets traverse the VPN with no problems at all since the fragments are
re-generated from scratch.
If my analysis of the above is correct, then my feeling is that either
ah_input() should re-calculate the fragment sizes, or some flag should be
set to tell ip_fragment() to use the slow method and recreate the
fragments.
Does this sound like a real problem, or have I missed something obvious?
Regards,
--
Chris Audley mailto:chris@navaho.co.uk
^ permalink raw reply
* Re: [3/5] [NET]: Add software TSOv4
From: YOSHIFUJI Hideaki / 吉藤英明 @ 2006-06-22 15:04 UTC (permalink / raw)
To: herbert; +Cc: davem, netdev, yoshfuji
In-Reply-To: <20060622081400.GC22671@gondor.apana.org.au>
In article <20060622081400.GC22671@gondor.apana.org.au> (at Thu, 22 Jun 2006 18:14:00 +1000), Herbert Xu <herbert@gondor.apana.org.au> says:
> [NET]: Add software TSOv4
>
> This patch adds the GSO implementation for IPv4 TCP.
>
> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
I'd appreciate if you code up IPv6 TCP as well. :-)
Regards,
--yoshfuji
^ permalink raw reply
* Re: [RFC 2/7] NetLabel: core network changes
From: Steve Grubb @ 2006-06-22 15:05 UTC (permalink / raw)
To: David Miller
Cc: jmorris, paul.moore, sds, redhat-lspp, linux-security-module,
selinux, netdev
In-Reply-To: <20060622.020055.115910616.davem@davemloft.net>
On Thursday 22 June 2006 05:00, David Miller wrote:
> > #define NETLINK_GENERIC 16
> > +#define NETLINK_NETLABEL 17 /* Network packet labeling */
> >
> > #define MAX_LINKS 32
>
> Please use generic netlink.
Since this is a security interface, shouldn't it be its own protocol so that
SE Linux can control commands being sent? Paul's patches do include a netlink
table in security/selinux/nlmsgtab.c. But I do not see any hooks to control
generic netlink messages. (There seems to be several protocols that SE Linux
is not controlling.) I could see that someone in secadm role should be able
to issue these commands, but someone at sysadm or auditadm would not.
If moving this over to generic is a must, then I think SE Linux will have to
clip into generic to control its packet flow.
-Steve
^ permalink raw reply
* Re: [PATCH 0/2][RFC] Network Event Notifier Mechanism
From: Steve Wise @ 2006-06-22 15:27 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <1150984380.27156.4.camel@stevo-desktop>
On Thu, 2006-06-22 at 08:53 -0500, Steve Wise wrote:
> On Thu, 2006-06-22 at 01:57 -0700, David Miller wrote:
> > From: Steve Wise <swise@opengridcomputing.com>
> > Date: Wed, 21 Jun 2006 13:45:19 -0500
> >
> > > This patch implements a mechanism that allows interested clients to
> > > register for notification of certain network events.
> >
> > We have a generic network event notification facility called
> > netlink, please use it and extend it for your needs if necessary.
>
> I'll investigate this.
>
> Thanks,
The in-kernel Infiniband subsystem needs to know when certain events
happen. For example, if the mac address of a neighbour changes. Any
rdma devices that are using said neighbour need to be notified of the
change. You are asking that I extend the netlink facility (if
necessary) to provide this functionality.
Are you suggesting, then, that the Infiniband subsystem should create an
in-kernel NETLINK socket and obtain these events (and the pertinent
information) via the socket?
I'm still learning about netlink, but my understanding to date is that
its a way to pass events/commands between the kernel and user
applications. It perhaps seems overkill to use this mechanism for
kernel->kernel event notifications. That's why I started with notifier
blocks and added a netevent_notifier mechanism.
Any help is greatly appreciated. Sorry if I'm being dense...
Steve.
^ permalink raw reply
* Re: [PATCH, RFT] bcm43xx: AccessPoint mode
From: Jiri Benc @ 2006-06-22 15:30 UTC (permalink / raw)
To: Michael Buesch
Cc: bcm43xx-dev, netdev, Alexander Tsvyashchenko, Francois Barre
In-Reply-To: <200606191107.34602.mb@bu3sch.de>
On Mon, 19 Jun 2006 11:07:34 +0200, Michael Buesch wrote:
> Well, it does not work 100%, but at least it's very promising.
> We are able to create a bssid and correctly send beacon frames out.
Great work! I was even able to ping. (Tried only open system
authentication for now, it seems it works quite well.)
> Please give it a testrun.
> Final note about hostapd:
> "hostapd snapshot 0.5-2006-06-10" seems to "work" in the sense
> that it is able to bring up the device.
> "hostapd snapshot 0.5-2006-06-11" seems to fail.
0.5-2006-06-19 works with the patch.
> Important notes from Alexander Tsvyashchenko's initial mail follow:
> [...]
> Although my previous patch to hostapd to make it interoperable with
> bcm43xx & dscape has been merged already in their CVS version, due to
> the subsequent changes in dscape stack current hostapd is again
> incompartible :-( So, to test this patch, the patch to hostapd should be
> applied.
Or, if you don't want to patch hostapd (untested, but should work):
iwpriv wlan0 param 1046 1
ip link set wmgmt0 name wmaster0ap
hostapd /path/to/hostapd.conf
iwpriv wlan0 param 1046 0
> I used hostapd snapshot 0.5-2006-06-10, patch for it is attached.
> The patch is very hacky and requires tricky way to bring everything up,
> but as dscape stack is changed quite constantly, I just do not want to
> waste time fixing it in "proper" way only to find a week later that
> dscape handling of master interface was changed completely once more and
> everything is broken again ;-)
Hopefully we will convert the whole hostapd<->stack communication to
netlink in some near future ;-)
> 2) Insert modules (80211, rate_control and bcm43xx-d80211)
modprobe bcm43xx-d80211 is enough, other modules will load
automatically.
> 4) "ifconfig wlan0 up" (this should be done by hostapd actually, but
> its operation with current dscape stack seems to be broken)
hostapd tries to open (put to 'up' state) wmgmt0 earlier than wlan0,
which is not possible. It should open wlan0 first; even more, opening of
wmgmt0 is not necessary as it will be opened automatically when wlan0 is
opened.
> 6) "iwconfig wlan0 essid <your-SSID-name>" (this also should not be
> required, but current combination of hostapd + dscape doesn't seem to
> generate config_interface callback when setting beacon, so this is
> required just to force call of config_interface).
The stack currently has very limited support for cards with "beacon
templates". ieee80211_beacon_get function is not designed in a way it is
used in bcm43xx. Although this seems to be easy to fix now, we will run
into other problems later (with TIM elements mainly).
I need to look how PS mode works in bcm chipsets to find a correct
solution for this. Do you have any ideas?
Thanks,
Jiri
--
Jiri Benc
SUSE Labs
^ permalink raw reply
* Binding a packet socket to interface down
From: Robert Iakobashvili @ 2006-06-22 15:32 UTC (permalink / raw)
To: netdev
Hi,
When packet socket (PF_SOCKET) is attempted in syscall bind () to the
network interface, which is down (no IF_UPP flag), packet_do_bind ()
sets error to the socket, but bind () does not fail.
When datagram, stream or raw socket fail to bind to some local
ip-port/ip, bind () fails.
Is this behavior of bind () for packet socket done deliberately, or
better to correct it so that bind will fail and return errno, e.g. ENODEV?
Thanks.
--
Sincerely,
------------------------------------------------------------------
Robert Iakobashvili, coroberti at gmail dot com
Navigare necesse est, vivere non est necesse.
------------------------------------------------------------------
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox