/*-
* Copyright (c) 2010-2012 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This material is based upon work partially supported by The
* NetBSD Foundation under a contract with Mindaugas Rasiukevicius.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
*    notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
*    notice, this list of conditions and the following disclaimer in the
*    documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
* TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
* PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
* BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*/

/*
* NPF state engine to track connection.
*/

#ifdef _KERNEL
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: npf_state.c,v 1.23 2020/05/30 14:16:56 rmind Exp $");

#include <sys/param.h>
#include <sys/systm.h>
#include <sys/mutex.h>
#endif

#include "npf_impl.h"

/*
* Generic connection states and timeout table.
*
* Note: used for connection-less protocols.
*/

#define NPF_ANY_CONN_CLOSED             0
#define NPF_ANY_CONN_NEW                1
#define NPF_ANY_CONN_ESTABLISHED        2
#define NPF_ANY_CONN_NSTATES            3

/*
* Parameters.
*/
typedef struct {
       int             timeouts[NPF_ANY_CONN_NSTATES];
       int             gre_timeout;
} npf_state_params_t;

/*
* Generic FSM.
*/
static const uint8_t npf_generic_fsm[NPF_ANY_CONN_NSTATES][2] = {
       [NPF_ANY_CONN_CLOSED] = {
               [NPF_FLOW_FORW]         = NPF_ANY_CONN_NEW,
       },
       [NPF_ANY_CONN_NEW] = {
               [NPF_FLOW_FORW]         = NPF_ANY_CONN_NEW,
               [NPF_FLOW_BACK]         = NPF_ANY_CONN_ESTABLISHED,
       },
       [NPF_ANY_CONN_ESTABLISHED] = {
               [NPF_FLOW_FORW]         = NPF_ANY_CONN_ESTABLISHED,
               [NPF_FLOW_BACK]         = NPF_ANY_CONN_ESTABLISHED,
       },
};

/*
* State sampler for debugging.
*/
#if defined(_NPF_TESTING)
static void (*npf_state_sample)(npf_state_t *, bool) = NULL;
#define NPF_STATE_SAMPLE(n, r) if (npf_state_sample) (*npf_state_sample)(n, r);
#else
#define NPF_STATE_SAMPLE(n, r)
#endif

void
npf_state_sysinit(npf_t *npf)
{
       npf_state_params_t *params = npf_param_allocgroup(npf,
           NPF_PARAMS_GENERIC_STATE, sizeof(npf_state_params_t));
       npf_param_t param_map[] = {
               /*
                * Generic timeout (in seconds).
                */
               {
                       "state.generic.timeout.closed",
                       &params->timeouts[NPF_ANY_CONN_CLOSED],
                       .default_val = 0,
                       .min = 0, .max = INT_MAX
               },
               {
                       "state.generic.timeout.new",
                       &params->timeouts[NPF_ANY_CONN_NEW],
                       .default_val = 30,
                       .min = 0, .max = INT_MAX
               },
               {
                       "state.generic.timeout.established",
                       &params->timeouts[NPF_ANY_CONN_ESTABLISHED],
                       .default_val = 60,
                       .min = 0, .max = INT_MAX
               },
               {
                       "state.generic.timeout.gre",
                       &params->gre_timeout,
                       .default_val = 24 * 60 * 60,
                       .min = 0, .max = INT_MAX
               },
       };
       npf_param_register(npf, param_map, __arraycount(param_map));
       npf_state_tcp_sysinit(npf);
}

void
npf_state_sysfini(npf_t *npf)
{
       const size_t len = sizeof(npf_state_params_t);
       npf_param_freegroup(npf, NPF_PARAMS_GENERIC_STATE, len);
       npf_state_tcp_sysfini(npf);
}

/*
* npf_state_init: initialise the state structure.
*
* Should normally be called on a first packet, which also determines the
* direction in a case of connection-orientated protocol.  Returns true on
* success and false otherwise (e.g. if protocol is not supported).
*/
bool
npf_state_init(npf_cache_t *npc, npf_state_t *nst)
{
       const int proto = npc->npc_proto;
       bool ret;

       KASSERT(npf_iscached(npc, NPC_IP46));
       KASSERT(npf_iscached(npc, NPC_LAYER4));

       memset(nst, 0, sizeof(npf_state_t));

       switch (proto) {
       case IPPROTO_TCP:
               /* Pass to TCP state tracking engine. */
               ret = npf_state_tcp(npc, nst, NPF_FLOW_FORW);
               break;
       case IPPROTO_UDP:
       case IPPROTO_ICMP:
       case IPPROTO_GRE:
               /* Generic. */
               nst->nst_state = npf_generic_fsm[nst->nst_state][NPF_FLOW_FORW];
               ret = true;
               break;
       default:
               ret = false;
       }
       NPF_STATE_SAMPLE(nst, ret);
       return ret;
}

void
npf_state_destroy(npf_state_t *nst)
{
       nst->nst_state = 0;
}

/*
* npf_state_inspect: inspect the packet according to the protocol state.
*
* Return true if packet is considered to match the state (e.g. for TCP,
* the packet belongs to the tracked connection) and false otherwise.
*/
bool
npf_state_inspect(npf_cache_t *npc, npf_state_t *nst, const npf_flow_t flow)
{
       const int proto = npc->npc_proto;
       bool ret;

       switch (proto) {
       case IPPROTO_TCP:
               /* Pass to TCP state tracking engine. */
               ret = npf_state_tcp(npc, nst, flow);
               break;
       case IPPROTO_UDP:
       case IPPROTO_ICMP:
       case IPPROTO_GRE:
               /* Generic. */
               nst->nst_state = npf_generic_fsm[nst->nst_state][flow];
               ret = true;
               break;
       default:
               ret = false;
       }
       NPF_STATE_SAMPLE(nst, ret);

       return ret;
}

/*
* npf_state_etime: return the expiration time depending on the state.
*/
int
npf_state_etime(npf_t *npf, const npf_state_t *nst, const int proto)
{
       const npf_state_params_t *params;
       const unsigned state = nst->nst_state;
       int timeout = 0;

       switch (proto) {
       case IPPROTO_TCP:
               /* Pass to TCP state tracking engine. */
               timeout = npf_state_tcp_timeout(npf, nst);
               break;
       case IPPROTO_UDP:
       case IPPROTO_ICMP:
               /* Generic. */
               params = npf->params[NPF_PARAMS_GENERIC_STATE];
               timeout = params->timeouts[state];
               break;
       case IPPROTO_GRE:
               params = npf->params[NPF_PARAMS_GENERIC_STATE];
               timeout = params->gre_timeout;
               break;
       default:
               KASSERT(false);
       }
       return timeout;
}

void
npf_state_dump(const npf_state_t *nst)
{
#if defined(DDB) || defined(_NPF_TESTING)
       const npf_tcpstate_t *fst = &nst->nst_tcpst[0];
       const npf_tcpstate_t *tst = &nst->nst_tcpst[1];

       printf("\tstate (%p) %d:\n\t\t"
           "F { end %u maxend %u mwin %u wscale %u }\n\t\t"
           "T { end %u maxend %u mwin %u wscale %u }\n",
           nst, nst->nst_state,
           fst->nst_end, fst->nst_maxend, fst->nst_maxwin, fst->nst_wscale,
           tst->nst_end, tst->nst_maxend, tst->nst_maxwin, tst->nst_wscale
       );
#endif
}

#if defined(_NPF_TESTING)
void
npf_state_setsampler(void (*func)(npf_state_t *, bool))
{
       npf_state_sample = func;
}
#endif