slirp: Mark pieces missing IPv6 support

[mirror_qemu.git] / net / filter-rewriter.c

diff --git a/net/filter-rewriter.c b/net/filter-rewriter.c
index dd323faf61cecda8f0631cad58beb4d149aa7707..b464abe5e883b2c1dbdbfb8603e3ea4ed2ed7350 100644 (file)
--- a/net/filter-rewriter.c
+++ b/net/filter-rewriter.c
@@ -20,11 +20,16 @@
 #include "qemu/main-loop.h"
 #include "qemu/iov.h"
 #include "net/checksum.h"
+#include "net/colo.h"
+#include "migration/colo.h"
+#include "util.h"
 
 #define FILTER_COLO_REWRITER(obj) \
     OBJECT_CHECK(RewriterState, (obj), TYPE_FILTER_REWRITER)
 
 #define TYPE_FILTER_REWRITER "filter-rewriter"
+#define FAILOVER_MODE_ON  true
+#define FAILOVER_MODE_OFF false
 
 typedef struct RewriterState {
     NetFilterState parent_obj;
@@ -32,8 +37,14 @@ typedef struct RewriterState {
     /* hashtable to save connection */
     GHashTable *connection_track_table;
     bool vnet_hdr;
+    bool failover_mode;
 } RewriterState;
 
+static void filter_rewriter_failover_mode(RewriterState *s)
+{
+    s->failover_mode = FAILOVER_MODE_ON;
+}
+
 static void filter_rewriter_flush(NetFilterState *nf)
 {
     RewriterState *s = FILTER_COLO_REWRITER(nf);
@@ -63,9 +74,9 @@ static int handle_primary_tcp_pkt(RewriterState *rf,
                                   Connection *conn,
                                   Packet *pkt, ConnectionKey *key)
 {
-    struct tcphdr *tcp_pkt;
+    struct tcp_hdr *tcp_pkt;
 
-    tcp_pkt = (struct tcphdr *)pkt->transport_header;
+    tcp_pkt = (struct tcp_hdr *)pkt->transport_header;
     if (trace_event_get_state_backends(TRACE_COLO_FILTER_REWRITER_DEBUG)) {
         trace_colo_filter_rewriter_pkt_info(__func__,
                     inet_ntoa(pkt->ip->ip_src), inet_ntoa(pkt->ip->ip_dst),
@@ -145,12 +156,13 @@ static int handle_primary_tcp_pkt(RewriterState *rf,
          * Active close step 2.
          */
         if (conn->tcp_state == TCPS_FIN_WAIT_1) {
-            conn->tcp_state = TCPS_TIME_WAIT;
             /*
              * For simplify implementation, we needn't wait 2MSL time
              * in filter rewriter. Because guest kernel will track the
              * TCP status and wait 2MSL time, if client resend the FIN
              * packet, guest will apply the last ACK too.
+             * So, we skip the TCPS_TIME_WAIT state here and go straight
+             * to TCPS_CLOSED state.
              */
             conn->tcp_state = TCPS_CLOSED;
             g_hash_table_remove(rf->connection_track_table, key);
@@ -165,9 +177,9 @@ static int handle_secondary_tcp_pkt(RewriterState *rf,
                                     Connection *conn,
                                     Packet *pkt, ConnectionKey *key)
 {
-    struct tcphdr *tcp_pkt;
+    struct tcp_hdr *tcp_pkt;
 
-    tcp_pkt = (struct tcphdr *)pkt->transport_header;
+    tcp_pkt = (struct tcp_hdr *)pkt->transport_header;
 
     if (trace_event_get_state_backends(TRACE_COLO_FILTER_REWRITER_DEBUG)) {
         trace_colo_filter_rewriter_pkt_info(__func__,
@@ -273,6 +285,13 @@ static ssize_t colo_rewriter_receive_iov(NetFilterState *nf,
              */
             reverse_connection_key(&key);
         }
+
+        /* After failover we needn't change new TCP packet */
+        if (s->failover_mode &&
+            !connection_has_tracked(s->connection_track_table, &key)) {
+            goto out;
+        }
+
         conn = connection_get(s->connection_track_table,
                               &key,
                               NULL);
@@ -306,11 +325,49 @@ static ssize_t colo_rewriter_receive_iov(NetFilterState *nf,
         }
     }
 
+out:
     packet_destroy(pkt, NULL);
     pkt = NULL;
     return 0;
 }
 
+static void reset_seq_offset(gpointer key, gpointer value, gpointer user_data)
+{
+    Connection *conn = (Connection *)value;
+
+    conn->offset = 0;
+}
+
+static gboolean offset_is_nonzero(gpointer key,
+                                  gpointer value,
+                                  gpointer user_data)
+{
+    Connection *conn = (Connection *)value;
+
+    return conn->offset ? true : false;
+}
+
+static void colo_rewriter_handle_event(NetFilterState *nf, int event,
+                                       Error **errp)
+{
+    RewriterState *rs = FILTER_COLO_REWRITER(nf);
+
+    switch (event) {
+    case COLO_EVENT_CHECKPOINT:
+        g_hash_table_foreach(rs->connection_track_table,
+                            reset_seq_offset, NULL);
+        break;
+    case COLO_EVENT_FAILOVER:
+        if (!g_hash_table_find(rs->connection_track_table,
+                              offset_is_nonzero, NULL)) {
+            filter_rewriter_failover_mode(rs);
+        }
+        break;
+    default:
+        break;
+    }
+}
+
 static void colo_rewriter_cleanup(NetFilterState *nf)
 {
     RewriterState *s = FILTER_COLO_REWRITER(nf);
@@ -354,6 +411,7 @@ static void filter_rewriter_init(Object *obj)
     RewriterState *s = FILTER_COLO_REWRITER(obj);
 
     s->vnet_hdr = false;
+    s->failover_mode = FAILOVER_MODE_OFF;
     object_property_add_bool(obj, "vnet_hdr_support",
                              filter_rewriter_get_vnet_hdr,
                              filter_rewriter_set_vnet_hdr, NULL);
@@ -366,6 +424,7 @@ static void colo_rewriter_class_init(ObjectClass *oc, void *data)
     nfc->setup = colo_rewriter_setup;
     nfc->cleanup = colo_rewriter_cleanup;
     nfc->receive_iov = colo_rewriter_receive_iov;
+    nfc->handle_event = colo_rewriter_handle_event;
 }
 
 static const TypeInfo colo_rewriter_info = {