Re: [PATCH 2/3] gdbserver: Remove gdb_id_to_thread_id

[Simon Marchi <simon.marchi@polymtl.ca>]

On 2017-09-15 15:07, Simon Marchi wrote:
> On 2017-09-15 12:55, Pedro Alves wrote:
>> On 09/10/2017 09:11 PM, Simon Marchi wrote:
>>> From what I understand, this function is not doing anything useful as 
>>> of
>>> today.
>>> 
>>> Here's the result of my archeological research:
>>> 
>> 
>> *nod*
>> 
>>> --- a/gdb/gdbserver/server.c
>>> +++ b/gdb/gdbserver/server.c
>>> @@ -4011,7 +4011,6 @@ process_serial_event (void)
>>>    unsigned int len;
>>>    int res;
>>>    CORE_ADDR mem_addr;
>>> -  int pid;
>>>    unsigned char sig;
>>>    int packet_len;
>>>    int new_packet_len = -1;
>>> @@ -4039,92 +4038,96 @@ process_serial_event (void)
>>>        handle_general_set (own_buf);
>>>        break;
>>>      case 'D':
>>> -      require_running (own_buf);
>>> +      {
>>> +	require_running (own_buf);
>>> 
>> 
>> The reindentation makes it hard to see the actual
>> change.  Is it just moving the int pid variable, or something else?
>> IMO, it'd be nicer to move the whole case 'D' body to a
>> handle_detach function.
> 
> Agreed, I'll do a preparatory patch that does that, and then a second
> version of this one (also taking into account all your other
> comments).

I realized I did not answer you question.  Yes it was just moving the 
pid variable.

Simon