Re: [PATCH v3] Disable get_ptrace_pid for NetBSD

[Simon Marchi <simark@simark.ca>]

On 2020-03-18 7:16 p.m., Kamil Rytarowski wrote:
> @@ -336,15 +350,14 @@ get_ptrace_pid (ptid_t ptid)
>  void
>  inf_ptrace_target::resume (ptid_t ptid, int step, enum gdb_signal signal)
>  {
> -  pid_t pid;
>    int request;
> 
>    if (minus_one_ptid == ptid)
>      /* Resume all threads.  Traditionally ptrace() only supports
>         single-threaded processes, so simply resume the inferior.  */
> -    pid = inferior_ptid.pid ();
> +    ptid = inferior_ptid;
>    else
> -    pid = get_ptrace_pid (ptid);
> +    ptid = ptid_t (inferior_ptid.pid ());

That's not what I meant.  To keep the existing behavior, I believe it should be:

  if (minus_one_ptid == ptid)
    /* Resume all threads.  Traditionally ptrace() only supports
       single-threaded processes, so simply resume the inferior.  */
    ptid = ptid_t (inferior_ptid.pid ());

> 
>    if (catch_syscall_enabled () > 0)
>      request = PT_SYSCALL;
> @@ -365,7 +378,7 @@ inf_ptrace_target::resume (ptid_t ptid, int step, enum gdb_signal signal)
>       where it was.  If GDB wanted it to start some other way, we have
>       already written a new program counter value to the child.  */
>    errno = 0;
> -  ptrace (request, pid, (PTRACE_TYPE_ARG3)1, gdb_signal_to_host (signal));
> +  gdb_ptrace (request, ptid, (PTRACE_TYPE_ARG3)1, gdb_signal_to_host (signal));
>    if (errno != 0)
>      perror_with_name (("ptrace"));
>  }

I'm getting this:

/home/simark/src/binutils-gdb/gdb/inf-ptrace.c: In member function ‘virtual void inf_ptrace_target::resume(ptid_t, int, gdb_signal)’:
/home/simark/src/binutils-gdb/gdb/inf-ptrace.c:379:15: error: invalid conversion from ‘int’ to ‘__ptrace_request’ [-fpermissive]
  379 |   gdb_ptrace (request, ptid, (PTRACE_TYPE_ARG3)1, gdb_signal_to_host (signal));
      |               ^~~~~~~
      |               |
      |               int
/home/simark/src/binutils-gdb/gdb/inf-ptrace.c:41:30: note:   initializing argument 1 of ‘int gdb_ptrace(__ptrace_request, ptid_t, long int, long int)’
   41 | gdb_ptrace (PTRACE_TYPE_ARG1 request, ptid_t ptid, PTRACE_TYPE_ARG3 addr,


We would need to change the type of the variable `ret` to PTRACE_TYPE_ARG1.

I'll make a test run on Linux with those changes to check if there's any regression.

Simon