run-tests.py

#! /usr/bin/env python3

import os
import subprocess
import sys
import sysconfig
import platform
import argparse
import json
import re
from glob import glob
import multiprocessing
from multiprocessing.pool import ThreadPool
import threading
import tempfile

from test_utils import (
    base_path,
    pyboard,
    TEST_TIMEOUT,
    TEST_MAXIMUM_RAW_REPL_FAILURES,
    MPYCROSS,
    test_instance_description,
    test_instance_epilog,
    test_directory_description,
    rm_f,
    normalize_newlines,
    set_injected_prologue,
    get_results_filename,
    convert_device_shortcut_to_real_device,
    get_test_instance,
    prepare_script_for_target,
    create_test_report,
    FLAKY_REASON_PREFIX,
)

RV32_ARCH_FLAGS = {
    "zba": 1 << 0,
    "zcmp": 1 << 1,
}

# Tests require at least CPython 3.3. If your default python3 executable
# is of lower version, you can point MICROPY_CPYTHON3 environment var
# to the correct executable.
if os.name == "nt":
    CPYTHON3 = os.getenv("MICROPY_CPYTHON3", "python")
    MICROPYTHON = os.getenv(
        "MICROPY_MICROPYTHON", base_path("../ports/windows/build-standard/micropython.exe")
    )
else:
    CPYTHON3 = os.getenv("MICROPY_CPYTHON3", "python3")
    MICROPYTHON = os.getenv(
        "MICROPY_MICROPYTHON", base_path("../ports/unix/build-standard/micropython")
    )

# Use CPython options to not save .pyc files, to only access the core standard library
# (not site packages which may clash with u-module names), and improve start up time.
CPYTHON3_CMD = [CPYTHON3, "-BS"]

# For diff'ing test output
DIFF = os.getenv("MICROPY_DIFF", "diff -u")

# Set PYTHONIOENCODING so that CPython will use utf-8 on systems which set another encoding in the locale
os.environ["PYTHONIOENCODING"] = "utf-8"

# Platforms associated with the unix port, values of `sys.platform`.
PC_PLATFORMS = ("darwin", "linux", "win32")

# Mapping from `sys.platform` to the port name, for special cases.
# See `platform_to_port()` function.
platform_to_port_map = {"pyboard": "stm32", "WiPy": "cc3200"}
platform_to_port_map.update({p: "unix" for p in PC_PLATFORMS})

# Tests to skip for values of the `--via-mpy` argument.
via_mpy_tests_to_skip = {
    # Skip the following when mpy is enabled.
    True: (
        # These print out the filename and that's expected to match the .py name.
        "import/import_file.py",
        "io/argv.py",
        "misc/sys_settrace_features.py",
        "misc/sys_settrace_generator.py",
        "misc/sys_settrace_loop.py",
    ),
}

# Tests to skip for specific emitters.
emitter_tests_to_skip = {
    # Some tests are known to fail with native emitter.
    # Remove them from the below when they work.
    "native": (
        # These require raise_varargs.
        "basics/gen_yield_from_close.py",
        "basics/try_finally_return2.py",
        "basics/try_reraise.py",
        "basics/try_reraise2.py",
        "misc/features.py",
        # These require checking for unbound local.
        "basics/annotate_var.py",
        "basics/del_deref.py",
        "basics/del_local.py",
        "basics/scope_implicit.py",
        "basics/unboundlocal.py",
        # These require "raise from".
        "basics/exception_chain.py",
        # These require stack-allocated slice optimisation.
        "micropython/heapalloc_slice.py",
        # These require implicitly running the scheduler between bytecodes.
        "micropython/schedule.py",
        # These require sys.exc_info().
        "misc/sys_exc_info.py",
        # These require sys.settrace().
        "misc/sys_settrace_cov.py",
        "misc/sys_settrace_features.py",
        "misc/sys_settrace_generator.py",
        "misc/sys_settrace_loop.py",
        # These are bytecode-specific tests.
        "stress/bytecode_limit.py",
    ),
}

# Tests to skip on specific targets.
# These are tests that are difficult to detect that they should not be run on the given target.
platform_tests_to_skip = {
    "esp8266": (
        "stress/list_sort.py",  # watchdog kicks in because it takes too long
    ),
    "minimal": (
        "basics/class_inplace_op.py",  # all special methods not supported
        "basics/subclass_native_init.py",  # native subclassing corner cases not support
        "micropython/opt_level.py",  # don't assume line numbers are stored
    ),
    "nrf": (
        "basics/io_buffered_writer.py",
        "basics/io_bytesio_cow.py",
        "basics/io_bytesio_ext.py",
        "basics/io_bytesio_ext2.py",
        "basics/io_iobase.py",
        "basics/io_stringio1.py",
        "basics/io_stringio_base.py",
        "basics/io_stringio_with.py",
        "basics/io_write_ext.py",
        "basics/memoryview1.py",  # no item assignment for memoryview
        "extmod/random_basic.py",  # unimplemented: random.seed
        "micropython/opt_level.py",  # no support for line numbers
        "misc/non_compliant.py",  # no item assignment for bytearray
    ),
    "renesas-ra": (
        "extmod/time_time_ns.py",  # RA fsp rtc function doesn't support nano sec info
    ),
    "rp2": (
        # Skip thread tests that require more that 2 threads.
        "thread/stress_heap.py",
        "thread/thread_lock2.py",
        "thread/thread_lock3.py",
        "thread/thread_shared2.py",
    ),
    "webassembly": (
        "basics/string_format_modulo.py",  # can't print nulls to stdout
        "basics/string_strip.py",  # can't print nulls to stdout
        "basics/weakref_callback_exception.py",  # has different exception printing output
        "basics/weakref_ref_collect.py",  # requires custom test due to GC behaviour
        "basics/weakref_finalize_collect.py",  # requires custom test due to GC behaviour
        "extmod/asyncio_basic2.py",
        "extmod/asyncio_cancel_self.py",
        "extmod/asyncio_current_task.py",
        "extmod/asyncio_exception.py",
        "extmod/asyncio_gather_finished_early.py",
        "extmod/asyncio_get_event_loop.py",
        "extmod/asyncio_heaplock.py",
        "extmod/asyncio_loop_stop.py",
        "extmod/asyncio_new_event_loop.py",
        "extmod/asyncio_threadsafeflag.py",
        "extmod/asyncio_wait_for_fwd.py",
        "extmod/asyncio_event_queue.py",
        "extmod/asyncio_iterator_event.py",
        "extmod/asyncio_wait_for_linked_task.py",
        "extmod/binascii_a2b_base64.py",
        "extmod/deflate_compress_memory_error.py",  # tries to allocate unlimited memory
        "extmod/re_stack_overflow.py",
        "extmod/re_stack_overflow2.py",
        "extmod/time_res.py",
        "extmod/vfs_posix.py",
        "extmod/vfs_posix_enoent.py",
        "extmod/vfs_posix_paths.py",
        "extmod/vfs_userfs.py",
        "micropython/emg_exc.py",
        "micropython/extreme_exc.py",
        "micropython/heapalloc_exc_compressed_emg_exc.py",
    ),
    "WiPy": (
        "misc/print_exception.py",  # requires error reporting full
    ),
    "zephyr": (
        # Skip thread tests that require more than 4 threads.
        "thread/stress_heap.py",
        "thread/thread_lock3.py",
    ),
}

# Tests with known intermittent failures. These tests still run, but failures
# are reclassified as "ignored" instead of "fail" so they don't affect the CI
# exit code. Paths are relative to the tests/ directory (must match test_file
# format used by run_one_test, which normalises backslashes to forward slashes).
#
# Values are (reason, platforms) tuples where platforms is None (all platforms)
# or a tuple of sys.platform strings to restrict ignoring to those platforms.
flaky_tests_to_ignore = {
    "thread/thread_gc1.py": ("GC race condition", None),
    "thread/stress_schedule.py": ("intermittent crash under QEMU", None),
    "thread/stress_recurse.py": ("stack overflow under emulation", None),
    "thread/stress_heap.py": ("flaky on macOS", ("darwin",)),
    "cmdline/repl_lock.py": ("REPL timing under QEMU", None),
    "cmdline/repl_cont.py": ("REPL escaping on macOS", ("darwin",)),
    "extmod/time_time_ns.py": ("CI runner clock precision", None),
}

# These tests don't test float explicitly but rather use it to perform the test.
tests_requiring_float = (
    "extmod/asyncio_basic.py",
    "extmod/asyncio_basic2.py",
    "extmod/asyncio_cancel_task.py",
    "extmod/asyncio_event.py",
    "extmod/asyncio_fair.py",
    "extmod/asyncio_gather.py",
    "extmod/asyncio_gather_notimpl.py",
    "extmod/asyncio_get_event_loop.py",
    "extmod/asyncio_iterator_event.py",
    "extmod/asyncio_lock.py",
    "extmod/asyncio_task_done.py",
    "extmod/asyncio_wait_for.py",
    "extmod/asyncio_wait_for_fwd.py",
    "extmod/asyncio_wait_for_linked_task.py",
    "extmod/asyncio_wait_task.py",
    "extmod/json_dumps_float.py",
    "extmod/json_loads_float.py",
    "extmod/random_extra_float.py",
    "extmod/select_poll_eintr.py",
    "extmod/tls_threads.py",
    "extmod/uctypes_le_float.py",
    "extmod/uctypes_native_float.py",
    "extmod/uctypes_sizeof_float.py",
    "misc/rge_sm.py",
    "ports/unix/ffi_float.py",
    "ports/unix/ffi_float2.py",
)

# These tests don't test slice explicitly but rather use it to perform the test.
tests_requiring_slice = (
    "basics/builtin_range.py",
    "basics/bytearray1.py",
    "basics/class_super.py",
    "basics/containment.py",
    "basics/errno1.py",
    "basics/fun_str.py",
    "basics/generator1.py",
    "basics/globals_del.py",
    "basics/memoryview1.py",
    "basics/memoryview_gc.py",
    "basics/object1.py",
    "basics/python34.py",
    "basics/struct_endian.py",
    "extmod/btree1.py",
    "extmod/deflate_decompress.py",
    "extmod/framebuf16.py",
    "extmod/framebuf4.py",
    "extmod/machine1.py",
    "extmod/time_mktime.py",
    "extmod/time_res.py",
    "extmod/tls_sslcontext_ciphers.py",
    "extmod/vfs_fat_fileio1.py",
    "extmod/vfs_fat_finaliser.py",
    "extmod/vfs_fat_more.py",
    "extmod/vfs_fat_ramdisk.py",
    "extmod/vfs_fat_ramdisklarge.py",
    "extmod/vfs_lfs.py",
    "extmod/vfs_rom.py",
    "float/string_format_modulo.py",
    "micropython/builtin_execfile.py",
    "micropython/extreme_exc.py",
    "micropython/heapalloc_fail_bytearray.py",
    "micropython/heapalloc_fail_list.py",
    "micropython/heapalloc_fail_memoryview.py",
    "micropython/import_mpy_invalid.py",
    "micropython/import_mpy_native.py",
    "micropython/import_mpy_native_gc.py",
    "misc/non_compliant.py",
    "misc/rge_sm.py",
)

# Tests that require `import target_wiring` to work.
tests_requiring_target_wiring = (
    "extmod/machine_spi_rate.py",
    "extmod/machine_uart_irq_txidle.py",
    "extmod/machine_uart_tx.py",
    "extmod_hardware/machine_can_timings.py",
    "extmod_hardware/machine_encoder.py",
    "extmod_hardware/machine_pwm.py",
    "extmod_hardware/machine_uart_irq_break.py",
    "extmod_hardware/machine_uart_irq_rx.py",
    "extmod_hardware/machine_uart_irq_rxidle.py",
)


# unescape wanted regex chars and escape unwanted ones
def convert_regex_escapes(line):
    cs = []
    escape = False
    for c in str(line, "utf8"):
        if escape:
            escape = False
            cs.append(c)
        elif c == "\\":
            escape = True
        elif c in ("(", ")", "[", "]", "{", "}", ".", "*", "+", "^", "$"):
            cs.append("\\" + c)
        else:
            cs.append(c)
    # accept carriage-return(s) before final newline
    if cs[-1] == "\n":
        cs[-1] = "\r*\n"
    return bytes("".join(cs), "utf8")


def platform_to_port(platform):
    return platform_to_port_map.get(platform, platform)


def detect_inline_asm_arch(pyb, args):
    for arch in ("rv32", "thumb", "xtensa"):
        output = run_feature_check(pyb, args, "inlineasm_{}.py".format(arch))
        if output.strip() == arch.encode():
            return arch
    return None


def map_rv32_arch_flags(flags):
    mapped_flags = []
    for extension, bit in RV32_ARCH_FLAGS.items():
        if flags & bit:
            mapped_flags.append(extension)
        flags &= ~bit
    if flags:
        raise Exception("Unexpected flag bits set in value {}".format(flags))
    return mapped_flags


def detect_test_platform(pyb, args):
    # Run a script to detect various bits of information about the target test instance.
    output = run_feature_check(pyb, args, "target_info.py")
    if output.endswith(b"CRASH"):
        raise ValueError("cannot detect platform: {}".format(output))
    platform, arch, arch_flags, build, thread, float_prec, unicode = (
        str(output, "ascii").strip().split()
    )
    if arch == "None":
        arch = None
    inlineasm_arch = detect_inline_asm_arch(pyb, args)
    if thread == "None":
        thread = None
    float_prec = int(float_prec)
    unicode = unicode == "True"
    if arch == "rv32imc":
        arch_flags = map_rv32_arch_flags(int(arch_flags))
    else:
        arch_flags = None

    args.platform = platform
    args.arch = arch
    args.arch_flags = arch_flags
    if arch and not args.mpy_cross_flags:
        args.mpy_cross_flags = "-march=" + arch
        if arch_flags:
            args.mpy_cross_flags += " -march-flags=" + ",".join(arch_flags)
    args.inlineasm_arch = inlineasm_arch
    args.build = build
    args.thread = thread
    args.float_prec = float_prec
    args.unicode = unicode

    # Print the detected information about the target.
    print("platform={}".format(platform), end="")
    if arch:
        print(" arch={}".format(arch), end="")
        if arch_flags:
            print(" arch_flags={}".format(",".join(arch_flags)), end="")
    if inlineasm_arch:
        print(" inlineasm={}".format(inlineasm_arch), end="")
    if thread:
        print(" thread={}".format(thread), end="")
    if float_prec:
        print(" float={}-bit".format(float_prec), end="")
    if unicode:
        print(" unicode", end="")


def detect_target_wiring_script(pyb, args):
    tw_data = b""
    tw_source = None
    if args.target_wiring:
        # A target_wiring path is explicitly provided, so use that.
        tw_source = args.target_wiring
        with open(tw_source, "rb") as f:
            tw_data = f.read()
    elif hasattr(pyb, "exec_raw") and pyb.exec_raw("import target_wiring") == (b"", b""):
        # The board already has a target_wiring module available, so use that.
        tw_source = "on-device"
    else:
        port = platform_to_port(args.platform)
        build = args.build
        tw_board_exact = None
        tw_board_partial = None
        tw_port = None
        for file in os.listdir("target_wiring"):
            file_base = file.removesuffix(".py")
            if file_base == build:
                # A file matching the target's board/build name.
                tw_board_exact = file
            elif file_base.endswith("x") and build.startswith(file_base.removesuffix("x")):
                # A file with a partial match to the target's board/build name.
                tw_board_partial = file
            elif file_base == port:
                # A file matching the target's port.
                tw_port = file
        tw_source = tw_board_exact or tw_board_partial or tw_port
        if tw_source:
            with open("target_wiring/" + tw_source, "rb") as f:
                tw_data = f.read()
    if tw_source:
        print(" target_wiring={}".format(tw_source), end="")
    pyb.target_wiring_script = tw_data


tests_with_regex_output = [
    base_path(file)
    for file in (
        "micropython/meminfo.py",
        "basics/bytes_compare3.py",
        "basics/builtin_help.py",
        "basics/weakref_callback_exception.py",
        "misc/sys_settrace_cov.py",
        "net_inet/tls_text_errors.py",
        "thread/thread_exc2.py",
        "ports/esp32/partition_ota.py",
    )
]


def run_micropython(pyb, args, test_file, test_file_abspath, is_special=False):
    had_crash = False
    if pyb is None:
        # run on PC
        if test_file_abspath.startswith((base_path("cmdline/"), base_path("feature_check/"))):
            # special handling for tests of the unix cmdline program
            is_special = True

        if is_special:
            # check for any cmdline options needed for this test
            args = [MICROPYTHON]
            with open(test_file, "rb") as f:
                line = f.readline()
                if line.startswith(b"# cmdline:"):
                    # subprocess.check_output on Windows only accepts strings, not bytes
                    args += [str(c, "utf-8") for c in line[10:].strip().split()]

            # run the test, possibly with redirected input
            try:
                if os.path.basename(test_file).startswith("repl_"):
                    # Need to use a PTY to test command line editing
                    try:
                        import pty
                    except ImportError:
                        # in case pty module is not available, like on Windows
                        return b"SKIP\n"
                    import select

                    # Even though these might have the pty module, it's unlikely to function.
                    if sys.platform in ["win32", "msys", "cygwin"]:
                        return b"SKIP\n"

                    def get(required=False):
                        rv = b""
                        while True:
                            ready = select.select([master], [], [], 0.02)
                            if ready[0] == [master]:
                                rv += os.read(master, 1024)
                            else:
                                if not required or rv:
                                    return rv

                    def send_get(what):
                        # Detect {\x00} pattern and convert to ctrl-key codes.
                        ctrl_code = lambda m: bytes([int(m.group(1))])
                        what = re.sub(rb"{\\x(\d\d)}", ctrl_code, what)

                        os.write(master, what)
                        return get()

                    with open(test_file, "rb") as f:
                        # instead of: output_mupy = subprocess.check_output(args, stdin=f)
                        master, slave = pty.openpty()
                        p = subprocess.Popen(
                            args, stdin=slave, stdout=slave, stderr=subprocess.STDOUT, bufsize=0
                        )
                        banner = get(True)
                        output_mupy = banner + b"".join(send_get(line) for line in f)
                        send_get(b"\x04")  # exit the REPL, so coverage info is saved
                        # At this point the process might have exited already, but trying to
                        # kill it 'again' normally doesn't result in exceptions as Python and/or
                        # the OS seem to try to handle this nicely. When running Linux on WSL
                        # though, the situation differs and calling Popen.kill after the process
                        # terminated results in a ProcessLookupError. Just catch that one here
                        # since we just want the process to be gone and that's the case.
                        try:
                            p.kill()
                        except ProcessLookupError:
                            pass
                        os.close(master)
                        os.close(slave)
                else:
                    output_mupy = subprocess.check_output(
                        args + [test_file], stderr=subprocess.STDOUT
                    )
            except subprocess.CalledProcessError:
                return b"CRASH"

        else:
            # a standard test run on PC

            # create system command
            cmdlist = [os.path.abspath(MICROPYTHON), "-X", "emit=" + args.emit]
            if args.heapsize is not None:
                cmdlist.extend(["-X", "heapsize=" + args.heapsize])
            if sys.platform == "darwin":
                cmdlist.extend(["-X", "realtime"])

            cwd = os.path.dirname(test_file)

            # if running via .mpy, first compile the .py file
            if args.via_mpy:
                mpy_filename = tempfile.mktemp(dir=cwd, suffix=".mpy")
                subprocess.check_output(
                    [MPYCROSS]
                    + args.mpy_cross_flags.split()
                    + ["-o", mpy_filename, "-X", "emit=" + args.emit, test_file]
                )
                mpy_modname = os.path.splitext(os.path.basename(mpy_filename))[0]
                cmdlist.extend(["-m", mpy_modname])
            else:
                cmdlist.append(test_file_abspath)

            # run the actual test
            try:
                output_mupy = subprocess.check_output(
                    cmdlist, stderr=subprocess.STDOUT, timeout=TEST_TIMEOUT, cwd=cwd
                )
            except subprocess.CalledProcessError as er:
                had_crash = True
                output_mupy = er.output + b"CRASH"
            except subprocess.TimeoutExpired as er:
                had_crash = True
                output_mupy = (er.output or b"") + b"TIMEOUT"

            # clean up if we had an intermediate .mpy file
            if args.via_mpy:
                rm_f(mpy_filename)

    else:
        # run via pyboard interface
        requires_target_wiring = test_file.endswith(tests_requiring_target_wiring)
        had_crash, output_mupy = pyb.run_script_on_remote_target(
            args, test_file_abspath, is_special, requires_target_wiring
        )

    # canonical form for all ports/platforms is to use \n for end-of-line
    output_mupy = normalize_newlines(output_mupy)

    # don't try to convert the output if we should skip this test
    if had_crash or output_mupy in (b"SKIP\n", b"SKIP-TOO-LARGE\n", b"CRASH"):
        return output_mupy

    # skipped special tests will output "SKIP" surrounded by other interpreter debug output
    if is_special and not had_crash and b"\nSKIP\n" in output_mupy:
        return b"SKIP\n"

    if is_special or test_file_abspath in tests_with_regex_output:
        # convert parts of the output that are not stable across runs
        with open(test_file + ".exp", "rb") as f:
            lines_exp = []
            for line in f.readlines():
                if line == b"########\n":
                    line = (line,)
                else:
                    line = (line, re.compile(convert_regex_escapes(line)))
                lines_exp.append(line)
        lines_mupy = [line + b"\n" for line in output_mupy.split(b"\n")]
        if output_mupy.endswith(b"\n"):
            lines_mupy = lines_mupy[:-1]  # remove erroneous last empty line
        i_mupy = 0
        for i in range(len(lines_exp)):
            if lines_exp[i][0] == b"########\n":
                # 8x #'s means match 0 or more whole lines
                line_exp = lines_exp[i + 1]
                skip = 0
                while i_mupy + skip < len(lines_mupy) and not line_exp[1].match(
                    lines_mupy[i_mupy + skip]
                ):
                    skip += 1
                if i_mupy + skip >= len(lines_mupy):
                    lines_mupy[i_mupy] = b"######## FAIL\n"
                    break
                del lines_mupy[i_mupy : i_mupy + skip]
                lines_mupy.insert(i_mupy, b"########\n")
                i_mupy += 1
            else:
                # a regex
                if lines_exp[i][1].match(lines_mupy[i_mupy]):
                    lines_mupy[i_mupy] = lines_exp[i][0]
                else:
                    # print("don't match: %r %s" % (lines_exp[i][1], lines_mupy[i_mupy])) # DEBUG
                    pass
                i_mupy += 1
            if i_mupy >= len(lines_mupy):
                break
        output_mupy = b"".join(lines_mupy)

    return output_mupy


def run_feature_check(pyb, args, test_file):
    if pyb is not None and test_file.startswith("repl_"):
        # REPL feature tests will not run via pyboard because they require prompt interactivity
        return b""
    test_file_path = base_path("feature_check", test_file)
    return run_micropython(pyb, args, test_file_path, test_file_path, is_special=True)


class TestError(Exception):
    pass


class ThreadSafeCounter:
    def __init__(self, start=0):
        self._value = start
        self._lock = threading.Lock()

    def increment(self):
        self.add(1)

    def add(self, to_add):
        with self._lock:
            self._value += to_add

    def append(self, arg):
        self.add([arg])

    @property
    def value(self):
        return self._value


def run_tests(pyb, tests, args, result_dir, num_threads=1):
    testcase_count = ThreadSafeCounter()
    raw_repl_failure_count = ThreadSafeCounter()
    test_results = ThreadSafeCounter([])

    skip_tests = set()
    skip_native = False
    skip_int_big = False
    skip_int_64 = False
    skip_bytearray = False
    skip_set_type = False
    skip_slice = False
    skip_async = False
    skip_const = False
    skip_revops = False
    skip_fstring = False
    skip_tstring = False
    skip_endian = False
    skip_inlineasm = False
    has_complex = True
    has_coverage = False

    if True:
        # Even if we run completely different tests in a different directory,
        # we need to access feature_checks from the same directory as the
        # run-tests.py script itself so use base_path.

        # Check if micropython.native is supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "native_check.py")
        if output != b"native\n":
            skip_native = True

        # Check if arbitrary-precision integers are supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "int_big.py")
        if output != b"1000000000000000000000000000000000000000000000\n":
            skip_int_big = True

        # Check if 'long long' precision integers are supported, even if arbitrary precision is not
        output = run_feature_check(pyb, args, "int_64.py")
        if output != b"4611686018427387904\n":
            skip_int_64 = True

        # Check if bytearray is supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "bytearray.py")
        if output != b"bytearray\n":
            skip_bytearray = True

        # Check if set type (and set literals) is supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "set_check.py")
        if output != b"{1}\n":
            skip_set_type = True

        # Check if slice is supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "slice.py")
        if output != b"slice\n":
            skip_slice = True

        # Check if async/await keywords are supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "async_check.py")
        if output != b"async\n":
            skip_async = True

        # Check if const keyword (MicroPython extension) is supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "const.py")
        if output != b"1\n":
            skip_const = True

        # Check if __rOP__ special methods are supported, and skip such tests if it's not
        output = run_feature_check(pyb, args, "reverse_ops.py")
        if output == b"TypeError\n":
            skip_revops = True

        # Check if fstring feature is enabled, and skip such tests if it doesn't
        output = run_feature_check(pyb, args, "fstring.py")
        if output != b"a=1\n":
            skip_fstring = True

        # Check if tstring feature is enabled, and skip such tests if it doesn't
        output = run_feature_check(pyb, args, "tstring.py")
        if output != b"tstring\n":
            skip_tstring = True

        if args.inlineasm_arch == "thumb":
            # Check if @micropython.asm_thumb supports Thumb2 instructions, and skip such tests if it doesn't
            output = run_feature_check(pyb, args, "inlineasm_thumb2.py")
            if output != b"thumb2\n":
                skip_tests.add("inlineasm/thumb/asmbcc.py")
                skip_tests.add("inlineasm/thumb/asmbitops.py")
                skip_tests.add("inlineasm/thumb/asmconst.py")
                skip_tests.add("inlineasm/thumb/asmdiv.py")
                skip_tests.add("inlineasm/thumb/asmit.py")
                skip_tests.add("inlineasm/thumb/asmspecialregs.py")
            if args.arch not in ("armv7emsp", "armv7emdp"):
                skip_tests.add("inlineasm/thumb/asmfpaddsub.py")
                skip_tests.add("inlineasm/thumb/asmfpcmp.py")
                skip_tests.add("inlineasm/thumb/asmfpldrstr.py")
                skip_tests.add("inlineasm/thumb/asmfpmuldiv.py")
                skip_tests.add("inlineasm/thumb/asmfpsqrt.py")

        if args.inlineasm_arch == "rv32":
            # Discover extension-specific inlineasm tests and add them to the
            # list of tests to run if applicable.
            for extension in RV32_ARCH_FLAGS:
                try:
                    output = run_feature_check(pyb, args, "inlineasm_rv32_{}.py".format(extension))
                    if output.strip() != "rv32_{}".format(extension).encode():
                        skip_tests.add("inlineasm/rv32/asm_ext_{}.py".format(extension))
                except FileNotFoundError:
                    pass

        # Check if emacs repl is supported, and skip such tests if it's not
        t = run_feature_check(pyb, args, "repl_emacs_check.py")
        if "True" not in str(t, "ascii"):
            skip_tests.add("cmdline/repl_emacs_keys.py")

        # Check if words movement in repl is supported, and skip such tests if it's not
        t = run_feature_check(pyb, args, "repl_words_move_check.py")
        if "True" not in str(t, "ascii"):
            skip_tests.add("cmdline/repl_words_move.py")

        upy_byteorder = run_feature_check(pyb, args, "byteorder.py")
        has_complex = run_feature_check(pyb, args, "complex.py") == b"complex\n"
        has_coverage = run_feature_check(pyb, args, "coverage.py") == b"coverage\n"
        cpy_byteorder = subprocess.check_output(
            CPYTHON3_CMD + [base_path("feature_check/byteorder.py")]
        )
        skip_endian = upy_byteorder != cpy_byteorder

        skip_inlineasm = args.inlineasm_arch is None

    # Some tests shouldn't be run on GitHub Actions
    if os.getenv("GITHUB_ACTIONS") == "true":
        if os.getenv("RUNNER_OS") == "Windows" and os.getenv("CI_BUILD_CONFIGURATION") == "Debug":
            # fails with stack overflow on Debug builds
            skip_tests.add("misc/sys_settrace_features.py")

    if args.float_prec == 0:
        skip_tests.update(tests_requiring_float)
    if args.float_prec < 32:
        skip_tests.add(
            "float/float2int_intbig.py"
        )  # requires fp32, there's float2int_fp30_intbig.py instead
        skip_tests.add(
            "float/float_struct_e.py"
        )  # requires fp32, there's float_struct_e_fp30.py instead
        skip_tests.add("float/bytes_construct.py")  # requires fp32
        skip_tests.add("float/bytearray_construct.py")  # requires fp32
        skip_tests.add("float/float_format_ints_power10.py")  # requires fp32
    if args.float_prec < 64:
        skip_tests.add("float/float_divmod.py")  # tested by float/float_divmod_relaxed.py instead
        skip_tests.add("float/float2int_doubleprec_intbig.py")
        skip_tests.add("float/float_struct_e_doubleprec.py")
        skip_tests.add("float/float_format_ints_doubleprec.py")
        skip_tests.add("float/float_parse_doubleprec.py")

    if not args.unicode:
        skip_tests.add("extmod/json_loads.py")  # tests loading a utf-8 character

    if skip_slice:
        skip_tests.update(tests_requiring_slice)

    if not has_complex:
        skip_tests.add("float/complex1.py")
        skip_tests.add("float/complex1_intbig.py")
        skip_tests.add("float/complex1_micropython.py")
        skip_tests.add("float/complex_reverse_op.py")
        skip_tests.add("float/complex_special_methods.py")
        skip_tests.add("float/int_big_float.py")
        skip_tests.add("float/true_value.py")
        skip_tests.add("float/types.py")
        skip_tests.add("float/complex_dunder.py")

    if not has_coverage:
        skip_tests.add("cmdline/cmd_parsetree.py")
        skip_tests.add("cmdline/repl_sys_ps1_ps2.py")
        skip_tests.add("extmod/ssl_poll.py")

    # Skip thread mutation tests on targets that have unsafe threading behaviour.
    if args.thread == "unsafe":
        for t in tests:
            if t.startswith("thread/mutate_"):
                skip_tests.add(t)

    # Some tests shouldn't be run on pyboard
    if args.platform not in PC_PLATFORMS:
        skip_tests.add("basics/exception_chain.py")  # warning is not printed
        skip_tests.add("micropython/meminfo.py")  # output is very different to PC output
        skip_tests.add("unicode/file1.py")  # requires local file access
        skip_tests.add("unicode/file2.py")  # requires local file access
        skip_tests.add("unicode/file_invalid.py")  # requires local file access

    # Skip certain tests when going via a .mpy file.
    skip_tests.update(via_mpy_tests_to_skip.get(args.via_mpy, ()))

    # Skip emitter-specific tests.
    skip_tests.update(emitter_tests_to_skip.get(args.emit, ()))

    # Skip platform-specific tests.
    skip_tests.update(platform_tests_to_skip.get(args.platform, ()))

    # Some tests are known to fail on 64-bit machines
    if pyb is None and platform.architecture()[0] == "64bit":
        pass

    # Some tests use unsupported features on Windows
    if os.name == "nt":
        if not sysconfig.get_platform().startswith("mingw"):
            # Works but CPython uses '\' path separator
            skip_tests.add("import/import_file.py")

    skip_tests = [os.path.realpath(base_path(skip_test)) for skip_test in skip_tests]

    def run_one_test(test_file):
        test_file_abspath = os.path.abspath(test_file).replace("\\", "/")
        # If test_file is one of our own tests always make it relative to our tests/ dir and
        # otherwise use the absolute path, regardless of actual path passed,
        # such that display and result output is always the same.
        try:
            test_file_relpath = os.path.relpath(test_file, start=base_path())
            if not test_file_relpath.startswith(".."):
                test_file = test_file_relpath
            else:
                test_file = test_file_abspath
        except ValueError:
            # Path on different drive on Windows.
            test_file = test_file_abspath
        test_file = test_file.replace("\\", "/")

        if args.filters:
            # Default verdict is the opposite of the first action
            verdict = "include" if args.filters[0][0] == "exclude" else "exclude"
            for action, pat in args.filters:
                if pat.search(test_file):
                    verdict = action
            if verdict == "exclude":
                return

        test_basename = test_file.replace("..", "_").replace("./", "").replace("/", "_")
        test_name = os.path.splitext(os.path.basename(test_file))[0]
        is_native = test_name.startswith("native_") or test_name.startswith("viper_")
        is_endian = test_name.endswith("_endian")
        is_int_big = (
            test_name.startswith("int_big")
            or test_name.endswith("_intbig")
            or test_name.startswith("ffi_int")  # these tests contain large integer literals
        )
        is_int_64 = test_name.startswith("int_64") or test_name.endswith("_int64")
        is_bytearray = test_name.startswith("bytearray") or test_name.endswith("_bytearray")
        is_set_type = test_name.startswith(("set_", "frozenset")) or test_name.endswith("_set")
        is_slice = test_name.find("slice") != -1
        is_async = test_name.startswith(("async_", "asyncio_")) or test_name.endswith("_async")
        is_const = test_name.startswith("const")
        is_fstring = test_name.startswith("string_fstring")
        is_tstring = test_name.startswith("string_tstring") or test_name.endswith("_tstring")
        is_inlineasm = test_name.startswith("asm")

        skip_it = os.path.realpath(test_file) in skip_tests
        skip_it |= skip_native and is_native
        skip_it |= skip_endian and is_endian
        skip_it |= skip_int_big and is_int_big
        skip_it |= skip_int_64 and is_int_64
        skip_it |= skip_bytearray and is_bytearray
        skip_it |= skip_set_type and is_set_type
        skip_it |= skip_slice and is_slice
        skip_it |= skip_async and is_async
        skip_it |= skip_const and is_const
        skip_it |= skip_revops and "reverse_op" in test_name
        skip_it |= skip_fstring and is_fstring
        skip_it |= skip_tstring and is_tstring
        skip_it |= skip_inlineasm and is_inlineasm

        if skip_it:
            print("skip ", test_file)
            test_results.append((test_file, "skip", ""))
            return
        elif args.dry_run:
            print("found", test_file)
            test_results.append((test_file, "found", ""))
            return

        # Run the test on the MicroPython target.
        output_mupy = run_micropython(pyb, args, test_file, test_file_abspath)

        # Check if the target requested to skip this test.
        if output_mupy == b"SKIP\n":
            if pyb is not None and hasattr(pyb, "read_until"):
                # Running on a target over a serial connection, and the target requested
                # to skip the test.  It does this via a SystemExit which triggers a soft
                # reset.  Wait for the soft reset to finish, so we don't interrupt the
                # start-up code (eg boot.py) when preparing to run the next test.
                pyb.read_until(1, b"raw REPL; CTRL-B to exit\r\n")
            print("skip ", test_file)
            test_results.append((test_file, "skip", ""))
            return
        elif output_mupy == b"SKIP-TOO-LARGE\n":
            print("lrge ", test_file)
            test_results.append((test_file, "skip", "too large"))
            return

        # Look at the output of the test to see if unittest was used.
        uses_unittest = False
        output_mupy_lines = output_mupy.splitlines()
        if any(
            line == b"ImportError: no module named 'unittest'" for line in output_mupy_lines[-3:]
        ):
            raise TestError(
                (
                    "error: test {} requires unittest".format(test_file),
                    "(eg run `mpremote mip install unittest` to install it)",
                )
            )
        elif (
            len(output_mupy_lines) > 4
            and output_mupy_lines[-4] == b"-" * 70
            and output_mupy_lines[-2] == b""
        ):
            # look for unittest summary
            unittest_ran_match = re.match(rb"Ran (\d+) tests$", output_mupy_lines[-3])
            unittest_result_match = re.match(
                b"("
                rb"(OK)( \(skipped=(\d+)\))?"
                b"|"
                rb"(FAILED) \(failures=(\d+), errors=(\d+)\)"
                b")$",
                output_mupy_lines[-1],
            )
            uses_unittest = unittest_ran_match and unittest_result_match

        # Determine the expected output.
        if uses_unittest:
            # Expected output is result of running unittest.
            output_expected = None
        else:
            # Prefer emitter-specific expected output.
            test_file_expected = test_file + "." + args.emit + ".exp"
            if not os.path.isfile(test_file_expected):
                # Fall back to generic expected output.
                test_file_expected = test_file + ".exp"
            if os.path.isfile(test_file_expected):
                # Expected output given by a file, so read that in.
                with open(test_file_expected, "rb") as f:
                    output_expected = f.read()
            else:
                # Run CPython to work out expected output.
                try:
                    output_expected = subprocess.check_output(
                        CPYTHON3_CMD + [test_file_abspath],
                        cwd=os.path.dirname(test_file),
                        stderr=subprocess.STDOUT,
                    )
                except subprocess.CalledProcessError as er:
                    output_expected = b"CPYTHON3 CRASH:\n" + er.output

            # Canonical form for all host platforms is to use \n for end-of-line.
            output_expected = output_expected.replace(b"\r\n", b"\n")

        # Work out if test passed or not.
        test_passed = False
        extra_info = ""
        if uses_unittest:
            test_passed = unittest_result_match.group(2) == b"OK"
            num_test_cases = int(unittest_ran_match.group(1))
            extra_info = "unittest: {} ran".format(num_test_cases)
            if test_passed and unittest_result_match.group(4) is not None:
                num_skipped = int(unittest_result_match.group(4))
                num_test_cases -= num_skipped
                extra_info += ", {} skipped".format(num_skipped)
            elif not test_passed:
                num_failures = int(unittest_result_match.group(6))
                num_errors = int(unittest_result_match.group(7))
                extra_info += ", {} failures, {} errors".format(num_failures, num_errors)
            extra_info = "(" + extra_info + ")"
            testcase_count.add(num_test_cases)
        else:
            testcase_count.add(len(output_expected.splitlines()))
            test_passed = output_expected == output_mupy

        filename_expected = os.path.join(result_dir, test_basename + ".exp")
        filename_mupy = os.path.join(result_dir, test_basename + ".out")

        # Print test summary, update counters, and save .exp/.out files if needed.
        if test_passed:
            print("pass ", test_file, extra_info)
            test_results.append((test_file, "pass", ""))
            rm_f(filename_expected)
            rm_f(filename_mupy)
        else:
            if output_mupy.startswith(b"could not enter raw repl"):
                extra_info = "raw REPL failed"
                raw_repl_failure_count.increment()
            print("FAIL ", test_file, extra_info)
            if output_expected is not None:
                with open(filename_expected, "wb") as f:
                    f.write(output_expected)
            else:
                rm_f(filename_expected)  # in case left over from previous failed run
            with open(filename_mupy, "wb") as f:
                f.write(output_mupy)
            test_results.append((test_file, "fail", ""))

        # Print a note if this looks like it might have been a misfired unittest
        if not uses_unittest and not test_passed:
            with open(test_file, "r") as f:
                if any(re.match("^import.+unittest", l) for l in f.readlines()):
                    print(
                        "NOTE: {} may be a unittest that doesn't run unittest.main()".format(
                            test_file
                        )
                    )

    if pyb:
        num_threads = 1

    try:
        if num_threads > 1:
            pool = ThreadPool(num_threads)
            pool.map(run_one_test, tests)
        else:
            for test in tests:
                run_one_test(test)
                if raw_repl_failure_count.value > TEST_MAXIMUM_RAW_REPL_FAILURES:
                    print("Too many raw REPL failures, aborting test run")
                    break
    except TestError as er:
        for line in er.args[0]:
            print(line)
        sys.exit(2)

    # Reclassify known-flaky test failures as ignored.
    # Safe to mutate: thread pool has joined.
    results = test_results.value
    for i, r in enumerate(results):
        if r[1] == "fail":
            reason, platforms = flaky_tests_to_ignore.get(r[0], (None, None))
            if reason is not None:
                if platforms is None or sys.platform in platforms:
                    results[i] = (r[0], "ignored", "{}: {}".format(FLAKY_REASON_PREFIX, reason))

    # Return test results.
    return test_results.value, testcase_count.value


class append_filter(argparse.Action):
    def __init__(self, option_strings, dest, **kwargs):
        super().__init__(option_strings, dest, default=[], **kwargs)

    def __call__(self, parser, args, value, option):
        if not hasattr(args, self.dest):
            args.filters = []
        if option.startswith(("-e", "--e")):
            option = "exclude"
        else:
            option = "include"
        args.filters.append((option, re.compile(value)))


def main():
    cmd_parser = argparse.ArgumentParser(
        formatter_class=argparse.RawDescriptionHelpFormatter,
        description=f"""Run and manage tests for MicroPython.

{test_instance_description}
{test_directory_description}

When running tests, run-tests.py compares the MicroPython output of the test with the output
produced by running the test through CPython unless a <test>.exp file is found (or a
<test>.native.exp file when using the native emitter), in which case it is used as comparison.

If a test fails, run-tests.py produces a pair of <test>.out and <test>.exp files in the result
directory with the MicroPython output and the expectations, respectively.
""",
        epilog=f"""\
{test_instance_epilog}
Options -i and -e can be multiple and processed in the order given. Regex
"search" (vs "match") operation is used. An action (include/exclude) of
the last matching regex is used:
  run-tests.py -i async - exclude all, then include tests containing "async" anywhere
  run-tests.py -e '/big.+int' - include all, then exclude by regex
  run-tests.py -e async -i async_foo - include all, exclude async, yet still include async_foo
""",
    )
    cmd_parser.add_argument(
        "-t", "--test-instance", default="unix", help="the MicroPython instance to test"
    )
    cmd_parser.add_argument(
        "-b", "--baudrate", default=115200, help="the baud rate of the serial device"
    )
    cmd_parser.add_argument("-u", "--user", default="micro", help="the telnet login username")
    cmd_parser.add_argument("-p", "--password", default="python", help="the telnet login password")
    cmd_parser.add_argument(
        "-d", "--test-dirs", nargs="*", help="input test directories (if no files given)"
    )
    cmd_parser.add_argument(
        "-r", "--result-dir", default=base_path("results"), help="directory for test results"
    )
    cmd_parser.add_argument(
        "-e",
        "--exclude",
        action=append_filter,
        metavar="REGEX",
        dest="filters",
        help="exclude test by regex on path/name.py",
    )
    cmd_parser.add_argument(
        "-i",
        "--include",
        action=append_filter,
        metavar="REGEX",
        dest="filters",
        help="include test by regex on path/name.py",
    )
    cmd_parser.add_argument(
        "--dry-run",
        action="store_true",
        help="Show tests which would run (though might still be skipped at runtime)",
    )
    cmd_parser.add_argument(
        "--emit", default="bytecode", help="MicroPython emitter to use (bytecode or native)"
    )
    cmd_parser.add_argument("--heapsize", help="heapsize to use (use default if not specified)")
    cmd_parser.add_argument(
        "--via-mpy", action="store_true", help="compile .py files to .mpy first"
    )
    cmd_parser.add_argument("--mpy-cross-flags", default="", help="flags to pass to mpy-cross")
    cmd_parser.add_argument(
        "--keep-path", action="store_true", help="do not clear MICROPYPATH when running tests"
    )
    cmd_parser.add_argument(
        "-j",
        "--jobs",
        default=multiprocessing.cpu_count(),
        metavar="N",
        type=int,
        help="Number of tests to run simultaneously",
    )
    cmd_parser.add_argument("files", nargs="*", help="input test files")
    cmd_parser.add_argument(
        "--print-failures",
        action="store_true",
        help="print the diff of expected vs. actual output for failed tests and exit",
    )
    cmd_parser.add_argument(
        "--clean-failures",
        action="store_true",
        help="delete the .exp and .out files from failed tests and exit",
    )
    cmd_parser.add_argument(
        "--run-failures",
        action="store_true",
        help="re-run only the failed tests",
    )
    cmd_parser.add_argument(
        "--begin",
        metavar="PROLOGUE",
        default=None,
        help="prologue python file to execute before module import",
    )
    cmd_parser.add_argument(
        "--target-wiring",
        default=None,
        help="force the given script to be used as target_wiring.py",
    )
    args = cmd_parser.parse_args()

    prologue = ""
    if args.begin:
        with open(args.begin, "rt") as source:
            prologue = source.read()
    set_injected_prologue(prologue)

    if args.print_failures:
        for out in glob(os.path.join(args.result_dir, "*.out")):
            testbase = out[:-4]
            print()
            print("FAILURE {0}".format(testbase))
            if os.path.exists(testbase + ".exp"):
                # Show diff of expected and actual output.
                os.system("{0} {1}.exp {1}.out".format(DIFF, testbase))
            else:
                # No expected output, just show the actual output (eg from a unittest).
                with open(out) as f:
                    for line in f:
                        print(line, end="")

        sys.exit(0)

    if args.clean_failures:
        for f in glob(os.path.join(args.result_dir, "*.exp")) + glob(
            os.path.join(args.result_dir, "*.out")
        ):
            os.remove(f)
        rm_f(get_results_filename(args))

        sys.exit(0)

    # Get the test instance to run on.
    pyb = get_test_instance(args.test_instance, args.baudrate, args.user, args.password)

    # Automatically detect the platform.
    detect_test_platform(pyb, args)

    if args.run_failures and (any(args.files) or args.test_dirs is not None):
        raise ValueError(
            "--run-failures cannot be used together with files or --test-dirs arguments"
        )

    if args.run_failures:
        results_file = get_results_filename(args)
        if os.path.exists(results_file):
            with open(results_file, "r") as f:
                tests = list(test[0] for test in json.load(f)["results"] if test[1] == "fail")
        else:
            tests = []
    elif len(args.files) == 0:
        test_extensions = ("*.py",)
        if args.platform == "webassembly":
            test_extensions += ("*.js", "*.mjs")

        all_test_dirs = []
        main_tests_dir_in_args = None
        if args.test_dirs is not None:
            # Run tests from given directories though if user explicitly passes this directory as argument
            # still do the normal test discovery to be consistent with running from within this directory.
            main_tests_dir = os.path.realpath(base_path())
            for test_dir in args.test_dirs:
                if os.path.realpath(test_dir) == main_tests_dir:
                    main_tests_dir_in_args = test_dir
                else:
                    all_test_dirs.append(test_dir)

        if args.test_dirs is None or main_tests_dir_in_args is not None:
            test_dirs = (
                "basics",
                "micropython",
                "misc",
                "extmod",
                "stress",
            )
            if args.inlineasm_arch is not None:
                test_dirs += ("inlineasm/{}".format(args.inlineasm_arch),)
            if args.thread is not None:
                test_dirs += ("thread",)
            if args.float_prec > 0:
                test_dirs += ("float",)
            if args.unicode:
                test_dirs += ("unicode",)
            port_specific_test_dir = "ports/{}".format(platform_to_port(args.platform))
            if os.path.isdir(port_specific_test_dir):
                test_dirs += (port_specific_test_dir,)
            if args.platform in PC_PLATFORMS:
                # run PC tests
                test_dirs += ("import",)
                if args.build != "minimal":
                    test_dirs += ("cmdline", "io")

            all_test_dirs.extend(
                test_dir
                if main_tests_dir_in_args is None
                else os.path.join(main_tests_dir_in_args, test_dir)
                for test_dir in test_dirs
            )

        tests = sorted(
            test_file
            for test_files in (
                glob(os.path.join(dir, ext)) for dir in all_test_dirs for ext in test_extensions
            )
            for test_file in test_files
        )
    else:
        # tests explicitly given
        tests = args.files

    # If any tests need it, prepare the target_wiring script for the target.
    if pyb and any(test.endswith(tests_requiring_target_wiring) for test in tests):
        detect_target_wiring_script(pyb, args)

    # End the target information line.
    print()

    if not args.keep_path:
        # Clear search path to make sure tests use only builtin modules, those in
        # extmod, and a path to unittest in case it's needed.
        os.environ["MICROPYPATH"] = (
            ".frozen"
            + os.pathsep
            + base_path("../extmod")
            + os.pathsep
            + base_path("../lib/micropython-lib/python-stdlib/unittest")
        )

    try:
        os.makedirs(args.result_dir, exist_ok=True)
        test_results, testcase_count = run_tests(pyb, tests, args, args.result_dir, args.jobs)
        res = create_test_report(args, test_results, testcase_count)
    finally:
        if pyb:
            pyb.close()

    if not res:
        sys.exit(1)


if __name__ == "__main__":
    main()