nihilist/tor
1
0
Fork 0
mirror of https://gitlab.torproject.org/tpo/core/tor.git synced 2024-11-11 13:43:47 +01:00
Code Issues Packages Projects Releases Wiki Activity
e3bf8854c8
tor/src/test/test.h

251 lines
9.5 KiB
C
Raw Normal View History

Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
/* Copyright (c) 2001-2003, Roger Dingledine.
* Copyright (c) 2004-2006, Roger Dingledine, Nick Mathewson.
Update the copyright year.
2016-02-27 18:48:19 +01:00
* Copyright (c) 2007-2016, The Tor Project, Inc. */
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
/* See LICENSE for licensing information */
Convert all include-guard macros to avoid reserved identifiers. In C, we technically aren't supposed to define our own things that start with an underscore. This is a purely machine-generated commit. First, I ran this script on all the headers in src/{common,or,test,tools/*}/*.h : ============================== use strict; my %macros = (); my %skipped = (); FILE: for my $fn (@ARGV) { my $f = $fn; if ($fn !~ /^\.\//) { $f = "./$fn"; } $skipped{$fn} = 0; open(F, $fn); while (<F>) { if (/^#ifndef ([A-Za-z0-9_]+)/) { $macros{$fn} = $1; next FILE; } } } print "#!/usr/bin/perl -w -i -p\n\n"; for my $fn (@ARGV) { if (! exists $macros{$fn}) { print "# No macro known for $fn!\n" if (!$skipped{$fn}); next; } if ($macros{$fn} !~ /_H_?$/) { print "# Weird macro for $fn...\n"; } my $goodmacro = uc $fn; $goodmacro =~ s#.*/##; $goodmacro =~ s#[\/\-\.]#_#g; print "s/(?<![A-Za-z0-9_])$macros{$fn}(?![A-Za-z0-9_])/TOR_${goodmacro}/g;\n" } ============================== It produced the following output, which I then re-ran on those same files: ============================== s/(?<![A-Za-z0-9_])_TOR_ADDRESS_H(?![A-Za-z0-9_])/TOR_ADDRESS_H/g; s/(?<![A-Za-z0-9_])_TOR_AES_H(?![A-Za-z0-9_])/TOR_AES_H/g; s/(?<![A-Za-z0-9_])_TOR_COMPAT_H(?![A-Za-z0-9_])/TOR_COMPAT_H/g; s/(?<![A-Za-z0-9_])_TOR_COMPAT_LIBEVENT_H(?![A-Za-z0-9_])/TOR_COMPAT_LIBEVENT_H/g; s/(?<![A-Za-z0-9_])_TOR_CONTAINER_H(?![A-Za-z0-9_])/TOR_CONTAINER_H/g; s/(?<![A-Za-z0-9_])_TOR_CRYPTO_H(?![A-Za-z0-9_])/TOR_CRYPTO_H/g; s/(?<![A-Za-z0-9_])TOR_DI_OPS_H(?![A-Za-z0-9_])/TOR_DI_OPS_H/g; s/(?<![A-Za-z0-9_])_TOR_MEMAREA_H(?![A-Za-z0-9_])/TOR_MEMAREA_H/g; s/(?<![A-Za-z0-9_])_TOR_MEMPOOL_H(?![A-Za-z0-9_])/TOR_MEMPOOL_H/g; s/(?<![A-Za-z0-9_])TOR_PROCMON_H(?![A-Za-z0-9_])/TOR_PROCMON_H/g; s/(?<![A-Za-z0-9_])_TOR_TORGZIP_H(?![A-Za-z0-9_])/TOR_TORGZIP_H/g; s/(?<![A-Za-z0-9_])_TOR_TORINT_H(?![A-Za-z0-9_])/TOR_TORINT_H/g; s/(?<![A-Za-z0-9_])_TOR_LOG_H(?![A-Za-z0-9_])/TOR_TORLOG_H/g; s/(?<![A-Za-z0-9_])_TOR_TORTLS_H(?![A-Za-z0-9_])/TOR_TORTLS_H/g; s/(?<![A-Za-z0-9_])_TOR_UTIL_H(?![A-Za-z0-9_])/TOR_UTIL_H/g; s/(?<![A-Za-z0-9_])_TOR_BUFFERS_H(?![A-Za-z0-9_])/TOR_BUFFERS_H/g; s/(?<![A-Za-z0-9_])_TOR_CHANNEL_H(?![A-Za-z0-9_])/TOR_CHANNEL_H/g; s/(?<![A-Za-z0-9_])_TOR_CHANNEL_TLS_H(?![A-Za-z0-9_])/TOR_CHANNELTLS_H/g; s/(?<![A-Za-z0-9_])_TOR_CIRCUITBUILD_H(?![A-Za-z0-9_])/TOR_CIRCUITBUILD_H/g; s/(?<![A-Za-z0-9_])_TOR_CIRCUITLIST_H(?![A-Za-z0-9_])/TOR_CIRCUITLIST_H/g; s/(?<![A-Za-z0-9_])_TOR_CIRCUITMUX_EWMA_H(?![A-Za-z0-9_])/TOR_CIRCUITMUX_EWMA_H/g; s/(?<![A-Za-z0-9_])_TOR_CIRCUITMUX_H(?![A-Za-z0-9_])/TOR_CIRCUITMUX_H/g; s/(?<![A-Za-z0-9_])_TOR_CIRCUITUSE_H(?![A-Za-z0-9_])/TOR_CIRCUITUSE_H/g; s/(?<![A-Za-z0-9_])_TOR_COMMAND_H(?![A-Za-z0-9_])/TOR_COMMAND_H/g; s/(?<![A-Za-z0-9_])_TOR_CONFIG_H(?![A-Za-z0-9_])/TOR_CONFIG_H/g; s/(?<![A-Za-z0-9_])TOR_CONFPARSE_H(?![A-Za-z0-9_])/TOR_CONFPARSE_H/g; s/(?<![A-Za-z0-9_])_TOR_CONNECTION_EDGE_H(?![A-Za-z0-9_])/TOR_CONNECTION_EDGE_H/g; s/(?<![A-Za-z0-9_])_TOR_CONNECTION_H(?![A-Za-z0-9_])/TOR_CONNECTION_H/g; s/(?<![A-Za-z0-9_])_TOR_CONNECTION_OR_H(?![A-Za-z0-9_])/TOR_CONNECTION_OR_H/g; s/(?<![A-Za-z0-9_])_TOR_CONTROL_H(?![A-Za-z0-9_])/TOR_CONTROL_H/g; s/(?<![A-Za-z0-9_])_TOR_CPUWORKER_H(?![A-Za-z0-9_])/TOR_CPUWORKER_H/g; s/(?<![A-Za-z0-9_])_TOR_DIRECTORY_H(?![A-Za-z0-9_])/TOR_DIRECTORY_H/g; s/(?<![A-Za-z0-9_])_TOR_DIRSERV_H(?![A-Za-z0-9_])/TOR_DIRSERV_H/g; s/(?<![A-Za-z0-9_])_TOR_DIRVOTE_H(?![A-Za-z0-9_])/TOR_DIRVOTE_H/g; s/(?<![A-Za-z0-9_])_TOR_DNS_H(?![A-Za-z0-9_])/TOR_DNS_H/g; s/(?<![A-Za-z0-9_])_TOR_DNSSERV_H(?![A-Za-z0-9_])/TOR_DNSSERV_H/g; s/(?<![A-Za-z0-9_])TOR_EVENTDNS_TOR_H(?![A-Za-z0-9_])/TOR_EVENTDNS_TOR_H/g; s/(?<![A-Za-z0-9_])_TOR_GEOIP_H(?![A-Za-z0-9_])/TOR_GEOIP_H/g; s/(?<![A-Za-z0-9_])_TOR_HIBERNATE_H(?![A-Za-z0-9_])/TOR_HIBERNATE_H/g; s/(?<![A-Za-z0-9_])_TOR_MAIN_H(?![A-Za-z0-9_])/TOR_MAIN_H/g; s/(?<![A-Za-z0-9_])_TOR_MICRODESC_H(?![A-Za-z0-9_])/TOR_MICRODESC_H/g; s/(?<![A-Za-z0-9_])_TOR_NETWORKSTATUS_H(?![A-Za-z0-9_])/TOR_NETWORKSTATUS_H/g; s/(?<![A-Za-z0-9_])_TOR_NODELIST_H(?![A-Za-z0-9_])/TOR_NODELIST_H/g; s/(?<![A-Za-z0-9_])_TOR_NTMAIN_H(?![A-Za-z0-9_])/TOR_NTMAIN_H/g; s/(?<![A-Za-z0-9_])_TOR_ONION_H(?![A-Za-z0-9_])/TOR_ONION_H/g; s/(?<![A-Za-z0-9_])_TOR_OR_H(?![A-Za-z0-9_])/TOR_OR_H/g; s/(?<![A-Za-z0-9_])_TOR_POLICIES_H(?![A-Za-z0-9_])/TOR_POLICIES_H/g; s/(?<![A-Za-z0-9_])_TOR_REASONS_H(?![A-Za-z0-9_])/TOR_REASONS_H/g; s/(?<![A-Za-z0-9_])_TOR_RELAY_H(?![A-Za-z0-9_])/TOR_RELAY_H/g; s/(?<![A-Za-z0-9_])_TOR_RENDCLIENT_H(?![A-Za-z0-9_])/TOR_RENDCLIENT_H/g; s/(?<![A-Za-z0-9_])_TOR_RENDCOMMON_H(?![A-Za-z0-9_])/TOR_RENDCOMMON_H/g; s/(?<![A-Za-z0-9_])_TOR_RENDMID_H(?![A-Za-z0-9_])/TOR_RENDMID_H/g; s/(?<![A-Za-z0-9_])_TOR_RENDSERVICE_H(?![A-Za-z0-9_])/TOR_RENDSERVICE_H/g; s/(?<![A-Za-z0-9_])_TOR_REPHIST_H(?![A-Za-z0-9_])/TOR_REPHIST_H/g; s/(?<![A-Za-z0-9_])_TOR_REPLAYCACHE_H(?![A-Za-z0-9_])/TOR_REPLAYCACHE_H/g; s/(?<![A-Za-z0-9_])_TOR_ROUTER_H(?![A-Za-z0-9_])/TOR_ROUTER_H/g; s/(?<![A-Za-z0-9_])_TOR_ROUTERLIST_H(?![A-Za-z0-9_])/TOR_ROUTERLIST_H/g; s/(?<![A-Za-z0-9_])_TOR_ROUTERPARSE_H(?![A-Za-z0-9_])/TOR_ROUTERPARSE_H/g; s/(?<![A-Za-z0-9_])TOR_ROUTERSET_H(?![A-Za-z0-9_])/TOR_ROUTERSET_H/g; s/(?<![A-Za-z0-9_])TOR_STATEFILE_H(?![A-Za-z0-9_])/TOR_STATEFILE_H/g; s/(?<![A-Za-z0-9_])_TOR_STATUS_H(?![A-Za-z0-9_])/TOR_STATUS_H/g; s/(?<![A-Za-z0-9_])TOR_TRANSPORTS_H(?![A-Za-z0-9_])/TOR_TRANSPORTS_H/g; s/(?<![A-Za-z0-9_])_TOR_TEST_H(?![A-Za-z0-9_])/TOR_TEST_H/g; s/(?<![A-Za-z0-9_])_TOR_FW_HELPER_H(?![A-Za-z0-9_])/TOR_TOR_FW_HELPER_H/g; s/(?<![A-Za-z0-9_])_TOR_FW_HELPER_NATPMP_H(?![A-Za-z0-9_])/TOR_TOR_FW_HELPER_NATPMP_H/g; s/(?<![A-Za-z0-9_])_TOR_FW_HELPER_UPNP_H(?![A-Za-z0-9_])/TOR_TOR_FW_HELPER_UPNP_H/g; ==============================
2012-10-12 18:13:10 +02:00
#ifndef TOR_TEST_H
#define TOR_TEST_H
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
/**
* \file test.h
Split crypto tests into a separate module.
2009-09-22 19:06:47 +02:00
* \brief Macros and functions used by unit tests.
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
*/
#include "compat.h"
#include "tinytest.h"
#define TT_EXIT_TEST_FUNCTION STMT_BEGIN goto done; STMT_END
#include "tinytest_macros.h"
#ifdef __GNUC__
#define PRETTY_FUNCTION __PRETTY_FUNCTION__
#else
#define PRETTY_FUNCTION ""
#endif
Improved fix for test_memeq_hex leak. The earlier fix would only handle the success case. In the failing case, test_mem_op does a goto done, which would leave the leak leaking.
2009-09-27 18:07:33 +02:00
/* As test_mem_op, but decodes 'hex' before comparing. There must be a
* local char* variable called mem_op_hex_tmp for this to work. */
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
#define test_mem_op_hex(expr1, op, hex) \
STMT_BEGIN \
size_t length = strlen(hex); \
Improved fix for test_memeq_hex leak. The earlier fix would only handle the success case. In the failing case, test_mem_op does a goto done, which would leave the leak leaking.
2009-09-27 18:07:33 +02:00
tor_free(mem_op_hex_tmp); \
mem_op_hex_tmp = tor_malloc(length/2); \
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
tor_assert((length&1)==0); \
Improved fix for test_memeq_hex leak. The earlier fix would only handle the success case. In the failing case, test_mem_op does a goto done, which would leave the leak leaking.
2009-09-27 18:07:33 +02:00
base16_decode(mem_op_hex_tmp, length/2, hex, length); \
Replace remaining uses of deprecated test_mem* functions
2014-09-16 03:27:23 +02:00
tt_mem_op(expr1, op, mem_op_hex_tmp, length/2); \
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
STMT_END
Replace operators used as macro arguments with OP_XX macros Part of fix for 13172
2014-11-12 19:28:07 +01:00
#define test_memeq_hex(expr1, hex) test_mem_op_hex(expr1, OP_EQ, hex)
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
Refactor the core of choosing by weights into a function This eliminates duplicated code, and lets us test a hairy piece of functionality.
2012-08-09 19:47:42 +02:00
#define tt_double_op(a,op,b) \
Update test_status.c to accommodate changes in heartbeat messages Fixes #15012; bug not in any released Tor
2015-02-24 16:05:34 +01:00
tt_assert_test_type(a,b,#a" "#op" "#b,double,(val1_ op val2_),"%g", \
Refactor the core of choosing by weights into a function This eliminates duplicated code, and lets us test a hairy piece of functionality.
2012-08-09 19:47:42 +02:00
TT_EXIT_TEST_FUNCTION)
Fix an overzealous compiler warning in the tests
2015-08-14 14:48:26 +02:00
/* Declare "double equal" in a sneaky way, so compiler won't complain about
* comparing floats with == or !=. Of course, only do this if you know what
* you're doing. */
#define tt_double_eq(a,b) \
STMT_BEGIN \
tt_double_op((a), >=, (b)); \
tt_double_op((a), <=, (b)); \
STMT_END
Fix numerous 64->32 errors in the unit tests Before the 11825 fix, these were all silently ignored.
2014-05-08 20:01:17 +02:00
#ifdef _MSC_VER
#define U64_PRINTF_TYPE uint64_t
Apply an MSVC compilation fix from Gisle Vanem This fixes a double-define introduced in 28538069b2f1909a7600ec6d
2014-08-13 21:11:00 +02:00
#define I64_PRINTF_TYPE int64_t
Fix numerous 64->32 errors in the unit tests Before the 11825 fix, these were all silently ignored.
2014-05-08 20:01:17 +02:00
#else
#define U64_PRINTF_TYPE unsigned long long
#define I64_PRINTF_TYPE long long
#endif
#define tt_size_op(a,op,b) \
tt_assert_test_fmt_type(a,b,#a" "#op" "#b,size_t,(val1_ op val2_), \
U64_PRINTF_TYPE, U64_FORMAT, \
fix whitespace
2014-05-12 05:40:48 +02:00
{print_ = (U64_PRINTF_TYPE) value_;}, {}, TT_EXIT_TEST_FUNCTION)
Fix numerous 64->32 errors in the unit tests Before the 11825 fix, these were all silently ignored.
2014-05-08 20:01:17 +02:00
#define tt_u64_op(a,op,b) \
tt_assert_test_fmt_type(a,b,#a" "#op" "#b,uint64_t,(val1_ op val2_), \
U64_PRINTF_TYPE, U64_FORMAT, \
fix whitespace
2014-05-12 05:40:48 +02:00
{print_ = (U64_PRINTF_TYPE) value_;}, {}, TT_EXIT_TEST_FUNCTION)
Fix numerous 64->32 errors in the unit tests Before the 11825 fix, these were all silently ignored.
2014-05-08 20:01:17 +02:00
#define tt_i64_op(a,op,b) \
tt_assert_test_fmt_type(a,b,#a" "#op" "#b,int64_t,(val1_ op val2_), \
I64_PRINTF_TYPE, I64_FORMAT, \
fix whitespace
2014-05-12 05:40:48 +02:00
{print_ = (I64_PRINTF_TYPE) value_;}, {}, TT_EXIT_TEST_FUNCTION)
Fix numerous 64->32 errors in the unit tests Before the 11825 fix, these were all silently ignored.
2014-05-08 20:01:17 +02:00
Split crypto tests into a separate module.
2009-09-22 19:06:47 +02:00
const char *get_fname(const char *name);
Basic 'handle' implementation and tests. This abstraction covers the case where one part of the program needs to refer to another object that is allowed to disappear.
2016-02-12 00:34:29 +01:00
struct crypto_pk_t *pk_generate(int idx);
Split crypto tests into a separate module.
2009-09-22 19:06:47 +02:00
Uplift status.c unit test coverage with new test cases and macros. A new set of unit test cases are provided, as well as introducing an alternative paradigm and macros to support it. Primarily, each test case is given its own namespace, in order to isolate tests from each other. We do this by in the usual fashion, by appending module and submodule names to our symbols. New macros assist by reducing friction for this and other tasks, like overriding a function in the global namespace with one in the current namespace, or declaring integer variables to assist tracking how many times a mock has been called. A set of tests for a small-scale module has been included in this commit, in order to highlight how the paradigm can be used. This suite gives 100% coverage to status.c in test execution.
2014-04-15 14:20:34 +02:00
#define US2_CONCAT_2__(a, b) a ## __ ## b
#define US_CONCAT_2__(a, b) a ## _ ## b
#define US_CONCAT_3__(a, b, c) a ## _ ## b ## _ ## c
#define US_CONCAT_2_(a, b) US_CONCAT_2__(a, b)
#define US_CONCAT_3_(a, b, c) US_CONCAT_3__(a, b, c)
/*
* These macros are helpful for streamlining the authorship of several test
* cases that use mocks.
*
* The pattern is as follows.
* * Declare a top level namespace:
* #define NS_MODULE foo
*
* * For each test case you want to write, create a new submodule in the
* namespace. All mocks and other information should belong to a single
* submodule to avoid interference with other test cases.
* You can simply name the submodule after the function in the module you
* are testing:
* #define NS_SUBMODULE some_function
* or, if you're wanting to write several tests against the same function,
* ie., you are testing an aspect of that function, you can use:
* #define NS_SUBMODULE ASPECT(some_function, behavior)
*
* * Declare all the mocks you will use. The NS_DECL macro serves to declare
* the mock in the current namespace (defined by NS_MODULE and NS_SUBMODULE).
* It behaves like MOCK_DECL:
* NS_DECL(int, dependent_function, (void *));
* Here, dependent_function must be declared and implemented with the
* MOCK_DECL and MOCK_IMPL macros. The NS_DECL macro also defines an integer
* global for use for tracking how many times a mock was called, and can be
* accessed by CALLED(mock_name). For example, you might put
* CALLED(dependent_function)++;
* in your mock body.
*
* * Define a function called NS(main) that will contain the body of the
* test case. The NS macro can be used to reference a name in the current
* namespace.
*
* * In NS(main), indicate that a mock function in the current namespace,
* declared with NS_DECL is to override that in the global namespace,
* with the NS_MOCK macro:
* NS_MOCK(dependent_function)
* Unmock with:
* NS_UNMOCK(dependent_function)
*
* * Define the mocks with the NS macro, eg.,
* int
* NS(dependent_function)(void *)
* {
* CALLED(dependent_function)++;
* }
*
* * In the struct testcase_t array, you can use the TEST_CASE and
* TEST_CASE_ASPECT macros to define the cases without having to do so
* explicitly nor without having to reset NS_SUBMODULE, eg.,
* struct testcase_t foo_tests[] = {
* TEST_CASE_ASPECT(some_function, behavior),
* ...
* END_OF_TESTCASES
* which will define a test case named "some_function__behavior".
*/
#define NAME_TEST_(name) #name
#define NAME_TEST(name) NAME_TEST_(name)
#define ASPECT(test_module, test_name) US2_CONCAT_2__(test_module, test_name)
#define TEST_CASE(function) \
{ \
NAME_TEST(function), \
NS_FULL(NS_MODULE, function, test_main), \
TT_FORK, \
NULL, \
NULL, \
}
#define TEST_CASE_ASPECT(function, aspect) \
{ \
NAME_TEST(ASPECT(function, aspect)), \
NS_FULL(NS_MODULE, ASPECT(function, aspect), test_main), \
TT_FORK, \
NULL, \
NULL, \
}
#define NS(name) US_CONCAT_3_(NS_MODULE, NS_SUBMODULE, name)
#define NS_FULL(module, submodule, name) US_CONCAT_3_(module, submodule, name)
#define CALLED(mock_name) US_CONCAT_2_(NS(mock_name), called)
#define NS_DECL(retval, mock_fn, args) \
Add -Wmissing-variable-declarations, with attendant fixes This is a big-ish patch, but it's very straightforward. Under this clang warning, we're not actually allowed to have a global variable without a previous extern declaration for it. The cases where we violated this rule fall into three roughly equal groups: * Stuff that should have been static. * Stuff that was global but where the extern was local to some other C file. * Stuff that was only global when built for the unit tests, that needed a conditional extern in the headers. The first two were IMO genuine problems; the last is a wart of how we build tests.
2016-06-02 15:46:12 +02:00
extern int CALLED(mock_fn); \
Uplift status.c unit test coverage with new test cases and macros. A new set of unit test cases are provided, as well as introducing an alternative paradigm and macros to support it. Primarily, each test case is given its own namespace, in order to isolate tests from each other. We do this by in the usual fashion, by appending module and submodule names to our symbols. New macros assist by reducing friction for this and other tasks, like overriding a function in the global namespace with one in the current namespace, or declaring integer variables to assist tracking how many times a mock has been called. A set of tests for a small-scale module has been included in this commit, in order to highlight how the paradigm can be used. This suite gives 100% coverage to status.c in test execution.
2014-04-15 14:20:34 +02:00
static retval NS(mock_fn) args; int CALLED(mock_fn) = 0
#define NS_MOCK(name) MOCK(name, NS(name))
#define NS_UNMOCK(name) UNMOCK(name)
Basic unit test for condition variables.
2013-09-28 05:15:53 +02:00
extern const struct testcase_setup_t passthrough_setup;
Add -Wmissing-variable-declarations, with attendant fixes This is a big-ish patch, but it's very straightforward. Under this clang warning, we're not actually allowed to have a global variable without a previous extern declaration for it. The cases where we violated this rule fall into three roughly equal groups: * Stuff that should have been static. * Stuff that was global but where the extern was local to some other C file. * Stuff that was only global when built for the unit tests, that needed a conditional extern in the headers. The first two were IMO genuine problems; the last is a wart of how we build tests.
2016-06-02 15:46:12 +02:00
extern struct testcase_t accounting_tests[];
extern struct testcase_t addr_tests[];
extern struct testcase_t address_tests[];
extern struct testcase_t buffer_tests[];
extern struct testcase_t cell_format_tests[];
extern struct testcase_t cell_queue_tests[];
extern struct testcase_t channel_tests[];
extern struct testcase_t channeltls_tests[];
extern struct testcase_t checkdir_tests[];
extern struct testcase_t circuitlist_tests[];
extern struct testcase_t circuitmux_tests[];
extern struct testcase_t compat_libevent_tests[];
extern struct testcase_t config_tests[];
extern struct testcase_t connection_tests[];
extern struct testcase_t container_tests[];
extern struct testcase_t controller_tests[];
extern struct testcase_t controller_event_tests[];
extern struct testcase_t crypto_tests[];
extern struct testcase_t dir_tests[];
extern struct testcase_t dir_handle_get_tests[];
extern struct testcase_t entryconn_tests[];
extern struct testcase_t entrynodes_tests[];
extern struct testcase_t guardfraction_tests[];
extern struct testcase_t extorport_tests[];
extern struct testcase_t hs_tests[];
extern struct testcase_t introduce_tests[];
extern struct testcase_t keypin_tests[];
extern struct testcase_t link_handshake_tests[];
extern struct testcase_t logging_tests[];
extern struct testcase_t microdesc_tests[];
extern struct testcase_t nodelist_tests[];
extern struct testcase_t oom_tests[];
Unit test for connection_handle_oos()
2016-07-02 06:39:45 +02:00
extern struct testcase_t oos_tests[];
Add -Wmissing-variable-declarations, with attendant fixes This is a big-ish patch, but it's very straightforward. Under this clang warning, we're not actually allowed to have a global variable without a previous extern declaration for it. The cases where we violated this rule fall into three roughly equal groups: * Stuff that should have been static. * Stuff that was global but where the extern was local to some other C file. * Stuff that was only global when built for the unit tests, that needed a conditional extern in the headers. The first two were IMO genuine problems; the last is a wart of how we build tests.
2016-06-02 15:46:12 +02:00
extern struct testcase_t options_tests[];
extern struct testcase_t policy_tests[];
extern struct testcase_t procmon_tests[];
extern struct testcase_t pubsub_tests[];
extern struct testcase_t pt_tests[];
extern struct testcase_t relay_tests[];
extern struct testcase_t relaycell_tests[];
extern struct testcase_t rend_cache_tests[];
extern struct testcase_t replaycache_tests[];
extern struct testcase_t router_tests[];
extern struct testcase_t routerkeys_tests[];
extern struct testcase_t routerlist_tests[];
extern struct testcase_t routerset_tests[];
extern struct testcase_t scheduler_tests[];
extern struct testcase_t socks_tests[];
extern struct testcase_t status_tests[];
extern struct testcase_t thread_tests[];
extern struct testcase_t tortls_tests[];
extern struct testcase_t util_tests[];
extern struct testcase_t util_format_tests[];
extern struct testcase_t util_process_tests[];
extern struct testcase_t dns_tests[];
extern struct testcase_t handle_tests[];
prop250: Add unit tests Signed-off-by: David Goulet <dgoulet@torproject.org> Signed-off-by: George Kadianakis <desnacked@riseup.net>
2016-05-03 17:42:50 +02:00
extern struct testcase_t sr_tests[];
Add -Wmissing-variable-declarations, with attendant fixes This is a big-ish patch, but it's very straightforward. Under this clang warning, we're not actually allowed to have a global variable without a previous extern declaration for it. The cases where we violated this rule fall into three roughly equal groups: * Stuff that should have been static. * Stuff that was global but where the extern was local to some other C file. * Stuff that was only global when built for the unit tests, that needed a conditional extern in the headers. The first two were IMO genuine problems; the last is a wart of how we build tests.
2016-06-02 15:46:12 +02:00
extern struct testcase_t slow_crypto_tests[];
extern struct testcase_t slow_util_tests[];
extern struct testgroup_t testgroups[];
extern const char AUTHORITY_CERT_1[];
extern const char AUTHORITY_SIGNKEY_1[];
extern const char AUTHORITY_SIGNKEY_A_DIGEST[];
extern const char AUTHORITY_SIGNKEY_A_DIGEST256[];
extern const char AUTHORITY_CERT_2[];
extern const char AUTHORITY_SIGNKEY_2[];
extern const char AUTHORITY_SIGNKEY_B_DIGEST[];
extern const char AUTHORITY_SIGNKEY_B_DIGEST256[];
extern const char AUTHORITY_CERT_3[];
extern const char AUTHORITY_SIGNKEY_3[];
extern const char AUTHORITY_SIGNKEY_C_DIGEST[];
extern const char AUTHORITY_SIGNKEY_C_DIGEST256[];
Refactor unit tests to use the tinytest framework. "Tinytest" is a minimalist C unit testing framework I wrote for Libevent. It supports some generally useful features, like being able to run separate unit tests in their own processes. I tried to do the refactoring to change test.c as little as possible. Thus, we mostly don't call the tinytest macros directly. Instead, the test.h header is now a wrapper on tinytest.h to make our existing test_foo() macros work. The next step(s) here will be: - To break test.c into separate files, each with its own test group. - To look into which things we can test - To refactor the more fiddly tests to use the tinytest macros directly and/or run forked. - To see about writing unit tests for things we couldn't previously test without forking.
2009-09-21 20:23:13 +02:00
#endif
Reference in New Issue Copy Permalink