Create a project roadmap for future development

[mjhouse]

This will involve some thought and research, but generally:

• Figure out a rough feature set that I would consider 1.0.0
• Create between 2 and 4 milestones to achieve 1.0.0
• Break milestones down into issues on the backlog

There will be a number of other issues necessary for an open source project but not directly tied to a particular feature- things like adding good-first-task labels and a CONTRIBUTION.md file. Those will be enumerated in #7.

[mjhouse]

Milestone 1 (v0.2.0): ELF Parsing

This milestone is to create a complete parser for the ELF format. The focus is on simplicity and ease of use rather than performance or memory efficiency. All decoded headers, arrays, tables etc. should be write-able as well as read-able. For example- a consumer of this parser should be able to load an ELF file, access the symbol tables and their contained symbols, identify a symbol to remove, remove it and write the resulting ELF file to a new location.

• [ ] An ELF file can be loaded from a given path, and all following elements can be parsed
• [x] A file header, including all fields, can be read/written from/to an ELF file
• [x] All program headers, including all fields, can be read/written from/to an ELF file
• [x] All section headers, including all fields, can be read/written from/to an ELF file
• [x] All sections can be parsed from a file
• [ ] Array items can be parsed from sections that are arrays
• [ ] Sections that are arrays can be converted to array structs containing array items
  • [x] Init array
  • [ ] Preinit array
  • [x] Fini array
• [x] Table items can be parsed from sections that are tables
  • [x] String items
  • [x] Symbol items
  • [x] Relocation items
  • [ ] Hash items (?)
• [ ] Sections that are tables can be converted to table structs containing table items
  • [x] String table
  • [x] Symbol table
  • [ ] Dynamic symbol table
  • [x] Relocation table
  • [x] Relocation addend table
  • [ ] Hash table
  • [ ] Symbols that are objects can be used to access object data
  • [ ] Symbols that are functions can be used to access function data
  • [x] Sections that have bodies can be used to access section data
  • [ ] Section bodies are written with a correct alignment after change

References

• http://www.skyfree.org/linux/references/ELF_Format.pdf
• https://docs.oracle.com/cd/E23824_01/html/819-0690/chapter6-79797.html
• https://docs.oracle.com/cd/E23824_01/html/819-0690/chapter6-54839.html
• https://docs.oracle.com/cd/E23824_01/html/819-0690/chapter3-8.html

Milestone 2 (v0.3.0): X86 Disassembly

This milestone is to create a simple disassembler for x86 64/32-bit assembly. The focus is on simplicity and ease of use rather than performance. All decoded instructions, flags and operands should be write-able as well as read-able. For example- a consumer of this disassembler should be able to iterate through mutable instructions, match on the MOV instruction type, check the target register (if there is one) and current instruction address, change the target to a different register and continue on.

• [ ] Instruction prefixes can be parsed from binary
  • [ ] Repeat / Lock prefix (instruction will have exclusive use of all shared memory)
    • [ ] LOCK (0xF0)
  • [ ] String manipulation prefix
    • [ ] REP (0xF3)
    • [ ] REPE (0xF3)
    • [ ] REPNE (0xF2)
  • [ ] Segment override prefix (memory access to use specified segment instead of default segment)
    • [ ] 0x2E (CS)
    • [ ] 0x36 (SS)
    • [ ] 0x3E (DS)
    • [ ] 0x26 (ES)
    • [ ] 0x64 (FS)
    • [ ] 0x65 (GS)
  • [ ] Operand override prefix (Changes size of data expected by default mode of the instruction e.g. 16-bit to 32-bit and vice versa)
    • [ ] 0x66
  • [ ] Address override prefix (Changes size of address expected by the instruction. 32-bit address could switch to 16-bit and vice versa)
    • [ ] 0x67
• [ ] Instruction opcodes can be parsed from binary
  • [ ] Extended opcodes (beginning with 0x0F) are recognized
  • [ ] Standard opcodes are recognized
• [ ] ModR/M byte can be decoded
• [ ] SIB (scale-index-base) byte can be decoded
• [ ] Displacement operands can be decoded
• [ ] Immediate operands can be decoded
• [ ] Register operands can be identified

References

• https://pdos.csail.mit.edu/6.828/2008/readings/i386/s17_02.htm
• http://www.c-jump.com/CIS77/CPU/x86/X77_0240_prefix.htm
• https://stackoverflow.com/questions/924303/how-to-write-a-disassembler

Milestone 3 (v0.4.0): ARM Disassembly

This milestone is to enhance the milestone 2 disassembler with support for ARM 64/32 assembly. The focus, again, should be on simplicity and ease-of-use rather than performance. All instructions should be both read-able and write-able.

[mjhouse]

I'm resolving this, since I've created the milestones above and roughed in issues necessary to complete Milestone 1