Awesome

Binary Ninja PCode

This plugin serves as a bridge between Binary Ninja and Ghidra's disassembler. This allows using all of Ghidra's supported architectures in Binary Ninja. It also makes a very poor attempt to convert p-code into Binary Ninja's Low Level IL. However, the outputted IL is very verbose and doesn't translate well into high level IL or Pseudo C.

Comparison of disassembly between Binary Ninja and Ghidra. Architecture x86-64 - Binary: /bin/ls.

Compilation

Ensure the repository is cloned with submodules (git submodule update --init).

mkdir build
cd build
CMAKE_PREFIX_PATH=<path to Qt6>/gcc_64/lib/cmake BN_INSTALL_DIR=<path to binary ninja> BN_API_PATH=<path to binaryninja-api> cmake ..
make -j$(nproc) install

Usage

Specify an architecture to be loaded using the LOAD_SLA environment variable. If no architecture (or all) is specified all architectures wil be loaded on startup, which might take a while.

LOAD_SLA="x86-64" ~/binaryninja/binaryninja

Known issues

• The Low Level IL is very verbose and produces unusable pseudo C.
• The plugin does not set the sleigh context specified in the .pspec. For x86-64 this is current hard-coded. If you get unexpected results like wrong instruction sizes, try setting the context for you architecture manually.
• No support for calling conventions or special registers (e.g. sp, lp).
• Overlapping registers are not handled. E.g. rax and eax are treated as separate registers.
• The disassembly is not tokenized into numbers and addresses, so clicking on an address doesn't work.
• Translation from pcode to IL is missing some instructions, and there are probably bugs.

Verbose Pseudo C output from Binary Ninja compared to Ghidra. Architecture V850 - Binary: some VW ecu dump.

Credits

• Ghidra - for the enormous amount of work that went into creating and supporting all these different architectures.
• pypcode - python library from the angr project to interface with pcode. Used some of their helper classes to simplify loading bytes into the pcode context.

Supported architectures

6502
65c02
68020
68030
68040
6805
6809
80251
80390
8048
8051
8085
AARCH64
AARCH64_AppleSilicon
AARCH64BE
ARM4_be
ARM4_le
ARM4t_be
ARM4t_le
ARM5_be
ARM5_le
ARM5t_be
ARM5t_le
ARM6_be
ARM6_le
ARM7_be
ARM7_le
ARM8_be
ARM8_le
avr32a
avr8
avr8e
avr8eind
avr8xmega
coldfire
CP1600
CR16B
CR16C
Dalvik_Base
Dalvik_DEX_Android10
Dalvik_DEX_Android11
Dalvik_DEX_Android12
Dalvik_DEX_KitKat
Dalvik_DEX_Lollipop
Dalvik_DEX_Marshmallow
Dalvik_DEX_Nougat
Dalvik_DEX_Oreo
Dalvik_DEX_Pie
Dalvik_ODEX_KitKat
data-be-64
data-le-64
dsPIC30F
dsPIC33C
dsPIC33E
dsPIC33F
H6309
HC05
HC08
HCS08
HCS12
JVM
m8c
MCS96
mips32be
mips32le
mips32R6be
mips32R6le
mips64be
mips64le
mx51
pa-risc32be
pic12c5xx
pic16
pic16c5x
pic16f
pic17c7xx
pic18
PIC24E
PIC24F
PIC24H
ppc_32_4xx_be
ppc_32_4xx_le
ppc_32_be
ppc_32_le
ppc_32_quicciii_be
ppc_32_quicciii_le
ppc_64_be
ppc_64_isa_altivec_be
ppc_64_isa_altivec_le
ppc_64_isa_altivec_vle_be
ppc_64_isa_be
ppc_64_isa_le
ppc_64_isa_vle_be
ppc_64_le
riscv
riscv
sh-1
sh-2
sh-2a
SparcV9_32
SparcV9_64
SuperH4_be
SuperH4_le
TI_MSP430
TI_MSP430X
toy_be
toy_be_posStack
toy_builder_be
toy_builder_be_align2
toy_builder_le
toy_builder_le_align2
toy_le
toy_wsz_be
toy_wsz_le
toy64_be
toy64_be_harvard
toy64_le
tricore
V850
x86
x86-64
z180
z80