summaryrefslogtreecommitdiffstats
path: root/src/plugins/gbp/gbp_policy.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/plugins/gbp/gbp_policy.c')
-rw-r--r--src/plugins/gbp/gbp_policy.c363
1 files changed, 1 insertions, 362 deletions
diff --git a/src/plugins/gbp/gbp_policy.c b/src/plugins/gbp/gbp_policy.c
index 5d5dcd0befc..fbdf3946d1d 100644
--- a/src/plugins/gbp/gbp_policy.c
+++ b/src/plugins/gbp/gbp_policy.c
@@ -16,368 +16,7 @@
#include <plugins/gbp/gbp.h>
#include <plugins/gbp/gbp_policy_dpo.h>
-#include <vnet/vxlan-gbp/vxlan_gbp_packet.h>
-
-#define foreach_gbp_policy \
- _(DENY, "deny")
-
-typedef enum
-{
-#define _(sym,str) GBP_ERROR_##sym,
- foreach_gbp_policy
-#undef _
- GBP_POLICY_N_ERROR,
-} gbp_policy_error_t;
-
-static char *gbp_policy_error_strings[] = {
-#define _(sym,string) string,
- foreach_gbp_policy
-#undef _
-};
-
-typedef enum
-{
-#define _(sym,str) GBP_POLICY_NEXT_##sym,
- foreach_gbp_policy
-#undef _
- GBP_POLICY_N_NEXT,
-} gbp_policy_next_t;
-
-/**
- * per-packet trace data
- */
-typedef struct gbp_policy_trace_t_
-{
- /* per-pkt trace data */
- u32 src_epg;
- u32 dst_epg;
- u32 acl_index;
- u32 allowed;
-} gbp_policy_trace_t;
-
-always_inline dpo_proto_t
-ethertype_to_dpo_proto (u16 etype)
-{
- etype = clib_net_to_host_u16 (etype);
-
- switch (etype)
- {
- case ETHERNET_TYPE_IP4:
- return (DPO_PROTO_IP4);
- case ETHERNET_TYPE_IP6:
- return (DPO_PROTO_IP6);
- }
-
- return (DPO_PROTO_NONE);
-}
-
-always_inline u32
-gbp_rule_l2_redirect (const gbp_rule_t * gu, vlib_buffer_t * b0)
-{
- const ethernet_header_t *eth0;
- const dpo_id_t *dpo;
- dpo_proto_t dproto;
-
- eth0 = vlib_buffer_get_current (b0);
- /* pop the ethernet header to prepare for L3 rewrite */
- vlib_buffer_advance (b0, vnet_buffer (b0)->l2.l2_len);
-
- dproto = ethertype_to_dpo_proto (eth0->type);
- dpo = &gu->gu_dpo[GBP_POLICY_NODE_L2][dproto];
-
- /* save the LB index for the next node and reset the IP flow hash
- * so it's recalculated */
- vnet_buffer (b0)->ip.adj_index[VLIB_TX] = dpo->dpoi_index;
- vnet_buffer (b0)->ip.flow_hash = 0;
-
- return (dpo->dpoi_next_node);
-}
-
-always_inline u8
-gbp_policy_is_ethertype_allowed (const gbp_contract_t * gc0, u16 ethertype)
-{
- u16 *et;
-
- vec_foreach (et, gc0->gc_allowed_ethertypes)
- {
- if (*et == ethertype)
- return (1);
- }
- return (0);
-}
-
-static uword
-gbp_policy_inline (vlib_main_t * vm,
- vlib_node_runtime_t * node,
- vlib_frame_t * frame, u8 is_port_based)
-{
- gbp_main_t *gm = &gbp_main;
- gbp_policy_main_t *gpm = &gbp_policy_main;
- u32 n_left_from, *from, *to_next;
- u32 next_index;
-
- next_index = 0;
- n_left_from = frame->n_vectors;
- from = vlib_frame_vector_args (frame);
-
- while (n_left_from > 0)
- {
- u32 n_left_to_next;
-
- vlib_get_next_frame (vm, node, next_index, to_next, n_left_to_next);
-
- while (n_left_from > 0 && n_left_to_next > 0)
- {
- const ethernet_header_t *h0;
- const gbp_endpoint_t *ge0;
- const gbp_contract_t *gc0;
- gbp_policy_next_t next0;
- gbp_contract_key_t key0;
- u32 bi0, sw_if_index0;
- vlib_buffer_t *b0;
- index_t gci0;
-
- gc0 = NULL;
- next0 = GBP_POLICY_NEXT_DENY;
- bi0 = from[0];
- to_next[0] = bi0;
- from += 1;
- to_next += 1;
- n_left_from -= 1;
- n_left_to_next -= 1;
-
- b0 = vlib_get_buffer (vm, bi0);
- h0 = vlib_buffer_get_current (b0);
- sw_if_index0 = vnet_buffer (b0)->sw_if_index[VLIB_TX];
-
- /*
- * If the A0bit is set then policy has already been applied
- * and we skip enforcement here.
- */
- if (vnet_buffer2 (b0)->gbp.flags & VXLAN_GBP_GPFLAGS_A)
- {
- next0 = vnet_l2_feature_next (b0,
- gpm->l2_output_feat_next
- [is_port_based],
- (is_port_based ?
- L2OUTPUT_FEAT_GBP_POLICY_PORT :
- L2OUTPUT_FEAT_GBP_POLICY_MAC));
- key0.as_u32 = ~0;
- goto trace;
- }
- /*
- * determine the src and dst EPG
- */
- if (is_port_based)
- ge0 = gbp_endpoint_find_itf (sw_if_index0);
- else
- ge0 = gbp_endpoint_find_mac (h0->dst_address,
- vnet_buffer (b0)->l2.bd_index);
-
- if (NULL != ge0)
- key0.gck_dst = ge0->ge_fwd.gef_epg_id;
- else
- /* If you cannot determine the destination EP then drop */
- goto trace;
-
- key0.gck_src = vnet_buffer2 (b0)->gbp.src_epg;
-
- if (EPG_INVALID != key0.gck_src)
- {
- if (PREDICT_FALSE (key0.gck_src == key0.gck_dst))
- {
- /*
- * intra-epg allowed
- */
- next0 =
- vnet_l2_feature_next (b0,
- gpm->l2_output_feat_next
- [is_port_based],
- (is_port_based ?
- L2OUTPUT_FEAT_GBP_POLICY_PORT :
- L2OUTPUT_FEAT_GBP_POLICY_MAC));
- vnet_buffer2 (b0)->gbp.flags |= VXLAN_GBP_GPFLAGS_A;
- }
- else
- {
- gci0 = gbp_contract_find (&key0);
-
- if (INDEX_INVALID != gci0)
- {
- u32 rule_match_p0, trace_bitmap0;
- fa_5tuple_opaque_t pkt_5tuple0;
- u32 acl_pos_p0, acl_match_p0;
- u8 is_ip60, l2_len0, action0;
- const gbp_rule_t *gu;
- u16 ether_type0;
- const u8 *h0;
-
- action0 = 0;
- gc0 = gbp_contract_get (gci0);
- l2_len0 = vnet_buffer (b0)->l2.l2_len;
- h0 = vlib_buffer_get_current (b0);
-
- ether_type0 = *(u16 *) (h0 + l2_len0 - 2);
-
- if (!gbp_policy_is_ethertype_allowed (gc0, ether_type0))
- {
- /*
- * black list model so drop
- */
- goto trace;
- }
-
- if ((ether_type0 ==
- clib_net_to_host_u16 (ETHERNET_TYPE_IP6))
- || (ether_type0 ==
- clib_net_to_host_u16 (ETHERNET_TYPE_IP4)))
- {
- is_ip60 =
- (ether_type0 ==
- clib_net_to_host_u16 (ETHERNET_TYPE_IP6)) ? 1 :
- 0;
- /*
- * tests against the ACL
- */
- acl_plugin_fill_5tuple_inline (gm->
- acl_plugin.p_acl_main,
- gc0->gc_lc_index, b0,
- is_ip60,
- /* is_input */ 0,
- /* is_l2_path */ 1,
- &pkt_5tuple0);
- acl_plugin_match_5tuple_inline (gm->
- acl_plugin.p_acl_main,
- gc0->gc_lc_index,
- &pkt_5tuple0,
- is_ip60, &action0,
- &acl_pos_p0,
- &acl_match_p0,
- &rule_match_p0,
- &trace_bitmap0);
-
- if (action0 > 0)
- {
- vnet_buffer2 (b0)->gbp.flags |=
- VXLAN_GBP_GPFLAGS_A;
- gu =
- gbp_rule_get (gc0->gc_rules[rule_match_p0]);
-
- switch (gu->gu_action)
- {
- case GBP_RULE_PERMIT:
- next0 = vnet_l2_feature_next
- (b0,
- gpm->l2_output_feat_next
- [is_port_based],
- (is_port_based ?
- L2OUTPUT_FEAT_GBP_POLICY_PORT :
- L2OUTPUT_FEAT_GBP_POLICY_MAC));
- break;
- case GBP_RULE_DENY:
- next0 = 0;
- break;
- case GBP_RULE_REDIRECT:
- next0 = gbp_rule_l2_redirect (gu, b0);
- break;
- }
- }
- }
- }
- }
- }
- else
- {
- /*
- * the src EPG is not set when the packet arrives on an EPG
- * uplink interface and we do not need to apply policy
- */
- next0 =
- vnet_l2_feature_next (b0,
- gpm->l2_output_feat_next[is_port_based],
- (is_port_based ?
- L2OUTPUT_FEAT_GBP_POLICY_PORT :
- L2OUTPUT_FEAT_GBP_POLICY_MAC));
- }
-
- trace:
- if (PREDICT_FALSE ((b0->flags & VLIB_BUFFER_IS_TRACED)))
- {
- gbp_policy_trace_t *t =
- vlib_add_trace (vm, node, b0, sizeof (*t));
- t->src_epg = key0.gck_src;
- t->dst_epg = key0.gck_dst;
- t->acl_index = (gc0 ? gc0->gc_acl_index : ~0),
- t->allowed = (next0 != GBP_POLICY_NEXT_DENY);
- }
-
- /* verify speculative enqueue, maybe switch current next frame */
- vlib_validate_buffer_enqueue_x1 (vm, node, next_index,
- to_next, n_left_to_next,
- bi0, next0);
- }
-
- vlib_put_next_frame (vm, node, next_index, n_left_to_next);
- }
-
- return frame->n_vectors;
-}
-
-VLIB_NODE_FN (gbp_policy_port_node) (vlib_main_t * vm,
- vlib_node_runtime_t * node,
- vlib_frame_t * frame)
-{
- return (gbp_policy_inline (vm, node, frame, 1));
-}
-
-VLIB_NODE_FN (gbp_policy_mac_node) (vlib_main_t * vm,
- vlib_node_runtime_t * node,
- vlib_frame_t * frame)
-{
- return (gbp_policy_inline (vm, node, frame, 0));
-}
-
-/* packet trace format function */
-static u8 *
-format_gbp_policy_trace (u8 * s, va_list * args)
-{
- CLIB_UNUSED (vlib_main_t * vm) = va_arg (*args, vlib_main_t *);
- CLIB_UNUSED (vlib_node_t * node) = va_arg (*args, vlib_node_t *);
- gbp_policy_trace_t *t = va_arg (*args, gbp_policy_trace_t *);
-
- s =
- format (s, "src:%d, dst:%d, acl:%d allowed:%d",
- t->src_epg, t->dst_epg, t->acl_index, t->allowed);
-
- return s;
-}
-
-/* *INDENT-OFF* */
-VLIB_REGISTER_NODE (gbp_policy_port_node) = {
- .name = "gbp-policy-port",
- .vector_size = sizeof (u32),
- .format_trace = format_gbp_policy_trace,
- .type = VLIB_NODE_TYPE_INTERNAL,
-
- .n_errors = ARRAY_LEN(gbp_policy_error_strings),
- .error_strings = gbp_policy_error_strings,
-
- .n_next_nodes = GBP_POLICY_N_NEXT,
-
- .next_nodes = {
- [GBP_POLICY_NEXT_DENY] = "error-drop",
- },
-};
-
-VLIB_REGISTER_NODE (gbp_policy_mac_node) = {
- .name = "gbp-policy-mac",
- .vector_size = sizeof (u32),
- .format_trace = format_gbp_policy_trace,
- .type = VLIB_NODE_TYPE_INTERNAL,
- .sibling_of = "gbp-policy-port",
-};
-
-/* *INDENT-ON* */
+gbp_policy_main_t gbp_policy_main;
static clib_error_t *
gbp_policy_init (vlib_main_t * vm)