casualx / lde Goto Github PK

Great lib! Thanks!!

I've been testing your library on a Windows binary (PE) and I noticed that depending where I start the iteration in the buffer I don't have the same number of iterations.

I would have expected that if I start at 0x600 (where the opcodes seem to start) I would iterate until the end of the file (size=67MB) but I got less iterations than if I start at some other places that was of interest for me.

I have the feeling the iteration stops when it encounters something it doesn't understand OR maybe there is something I don't understand about the use of this library itself. Am I correct to think that the point of the iter function is to read the buffer to the end and provide an opcode for every iteration? Or am I wrong? (might sounds like a stupid question but I'm a beginner on this)

I don't mind writing some small code to help to debug. I'm not blocked by this issue but I'm interested to know more and participate on the development.

It would be nice to see a version of this go up on crates.io. Github pages documentation via travis-cargo would be useful too, as it's so nicely documented!

log

BUG -> bytes: [0F, 0E] -> must have len: 2 -> have len: 0 -> mnemonic: femms
BUG -> bytes: [0F, C9] -> must have len: 2 -> have len: 0 -> mnemonic: bswap ecx
BUG -> bytes: [0F, CA] -> must have len: 2 -> have len: 0 -> mnemonic: bswap edx
BUG -> bytes: [0F, CB] -> must have len: 2 -> have len: 0 -> mnemonic: bswap ebx
BUG -> bytes: [0F, CC] -> must have len: 2 -> have len: 0 -> mnemonic: bswap esp
BUG -> bytes: [0F, CD] -> must have len: 2 -> have len: 0 -> mnemonic: bswap ebp
BUG -> bytes: [0F, CE] -> must have len: 2 -> have len: 0 -> mnemonic: bswap esi
BUG -> bytes: [0F, CF] -> must have len: 2 -> have len: 0 -> mnemonic: bswap edi

objdump:
49 b8 00 00 00 00 01 00 00 00
movabs $0x100000000,%r8

lde stops the instruction at 49 b8 00 00 00 00.

Probably the same issue with 48 b8 4a ba e2 b9 29 ab b9 bf.