#!/usr/bin/env python3
#
# Copyright (c) 2017 Intel Corporation
#
# SPDX-License-Identifier: Apache-2.0

"""
Script to generate system call invocation macros

This script parses the system call metadata JSON file emitted by
parse_syscalls.py to create several files:

- A file containing weak aliases of any potentially unimplemented system calls,
  as well as the system call dispatch table, which maps system call type IDs
  to their handler functions.

- A header file defining the system call type IDs, as well as function
  prototypes for all system call handler functions.

- A directory containing header files. Each header corresponds to a header
  that was identified as containing system call declarations. These
  generated headers contain the inline invocation functions for each system
  call in that header.
"""

import argparse
import json
import os
import re
import sys

# Some kernel headers cannot include automated tracing without causing unintended recursion or
# other serious issues.
# These headers typically already have very specific tracing hooks for all relevant things
# written by hand so are excluded.
notracing = ["kernel.h", "zephyr/kernel.h", "errno_private.h", "zephyr/errno_private.h"]

types64 = ["int64_t", "uint64_t"]

# The kernel linkage is complicated.  These functions from
# userspace_handlers.c are present in the kernel .a library after
# userspace.c, which contains the weak fallbacks defined here.  So the
# linker finds the weak one first and stops searching, and thus won't
# see the real implementation which should override.  Yet changing the
# order runs afoul of a comment in CMakeLists.txt that the order is
# critical.  These are core syscalls that won't ever be unconfigured,
# just disable the fallback mechanism as a simple workaround.
noweak = ["z_mrsh_k_object_release", "z_mrsh_k_object_access_grant", "z_mrsh_k_object_alloc"]

table_template = """/* auto-generated by gen_syscalls.py, don't edit */

#include <zephyr/llext/symbol.h>

/* Weak handler functions that get replaced by the real ones unless a system
 * call is not implemented due to kernel configuration.
 */
%s

const _k_syscall_handler_t _k_syscall_table[K_SYSCALL_LIMIT] = {
\t%s
};
"""

list_template = """/* auto-generated by gen_syscalls.py, don't edit */

#ifndef ZEPHYR_SYSCALL_LIST_H
#define ZEPHYR_SYSCALL_LIST_H

%s

#ifndef _ASMLANGUAGE

#include <stdarg.h>
#include <stdint.h>

#endif /* _ASMLANGUAGE */

#endif /* ZEPHYR_SYSCALL_LIST_H */
"""

syscall_template = """/* auto-generated by gen_syscalls.py, don't edit */

{include_guard}

{tracing_include}

#ifndef _ASMLANGUAGE

#include <stdarg.h>

#include <zephyr/syscall_list.h>
#include <zephyr/syscall.h>

#include <zephyr/linker/sections.h>


#ifdef __cplusplus
extern "C" {{
#endif

{invocations}

#ifdef __cplusplus
}}
#endif

#endif
#endif /* include guard */
"""

handler_template = """
extern uintptr_t z_hdlr_%s(uintptr_t arg1, uintptr_t arg2, uintptr_t arg3,
                uintptr_t arg4, uintptr_t arg5, uintptr_t arg6, void *ssf);
"""

weak_template = """
__weak ALIAS_OF(handler_no_syscall)
uintptr_t %s(uintptr_t arg1, uintptr_t arg2, uintptr_t arg3,
         uintptr_t arg4, uintptr_t arg5, uintptr_t arg6, void *ssf);
"""

# defines a macro wrapper which supersedes the syscall when used
# and provides tracing enter/exit hooks while allowing per compilation unit
# enable/disable of syscall tracing. Used for returning functions
# Note that the last argument to the exit macro is the return value.
syscall_tracer_with_return_template = """
#if defined(CONFIG_TRACING_SYSCALL)
#ifndef DISABLE_SYSCALL_TRACING
{trace_diagnostic}
#define {func_name}({argnames}) ({{ \
	{func_type} syscall__retval; \
	sys_port_trace_syscall_enter({syscall_id}, {func_name}{trace_argnames}); \
	syscall__retval = {func_name}({argnames}); \
	sys_port_trace_syscall_exit({syscall_id}, {func_name}{trace_argnames}, syscall__retval); \
	syscall__retval; \
}})
#endif
#endif
"""

# defines a macro wrapper which supersedes the syscall when used
# and provides tracing enter/exit hooks while allowing per compilation unit
# enable/disable of syscall tracing. Used for non-returning (void) functions
syscall_tracer_void_template = """
#if defined(CONFIG_TRACING_SYSCALL)
#ifndef DISABLE_SYSCALL_TRACING
{trace_diagnostic}
#define {func_name}({argnames}) do {{ \
	sys_port_trace_syscall_enter({syscall_id}, {func_name}{trace_argnames}); \
	{func_name}({argnames}); \
	sys_port_trace_syscall_exit({syscall_id}, {func_name}{trace_argnames}); \
}} while(false)
#endif
#endif
"""


llext_weakdefs_template = """/* auto-generated by gen_syscalls.py, don't edit */

#include <zephyr/toolchain.h>
#include <zephyr/llext/symbol.h>

/*
 * This symbol is placed at address 0 by llext-sections.ld. Its value and
 * type is not important, we are only interested in its location.
 */
static void * const no_syscall_impl Z_GENERIC_SECTION(llext_no_syscall_impl);

/*
 * Weak references to all syscall implementations. Those not found by the
 * linker outside this file will be exported as NULL and simply fail when
 * an extension requiring them is loaded.
 */
%s
"""


llext_exports_template = """/* auto-generated by gen_syscalls.py, don't edit */

/*
 * Export the implementation functions of all emitted syscalls.
 * Only the symbol names are relevant in this file, they will be
 * resolved to the actual implementation functions by the linker.
 */

/* Symbol declarations */
%s

/* Exported symbols */
%s
"""

typename_regex = re.compile(r'(.*?)([A-Za-z0-9_]+)$')


class SyscallParseException(Exception):
    pass


def typename_split(item):
    item = item.strip().replace("\n", " ")
    if "[" in item:
        raise SyscallParseException(
            f"Please pass arrays to syscalls as pointers, unable to process '{item}'"
        )

    if "(" in item:
        raise SyscallParseException("Please use typedefs for function pointers")

    mo = typename_regex.match(item)
    if not mo:
        raise SyscallParseException("Malformed system call invocation")

    m = mo.groups()
    return (m[0].strip(), m[1])


def need_split(argtype):
    return (not args.long_registers) and (argtype in types64)


# Note: "lo" and "hi" are named in little endian conventions,
# but it doesn't matter as long as they are consistently
# generated.
def union_decl(ctype, split):
    middle = "struct { uintptr_t lo, hi; } split" if split else "uintptr_t x"
    return f"union {{ {middle}; {ctype} val; }}"


def wrapper_defs(func_name, func_type, args, fn, userspace_only):
    ret64 = need_split(func_type)
    mrsh_args = []  # List of rvalue expressions for the marshalled invocation

    decl_arglist = ", ".join([" ".join(argrec) for argrec in args]) or "void"
    syscall_id = "K_SYSCALL_" + func_name.upper()

    wrap = ''
    if not userspace_only:
        wrap += f"extern {func_type} z_impl_{func_name}({decl_arglist});\n"
        wrap += "\n"

    wrap += "__pinned_func\n"
    wrap += f"static inline {func_type} {func_name}({decl_arglist})\n"
    wrap += "{\n"
    if not userspace_only:
        wrap += "#ifdef CONFIG_USERSPACE\n"

    wrap += ("\t" + "uint64_t ret64;\n") if ret64 else ""
    if not userspace_only:
        wrap += "\t" + "if (z_syscall_trap()) {\n"

    valist_args = []
    for argnum, (argtype, argname) in enumerate(args):
        split = need_split(argtype)
        decl = union_decl(argtype, split)
        wrap += f"\t\t{decl} parm{argnum}"
        if argtype != "va_list":
            wrap += f" = {{ .val = {argname} }};\n"
        else:
            # va_list objects are ... peculiar.
            wrap += ";\n" + "\t\t" + f"va_copy(parm{argnum}.val, {argname});\n"
            valist_args.append(f"parm{argnum}.val")
        if split:
            mrsh_args.append(f"parm{argnum}.split.lo")
            mrsh_args.append(f"parm{argnum}.split.hi")
        else:
            mrsh_args.append(f"parm{argnum}.x")

    if ret64:
        mrsh_args.append("(uintptr_t)&ret64")

    if len(mrsh_args) > 6:
        wrap += "\t\t" + "uintptr_t more[] = {\n"
        wrap += "\t\t\t" + (",\n\t\t\t".join(mrsh_args[5:])) + "\n"
        wrap += "\t\t" + "};\n"
        mrsh_args[5:] = ["(uintptr_t) &more"]

    invoke_args = ", ".join(mrsh_args + [syscall_id])
    invoke = f"arch_syscall_invoke{len(mrsh_args)}({invoke_args})"

    if ret64:
        invoke = "\t\t" + f"(void) {invoke};\n"
        retcode = "\t\t" + f"return ({func_type}) ret64;\n"
    elif func_type == "void":
        invoke = "\t\t" + f"(void) {invoke};\n"
        retcode = "\t\t" + "return;\n"
    elif valist_args:
        invoke = "\t\t" + f"{func_type} invoke__retval = {invoke};\n"
        retcode = "\t\t" + "return invoke__retval;\n"
    else:
        invoke = "\t\t" + f"return ({func_type}) {invoke};\n"
        retcode = ""

    wrap += invoke
    for argname in valist_args:
        wrap += "\t\t" + f"va_end({argname});\n"
    wrap += retcode
    if not userspace_only:
        wrap += "\t" + "}\n"
        wrap += "#endif\n"

        # Otherwise fall through to direct invocation of the impl func.
        # Note the compiler barrier: that is required to prevent code from
        # the impl call from being hoisted above the check for user
        # context.
        impl_arglist = ", ".join([argrec[1] for argrec in args])
        impl_call = f"z_impl_{func_name}({impl_arglist})"
        wrap += "\t" + "compiler_barrier();\n"
        ret = "return " if func_type != "void" else ""
        wrap += "\t" + f"{ret}{impl_call};\n"

    wrap += "}\n"

    if fn not in notracing:
        argnames = ", ".join([f"{argname}" for _, argname in args])
        trace_argnames = ""
        if len(args) > 0:
            trace_argnames = ", " + argnames
        trace_diagnostic = ""
        if os.getenv('TRACE_DIAGNOSTICS'):
            trace_diagnostic = f"#warning Tracing {func_name}"
        if func_type != "void":
            wrap += syscall_tracer_with_return_template.format(
                func_type=func_type,
                func_name=func_name,
                argnames=argnames,
                trace_argnames=trace_argnames,
                syscall_id=syscall_id,
                trace_diagnostic=trace_diagnostic,
            )
        else:
            wrap += syscall_tracer_void_template.format(
                func_type=func_type,
                func_name=func_name,
                argnames=argnames,
                trace_argnames=trace_argnames,
                syscall_id=syscall_id,
                trace_diagnostic=trace_diagnostic,
            )

    return wrap


# Returns an expression for the specified (zero-indexed!) marshalled
# parameter to a syscall, with handling for a final "more" parameter.
def mrsh_rval(mrsh_num, total):
    if mrsh_num < 5 or total <= 6:
        return f"arg{mrsh_num}"
    else:
        return f"(((uintptr_t *)more)[{mrsh_num - 5}])"


def marshall_defs(func_name, func_type, args):
    mrsh_name = "z_mrsh_" + func_name

    nmrsh = 0  # number of marshalled uintptr_t parameter
    vrfy_parms = []  # list of (argtype, bool_is_split)
    for argtype, _ in args:
        split = need_split(argtype)
        vrfy_parms.append((argtype, split))
        nmrsh += 2 if split else 1

    # Final argument for a 64 bit return value?
    if need_split(func_type):
        nmrsh += 1

    decl_arglist = ", ".join([" ".join(argrec) for argrec in args])
    mrsh = f"extern {func_type} z_vrfy_{func_name}({decl_arglist});\n"

    mrsh += f"uintptr_t {mrsh_name}(uintptr_t arg0, uintptr_t arg1, uintptr_t arg2,\n"
    if nmrsh <= 6:
        mrsh += "\t\t" + "uintptr_t arg3, uintptr_t arg4, uintptr_t arg5, void *ssf)\n"
    else:
        mrsh += "\t\t" + "uintptr_t arg3, uintptr_t arg4, void *more, void *ssf)\n"
    mrsh += "{\n"
    mrsh += "\t" + "_current->syscall_frame = ssf;\n"

    for unused_arg in range(nmrsh, 6):
        mrsh += f"\t(void) arg{unused_arg};\t/* unused */\n"

    if nmrsh > 6:
        mrsh += (
            "\tK_OOPS(K_SYSCALL_MEMORY_READ(more, " + str(nmrsh - 5) + " * sizeof(uintptr_t)));\n"
        )

    argnum = 0
    for i, (argtype, split) in enumerate(vrfy_parms):
        decl = union_decl(argtype, split)
        mrsh += f"\t{decl} parm{i};\n"
        rval = mrsh_rval(argnum, nmrsh)
        if split:
            mrsh += "\t" + f"parm{i}.split.lo = {rval};\n"
            argnum += 1
            rval = mrsh_rval(argnum, nmrsh)
            mrsh += "\t" + f"parm{i}.split.hi = {rval};\n"
        else:
            mrsh += "\t" + f"parm{i}.x = {rval};\n"
        argnum += 1

    # Finally, invoke the verify function
    out_args = ", ".join([f"parm{i}.val" for i in range(len(args))])
    vrfy_call = f"z_vrfy_{func_name}({out_args})"

    if func_type == "void":
        mrsh += "\t" + f"{vrfy_call};\n"
        mrsh += "\t" + "_current->syscall_frame = NULL;\n"
        mrsh += "\t" + "return 0;\n"
    else:
        mrsh += "\t" + f"{func_type} ret = {vrfy_call};\n"

        if need_split(func_type):
            rval = mrsh_rval(nmrsh - 1, nmrsh)
            ptr = f"((uint64_t *){rval})"
            mrsh += "\t" + f"K_OOPS(K_SYSCALL_MEMORY_WRITE({ptr}, 8));\n"
            mrsh += "\t" + f"*{ptr} = ret;\n"
            mrsh += "\t" + "_current->syscall_frame = NULL;\n"
            mrsh += "\t" + "return 0;\n"
        else:
            mrsh += "\t" + "_current->syscall_frame = NULL;\n"
            mrsh += "\t" + "return (uintptr_t) ret;\n"

    mrsh += "}\n"

    return mrsh, mrsh_name


def analyze_fn(match_group, fn, userspace_only):
    func, args = match_group

    try:
        if args == "void":
            args = []
        else:
            args = [typename_split(a) for a in args.split(",")]

        func_type, func_name = typename_split(func)
    except SyscallParseException:
        sys.stderr.write(f"In declaration of {func}\n")
        raise

    sys_id = "K_SYSCALL_" + func_name.upper()

    marshaller = None
    marshaller, handler = marshall_defs(func_name, func_type, args)
    invocation = wrapper_defs(func_name, func_type, args, fn, userspace_only)

    # Entry in _k_syscall_table
    table_entry = f"[{sys_id}] = {handler}"

    return (handler, invocation, marshaller, sys_id, table_entry)


def parse_args():
    global args
    parser = argparse.ArgumentParser(
        description=__doc__,
        formatter_class=argparse.RawDescriptionHelpFormatter,
        allow_abbrev=False,
    )

    parser.add_argument(
        "-i", "--json-file", required=True, help="Read syscall information from json file"
    )
    parser.add_argument(
        "-d", "--syscall-dispatch", required=True, help="output C system call dispatch table file"
    )
    parser.add_argument(
        "-l", "--syscall-list", required=True, help="output C system call list header"
    )
    parser.add_argument(
        "-o", "--base-output", required=True, help="Base output directory for syscall macro headers"
    )
    parser.add_argument(
        "-s",
        "--split-type",
        action="append",
        help="A long type that must be split/marshalled on 32-bit systems",
    )
    parser.add_argument(
        "-x",
        "--long-registers",
        action="store_true",
        help="Indicates we are on system with 64-bit registers",
    )
    parser.add_argument(
        "--gen-mrsh-files", action="store_true", help="Generate marshalling files (*_mrsh.c)"
    )
    parser.add_argument(
        "-e", "--syscall-exports-llext", help="output C system call export for extensions"
    )
    parser.add_argument(
        "-w", "--syscall-weakdefs-llext", help="output C system call weak definitions"
    )
    parser.add_argument(
        "-u",
        "--userspace-only",
        action="store_true",
        help="Only generate the userpace path of wrappers",
    )
    args = parser.parse_args()


def main():
    parse_args()

    if args.split_type is not None:
        for t in args.split_type:
            types64.append(t)

    with open(args.json_file) as fd:
        syscalls = json.load(fd)

    invocations = {}
    mrsh_defs = {}
    mrsh_includes = {}
    ids_emit = []
    ids_not_emit = []
    table_entries = []
    handlers = []
    emit_list = []
    exported = []

    for match_group, fn, to_emit in syscalls:
        handler, inv, mrsh, sys_id, entry = analyze_fn(match_group, fn, args.userspace_only)

        if fn not in invocations:
            invocations[fn] = []

        invocations[fn].append(inv)
        handlers.append(handler)

        if to_emit:
            ids_emit.append(sys_id)
            table_entries.append(entry)
            emit_list.append(handler)
            exported.append(handler.replace("z_mrsh_", "z_impl_"))
        else:
            ids_not_emit.append(sys_id)

        if mrsh and to_emit:
            syscall = typename_split(match_group[0])[1]
            mrsh_defs[syscall] = mrsh
            mrsh_includes[syscall] = f"#include <zephyr/syscalls/{fn}>"

    with open(args.syscall_dispatch, "w") as fp:
        table_entries.append("[K_SYSCALL_BAD] = handler_bad_syscall")

        weak_defines = "".join(
            [weak_template % name for name in handlers if name not in noweak and name in emit_list]
        )

        # The "noweak" ones just get a regular declaration
        noweak_args = ", ".join([f"uintptr_t arg{i + 1}" for i in range(6)]) + ", void *ssf"
        weak_defines += "\n".join([f"extern uintptr_t {s}({noweak_args});" for s in noweak])

        fp.write(table_template % (weak_defines, ",\n\t".join(table_entries)))

    exported.sort()

    if args.syscall_weakdefs_llext:
        with open(args.syscall_weakdefs_llext, "w") as fp:
            # Provide weak definitions for all emitted syscalls
            weak_refs = "\n".join(
                f"extern __weak ALIAS_OF(no_syscall_impl) void * const {e};" for e in exported
            )
            fp.write(llext_weakdefs_template % weak_refs)

    if args.syscall_exports_llext:
        with open(args.syscall_exports_llext, "w") as fp:
            # Export symbols for emitted syscalls
            extern_refs = "\n".join(f"extern void * const {e};" for e in exported)
            exported_symbols = "\n".join(f"EXPORT_GROUP_SYMBOL(SYSCALL, {e});" for e in exported)
            fp.write(llext_exports_template % (extern_refs, exported_symbols))

    # Listing header emitted to stdout
    ids_emit.sort()
    ids_emit.extend(["K_SYSCALL_BAD", "K_SYSCALL_LIMIT"])

    ids_as_defines = ""
    for i, item in enumerate(ids_emit):
        ids_as_defines += f"#define {item} {i}\n"

    if ids_not_emit:
        # There are syscalls that are not used in the image but
        # their IDs are used in the generated stubs. So need to
        # make them usable but outside the syscall ID range.
        ids_as_defines += "\n\n/* Following syscalls are not used in image */\n"
        ids_not_emit.sort()
        num_emitted_ids = len(ids_emit)
        for i, item in enumerate(ids_not_emit):
            ids_as_defines += f"#define {item} {i + num_emitted_ids}\n"

    with open(args.syscall_list, "w") as fp:
        fp.write(list_template % ids_as_defines)

    os.makedirs(args.base_output, exist_ok=True)
    for fn, invo_list in invocations.items():
        out_fn = os.path.join(args.base_output, fn)

        ig = re.sub("[^a-zA-Z0-9]", "_", "Z_INCLUDE_SYSCALLS_" + fn).upper()
        include_guard = f"#ifndef {ig}\n#define {ig}\n"
        tracing_include = ""
        if fn not in notracing:
            tracing_include = "#include <zephyr/tracing/tracing_syscall.h>"
        header = syscall_template.format(
            include_guard=include_guard,
            tracing_include=tracing_include,
            invocations="\n\n".join(invo_list),
        )

        with open(out_fn, "w") as fp:
            fp.write(header)

    # Likewise emit _mrsh.c files for syscall inclusion
    if args.gen_mrsh_files:
        for fn in mrsh_defs:
            mrsh_fn = os.path.join(args.base_output, fn + "_mrsh.c")

            with open(mrsh_fn, "w") as fp:
                fp.write("/* auto-generated by gen_syscalls.py, don't edit */\n\n")
                fp.write(mrsh_includes[fn] + "\n")
                fp.write("\n")
                fp.write(mrsh_defs[fn] + "\n")


if __name__ == "__main__":
    main()