This is the mail archive of the binutils@sources.redhat.com mailing list for the binutils project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Getting rid of BFD section->_raw_size and section->_cooked_size


Ya! And nice timing. Its really good to see BFD tightening the hatch on its interfaces! The tighter the iterfaces, the greater the freedom to inovate the internals.

Some, slightly mindless, suggestions taken from an external point of view follow.

- bfd_section_size (abfd, sec)
- bfd_unaltered_section_size (abfd, sec)

Per a recent post from me, if the BFD isn't needed, don't include it. GDB, in various places, is draging around "struct bfd"s just so that it can it pass (unused) into these methods. Also, an alternative name, such as bfd_section_current_size (sec), would drop a stronger hint that the former can and will change. However, whichever.

bfd_set_section_size (abfd, sec, size)

If the intent is for this method to only be used when constructing a section then a full initialization method, such as bfd_section_init VARIANT (sec, variant argument list), will be better. No matter how hard you try, people will subvert the _raw_size-write method in ways you never intend, "trust me" :-).

bfd_incr_section_size
bfd_incr_set_section_size

Suggest bfd_section_adjust_size(sec). "incr" strongly suggests only positive adjustment is permitted, and that negative adjustments requires the use of the [missing] "decr" method :-) A word like "adjust", being more vague, should make the general intent clear.

Happy hacking!
Andrew


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]