Hunk parser (and general hunk info)

Aim is to add support for Amiga Hunk format to Clang/LLVM.
we should be able to cross compile for Amiga on any modern computer

The parser is simply an exercise to understand the format but could be used to validate generated hunks

unanswered questions

• how do reloc hunks actually work? look at the Amigados manual page 371
• how is the entry point defined? is there a convention?
• what are external symbols?
• how are the code/data/bss sections represented by LLVM/Clang?
• what is the minimum viable support needed?
• how would we control things like hunk names, chip/fast mem requirements? config file/pragmas?

plan of attack

• create simple hello world application (no external lib deps)
• compile and link with sas/c
• examine hunk output
• disassemble / decompile code hunk
• repeat for increasingly complex examples
  • strings / constants
  • variables (inintialized / uninitialized)
  • structs

references

• SAS/C 650 vol1 chapter 12 for how the hunk sections are generated / controlled
• http://amiga-dev.wikidot.com/file-format:hunk
• http://fileformats.archiveteam.org/wiki/Amiga_Hunk
• https://en.wikipedia.org/wiki/Amiga_Hunk
• The AmigaDos Manual chapter 10 for the hunk format definition
• The AmigaDos Manual chapter 7 for details on what the linker does to create load files