libeio/etp.c

/*
 * libetp implementation
 *
 * Copyright (c) 2007,2008,2009,2010,2011,2012,2013 Marc Alexander Lehmann <libetp@schmorp.de>
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modifica-
 * tion, 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 AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER-
 * CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO
 * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE-
 * CIAL, 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 OTH-
 * ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
 * OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * Alternatively, the contents of this file may be used under the terms of
 * the GNU General Public License ("GPL") version 2 or any later version,
 * in which case the provisions of the GPL are applicable instead of
 * the above. If you wish to allow the use of your version of this file
 * only under the terms of the GPL and not to allow others to use your
 * version of this file under the BSD license, indicate your decision
 * by deleting the provisions above and replace them with the notice
 * and other provisions required by the GPL. If you do not delete the
 * provisions above, a recipient may use your version of this file under
 * either the BSD or the GPL.
 */

#ifndef ETP_API_DECL
# define ETP_API_DECL static
#endif

#ifndef ETP_PRI_MIN
# define ETP_PRI_MIN 0
# define ETP_PRI_MAX 0
#endif

#ifndef ETP_TYPE_QUIT
# define ETP_TYPE_QUIT 0
#endif

#ifndef ETP_TYPE_GROUP
# define ETP_TYPE_GROUP 1
#endif

#define ETP_NUM_PRI (ETP_PRI_MAX - ETP_PRI_MIN + 1)

#define ETP_TICKS ((1000000 + 1023) >> 10)

enum {
  ETP_FLAG_GROUPADD = 0x04, /* some request was added to the group */
  ETP_FLAG_DELAYED  = 0x08, /* groiup request has been delayed */
};

/* calculate time difference in ~1/ETP_TICKS of a second */
ecb_inline int
etp_tvdiff (struct timeval *tv1, struct timeval *tv2)
{
  return  (tv2->tv_sec  - tv1->tv_sec ) * ETP_TICKS
       + ((tv2->tv_usec - tv1->tv_usec) >> 10);
}

static unsigned int started, idle, wanted = 4;

static void (*want_poll_cb) (void);
static void (*done_poll_cb) (void);
 
static unsigned int max_poll_time;     /* reslock */
static unsigned int max_poll_reqs;     /* reslock */

static unsigned int nreqs;    /* reqlock */
static unsigned int nready;   /* reqlock */
static unsigned int npending; /* reqlock */
static unsigned int max_idle = 4;      /* maximum number of threads that can idle indefinitely */
static unsigned int idle_timeout = 10; /* number of seconds after which an idle threads exit */

static xmutex_t wrklock;
static xmutex_t reslock;
static xmutex_t reqlock;
static xcond_t  reqwait;

struct etp_tmpbuf
{
  void *ptr;
  int len;
};

static void *
etp_tmpbuf_get (struct etp_tmpbuf *buf, int len)
{
  if (buf->len < len)
    {
      free (buf->ptr);
      buf->ptr = malloc (buf->len = len);
    }

  return buf->ptr;
}

/*
 * a somewhat faster data structure might be nice, but
 * with 8 priorities this actually needs <20 insns
 * per shift, the most expensive operation.
 */
typedef struct
{
  ETP_REQ *qs[ETP_NUM_PRI], *qe[ETP_NUM_PRI]; /* qstart, qend */
  int size;
} etp_reqq;

struct etp_pool
{
   etp_reqq req_queue;
   etp_reqq res_queue;
};

typedef struct etp_pool *etp_pool;

typedef struct etp_worker
{
  etp_pool pool;

  struct etp_tmpbuf tmpbuf;

  /* locked by wrklock */
  struct etp_worker *prev, *next;

  xthread_t tid;

#ifdef ETP_WORKER_COMMON
  ETP_WORKER_COMMON
#endif
} etp_worker;

static etp_worker wrk_first; /* NOT etp */

#define ETP_WORKER_LOCK(wrk)   X_LOCK   (wrklock)
#define ETP_WORKER_UNLOCK(wrk) X_UNLOCK (wrklock)

/* worker threads management */

static void
etp_worker_clear (etp_worker *wrk)
{
}

static void ecb_cold
etp_worker_free (etp_worker *wrk)
{
  free (wrk->tmpbuf.ptr);

  wrk->next->prev = wrk->prev;
  wrk->prev->next = wrk->next;

  free (wrk);
}

ETP_API_DECL unsigned int
etp_nreqs (etp_pool pool)
{
  int retval;
  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  retval = nreqs;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
  return retval;
}

ETP_API_DECL unsigned int
etp_nready (etp_pool pool)
{
  unsigned int retval;

  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  retval = nready;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);

  return retval;
}

ETP_API_DECL unsigned int
etp_npending (etp_pool pool)
{
  unsigned int retval;

  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  retval = npending;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);

  return retval;
}

ETP_API_DECL unsigned int
etp_nthreads (etp_pool pool)
{
  unsigned int retval;

  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  retval = started;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);

  return retval;
}

static void ecb_noinline ecb_cold
reqq_init (etp_reqq *q)
{
  int pri;

  for (pri = 0; pri < ETP_NUM_PRI; ++pri)
    q->qs[pri] = q->qe[pri] = 0;

  q->size = 0;
}

static int ecb_noinline
reqq_push (etp_reqq *q, ETP_REQ *req)
{
  int pri = req->pri;
  req->next = 0;

  if (q->qe[pri])
    {
      q->qe[pri]->next = req;
      q->qe[pri] = req;
    }
  else
    q->qe[pri] = q->qs[pri] = req;

  return q->size++;
}

static ETP_REQ * ecb_noinline
reqq_shift (etp_reqq *q)
{
  int pri;

  if (!q->size)
    return 0;

  --q->size;

  for (pri = ETP_NUM_PRI; pri--; )
    {
      ETP_REQ *req = q->qs[pri];

      if (req)
        {
          if (!(q->qs[pri] = (ETP_REQ *)req->next))
            q->qe[pri] = 0;

          return req;
        }
    }

  abort ();
}

ETP_API_DECL int ecb_cold
etp_init (etp_pool pool, void (*want_poll)(void), void (*done_poll)(void))
{
  X_MUTEX_CREATE (wrklock);
  X_MUTEX_CREATE (reslock);
  X_MUTEX_CREATE (reqlock);
  X_COND_CREATE  (reqwait);

  reqq_init (&pool->req_queue);
  reqq_init (&pool->res_queue);

  wrk_first.next =
  wrk_first.prev = &wrk_first;

  started  = 0;
  idle     = 0;
  nreqs    = 0;
  nready   = 0;
  npending = 0;

  want_poll_cb = want_poll;
  done_poll_cb = done_poll;

  return 0;
}

static void ecb_noinline ecb_cold
etp_proc_init (void)
{
#if HAVE_PRCTL_SET_NAME
  /* provide a more sensible "thread name" */
  char name[16 + 1];
  const int namelen = sizeof (name) - 1;
  int len;

  prctl (PR_GET_NAME, (unsigned long)name, 0, 0, 0);
  name [namelen] = 0;
  len = strlen (name);
  strcpy (name + (len <= namelen - 4 ? len : namelen - 4), "/eio");
  prctl (PR_SET_NAME, (unsigned long)name, 0, 0, 0);
#endif
}

X_THREAD_PROC (etp_proc)
{
  ETP_REQ *req;
  struct timespec ts;
  etp_worker *self = (etp_worker *)thr_arg;
  etp_pool pool = self->pool;

  etp_proc_init ();

  /* try to distribute timeouts somewhat evenly */
  ts.tv_nsec = ((unsigned long)self & 1023UL) * (1000000000UL / 1024UL);

  for (;;)
    {
      ts.tv_sec = 0;

      X_LOCK (reqlock);

      for (;;)
        {
          req = reqq_shift (&pool->req_queue);

          if (req)
            break;

          if (ts.tv_sec == 1) /* no request, but timeout detected, let's quit */
            {
              X_UNLOCK (reqlock);
              X_LOCK (wrklock);
              --started;
              X_UNLOCK (wrklock);
              goto quit;
            }

          ++idle;

          if (idle <= max_idle)
            /* we are allowed to idle, so do so without any timeout */
            X_COND_WAIT (reqwait, reqlock);
          else
            {
              /* initialise timeout once */
              if (!ts.tv_sec)
                ts.tv_sec = time (0) + idle_timeout;

              if (X_COND_TIMEDWAIT (reqwait, reqlock, ts) == ETIMEDOUT)
                ts.tv_sec = 1; /* assuming this is not a value computed above.,.. */
            }

          --idle;
        }

      --nready;

      X_UNLOCK (reqlock);
     
      if (req->type == ETP_TYPE_QUIT)
        goto quit;

      ETP_EXECUTE (self, req);

      X_LOCK (reslock);

      ++npending;

      if (!reqq_push (&pool->res_queue, req) && want_poll_cb)
        want_poll_cb ();

      etp_worker_clear (self);

      X_UNLOCK (reslock);
    }

quit:
  free (req);

  X_LOCK (wrklock);
  etp_worker_free (self);
  X_UNLOCK (wrklock);

  return 0;
}

static void ecb_cold
etp_start_thread (etp_pool pool)
{
  etp_worker *wrk = calloc (1, sizeof (etp_worker));

  /*TODO*/
  assert (("unable to allocate worker thread data", wrk));

  wrk->pool = pool;

  X_LOCK (wrklock);

  if (xthread_create (&wrk->tid, etp_proc, (void *)wrk))
    {
      wrk->prev = &wrk_first;
      wrk->next = wrk_first.next;
      wrk_first.next->prev = wrk;
      wrk_first.next = wrk;
      ++started;
    }
  else
    free (wrk);

  X_UNLOCK (wrklock);
}

static void
etp_maybe_start_thread (etp_pool pool)
{
  if (ecb_expect_true (etp_nthreads (pool) >= wanted))
    return;
  
  /* todo: maybe use idle here, but might be less exact */
  if (ecb_expect_true (0 <= (int)etp_nthreads (pool) + (int)etp_npending (pool) - (int)etp_nreqs (pool)))
    return;

  etp_start_thread (pool);
}

static void ecb_cold
etp_end_thread (etp_pool pool)
{
  ETP_REQ *req = calloc (1, sizeof (ETP_REQ)); /* will be freed by worker */

  req->type = ETP_TYPE_QUIT;
  req->pri  = ETP_PRI_MAX - ETP_PRI_MIN;

  X_LOCK (reqlock);
  reqq_push (&pool->req_queue, req);
  X_COND_SIGNAL (reqwait);
  X_UNLOCK (reqlock);

  X_LOCK (wrklock);
  --started;
  X_UNLOCK (wrklock);
}

ETP_API_DECL int
etp_poll (etp_pool pool)
{
  unsigned int maxreqs;
  unsigned int maxtime;
  struct timeval tv_start, tv_now;

  X_LOCK (reslock);
  maxreqs = max_poll_reqs;
  maxtime = max_poll_time;
  X_UNLOCK (reslock);

  if (maxtime)
    gettimeofday (&tv_start, 0);

  for (;;)
    {
      ETP_REQ *req;

      etp_maybe_start_thread (pool);

      X_LOCK (reslock);
      req = reqq_shift (&pool->res_queue);

      if (req)
        {
          --npending;

          if (!pool->res_queue.size && done_poll_cb)
            done_poll_cb ();
        }

      X_UNLOCK (reslock);

      if (!req)
        return 0;

      X_LOCK (reqlock);
      --nreqs;
      X_UNLOCK (reqlock);

      if (ecb_expect_false (req->type == ETP_TYPE_GROUP && req->size))
        {
          req->flags |= ETP_FLAG_DELAYED; /* mark request as delayed */
          continue;
        }
      else
        {
          int res = ETP_FINISH (req);
          if (ecb_expect_false (res))
            return res;
        }

      if (ecb_expect_false (maxreqs && !--maxreqs))
        break;

      if (maxtime)
        {
          gettimeofday (&tv_now, 0);

          if (etp_tvdiff (&tv_start, &tv_now) >= maxtime)
            break;
        }
    }

  errno = EAGAIN;
  return -1;
}

ETP_API_DECL void
etp_grp_cancel (etp_pool pool, ETP_REQ *grp);

ETP_API_DECL void
etp_cancel (etp_pool pool, ETP_REQ *req)
{
  req->cancelled = 1;

  etp_grp_cancel (pool, req);
}

ETP_API_DECL void
etp_grp_cancel (etp_pool pool, ETP_REQ *grp)
{
  for (grp = grp->grp_first; grp; grp = grp->grp_next)
    etp_cancel (pool, grp);
}

ETP_API_DECL void
etp_submit (etp_pool pool, ETP_REQ *req)
{
  req->pri -= ETP_PRI_MIN;

  if (ecb_expect_false (req->pri < ETP_PRI_MIN - ETP_PRI_MIN)) req->pri = ETP_PRI_MIN - ETP_PRI_MIN;
  if (ecb_expect_false (req->pri > ETP_PRI_MAX - ETP_PRI_MIN)) req->pri = ETP_PRI_MAX - ETP_PRI_MIN;

  if (ecb_expect_false (req->type == ETP_TYPE_GROUP))
    {
      /* I hope this is worth it :/ */
      X_LOCK (reqlock);
      ++nreqs;
      X_UNLOCK (reqlock);

      X_LOCK (reslock);

      ++npending;

      if (!reqq_push (&pool->res_queue, req) && want_poll_cb)
        want_poll_cb ();

      X_UNLOCK (reslock);
    }
  else
    {
      X_LOCK (reqlock);
      ++nreqs;
      ++nready;
      reqq_push (&pool->req_queue, req);
      X_COND_SIGNAL (reqwait);
      X_UNLOCK (reqlock);

      etp_maybe_start_thread (pool);
    }
}

ETP_API_DECL void ecb_cold
etp_set_max_poll_time (etp_pool pool, double nseconds)
{
  if (WORDACCESS_UNSAFE) X_LOCK   (reslock);
  max_poll_time = nseconds * ETP_TICKS;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
}

ETP_API_DECL void ecb_cold
etp_set_max_poll_reqs (etp_pool pool, unsigned int maxreqs)
{
  if (WORDACCESS_UNSAFE) X_LOCK   (reslock);
  max_poll_reqs = maxreqs;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
}

ETP_API_DECL void ecb_cold
etp_set_max_idle (etp_pool pool, unsigned int nthreads)
{
  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  max_idle = nthreads;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
}

ETP_API_DECL void ecb_cold
etp_set_idle_timeout (etp_pool pool, unsigned int seconds)
{
  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
  idle_timeout = seconds;
  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
}

ETP_API_DECL void ecb_cold
etp_set_min_parallel (etp_pool pool, unsigned int nthreads)
{
  if (wanted < nthreads)
    wanted = nthreads;
}

ETP_API_DECL void ecb_cold
etp_set_max_parallel (etp_pool pool, unsigned int nthreads)
{
  if (wanted > nthreads)
    wanted = nthreads;

  while (started > wanted)
    etp_end_thread (pool);
}

Revision:	1.5
Committed:	Thu Jun 25 17:40:24 2015 UTC (8 years, 11 months ago) by root
Content type:	text/plain
Branch:	MAIN
Changes since 1.4:	+60 -49 lines
Log Message:	* empty log message *
#	Content
1	/*
2	* libetp implementation
3	*
4	* Copyright (c) 2007,2008,2009,2010,2011,2012,2013 Marc Alexander Lehmann <libetp@schmorp.de>
5	* All rights reserved.
6	*
7	* Redistribution and use in source and binary forms, with or without modifica-
8	* tion, are permitted provided that the following conditions are met:
9	*
10	* 1. Redistributions of source code must retain the above copyright notice,
11	* this list of conditions and the following disclaimer.
12	*
13	* 2. Redistributions in binary form must reproduce the above copyright
14	* notice, this list of conditions and the following disclaimer in the
15	* documentation and/or other materials provided with the distribution.
16	*
17	* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
18	* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER-
19	* CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
20	* EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE-
21	* CIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
22	* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
23	* OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
24	* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTH-
25	* ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
26	* OF THE POSSIBILITY OF SUCH DAMAGE.
27	*
28	* Alternatively, the contents of this file may be used under the terms of
29	* the GNU General Public License ("GPL") version 2 or any later version,
30	* in which case the provisions of the GPL are applicable instead of
31	* the above. If you wish to allow the use of your version of this file
32	* only under the terms of the GPL and not to allow others to use your
33	* version of this file under the BSD license, indicate your decision
34	* by deleting the provisions above and replace them with the notice
35	* and other provisions required by the GPL. If you do not delete the
36	* provisions above, a recipient may use your version of this file under
37	* either the BSD or the GPL.
38	*/
39
40	#ifndef ETP_API_DECL
41	# define ETP_API_DECL static
42	#endif
43
44	#ifndef ETP_PRI_MIN
45	# define ETP_PRI_MIN 0
46	# define ETP_PRI_MAX 0
47	#endif
48
49	#ifndef ETP_TYPE_QUIT
50	# define ETP_TYPE_QUIT 0
51	#endif
52
53	#ifndef ETP_TYPE_GROUP
54	# define ETP_TYPE_GROUP 1
55	#endif
56
57	#define ETP_NUM_PRI (ETP_PRI_MAX - ETP_PRI_MIN + 1)
58
59	#define ETP_TICKS ((1000000 + 1023) >> 10)
60
61	enum {
62	ETP_FLAG_GROUPADD = 0x04, /* some request was added to the group */
63	ETP_FLAG_DELAYED = 0x08, /* groiup request has been delayed */
64	};
65
66	/* calculate time difference in ~1/ETP_TICKS of a second */
67	ecb_inline int
68	etp_tvdiff (struct timeval tv1, struct timeval tv2)
69	{
70	return (tv2->tv_sec - tv1->tv_sec ) * ETP_TICKS
71	+ ((tv2->tv_usec - tv1->tv_usec) >> 10);
72	}
73
74	static unsigned int started, idle, wanted = 4;
75
76	static void (*want_poll_cb) (void);
77	static void (*done_poll_cb) (void);
78
79	static unsigned int max_poll_time; /* reslock */
80	static unsigned int max_poll_reqs; /* reslock */
81
82	static unsigned int nreqs; /* reqlock */
83	static unsigned int nready; /* reqlock */
84	static unsigned int npending; /* reqlock */
85	static unsigned int max_idle = 4; /* maximum number of threads that can idle indefinitely */
86	static unsigned int idle_timeout = 10; /* number of seconds after which an idle threads exit */
87
88	static xmutex_t wrklock;
89	static xmutex_t reslock;
90	static xmutex_t reqlock;
91	static xcond_t reqwait;
92
93	struct etp_tmpbuf
94	{
95	void *ptr;
96	int len;
97	};
98
99	static void *
100	etp_tmpbuf_get (struct etp_tmpbuf *buf, int len)
101	{
102	if (buf->len < len)
103	{
104	free (buf->ptr);
105	buf->ptr = malloc (buf->len = len);
106	}
107
108	return buf->ptr;
109	}
110
111	/*
112	* a somewhat faster data structure might be nice, but
113	* with 8 priorities this actually needs <20 insns
114	* per shift, the most expensive operation.
115	*/
116	typedef struct
117	{
118	ETP_REQ qs[ETP_NUM_PRI], qe[ETP_NUM_PRI]; /* qstart, qend */
119	int size;
120	} etp_reqq;
121
122	struct etp_pool
123	{
124	etp_reqq req_queue;
125	etp_reqq res_queue;
126	};
127
128	typedef struct etp_pool *etp_pool;
129
130	typedef struct etp_worker
131	{
132	etp_pool pool;
133
134	struct etp_tmpbuf tmpbuf;
135
136	/* locked by wrklock */
137	struct etp_worker prev, next;
138
139	xthread_t tid;
140
141	#ifdef ETP_WORKER_COMMON
142	ETP_WORKER_COMMON
143	#endif
144	} etp_worker;
145
146	static etp_worker wrk_first; /* NOT etp */
147
148	#define ETP_WORKER_LOCK(wrk) X_LOCK (wrklock)
149	#define ETP_WORKER_UNLOCK(wrk) X_UNLOCK (wrklock)
150
151	/* worker threads management */
152
153	static void
154	etp_worker_clear (etp_worker *wrk)
155	{
156	}
157
158	static void ecb_cold
159	etp_worker_free (etp_worker *wrk)
160	{
161	free (wrk->tmpbuf.ptr);
162
163	wrk->next->prev = wrk->prev;
164	wrk->prev->next = wrk->next;
165
166	free (wrk);
167	}
168
169	ETP_API_DECL unsigned int
170	etp_nreqs (etp_pool pool)
171	{
172	int retval;
173	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
174	retval = nreqs;
175	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
176	return retval;
177	}
178
179	ETP_API_DECL unsigned int
180	etp_nready (etp_pool pool)
181	{
182	unsigned int retval;
183
184	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
185	retval = nready;
186	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
187
188	return retval;
189	}
190
191	ETP_API_DECL unsigned int
192	etp_npending (etp_pool pool)
193	{
194	unsigned int retval;
195
196	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
197	retval = npending;
198	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
199
200	return retval;
201	}
202
203	ETP_API_DECL unsigned int
204	etp_nthreads (etp_pool pool)
205	{
206	unsigned int retval;
207
208	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
209	retval = started;
210	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
211
212	return retval;
213	}
214
215	static void ecb_noinline ecb_cold
216	reqq_init (etp_reqq *q)
217	{
218	int pri;
219
220	for (pri = 0; pri < ETP_NUM_PRI; ++pri)
221	q->qs[pri] = q->qe[pri] = 0;
222
223	q->size = 0;
224	}
225
226	static int ecb_noinline
227	reqq_push (etp_reqq q, ETP_REQ req)
228	{
229	int pri = req->pri;
230	req->next = 0;
231
232	if (q->qe[pri])
233	{
234	q->qe[pri]->next = req;
235	q->qe[pri] = req;
236	}
237	else
238	q->qe[pri] = q->qs[pri] = req;
239
240	return q->size++;
241	}
242
243	static ETP_REQ * ecb_noinline
244	reqq_shift (etp_reqq *q)
245	{
246	int pri;
247
248	if (!q->size)
249	return 0;
250
251	--q->size;
252
253	for (pri = ETP_NUM_PRI; pri--; )
254	{
255	ETP_REQ *req = q->qs[pri];
256
257	if (req)
258	{
259	if (!(q->qs[pri] = (ETP_REQ *)req->next))
260	q->qe[pri] = 0;
261
262	return req;
263	}
264	}
265
266	abort ();
267	}
268
269	ETP_API_DECL int ecb_cold
270	etp_init (etp_pool pool, void (want_poll)(void), void (done_poll)(void))
271	{
272	X_MUTEX_CREATE (wrklock);
273	X_MUTEX_CREATE (reslock);
274	X_MUTEX_CREATE (reqlock);
275	X_COND_CREATE (reqwait);
276
277	reqq_init (&pool->req_queue);
278	reqq_init (&pool->res_queue);
279
280	wrk_first.next =
281	wrk_first.prev = &wrk_first;
282
283	started = 0;
284	idle = 0;
285	nreqs = 0;
286	nready = 0;
287	npending = 0;
288
289	want_poll_cb = want_poll;
290	done_poll_cb = done_poll;
291
292	return 0;
293	}
294
295	static void ecb_noinline ecb_cold
296	etp_proc_init (void)
297	{
298	#if HAVE_PRCTL_SET_NAME
299	/* provide a more sensible "thread name" */
300	char name[16 + 1];
301	const int namelen = sizeof (name) - 1;
302	int len;
303
304	prctl (PR_GET_NAME, (unsigned long)name, 0, 0, 0);
305	name [namelen] = 0;
306	len = strlen (name);
307	strcpy (name + (len <= namelen - 4 ? len : namelen - 4), "/eio");
308	prctl (PR_SET_NAME, (unsigned long)name, 0, 0, 0);
309	#endif
310	}
311
312	X_THREAD_PROC (etp_proc)
313	{
314	ETP_REQ *req;
315	struct timespec ts;
316	etp_worker self = (etp_worker )thr_arg;
317	etp_pool pool = self->pool;
318
319	etp_proc_init ();
320
321	/* try to distribute timeouts somewhat evenly */
322	ts.tv_nsec = ((unsigned long)self & 1023UL) * (1000000000UL / 1024UL);
323
324	for (;;)
325	{
326	ts.tv_sec = 0;
327
328	X_LOCK (reqlock);
329
330	for (;;)
331	{
332	req = reqq_shift (&pool->req_queue);
333
334	if (req)
335	break;
336
337	if (ts.tv_sec == 1) /* no request, but timeout detected, let's quit */
338	{
339	X_UNLOCK (reqlock);
340	X_LOCK (wrklock);
341	--started;
342	X_UNLOCK (wrklock);
343	goto quit;
344	}
345
346	++idle;
347
348	if (idle <= max_idle)
349	/* we are allowed to idle, so do so without any timeout */
350	X_COND_WAIT (reqwait, reqlock);
351	else
352	{
353	/* initialise timeout once */
354	if (!ts.tv_sec)
355	ts.tv_sec = time (0) + idle_timeout;
356
357	if (X_COND_TIMEDWAIT (reqwait, reqlock, ts) == ETIMEDOUT)
358	ts.tv_sec = 1; /* assuming this is not a value computed above.,.. */
359	}
360
361	--idle;
362	}
363
364	--nready;
365
366	X_UNLOCK (reqlock);
367
368	if (req->type == ETP_TYPE_QUIT)
369	goto quit;
370
371	ETP_EXECUTE (self, req);
372
373	X_LOCK (reslock);
374
375	++npending;
376
377	if (!reqq_push (&pool->res_queue, req) && want_poll_cb)
378	want_poll_cb ();
379
380	etp_worker_clear (self);
381
382	X_UNLOCK (reslock);
383	}
384
385	quit:
386	free (req);
387
388	X_LOCK (wrklock);
389	etp_worker_free (self);
390	X_UNLOCK (wrklock);
391
392	return 0;
393	}
394
395	static void ecb_cold
396	etp_start_thread (etp_pool pool)
397	{
398	etp_worker *wrk = calloc (1, sizeof (etp_worker));
399
400	/TODO/
401	assert (("unable to allocate worker thread data", wrk));
402
403	wrk->pool = pool;
404
405	X_LOCK (wrklock);
406
407	if (xthread_create (&wrk->tid, etp_proc, (void *)wrk))
408	{
409	wrk->prev = &wrk_first;
410	wrk->next = wrk_first.next;
411	wrk_first.next->prev = wrk;
412	wrk_first.next = wrk;
413	++started;
414	}
415	else
416	free (wrk);
417
418	X_UNLOCK (wrklock);
419	}
420
421	static void
422	etp_maybe_start_thread (etp_pool pool)
423	{
424	if (ecb_expect_true (etp_nthreads (pool) >= wanted))
425	return;
426
427	/* todo: maybe use idle here, but might be less exact */
428	if (ecb_expect_true (0 <= (int)etp_nthreads (pool) + (int)etp_npending (pool) - (int)etp_nreqs (pool)))
429	return;
430
431	etp_start_thread (pool);
432	}
433
434	static void ecb_cold
435	etp_end_thread (etp_pool pool)
436	{
437	ETP_REQ req = calloc (1, sizeof (ETP_REQ)); / will be freed by worker */
438
439	req->type = ETP_TYPE_QUIT;
440	req->pri = ETP_PRI_MAX - ETP_PRI_MIN;
441
442	X_LOCK (reqlock);
443	reqq_push (&pool->req_queue, req);
444	X_COND_SIGNAL (reqwait);
445	X_UNLOCK (reqlock);
446
447	X_LOCK (wrklock);
448	--started;
449	X_UNLOCK (wrklock);
450	}
451
452	ETP_API_DECL int
453	etp_poll (etp_pool pool)
454	{
455	unsigned int maxreqs;
456	unsigned int maxtime;
457	struct timeval tv_start, tv_now;
458
459	X_LOCK (reslock);
460	maxreqs = max_poll_reqs;
461	maxtime = max_poll_time;
462	X_UNLOCK (reslock);
463
464	if (maxtime)
465	gettimeofday (&tv_start, 0);
466
467	for (;;)
468	{
469	ETP_REQ *req;
470
471	etp_maybe_start_thread (pool);
472
473	X_LOCK (reslock);
474	req = reqq_shift (&pool->res_queue);
475
476	if (req)
477	{
478	--npending;
479
480	if (!pool->res_queue.size && done_poll_cb)
481	done_poll_cb ();
482	}
483
484	X_UNLOCK (reslock);
485
486	if (!req)
487	return 0;
488
489	X_LOCK (reqlock);
490	--nreqs;
491	X_UNLOCK (reqlock);
492
493	if (ecb_expect_false (req->type == ETP_TYPE_GROUP && req->size))
494	{
495	req->flags \|= ETP_FLAG_DELAYED; /* mark request as delayed */
496	continue;
497	}
498	else
499	{
500	int res = ETP_FINISH (req);
501	if (ecb_expect_false (res))
502	return res;
503	}
504
505	if (ecb_expect_false (maxreqs && !--maxreqs))
506	break;
507
508	if (maxtime)
509	{
510	gettimeofday (&tv_now, 0);
511
512	if (etp_tvdiff (&tv_start, &tv_now) >= maxtime)
513	break;
514	}
515	}
516
517	errno = EAGAIN;
518	return -1;
519	}
520
521	ETP_API_DECL void
522	etp_grp_cancel (etp_pool pool, ETP_REQ *grp);
523
524	ETP_API_DECL void
525	etp_cancel (etp_pool pool, ETP_REQ *req)
526	{
527	req->cancelled = 1;
528
529	etp_grp_cancel (pool, req);
530	}
531
532	ETP_API_DECL void
533	etp_grp_cancel (etp_pool pool, ETP_REQ *grp)
534	{
535	for (grp = grp->grp_first; grp; grp = grp->grp_next)
536	etp_cancel (pool, grp);
537	}
538
539	ETP_API_DECL void
540	etp_submit (etp_pool pool, ETP_REQ *req)
541	{
542	req->pri -= ETP_PRI_MIN;
543
544	if (ecb_expect_false (req->pri < ETP_PRI_MIN - ETP_PRI_MIN)) req->pri = ETP_PRI_MIN - ETP_PRI_MIN;
545	if (ecb_expect_false (req->pri > ETP_PRI_MAX - ETP_PRI_MIN)) req->pri = ETP_PRI_MAX - ETP_PRI_MIN;
546
547	if (ecb_expect_false (req->type == ETP_TYPE_GROUP))
548	{
549	/* I hope this is worth it :/ */
550	X_LOCK (reqlock);
551	++nreqs;
552	X_UNLOCK (reqlock);
553
554	X_LOCK (reslock);
555
556	++npending;
557
558	if (!reqq_push (&pool->res_queue, req) && want_poll_cb)
559	want_poll_cb ();
560
561	X_UNLOCK (reslock);
562	}
563	else
564	{
565	X_LOCK (reqlock);
566	++nreqs;
567	++nready;
568	reqq_push (&pool->req_queue, req);
569	X_COND_SIGNAL (reqwait);
570	X_UNLOCK (reqlock);
571
572	etp_maybe_start_thread (pool);
573	}
574	}
575
576	ETP_API_DECL void ecb_cold
577	etp_set_max_poll_time (etp_pool pool, double nseconds)
578	{
579	if (WORDACCESS_UNSAFE) X_LOCK (reslock);
580	max_poll_time = nseconds * ETP_TICKS;
581	if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
582	}
583
584	ETP_API_DECL void ecb_cold
585	etp_set_max_poll_reqs (etp_pool pool, unsigned int maxreqs)
586	{
587	if (WORDACCESS_UNSAFE) X_LOCK (reslock);
588	max_poll_reqs = maxreqs;
589	if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
590	}
591
592	ETP_API_DECL void ecb_cold
593	etp_set_max_idle (etp_pool pool, unsigned int nthreads)
594	{
595	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
596	max_idle = nthreads;
597	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
598	}
599
600	ETP_API_DECL void ecb_cold
601	etp_set_idle_timeout (etp_pool pool, unsigned int seconds)
602	{
603	if (WORDACCESS_UNSAFE) X_LOCK (reqlock);
604	idle_timeout = seconds;
605	if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
606	}
607
608	ETP_API_DECL void ecb_cold
609	etp_set_min_parallel (etp_pool pool, unsigned int nthreads)
610	{
611	if (wanted < nthreads)
612	wanted = nthreads;
613	}
614
615	ETP_API_DECL void ecb_cold
616	etp_set_max_parallel (etp_pool pool, unsigned int nthreads)
617	{
618	if (wanted > nthreads)
619	wanted = nthreads;
620
621	while (started > wanted)
622	etp_end_thread (pool);
623	}
624