TP3/shell/tests/fcml_vs_objdump.py

#!/usr/bin/env python3
"""fcml_vs_objdump.py -- differential test between two 16-bit disassemblors.

Compares the *length of the first instruction* over many byte windows.  A
length disagreement is the sharpest possible signal that the two tools read
the ModRM/displacement/SIB fields differently, because if the lengths agree
the tools walked the same opcode table.

Usage:  fcml_vs_objdump.py [trials] [--seed N]
"""

import os
import random
import re
import subprocess
import sys
import tempfile

RE_LEN = re.compile(r"^\s*Instruction code length:\s*(\d+)\s*$", re.M)
RE_ADDR = re.compile(r"^\s*([0-9a-f]+):\t", re.M)


def fcml_first_len(win):
    hexs = "".join("%02X" % b for b in win)
    out = subprocess.run(["fcml-disasm", "-m16", "-rh", "-rz", "0x" + hexs],
                         capture_output=True, text=True)
    if out.returncode != 0:
        return None
    m = RE_LEN.search(out.stdout)
    return int(m.group(1)) if m else None


def objdump_first_len(win, tmp):
    with open(tmp, "wb") as f:
        f.write(win)
    out = subprocess.run(["objdump", "-D", "-b", "binary", "-m", "i8086", tmp],
                         capture_output=True, text=True)
    if out.returncode != 0:
        return None
    addrs = [int(a, 16) for a in RE_ADDR.findall(out.stdout)]
    if not addrs:
        return None
    if len(addrs) == 1:
        # single instruction covering the whole window
        return len(win)
    return addrs[1] - addrs[0]


def main(argv):
    trials = 3000
    seed = 1
    args = [a for a in argv[1:]]
    if "--seed" in args:
        i = args.index("--seed")
        seed = int(args[i + 1])
        del args[i:i + 2]
    if args:
        trials = int(args[0])
    rnd = random.Random(seed)
    tmp = os.path.join(tempfile.gettempdir(), "diffprobe.bin")
    agree = 0
    one_err = 0
    diffs = []
    for t in range(trials):
        n = rnd.choice([4, 6, 8, 10, 12, 15])
        win = bytes(rnd.randrange(256) for _ in range(n))
        a = fcml_first_len(win)
        b = objdump_first_len(win, tmp)
        if a is None or b is None:
            one_err += 1
            continue
        if a == b:
            agree += 1
        else:
            diffs.append((win.hex(" ").upper(), a, b))
    total = agree + len(diffs)
    print("seed=%d trials=%d compared=%d agree=%d (%d skipped: one tool "
          "rejected the input) disagreements=%d"
          % (seed, trials, total, agree, one_err, len(diffs)))
    for h, a, b in diffs[:25]:
        print("  %-32s fcml=%-3s objdump=%-3s" % (h, a, b))
    return 1 if diffs else 0


if __name__ == "__main__":
    sys.exit(main(sys.argv))