Anforderungen  |   Konzepte  |   Entwurf  |   Entwicklung  |   Qualitätssicherung  |   Lebenszyklus  |   Steuerung
 
 
 
 


Quelle  jittests.py   Sprache: Python

 
#!/usr/bin/env python
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.


# jit_test.py -- Python harness for JavaScript trace tests.

import os
import re
import sys
import traceback
from collections import namedtuple
from enum import Enum

if sys.platform.startswith("linux"or sys.platform.startswith("darwin"):
    from .tasks_unix import run_all_tests
else:
    from .tasks_win import run_all_tests

from .progressbar import NullProgressBar, ProgressBar
from .results import escape_cmdline
from .structuredlog import TestLogger
from .tempfile import TemporaryDirectory

TESTS_LIB_DIR = os.path.dirname(os.path.abspath(__file__))
JS_DIR = os.path.dirname(os.path.dirname(TESTS_LIB_DIR))
TOP_SRC_DIR = os.path.dirname(os.path.dirname(JS_DIR))
TEST_DIR = os.path.join(JS_DIR, "jit-test""tests")
LIB_DIR = os.path.join(JS_DIR, "jit-test""lib") + os.path.sep
MODULE_DIR = os.path.join(JS_DIR, "jit-test""modules") + os.path.sep
SHELL_XDR = "shell.xdr"


class OutputStatus(Enum):
    OK = 1
    SKIPPED = 2
    FAILED = 3

    def __bool__(self):
        return self != OutputStatus.FAILED


# Backported from Python 3.1 posixpath.py


def _relpath(path, start=None):
    """Return a relative version of a path"""

    if not path:
        raise ValueError("no path specified")

    if start is None:
        start = os.curdir

    start_list = os.path.abspath(start).split(os.sep)
    path_list = os.path.abspath(path).split(os.sep)

    # Work out how much of the filepath is shared by start and path.
    i = len(os.path.commonprefix([start_list, path_list]))

    rel_list = [os.pardir] * (len(start_list) - i) + path_list[i:]
    if not rel_list:
        return os.curdir
    return os.path.join(*rel_list)


# Mapping of Python chars to their javascript string representation.
QUOTE_MAP = {
    "\\""\\\\",
    "\b""\\b",
    "\f""\\f",
    "\n""\\n",
    "\r""\\r",
    "\t""\\t",
    "\v""\\v",
}

# Quote the string S, javascript style.


def js_quote(quote, s):
    result = quote
    for c in s:
        if c == quote:
            result += "\\" + quote
        elif c in QUOTE_MAP:
            result += QUOTE_MAP[c]
        else:
            result += c
    result += quote
    return result


os.path.relpath = _relpath


def extend_condition(condition, value):
    if condition:
        condition += " || "
    condition += "({})".format(value)
    return condition


class JitTest:
    VALGRIND_CMD = []
    paths = (d for d in os.environ["PATH"].split(os.pathsep))
    valgrinds = (os.path.join(d, "valgrind"for d in paths)
    if any(os.path.exists(p) for p in valgrinds):
        VALGRIND_CMD = [
            "valgrind",
            "-q",
            "--smc-check=all-non-file",
            "--error-exitcode=1",
            "--gen-suppressions=all",
            "--show-possibly-lost=no",
            "--leak-check=full",
        ]
        if os.uname()[0] == "Darwin":
            VALGRIND_CMD.append("--dsymutil=yes")

    del paths
    del valgrinds

    def __init__(self, path):
        # Absolute path of the test file.
        self.path = path

        # Path relative to the top mozilla/ directory.
        self.relpath_top = os.path.relpath(path, TOP_SRC_DIR)

        # Path relative to mozilla/js/src/jit-test/tests/.
        self.relpath_tests = os.path.relpath(path, TEST_DIR)

        # jit flags to enable
        self.jitflags = []
        # True means the test is slow-running
        self.slow = False
        # Heavy tests will never run alongside other heavy tests
        self.heavy = False
        # True means that OOM is not considered a failure
        self.allow_oom = False
        # True means CrashAtUnhandlableOOM is not considered a failure
        self.allow_unhandlable_oom = False
        # True means that hitting recursion the limits is not considered a failure.
        self.allow_overrecursed = False
        # True means run under valgrind
        self.valgrind = False
        # True means force Pacific time for the test
        self.tz_pacific = False
        # Additional files to include, in addition to prologue.js
        self.other_lib_includes = []
        self.other_script_includes = []
        # List of other configurations to test with.
        self.test_also = []
        # List of other configurations to test with all existing variants.
        self.test_join = []
        # Errors to expect and consider passing
        self.expect_error = ""
        # Exit status to expect from shell
        self.expect_status = 0
        # Exit status or error output.
        self.expect_crash = False
        self.is_module = False
        # Reflect.stringify implementation to test
        self.test_reflect_stringify = None
        # Use self-hosted XDR instead of parsing the source stored in the binary.
        self.selfhosted_xdr_path = None
        self.selfhosted_xdr_mode = "off"

        # Skip-if condition. We don't have a xulrunner, but we can ask the shell
        # directly.
        self.skip_if_cond = ""
        self.skip_variant_if_cond = {}

        # Expected by the test runner. Always true for jit-tests.
        self.enable = True

    def copy(self):
        t = JitTest(self.path)
        t.jitflags = self.jitflags[:]
        t.slow = self.slow
        t.heavy = self.heavy
        t.allow_oom = self.allow_oom
        t.allow_unhandlable_oom = self.allow_unhandlable_oom
        t.allow_overrecursed = self.allow_overrecursed
        t.valgrind = self.valgrind
        t.tz_pacific = self.tz_pacific
        t.other_lib_includes = self.other_lib_includes[:]
        t.other_script_includes = self.other_script_includes[:]
        t.test_also = self.test_also
        t.test_join = self.test_join
        t.expect_error = self.expect_error
        t.expect_status = self.expect_status
        t.expect_crash = self.expect_crash
        t.test_reflect_stringify = self.test_reflect_stringify
        t.selfhosted_xdr_path = self.selfhosted_xdr_path
        t.selfhosted_xdr_mode = self.selfhosted_xdr_mode
        t.enable = True
        t.is_module = self.is_module
        t.skip_if_cond = self.skip_if_cond
        t.skip_variant_if_cond = self.skip_variant_if_cond
        return t

    def copy_and_extend_jitflags(self, variant):
        t = self.copy()
        t.jitflags.extend(variant)
        for flags in variant:
            if flags in self.skip_variant_if_cond:
                t.skip_if_cond = extend_condition(
                    t.skip_if_cond, self.skip_variant_if_cond[flags]
                )
        return t

    def copy_variants(self, variants):
        # Append variants to be tested in addition to the current set of tests.
        variants = variants + self.test_also

        # For each existing variant, duplicates it for each list of options in
        # test_join.  This will multiply the number of variants by 2 for set of
        # options.
        for join_opts in self.test_join:
            variants = variants + [opts + join_opts for opts in variants]

        # For each list of jit flags, make a copy of the test.
        return [self.copy_and_extend_jitflags(v) for v in variants]

    COOKIE = b"|jit-test|"

    # We would use 500019 (5k19), but quit() only accepts values up to 127, due to fuzzers
    SKIPPED_EXIT_STATUS = 59
    Directives = {}

    @classmethod
    def find_directives(cls, file_name):
        meta = ""
        line = open(file_name, "rb").readline()
        i = line.find(cls.COOKIE)
        if i != -1:
            meta = ";" + line[i + len(cls.COOKIE) :].decode(errors="strict").strip("\n")
        return meta

    @classmethod
    def from_file(cls, path, options):
        test = cls(path)

        # If directives.txt exists in the test's directory then it may
        # contain metainformation that will be catenated with
        # whatever's in the test file.  The form of the directive in
        # the directive file is the same as in the test file.  Only
        # the first line is considered, just as for the test file.

        dir_meta = ""
        dir_name = os.path.dirname(path)
        if dir_name in cls.Directives:
            dir_meta = cls.Directives[dir_name]
        else:
            meta_file_name = os.path.join(dir_name, "directives.txt")
            if os.path.exists(meta_file_name):
                dir_meta = cls.find_directives(meta_file_name)
            cls.Directives[dir_name] = dir_meta

        filename, file_extension = os.path.splitext(path)
        meta = cls.find_directives(path)

        if meta != "" or dir_meta != "":
            meta = meta + dir_meta
            parts = meta.split(";")
            for part in parts:
                part = part.strip()
                if not part:
                    continue
                name, _, value = part.partition(":")
                if value:
                    value = value.strip()
                    if name == "error":
                        test.expect_error = value
                    elif name == "exitstatus":
                        try:
                            status = int(value, 0)
                            if status == test.SKIPPED_EXIT_STATUS:
                                print(
                                    "warning: jit-tests uses {} as a sentinel"
                                    " return value {}",
                                    test.SKIPPED_EXIT_STATUS,
                                    path,
                                )
                            else:
                                test.expect_status = status
                        except ValueError:
                            print(
                                "warning: couldn't parse exit status"
                                " {}".format(value)
                            )
                    elif name == "thread-count":
                        try:
                            test.jitflags.append(
                                "--thread-count={}".format(int(value, 0))
                            )
                        except ValueError:
                            print(
                                "warning: couldn't parse thread-count"
                                " {}".format(value)
                            )
                    elif name == "include":
                        test.other_lib_includes.append(value)
                    elif name == "local-include":
                        test.other_script_includes.append(value)
                    elif name == "skip-if":
                        test.skip_if_cond = extend_condition(test.skip_if_cond, value)
                    elif name == "skip-variant-if":
                        try:
                            [variant, condition] = value.split(",")
                            test.skip_variant_if_cond[variant] = extend_condition(
                                test.skip_if_cond, condition
                            )
                        except ValueError:
                            print("warning: couldn't parse skip-variant-if")
                    else:
                        print(
                            "{}: warning: unrecognized |jit-test| attribute"
                            " {}".format(path, part)
                        )
                else:
                    if name == "slow":
                        test.slow = True
                    elif name == "heavy":
                        test.heavy = True
                    elif name == "allow-oom":
                        test.allow_oom = True
                    elif name == "allow-unhandlable-oom":
                        test.allow_unhandlable_oom = True
                    elif name == "allow-overrecursed":
                        test.allow_overrecursed = True
                    elif name == "valgrind":
                        test.valgrind = options.valgrind
                    elif name == "tz-pacific":
                        test.tz_pacific = True
                    elif name.startswith("test-also="):
                        test.test_also.append(
                            re.split(r"\s+", name[len("test-also=") :])
                        )
                    elif name.startswith("test-join="):
                        test.test_join.append(
                            re.split(r"\s+", name[len("test-join=") :])
                        )
                    elif name == "module":
                        test.is_module = True
                    elif name == "crash":
                        # Crashes are only allowed in self-test, as it is
                        # intended to verify that our testing infrastructure
                        # works, and not meant as a way to accept temporary
                        # failing tests. These tests should either be fixed or
                        # skipped.
                        assert (
                            "self-test" in path
                        ), "{}: has an unexpected crash annotation.".format(path)
                        test.expect_crash = True
                    elif name.startswith("--"):
                        # // |jit-test| --ion-gvn=off; --no-sse4
                        test.jitflags.append(name)
                    elif name.startswith("-P"):
                        prefAndValue = name.split()
                        assert (
                            len(prefAndValue) == 2
                        ), f"{name}: failed to parse preference"
                        # // |jit-test| -P pref(=value)?
                        test.jitflags.append("--setpref=" + prefAndValue[1])
                    else:
                        print(
                            "{}: warning: unrecognized |jit-test| attribute"
                            " {}".format(path, part)
                        )

        if options.valgrind_all:
            test.valgrind = True

        if options.test_reflect_stringify is not None:
            test.expect_error = ""
            test.expect_status = 0

        return test

    def command(self, prefix, libdir, moduledir, tempdir, remote_prefix=None):
        path = self.path
        if remote_prefix:
            path = self.path.replace(TEST_DIR, remote_prefix)

        scriptdir_var = os.path.dirname(path)
        if not scriptdir_var.endswith("/"):
            scriptdir_var += "/"

        # Note: The tempdir provided as argument is managed by the caller
        # should remain alive as long as the test harness. Therefore, the XDR
        # content of the self-hosted code would be accessible to all JS Shell
        # instances.
        self.selfhosted_xdr_path = os.path.join(tempdir, SHELL_XDR)

        # Platforms where subprocess immediately invokes exec do not care
        # whether we use double or single quotes. On windows and when using
        # a remote device, however, we have to be careful to use the quote
        # style that is the opposite of what the exec wrapper uses.
        if remote_prefix:
            quotechar = '"'
        else:
            quotechar = "'"

        # Don't merge the expressions: We want separate -e arguments to avoid
        # semicolons in the command line, bug 1351607.
        exprs = [
            "const platform={}".format(js_quote(quotechar, sys.platform)),
            "const libdir={}".format(js_quote(quotechar, libdir)),
            "const scriptdir={}".format(js_quote(quotechar, scriptdir_var)),
        ]

        # We may have specified '-a' or '-d' twice: once via --jitflags, once
        # via the "|jit-test|" line.  Remove dups because they are toggles.
        # Note: |dict.fromkeys(flags)| is similar to |set(flags)| but it
        # preserves order.
        cmd = prefix + []
        cmd += list(dict.fromkeys(self.jitflags))
        # Handle selfhosted XDR file.
        if self.selfhosted_xdr_mode != "off":
            cmd += [
                "--selfhosted-xdr-path",
                self.selfhosted_xdr_path,
                "--selfhosted-xdr-mode",
                self.selfhosted_xdr_mode,
            ]
        for expr in exprs:
            cmd += ["-e", expr]
        for inc in self.other_lib_includes:
            cmd += ["-f", libdir + inc]
        for inc in self.other_script_includes:
            cmd += ["-f", scriptdir_var + inc]
        if self.skip_if_cond:
            cmd += [
                "-e",
                "if ({}) quit({})".format(self.skip_if_cond, self.SKIPPED_EXIT_STATUS),
            ]
        cmd += ["--module-load-path", moduledir]
        if self.is_module:
            cmd += ["--module", path]
        elif self.test_reflect_stringify is None:
            cmd += ["-f", path]
        else:
            cmd += ["--", self.test_reflect_stringify, "--check", path]

        if self.valgrind:
            cmd = self.VALGRIND_CMD + cmd

        if self.allow_unhandlable_oom or self.expect_crash:
            cmd += ["--suppress-minidump"]

        return cmd

    def get_command(self, prefix, tempdir):
        """Shim for the test runner."""
        return self.command(prefix, LIB_DIR, MODULE_DIR, tempdir)


def find_tests(substring=None):
    ans = []
    for dirpath, dirnames, filenames in os.walk(TEST_DIR):
        dirnames.sort()
        filenames.sort()
        if dirpath == ".":
            continue

        for filename in filenames:
            if not filename.endswith(".js"):
                continue
            if filename in ("shell.js""browser.js"):
                continue
            test = os.path.join(dirpath, filename)
            if substring is None or substring in os.path.relpath(test, TEST_DIR):
                ans.append(test)
    return ans


def check_output(out, err, rc, timed_out, test, options):
    # Allow skipping to compose with other expected results
    if test.skip_if_cond:
        if rc == test.SKIPPED_EXIT_STATUS:
            return OutputStatus.SKIPPED

    if timed_out:
        relpath = os.path.normpath(test.relpath_tests).replace(os.sep, "/")
        if relpath in options.ignore_timeouts:
            return OutputStatus.OK
        return OutputStatus.FAILED

    if test.expect_error:
        # The shell exits with code 3 on uncaught exceptions.
        if rc != 3:
            return OutputStatus.FAILED

        return test.expect_error in err

    for line in out.split("\n"):
        if line.startswith("Trace stats check failed"):
            return OutputStatus.FAILED

    for line in err.split("\n"):
        if "Assertion failed:" in line:
            return OutputStatus.FAILED

    if test.expect_crash:
        # Python 3 on Windows interprets process exit codes as unsigned
        # integers, where Python 2 used to allow signed integers. Account for
        # each possibility here.
        if sys.platform == "win32" and rc in (3 - 2**31, 3 + 2**31):
            return OutputStatus.OK

        if sys.platform != "win32" and rc == -11:
            return OutputStatus.OK

        # When building with ASan enabled, ASan will convert the -11 returned
        # value to 1. As a work-around we look for the error output which
        # includes the crash reason.
        if rc == 1 and ("Hit MOZ_CRASH" in err or "Assertion failure:" in err):
            return OutputStatus.OK

        # When running jittests on Android, SEGV results in a return code of
        # 128 + 11 = 139. Due to a bug in tinybox, we have to check for 138 as
        # well.
        if rc == 139 or rc == 138:
            return OutputStatus.OK

        # Crashing test should always crash as expected, otherwise this is an
        # error. The JS shell crash() function can be used to force the test
        # case to crash in unexpected configurations.
        return OutputStatus.FAILED

    if rc != test.expect_status:
        # Allow a non-zero exit code if we want to allow OOM, but only if we
        # actually got OOM.
        if (
            test.allow_oom
            and "out of memory" in err
            and "Assertion failure" not in err
            and "MOZ_CRASH" not in err
        ):
            return OutputStatus.OK

        # Allow a non-zero exit code if we want to allow unhandlable OOM, but
        # only if we actually got unhandlable OOM.
        if test.allow_unhandlable_oom and "MOZ_CRASH([unhandlable oom]" in err:
            return OutputStatus.OK

        # Allow a non-zero exit code if we want to all too-much-recursion and
        # the test actually over-recursed.
        if (
            test.allow_overrecursed
            and "too much recursion" in err
            and "Assertion failure" not in err
        ):
            return OutputStatus.OK

        # Allow a zero exit code if we are running under a sanitizer that
        # forces the exit status.
        if test.expect_status != 0 and options.unusable_error_status:
            return OutputStatus.OK

        return OutputStatus.FAILED

    return OutputStatus.OK


def print_automation_format(ok, res, slog):
    # Output test failures in a parsable format suitable for automation, eg:
    # TEST-RESULT | filename.js | Failure description (code N, args "--foobar")
    #
    # Example:
    # TEST-PASS | foo/bar/baz.js | (code 0, args "--ion-eager")
    # TEST-UNEXPECTED-FAIL | foo/bar/baz.js | TypeError: or something (code -9, args "--no-ion")
    # INFO exit-status     : 3
    # INFO timed-out       : False
    # INFO stdout          > foo
    # INFO stdout          > bar
    # INFO stdout          > baz
    # INFO stderr         2> TypeError: or something
    # TEST-UNEXPECTED-FAIL | jit_test.py: Test execution interrupted by user
    result = "TEST-PASS" if ok else "TEST-UNEXPECTED-FAIL"
    message = "Success" if ok else res.describe_failure()
    jitflags = " ".join(res.test.jitflags)
    print(
        '{} | {} | {} (code {}, args "{}") [{:.1f} s]'.format(
            result, res.test.relpath_top, message, res.rc, jitflags, res.dt
        )
    )

    details = {
        "message": message,
        "extra": {
            "jitflags": jitflags,
        },
    }
    if res.extra:
        details["extra"].update(res.extra)
    slog.test(res.test.relpath_tests, "PASS" if ok else "FAIL", res.dt, **details)

    # For failed tests, print as much information as we have, to aid debugging.
    if ok:
        return
    print("INFO exit-status : {}".format(res.rc))
    print("INFO timed-out : {}".format(res.timed_out))
    warnings = []
    for line in res.out.splitlines():
        # See Bug 1868693
        if line.startswith("WARNING"and "unused DT entry" in line:
            warnings.append(line)
            continue
        print("INFO stdout > " + line.strip())
    for line in res.err.splitlines():
        # See Bug 1868693
        if line.startswith("WARNING"and "unused DT entry" in line:
            warnings.append(line)
            continue
        print("INFO stderr 2> " + line.strip())
    for line in warnings:
        print("INFO (warn-stderr) 2> " + line.strip())


def print_test_summary(num_tests, failures, complete, slow_tests, doing, options):
    def test_details(res):
        if options.show_failed:
            return escape_cmdline(res.cmd)
        return " ".join(res.test.jitflags + [res.test.relpath_tests])

    if failures:
        if options.write_failures:
            try:
                out = open(options.write_failures, "w")
                # Don't write duplicate entries when we are doing multiple
                # failures per job.
                written = set()
                for res in failures:
                    if res.test.path not in written:
                        out.write(os.path.relpath(res.test.path, TEST_DIR) + "\n")
                        if options.write_failure_output:
                            out.write(res.out)
                            out.write(res.err)
                            out.write("Exit code: " + str(res.rc) + "\n")
                        written.add(res.test.path)
                out.close()
            except IOError:
                sys.stderr.write(
                    "Exception thrown trying to write failure"
                    " file '{}'\n".format(options.write_failures)
                )
                traceback.print_exc()
                sys.stderr.write("---\n")

        print("FAILURES:")
        for res in failures:
            if not res.timed_out:
                print(" " + test_details(res))

        print("TIMEOUTS:")
        for res in failures:
            if res.timed_out:
                print(" " + test_details(res))
    else:
        print(
            "PASSED ALL"
            + (
                ""
                if complete
                else " (partial run -- interrupted by user {})".format(doing)
            )
        )

    if options.format == "automation":
        num_failures = len(failures) if failures else 0
        print("Result summary:")
        print("Passed: {:d}".format(num_tests - num_failures))
        print("Failed: {:d}".format(num_failures))

    if num_tests != 0 and options.show_slow:
        threshold = options.slow_test_threshold
        fraction_fast = 1 - len(slow_tests) / num_tests
        print(
            "{:5.2f}% of tests ran in under {}s".format(fraction_fast * 100, threshold)
        )

        print("Slowest tests that took longer than {}s:".format(threshold))
        slow_tests.sort(key=lambda res: res.dt, reverse=True)
        any = False
        for i in range(min(len(slow_tests), 20)):
            res = slow_tests[i]
            print(" {:6.2f} {}".format(res.dt, test_details(res)))
            any = True
        if not any:
            print("None")

    return not failures


def create_progressbar(num_tests, options):
    if (
        not options.hide_progress
        and not options.show_cmd
        and ProgressBar.conservative_isatty()
    ):
        fmt = [
            {"value""PASS""color""green"},
            {"value""FAIL""color""red"},
            {"value""TIMEOUT""color""blue"},
            {"value""SKIP""color""brightgray"},
        ]
        return ProgressBar(num_tests, fmt)
    return NullProgressBar()


def process_test_results(results, num_tests, pb, options, slog):
    failures = []
    timeouts = 0
    skipped = 0
    complete = False
    output_dict = {}
    doing = "before starting"
    slow_tests = []

    if num_tests == 0:
        pb.finish(True)
        complete = True
        return print_test_summary(
            num_tests, failures, complete, slow_tests, doing, options
        )

    try:
        for i, res in enumerate(results):
            status = check_output(
                res.out, res.err, res.rc, res.timed_out, res.test, options
            )

            if status:
                show_output = options.show_output and not options.failed_only
            else:
                show_output = options.show_output or not options.no_show_failed

            if show_output:
                pb.beginline()
                sys.stdout.write(res.out)
                sys.stdout.write(res.err)
                sys.stdout.write("Exit code: {}\n".format(res.rc))

            if res.test.valgrind and not show_output:
                pb.beginline()
                sys.stdout.write(res.err)

            if options.check_output:
                if res.test.path in output_dict.keys():
                    if output_dict[res.test.path] != res.out:
                        pb.message(
                            "FAIL - OUTPUT DIFFERS {}".format(res.test.relpath_tests)
                        )
                else:
                    output_dict[res.test.path] = res.out

            doing = "after {}".format(res.test.relpath_tests)
            if status == OutputStatus.SKIPPED:
                skipped += 1
            elif status == OutputStatus.FAILED:
                failures.append(res)
                if res.timed_out:
                    pb.message("TIMEOUT - {}".format(res.test.relpath_tests))
                    timeouts += 1
                else:
                    pb.message("FAIL - {}".format(res.test.relpath_tests))

            if options.format == "automation":
                print_automation_format(status, res, slog)

            n = i + 1
            pb.update(
                n,
                {
                    "PASS": n - len(failures),
                    "FAIL": len(failures),
                    "TIMEOUT": timeouts,
                    "SKIP": skipped,
                },
            )

            if res.dt > options.slow_test_threshold:
                slow_tests.append(res)
        complete = True
    except KeyboardInterrupt:
        print(
            "TEST-UNEXPECTED-FAIL | jit_test.py"
            + " : Test execution interrupted by user"
        )

    pb.finish(True)
    return print_test_summary(num_tests, failures, complete, slow_tests, doing, options)


def run_tests(tests, num_tests, prefix, options, remote=False):
    slog = None
    if options.format == "automation":
        slog = TestLogger("jittests")
        slog.suite_start()

    if remote:
        ok = run_tests_remote(tests, num_tests, prefix, options, slog)
    else:
        ok = run_tests_local(tests, num_tests, prefix, options, slog)

    if slog:
        slog.suite_end()

    return ok


def run_tests_local(tests, num_tests, prefix, options, slog):
    # The jstests tasks runner requires the following options. The names are
    # taken from the jstests options processing code, which are frequently
    # subtly different from the options jit-tests expects. As such, we wrap
    # them here, as needed.
    AdaptorOptions = namedtuple(
        "AdaptorOptions",
        [
            "worker_count",
            "passthrough",
            "timeout",
            "output_fp",
            "hide_progress",
            "run_skipped",
            "show_cmd",
            "use_xdr",
        ],
    )
    shim_options = AdaptorOptions(
        options.max_jobs,
        False,
        options.timeout,
        sys.stdout,
        False,
        True,
        options.show_cmd,
        options.use_xdr,
    )

    with TemporaryDirectory() as tempdir:
        pb = create_progressbar(num_tests, options)
        gen = run_all_tests(tests, prefix, tempdir, pb, shim_options)
        ok = process_test_results(gen, num_tests, pb, options, slog)
    return ok


def run_tests_remote(tests, num_tests, prefix, options, slog):
    # Setup device with everything needed to run our tests.
    from mozdevice import ADBError, ADBTimeoutError

    from .tasks_adb_remote import get_remote_results

    # Run all tests.
    pb = create_progressbar(num_tests, options)
    try:
        gen = get_remote_results(tests, prefix, pb, options)
        ok = process_test_results(gen, num_tests, pb, options, slog)
    except (ADBError, ADBTimeoutError):
        print("TEST-UNEXPECTED-FAIL | jit_test.py" + " : Device error during test")
        raise
    return ok


if __name__ == "__main__":
    print("Use ../jit-test/jit_test.py to run these tests.")

Messung V0.5
C=89 H=89 G=88

¤ Dauer der Verarbeitung: 0.13 Sekunden  ¤

*© Formatika GbR, Deutschland






Wurzel

Suchen

Beweissystem der NASA

Beweissystem Isabelle

NIST Cobol Testsuite

Cephes Mathematical Library

Wiener Entwicklungsmethode

Haftungshinweis

Die Informationen auf dieser Webseite wurden nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit, noch Qualität der bereit gestellten Informationen zugesichert.

Bemerkung:

Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.






                                                                                                                                                                                                                                                                                                                                                                                                     


Neuigkeiten

     Aktuelles
     Motto des Tages

Software

     Produkte
     Quellcodebibliothek

Aktivitäten

     Artikel über Sicherheit
     Anleitung zur Aktivierung von SSL

Muße

     Gedichte
     Musik
     Bilder

Jenseits des Üblichen ....
    

Besucherstatistik

Besucherstatistik

Monitoring

Montastic status badge