[PATCH v3 5/7] gdb/remote.c: refactor pending fork status functions

[Pedro Alves]

On 2021-12-01 14:44, Simon Marchi via Gdb-patches wrote:
> From: Simon Marchi <simon.marchi@efficios.com>
> 
> In preparation for a following patch, refactor a few things that I did
> find a bit awkward, and to make them a bit more reusable.
> 
>  - Pass an inferior to kill_new_fork_children instead of a pid.  That
>    allows iterating on only this inferior's threads and avoid further
>    filtering on the thread's pid.
>  - Change thread_pending_fork_status to return a non-nullptr value only
>    if the thread does have a pending fork status.
>  - Remove is_pending_fork_parent_thread, as one can just use
>    thread_pending_fork_status and check for nullptr.
>  - Replace is_pending_fork_parent with is_fork_status, which just
>    returns if the given targeet_waitkind if a fork or a vfork.  Push

typo "targeet_waitkind".

>  
> -/* Return the thread's pending status used to determine whether the
> -   thread is a fork parent stopped at a fork event.  */
> +/* Return THREAD's pending status if the it is a pending fork parent, else
> +   return nullptr.  */

Some typo in "if the it is".

>  
> -static const target_waitstatus &
> +static const target_waitstatus *
>  thread_pending_fork_status (struct thread_info *thread)
>  {


> -/* Kill any new fork children of process PID that haven't been
> +/* Kill any new fork children of inferior INF that haven't been
>     processed by follow_fork.  */
>  
>  void
> -remote_target::kill_new_fork_children (int pid)
> +remote_target::kill_new_fork_children (inferior *inf)
>  {
>    remote_state *rs = get_remote_state ();
>    struct notif_client *notif = &notif_client_stop;
>  
>    /* Kill the fork child threads of any threads in process PID
>       that are stopped at a fork event.  */

Stale comment, refers to PID.

> -  for (thread_info *thread : all_non_exited_threads (this))
> +  for (thread_info *thread : inf->non_exited_threads ())
>      {
> -      const target_waitstatus &ws = thread->pending_follow;
> +      const target_waitstatus *ws = thread_pending_fork_status (thread);
>  
> -      if (is_pending_fork_parent (ws, pid, thread->ptid))
> -	{
> -	  int child_pid = ws.child_ptid ().pid ();
> -	  int res;
> +      if (ws == nullptr)
> +	continue;
>  
> -	  res = remote_vkill (child_pid);
> -	  if (res != 0)
> -	    error (_("Can't kill fork child process %d"), child_pid);
> -	}
> +      int child_pid = ws->child_ptid ().pid ();
> +      int res = remote_vkill (child_pid);
> +
> +      if (res != 0)
> +	error (_("Can't kill fork child process %d"), child_pid);
>      }
>  
>    /* Check for any pending fork events (not reported or processed yet)
>       in process PID and kill those fork child threads as well.  */

Ditto.

Otherwise LGTM.