[Dwarf-Discuss] decoding of form entries

Michael Eager eager at eagercon.com
Tue Aug 3 11:36:38 PDT 2010


I think that when we start discussing changes in the DWARF standard,
we should be working from a proposal and hold those discussions on
the committee mailing list.

Ron Brender wrote:
> I am in agreement with Todd and David G that the intent for the 
> attributes discussed in 2.19 is that the reference will identify a DIE 
> for an object (variable, parameter, constant), whose contents is the 
> value of the original attribute.
> 
> In no case was it ever envisioned, so far as I recall, that the target 
> DIE would need to have a "matching" attribute.
> 
> David A also has a point about the wording being not quite consistent 
> with intended usage, mixing up our conventions for entry versus entity. 
> Something like the following as a replacement for the second bullet in 
> 2.19 might be better, or at least a starting point for discussion:
> 
>   . "For a reference, the reference identifies another debugging 
> information entry, typically corresponding to a variable, constant or 
> other object; the value of the other entity specifies the value of the 
> attribute."
> 
> Moreover,scanning through the list of DIE kinds, I actually don't see 
> any ones other than the various object kinds that can actually make 
> sense in the context of Section 2.19. This suggests that without loss of 
> generality the second bullet can be made specific to objects as in
> 
>   . "For a reference, the reference identifies another debugging 
> information entry corresponding to a variable, constant or other object; 
> the contents of that object specifies the value of the attribute."
> 
> Note that this does not restrict what references can refer to in 
> general, just what they can refer for the attributes discussed in 
> Section 2.19.
> 
> Such are my 2 bits...
> 
> Ron
> 
> ---------------------
> David Anderson wrote:
>> On 08/02/2010 12:58 AM, Mathieu Lacage wrote:
>>> Something that has been bugging me for a while is that I am not sure I
>>> understand the meaning of 2.19  p 40 (dw4) that talks about how to
>>> retrieve the value associated to an attribute in an entry. It says:
>>>
>>> "For a reference, the value is a reference to another entity whose
>>> value is the value of the
>>> attribute."
>>
>> The quoted sentence from DWARF4 confuses me too.
>>
>> A 'reference' is defined in 'classes of attribute value', Figure 3, as 
>> referring to
>> a debugging information entry or a type signature.
>>
>> The word 'entity' is used to refer to data in the running application, 
>> as in
>> Sec 2.1:
>> " An entry, or group of entries together, provide a description of a
>> corresponding entity in the source program. "
>> Or elsewhere "entity in memory"  or "Program entity".
>>
>> Others seem to think the 'entity' is another DIE (indeed
>> I think that was the intent), but I wonder what
>> in DWARF4 terminology supports that interpretation of that word 
>> 'entity' in 2.19.
>>
>> The sentence seems(?) to be inconsistent with the rest of DWARF4 in its
>> use of 'entity'.
>>
>> David Anderson
>>
> 
> 
> 
> _______________________________________________
> Dwarf-Discuss mailing list
> Dwarf-Discuss at lists.dwarfstd.org
> http://lists.dwarfstd.org/listinfo.cgi/dwarf-discuss-dwarfstd.org
> 


-- 
Michael Eager	 eager at eagercon.com
1960 Park Blvd., Palo Alto, CA 94306  650-325-8077



More information about the Dwarf-Discuss mailing list