fib: Table Replace

Type: feature

from the API doc, a table replace is:

"
    The use-case is that, for some unspecified reason, the control plane
    has a very different set of entries it wants in the table than VPP
    currently has. The CP would thus like to 'replace' VPP's current table
    only by specifying what the new set of entries shall be, i.e. it is not
    going to delete anything that already eixts.
    the CP delcartes the start of this procedure with this begin_replace
    API Call, and when it has populated all the entries it wants, it calls
    the below end_replace API. From this point on it is of coursce free
    to add and delete entries as usual.
    The underlying mechanism by which VPP implements this replace is
    purposefully left unspecified.
"

In the FIB, the algorithm is implemented using mark and sweep.

Algorithm goes:
1) replace_begin: this marks all the entries in that table as 'stale'
2) download all the entries that should be in this table
 - this clears the stale flag on those entries
3) signal the table converged: ip_table_replace_end
 - this removes all entries that are still stale

this procedure can be used when an agent first connects to VPP,
as an alternative to dump and diff state reconciliation.

Change-Id: I168edec10cf7670866076b129ebfe6149ea8222e
Signed-off-by: Neale Ranns <nranns@cisco.com>

1 files changed, 88 insertions, 2 deletions

diff --git a/src/vnet/mfib/mfib_table.h b/src/vnet/mfib/mfib_table.h
index 47461375a10..d6b6dc3845d 100644
--- a/src/vnet/mfib/mfib_table.h
+++ b/src/vnet/mfib/mfib_table.h
@@ -29,6 +29,42 @@
 #define MFIB_TABLE_TOTAL_LOCKS MFIB_N_SOURCES
 
 /**
+ * Flags for the source data
+ */
+typedef enum mfib_table_attribute_t_ {
+    /**
+     * Marker. Add new values after this one.
+     */
+    MFIB_TABLE_ATTRIBUTE_FIRST,
+    /**
+     * the table is currently resync-ing
+     */
+    MFIB_TABLE_ATTRIBUTE_RESYNC = MFIB_TABLE_ATTRIBUTE_FIRST,
+    /**
+     * Marker. add new entries before this one.
+     */
+    MFIB_TABLE_ATTRIBUTE_LAST = MFIB_TABLE_ATTRIBUTE_RESYNC,
+} mfib_table_attribute_t;
+
+#define MFIB_TABLE_ATTRIBUTE_MAX (MFIB_TABLE_ATTRIBUTE_LAST+1)
+
+#define MFIB_TABLE_ATTRIBUTES {		         \
+    [MFIB_TABLE_ATTRIBUTE_RESYNC]  = "resync",    \
+}
+
+#define FOR_EACH_MFIB_TABLE_ATTRIBUTE(_item)      	\
+    for (_item = MFIB_TABLE_ATTRIBUTE_FIRST;		\
+	 _item < MFIB_TABLE_ATTRIBUTE_MAX;		\
+	 _item++)
+
+typedef enum mfib_table_flags_t_ {
+    MFIB_TABLE_FLAG_NONE   = 0,
+    MFIB_TABLE_FLAG_RESYNC  = (1 << MFIB_TABLE_ATTRIBUTE_RESYNC),
+} __attribute__ ((packed)) mfib_table_flags_t;
+
+extern u8* format_mfib_table_flags(u8 *s, va_list *args);
+
+/**
  * @brief
  *   A protocol Independent IP multicast FIB table
  */
@@ -56,6 +92,11 @@ typedef struct mfib_table_t_
     fib_protocol_t mft_proto;
 
     /**
+     * table falgs
+     */
+    mfib_table_flags_t mft_flags;
+
+    /**
      * number of locks on the table
      */
     u16 mft_locks[MFIB_TABLE_N_LOCKS];
@@ -66,6 +107,11 @@ typedef struct mfib_table_t_
     u32 mft_table_id;
 
     /**
+     * resync epoch
+     */
+    u32 mft_epoch;
+
+    /**
      * Index into FIB vector.
      */
     fib_node_index_t mft_index;
@@ -282,6 +328,46 @@ extern void mfib_table_flush(u32 fib_index,
 
 /**
  * @brief
+ *  Resync all entries from a table for the source
+ *  this is the mark part of the mark and sweep algorithm.
+ *  All entries in this FIB that are sourced by 'source' are marked
+ *  as stale.
+ *
+ * @param fib_index
+ *  The index of the FIB
+ *
+ * @paran proto
+ *  The protocol of the entries in the table
+ *
+ * @param source
+ *  the source to flush
+ */
+extern void mfib_table_mark(u32 fib_index,
+                            fib_protocol_t proto,
+                            mfib_source_t source);
+
+/**
+ * @brief
+ *  Signal that the table has converged, i.e. all updates are complete.
+ *  this is the sweep part of the mark and sweep algorithm.
+ *  All entries in this FIB that are sourced by 'source' and marked
+ *  as stale are flushed.
+ *
+ * @param fib_index
+ *  The index of the FIB
+ *
+ * @paran proto
+ *  The protocol of the entries in the table
+ *
+ * @param source
+ *  the source to flush
+ */
+extern void mfib_table_sweep(u32 fib_index,
+                             fib_protocol_t proto,
+                             mfib_source_t source);
+
+/**
+ * @brief
  *  Get the index of the FIB bound to the interface
  *
  * @paran proto
@@ -465,8 +551,8 @@ extern mfib_table_t *mfib_table_get(fib_node_index_t index,
 /**
  * @brief Call back function when walking entries in a FIB table
  */
-typedef int (*mfib_table_walk_fn_t)(fib_node_index_t fei,
-                                    void *ctx);
+typedef walk_rc_t (*mfib_table_walk_fn_t)(fib_node_index_t fei,
+                                          void *ctx);
 
 /**
  * @brief Walk all entries in a FIB table