StarForth/scripts/extract_doe_region.py

#!/usr/bin/env python3
"""
extract_doe_region.py -- spool the DOE-PERSIST-TRIAL binary ring off
disk/artemis.img into a CSV, for the multiuser/multitasking DoE campaign
report (FABRIC-3.md §XXXV.10).

Reads the SAME devblock_from_top-addressed ring doe_region.c writes
(include/starkernel/doe_region.h has the authoritative layout: control
header at devblock_from_top=98, slot devblocks growable from 99, 64-byte
doe_trial_slot_t records, 64 slots/devblock). Two access modes:

  --img PATH        Read the image file directly (default). No privilege
                     needed, works on a plain regular file. This is what
                     "reading the drive" reduces to when the image is
                     just a file, which is the normal QEMU-dev case.

  --loop PATH        Attach PATH via losetup first (needs sudo), read the
                     resulting /dev/loopN node, then detach it again on
                     exit -- the literal "mount the virtual drive on a
                     loop device" step for when this runs against a real
                     block device or a copy pulled off real hardware. The
                     ring is raw binary at a fixed byte offset, not a
                     filesystem -- there is nothing to `mount` in the
                     POSIX-filesystem sense, so this attaches the loop
                     device and reads it exactly like a regular file.

Either mode is read-only and never modifies the image. Safe to run
against a live campaign's artemis.img (the ring's control header CRC
means a torn read mid-write is simply rejected as "ring not yet
initialized" by the same discipline doe_region.c itself uses -- see
doe_region_ctrl_read()); worst case is missing the very latest record
until the next run of this script.

CRC verification (§XXXV.12): both the control header and each individual
trial slot carry a CRC-64/XZ checksum (same polynomial and algorithm as
src/block_subsystem.c's compute_crc64() -- reimplemented here in Python,
not guessed at; verified bit-for-bit against a live kernel-written
record). A bad header CRC is treated as "ring not yet initialized" (same
as a bad magic/version). A bad slot CRC means that record is skipped and
reported, not silently included as if it were real data -- since
DOE_SLOTS_PER_DEVBLOCK=64 slots share one devblock via a read-modify-
write, a crash mid-write to any slot can corrupt its siblings too.
"""

import argparse
import csv
import struct
import subprocess
import sys
from contextlib import contextmanager
from pathlib import Path

DEVBLOCK_SIZE = 4096

# Mirrors include/starkernel/doe_region.h and include/starkernel/log_region.h
ARTEMIS_SIG_DEVBLOCK_FROM_TOP = 64
LOG_REGION_DEVBLOCK_FROM_TOP_BASE = ARTEMIS_SIG_DEVBLOCK_FROM_TOP + 1  # 65
LOG_REGION_MAX_DEVBLOCKS = 32
DOE_REGION_DEVBLOCK_FROM_TOP_BASE = (
    LOG_REGION_DEVBLOCK_FROM_TOP_BASE + 1 + LOG_REGION_MAX_DEVBLOCKS
)  # 98

DOE_SLOT_SIZE = 64
DOE_SLOTS_PER_DEVBLOCK = DEVBLOCK_SIZE // DOE_SLOT_SIZE  # 64
DOE_REGION_MAGIC = 0x44454F44  # 'DOED'
DOE_REGION_VERSION_0 = 0

# doe_region_ctrl_t: magic(u64) devblocks(u32) head_slot(u32) tail_slot(u32)
# record_count(u32) hdr_crc(u64), little-endian, rest is padding to 4096.
CTRL_FMT = "<QIIIIQ"
CTRL_SIZE = struct.calcsize(CTRL_FMT)
CTRL_CRC_SPAN = struct.calcsize("<QIIII")  # everything before hdr_crc

# doe_trial_slot_t: fleet_k_q48(u64) switch_count_cumulative(u64)
# run_id(u32) cfg(u32) rep(u32) nw(u32) mode(u32) fail_count(u32)
# fleet_conserved(u32) isa(8s) slot_crc(u64), rest is padding to 64.
SLOT_FMT = "<QQIIIIIII8sQ"
SLOT_SIZE = struct.calcsize(SLOT_FMT)
SLOT_CRC_SPAN = struct.calcsize("<QQIIIIIII8s")  # everything before slot_crc
assert SLOT_SIZE <= DOE_SLOT_SIZE


# CRC-64/XZ, same polynomial/algorithm as compute_crc64() in
# src/block_subsystem.c (crc64_init()'s poly=0x42F0E1EBA9EA3693, table-
# driven, init/final XOR 0xFFFFFFFFFFFFFFFF) -- reimplemented here so
# host-side verification genuinely matches the kernel side, not guessed.
_CRC64_POLY = 0x42F0E1EBA9EA3693
_CRC64_TABLE = None


def _crc64_table():
    global _CRC64_TABLE
    if _CRC64_TABLE is None:
        table = []
        for i in range(256):
            crc = i
            for _ in range(8):
                crc = (crc >> 1) ^ _CRC64_POLY if (crc & 1) else (crc >> 1)
            table.append(crc & 0xFFFFFFFFFFFFFFFF)
        _CRC64_TABLE = table
    return _CRC64_TABLE


def compute_crc64(data: bytes) -> int:
    table = _crc64_table()
    crc = 0xFFFFFFFFFFFFFFFF
    for byte in data:
        idx = (crc ^ byte) & 0xFF
        crc = table[idx] ^ (crc >> 8)
    return crc ^ 0xFFFFFFFFFFFFFFFF


def dft_to_offset(devblock_from_top: int, total_devblocks: int) -> int:
    idx = total_devblocks - 1 - devblock_from_top
    if idx < 0:
        raise ValueError(
            f"devblock_from_top={devblock_from_top} exceeds total_devblocks={total_devblocks}"
        )
    return idx * DEVBLOCK_SIZE


@contextmanager
def loop_attach(img_path: Path):
    """Attach img_path via losetup (read-only), yield the /dev/loopN path,
    detach on exit. Requires sudo -- the literal 'mount on a loop device'
    step; nothing here is a filesystem mount."""
    result = subprocess.run(
        ["sudo", "losetup", "--find", "--show", "--read-only", str(img_path)],
        capture_output=True, text=True, check=True,
    )
    loop_dev = result.stdout.strip()
    print(f"  attached {img_path} -> {loop_dev}", file=sys.stderr)
    try:
        yield loop_dev
    finally:
        subprocess.run(["sudo", "losetup", "--detach", loop_dev], check=False)
        print(f"  detached {loop_dev}", file=sys.stderr)


def read_ring(dev_path: str):
    """Returns a list of dicts, one per live trial record, in ring order
    (oldest first). Empty list if the ring has never been written."""
    size = Path(dev_path).stat().st_size if not dev_path.startswith("/dev/") else None
    if size is None:
        with open(dev_path, "rb") as f:
            f.seek(0, 2)
            size = f.tell()

    total_devblocks = size // DEVBLOCK_SIZE
    if total_devblocks <= DOE_REGION_DEVBLOCK_FROM_TOP_BASE:
        raise ValueError(
            f"image too small ({size} bytes, {total_devblocks} devblocks) "
            f"for the DOE region fence (needs > {DOE_REGION_DEVBLOCK_FROM_TOP_BASE})"
        )

    ctrl_off = dft_to_offset(DOE_REGION_DEVBLOCK_FROM_TOP_BASE, total_devblocks)
    with open(dev_path, "rb") as f:
        f.seek(ctrl_off)
        ctrl_raw = f.read(DEVBLOCK_SIZE)

    magic, devblocks, head_slot, tail_slot, record_count, hdr_crc = struct.unpack_from(
        CTRL_FMT, ctrl_raw, 0
    )
    magic_lo = magic & 0xFFFFFFFF
    version = (magic >> 32) & 0xFF
    if magic_lo != DOE_REGION_MAGIC or version != DOE_REGION_VERSION_0:
        print("  DOE region not yet initialized (no trials persisted yet)", file=sys.stderr)
        return []

    want_hdr_crc = compute_crc64(ctrl_raw[:CTRL_CRC_SPAN])
    if want_hdr_crc != hdr_crc:
        print("  DOE region control header CRC mismatch -- treating as "
              "'not yet initialized' (same discipline doe_region_ctrl_read() "
              "uses on the kernel side)", file=sys.stderr)
        return []

    total_slots = devblocks * DOE_SLOTS_PER_DEVBLOCK
    records = []
    corrupt = 0
    with open(dev_path, "rb") as f:
        idx = head_slot
        for _ in range(record_count):
            slot_devblock_dft = DOE_REGION_DEVBLOCK_FROM_TOP_BASE + 1 + (idx // DOE_SLOTS_PER_DEVBLOCK)
            slot_off_in_devblock = (idx % DOE_SLOTS_PER_DEVBLOCK) * DOE_SLOT_SIZE
            blk_off = dft_to_offset(slot_devblock_dft, total_devblocks)
            f.seek(blk_off + slot_off_in_devblock)
            raw = f.read(SLOT_SIZE)
            (fleet_k_q48, switch_count, run_id, cfg, rep, nw, mode,
             fail_count, fleet_conserved, isa_raw, slot_crc) = struct.unpack(SLOT_FMT, raw)

            want_slot_crc = compute_crc64(raw[:SLOT_CRC_SPAN])
            if want_slot_crc != slot_crc:
                corrupt += 1
                print(f"  slot {idx} CRC mismatch (devblock-sharing torn write?) -- skipped", file=sys.stderr)
                idx = (idx + 1) % total_slots
                continue

            isa = isa_raw.split(b"\x00", 1)[0].decode("ascii", errors="replace")
            records.append({
                "isa": isa,
                "run_id": run_id,
                "cfg": cfg,
                "rep": rep,
                "nw": nw,
                "mode": mode,
                "fail_count": fail_count,
                "fleet_k_q48": fleet_k_q48,
                "fleet_conserved": fleet_conserved,
                "switch_count_cumulative": switch_count,
            })
            idx = (idx + 1) % total_slots

    print(f"  ring: devblocks={devblocks} head={head_slot} tail={tail_slot} "
          f"record_count={record_count} -> {len(records)} valid records read"
          + (f", {corrupt} corrupt slot(s) skipped" if corrupt else ""),
          file=sys.stderr)
    return records


def main():
    ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
    ap.add_argument("--img", default="disk/artemis.img", help="Path to the disk image (default: disk/artemis.img)")
    ap.add_argument("--loop", action="store_true", help="Attach --img via losetup first (needs sudo)")
    ap.add_argument("-o", "--out", default=None, help="Output CSV path (default: stdout)")
    args = ap.parse_args()

    img_path = Path(args.img)
    if not img_path.exists():
        print(f"extract_doe_region: image not found: {img_path}", file=sys.stderr)
        sys.exit(1)

    if args.loop:
        with loop_attach(img_path) as loop_dev:
            records = read_ring(loop_dev)
    else:
        records = read_ring(str(img_path))

    fieldnames = ["isa", "run_id", "cfg", "rep", "nw", "mode", "fail_count",
                  "fleet_k_q48", "fleet_conserved", "switch_count_cumulative"]

    out_f = open(args.out, "w", newline="") if args.out else sys.stdout
    try:
        writer = csv.DictWriter(out_f, fieldnames=fieldnames)
        writer.writeheader()
        for rec in records:
            writer.writerow(rec)
    finally:
        if args.out:
            out_f.close()
            print(f"  wrote {len(records)} rows -> {args.out}", file=sys.stderr)


if __name__ == "__main__":
    main()