run_project_tests: Clean on backends that can't install

Also sets more groundwork for running unit tests with backends other
that Ninja.

Transferring global state to executors is totally broken in Python 3.4
so just serialize all the commands.

run_cross_test.py

@@ -43,7 +43,6 @@ def runtests(cross_file):
     sys.exit(failing_tests)
 
 if __name__ == '__main__':
-    backend = 'ninja'
+    setup_commands('ninja')
-    setup_commands()
     cross_file = sys.argv[1]
     runtests(cross_file)

run_project_tests.py

@@ -34,7 +34,7 @@ import multiprocessing
 import concurrent.futures as conc
 import re
 
-from run_tests import get_backend_commands
+from run_tests import get_backend_commands, get_backend_args_for_dir
 
 
 class BuildStep(Enum):
@@ -155,18 +155,13 @@ def stop_handler(signal, frame):
 signal.signal(signal.SIGINT, stop_handler)
 signal.signal(signal.SIGTERM, stop_handler)
 
-# Must define these here for compatibility with Python 3.4
+# Needed when running cross tests because we don't generate prebuilt files
-backend = None
+compiler = None
-backend_flags = None
-compile_commands = None
-clean_commands = []
-test_commands = None
-install_commands = []
-uninstall_commands = None
 
-def setup_commands():
+def setup_commands(optbackend):
     global do_debug, backend, backend_flags
     global compile_commands, clean_commands, test_commands, install_commands, uninstall_commands
+    backend = optbackend
     msbuild_exe = shutil.which('msbuild')
     # Auto-detect backend if unspecified
     if backend is None:
@@ -188,14 +183,6 @@ def setup_commands():
     compile_commands, clean_commands, test_commands, install_commands, \
         uninstall_commands = get_backend_commands(backend, do_debug)
 
-def get_compile_commands_for_dir(compile_commands, test_build_dir):
-    if 'msbuild' in compile_commands[0]:
-        sln_name = glob(os.path.join(test_build_dir, '*.sln'))[0]
-        comp = compile_commands + [os.path.split(sln_name)[-1]]
-    else:
-        comp = compile_commands
-    return comp
-
 def get_relative_files_list_from_dir(fromdir):
     paths = []
     for (root, _, files) in os.walk(fromdir):
@@ -221,7 +208,7 @@ def platform_fix_name(fname):
 
     return fname
 
-def validate_install(srcdir, installdir):
+def validate_install(srcdir, installdir, compiler):
     # List of installed files
     info_file = os.path.join(srcdir, 'installed_files.txt')
     # If this exists, the test does not install any other files
@@ -315,18 +302,18 @@ def parse_test_args(testdir):
         pass
     return args
 
-def run_test(skipped, testdir, extra_args, flags, compile_commands, should_fail):
+def run_test(skipped, testdir, extra_args, compiler, backend, flags, commands, should_fail):
     if skipped:
         return None
     with AutoDeletedDir(tempfile.mkdtemp(prefix='b ', dir='.')) as build_dir:
         with AutoDeletedDir(tempfile.mkdtemp(prefix='i ', dir=os.getcwd())) as install_dir:
             try:
-                return _run_test(testdir, build_dir, install_dir, extra_args, flags, compile_commands, should_fail)
+                return _run_test(testdir, build_dir, install_dir, extra_args, compiler, backend, flags, commands, should_fail)
             finally:
                 mlog.shutdown() # Close the log file because otherwise Windows wets itself.
 
-def _run_test(testdir, test_build_dir, install_dir, extra_args, flags, compile_commands, should_fail):
+def _run_test(testdir, test_build_dir, install_dir, extra_args, compiler, backend, flags, commands, should_fail):
-    global install_commands, clean_commands
+    compile_commands, clean_commands, install_commands, uninstall_commands = commands
     test_args = parse_test_args(testdir)
     gen_start = time.time()
     # Configure in-process
@@ -347,9 +334,9 @@ def _run_test(testdir, test_build_dir, install_dir, extra_args, flags, compile_c
     if returncode != 0:
         return TestResult('Generating the build system failed.', BuildStep.configure, stdo, stde, mesonlog, gen_time)
     # Build with subprocess
-    comp = get_compile_commands_for_dir(compile_commands, test_build_dir)
+    dir_args = get_backend_args_for_dir(backend, test_build_dir)
     build_start = time.time()
-    pc, o, e = Popen_safe(comp, cwd=test_build_dir)
+    pc, o, e = Popen_safe(compile_commands + dir_args, cwd=test_build_dir)
     build_time = time.time() - build_start
     stdo += o
     stde += e
@@ -376,8 +363,8 @@ def _run_test(testdir, test_build_dir, install_dir, extra_args, flags, compile_c
         return TestResult('Test that should have failed to run unit tests succeeded', BuildStep.test, stdo, stde, mesonlog, gen_time)
     if returncode != 0:
         return TestResult('Running unit tests failed.', BuildStep.test, stdo, stde, mesonlog, gen_time, build_time, test_time)
-    if len(install_commands) == 0:
+    # Do installation, if the backend supports it
-        return TestResult('', BuildStep.install, '', '', mesonlog, gen_time, build_time, test_time)
+    if len(install_commands) != 0:
         env = os.environ.copy()
         env['DESTDIR'] = install_dir
         # Install with subprocess
@@ -386,15 +373,16 @@ def _run_test(testdir, test_build_dir, install_dir, extra_args, flags, compile_c
         stde += e
         if pi.returncode != 0:
             return TestResult('Running install failed.', BuildStep.install, stdo, stde, mesonlog, gen_time, build_time, test_time)
-    if len(clean_commands) != 0:
-        env = os.environ.copy()
     # Clean with subprocess
-        pi, o, e = Popen_safe(clean_commands, cwd=test_build_dir, env=env)
+    env = os.environ.copy()
+    pi, o, e = Popen_safe(clean_commands + dir_args, cwd=test_build_dir, env=env)
     stdo += o
     stde += e
     if pi.returncode != 0:
         return TestResult('Running clean failed.', BuildStep.clean, stdo, stde, mesonlog, gen_time, build_time, test_time)
-    return TestResult(validate_install(testdir, install_dir), BuildStep.validate, stdo, stde, mesonlog, gen_time, build_time, test_time)
+    if len(install_commands) == 0:
+        return TestResult('', BuildStep.install, '', '', mesonlog, gen_time, build_time, test_time)
+    return TestResult(validate_install(testdir, install_dir, compiler), BuildStep.validate, stdo, stde, mesonlog, gen_time, build_time, test_time)
 
 def gather_tests(testdir):
     tests = [t.replace('\\', '/').split('/', 2)[2] for t in glob(os.path.join(testdir, '*'))]
@@ -454,6 +442,7 @@ def run_tests(all_tests, log_name_base, extra_args):
     passing_tests = 0
     failing_tests = 0
     skipped_tests = 0
+    commands = (compile_commands, clean_commands, install_commands, uninstall_commands)
 
     try:
         # This fails in some CI environments for unknown reasons.
@@ -483,7 +472,7 @@ def run_tests(all_tests, log_name_base, extra_args):
             should_fail = False
             if name.startswith('failing'):
                 should_fail = name.split('failing-')[1]
-            result = executor.submit(run_test, skipped, t, extra_args, backend_flags, compile_commands, should_fail)
+            result = executor.submit(run_test, skipped, t, extra_args, compiler, backend, backend_flags, commands, should_fail)
             futures.append((testname, t, result))
         for (testname, t, result) in futures:
             sys.stdout.flush()
@@ -598,7 +587,7 @@ def generate_prebuilt():
     return objectfile, stlibfile
 
 def check_meson_commands_work():
-    global meson_command, compile_commands, test_commands, install_commands
+    global backend, meson_command, compile_commands, test_commands, install_commands
     testdir = 'test cases/common/1 trivial'
     with AutoDeletedDir(tempfile.mkdtemp(prefix='b ', dir='.')) as build_dir:
         print('Checking that configuring works...')
@@ -607,8 +596,8 @@ def check_meson_commands_work():
         if pc.returncode != 0:
             raise RuntimeError('Failed to configure {!r}:\n{}\n{}'.format(testdir, e, o))
         print('Checking that building works...')
-        compile_cmd = get_compile_commands_for_dir(compile_commands, build_dir)
+        dir_args = get_backend_args_for_dir(backend, build_dir)
-        pc, o, e = Popen_safe(compile_cmd, cwd=build_dir)
+        pc, o, e = Popen_safe(compile_commands + dir_args, cwd=build_dir)
         if pc.returncode != 0:
             raise RuntimeError('Failed to build {!r}:\n{}\n{}'.format(testdir, e, o))
         print('Checking that testing works...')
@@ -628,8 +617,7 @@ if __name__ == '__main__':
     parser.add_argument('--backend', default=None, dest='backend',
                         choices=backendlist)
     options = parser.parse_args()
-    backend = options.backend
+    setup_commands(options.backend)
-    setup_commands()
 
     # Appveyor sets the `platform` environment variable which completely messes
     # up building with the vs2010 and vs2015 backends.

run_tests.py

@@ -22,6 +22,7 @@ import platform
 from mesonbuild import mesonlib
 from mesonbuild.environment import detect_ninja
 from enum import Enum
+from glob import glob
 
 Backend = Enum('Backend', 'ninja vs xcode')
 
@@ -30,6 +31,15 @@ if mesonlib.is_windows():
 else:
     exe_suffix = ''
 
+def get_backend_args_for_dir(backend, builddir):
+    '''
+    Visual Studio backend needs to be given the solution to build
+    '''
+    if backend.startswith('vs'):
+        sln_name = glob(os.path.join(builddir, '*.sln'))[0]
+        return [os.path.split(sln_name)[-1]]
+    return []
+
 def get_build_target_args(backend, target):
     if target is None:
         return []
@@ -44,7 +54,7 @@ def get_backend_commands(backend, debug=False):
     uninstall_cmd = []
     if backend.startswith('vs'):
         cmd = ['msbuild']
-        clean_cmd = []
+        clean_cmd = cmd + ['/target:Clean']
         test_cmd = cmd + ['RUN_TESTS.vcxproj']
     elif backend == 'xcode':
         cmd = ['xcodebuild']