Enable sending and removal of version 0 sending
Affected functions,
ip_vs_sync_buff_create()
ip_vs_sync_conn()
ip_vs_core.c removal of IPv4 check.
*v4
moved sanity check and pe_name_len after sloop.
use cp->pe instead of cp->dest->svc->pe
real length in each sync_conn, not paded length
however total size of a sync_msg includes pading.
*v3
Sending ip_vs_sync_conn_options in network order.
Sending Templates for ONE_PACKET conn.
Renaming of ip_vs_sync_mesg to ip_vs_sync_mesg_v0
Signed-off-by: Hans Schillstrom <hans.schillstrom@xxxxxxxxxxxx>
---
include/net/ip_vs.h | 2 +-
net/netfilter/ipvs/ip_vs_core.c | 13 ++-
net/netfilter/ipvs/ip_vs_sync.c | 182 +++++++++++++++++++++++++++++---------
3 files changed, 149 insertions(+), 48 deletions(-)
diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h
index 4069484..a715f3d 100644
--- a/include/net/ip_vs.h
+++ b/include/net/ip_vs.h
@@ -919,7 +919,7 @@ extern char ip_vs_master_mcast_ifn[IP_VS_IFNAME_MAXLEN];
extern char ip_vs_backup_mcast_ifn[IP_VS_IFNAME_MAXLEN];
extern int start_sync_thread(int state, char *mcast_ifn, __u8 syncid);
extern int stop_sync_thread(int state);
-extern void ip_vs_sync_conn(const struct ip_vs_conn *cp);
+extern void ip_vs_sync_conn(struct ip_vs_conn *cp);
/*
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 9acdd79..6c5775d 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -1536,9 +1536,15 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
*
* Sync connection if it is about to close to
* encorage the standby servers to update the connections timeout
+ *
+ * For ONE_PKT let ip_vs_sync_conn() do the filter work.
*/
- pkts = atomic_add_return(1, &cp->in_pkts);
- if (af == AF_INET && (ip_vs_sync_state & IP_VS_STATE_MASTER) &&
+ if (cp->flags & IP_VS_CONN_F_ONE_PACKET)
+ pkts = sysctl_ip_vs_sync_threshold[0];
+ else
+ pkts = atomic_add_return(1, &cp->in_pkts);
+
+ if ((ip_vs_sync_state & IP_VS_STATE_MASTER) &&
cp->protocol == IPPROTO_SCTP) {
if ((cp->state == IP_VS_SCTP_S_ESTABLISHED &&
(pkts % sysctl_ip_vs_sync_threshold[1]
@@ -1553,8 +1559,7 @@ ip_vs_in(unsigned int hooknum, struct sk_buff *skb, int af)
}
/* Keep this block last: TCP and others with pp->num_states <= 1 */
- else if (af == AF_INET &&
- (ip_vs_sync_state & IP_VS_STATE_MASTER) &&
+ else if ((ip_vs_sync_state & IP_VS_STATE_MASTER) &&
(((cp->protocol != IPPROTO_TCP ||
cp->state == IP_VS_TCP_S_ESTABLISHED) &&
(pkts % sysctl_ip_vs_sync_threshold[1]
diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c
index f8bd8fd..083501b 100644
--- a/net/netfilter/ipvs/ip_vs_sync.c
+++ b/net/netfilter/ipvs/ip_vs_sync.c
@@ -226,7 +226,7 @@ struct ip_vs_sync_thread_data {
#define MAX_CONNS_PER_SYNCBUFF 255 /* nr_conns in ip_vs_sync_mesg is 8 bit */
/* Version 0 header */
-struct ip_vs_sync_mesg {
+struct ip_vs_sync_mesg_v0 {
__u8 nr_conns;
__u8 syncid;
__u16 size;
@@ -235,7 +235,7 @@ struct ip_vs_sync_mesg {
};
/* Version 1 header */
-struct ip_vs_sync_mesg_v2 {
+struct ip_vs_sync_mesg {
__u8 reserved; /* must be zero */
__u8 syncid;
__u16 size;
@@ -299,6 +299,17 @@ static void ntoh_seq(struct ip_vs_seq *no, struct ip_vs_seq *ho)
ho->previous_delta = get_unaligned_be32(&no->previous_delta);
}
+/*
+ * Copy of struct ip_vs_seq
+ * From Aligned host order to unaligned network order
+ */
+static void hton_seq(struct ip_vs_seq *ho, struct ip_vs_seq *no)
+{
+ put_unaligned_be32(ho->init_seq, &no->init_seq);
+ put_unaligned_be32(ho->delta, &no->delta);
+ put_unaligned_be32(ho->previous_delta, &no->previous_delta);
+}
+
static inline struct ip_vs_sync_buff *sb_dequeue(void)
{
struct ip_vs_sync_buff *sb;
@@ -317,6 +328,9 @@ static inline struct ip_vs_sync_buff *sb_dequeue(void)
return sb;
}
+/*
+ * Create a new sync buffer for Version 1 proto.
+ */
static inline struct ip_vs_sync_buff * ip_vs_sync_buff_create(void)
{
struct ip_vs_sync_buff *sb;
@@ -328,11 +342,15 @@ static inline struct ip_vs_sync_buff * ip_vs_sync_buff_create(void)
kfree(sb);
return NULL;
}
- sb->mesg->nr_conns = 0;
+ sb->mesg->reserved = 0; /* old nr_conns i.e. must be zeo now */
+ sb->mesg->version = SYNC_PROTO_VER;
sb->mesg->syncid = ip_vs_master_syncid;
- sb->mesg->size = 4;
- sb->head = (unsigned char *)sb->mesg + 4;
+ sb->mesg->size = sizeof(struct ip_vs_sync_mesg);
+ sb->mesg->nr_conns = 0;
+ sb->mesg->spare = 0;
+ sb->head = (unsigned char *)sb->mesg + sizeof(struct ip_vs_sync_mesg);
sb->end = (unsigned char *)sb->mesg + sync_send_mesg_maxlen;
+
sb->firstuse = jiffies;
return sb;
}
@@ -373,18 +391,56 @@ get_curr_sync_buff(unsigned long time)
return sb;
}
-
/*
* Add an ip_vs_conn information into the current sync_buff.
* Called by ip_vs_in.
+ * Sending Version 1 messages
*/
-void ip_vs_sync_conn(const struct ip_vs_conn *cp)
+void ip_vs_sync_conn(struct ip_vs_conn *cp)
{
struct ip_vs_sync_mesg *m;
- struct ip_vs_sync_conn_v0 *s;
- int len;
+ union ip_vs_sync_conn *s;
+ __u8 *p;
+ unsigned int len, pe_name_len, pad;
+
+ /* Do not sync ONE PACKET */
+ if (cp->flags & IP_VS_CONN_F_ONE_PACKET)
+ goto control;
+sloop:
+ /* Sanity checks */
+ if (cp->pe_data_len && (!cp->pe_data || !cp->dest)) {
+ IP_VS_ERR_RL("SYNC, connection pe_data invalid\n");
+ return;
+ }
+ pe_name_len = 0;
+ if (cp->pe_data_len && cp->dest->svc && cp->pe && cp->pe->name)
+ pe_name_len = strnlen(cp->pe->name,
+ IP_VS_PENAME_MAXLEN);
spin_lock(&curr_sb_lock);
+#ifdef CONFIG_IP_VS_IPV6
+ /* Assumption,if not configured for IPv6 no packets should enter here */
+ if (cp->af == AF_INET6)
+ len = sizeof(struct ip_vs_sync_v6);
+ else
+#endif
+ len = sizeof(struct ip_vs_sync_v4);
+
+ if (cp->flags & IP_VS_CONN_F_SEQ_MASK)
+ len += sizeof(struct ip_vs_sync_conn_options) + 2;
+
+ if (cp->pe_data_len)
+ len += cp->pe_data_len + 2; /* + Param hdr field */
+ if (pe_name_len)
+ len += pe_name_len + 2;
+ /* No of padding bytes */
+ pad = (4 - len) & 3;
+ /* check if there is a space for this one */
+ if (curr_sb && (curr_sb->head+len+pad > curr_sb->end)) {
+ sb_queue_tail(curr_sb);
+ curr_sb = NULL;
+ }
+
if (!curr_sb) {
if (!(curr_sb=ip_vs_sync_buff_create())) {
spin_unlock(&curr_sb_lock);
@@ -393,41 +449,81 @@ void ip_vs_sync_conn(const struct ip_vs_conn *cp)
}
}
- len = (cp->flags & IP_VS_CONN_F_SEQ_MASK) ? FULL_CONN_SIZE :
- SIMPLE_CONN_SIZE;
m = curr_sb->mesg;
- s = (struct ip_vs_sync_conn_v0 *)curr_sb->head;
-
- /* copy members */
- s->protocol = cp->protocol;
- s->cport = cp->cport;
- s->vport = cp->vport;
- s->dport = cp->dport;
- s->caddr = cp->caddr.ip;
- s->vaddr = cp->vaddr.ip;
- s->daddr = cp->daddr.ip;
- s->flags = htons(cp->flags & ~IP_VS_CONN_F_HASHED);
- s->state = htons(cp->state);
- if (cp->flags & IP_VS_CONN_F_SEQ_MASK) {
- struct ip_vs_sync_conn_options *opt =
- (struct ip_vs_sync_conn_options *)&s[1];
- memcpy(opt, &cp->in_seq, sizeof(*opt));
- }
-
+ s = (union ip_vs_sync_conn *)curr_sb->head;
+ p = (char *)s;
+
+ /* Set message type & copy members */
+ s->v4.type = (cp->af == AF_INET6 ? STYPE_F_INET6 : 0);
+ s->v4.ver_size = htons(len & SVER_MASK); /* Version 0 */
+ s->v4.flags = htonl(cp->flags & ~IP_VS_CONN_F_HASHED);
+ s->v4.state = htons(cp->state);
+ s->v4.protocol = cp->protocol;
+ s->v4.cport = cp->cport;
+ s->v4.vport = cp->vport;
+ s->v4.dport = cp->dport;
+ s->v4.fwmark = htonl(cp->fwmark);
+ s->v4.timeout = htonl(cp->timeout / HZ);
m->nr_conns++;
- m->size += len;
- curr_sb->head += len;
- /* check if there is a space for next one */
- if (curr_sb->head+FULL_CONN_SIZE > curr_sb->end) {
- sb_queue_tail(curr_sb);
- curr_sb = NULL;
+#ifdef CONFIG_IP_VS_IPV6
+ if (cp->af == AF_INET6) {
+ p += sizeof(struct ip_vs_sync_v6);
+ ipv6_addr_copy(&s->v6.caddr, &cp->caddr.in6);
+ ipv6_addr_copy(&s->v6.vaddr, &cp->vaddr.in6);
+ ipv6_addr_copy(&s->v6.daddr, &cp->daddr.in6);
+ } else
+#endif
+ {
+ p += sizeof(struct ip_vs_sync_v4); /* options ptr */
+ s->v4.caddr = cp->caddr.ip;
+ s->v4.vaddr = cp->vaddr.ip;
+ s->v4.daddr = cp->daddr.ip;
}
+ if (cp->flags & IP_VS_CONN_F_SEQ_MASK) {
+ *(p++) = IPVS_OPT_SEQ_DATA;
+ *(p++) = sizeof(struct ip_vs_sync_conn_options);
+ hton_seq((struct ip_vs_seq *)p, &cp->in_seq);
+ p += sizeof(struct ip_vs_seq);
+ hton_seq((struct ip_vs_seq *)p, &cp->out_seq);
+ p += sizeof(struct ip_vs_seq);
+ }
+ /* Handle pe data */
+ if (cp->pe_data_len && cp->pe_data) {
+ *(p++) = IPVS_OPT_PE_DATA;
+ *(p++) = cp->pe_data_len;
+ memcpy(p, cp->pe_data,cp->pe_data_len);
+ p += cp->pe_data_len;
+ if (pe_name_len) {
+ /* Add PE_NAME */
+ *(p++) = IPVS_OPT_PE_NAME;
+ *(p++) = pe_name_len;
+ memcpy(p, cp->pe->name, pe_name_len);
+ p += pe_name_len;
+ }
+ }
+ m->size += len+pad;
+ curr_sb->head += len+pad;
+ while (pad--)
+ *(p++) = 0; /* Dont leave random bytes at end */
spin_unlock(&curr_sb_lock);
+control:
/* synchronize its controller if it has */
- if (cp->control)
- ip_vs_sync_conn(cp->control);
+ cp = cp->control;
+ if (!cp)
+ return;
+ /*
+ * Reduce sync rate for templates
+ * i.e only increment in_pkts for Templates.
+ */
+ if (cp->flags & IP_VS_CONN_F_TEMPLATE) {
+ int pkts = atomic_add_return(1, &cp->in_pkts);
+
+ if (pkts % sysctl_ip_vs_sync_threshold[1] != 1)
+ return;
+ }
+ goto sloop;
}
/*
@@ -596,7 +692,7 @@ static void ip_vs_proc_conn(struct ip_vs_conn_param *param, unsigned flags,
*/
static void ip_vs_process_message_v0(const char *buffer, const size_t buflen)
{
- struct ip_vs_sync_mesg *m = (struct ip_vs_sync_mesg *)buffer;
+ struct ip_vs_sync_mesg_v0 *m = (struct ip_vs_sync_mesg_v0 *)buffer;
struct ip_vs_sync_conn_v0 *s;
struct ip_vs_sync_conn_options *opt;
struct ip_vs_protocol *pp;
@@ -604,7 +700,7 @@ static void ip_vs_process_message_v0(const char *buffer, const size_t buflen)
char *p;
int i;
- p = (char *)buffer + sizeof(struct ip_vs_sync_mesg);
+ p = (char *)buffer + sizeof(struct ip_vs_sync_mesg_v0);
for (i=0; i<m->nr_conns; i++) {
unsigned flags, state;
@@ -848,11 +944,11 @@ out:
*/
static void ip_vs_process_message(__u8 *buffer, const size_t buflen)
{
- struct ip_vs_sync_mesg_v2 *m2 = (struct ip_vs_sync_mesg_v2 *)buffer;
+ struct ip_vs_sync_mesg *m2 = (struct ip_vs_sync_mesg *)buffer;
__u8 *p, *msg_end;
- unsigned int i, nr_conns;
+ int i, nr_conns;
- if (buflen < sizeof(struct ip_vs_sync_mesg)) {
+ if (buflen < sizeof(struct ip_vs_sync_mesg_v0)) {
IP_VS_DBG(2, "BACKUP, message header too short\n");
return;
}
@@ -872,7 +968,7 @@ static void ip_vs_process_message(__u8 *buffer, const size_t buflen)
if ((m2->version == SYNC_PROTO_VER) && (m2->reserved == 0)
&& (m2->spare == 0)) {
- msg_end = buffer + sizeof(struct ip_vs_sync_mesg_v2);
+ msg_end = buffer + sizeof(struct ip_vs_sync_mesg);
nr_conns = m2->nr_conns;
for (i=0; i<nr_conns; i++) {
--
1.7.2.3
--
To unsubscribe from this list: send the line "unsubscribe lvs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
