Re: [PATCH] ada-lang.c:coerce_unspec_val_to_type: Preserve laziness.

[Pedro Alves <palves at redhat dot com>]

On 07/05/2013 11:29 AM, Andrew Burgess wrote:
> On 04/07/2013 7:22 PM, Pedro Alves wrote:
>> ada-lang.c:coerce_unspec_val_to_type does:
>>
>>       if (value_lazy (val)
>> 	  || TYPE_LENGTH (type) > TYPE_LENGTH (value_type (val)))
>> 	result = allocate_value_lazy (type);
>>       else
>> 	{
>> 	  result = allocate_value (type);
>> 	  memcpy (value_contents_raw (result), value_contents (val),
>> 		  TYPE_LENGTH (type));
>> 	}
>>       set_value_component_location (result, val);
>>       set_value_bitsize (result, value_bitsize (val));
>>       set_value_bitpos (result, value_bitpos (val));
>>       set_value_address (result, value_address (val));
>>       set_value_optimized_out (result, value_optimized_out (val));
>>
>> Notice that before value_optimized_out was made to auto-fetch lazy
>> values, VAL would end up still lazy if it was lazy on entry.  It's not
>> really a problem here if VAL is lazy, and VAL->optimized_out is 0,
>> because RESULT is also left lazy.  IOW, this just wants to copy the
>> VAL->optimized_out flag to RESULT->optimized_out, nothing else.
>>
>> The patch adds the value_optimized_out_const function for that.
> 
> As an alternative, this patch avoids the new _const function,
> I move the setting of the optimized_out value into the else case where we know val is no longer lazy.
> 
> Cheers,
> Andrew
> 
> 
> diff --git a/gdb/ada-lang.c b/gdb/ada-lang.c
> index 8240fee..ea7b579 100644
> --- a/gdb/ada-lang.c
> +++ b/gdb/ada-lang.c
> @@ -576,12 +576,14 @@ coerce_unspec_val_to_type (struct value *val, struct type *type)
>  	  result = allocate_value (type);
>  	  memcpy (value_contents_raw (result), value_contents (val),
>  		  TYPE_LENGTH (type));
> +	  /* Only copy the optimized out setting if val is not lazy.  If it
> +	     is lazy then the optimized_out flag will always be false.  */
> +	  set_value_optimized_out (result, value_optimized_out (val));
>  	}
>        set_value_component_location (result, val);
>        set_value_bitsize (result, value_bitsize (val));
>        set_value_bitpos (result, value_bitpos (val));
>        set_value_address (result, value_address (val));
> -      set_value_optimized_out (result, value_optimized_out (val));
>        return result;

That works, but I had originally discarded such an approach
because it looked brittle to me.  E.g., I was toying with optimizing
value_optimized_out by only fetching lazy values if it really can't
avoid it, and such a change would render this subtly broken.

-- 
Pedro Alves