summaryrefslogtreecommitdiffstats
path: root/build-root
diff options
context:
space:
mode:
authorVincent JARDIN <vincent.jardin@6wind.com>2016-01-14 17:01:08 -0800
committerEd Warnicke <eaw@cisco.com>2016-01-18 13:25:08 +0000
commit1d3be19c77e965309ba7f2fd43994b4a7952aade (patch)
treee25c8bfdc3e4f99dc484a03a5e73d35e0d519a83 /build-root
parent36de03819fd017a464dbd9eedfbb0bf492863fb5 (diff)
build - Vagrant template to get more emulated PCI NICs
For some testing, it an be usefull to have more physical NICS that DPDK's PMDs can bind to. Example to run vpp within a VM with 3 emulated NICs: export VPP_VAGRANT_NICS=3 vagrant up Change-Id: I82d70f21c0a9ceba126ab6620c3b869d590d8de1 Signed-off-by: Vincent JARDIN <vincent.jardin@6wind.com>
Diffstat (limited to 'build-root')
-rw-r--r--build-root/vagrant/Vagrantfile9
1 files changed, 9 insertions, 0 deletions
diff --git a/build-root/vagrant/Vagrantfile b/build-root/vagrant/Vagrantfile
index 4e2669b513a..f00233d632c 100644
--- a/build-root/vagrant/Vagrantfile
+++ b/build-root/vagrant/Vagrantfile
@@ -22,6 +22,15 @@ Vagrant.configure(2) do |config|
config.cache.scope = :box
end
+ # Define some physical ports for your VMs to be used by DPDK
+ nics = 0
+ if ENV.key?('VPP_VAGRANT_NICS')
+ nics = ENV['VPP_VAGRANT_NICS'].to_i(10)
+ end
+ for i in 1..nics
+ config.vm.network "private_network", type: "dhcp"
+ end
+
# use http proxy if avaiable
if ENV['http_proxy'] && Vagrant.has_plugin?("vagrant-proxyconf")
config.proxy.http = "$http_proxy"
href='#n183'>183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 
/*
 *------------------------------------------------------------------
 * Copyright (c) 2017 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 <stdlib.h>
#include <stdio.h>
#include <stdint.h>
#include <arpa/inet.h>
#include <stddef.h>
#include <assert.h>

#include <vpp-api/vapi/vapi_dbg.h>
#include <vpp-api/vapi/vapi.h>
#include <vpp-api/vapi/vapi_internal.h>
#include <vppinfra/types.h>
#include <vppinfra/pool.h>
#include <vlib/vlib.h>
#include <vlibapi/api_common.h>
#include <vlibmemory/memory_client.h>

#include <vapi/memclnt.api.vapi.h>

/* we need to use control pings for some stuff and because we're forced to put
 * the code in headers, we need a way to be able to grab the ids of these
 * messages - so declare them here as extern */
vapi_msg_id_t vapi_msg_id_control_ping = 0;
vapi_msg_id_t vapi_msg_id_control_ping_reply = 0;

DEFINE_VAPI_MSG_IDS_MEMCLNT_API_JSON;
DEFINE_VAPI_MSG_IDS_VPE_API_JSON;

struct
{
  size_t count;
  vapi_message_desc_t **msgs;
  size_t max_len_name_with_crc;
} __vapi_metadata;

typedef struct
{
  u32 context;
  vapi_cb_t callback;
  void *callback_ctx;
  bool is_dump;
} vapi_req_t;

static const u32 context_counter_mask = (1 << 31);

typedef struct
{
  vapi_error_e (*cb) (vapi_ctx_t ctx, void *callback_ctx, vapi_msg_id_t id,
		      void *payload);
  void *ctx;
} vapi_generic_cb_with_ctx;

typedef struct
{
  vapi_error_e (*cb) (vapi_ctx_t ctx, void *callback_ctx, void *payload);
  void *ctx;
} vapi_event_cb_with_ctx;

struct vapi_ctx_s
{
  vapi_mode_e mode;
  int requests_size;		/* size of the requests array (circular queue) */
  int requests_start;		/* index of first request */
  int requests_count;		/* number of used slots */
  vapi_req_t *requests;
  u32 context_counter;
  vapi_generic_cb_with_ctx generic_cb;
  vapi_event_cb_with_ctx *event_cbs;
  u16 *vapi_msg_id_t_to_vl_msg_id;
  u16 vl_msg_id_max;
  vapi_msg_id_t *vl_msg_id_to_vapi_msg_t;
  bool connected;
  bool handle_keepalives;
  pthread_mutex_t requests_mutex;
};

u32
vapi_gen_req_context (vapi_ctx_t ctx)
{
  ++ctx->context_counter;
  ctx->context_counter %= context_counter_mask;
  return ctx->context_counter | context_counter_mask;
}

size_t
vapi_get_request_count (vapi_ctx_t ctx)
{
  return ctx->requests_count;
}

bool
vapi_requests_full (vapi_ctx_t ctx)
{
  return (ctx->requests_count == ctx->requests_size);
}

bool
vapi_requests_empty (vapi_ctx_t ctx)
{
  return (0 == ctx->requests_count);
}

static int
vapi_requests_end (vapi_ctx_t ctx)
{
  return (ctx->requests_start + ctx->requests_count) % ctx->requests_size;
}

void
vapi_store_request (vapi_ctx_t ctx, u32 context, bool is_dump,
		    vapi_cb_t callback, void *callback_ctx)
{
  assert (!vapi_requests_full (ctx));
  /* if the mutex is not held, bad things will happen */
  assert (0 != pthread_mutex_trylock (&ctx->requests_mutex));
  const int requests_end = vapi_requests_end (ctx);
  vapi_req_t *slot = &ctx->requests[requests_end];
  slot->is_dump = is_dump;
  slot->context = context;
  slot->callback = callback;
  slot->callback_ctx = callback_ctx;
  VAPI_DBG ("stored@%d: context:%x (start is @%d)", requests_end, context,
	    ctx->requests_start);
  ++ctx->requests_count;
  assert (!vapi_requests_empty (ctx));
}

#if VAPI_DEBUG_ALLOC
struct to_be_freed_s;
struct to_be_freed_s
{
  void *v;
  struct to_be_freed_s *next;
};

static struct to_be_freed_s *to_be_freed = NULL;

void
vapi_add_to_be_freed (void *v)
{
  struct to_be_freed_s *prev = NULL;
  struct to_be_freed_s *tmp;
  tmp = to_be_freed;
  while (tmp && tmp->v)
    {
      prev = tmp;
      tmp = tmp->next;
    }
  if (!tmp)
    {
      if (!prev)
	{
	  tmp = to_be_freed = calloc (1, sizeof (*to_be_freed));
	}
      else
	{
	  tmp = prev->next = calloc (1, sizeof (*to_be_freed));
	}
    }
  VAPI_DBG ("To be freed %p", v);
  tmp->v = v;
}

void
vapi_trace_free (void *v)
{
  struct to_be_freed_s *tmp = to_be_freed;
  while (tmp && tmp->v != v)
    {
      tmp = tmp->next;
    }
  if (tmp && tmp->v == v)
    {
      VAPI_DBG ("Freed %p", v);
      tmp->v = NULL;
    }
  else
    {
      VAPI_ERR ("Trying to free untracked pointer %p", v);
      abort ();
    }
}

void
vapi_to_be_freed_validate ()
{
  struct to_be_freed_s *tmp = to_be_freed;
  while (tmp)
    {
      if (tmp->v)
	{
	  VAPI_ERR ("Unfreed msg %p!", tmp->v);
	}
      tmp = tmp->next;
    }
}

#endif

void *
vapi_msg_alloc (vapi_ctx_t ctx, size_t size)
{
  if (!ctx->connected)
    {
      return NULL;
    }
  void *rv = vl_msg_api_alloc_or_null (size);
  if (rv)
    {
      clib_memset (rv, 0, size);
    }
  return rv;
}

void
vapi_msg_free (vapi_ctx_t ctx, void *msg)
{
  if (!ctx->connected)
    {
      return;
    }
#if VAPI_DEBUG_ALLOC
  vapi_trace_free (msg);
#endif
  vl_msg_api_free (msg);
}

vapi_msg_id_t
vapi_lookup_vapi_msg_id_t (vapi_ctx_t ctx, u16 vl_msg_id)
{
  if (vl_msg_id <= ctx->vl_msg_id_max)
    {
      return ctx->vl_msg_id_to_vapi_msg_t[vl_msg_id];
    }
  return VAPI_INVALID_MSG_ID;
}

vapi_error_e
vapi_ctx_alloc (vapi_ctx_t * result)
{
  vapi_ctx_t ctx = calloc (1, sizeof (struct vapi_ctx_s));
  if (!ctx)
    {
      return VAPI_ENOMEM;
    }
  ctx->context_counter = 0;
  ctx->vapi_msg_id_t_to_vl_msg_id =
    malloc (__vapi_metadata.count *
	    sizeof (*ctx->vapi_msg_id_t_to_vl_msg_id));
  if (!ctx->vapi_msg_id_t_to_vl_msg_id)
    {
      goto fail;
    }
  clib_memset (ctx->vapi_msg_id_t_to_vl_msg_id, ~0,
	       __vapi_metadata.count *
	       sizeof (*ctx->vapi_msg_id_t_to_vl_msg_id));
  ctx->event_cbs = calloc (__vapi_metadata.count, sizeof (*ctx->event_cbs));
  if (!ctx->event_cbs)
    {
      goto fail;
    }
  pthread_mutex_init (&ctx->requests_mutex, NULL);
  *result = ctx;
  return VAPI_OK;
fail:
  vapi_ctx_free (ctx);
  return VAPI_ENOMEM;
}

void
vapi_ctx_free (vapi_ctx_t ctx)
{
  assert (!ctx->connected);
  free (ctx->requests);
  free (ctx->vapi_msg_id_t_to_vl_msg_id);
  free (ctx->event_cbs);
  free (ctx->vl_msg_id_to_vapi_msg_t);
  pthread_mutex_destroy (&ctx->requests_mutex);
  free (ctx);
}

bool
vapi_is_msg_available (vapi_ctx_t ctx, vapi_msg_id_t id)
{
  return vapi_lookup_vl_msg_id (ctx, id) != UINT16_MAX;
}

vapi_error_e
vapi_connect (vapi_ctx_t ctx, const char *name,
	      const char *chroot_prefix,
	      int max_outstanding_requests,
	      int response_queue_size, vapi_mode_e mode,
	      bool handle_keepalives)
{
  if (response_queue_size <= 0 || max_outstanding_requests <= 0)
    {
      return VAPI_EINVAL;
    }
  if (!clib_mem_get_per_cpu_heap () && !clib_mem_init (0, 1024 * 1024 * 32))
    {
      return VAPI_ENOMEM;
    }
  ctx->requests_size = max_outstanding_requests;
  const size_t size = ctx->requests_size * sizeof (*ctx->requests);
  void *tmp = realloc (ctx->requests, size);
  if (!tmp)
    {
      return VAPI_ENOMEM;
    }
  ctx->requests = tmp;
  clib_memset (ctx->requests, 0, size);
  /* coverity[MISSING_LOCK] - 177211 requests_mutex is not needed here */
  ctx->requests_start = ctx->requests_count = 0;
  if (chroot_prefix)
    {
      VAPI_DBG ("set memory root path `%s'", chroot_prefix);
      vl_set_memory_root_path ((char *) chroot_prefix);
    }
  static char api_map[] = "/vpe-api";
  VAPI_DBG ("client api map `%s'", api_map);
  if ((vl_client_api_map (api_map)) < 0)
    {
      return VAPI_EMAP_FAIL;
    }
  VAPI_DBG ("connect client `%s'", name);
  if (vl_client_connect ((char *) name, 0, response_queue_size) < 0)
    {
      vl_client_api_unmap ();
      return VAPI_ECON_FAIL;
    }
#if VAPI_DEBUG_CONNECT
  VAPI_DBG ("start probing messages");
#endif
  int rv;
  int i;
  for (i = 0; i < __vapi_metadata.count; ++i)
    {
      vapi_message_desc_t *m = __vapi_metadata.msgs[i];
      u8 scratch[m->name_with_crc_len + 1];
      memcpy (scratch, m->name_with_crc, m->name_with_crc_len + 1);
      u32 id = vl_msg_api_get_msg_index (scratch);
      if (VAPI_INVALID_MSG_ID != id)
	{
	  if (id > UINT16_MAX)
	    {
	      VAPI_ERR ("Returned vl_msg_id `%u' > UINT16MAX `%u'!", id,
			UINT16_MAX);
	      rv = VAPI_EINVAL;
	      goto fail;
	    }
	  if (id > ctx->vl_msg_id_max)
	    {
	      vapi_msg_id_t *tmp = realloc (ctx->vl_msg_id_to_vapi_msg_t,
					    sizeof
					    (*ctx->vl_msg_id_to_vapi_msg_t) *
					    (id + 1));
	      if (!tmp)
		{
		  rv = VAPI_ENOMEM;
		  goto fail;
		}
	      ctx->vl_msg_id_to_vapi_msg_t = tmp;
	      ctx->vl_msg_id_max = id;
	    }
	  ctx->vl_msg_id_to_vapi_msg_t[id] = m->id;
	  ctx->vapi_msg_id_t_to_vl_msg_id[m->id] = id;
#if VAPI_DEBUG_CONNECT
	  VAPI_DBG ("Message `%s' has vl_msg_id `%u'", m->name_with_crc,
		    (unsigned) id);
#endif
	}
      else
	{
	  ctx->vapi_msg_id_t_to_vl_msg_id[m->id] = UINT16_MAX;
	  VAPI_DBG ("Message `%s' not available", m->name_with_crc);
	}
    }
#if VAPI_DEBUG_CONNECT
  VAPI_DBG ("finished probing messages");
#endif
  if (!vapi_is_msg_available (ctx, vapi_msg_id_control_ping) ||
      !vapi_is_msg_available (ctx, vapi_msg_id_control_ping_reply))
    {
      VAPI_ERR
	("control ping or control ping reply not available, cannot connect");
      rv = VAPI_EINCOMPATIBLE;
      goto fail;
    }
  ctx->mode = mode;
  ctx->connected = true;
  if (vapi_is_msg_available (ctx, vapi_msg_id_memclnt_keepalive))
    {
      ctx->handle_keepalives = handle_keepalives;
    }
  else
    {
      ctx->handle_keepalives = false;
    }
  return VAPI_OK;
fail:
  vl_client_disconnect ();
  vl_client_api_unmap ();
  return rv;
}

vapi_error_e
vapi_disconnect (vapi_ctx_t ctx)
{
  if (!ctx->connected)
    {
      return VAPI_EINVAL;
    }
  vl_client_disconnect ();
  vl_client_api_unmap ();
#if VAPI_DEBUG_ALLOC
  vapi_to_be_freed_validate ();
#endif
  ctx->connected = false;
  return VAPI_OK;
}

vapi_error_e
vapi_get_fd (vapi_ctx_t ctx, int *fd)
{
  return VAPI_ENOTSUP;
}

vapi_error_e
vapi_send (vapi_ctx_t ctx, void *msg)
{
  vapi_error_e rv = VAPI_OK;
  if (!ctx || !msg || !ctx->connected)
    {
      rv = VAPI_EINVAL;
      goto out;
    }
  int tmp;
  svm_queue_t *q = vlibapi_get_main ()->shmem_hdr->vl_input_queue;
#if VAPI_DEBUG
  unsigned msgid = be16toh (*(u16 *) msg);
  if (msgid <= ctx->vl_msg_id_max)
    {
      vapi_msg_id_t id = ctx->vl_msg_id_to_vapi_msg_t[msgid];
      if (id < __vapi_metadata.count)
	{
	  VAPI_DBG ("send msg@%p:%u[%s]", msg, msgid,
		    __vapi_metadata.msgs[id]->name);
	}
      else
	{
	  VAPI_DBG ("send msg@%p:%u[UNKNOWN]", msg, msgid);
	}
    }
  else
    {
      VAPI_DBG ("send msg@%p:%u[UNKNOWN]", msg, msgid);
    }
#endif
  tmp = svm_queue_add (q, (u8 *) & msg,
		       VAPI_MODE_BLOCKING == ctx->mode ? 0 : 1);
  if (tmp < 0)
    {
      rv = VAPI_EAGAIN;
    }
  else
    VL_MSG_API_POISON (msg);
out:
  VAPI_DBG ("vapi_send() rv = %d", rv);
  return rv;
}

vapi_error_e
vapi_send2 (vapi_ctx_t ctx, void *msg1, void *msg2)
{
  vapi_error_e rv = VAPI_OK;
  if (!ctx || !msg1 || !msg2 || !ctx->connected)
    {
      rv = VAPI_EINVAL;
      goto out;
    }
  svm_queue_t *q = vlibapi_get_main ()->shmem_hdr->vl_input_queue;
#if VAPI_DEBUG
  unsigned msgid1 = be16toh (*(u16 *) msg1);
  unsigned msgid2 = be16toh (*(u16 *) msg2);
  const char *name1 = "UNKNOWN";
  const char *name2 = "UNKNOWN";
  if (msgid1 <= ctx->vl_msg_id_max)
    {
      vapi_msg_id_t id = ctx->vl_msg_id_to_vapi_msg_t[msgid1];
      if (id < __vapi_metadata.count)
	{
	  name1 = __vapi_metadata.msgs[id]->name;
	}
    }
  if (msgid2 <= ctx->vl_msg_id_max)
    {
      vapi_msg_id_t id = ctx->vl_msg_id_to_vapi_msg_t[msgid2];
      if (id < __vapi_metadata.count)
	{
	  name2 = __vapi_metadata.msgs[id]->name;
	}
    }
  VAPI_DBG ("send two: %u[%s], %u[%s]", msgid1, name1, msgid2, name2);
#endif
  int tmp = svm_queue_add2 (q, (u8 *) & msg1, (u8 *) & msg2,
			    VAPI_MODE_BLOCKING == ctx->mode ? 0 : 1);
  if (tmp < 0)
    {
      rv = VAPI_EAGAIN;
    }
  else
    VL_MSG_API_POISON (msg1);
out:
  VAPI_DBG ("vapi_send() rv = %d", rv);
  return rv;
}

vapi_error_e
vapi_recv (vapi_ctx_t ctx, void **msg, size_t * msg_size,
	   svm_q_conditional_wait_t cond, u32 time)
{
  if (!ctx || !ctx->connected || !msg || !msg_size)
    {
      return VAPI_EINVAL;
    }
  vapi_error_e rv = VAPI_OK;
  api_main_t *am = vlibapi_get_main ();
  uword data;

  if (am->our_pid == 0)
    {
      return VAPI_EINVAL;
    }

  svm_queue_t *q = am->vl_input_queue;
again:
  VAPI_DBG ("doing shm queue sub");

  int tmp = svm_queue_sub (q, (u8 *) & data, cond, time);

  if (tmp == 0)
    {
      VL_MSG_API_UNPOISON ((void *) data);
#if VAPI_DEBUG_ALLOC
      vapi_add_to_be_freed ((void *) data);
#endif
      msgbuf_t *msgbuf =
	(msgbuf_t *) ((u8 *) data - offsetof (msgbuf_t, data));
      if (!msgbuf->data_len)
	{
	  vapi_msg_free (ctx, (u8 *) data);
	  return VAPI_EAGAIN;
	}
      *msg = (u8 *) data;
      *msg_size = ntohl (msgbuf->data_len);
#if VAPI_DEBUG
      unsigned msgid = be16toh (*(u16 *) * msg);
      if (msgid <= ctx->vl_msg_id_max)
	{
	  vapi_msg_id_t id = ctx->vl_msg_id_to_vapi_msg_t[msgid];
	  if (id < __vapi_metadata.count)
	    {
	      VAPI_DBG ("recv msg@%p:%u[%s]", *msg, msgid,
			__vapi_metadata.msgs[id]->name);
	    }
	  else
	    {
	      VAPI_DBG ("recv msg@%p:%u[UNKNOWN]", *msg, msgid);
	    }
	}
      else
	{
	  VAPI_DBG ("recv msg@%p:%u[UNKNOWN]", *msg, msgid);
	}
#endif
      if (ctx->handle_keepalives)
	{
	  unsigned msgid = be16toh (*(u16 *) * msg);
	  if (msgid ==
	      vapi_lookup_vl_msg_id (ctx, vapi_msg_id_memclnt_keepalive))
	    {
	      vapi_msg_memclnt_keepalive_reply *reply = NULL;
	      do
		{
		  reply = vapi_msg_alloc (ctx, sizeof (*reply));
		}
	      while (!reply);
	      reply->header.context = vapi_get_client_index (ctx);
	      reply->header._vl_msg_id =
		vapi_lookup_vl_msg_id (ctx,
				       vapi_msg_id_memclnt_keepalive_reply);
	      reply->payload.retval = 0;
	      vapi_msg_memclnt_keepalive_reply_hton (reply);
	      while (VAPI_EAGAIN == vapi_send (ctx, reply));
	      vapi_msg_free (ctx, *msg);
	      VAPI_DBG ("autohandled memclnt_keepalive");
	      goto again;
	    }
	}
    }
  else
    {
      rv = VAPI_EAGAIN;
    }
  return rv;
}

vapi_error_e
vapi_wait (vapi_ctx_t ctx, vapi_wait_mode_e mode)
{
  return VAPI_ENOTSUP;
}

static vapi_error_e
vapi_dispatch_response (vapi_ctx_t ctx, vapi_msg_id_t id,
			u32 context, void *msg)
{
  int mrv;
  if (0 != (mrv = pthread_mutex_lock (&ctx->requests_mutex)))
    {
      VAPI_DBG ("pthread_mutex_lock() failed, rv=%d:%s", mrv, strerror (mrv));
      return VAPI_MUTEX_FAILURE;
    }
  int tmp = ctx->requests_start;
  const int requests_end = vapi_requests_end (ctx);
  while (ctx->requests[tmp].context != context && tmp != requests_end)
    {
      ++tmp;
      if (tmp == ctx->requests_size)
	{
	  tmp = 0;
	}
    }
  VAPI_DBG ("dispatch, search from %d, %s at %d", ctx->requests_start,
	    ctx->requests[tmp].context == context ? "matched" : "stopped",
	    tmp);
  vapi_error_e rv = VAPI_OK;
  if (ctx->requests[tmp].context == context)
    {
      while (ctx->requests_start != tmp)
	{
	  VAPI_ERR ("No response to req with context=%u",
		    (unsigned) ctx->requests[tmp].context);
	  ctx->requests[ctx->requests_start].callback (ctx, ctx->requests
						       [ctx->
							requests_start].callback_ctx,
						       VAPI_ENORESP, true,
						       NULL);
	  clib_memset (&ctx->requests[ctx->requests_start], 0,
		       sizeof (ctx->requests[ctx->requests_start]));
	  ++ctx->requests_start;
	  --ctx->requests_count;
	  if (ctx->requests_start == ctx->requests_size)
	    {
	      ctx->requests_start = 0;
	    }
	}
      // now ctx->requests_start == tmp
      int payload_offset = vapi_get_payload_offset (id);
      void *payload = ((u8 *) msg) + payload_offset;
      bool is_last = true;
      if (ctx->requests[tmp].is_dump)
	{
	  if (vapi_msg_id_control_ping_reply == id)
	    {
	      payload = NULL;
	    }
	  else
	    {
	      is_last = false;
	    }
	}
      if (payload_offset != -1)
	{
	  rv =
	    ctx->requests[tmp].callback (ctx, ctx->requests[tmp].callback_ctx,
					 VAPI_OK, is_last, payload);
	}
      else
	{
	  /* this is a message without payload, so bend the callback a little
	   */
	  rv =
	    ((vapi_error_e (*)(vapi_ctx_t, void *, vapi_error_e, bool))
	     ctx->requests[tmp].callback) (ctx,
					   ctx->requests[tmp].callback_ctx,
					   VAPI_OK, is_last);
	}
      if (is_last)
	{
	  clib_memset (&ctx->requests[ctx->requests_start], 0,
		       sizeof (ctx->requests[ctx->requests_start]));
	  ++ctx->requests_start;
	  --ctx->requests_count;
	  if (ctx->requests_start == ctx->requests_size)
	    {
	      ctx->requests_start = 0;
	    }
	}
      VAPI_DBG ("after dispatch, req start = %d, end = %d, count = %d",
		ctx->requests_start, requests_end, ctx->requests_count);
    }
  if (0 != (mrv = pthread_mutex_unlock (&ctx->requests_mutex)))
    {
      VAPI_DBG ("pthread_mutex_unlock() failed, rv=%d:%s", mrv,
		strerror (mrv));
      abort ();			/* this really shouldn't happen */
    }
  return rv;
}

static vapi_error_e
vapi_dispatch_event (vapi_ctx_t ctx, vapi_msg_id_t id, void *msg)
{
  if (ctx->event_cbs[id].cb)
    {
      return ctx->event_cbs[id].cb (ctx, ctx->event_cbs[id].ctx, msg);
    }
  else if (ctx->generic_cb.cb)
    {
      return ctx->generic_cb.cb (ctx, ctx->generic_cb.ctx, id, msg);
    }
  else
    {
      VAPI_DBG
	("No handler/generic handler for msg id %u[%s], message ignored",
	 (unsigned) id, __vapi_metadata.msgs[id]->name);
    }
  return VAPI_OK;
}

bool
vapi_msg_is_with_context (vapi_msg_id_t id)
{
  assert (id <= __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->has_context;
}

vapi_error_e
vapi_dispatch_one (vapi_ctx_t ctx)
{
  VAPI_DBG ("vapi_dispatch_one()");
  void *msg;
  size_t size;
  vapi_error_e rv = vapi_recv (ctx, &msg, &size, SVM_Q_WAIT, 0);
  if (VAPI_OK != rv)
    {
      VAPI_DBG ("vapi_recv failed with rv=%d", rv);
      return rv;
    }
  u16 vpp_id = be16toh (*(u16 *) msg);
  if (vpp_id > ctx->vl_msg_id_max)
    {
      VAPI_ERR ("Unknown msg ID received, id `%u', out of range <0,%u>",
		(unsigned) vpp_id, (unsigned) ctx->vl_msg_id_max);
      vapi_msg_free (ctx, msg);
      return VAPI_EINVAL;
    }
  if (VAPI_INVALID_MSG_ID == (unsigned) ctx->vl_msg_id_to_vapi_msg_t[vpp_id])
    {
      VAPI_ERR ("Unknown msg ID received, id `%u' marked as not supported",
		(unsigned) vpp_id);
      vapi_msg_free (ctx, msg);
      return VAPI_EINVAL;
    }
  const vapi_msg_id_t id = ctx->vl_msg_id_to_vapi_msg_t[vpp_id];
  const size_t expect_size = vapi_get_message_size (id);
  if (size < expect_size)
    {
      VAPI_ERR
	("Invalid msg received, unexpected size `%zu' < expected min `%zu'",
	 size, expect_size);
      vapi_msg_free (ctx, msg);
      return VAPI_EINVAL;
    }
  u32 context;
  vapi_get_swap_to_host_func (id) (msg);
  if (vapi_msg_is_with_context (id))
    {
      context = *(u32 *) (((u8 *) msg) + vapi_get_context_offset (id));
      /* is this a message originating from VAPI? */
      VAPI_DBG ("dispatch, context is %x", context);
      if (context & context_counter_mask)
	{
	  rv = vapi_dispatch_response (ctx, id, context, msg);
	  goto done;
	}
    }
  rv = vapi_dispatch_event (ctx, id, msg);

done:
  vapi_msg_free (ctx, msg);
  return rv;
}

vapi_error_e
vapi_dispatch (vapi_ctx_t ctx)
{
  vapi_error_e rv = VAPI_OK;
  while (!vapi_requests_empty (ctx))
    {
      rv = vapi_dispatch_one (ctx);
      if (VAPI_OK != rv)
	{
	  return rv;
	}
    }
  return rv;
}

void
vapi_set_event_cb (vapi_ctx_t ctx, vapi_msg_id_t id,
		   vapi_event_cb callback, void *callback_ctx)
{
  vapi_event_cb_with_ctx *c = &ctx->event_cbs[id];
  c->cb = callback;
  c->ctx = callback_ctx;
}

void
vapi_clear_event_cb (vapi_ctx_t ctx, vapi_msg_id_t id)
{
  vapi_set_event_cb (ctx, id, NULL, NULL);
}

void
vapi_set_generic_event_cb (vapi_ctx_t ctx, vapi_generic_event_cb callback,
			   void *callback_ctx)
{
  ctx->generic_cb.cb = callback;
  ctx->generic_cb.ctx = callback_ctx;
}

void
vapi_clear_generic_event_cb (vapi_ctx_t ctx)
{
  ctx->generic_cb.cb = NULL;
  ctx->generic_cb.ctx = NULL;
}

u16
vapi_lookup_vl_msg_id (vapi_ctx_t ctx, vapi_msg_id_t id)
{
  assert (id < __vapi_metadata.count);
  return ctx->vapi_msg_id_t_to_vl_msg_id[id];
}

int
vapi_get_client_index (vapi_ctx_t ctx)
{
  return vlibapi_get_main ()->my_client_index;
}

bool
vapi_is_nonblocking (vapi_ctx_t ctx)
{
  return (VAPI_MODE_NONBLOCKING == ctx->mode);
}

size_t
vapi_get_max_request_count (vapi_ctx_t ctx)
{
  return ctx->requests_size - 1;
}

int
vapi_get_payload_offset (vapi_msg_id_t id)
{
  assert (id < __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->payload_offset;
}

void (*vapi_get_swap_to_host_func (vapi_msg_id_t id)) (void *msg)
{
  assert (id < __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->swap_to_host;
}

void (*vapi_get_swap_to_be_func (vapi_msg_id_t id)) (void *msg)
{
  assert (id < __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->swap_to_be;
}

size_t
vapi_get_message_size (vapi_msg_id_t id)
{
  assert (id < __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->size;
}

size_t
vapi_get_context_offset (vapi_msg_id_t id)
{
  assert (id < __vapi_metadata.count);
  return __vapi_metadata.msgs[id]->context_offset;
}

vapi_msg_id_t
vapi_register_msg (vapi_message_desc_t * msg)
{
  int i = 0;
  for (i = 0; i < __vapi_metadata.count; ++i)
    {
      if (!strcmp
	  (msg->name_with_crc, __vapi_metadata.msgs[i]->name_with_crc))
	{
	  /* this happens if somebody is linking together several objects while
	   * using the static inline headers, just fill in the already
	   * assigned id here so that all the objects are in sync */
	  msg->id = __vapi_metadata.msgs[i]->id;
	  return msg->id;
	}
    }
  vapi_msg_id_t id = __vapi_metadata.count;
  ++__vapi_metadata.count;
  __vapi_metadata.msgs =
    realloc (__vapi_metadata.msgs,
	     sizeof (*__vapi_metadata.msgs) * __vapi_metadata.count);
  __vapi_metadata.msgs[id] = msg;
  size_t s = strlen (msg->name_with_crc);
  if (s > __vapi_metadata.max_len_name_with_crc)
    {
      __vapi_metadata.max_len_name_with_crc = s;
    }
  msg->id = id;
  return id;
}

vapi_error_e
vapi_producer_lock (vapi_ctx_t ctx)
{
  int mrv;
  if (0 != (mrv = pthread_mutex_lock (&ctx->requests_mutex)))
    {
      VAPI_DBG ("pthread_mutex_lock() failed, rv=%d:%s", mrv, strerror (mrv));
      (void) mrv;		/* avoid warning if the above debug is not enabled */
      return VAPI_MUTEX_FAILURE;
    }
  return VAPI_OK;
}

vapi_error_e
vapi_producer_unlock (vapi_ctx_t ctx)
{
  int mrv;
  if (0 != (mrv = pthread_mutex_unlock (&ctx->requests_mutex)))
    {
      VAPI_DBG ("pthread_mutex_unlock() failed, rv=%d:%s", mrv,
		strerror (mrv));
      (void) mrv;		/* avoid warning if the above debug is not enabled */
      return VAPI_MUTEX_FAILURE;
    }
  return VAPI_OK;
}

size_t
vapi_get_message_count ()
{
  return __vapi_metadata.count;
}

const char *
vapi_get_msg_name (vapi_msg_id_t id)
{
  return __vapi_metadata.msgs[id]->name;
}

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