zebra: Sweep our nexthop objects out on restart

On restart, if we failed to remove any nexthop objects due
to a kill -9 or such event, sweep them if we aren't using them.
Add a proto field to handle this and remove the is_kernel bool.

Add a dupicate flag that indicates this nexthop group is only
present in our ID hashtable. It is a dupicate nexthop we received
from the kernel, therefore we cannot hash on it.

Make the idcounter globally accessible so that kernel updates
increment it as soon as we receive them, not when we handle them.

Signed-off-by: Stephen Worley <sworley@cumulusnetworks.com>

1 files changed, 23 insertions, 8 deletions

diff --git a/zebra/zebra_nhg.h b/zebra/zebra_nhg.h
index e06d415b28..1fdda276cb 100644
--- a/zebra/zebra_nhg.h
+++ b/zebra/zebra_nhg.h
@@ -45,7 +45,7 @@ struct nhg_hash_entry {
 	uint32_t id;
 	afi_t afi;
 	vrf_id_t vrf_id;
-	bool is_kernel_nh;
+	int type;
 
 	struct nexthop_group *nhg;
 
@@ -77,24 +77,32 @@ struct nhg_hash_entry {
  * and correct( ie no recursive pointer ) or a nexthop that is recursively
  * resolved and correct.
  */
-#define NEXTHOP_GROUP_VALID 0x1
+#define NEXTHOP_GROUP_VALID (1 << 0)
 /*
  * Has this nexthop group been installed?  At this point in time, this
  * means that the data-plane has been told about this nexthop group
  * and it's possible usage by a route entry.
  */
-#define NEXTHOP_GROUP_INSTALLED 0x2
+#define NEXTHOP_GROUP_INSTALLED (1 << 1)
 /*
  * Has the nexthop group been queued to be send to the FIB?
  * The NEXTHOP_GROUP_VALID flag should also be set by this point.
  */
-#define NEXTHOP_GROUP_QUEUED 0x4
+#define NEXTHOP_GROUP_QUEUED (1 << 2)
 /*
  * Is this a nexthop that is recursively resolved?
  */
-#define NEXTHOP_GROUP_RECURSIVE 0x8
+#define NEXTHOP_GROUP_RECURSIVE (1 << 3)
+/*
+ * This is a duplicate nexthop we got from the kernel, we are only tracking
+ * it in our ID hash table, it is unusable by our routes.
+ */
+#define NEXTHOP_GROUP_DUPLICATE (1 << 4)
 };
 
+/* Was this one we created, either this session or previously? */
+#define ZEBRA_NHG_CREATED(NHE) ((NHE->type) == ZEBRA_ROUTE_NHG)
+
 /* Abstraction for connected trees */
 struct nhg_connected {
 	struct nhg_connected_tree_item tree_item;
@@ -135,7 +143,11 @@ struct nhg_ctx {
 
 	vrf_id_t vrf_id;
 	afi_t afi;
-	bool is_kernel_nh;
+	/*
+	 * This should only every be ZEBRA_ROUTE_NHG unless we get a a kernel
+	 * created nexthop not made by us.
+	 */
+	int type;
 
 	/* If its a group array, how many? */
 	uint8_t count;
@@ -216,7 +228,8 @@ extern int nhg_ctx_process(struct nhg_ctx *ctx);
 /* Find via kernel nh creation */
 extern int zebra_nhg_kernel_find(uint32_t id, struct nexthop *nh,
 				 struct nh_grp *grp, uint8_t count,
-				 vrf_id_t vrf_id, afi_t afi);
+				 vrf_id_t vrf_id, afi_t afi, int type,
+				 int startup);
 
 /* Find via route creation */
 extern struct nhg_hash_entry *
@@ -242,9 +255,11 @@ extern uint8_t zebra_nhg_nhe2grp(struct nh_grp *grp, struct nhg_hash_entry *nhe,
 void zebra_nhg_install_kernel(struct nhg_hash_entry *nhe);
 void zebra_nhg_uninstall_kernel(struct nhg_hash_entry *nhe);
 
-void zebra_nhg_cleanup_tables(void);
+void zebra_nhg_cleanup_tables(struct hash *hash);
 
 /* Forward ref of dplane update context type */
 struct zebra_dplane_ctx;
 void zebra_nhg_dplane_result(struct zebra_dplane_ctx *ctx);
+
+void zebra_nhg_sweep_table(struct hash *hash);
 #endif