Decode symbols

[thetheodor]

Hello, I'm trying to recover the symbol names of disassembled programs. For example:

cat test.c
long b;
long c;
int foo() { return b + c; }
gcc -O1  test.c -c 
objdump -D test.o  

test.o:     file format elf64-x86-64


Disassembly of section .text:

0000000000000000 <foo>:
   0:	48 8b 05 00 00 00 00 	mov    0x0(%rip),%rax        # 7 <foo+0x7>
   7:	03 05 00 00 00 00    	add    0x0(%rip),%eax        # d <foo+0xd>
   d:	c3                   	ret

Disassembly of section .bss:

0000000000000000 <c>:
	...

0000000000000008 <b>:
	...

Here we can do some math to figure out that the offsets <foo+0x7> and <food+0xd> are reffering to symbols c and b.

My question (related to #300) is whether there is a way to do this within iced. If I just copy the .text section to a new object file and then within iced I disassemble it and print the instructions, I get:

mov 7,%rax
add 0xD,%eax
ret

If I include the .bss section then I get a lot of garbage instructions. Is there a way to tell iced that it's reading a .bss section and resolve the addresses to symbol names?

Bonus question: can this also be extended to cases where the assembler doesn't resolve the address but emits an actual rip-relative operand?

Thanks!

[thetheodor]

Answering my own question, I just need to read the relocation table (the same output as objdump -r) and figure out the address correspondence.