l2/tools/gen_linux_sl.py

#!/usr/bin/env python3
"""Generate stdlib/linux.sl from syscall_64.tbl metadata."""

from __future__ import annotations

from dataclasses import dataclass
from pathlib import Path
import re

ROOT = Path(__file__).resolve().parent.parent
SRC = ROOT / "syscall_64.tbl"
DST = ROOT / "stdlib" / "linux.sl"


def _sanitize_alias(alias: str) -> str:
    name = alias.strip()
    if not name:
        return ""
    if name.startswith("__x64_sys_"):
        name = name[len("__x64_sys_") :]
    elif name.startswith("sys_"):
        name = name[len("sys_") :]
    name = re.sub(r"[^A-Za-z0-9_]", "_", name)
    name = re.sub(r"_+", "_", name).strip("_")
    if not name:
        return ""
    if name[0].isdigit():
        name = "n_" + name
    return name


@dataclass(frozen=True)
class SyscallEntry:
    argc: int
    num: int
    aliases: tuple[str, ...]


def _parse_table(path: Path) -> list[SyscallEntry]:
    entries: list[SyscallEntry] = []
    for raw in path.read_text(encoding="utf-8").splitlines():
        line = raw.strip()
        if not line or line.startswith("#"):
            continue
        parts = line.split(maxsplit=2)
        if len(parts) < 3:
            continue
        try:
            argc = int(parts[0])
            num = int(parts[1])
        except ValueError:
            continue
        aliases = tuple(a for a in parts[2].split("/") if a)
        if not aliases:
            continue
        entries.append(SyscallEntry(argc=argc, num=num, aliases=aliases))
    return entries


def _emit_header(lines: list[str]) -> None:
    lines.extend(
        [
            "# Autogenerated from syscall_64.tbl",
            "# Generated by tools/gen_linux_sl.py",
            "# Linux syscall constants + convenience wrappers for L2",
            "",
            "# File descriptor constants",
            "macro FD_STDIN 0 0 ;",
            "macro FD_STDOUT 0 1 ;",
            "macro FD_STDERR 0 2 ;",
            "",
            "# Common open(2) flags",
            "macro O_RDONLY 0 0 ;",
            "macro O_WRONLY 0 1 ;",
            "macro O_RDWR 0 2 ;",
            "macro O_CREAT 0 64 ;",
            "macro O_EXCL 0 128 ;",
            "macro O_NOCTTY 0 256 ;",
            "macro O_TRUNC 0 512 ;",
            "macro O_APPEND 0 1024 ;",
            "macro O_NONBLOCK 0 2048 ;",
            "macro O_CLOEXEC 0 524288 ;",
            "",
            "# lseek(2)",
            "macro SEEK_SET 0 0 ;",
            "macro SEEK_CUR 0 1 ;",
            "macro SEEK_END 0 2 ;",
            "",
            "# mmap(2)",
            "macro PROT_NONE 0 0 ;",
            "macro PROT_READ 0 1 ;",
            "macro PROT_WRITE 0 2 ;",
            "macro PROT_EXEC 0 4 ;",
            "macro MAP_PRIVATE 0 2 ;",
            "macro MAP_ANONYMOUS 0 32 ;",
            "macro MAP_SHARED 0 1 ;",
            "",
            "# Socket constants",
            "macro AF_UNIX 0 1 ;",
            "macro AF_INET 0 2 ;",
            "macro AF_INET6 0 10 ;",
            "macro SOCK_STREAM 0 1 ;",
            "macro SOCK_DGRAM 0 2 ;",
            "macro SOCK_NONBLOCK 0 2048 ;",
            "macro SOCK_CLOEXEC 0 524288 ;",
            "",
            "macro INADDR_ANY 0 0 ;",
            "",
            "# Generic syscall helpers with explicit argument count",
            "# Stack form:",
            "#   syscall  -> <argN> ... <arg0> <argc> <nr> syscall",
            "#   syscallN -> <argN-1> ... <arg0> <nr> syscallN",
            "",
            "# swap impl is provided so this can be used without stdlib",
            "# ___linux_swap [*, x1 | x2] -> [*, x2 | x1]",
            ":asm ___linux_swap {",
            "    mov rax, [r12]",
            "    mov rbx, [r12 + 8]",
            "    mov [r12], rbx",
            "    mov [r12 + 8], rax",
            "}",
            ";",
            "",
            "macro syscall0 0",
            "    0",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall1 0",
            "    1",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall2 0",
            "    2",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall3 0",
            "    3",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall4 0",
            "    4",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall5 0",
            "    5",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
            "macro syscall6 0",
            "    6",
            "    ___linux_swap",
            "    syscall",
            ";",
            "",
        ]
    )


def _emit_entry(lines: list[str], alias: str, argc: int, num: int) -> None:
    safe_argc = max(0, min(argc, 6))
    lines.extend(
        [
            f"macro syscall.{alias} 0",
            f"    {num}",
            f"    syscall{safe_argc}",
            ";",
            "",
            f"macro syscall.{alias}.num 0",
            f"    {num}",
            ";",
            "",
            f"macro syscall.{alias}.argc 0",
            f"    {safe_argc}",
            ";",
            "",
        ]
    )


def generate() -> str:
    entries = _parse_table(SRC)

    lines: list[str] = []
    _emit_header(lines)

    emitted: set[str] = set()
    for entry in sorted(entries, key=lambda e: (e.num, e.aliases[0])):
        for alias in entry.aliases:
            name = _sanitize_alias(alias)
            if not name:
                continue
            key = f"syscall.{name}"
            if key in emitted:
                continue
            _emit_entry(lines, name, entry.argc, entry.num)
            emitted.add(key)

    return "\n".join(lines).rstrip() + "\n"


def main() -> None:
    output = generate()
    DST.parent.mkdir(parents=True, exist_ok=True)
    DST.write_text(output, encoding="utf-8")
    print(f"wrote {DST}")


if __name__ == "__main__":
    main()