libuv/src/threadpool.c

/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this software and associated documentation files (the "Software"), to
 * deal in the Software without restriction, including without limitation the
 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
 * sell copies of the Software, and to permit persons to whom the Software is
 * furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

#include "uv-common.h"
#include "uv_log.h"
#include "uv_trace.h"

#if !defined(_WIN32)
# include "unix/internal.h"
#else
# include "win/internal.h"
#endif

#include <stdlib.h>
#ifdef USE_FFRT
#include <assert.h>
#include "ffrt_inner.h"
#endif
#include <stdio.h>
#ifdef ASYNC_STACKTRACE
#include "dfx/async_stack/libuv_async_stack.h"
#endif

#define MAX_THREADPOOL_SIZE 1024
#define UV_TRACE_NAME "UV_TRACE"

static uv_rwlock_t g_closed_uv_loop_rwlock;
static uv_once_t once = UV_ONCE_INIT;
static uv_cond_t cond;
static uv_mutex_t mutex;
static unsigned int idle_threads;
static unsigned int nthreads;
static uv_thread_t* threads;
static uv_thread_t default_threads[4];
static struct uv__queue exit_message;
static struct uv__queue wq;
static struct uv__queue run_slow_work_message;
static struct uv__queue slow_io_pending_wq;


#ifdef UV_STATISTIC
#define MAX_DUMP_QUEUE_SIZE 200
static uv_mutex_t dump_queue_mutex;
static QUEUE dump_queue;
static unsigned int dump_queue_size;

static int statistic_idle;
static uv_mutex_t statistic_mutex;
static QUEUE statistic_works;
static uv_cond_t dump_cond;
static uv_thread_t dump_thread;

static void uv_dump_worker(void*  arg) {
  struct uv__statistic_work* w;
  struct uv__queue* q;
  uv_sem_post((uv_sem_t*) arg);
  arg = NULL;
  uv_mutex_lock(&statistic_mutex);
  for (;;) {
    while (uv__queue_empty(&statistic_works)) {
      statistic_idle = 1;
      uv_cond_wait(&dump_cond, &statistic_mutex);
      statistic_idle = 0;
    }
    q = uv__queue_head(&statistic_works);
    if (q == &exit_message) {
      uv_cond_signal(&dump_cond);
      uv_mutex_unlock(&statistic_mutex);
      break;
    }
    uv__queue_remove(q);
    uv__queue_init(q);
    uv_mutex_unlock(&statistic_mutex);
    w = uv__queue_data(q, struct uv__statistic_work, wq);
    w->work(w);
    free(w);
    uv_mutex_lock(&statistic_mutex);
  }
}

static void post_statistic_work(struct uv__queue* q) {
  uv_mutex_lock(&statistic_mutex);
  uv__queue_insert_tail(&statistic_works, q);
  if (statistic_idle)
    uv_cond_signal(&dump_cond);
  uv_mutex_unlock(&statistic_mutex);
}

static void uv__queue_work_info(struct uv__statistic_work *work) {
  uv_mutex_lock(&dump_queue_mutex);
  if (dump_queue_size + 1 > MAX_DUMP_QUEUE_SIZE) { /* release works already done */
    struct uv__queue* q;
    uv__queue_foreach(q, &dump_queue) {
      struct uv_work_dump_info* info = uv__queue_data(q, struct uv_work_dump_info, wq);
      if (info->state == DONE_END) {
        uv__queue_remove(q);
        free(info);
        dump_queue_size--;
      }
    }
    if (dump_queue_size + 1 > MAX_DUMP_QUEUE_SIZE) {
      abort(); /* too many works not done. */
    }
  }

  uv__queue_insert_head(&dump_queue,  &work->info->wq);
  dump_queue_size++;
  uv_mutex_unlock(&dump_queue_mutex);
}

static void uv__update_work_info(struct uv__statistic_work *work) {
  uv_mutex_lock(&dump_queue_mutex);
  if (work != NULL && work->info != NULL) {
    work->info->state = work->state;
    switch (work->state) {
      case WAITING:
        work->info->queue_time = work->time;
        break;
      case WORK_EXECUTING:
        work->info->execute_start_time = work->time;
        break;
      case WORK_END:
        work->info->execute_end_time = work->time;
        break;
      case DONE_EXECUTING:
        work->info->done_start_time = work->time;
        break;
      case DONE_END:
        work->info->done_end_time = work->time;
        break;
      default:
        break;
    }
  }
  uv_mutex_unlock(&dump_queue_mutex);
}


// return the timestamp in millisecond
static uint64_t uv__now_timestamp() {
  uv_timeval64_t tv;
  int r = uv_gettimeofday(&tv);
  if (r != 0) {
    return 0;
  }
  return (uint64_t)tv.tv_sec * 1000 + tv.tv_usec / 1000;
}


static void uv__post_statistic_work(struct uv__work *w, enum uv_work_state state) {
  struct uv__statistic_work* dump_work = (struct uv__statistic_work*)malloc(sizeof(struct uv__statistic_work));
  if (dump_work == NULL) {
    return;
  }
  dump_work->info = w->info;
  dump_work->work = uv__update_work_info;
  dump_work->time = uv__now_timestamp();
  dump_work->state = state;
  uv__queue_init(&dump_work->wq);
  post_statistic_work(&dump_work->wq);
}


static void init_work_dump_queue()
{
  if (uv_mutex_init(&dump_queue_mutex))
    abort();
  uv_mutex_lock(&dump_queue_mutex);
  uv__queue_init(&dump_queue);
  dump_queue_size = 0;
  uv_mutex_unlock(&dump_queue_mutex);

  /* init dump thread */
  statistic_idle = 1;
  if (uv_mutex_init(&statistic_mutex))
    abort();
  uv__queue_init(&statistic_works);
  uv_sem_t sem;
  if (uv_cond_init(&dump_cond))
    abort();
  if (uv_sem_init(&sem, 0))
    abort();
  if (uv_thread_create(&dump_thread, uv_dump_worker, &sem))
      abort();
  uv_sem_wait(&sem);
  uv_sem_destroy(&sem);
}


void uv_init_dump_info(struct uv_work_dump_info* info, struct uv__work* w) {
  if (info == NULL)
    return;
  info->queue_time = 0;
  info->state = WAITING;
  info->execute_start_time = 0;
  info->execute_end_time = 0;
  info->done_start_time = 0;
  info->done_end_time = 0;
  info->work = w;
  uv__queue_init(&info->wq);
}


void uv_queue_statics(struct uv_work_dump_info* info) {
  struct uv__statistic_work* dump_work = (struct uv__statistic_work*)malloc(sizeof(struct uv__statistic_work));
  if (dump_work == NULL) {
    abort();
  }
  dump_work->info = info;
  dump_work->work = uv__queue_work_info;
  info->queue_time = uv__now_timestamp();
  dump_work->state = WAITING;
  uv__queue_init(&dump_work->wq);
  post_statistic_work(&dump_work->wq);
}


uv_worker_info_t* uv_dump_work_queue(int* size) {
#ifdef UV_STATISTIC
  uv_mutex_lock(&dump_queue_mutex);
  if (uv__queue_empty(&dump_queue)) {
    return NULL;
  }
  *size = dump_queue_size;
  uv_worker_info_t* dump_info = (uv_worker_info_t*) malloc(sizeof(uv_worker_info_t) * dump_queue_size);
  struct uv__queue* q;
  int i = 0;
  uv__queue_foreach(q, &dump_queue) {
    struct uv_work_dump_info* info = uv__queue_data(q, struct uv_work_dump_info, wq);
    dump_info[i].queue_time = info->queue_time;
    dump_info[i].builtin_return_address[0] = info->builtin_return_address[0];
    dump_info[i].builtin_return_address[1] = info->builtin_return_address[1];
    dump_info[i].builtin_return_address[2] = info->builtin_return_address[2];
    switch (info->state) {
      case WAITING:
        strcpy(dump_info[i].state, "waiting");
        break;
      case WORK_EXECUTING:
        strcpy(dump_info[i].state, "work_executing");
        break;
      case WORK_END:
        strcpy(dump_info[i].state, "work_end");
        break;
      case DONE_EXECUTING:
        strcpy(dump_info[i].state, "done_executing");
        break;
      case DONE_END:
        strcpy(dump_info[i].state, "done_end");
        break;
      default:
        break;
    }
    dump_info[i].execute_start_time = info->execute_start_time;
    dump_info[i].execute_end_time = info->execute_end_time;
    dump_info[i].done_start_time = info->done_start_time;
    dump_info[i].done_end_time = info->done_end_time;
    ++i;
  }
  uv_mutex_unlock(&dump_queue_mutex);
  return dump_info;
#else
  size = 0;
  return NULL;
#endif
}
#endif


static void init_closed_uv_loop_rwlock_once(void) {
  uv_rwlock_init(&g_closed_uv_loop_rwlock);
}


void rdlock_closed_uv_loop_rwlock(void) {
  uv_rwlock_rdlock(&g_closed_uv_loop_rwlock);
}


void rdunlock_closed_uv_loop_rwlock(void) {
  uv_rwlock_rdunlock(&g_closed_uv_loop_rwlock);
}


int is_uv_loop_good_magic(const uv_loop_t* loop) {
  if (loop->magic == UV_LOOP_MAGIC) {
    return 1;
  }
  UV_LOGE("loop:(%{public}zu:%{public}#x) invalid", (size_t)loop, loop->magic);
  return 0;
}


void on_uv_loop_close(uv_loop_t* loop) {
  time_t t1, t2;
  time(&t1);
  uv_start_trace(UV_TRACE_TAG, "Get Write Lock");
  uv_rwlock_wrlock(&g_closed_uv_loop_rwlock);
  uv_end_trace(UV_TRACE_TAG);
  loop->magic = ~UV_LOOP_MAGIC;
  uv_start_trace(UV_TRACE_TAG, "Release Write Lock");
  uv_rwlock_wrunlock(&g_closed_uv_loop_rwlock);
  uv_end_trace(UV_TRACE_TAG);
  time(&t2);
  UV_LOGI("loop:(%{public}zu) closed in %{public}zds", (size_t)loop, (ssize_t)(t2 - t1));
}


static void uv__cancelled(struct uv__work* w) {
  abort();
}


#ifndef USE_FFRT
static unsigned int slow_io_work_running;

static unsigned int slow_work_thread_threshold(void) {
  return (nthreads + 1) / 2;
}


/* To avoid deadlock with uv_cancel() it's crucial that the worker
 * never holds the global mutex and the loop-local mutex at the same time.
 */
static void worker(void* arg) {
  struct uv__work* w;
  struct uv__queue* q;
  int is_slow_work;

  uv_sem_post((uv_sem_t*) arg);
  arg = NULL;

  uv_mutex_lock(&mutex);
  for (;;) {
    /* `mutex` should always be locked at this point. */

    /* Keep waiting while either no work is present or only slow I/O
       and we're at the threshold for that. */
    while (uv__queue_empty(&wq) ||
           (uv__queue_head(&wq) == &run_slow_work_message &&
            uv__queue_next(&run_slow_work_message) == &wq &&
            slow_io_work_running >= slow_work_thread_threshold())) {
      idle_threads += 1;
      uv_cond_wait(&cond, &mutex);
      idle_threads -= 1;
    }

    q = uv__queue_head(&wq);
    if (q == &exit_message) {
      uv_cond_signal(&cond);
      uv_mutex_unlock(&mutex);
      break;
    }

    uv__queue_remove(q);
    uv__queue_init(q);  /* Signal uv_cancel() that the work req is executing. */

    is_slow_work = 0;
    if (q == &run_slow_work_message) {
      /* If we're at the slow I/O threshold, re-schedule until after all
         other work in the queue is done. */
      if (slow_io_work_running >= slow_work_thread_threshold()) {
        uv__queue_insert_tail(&wq, q);
        continue;
      }

      /* If we encountered a request to run slow I/O work but there is none
         to run, that means it's cancelled => Start over. */
      if (uv__queue_empty(&slow_io_pending_wq))
        continue;

      is_slow_work = 1;
      slow_io_work_running++;

      q = uv__queue_head(&slow_io_pending_wq);
      uv__queue_remove(q);
      uv__queue_init(q);

      /* If there is more slow I/O work, schedule it to be run as well. */
      if (!uv__queue_empty(&slow_io_pending_wq)) {
        uv__queue_insert_tail(&wq, &run_slow_work_message);
        if (idle_threads > 0)
          uv_cond_signal(&cond);
      }
    }

    uv_mutex_unlock(&mutex);

    w = uv__queue_data(q, struct uv__work, wq);
#ifdef UV_STATISTIC
    uv__post_statistic_work(w, WORK_EXECUTING);
#endif
#ifdef ASYNC_STACKTRACE
    uv_work_t* req = container_of(w, uv_work_t, work_req);
    LibuvSetStackId((uint64_t)req->reserved[3]);
#endif
    w->work(w);
#ifdef UV_STATISTIC
    uv__post_statistic_work(w, WORK_END);
#endif
    uv_mutex_lock(&w->loop->wq_mutex);
    w->work = NULL;  /* Signal uv_cancel() that the work req is done
                        executing. */
    uv__queue_insert_tail(&w->loop->wq, &w->wq);
    uv_async_send(&w->loop->wq_async);
    uv_mutex_unlock(&w->loop->wq_mutex);

    /* Lock `mutex` since that is expected at the start of the next
     * iteration. */
    uv_mutex_lock(&mutex);
    if (is_slow_work) {
      /* `slow_io_work_running` is protected by `mutex`. */
      slow_io_work_running--;
    }
  }
}
#endif


static void post(struct uv__queue* q, enum uv__work_kind kind) {
  uv_mutex_lock(&mutex);
  if (kind == UV__WORK_SLOW_IO) {
    /* Insert into a separate queue. */
    uv__queue_insert_tail(&slow_io_pending_wq, q);
    if (!uv__queue_empty(&run_slow_work_message)) {
      /* Running slow I/O tasks is already scheduled => Nothing to do here.
         The worker that runs said other task will schedule this one as well. */
      uv_mutex_unlock(&mutex);
      return;
    }
    q = &run_slow_work_message;
  }

  uv__queue_insert_tail(&wq, q);
  if (idle_threads > 0)
    uv_cond_signal(&cond);
  uv_mutex_unlock(&mutex);
}


#ifdef __MVS__
/* TODO(itodorov) - zos: revisit when Woz compiler is available. */
__attribute__((destructor))
#endif
void uv__threadpool_cleanup(void) {
  unsigned int i;

  if (nthreads == 0)
    return;

#ifndef __MVS__
  /* TODO(gabylb) - zos: revisit when Woz compiler is available. */
  post(&exit_message, UV__WORK_CPU);
#endif

  for (i = 0; i < nthreads; i++)
    if (uv_thread_join(threads + i))
      abort();

  if (threads != default_threads)
    uv__free(threads);

  uv_mutex_destroy(&mutex);
  uv_cond_destroy(&cond);

  threads = NULL;
  nthreads = 0;
#ifdef UV_STATISTIC
  post_statistic_work(&exit_message);
  uv_thread_join(dump_thread);
  uv_mutex_destroy(&statistic_mutex);
  uv_cond_destroy(&dump_cond);
#endif
}


#ifndef USE_FFRT
static void init_threads(void) {
  uv_thread_options_t config;
  unsigned int i;
  const char* val;
  uv_sem_t sem;

  nthreads = ARRAY_SIZE(default_threads);
  val = getenv("UV_THREADPOOL_SIZE");
  if (val != NULL)
    nthreads = atoi(val);
  if (nthreads == 0)
    nthreads = 1;
  if (nthreads > MAX_THREADPOOL_SIZE)
    nthreads = MAX_THREADPOOL_SIZE;

  threads = default_threads;
  if (nthreads > ARRAY_SIZE(default_threads)) {
    threads = uv__malloc(nthreads * sizeof(threads[0]));
    if (threads == NULL) {
      nthreads = ARRAY_SIZE(default_threads);
      threads = default_threads;
    }
  }

  if (uv_cond_init(&cond))
    abort();

  if (uv_mutex_init(&mutex))
    abort();

  uv__queue_init(&wq);
  uv__queue_init(&slow_io_pending_wq);
  uv__queue_init(&run_slow_work_message);

  if (uv_sem_init(&sem, 0))
    abort();

  config.flags = UV_THREAD_HAS_STACK_SIZE;
  config.stack_size = 8u << 20;  /* 8 MB */

  for (i = 0; i < nthreads; i++)
    if (uv_thread_create_ex(threads + i, &config, worker, &sem))
      abort();

  for (i = 0; i < nthreads; i++)
    uv_sem_wait(&sem);

  uv_sem_destroy(&sem);
}


#ifndef _WIN32
static void reset_once(void) {
  uv_once_t child_once = UV_ONCE_INIT;
  memcpy(&once, &child_once, sizeof(child_once));
}
#endif


static void init_once(void) {
#ifndef _WIN32
  /* Re-initialize the threadpool after fork.
   * Note that this discards the global mutex and condition as well
   * as the work queue.
   */
  if (pthread_atfork(NULL, NULL, &reset_once))
    abort();
#endif
  init_closed_uv_loop_rwlock_once();
#ifdef UV_STATISTIC
  init_work_dump_queue();
#endif
  init_threads();
}


void uv__work_submit(uv_loop_t* loop,
                     struct uv__work* w,
                     enum uv__work_kind kind,
                     void (*work)(struct uv__work* w),
                     void (*done)(struct uv__work* w, int status)) {
  uv_once(&once, init_once);
  w->loop = loop;
  w->work = work;
  w->done = done;
  post(&w->wq, kind);
}
#endif


#ifdef USE_FFRT
static void uv__task_done_wrapper(void* work, int status) {
  struct uv__work* w = (struct uv__work*)work;
  w->done(w, status);
}
#endif


/* TODO(bnoordhuis) teach libuv how to cancel file operations
 * that go through io_uring instead of the thread pool.
 */
static int uv__work_cancel(uv_loop_t* loop, uv_req_t* req, struct uv__work* w) {
  int cancelled;

  rdlock_closed_uv_loop_rwlock();
  if (!is_uv_loop_good_magic(w->loop)) {
    rdunlock_closed_uv_loop_rwlock();
    return 0;
  }

#ifndef USE_FFRT
  uv_mutex_lock(&mutex);
  uv_mutex_lock(&w->loop->wq_mutex);

  cancelled = !uv__queue_empty(&w->wq) && w->work != NULL;
  if (cancelled)
    uv__queue_remove(&w->wq);

  uv_mutex_unlock(&w->loop->wq_mutex);
  uv_mutex_unlock(&mutex);
#else
  uv_mutex_lock(&w->loop->wq_mutex);
  cancelled = !uv__queue_empty(&w->wq) && w->work != NULL
    && ffrt_executor_task_cancel(w, (ffrt_qos_t)(intptr_t)req->reserved[0]);
  uv_mutex_unlock(&w->loop->wq_mutex);
#endif

  if (!cancelled) {
    rdunlock_closed_uv_loop_rwlock();
    return UV_EBUSY;
  }

  w->work = uv__cancelled;
  uv_mutex_lock(&loop->wq_mutex);
#ifndef USE_FFRT
  uv__queue_insert_tail(&loop->wq, &w->wq);
  uv_async_send(&loop->wq_async);
#else
  uv__loop_internal_fields_t* lfields = uv__get_internal_fields(w->loop);
  int qos = (ffrt_qos_t)(intptr_t)req->reserved[0];

  if (uv_check_data_valid((struct uv_loop_data*)(w->loop->data)) == 0) {
    int status = (w->work == uv__cancelled) ? UV_ECANCELED : 0;
    struct uv_loop_data* addr = (struct uv_loop_data*)((uint64_t)w->loop->data -
      (UV_EVENT_MAGIC_OFFSET << UV_EVENT_MAGIC_OFFSETBITS));
    addr->post_task_func(addr->event_handler, uv__task_done_wrapper, (void*)w, status, qos);
  } else {
    uv__queue_insert_tail(&(lfields->wq_sub[qos]), &w->wq);
    uv_async_send(&loop->wq_async);
  }
#endif
  uv_mutex_unlock(&loop->wq_mutex);
  rdunlock_closed_uv_loop_rwlock();

  return 0;
}


void uv__work_done(uv_async_t* handle) {
  struct uv__work* w;
  uv_loop_t* loop;
  struct uv__queue* q;
  struct uv__queue wq;
  int err;
  int nevents;

  loop = container_of(handle, uv_loop_t, wq_async);
  rdlock_closed_uv_loop_rwlock();
  if (!is_uv_loop_good_magic(loop)) {
    rdunlock_closed_uv_loop_rwlock();
    return;
  }
  rdunlock_closed_uv_loop_rwlock();

  uv_mutex_lock(&loop->wq_mutex);
#ifndef USE_FFRT
  uv__queue_move(&loop->wq, &wq);
#else
  uv__loop_internal_fields_t* lfields = uv__get_internal_fields(loop);
  int i;
  uv__queue_init(&wq);
  for (i = 3; i >= 0; i--) {
    if (!uv__queue_empty(&lfields->wq_sub[i])) {
      uv__queue_append(&lfields->wq_sub[i], &wq);
    }
  }
#endif
  uv_mutex_unlock(&loop->wq_mutex);

  nevents = 0;
  uv_start_trace(UV_TRACE_TAG, UV_TRACE_NAME);
  while (!uv__queue_empty(&wq)) {
    q = uv__queue_head(&wq);
    uv__queue_remove(q);

    w = container_of(q, struct uv__work, wq);
    err = (w->work == uv__cancelled) ? UV_ECANCELED : 0;
#ifdef UV_STATISTIC
    uv__post_statistic_work(w, DONE_EXECUTING);
    struct uv__statistic_work* dump_work = (struct uv__statistic_work*)malloc(sizeof(struct uv__statistic_work));
    if (dump_work == NULL) {
      UV_LOGE("malloc(%{public}zu) failed: %{public}s", sizeof(struct uv__statistic_work), strerror(errno));
      break;
    }
    dump_work->info = w->info;
    dump_work->work = uv__update_work_info;
#endif
#ifdef ASYNC_STACKTRACE
    uv_work_t* req = container_of(w, uv_work_t, work_req);
    LibuvSetStackId((uint64_t)req->reserved[3]);
#endif
    w->done(w, err);
    nevents++;
#ifdef UV_STATISTIC
    dump_work->time = uv__now_timestamp();
    dump_work->state = DONE_END;
    QUEUE_INIT(&dump_work->wq);
    post_statistic_work(&dump_work->wq);
#endif
  }
  uv_end_trace(UV_TRACE_TAG);

  /* This check accomplishes 2 things:
   * 1. Even if the queue was empty, the call to uv__work_done() should count
   *    as an event. Which will have been added by the event loop when
   *    calling this callback.
   * 2. Prevents accidental wrap around in case nevents == 0 events == 0.
   */
  if (nevents > 1) {
    /* Subtract 1 to counter the call to uv__work_done(). */
    uv__metrics_inc_events(loop, nevents - 1);
    if (uv__get_internal_fields(loop)->current_timeout == 0)
      uv__metrics_inc_events_waiting(loop, nevents - 1);
  }
}


static void uv__queue_work(struct uv__work* w) {
  uv_work_t* req = container_of(w, uv_work_t, work_req);

  req->work_cb(req);
}


static void uv__queue_done(struct uv__work* w, int err) {
  uv_work_t* req;

  if (w == NULL) {
    UV_LOGE("uv_work_t is NULL");
    return;
  }

  req = container_of(w, uv_work_t, work_req);
  uv__req_unregister(req->loop, req);

  if (req->after_work_cb == NULL)
    return;

  req->after_work_cb(req, err);
}


#ifdef USE_FFRT
void uv__ffrt_work(ffrt_executor_task_t* data, ffrt_qos_t qos)
{
  struct uv__work* w = (struct uv__work *)data;
  uv_loop_t* loop = w->loop;
#ifdef UV_STATISTIC
  uv__post_statistic_work(w, WORK_EXECUTING);
#endif
  uv_work_t* req = container_of(w, uv_work_t, work_req);
#ifdef ASYNC_STACKTRACE
  LibuvSetStackId((uint64_t)req->reserved[3]);
#endif
  w->work(w);
#ifdef UV_STATISTIC
  uv__post_statistic_work(w, WORK_END);
#endif
  rdlock_closed_uv_loop_rwlock();
  if (loop->magic != UV_LOOP_MAGIC) {
    rdunlock_closed_uv_loop_rwlock();
    UV_LOGE("uv_loop(%{public}zu:%{public}#x) in task(%p:%p) is invalid",
            (size_t)loop, loop->magic, req->work_cb, req->after_work_cb);
    return;
  }

  uv_mutex_lock(&loop->wq_mutex);
  w->work = NULL; /* Signal uv_cancel() that the work req is done executing. */

  if (uv_check_data_valid((struct uv_loop_data*)(loop->data)) == 0) {
    int status = (w->work == uv__cancelled) ? UV_ECANCELED : 0;
    struct uv_loop_data* addr = (struct uv_loop_data*)((uint64_t)loop->data -
      (UV_EVENT_MAGIC_OFFSET << UV_EVENT_MAGIC_OFFSETBITS));
    addr->post_task_func(addr->event_handler, uv__task_done_wrapper, (void*)w, status, qos);
  } else {
    uv__loop_internal_fields_t* lfields = uv__get_internal_fields(loop);
    uv__queue_insert_tail(&(lfields->wq_sub[qos]), &w->wq);
    uv_async_send(&loop->wq_async);
  }
  uv_mutex_unlock(&loop->wq_mutex);
  rdunlock_closed_uv_loop_rwlock();
}

static void init_once(void)
{
  init_closed_uv_loop_rwlock_once();
  /* init uv work statics queue */
#ifdef UV_STATISTIC
  init_work_dump_queue();
#endif
  ffrt_executor_task_register_func(uv__ffrt_work, ffrt_uv_task);
}


/* ffrt uv__work_submit */
void uv__work_submit(uv_loop_t* loop,
                     uv_req_t* req,
                     struct uv__work* w,
                     enum uv__work_kind kind,
                     void (*work)(struct uv__work *w),
                     void (*done)(struct uv__work *w, int status)) {
  uv_once(&once, init_once);
  ffrt_task_attr_t attr;
  ffrt_task_attr_init(&attr);

  switch(kind) {
    case UV__WORK_CPU:
      ffrt_task_attr_set_qos(&attr, ffrt_qos_default);
      break;
    case UV__WORK_FAST_IO:
      ffrt_task_attr_set_qos(&attr, ffrt_qos_default);
      break;
    case UV__WORK_SLOW_IO:
      ffrt_task_attr_set_qos(&attr, ffrt_qos_background);
      break;
    default:
#ifdef USE_OHOS_DFX
      UV_LOGI("Unknown work kind");
#endif
      return;
  }

  w->loop = loop;
  w->work = work;
  w->done = done;

  req->reserved[0] = (void *)(intptr_t)ffrt_task_attr_get_qos(&attr);
  ffrt_executor_task_submit((ffrt_executor_task_t *)w, &attr);
  ffrt_task_attr_destroy(&attr);
}


/* ffrt uv__work_submit */
void uv__work_submit_with_qos(uv_loop_t* loop,
                     uv_req_t* req,
                     struct uv__work* w,
                     ffrt_qos_t qos,
                     void (*work)(struct uv__work *w),
                     void (*done)(struct uv__work *w, int status)) {
    uv_once(&once, init_once);
    ffrt_task_attr_t attr;
    ffrt_task_attr_init(&attr);
    ffrt_task_attr_set_qos(&attr, qos);

    w->loop = loop;
    w->work = work;
    w->done = done;

    req->reserved[0] = (void *)(intptr_t)ffrt_task_attr_get_qos(&attr);
    ffrt_executor_task_submit((ffrt_executor_task_t *)w, &attr);
    ffrt_task_attr_destroy(&attr);
}
#endif


int uv_queue_work(uv_loop_t* loop,
                  uv_work_t* req,
                  uv_work_cb work_cb,
                  uv_after_work_cb after_work_cb) {
  if (work_cb == NULL)
    return UV_EINVAL;

  uv__req_init(loop, req, UV_WORK);
  req->loop = loop;
  req->work_cb = work_cb;
  req->after_work_cb = after_work_cb;

#ifdef UV_STATISTIC
  struct uv_work_dump_info* info = (struct uv_work_dump_info*) malloc(sizeof(struct uv_work_dump_info));
  if (info == NULL) {
    abort();
  }
  uv_init_dump_info(info, &req->work_req);
  info->builtin_return_address[0] = __builtin_return_address(0);
  info->builtin_return_address[1] = __builtin_return_address(1);
  info->builtin_return_address[2] = __builtin_return_address(2);
  (req->work_req).info = info;
#endif
#ifdef ASYNC_STACKTRACE
  req->reserved[3] = (void*)LibuvCollectAsyncStack();
#endif
  uv__work_submit(loop,
#ifdef USE_FFRT
                  (uv_req_t*)req,
#endif
                  &req->work_req,
                  UV__WORK_CPU,
                  uv__queue_work,
                  uv__queue_done
);
#ifdef UV_STATISTIC
  uv_queue_statics(info);
#endif
  return 0;
}


int uv_queue_work_with_qos(uv_loop_t* loop,
                  uv_work_t* req,
                  uv_work_cb work_cb,
                  uv_after_work_cb after_work_cb,
                  uv_qos_t qos) {
#ifdef USE_FFRT
  if (work_cb == NULL)
    return UV_EINVAL;

  STATIC_ASSERT(uv_qos_background == ffrt_qos_background);
  STATIC_ASSERT(uv_qos_utility == ffrt_qos_utility);
  STATIC_ASSERT(uv_qos_default == ffrt_qos_default);
  STATIC_ASSERT(uv_qos_user_initiated == ffrt_qos_user_initiated);
  if (qos < ffrt_qos_background || qos > ffrt_qos_user_initiated) {
    return UV_EINVAL;
  }

  uv__req_init(loop, req, UV_WORK);
  req->loop = loop;
  req->work_cb = work_cb;
  req->after_work_cb = after_work_cb;
#ifdef UV_STATISTIC
  struct uv_work_dump_info* info = (struct uv_work_dump_info*)malloc(sizeof(struct uv_work_dump_info));
  if (info == NULL) {
    abort();
  }
  uv_init_dump_info(info, &req->work_req);
  info->builtin_return_address[0] = __builtin_return_address(0);
  info->builtin_return_address[1] = __builtin_return_address(1);
  info->builtin_return_address[2] = __builtin_return_address(2);
  (req->work_req).info = info;
#endif
  uv__work_submit_with_qos(loop,
                  (uv_req_t*)req,
                  &req->work_req,
                  (ffrt_qos_t)qos,
                  uv__queue_work,
                  uv__queue_done);
#ifdef UV_STATISTIC
  uv_queue_statics(info);
#endif
  return 0;
#else
  return uv_queue_work(loop, req, work_cb, after_work_cb);
#endif
}


int uv_cancel(uv_req_t* req) {
  struct uv__work* wreq;
  uv_loop_t* loop;

  switch (req->type) {
  case UV_FS:
    loop =  ((uv_fs_t*) req)->loop;
    wreq = &((uv_fs_t*) req)->work_req;
    break;
  case UV_GETADDRINFO:
    loop =  ((uv_getaddrinfo_t*) req)->loop;
    wreq = &((uv_getaddrinfo_t*) req)->work_req;
    break;
  case UV_GETNAMEINFO:
    loop = ((uv_getnameinfo_t*) req)->loop;
    wreq = &((uv_getnameinfo_t*) req)->work_req;
    break;
  case UV_RANDOM:
    loop = ((uv_random_t*) req)->loop;
    wreq = &((uv_random_t*) req)->work_req;
    break;
  case UV_WORK:
    loop =  ((uv_work_t*) req)->loop;
    wreq = &((uv_work_t*) req)->work_req;
    break;
  default:
    return UV_EINVAL;
  }

  return uv__work_cancel(loop, req, wreq);
}