[Dwarf-discuss] Re: LSB DWARF Chapter

[David Anderson]

Thanks for sending the notice to the dwarf-discuss mailing list.
I am quite happy to learn work is going forward on such a
document.

The proposed Elf LSB document uses the term
        reletive
in at least 3 places (Dwarf Execption Header Encoding).
I believe the word in US English is 'relative' with just 2 'e', not 3.
(British English could be different, I don't know.)

Section 7.1 DWARF Exception Header Encoding
wades right into differences with DWARF 2/3 with no
overview or context.   Later the augmentation of 'P'
mentions 'personality' and the terminology here
in 7.1 does not use that term, so a reader has (I think)
a very difficult time understanding that section 7.1
relates to  the 'P' augmentation byte.
Using identical terms in 7.1 and the 'P' documentation
would be helpful.

All of section 7 refers freely to DWARF3 (which, IMO, is just fine)
but seems not to make clear that .eh_frame is a GNU Elf
feature which uses DWARF3 .debug_frame as it's base design,
but is, nonetheless, a section 
I fear failing to make this clear will confuse readers
of the Elf document who do follow up by reading the DWARF3 document.
The detail of the section-name  being different from .debug_frame
won't, I fear, leap out at readers.


The 7.2. DWARF CFI Extensions section
is odd in that it names DWARF2 and links to it, but
adds DWARF3 operations.  It seems to me that the DWARF3
document on FreeStandards.org would be the better document
to reference, but also mentioning the DWARF2 document
would be fine.  With DWARF3 referenced, calling the *_sf
operations 'additional' seems odd.
Actually this comment applies generally.  Linking to
the DWARF3 document as a base seems more applicable than
DWARF2. 




Typo alert:
"The Agumentation String indicates the presence of some
optional fields, and how those fileds should be intepreted.
                               ^^^^^^  spelling error here.
Each character in the augmentation string in the CIE can be
interpreted as below:"


The CIE entry Frame Description Entry have some minor problems too.
Mainly: neither it nor the CIE of .eh_frame give lengths of the fields
for most fields.  Nor do they refer to the DWARF2/3 documents
as additional guidance (where such applies).
Nor to they make clear which fields are 'just like DWARF3' and
which are .eh_frame specific.



Regards,
David B. Anderson davea at sgi dot com http://reality.sgiweb.org/davea