[lwip-devel] [bug #23408] Deadlock on sys_mbox_post sys_mbox_fetch

[Vadim G Melihow]

Follow-up Comment #5, bug #23408 (project lwip):

Ok I try again. 8-). Sorry for my english.

I use lwip v 1.3.0 under Linux.

my sys_arch file is contrib/ports/unix/sys_arch.c


void
sys_mbox_post(struct sys_mbox *mbox, void *msg)
{
  u8_t first;

  sys_sem_wait(mbox->mutex);

  LWIP_DEBUGF(SYS_DEBUG, ("sys_mbox_post: mbox %p msg %pn", (void *)mbox,
(void *)msg));

  while ((mbox->last + 1) >= (mbox->first + SYS_MBOX_SIZE)) {
    mbox->wait_send++;
    sys_sem_signal(mbox->mutex); <------ 1st operation

/*
I want to say that between these two operations (1st and 2nd) possibly switch
CPU context to writer thread and filling the whole messagebox, followed by
blocking semaphore mbox->mail.

As a result we get a deadlock.

When a large amount of msgbox elements (SYS_MBOX_SIZE) that is likely to
happen is small. 

*/

    sys_arch_sem_wait(mbox->mail, 0); <-------- 2nd operation
    sys_arch_sem_wait(mbox->mutex, 0);
    mbox->wait_send--;
  }

  mbox->msgs[mbox->last % SYS_MBOX_SIZE] = msg;

  if (mbox->last == mbox->first) {
    first = 1;
  } else {
    first = 0;
  }

  mbox->last++;

  if (first) {
    sys_sem_signal(mbox->mail);
  }

  sys_sem_signal(mbox->mutex);
}






    _______________________________________________________

Reply to this item at:

  <http://savannah.nongnu.org/bugs/?23408>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.nongnu.org/