tools: tc-testing: Introduce plugin architecture

This should be a general test architecture, and yet allow specific
tests to be done.  Introduce a plugin architecture.

An individual test has 4 stages, setup/execute/verify/teardown.  Each
plugin gets a chance to run a function at each stage, plus one call
before all the tests are called ("pre" suite) and one after all the
tests are called ("post" suite).  In addition, just before each
command is executed, the plugin gets a chance to modify the command
using the "adjust_command" hook.  This makes the test suite quite
flexible.

Future patches will take some functionality out of the tdc.py script and
place it in plugins.

To use the plugins, place the implementation in the plugins directory
and run tdc.py.  It will notice the plugins and use them.

Signed-off-by: Brenda J. Butler <bjb@mojatatu.com>
Acked-by: Lucas Bates <lucasb@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

tools/testing/selftests/tc-testing/TdcPlugin.py

@@ -0,0 +1,74 @@
+#!/usr/bin/env python3
+
+class TdcPlugin:
+    def __init__(self):
+        super().__init__()
+        print(' -- {}.__init__'.format(self.sub_class))
+
+    def pre_suite(self, testcount, testidlist):
+        '''run commands before test_runner goes into a test loop'''
+        self.testcount = testcount
+        self.testidlist = testidlist
+        if self.args.verbose > 1:
+            print(' -- {}.pre_suite'.format(self.sub_class))
+
+    def post_suite(self, index):
+        '''run commands after test_runner completes the test loop
+        index is the last ordinal number of test that was attempted'''
+        if self.args.verbose > 1:
+            print(' -- {}.post_suite'.format(self.sub_class))
+
+    def pre_case(self, test_ordinal, testid):
+        '''run commands before test_runner does one test'''
+        if self.args.verbose > 1:
+            print(' -- {}.pre_case'.format(self.sub_class))
+        self.args.testid = testid
+        self.args.test_ordinal = test_ordinal
+
+    def post_case(self):
+        '''run commands after test_runner does one test'''
+        if self.args.verbose > 1:
+            print(' -- {}.post_case'.format(self.sub_class))
+
+    def pre_execute(self):
+        '''run command before test-runner does the execute step'''
+        if self.args.verbose > 1:
+            print(' -- {}.pre_execute'.format(self.sub_class))
+
+    def post_execute(self):
+        '''run command after test-runner does the execute step'''
+        if self.args.verbose > 1:
+            print(' -- {}.post_execute'.format(self.sub_class))
+
+    def adjust_command(self, stage, command):
+        '''adjust the command'''
+        if self.args.verbose > 1:
+            print(' -- {}.adjust_command {}'.format(self.sub_class, stage))
+
+        # if stage == 'pre':
+        #     pass
+        # elif stage == 'setup':
+        #     pass
+        # elif stage == 'execute':
+        #     pass
+        # elif stage == 'verify':
+        #     pass
+        # elif stage == 'teardown':
+        #     pass
+        # elif stage == 'post':
+        #     pass
+        # else:
+        #     pass
+
+        return command
+
+    def add_args(self, parser):
+        '''Get the plugin args from the command line'''
+        self.argparser = parser
+        return self.argparser
+
+    def check_args(self, args, remaining):
+        '''Check that the args are set correctly'''
+        self.args = args
+        if self.args.verbose > 1:
+            print(' -- {}.check_args'.format(self.sub_class))

tools/testing/selftests/tc-testing/creating-plugins/AddingPlugins.txt

@@ -0,0 +1,104 @@
+tdc - Adding plugins for tdc
+
+Author: Brenda J. Butler - bjb@mojatatu.com
+
+ADDING PLUGINS
+--------------
+
+A new plugin should be written in python as a class that inherits from TdcPlugin.
+There are some examples in plugin-lib.
+
+The plugin can be used to add functionality to the test framework,
+such as:
+
+- adding commands to be run before and/or after the test suite
+- adding commands to be run before and/or after the test cases
+- adding commands to be run before and/or after the execute phase of the test cases
+- ability to alter the command to be run in any phase:
+    pre        (the pre-suite stage)
+    prepare
+    execute
+    verify
+    teardown
+    post       (the post-suite stage)
+- ability to add to the command line args, and use them at run time
+
+
+The functions in the class should follow the following interfaces:
+
+    def __init__(self)
+    def pre_suite(self, testcount, testidlist)     # see "PRE_SUITE" below
+    def post_suite(self, ordinal)                  # see "SKIPPING" below
+    def pre_case(self, test_ordinal, testid)       # see "PRE_CASE" below
+    def post_case(self)
+    def pre_execute(self)
+    def post_execute(self)
+    def adjust_command(self, stage, command)       # see "ADJUST" below
+    def add_args(self, parser)                     # see "ADD_ARGS" below
+    def check_args(self, args, remaining)          # see "CHECK_ARGS" below
+
+
+PRE_SUITE
+
+This method takes a testcount (number of tests to be run) and
+testidlist (array of test ids for tests that will be run).  This is
+useful for various things, including when an exception occurs and the
+rest of the tests must be skipped.  The info is stored in the object,
+and the post_suite method can refer to it when dumping the "skipped"
+TAP output.  The tdc.py script will do that for the test suite as
+defined in the test case, but if the plugin is being used to run extra
+tests on each test (eg, check for memory leaks on associated
+co-processes) then that other tap output can be generated in the
+post-suite method using this info passed in to the pre_suite method.
+
+
+SKIPPING
+
+The post_suite method will receive the ordinal number of the last
+test to be attempted.  It can use this info when outputting
+the TAP output for the extra test cases.
+
+
+PRE_CASE
+
+The pre_case method will receive the ordinal number of the test
+and the test id.  Useful for outputing the extra test results.
+
+
+ADJUST
+
+The adjust_command method receives a string representing
+the execution stage and a string which is the actual command to be
+executed.  The plugin can adjust the command, based on the stage of
+execution.
+
+The stages are represented by the following strings:
+
+    'pre'
+    'setup'
+    'command'
+    'verify'
+    'teardown'
+    'post'
+
+The adjust_command method must return the adjusted command so tdc
+can use it.
+
+
+ADD_ARGS
+
+The add_args method receives the argparser object and can add
+arguments to it.  Care should be taken that the new arguments do not
+conflict with any from tdc.py or from other plugins that will be used
+concurrently.
+
+The add_args method should return the argparser object.
+
+
+CHECK_ARGS
+
+The check_args method is so that the plugin can do validation on
+the args, if needed.  If there is a problem, and Exception should
+be raised, with a string that explains the problem.
+
+eg:  raise Exception('plugin xxx, arg -y is wrong, fix it')

tools/testing/selftests/tc-testing/plugin-lib/README-PLUGINS

@@ -0,0 +1,27 @@
+tdc.py will look for plugins in a directory plugins off the cwd.
+Make a set of numbered symbolic links from there to the actual plugins.
+Eg:
+
+tdc.py
+plugin-lib/
+plugins/
+    __init__.py
+    10-rootPlugin.py -> ../plugin-lib/rootPlugin.py
+    20-valgrindPlugin.py -> ../plugin-lib/valgrindPlugin.py
+    30-nsPlugin.py -> ../plugin-lib/nsPlugin.py
+
+
+tdc.py will find them and use them.
+
+
+rootPlugin
+    Check if the uid is root.  If not, bail out.
+
+valgrindPlugin
+    Run the command under test with valgrind, and produce an extra set of TAP results for the memory tests.
+    This plugin will write files to the cwd, called vgnd-xxx.log.  These will contain
+    the valgrind output for test xxx.  Any file matching the glob 'vgnd-*.log' will be
+    deleted at the end of the run.
+
+nsPlugin
+    Run all the commands in a network namespace.

tools/testing/selftests/tc-testing/tdc.py

@@ -11,17 +11,91 @@ import re
 import os
 import sys
 import argparse
+import importlib
 import json
 import subprocess
+import time
 from collections import OrderedDict
 from string import Template
 
 from tdc_config import *
 from tdc_helper import *
 
+import TdcPlugin
 
 USE_NS = True
 
+class PluginMgr:
+    def __init__(self, argparser):
+        super().__init__()
+        self.plugins = {}
+        self.plugin_instances = []
+        self.args = []
+        self.argparser = argparser
+
+        # TODO, put plugins in order
+        plugindir = os.getenv('TDC_PLUGIN_DIR', './plugins')
+        for dirpath, dirnames, filenames in os.walk(plugindir):
+            for fn in filenames:
+                if (fn.endswith('.py') and
+                    not fn == '__init__.py' and
+                    not fn.startswith('#') and
+                    not fn.startswith('.#')):
+                    mn = fn[0:-3]
+                    foo = importlib.import_module('plugins.' + mn)
+                    self.plugins[mn] = foo
+                    self.plugin_instances.append(foo.SubPlugin())
+
+    def call_pre_suite(self, testcount, testidlist):
+        for pgn_inst in self.plugin_instances:
+            pgn_inst.pre_suite(testcount, testidlist)
+
+    def call_post_suite(self, index):
+        for pgn_inst in reversed(self.plugin_instances):
+            pgn_inst.post_suite(index)
+
+    def call_pre_case(self, test_ordinal, testid):
+        for pgn_inst in self.plugin_instances:
+            try:
+                pgn_inst.pre_case(test_ordinal, testid)
+            except Exception as ee:
+                print('exception {} in call to pre_case for {} plugin'.
+                      format(ee, pgn_inst.__class__))
+                print('test_ordinal is {}'.format(test_ordinal))
+                print('testid is {}'.format(testid))
+                raise
+
+    def call_post_case(self):
+        for pgn_inst in reversed(self.plugin_instances):
+            pgn_inst.post_case()
+
+    def call_pre_execute(self):
+        for pgn_inst in self.plugin_instances:
+            pgn_inst.pre_execute()
+
+    def call_post_execute(self):
+        for pgn_inst in reversed(self.plugin_instances):
+            pgn_inst.post_execute()
+
+    def call_add_args(self, parser):
+        for pgn_inst in self.plugin_instances:
+            parser = pgn_inst.add_args(parser)
+        return parser
+
+    def call_check_args(self, args, remaining):
+        for pgn_inst in self.plugin_instances:
+            pgn_inst.check_args(args, remaining)
+
+    def call_adjust_command(self, stage, command):
+        for pgn_inst in self.plugin_instances:
+            command = pgn_inst.adjust_command(stage, command)
+        return command
+
+    @staticmethod
+    def _make_argparser(args):
+        self.argparser = argparse.ArgumentParser(
+            description='Linux TC unit tests')
+
 
 def replace_keywords(cmd):
     """
@@ -33,21 +107,27 @@ def replace_keywords(cmd):
     return subcmd
 
 
-def exec_cmd(command, nsonly=True):
+def exec_cmd(args, pm, stage, command, nsonly=True):
     """
     Perform any required modifications on an executable command, then run
     it in a subprocess and return the results.
     """
+    if len(command.strip()) == 0:
+        return None, None
     if (USE_NS and nsonly):
         command = 'ip netns exec $NS ' + command
 
     if '$' in command:
         command = replace_keywords(command)
 
+    command = pm.call_adjust_command(stage, command)
+    if args.verbose > 0:
+        print('command "{}"'.format(command))
     proc = subprocess.Popen(command,
         shell=True,
         stdout=subprocess.PIPE,
-        stderr=subprocess.PIPE)
+        stderr=subprocess.PIPE,
+        env=ENVIR)
     (rawout, serr) = proc.communicate()
 
     if proc.returncode != 0 and len(serr) > 0:
@@ -60,69 +140,85 @@ def exec_cmd(command, nsonly=True):
     return proc, foutput
 
 
-def prepare_env(cmdlist):
+def prepare_env(args, pm, stage, prefix, cmdlist):
     """
-    Execute the setup/teardown commands for a test case. Optionally
+    Execute the setup/teardown commands for a test case.
-    terminate test execution if the command fails.
+    Optionally terminate test execution if the command fails.
     """
+    if args.verbose > 0:
+        print('{}'.format(prefix))
     for cmdinfo in cmdlist:
-        if (type(cmdinfo) == list):
+        if isinstance(cmdinfo, list):
             exit_codes = cmdinfo[1:]
             cmd = cmdinfo[0]
         else:
             exit_codes = [0]
             cmd = cmdinfo
 
-        if (len(cmd) == 0):
+        if not cmd:
             continue
 
-        (proc, foutput) = exec_cmd(cmd)
+        (proc, foutput) = exec_cmd(args, pm, stage, cmd)
 
-        if proc.returncode not in exit_codes:
+        if proc and (proc.returncode not in exit_codes):
-            print
+            print('', file=sys.stderr)
-            print("Could not execute:")
+            print("{} *** Could not execute: \"{}\"".format(prefix, cmd),
-            print(cmd)
+                  file=sys.stderr)
-            print("\nError message:")
+            print("\n{} *** Error message: \"{}\"".format(prefix, foutput),
-            print(foutput)
+                  file=sys.stderr)
-            print("\nAborting test run.")
+            print("\n{} *** Aborting test run.".format(prefix), file=sys.stderr)
-            # ns_destroy()
+            print("\n\n{} *** stdout ***".format(proc.stdout), file=sys.stderr)
-            raise Exception('prepare_env did not complete successfully')
+            print("\n\n{} *** stderr ***".format(proc.stderr), file=sys.stderr)
+            raise Exception('"{}" did not complete successfully'.format(prefix))
 
-def run_one_test(index, tidx):
+def run_one_test(pm, args, index, tidx):
     result = True
     tresult = ""
     tap = ""
+    if args.verbose > 0:
+        print("\t====================\n=====> ", end="")
     print("Test " + tidx["id"] + ": " + tidx["name"])
-    prepare_env(tidx["setup"])
+
-    (p, procout) = exec_cmd(tidx["cmdUnderTest"])
+    pm.call_pre_case(index, tidx['id'])
+    prepare_env(args, pm, 'setup', "-----> prepare stage", tidx["setup"])
+
+    if (args.verbose > 0):
+        print('-----> execute stage')
+    pm.call_pre_execute()
+    (p, procout) = exec_cmd(args, pm, 'execute', tidx["cmdUnderTest"])
     exit_code = p.returncode
+    pm.call_post_execute()
 
     if (exit_code != int(tidx["expExitCode"])):
         result = False
         print("exit:", exit_code, int(tidx["expExitCode"]))
         print(procout)
     else:
-        match_pattern = re.compile(str(tidx["matchPattern"]),
+        if args.verbose > 0:
-                                   re.DOTALL | re.MULTILINE)
+            print('-----> verify stage')
-        (p, procout) = exec_cmd(tidx["verifyCmd"])
+        match_pattern = re.compile(
+            str(tidx["matchPattern"]), re.DOTALL | re.MULTILINE)
+        (p, procout) = exec_cmd(args, pm, 'verify', tidx["verifyCmd"])
         match_index = re.findall(match_pattern, procout)
         if len(match_index) != int(tidx["matchCount"]):
             result = False
 
     if not result:
-        tresult += "not "
+        tresult += 'not '
-    tresult += "ok {} - {} # {}\n".format(str(index), tidx['id'], tidx["name"])
+    tresult += 'ok {} - {} # {}\n'.format(str(index), tidx['id'], tidx['name'])
     tap += tresult
 
     if result == False:
         tap += procout
 
-    prepare_env(tidx["teardown"])
+    prepare_env(args, pm, 'teardown', '-----> teardown stage', tidx['teardown'])
+    pm.call_post_case()
+
     index += 1
 
     return tap
 
-def test_runner(filtered_tests, args):
+def test_runner(pm, args, filtered_tests):
     """
     Driver function for the unit tests.
 
@@ -135,63 +231,71 @@ def test_runner(filtered_tests, args):
     tcount = len(testlist)
     index = 1
     tap = str(index) + ".." + str(tcount) + "\n"
+    badtest = None
 
+    pm.call_pre_suite(tcount, [tidx['id'] for tidx in testlist])
+
+    if args.verbose > 1:
+        print('Run tests here')
     for tidx in testlist:
         if "flower" in tidx["category"] and args.device == None:
             continue
         try:
             badtest = tidx  # in case it goes bad
-            tap += run_one_test(index, tidx)
+            tap += run_one_test(pm, args, index, tidx)
         except Exception as ee:
             print('Exception {} (caught in test_runner, running test {} {} {})'.
                   format(ee, index, tidx['id'], tidx['name']))
             break
         index += 1
 
+    # if we failed in setup or teardown,
+    # fill in the remaining tests with not ok
     count = index
     tap += 'about to flush the tap output if tests need to be skipped\n'
     if tcount + 1 != index:
         for tidx in testlist[index - 1:]:
             msg = 'skipped - previous setup or teardown failed'
-            tap += 'ok {} - {} # {} {} {} \n'.format(
+            tap += 'ok {} - {} # {} {} {}\n'.format(
                 count, tidx['id'], msg, index, badtest.get('id', '--Unknown--'))
             count += 1
 
     tap += 'done flushing skipped test tap output\n'
+    pm.call_post_suite(index)
 
     return tap
 
 
-def ns_create():
+def ns_create(args, pm):
     """
     Create the network namespace in which the tests will be run and set up
     the required network devices for it.
     """
     if (USE_NS):
         cmd = 'ip netns add $NS'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip link add $DEV0 type veth peer name $DEV1'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip link set $DEV1 netns $NS'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip link set $DEV0 up'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip -n $NS link set $DEV1 up'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip link set $DEV2 netns $NS'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
         cmd = 'ip -n $NS link set $DEV2 up'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'pre', cmd, False)
 
 
-def ns_destroy():
+def ns_destroy(args, pm):
     """
     Destroy the network namespace for testing (and any associated network
     devices as well)
     """
     if (USE_NS):
         cmd = 'ip netns delete $NS'
-        exec_cmd(cmd, False)
+        exec_cmd(args, pm, 'post', cmd, False)
 
 
 def has_blank_ids(idlist):
@@ -272,10 +376,10 @@ def set_args(parser):
     return parser
 
 
-def check_default_settings(args):
+def check_default_settings(args, remaining, pm):
     """
-    Process any arguments overriding the default settings, and ensure the
+    Process any arguments overriding the default settings,
-    settings are correct.
+    and ensure the settings are correct.
     """
     # Allow for overriding specific settings
     global NAMES
@@ -288,6 +392,8 @@ def check_default_settings(args):
         print("The specified tc path " + NAMES['TC'] + " does not exist.")
         exit(1)
 
+    pm.call_check_args(args, remaining)
+
 
 def get_id_list(alltests):
     """
@@ -301,16 +407,7 @@ def check_case_id(alltests):
     Check for duplicate test case IDs.
     """
     idl = get_id_list(alltests)
-    # print('check_case_id:  idl is {}'.format(idl))
-    # answer = list()
-    # for x in idl:
-    #     print('Looking at {}'.format(x))
-    #     print('what the heck is idl.count(x)???   {}'.format(idl.count(x)))
-    #     if idl.count(x) > 1:
-    #         answer.append(x)
-    #         print(' ... append it {}'.format(x))
     return [x for x in idl if idl.count(x) > 1]
-    return answer
 
 
 def does_id_exist(alltests, newid):
@@ -403,7 +500,7 @@ def get_test_cases(args):
 
         for ff in args.file:
             if not os.path.isfile(ff):
-                print("IGNORING file " + ff + " \n\tBECAUSE does not exist.")
+                print("IGNORING file " + ff + "\n\tBECAUSE does not exist.")
             else:
                 flist.append(os.path.abspath(ff))
 
@@ -445,7 +542,7 @@ def get_test_cases(args):
     return allcatlist, allidlist, testcases_by_cats, alltestcases
 
 
-def set_operation_mode(args):
+def set_operation_mode(pm, args):
     """
     Load the test case data and process remaining arguments to determine
     what the script should do for this run, and call the appropriate
@@ -486,12 +583,15 @@ def set_operation_mode(args):
         print("This script must be run with root privileges.\n")
         exit(1)
 
-    ns_create()
+    ns_create(args, pm)
 
-    catresults = test_runner(alltests, args)
+    if len(alltests):
+        catresults = test_runner(pm, args, alltests)
+    else:
+        catresults = 'No tests found\n'
     print('All test results: \n\n{}'.format(catresults))
 
-    ns_destroy()
+    ns_destroy(args, pm)
 
 
 def main():
@@ -501,10 +601,15 @@ def main():
     """
     parser = args_parse()
     parser = set_args(parser)
+    pm = PluginMgr(parser)
+    parser = pm.call_add_args(parser)
     (args, remaining) = parser.parse_known_args()
-    check_default_settings(args)
+    args.NAMES = NAMES
+    check_default_settings(args, remaining, pm)
+    if args.verbose > 2:
+        print('args is {}'.format(args))
 
-    set_operation_mode(args)
+    set_operation_mode(pm, args)
 
     exit(0)