/*-
* Copyright (c) 2010-2020 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.
*/

#ifdef _KERNEL
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: npf_worker.c,v 1.10 2020/08/27 18:49:36 riastradh Exp $");

#include <sys/param.h>
#include <sys/types.h>

#include <sys/mutex.h>
#include <sys/kmem.h>
#include <sys/kernel.h>
#include <sys/kthread.h>
#include <sys/cprng.h>
#endif

#include "npf_impl.h"

typedef struct npf_worker {
       kmutex_t                lock;
       kcondvar_t              cv;
       kcondvar_t              exit_cv;
       bool                    exit;
       LIST_HEAD(, npf)        instances;
       unsigned                worker_count;
       lwp_t *                 worker[];
} npf_workerinfo_t;

#define NPF_GC_MINWAIT          (10)            // 10 ms
#define NPF_GC_MAXWAIT          (10 * 1000)     // 10 sec

/*
* Flags for the npf_t::worker_flags field.
*/
#define WFLAG_ACTIVE            0x01    // instance enqueued for workers
#define WFLAG_INITED            0x02    // worker setup the instance
#define WFLAG_REMOVE            0x04    // remove the instance

static void                     npf_worker(void *)      __dead;
static npf_workerinfo_t *       worker_info             __read_mostly;

int
npf_worker_sysinit(unsigned nworkers)
{
       const size_t len = offsetof(npf_workerinfo_t, worker[nworkers]);
       npf_workerinfo_t *winfo;

       KASSERT(worker_info == NULL);

       if (!nworkers) {
               return 0;
       }

       winfo = kmem_zalloc(len, KM_SLEEP);
       winfo->worker_count = nworkers;
       mutex_init(&winfo->lock, MUTEX_DEFAULT, IPL_SOFTNET);
       cv_init(&winfo->exit_cv, "npfgcx");
       cv_init(&winfo->cv, "npfgcw");
       LIST_INIT(&winfo->instances);
       worker_info = winfo;

       for (unsigned i = 0; i < nworkers; i++) {
               if (kthread_create(PRI_NONE, KTHREAD_MPSAFE | KTHREAD_MUSTJOIN,
                   NULL, npf_worker, winfo, &winfo->worker[i], "npfgc%u", i)) {
                       npf_worker_sysfini();
                       return ENOMEM;
               }
       }
       return 0;
}

void
npf_worker_sysfini(void)
{
       npf_workerinfo_t *winfo = worker_info;
       unsigned nworkers;

       if (!winfo) {
               return;
       }

       /* Notify the workers to exit. */
       mutex_enter(&winfo->lock);
       winfo->exit = true;
       cv_broadcast(&winfo->cv);
       mutex_exit(&winfo->lock);

       /* Wait for them to finish and then destroy. */
       nworkers = winfo->worker_count;
       for (unsigned i = 0; i < nworkers; i++) {
               lwp_t *worker;

               if ((worker = winfo->worker[i]) != NULL) {
                       kthread_join(worker);
               }
       }
       cv_destroy(&winfo->cv);
       cv_destroy(&winfo->exit_cv);
       mutex_destroy(&winfo->lock);
       kmem_free(winfo, offsetof(npf_workerinfo_t, worker[nworkers]));
       worker_info = NULL;
}

int
npf_worker_addfunc(npf_t *npf, npf_workfunc_t work)
{
       KASSERTMSG(npf->worker_flags == 0,
           "the task must be added before the npf_worker_enlist() call");

       for (unsigned i = 0; i < NPF_MAX_WORKS; i++) {
               if (npf->worker_funcs[i] == NULL) {
                       npf->worker_funcs[i] = work;
                       return 0;
               }
       }
       return -1;
}

void
npf_worker_signal(npf_t *npf)
{
       npf_workerinfo_t *winfo = worker_info;

       if ((npf->worker_flags & WFLAG_ACTIVE) == 0) {
               return;
       }
       KASSERT(winfo != NULL);

       mutex_enter(&winfo->lock);
       cv_signal(&winfo->cv);
       mutex_exit(&winfo->lock);
}

/*
* npf_worker_enlist: add the NPF instance for worker(s) to process.
*/
void
npf_worker_enlist(npf_t *npf)
{
       npf_workerinfo_t *winfo = worker_info;

       KASSERT(npf->worker_flags == 0);
       if (!winfo) {
               return;
       }

       mutex_enter(&winfo->lock);
       LIST_INSERT_HEAD(&winfo->instances, npf, worker_entry);
       npf->worker_flags |= WFLAG_ACTIVE;
       mutex_exit(&winfo->lock);
}

/*
* npf_worker_discharge: remove the NPF instance the list for workers.
*
* => May block waiting for a worker to finish processing the instance.
*/
void
npf_worker_discharge(npf_t *npf)
{
       npf_workerinfo_t *winfo = worker_info;

       if ((npf->worker_flags & WFLAG_ACTIVE) == 0) {
               return;
       }
       KASSERT(winfo != NULL);

       /*
        * Notify the worker(s) that we are removing this instance.
        */
       mutex_enter(&winfo->lock);
       KASSERT(npf->worker_flags & WFLAG_ACTIVE);
       npf->worker_flags |= WFLAG_REMOVE;
       cv_broadcast(&winfo->cv);

       /* Wait for a worker to process this request. */
       while (npf->worker_flags & WFLAG_ACTIVE) {
               cv_wait(&winfo->exit_cv, &winfo->lock);
       }
       mutex_exit(&winfo->lock);
       KASSERT(npf->worker_flags == 0);
}

static void
remove_npf_instance(npf_workerinfo_t *winfo, npf_t *npf)
{
       KASSERT(mutex_owned(&winfo->lock));
       KASSERT(npf->worker_flags & WFLAG_ACTIVE);
       KASSERT(npf->worker_flags & WFLAG_REMOVE);

       /*
        * Remove the NPF instance:
        * - Release any structures owned by the worker.
        * - Remove the instance from the list.
        * - Notify any thread waiting for removal to complete.
        */
       if (npf->worker_flags & WFLAG_INITED) {
               npfk_thread_unregister(npf);
       }
       LIST_REMOVE(npf, worker_entry);
       npf->worker_flags = 0;
       cv_broadcast(&winfo->exit_cv);
}

static unsigned
process_npf_instance(npf_workerinfo_t *winfo, npf_t *npf)
{
       npf_workfunc_t work;

       KASSERT(mutex_owned(&winfo->lock));

       if (npf->worker_flags & WFLAG_REMOVE) {
               remove_npf_instance(winfo, npf);
               return NPF_GC_MAXWAIT;
       }

       if ((npf->worker_flags & WFLAG_INITED) == 0) {
               npfk_thread_register(npf);
               npf->worker_flags |= WFLAG_INITED;
       }

       /* Run the jobs. */
       for (unsigned i = 0; i < NPF_MAX_WORKS; i++) {
               if ((work = npf->worker_funcs[i]) == NULL) {
                       break;
               }
               work(npf);
       }

       return MAX(MIN(npf->worker_wait_time, NPF_GC_MAXWAIT), NPF_GC_MINWAIT);
}

/*
* npf_worker: the main worker loop, processing enlisted NPF instances.
*
* XXX: Currently, npf_workerinfo_t::lock would serialize all workers,
* so there is no point to have more than one worker; but there might
* not be much point anyway.
*/
static void
npf_worker(void *arg)
{
       npf_workerinfo_t *winfo = arg;
       npf_t *npf;

       mutex_enter(&winfo->lock);
       for (;;) {
               unsigned wait_time = NPF_GC_MAXWAIT;

               /*
                * Iterate all instances.  We do not use LIST_FOREACH here,
                * since the instance can be removed.
                */
               npf = LIST_FIRST(&winfo->instances);
               while (npf) {
                       npf_t *next = LIST_NEXT(npf, worker_entry);
                       unsigned i_wait_time = process_npf_instance(winfo, npf);
                       wait_time = MIN(wait_time, i_wait_time);
                       npf = next;
               }

               /*
                * Sleep and periodically wake up, unless we get notified.
                */
               if (winfo->exit) {
                       break;
               }
               cv_timedwait(&winfo->cv, &winfo->lock, mstohz(wait_time));
       }
       mutex_exit(&winfo->lock);

       KASSERTMSG(LIST_EMPTY(&winfo->instances),
           "NPF instances must be discharged before the npfk_sysfini() call");

       kthread_exit(0);
}