TODO

## TOP PRIORITY

Fix all parsing bugs in crashes. I have some more fixing and testing to go before I update
the release tag of this alpha version, so this last commit is not a complete update to
the code base, but resolves quite a number of the existing issues.

## Handle SCOP (Secure code partitioning)

We must determine how functions such as elf_text_base are going to work
when there are multiple LOAD segments in SCOP linked binaries.

## Merge in Ignacios work for code injection

## 9/17/2018

Make sure that we take into account dynamic segments that exist in
gcc -nostdlib linked binaries. The dynamic segment exists but only
partially. We must account for this in various spots. Otherwise it
can lead to assuming we found DT_PLTREL* etc. which has led to FPE
bugs.

## TODO 5/24/2017

* If there are no section headers then make sure to use dynamic segment to setup
* symtab and other items (.i.e plt/got)
-- status: complete

* We should be able to load any file that the kernel can load, and reconstruct
* finer-grained meta data such as sections
based on techniques we used in ECFS, etc.
-- status: complete

## FINISHED - * elf_plt_by_name()

## FINISHED - * Write a tool for printing the PLT entries for every JUMP_SLOT
symbol

## FINISHED - * elf_plt_iterator_next prints "PLT-0" as its final (extra)
iteration.

* userland execve based debugging (userland ptrace) with breakpoint, peek, poke,
* continue, attach, etc. (20% complete).

* Display forensically reconstructed sections in ascending order by sh_addr
rather than the order they were reconstructed in.

# eh_frame symbol reconstruction with text padding viruses

With a text padding Virus, the parasite ends up in the last section header, which is
on x86 usually .eh_frame. Because of this, libelfmaster sees the .eh_frame section
and continues parsing into the parasite code resulting in corrupted FDE data and
therefore fails. We must handle this situation somehow!