[Dwarf-Discuss] Ambiguity in DWARF4 of debug_info_offset in .debug_aranges

Ron Brender ron.brender at charter.net
Sun Aug 15 10:45:35 PDT 2010


Keith,

Re section 6.1.2, Lookup by Address, I believe you are correct that it 
is ambiguous whether the debug_info_offset refers to the .debug_info or 
.debug_types section.

I suspect this is an error that was introduced when the type signature 
concept was introduced into DWARF and most mentions of .debug_info 
needed to be augmented to mention .debug_types as well. In this case, 
.debug_types should not be mentioned.

In practice, the .debug_types section by its nature will usually not 
contain any entries for an entity that has an associated address in 
memory--static data members might be the exception. Even in this case, 
the data member definition should not be in the shared type description.

Re Section 2.5.1.5, Control Flow Operations, I think this is am 
ambiguity whose origin is the same to the above. Most implementations 
link the DWARF sections such that they are not loaded in memory and all 
have a base address of zero; as a result, relocation information will 
help to determine whether the offset is within .debug_info or debug_types.

These problems should be reported as Issues for resolution in DWARF 5.

Ron

-------------------
Keith Walker wrote:
> In section 6.1.2 of the DWARF3 spec. the debug_info_offset was defined as:
>  
>   "3.  debug_info_offset (section offset)
>  
>    The offset from the beginning of the .debug_info section of the 
> compilation unit header
>    referenced by the set."
>  
>  
> However in section 6.1.2 of the DWARF4 spec. the definition has been 
> changed to:
>  
>   "3.  debug_info_offset (section offset)
>  
>    The offset from the beginning of the .debug_info or .debug_types 
> section of the
>    compilation unit header referenced by the set."
>  
>  
> However there doesn't appear to be any obvious way of determing whether 
> the offset is into the .debug_info section or the .debug_types 
> section.    Have I missed something?
>  
> [A similar issue also appears in section 2.5.1.5 for the definition of 
> the DW_OP_call_ref operator;   but in that case you might expect there 
> to be some additional relocation information available to help determine 
> which section is being referenced.]
>  
> Keith Walker
>  
>  
> 
> -- IMPORTANT NOTICE: The contents of this email and any attachments are 
> confidential and may also be privileged. If you are not the intended 
> recipient, please notify the sender immediately and do not disclose the 
> contents to any other person, use it for any purpose, or store or copy 
> the information in any medium. Thank you.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Dwarf-Discuss mailing list
> Dwarf-Discuss at lists.dwarfstd.org
> http://lists.dwarfstd.org/listinfo.cgi/dwarf-discuss-dwarfstd.org





More information about the Dwarf-Discuss mailing list