summaryrefslogtreecommitdiffstats
path: root/src/vnet/mpls/mpls_api.c
blob: 4efb61786ad20f4c33265a5cbd1e85fd0990dbde (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
<
/*
 * Copyright (c) 2016 Cisco and/or its affiliates.
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
/**
 * @brief
 *
 */

#ifndef __REPLICATE_DPO_H__
#define __REPLICATE_DPO_H__

#include <vlib/vlib.h>
#include <vnet/ip/lookup.h>
#include <vnet/dpo/dpo.h>
#include <vnet/dpo/load_balance.h>
#include <vnet/fib/fib_types.h>
#include <vnet/mpls/mpls_types.h>

/**
 * replicate main
 */
typedef struct replicate_main_t_
{
    vlib_combined_counter_main_t repm_counters;

    /* per-cpu vector of cloned packets */
    u32 **clones;
} replicate_main_t;

extern replicate_main_t replicate_main;

/**
 * The number of buckets that a load-balance object can have and still
 * fit in one cache-line
 */
#define REP_NUM_INLINE_BUCKETS 4

/**
 * The FIB DPO provieds;
 *  - load-balancing over the next DPOs in the chain/graph
 *  - per-route counters
 */
typedef struct replicate_t_ {
    /**
     * number of buckets in the load-balance. always a power of 2.
     */
    u16 rep_n_buckets;

   /**
     * The protocol of packets that traverse this REP.
     * need in combination with the flow hash config to determine how to hash.
     * u8.
     */
    dpo_proto_t rep_proto;

    /**
     * The number of locks, which is approximately the number of users,
/*
 *------------------------------------------------------------------
 * mpls_api.c - mpls api
 *
 * Copyright (c) 2016 Cisco and/or its affiliates.
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 *------------------------------------------------------------------
 */

#include <vnet/vnet.h>
#include <vlibmemory/api.h>

#include <vnet/interface.h>
#include <vnet/api_errno.h>
#include <vnet/mpls/mpls.h>
#include <vnet/mpls/mpls_tunnel.h>
#include <vnet/fib/fib_table.h>
#include <vnet/fib/fib_api.h>
#include <vnet/fib/mpls_fib.h>
#include <vnet/fib/fib_path_list.h>
#include <vnet/ip/ip_types_api.h>

#include <vnet/format_fns.h>
#include <vnet/mpls/mpls.api_enum.h>
#include <vnet/mpls/mpls.api_types.h>

#define REPLY_MSG_ID_BASE mpls_main.msg_id_base
#include <vlibapi/api_helper_macros.h>

void
mpls_table_delete (u32 table_id, u8 is_api)
{
  u32 fib_index;

  /*
   * The MPLS defult table must also be explicitly created via the API.
   * So in contrast to IP, it gets no special treatment here.
   *
   * The API holds only one lock on the table.
   * i.e. it can be added many times via the API but needs to be
   * deleted only once.
   */
  fib_index = fib_table_find (FIB_PROTOCOL_MPLS, table_id);

  if (~0 != fib_index)
    {
      fib_table_unlock (fib_index,
			FIB_PROTOCOL_MPLS,
			(is_api ? FIB_SOURCE_API : FIB_SOURCE_CLI));
    }
}

void
vl_api_mpls_table_add_del_t_handler (vl_api_mpls_table_add_del_t * mp)
{
  vl_api_mpls_table_add_del_reply_t *rmp;
  vnet_main_t *vnm;
  int rv = 0;

  vnm = vnet_get_main ();
  vnm->api_errno = 0;

  if (mp->mt_is_add)
    mpls_table_create (ntohl (mp->mt_table.mt_table_id),
		       1, mp->mt_table.mt_name);
  else
    mpls_table_delete (ntohl (mp->mt_table.mt_table_id), 1);

  // NB: Nothing sets rv; none of the above returns an error

  REPLY_MACRO (VL_API_MPLS_TABLE_ADD_DEL_REPLY);
}

static int
mpls_ip_bind_unbind_handler (vnet_main_t * vnm,
			     vl_api_mpls_ip_bind_unbind_t * mp)
{
  u32 mpls_fib_index, ip_fib_index;
  fib_prefix_t pfx;

  mpls_fib_index =
    fib_table_find (FIB_PROTOCOL_MPLS, ntohl (mp->mb_mpls_table_id));

  if (~0 == mpls_fib_index)
    {
      return VNET_API_ERROR_NO_SUCH_FIB;
    }

  ip_prefix_decode (&mp->mb_prefix, &pfx);

  ip_fib_index = fib_table_find (pfx.fp_proto, ntohl (mp->mb_ip_table_id));
  if (~0 == ip_fib_index)
    return VNET_API_ERROR_NO_SUCH_FIB;

  if (mp->mb_is_bind)
    fib_table_entry_local_label_add (ip_fib_index, &pfx,
				     ntohl (mp->mb_label));
  else
    fib_table_entry_local_label_remove (ip_fib_index, &pfx,
					ntohl (mp->mb_label));

  return (0);
}

void
vl_api_mpls_ip_bind_unbind_t_handler (vl_api_mpls_ip_bind_unbind_t * mp)
{
  vl_api_mpls_ip_bind_unbind_reply_t *rmp;
  vnet_main_t *vnm;
  int rv;

  vnm = vnet_get_main ();
  vnm->api_errno = 0;

  rv = mpls_ip_bind_unbind_handler (vnm, mp);
  rv = (rv == 0) ? vnm->api_errno : rv;

  REPLY_MACRO (VL_API_MPLS_IP_BIND_UNBIND_REPLY);
}

static int
mpls_route_add_del_t_handler (vnet_main_t * vnm,
			      vl_api_mpls_route_add_del_t * mp,
			      u32 * stats_index)
{
  fib_route_path_t *rpaths = NULL, *rpath;
  vl_api_fib_path_t *apath;
  u32 fib_index;
  int rv, ii;

  fib_prefix_t pfx = {
    .fp_len = 21,
    .fp_proto = FIB_PROTOCOL_MPLS,
    .fp_eos = mp->mr_route.mr_eos,
    .fp_label = ntohl (mp->mr_route.mr_label),
  };
  if (pfx.fp_eos)
    {
      pfx.fp_payload_proto = mp->mr_route.mr_eos_proto;
    }
  else
    {
      pfx.fp_payload_proto = DPO_PROTO_MPLS;
    }

  rv = fib_api_table_id_decode (FIB_PROTOCOL_MPLS,
				ntohl (mp->mr_route.mr_table_id), &fib_index);
  if (0 != rv)
    goto out;

  vec_validate (rpaths, mp->mr_route.mr_n_paths - 1);

  for (ii = 0; ii < mp->mr_route.mr_n_paths; ii++)
    {
      apath = &mp->mr_route.mr_paths[ii];
      rpath = &rpaths[ii];

      rv = fib_api_path_decode (apath, rpath);

      if (0 != rv)
	goto out;
    }

  rv = fib_api_route_add_del (
    mp->mr_is_add, mp->mr_is_multipath, fib_index, &pfx, FIB_SOURCE_API,
    (mp->mr_route.mr_is_multicast ? FIB_ENTRY_FLAG_MULTICAST :
				    FIB_ENTRY_FLAG_NONE),
    rpaths);

  if (mp->mr_is_add && 0 == rv)
    *stats_index = fib_table_entry_get_stats_index (fib_index, &pfx);

out:
  vec_free (rpaths);

  return (rv);
}

void
vl_api_mpls_route_add_del_t_handler (vl_api_mpls_route_add_del_t * mp)
{
  vl_api_mpls_route_add_del_reply_t *rmp;
  vnet_main_t *vnm;
  u32 stats_index;
  int rv;

  vnm = vnet_get_main ();
  stats_index = ~0;

  rv = mpls_route_add_del_t_handler (vnm, mp, &stats_index);

  /* *INDENT-OFF* */
  REPLY_MACRO2 (VL_API_MPLS_ROUTE_ADD_DEL_REPLY,
  ({
    rmp->stats_index = htonl (stats_index);
  }));
  /* *INDENT-ON* */
}

void
mpls_table_create (u32 table_id, u8 is_api, const u8 * name)
{
  /*
   * The MPLS defult table must also be explicitly created via the API.
   * So in contrast to IP, it gets no special treatment here.
   */

  /*
   * The API holds only one lock on the table.
   * i.e. it can be added many times via the API but needs to be
   * deleted only once.
   */
      fib_table_find_or_create_and_lock_w_name (FIB_PROTOCOL_MPLS,
						table_id,
						(is_api ?
						 FIB_SOURCE_API :
						 FIB_SOURCE_CLI), name);
}

static void
vl_api_mpls_tunnel_add_del_t_handler (vl_api_mpls_tunnel_add_del_t * mp)
{
  u32 tunnel_sw_if_index = ~0, tunnel_index = ~0;
  vl_api_mpls_tunnel_add_del_reply_t *rmp;
  fib_route_path_t *rpath, *rpaths = NULL;
  int ii, rv = 0;

  vec_validate (rpaths, mp->mt_tunnel.mt_n_paths - 1);

  for (ii = 0; ii < mp->mt_tunnel.mt_n_paths; ii++)
    {
      rpath = &rpaths[ii];

      rv = fib_api_path_decode (&mp->mt_tunnel.mt_paths[ii], rpath);

      if (0 != rv)
	goto out;
    }
  tunnel_sw_if_index = ntohl (mp->mt_tunnel.mt_sw_if_index);

  if (mp->mt_is_add)
    {
      if (~0 == tunnel_sw_if_index)
	tunnel_sw_if_index =
	  vnet_mpls_tunnel_create (mp->mt_tunnel.mt_l2_only,
				   mp->mt_tunnel.mt_is_multicast,
				   mp->mt_tunnel.mt_tag);
      vnet_mpls_tunnel_path_add (tunnel_sw_if_index, rpaths);

      tunnel_index = vnet_mpls_tunnel_get_index (tunnel_sw_if_index);
    }
  else
    {
      tunnel_index = vnet_mpls_tunnel_get_index (tunnel_sw_if_index);
      tunnel_sw_if_index = ntohl (mp->mt_tunnel.mt_sw_if_index);
      if (!vnet_mpls_tunnel_path_remove (tunnel_sw_if_index, rpaths))
	vnet_mpls_tunnel_del (tunnel_sw_if_index);
    }

  vec_free (rpaths);

out:
  /* *INDENT-OFF* */
  REPLY_MACRO2(VL_API_MPLS_TUNNEL_ADD_DEL_REPLY,
  ({
    rmp->sw_if_index = ntohl(tunnel_sw_if_index);
    rmp->tunnel_index = ntohl(tunnel_index);
  }));
  /* *INDENT-ON* */
}

static void
  vl_api_sw_interface_set_mpls_enable_t_handler
  (vl_api_sw_interface_set_mpls_enable_t * mp)
{
  vl_api_sw_interface_set_mpls_enable_reply_t *rmp;
  int rv = 0;

  VALIDATE_SW_IF_INDEX (mp);

  rv = mpls_sw_interface_enable_disable (&mpls_main, ntohl (mp->sw_if_index),
					 mp->enable);

  BAD_SW_IF_INDEX_LABEL;
  REPLY_MACRO (VL_API_SW_INTERFACE_SET_MPLS_ENABLE_REPLY);
}

typedef struct mpls_tunnel_send_walk_ctx_t_
{
  vl_api_registration_t *reg;
  u32 sw_if_index;
  u32 context;
} mpls_tunnel_send_walk_ctx_t;

static void
send_mpls_tunnel_entry (u32 mti, void *arg)
{
  mpls_tunnel_send_walk_ctx_t *ctx;
  vl_api_mpls_tunnel_details_t *mp;
  fib_path_encode_ctx_t path_ctx = {
    .rpaths = NULL,
  };
  const mpls_tunnel_t *mt;
  fib_route_path_t *rpath;
  vl_api_fib_path_t *fp;
  u32 n;

  ctx = arg;

  mt = mpls_tunnel_get (mti);

  if (~0 != ctx->sw_if_index && mt->mt_sw_if_index != ctx->sw_if_index)
    return;

  n = fib_path_list_get_n_paths (mt->mt_path_list);

  mp = vl_msg_api_alloc (sizeof (*mp) + n * sizeof (vl_api_fib_path_t));
  clib_memset (mp, 0, sizeof (*mp) + n * sizeof (vl_api_fib_path_t));

  mp->_vl_msg_id = ntohs (REPLY_MSG_ID_BASE + VL_API_MPLS_TUNNEL_DETAILS);
  mp->context = ctx->context;

  mp->mt_tunnel.mt_n_paths = n;
  mp->mt_tunnel.mt_sw_if_index = ntohl (mt->mt_sw_if_index);
  mp->mt_tunnel.mt_tunnel_index = ntohl (mti);
  mp->mt_tunnel.mt_l2_only = ! !(MPLS_TUNNEL_FLAG_L2 & mt->mt_flags);
  mp->mt_tunnel.mt_is_multicast = ! !(MPLS_TUNNEL_FLAG_MCAST & mt->mt_flags);
  memcpy (mp->mt_tunnel.mt_tag, mt->mt_tag, sizeof (mp->mt_tunnel.mt_tag));

  fib_path_list_walk_w_ext (mt->mt_path_list,
			    &mt->mt_path_exts, fib_path_encode, &path_ctx);

  fp = mp->mt_tunnel.mt_paths;
  vec_foreach (rpath, path_ctx.rpaths)
  {
    fib_api_path_encode (rpath, fp);
    fp++;
  }

  vl_api_send_msg (ctx->reg, (u8 *) mp);

  vec_free (path_ctx.rpaths);
}

static void
vl_api_mpls_tunnel_dump_t_handler (vl_api_mpls_tunnel_dump_t * mp)
{
  vl_api_registration_t *reg;

  reg = vl_api_client_index_to_registration (mp->client_index);
  if (!reg)
    return;

  mpls_tunnel_send_walk_ctx_t ctx = {
    .reg = reg,
    .sw_if_index = ntohl (mp->sw_if_index),
    .context = mp->context,
  };
  mpls_tunnel_walk (send_mpls_tunnel_entry, &ctx);
}

static void
send_mpls_table_details (vpe_api_main_t * am,
			 vl_api_registration_t * reg,
			 u32 context, const fib_table_t * table)
{
  vl_api_mpls_table_details_t *mp;

  mp = vl_msg_api_alloc (sizeof (*mp));
  memset (mp, 0, sizeof (*mp));
  mp->_vl_msg_id = ntohs (REPLY_MSG_ID_BASE + VL_API_MPLS_TABLE_DETAILS);
  mp->context = context;

  mp->mt_table.mt_table_id = htonl (table->ft_table_id);
  memcpy (mp->mt_table.mt_name,
	  table->ft_desc,
	  clib_min (vec_len (table->ft_desc), sizeof (mp->mt_table.mt_name)));

  vl_api_send_msg (reg, (u8 *) mp);
}

static void
vl_api_mpls_table_dump_t_handler (vl_api_mpls_table_dump_t * mp)
{
  vpe_api_main_t *am = &vpe_api_main;
  vl_api_registration_t *reg;
  mpls_main_t *mm = &mpls_main;
  fib_table_t *fib_table;

  reg = vl_api_client_index_to_registration (mp->client_index);
  if (!reg)
    return;

  /* *INDENT-OFF* */
  pool_foreach (fib_table, mm->fibs)
   {
    send_mpls_table_details(am, reg, mp->context, fib_table);
  }
  /* *INDENT-ON* */
}

static void
send_mpls_route_details (vpe_api_main_t * am,
			 vl_api_registration_t * reg,
			 u32 context, fib_node_index_t fib_entry_index)
{
  fib_route_path_t *rpaths, *rpath;
  vl_api_mpls_route_details_t *mp;
  const fib_prefix_t *pfx;
  vl_api_fib_path_t *fp;
  int path_count;

  rpaths = fib_entry_encode (fib_entry_index);
  pfx = fib_entry_get_prefix (fib_entry_index);

  path_count = vec_len (rpaths);
  mp = vl_msg_api_alloc (sizeof (*mp) + path_count * sizeof (*fp));
  if (!mp)
    return;
  clib_memset (mp, 0, sizeof (*mp));
  mp->_vl_msg_id = ntohs (REPLY_MSG_ID_BASE + VL_API_MPLS_ROUTE_DETAILS);
  mp->context = context;

  mp->mr_route.mr_table_id =
    htonl (fib_table_get_table_id
	   (fib_entry_get_fib_index (fib_entry_index), pfx->fp_proto));
  mp->mr_route.mr_eos = pfx->fp_eos;
  mp->mr_route.mr_eos_proto = pfx->fp_payload_proto;
  mp->mr_route.mr_label = htonl (pfx->fp_label);

  mp->mr_route.mr_n_paths = path_count;
  fp = mp->mr_route.mr_paths;
  vec_foreach (rpath, rpaths)
  {
    fib_api_path_encode (rpath, fp);
    fp++;
  }

  vec_free (rpaths);
  vl_api_send_msg (reg, (u8 *) mp);
}

typedef struct vl_api_mpls_route_dump_table_walk_ctx_t_
{
  fib_node_index_t *lfeis;
} vl_api_mpls_route_dump_table_walk_ctx_t;

static fib_table_walk_rc_t
vl_api_mpls_route_dump_table_walk (fib_node_index_t fei, void *arg)
{
  vl_api_mpls_route_dump_table_walk_ctx_t *ctx = arg;

  vec_add1 (ctx->lfeis, fei);

  return (FIB_TABLE_WALK_CONTINUE);
}

static void
vl_api_mpls_route_dump_t_handler (vl_api_mpls_route_dump_t * mp)
{
  vpe_api_main_t *am = &vpe_api_main;
  vl_api_registration_t *reg;
  fib_node_index_t *lfeip = NULL;
  vl_api_mpls_route_dump_table_walk_ctx_t ctx = {
    .lfeis = NULL,
  };
  u32 fib_index;

  reg = vl_api_client_index_to_registration (mp->client_index);
  if (!reg)
    return;

  fib_index = fib_table_find (FIB_PROTOCOL_MPLS,
			      ntohl (mp->table.mt_table_id));

  if (INDEX_INVALID != fib_index)
    {
      fib_table_walk (fib_index,
		      FIB_PROTOCOL_MPLS,
		      vl_api_mpls_route_dump_table_walk, &ctx);
      vec_sort_with_function (ctx.lfeis, fib_entry_cmp_for_sort);

      vec_foreach (lfeip, ctx.lfeis)
      {
	send_mpls_route_details (am, reg, mp->context, *lfeip);
      }

      vec_free (ctx.lfeis);
    }
}

#include <vnet/mpls/mpls.api.c>
static clib_error_t *
mpls_api_hookup (vlib_main_t * vm)
{
  api_main_t *am = vlibapi_get_main ();

  /*
   * Trace space for 8 MPLS encap labels
   */
  am->api_trace_cfg[VL_API_MPLS_TUNNEL_ADD_DEL].size += 8 * sizeof (u32);

  /*
   * Set up the (msg_name, crc, message-id) table
   */
  REPLY_MSG_ID_BASE = setup_message_id_table ();

  return 0;
}

VLIB_API_INIT_FUNCTION (mpls_api_hookup);

/*
 * fd.io coding-style-patch-verification: ON
 *
 * Local Variables:
 * eval: (c-set-style "gnu")
 * End:
 */