* [Tarantool-patches] [PATCH v1 luajit 1/5] test: fix setting of {DY}LD_LIBRARY_PATH variables
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
@ 2023-03-15 16:11 ` Sergey Kaplun via Tarantool-patches
2023-03-20 13:54 ` Maxim Kokryashkin via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 2/5] test: introduce module for C tests Sergey Kaplun via Tarantool-patches
` (4 subsequent siblings)
5 siblings, 1 reply; 12+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2023-03-15 16:11 UTC (permalink / raw)
To: Igor Munkin, Maxim Kokryashkin; +Cc: tarantool-patches
When we set `LUA_TEST_ENV_MORE` variable to be used in the additional
env command for run testing if `"` is used to wrap `LD_LIBRARY_PATH`
value the content of this environment variable is literally
`"/abs/path1:/abs/path2:...:". So, the first entry is treated as the
relative path starting with `"`. In that case if we need to library to
be loaded via FFI for this particular test, that loading fails with the
error "cannot open shared object file", since the path to it is
incorrect.
This patch removes `"` wrapping for the aforementioned variables.
---
You can test the behaviour of the patch with the following diff:
| diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
| index a428d009..86dc916a 100644
| --- a/test/tarantool-tests/CMakeLists.txt
| +++ b/test/tarantool-tests/CMakeLists.txt
| @@ -65,9 +65,9 @@ add_subdirectory(lj-49-bad-lightuserdata)
| add_subdirectory(lj-416-xor-before-jcc)
| add_subdirectory(lj-601-fix-gc-finderrfunc)
| add_subdirectory(lj-727-lightuserdata-itern)
| -add_subdirectory(lj-flush-on-trace)
| add_subdirectory(misclib-getmetrics-capi)
| add_subdirectory(misclib-sysprof-capi)
| +add_subdirectory(lj-flush-on-trace)
|
| # The part of the memory profiler toolchain is located in tools
| # directory, jit, profiler, and bytecode toolchains are located
test/tarantool-tests/CMakeLists.txt | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
index a428d009..38d6ae49 100644
--- a/test/tarantool-tests/CMakeLists.txt
+++ b/test/tarantool-tests/CMakeLists.txt
@@ -102,6 +102,11 @@ endif()
# loaded modules on MacOS instead of shared libraries as it is
# done on Linux and BSD, another environment variable should be
# used to guide <ffi.load> while searching the extension.
+# XXX: Be noticed that we shouldn't use `"` here to wrap
+# the variable's content. If we do this, the variable value will
+# contain `"` at the beginning and the end, so this `"` at the
+# beginning will be treated as the directory for the first entry
+# (the last subdirectory added).
if(CMAKE_SYSTEM_NAME STREQUAL "Darwin")
# XXX: Apple tries their best to "protect their users from
# malware". As a result SIP (see the link[1] below) has been
@@ -122,9 +127,9 @@ if(CMAKE_SYSTEM_NAME STREQUAL "Darwin")
#
# [1]: https://support.apple.com/en-us/HT204899
# [2]: https://developer.apple.com/library/archive/documentation/Security/Conceptual/System_Integrity_Protection_Guide/RuntimeProtections/RuntimeProtections.html
- list(APPEND LUA_TEST_ENV_MORE DYLD_LIBRARY_PATH="${LD_LIBRARY_PATH}")
+ list(APPEND LUA_TEST_ENV_MORE DYLD_LIBRARY_PATH=${LD_LIBRARY_PATH})
else()
- list(APPEND LUA_TEST_ENV_MORE LD_LIBRARY_PATH="${LD_LIBRARY_PATH}")
+ list(APPEND LUA_TEST_ENV_MORE LD_LIBRARY_PATH=${LD_LIBRARY_PATH})
endif()
# LUA_CPATH and LD_LIBRARY_PATH variables and also TESTLIBS list
--
2.34.1
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [Tarantool-patches] [PATCH v1 luajit 1/5] test: fix setting of {DY}LD_LIBRARY_PATH variables
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 1/5] test: fix setting of {DY}LD_LIBRARY_PATH variables Sergey Kaplun via Tarantool-patches
@ 2023-03-20 13:54 ` Maxim Kokryashkin via Tarantool-patches
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-20 13:54 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 866 bytes --]
Hi!
Thanks for the patch!
LGTM, except for the commit message.
>Среда, 15 марта 2023, 19:14 +03:00 от Sergey Kaplun <skaplun@tarantool.org>:
>
>When we set `LUA_TEST_ENV_MORE` variable to be used in the additional
>env command for run testing if `"` is used to wrap `LD_LIBRARY_PATH`
Typo: s/wrap/wrap the
>value the content of this environment variable is literally
>`"/abs/path1:/abs/path2:...:". So, the first entry is treated as the
Typo: I guess you have forgot a ` here.
>relative path starting with `"`. In that case if we need to library to
Typo: s/to library/the library
>be loaded via FFI for this particular test, that loading fails with the
>error "cannot open shared object file", since the path to it is
>incorrect.
>
>This patch removes `"` wrapping for the aforementioned variables.
>---
--
Best regards,
Maxim Kokryashkin
[-- Attachment #2: Type: text/html, Size: 1798 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* [Tarantool-patches] [PATCH v1 luajit 2/5] test: introduce module for C tests
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 1/5] test: fix setting of {DY}LD_LIBRARY_PATH variables Sergey Kaplun via Tarantool-patches
@ 2023-03-15 16:11 ` Sergey Kaplun via Tarantool-patches
2023-03-20 15:17 ` Maxim Kokryashkin via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 3/5] test: introduce utils.h helper " Sergey Kaplun via Tarantool-patches
` (3 subsequent siblings)
5 siblings, 1 reply; 12+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2023-03-15 16:11 UTC (permalink / raw)
To: Igor Munkin, Maxim Kokryashkin; +Cc: tarantool-patches
We need an instrument to write tests in plain C for LuaJIT, to be able:
* easily test LuaC API
* test patches without usage plain Lua
* write unit tests
* startup LuaJIT with custom memory allocator, to test some GC issues
* maybe, in future, use custom hashing function to test a behavior
of LuaJIT tables
and so on.
The <test.c> module serves to achieve these goals without too fancy
features.
It's functionality inspired by cmoka API [1], but only TAP14 [2]
protocol is supported (Version of TAP set to 13 to be compatible with
old TAP13 harnesses).
The group of unit tests is declared like the following:
| void *t_state = NULL;
| const struct test_unit tgroup[] = {
| test_unit_new(test_base),
| test_unit_new(test_subtest),
| };
| return test_run_group(tgroup, t_state);
`test_run_group()` runs the whole group of tests, returns
`TEST_EXIT_SUCCESS` or `TEST_EXIT_FAILURE`.
If a similar group is declared inside unit test, this group will be
considered as a subtest.
This library provides an API similar to glibc (3) `assert()` to use
inside unit tests. `assert_[true,false]()` are useful for condition
checks and `assert_{type}_[not_,]_equal()` are useful for value
comparisons. If some assertion fails diagnostic is set, all test
considered as failing and finished via `longjmp()`, so these assertions
can be used inside custom subroutines.
Also, this module provides ability to skip one test or all tests, mark
test as todo, bail out all tests. `skip()`, `skip_all()` and `todo()`
macros are implemented via an early return to be used only in the test
body to make skipping clear. `skip_all()` may be used both for the
parent test and for a subtest.
As a part of this commit, tarantool-c-tests directory is created with
the corresponding CMakeLists.txt file to build this test library.
Tests to be rewritten in C with this library in the next commit and
placed as unit tests are:
* misclib-getmetrics-capi.test.lua
* misclib-sysprof-capi.test.lua
For now the tarantool-c-tests target just build the test library without
new tests to run.
[1]: https://github.com/clibs/cmocka
[2]: https://testanything.org/tap-version-14-specification.html
Part of tarantool/tarantool#7900
---
I left some notes about this test module and I'll be happy to read your
thoughts about them.
* Should we cast to `(void *)` in `assert_ptr_[not_]equal()`? Or it will
be better to notice user about bad type comparisons?
* How often should we flush stdout?
* Obviously we can use `_test_run_group(__func__, NULL, 0, NULL)` with
`test_set_skip_reason()` set to implement `skip_all()` functionality.
Nevertheless, I decided to reimpliment it's logic separately to be
more easily maintained in the future.
test/CMakeLists.txt | 2 +
test/tarantool-c-tests/CMakeLists.txt | 43 +++++
test/tarantool-c-tests/test.c | 251 ++++++++++++++++++++++++++
test/tarantool-c-tests/test.h | 251 ++++++++++++++++++++++++++
4 files changed, 547 insertions(+)
create mode 100644 test/tarantool-c-tests/CMakeLists.txt
create mode 100644 test/tarantool-c-tests/test.c
create mode 100644 test/tarantool-c-tests/test.h
diff --git a/test/CMakeLists.txt b/test/CMakeLists.txt
index a8262b12..47296a22 100644
--- a/test/CMakeLists.txt
+++ b/test/CMakeLists.txt
@@ -48,12 +48,14 @@ separate_arguments(LUAJIT_TEST_COMMAND)
add_subdirectory(LuaJIT-tests)
add_subdirectory(PUC-Rio-Lua-5.1-tests)
add_subdirectory(lua-Harness-tests)
+add_subdirectory(tarantool-c-tests)
add_subdirectory(tarantool-tests)
add_custom_target(${PROJECT_NAME}-test DEPENDS
LuaJIT-tests
PUC-Rio-Lua-5.1-tests
lua-Harness-tests
+ tarantool-c-tests
tarantool-tests
)
diff --git a/test/tarantool-c-tests/CMakeLists.txt b/test/tarantool-c-tests/CMakeLists.txt
new file mode 100644
index 00000000..5ebea441
--- /dev/null
+++ b/test/tarantool-c-tests/CMakeLists.txt
@@ -0,0 +1,43 @@
+find_program(PROVE prove)
+if(NOT PROVE)
+ message(WARNING "`prove' is not found, so tarantool-c-tests target is not generated")
+ return()
+endif()
+
+set(C_TEST_SUFFIX .c_test)
+set(C_TEST_FLAGS --failures --shuffle)
+
+if(CMAKE_VERBOSE_MAKEFILE)
+ list(APPEND C_TEST_FLAGS --verbose)
+endif()
+
+# Build libtest.
+
+set(TEST_LIB_NAME "test")
+add_library(libtest STATIC EXCLUDE_FROM_ALL ${CMAKE_CURRENT_SOURCE_DIR}/test.c)
+target_include_directories(libtest PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})
+set_target_properties(libtest PROPERTIES
+ COMPILE_FLAGS "-Wall -Wextra"
+ OUTPUT_NAME "${TEST_LIB_NAME}"
+ LIBRARY_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
+)
+
+# XXX: For now, just build libtest. The tests to be depended on
+# will be added at the next commit.
+add_custom_target(tarantool-c-tests
+ DEPENDS libluajit libtest
+)
+
+# XXX: For now, run 0 tests. Just verify that libtest was build.
+add_custom_command(TARGET tarantool-c-tests
+ COMMENT "Running Tarantool C tests"
+ COMMAND
+ ${PROVE}
+ ${CMAKE_CURRENT_BINARY_DIR}
+ --ext ${C_TEST_SUFFIX}
+ --jobs ${CMAKE_BUILD_PARALLEL_LEVEL}
+ ${C_TEST_FLAGS}
+ WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
+)
+
+# vim: ft=cmake expandtab shiftwidth=2: tabstop=2:
diff --git a/test/tarantool-c-tests/test.c b/test/tarantool-c-tests/test.c
new file mode 100644
index 00000000..dc63cf3f
--- /dev/null
+++ b/test/tarantool-c-tests/test.c
@@ -0,0 +1,251 @@
+#include "test.h"
+
+/*
+ * Test module, based on TAP 14 specification [1].
+ * [1]: https://testanything.org/tap-version-14-specification.html
+ */
+
+/* Need for `PATH_MAX` in diagnostic definition. */
+#include <limits.h>
+#include <setjmp.h>
+#include <stdarg.h>
+/* Need for `strchr()` in diagnostic parsing. */
+#include <string.h>
+
+/*
+ * Test level: 0 for the parent test, >0 for any subtests.
+ */
+static int level = -1;
+
+/*
+ * The last diagnostic data to be used in the YAML Diagnostic
+ * block.
+ *
+ * Contains filename, line number and failed expression or assert
+ * name and "got" and "expected" fields. All entries are separated
+ * by \n.
+ * The longest field is filename here, so PATH_MAX * 3 as
+ * the diagnostic string length should be enough.
+ *
+ * The first \0 means the end of diagnostic data.
+ *
+ * As far as `strchr()` searches until \0, all previous entries
+ * are suppressed by the last one. If the first byte is \0 --
+ * diagnostic is empty.
+ */
+#define TEST_DIAG_DATA_MAX (PATH_MAX * 3)
+char test_diag_buf[TEST_DIAG_DATA_MAX] = {0};
+
+const char *skip_reason = NULL;
+const char *todo_reason = NULL;
+
+/* Indent for the TAP. 4 spaces is default for subtest. */
+static void indent(void)
+{
+ int i;
+ for (i = 0; i < level; i++)
+ printf(" ");
+}
+
+void test_message(const char *fmt, ...)
+{
+ va_list ap;
+ indent();
+ va_start(ap, fmt);
+ vprintf(fmt, ap);
+ printf("\n");
+ va_end(ap);
+}
+
+static void test_print_tap_version(void)
+{
+ /*
+ * Since several TAP13 parsers in popular usage treat
+ * a repeated Version declaration as an error, even if the
+ * Version is indented, Subtests _should not_ include a
+ * Version, if TAP13 Harness compatibility is
+ * desirable [1].
+ */
+ if (level == 0)
+ test_message("TAP version %d", TAP_VERSION);
+}
+
+static void test_start_comment(const char *t_name)
+{
+ if (level > -1)
+ /*
+ * Inform about starting subtest, easier for
+ * humans to read.
+ * Subtest with a name must be terminated by a
+ * Test Point with a matching Description [1].
+ */
+ test_comment("Subtest: %s", t_name);
+}
+
+void _test_print_skip_all(const char *group_name, const char *reason)
+{
+ test_start_comment(group_name);
+ /*
+ * XXX: This test isn't started yet, so set indent level
+ * manually.
+ */
+ level++;
+ test_print_tap_version();
+ /*
+ * XXX: `SKIP_DIRECTIVE` is not necessary here according
+ * to the TAP14 specification [1], but some harnesses may
+ * fail to parse the output without it.
+ */
+ test_message("1..0" SKIP_DIRECTIVE "%s", reason);
+ level--;
+}
+
+/* Just inform TAP parser how many tests we want to run. */
+static void test_plan(size_t planned)
+{
+ test_message("1..%lu", planned);
+}
+
+/* Human-readable output how many tests/subtests are failed. */
+static void test_finish(size_t planned, size_t failed)
+{
+ const char *t_type = level == 0 ? "tests" : "subtests";
+ if (failed > 0)
+ test_comment("Looks like you failed %lu %s out of %lu",
+ failed, t_type, planned);
+ fflush(stdout);
+}
+
+void test_set_skip_reason(const char *reason)
+{
+ skip_reason = reason;
+}
+
+void test_set_todo_reason(const char *reason)
+{
+ todo_reason = reason;
+}
+
+void test_save_diag_data(const char *fmt, ...)
+{
+ va_list ap;
+ va_start(ap, fmt);
+ vsnprintf(test_diag_buf, TEST_DIAG_DATA_MAX, fmt, ap);
+ va_end(ap);
+}
+
+static void test_clear_diag_data(void)
+{
+ /*
+ * Limit buffer with zero byte to show that there is no
+ * any entry.
+ */
+ test_diag_buf[0] = '\0';
+}
+
+static int test_diagnostic_is_set(void)
+{
+ return test_diag_buf[0] != '\0';
+}
+
+/*
+ * Parse the last diagnostic data entry and print it in YAML
+ * format with the corresponding additional half-indent in TAP
+ * (2 spaces).
+ * Clear diagnostic message to be sure that it's printed once.
+ * XXX: \n separators are changed to \0 during parsing and
+ * printing output for convenience in usage.
+ */
+static void test_diagnostic(void)
+{
+ test_message(" ---");
+ char *ent = test_diag_buf;
+ char *ent_end = NULL;
+ while ((ent_end = strchr(ent, '\n')) != NULL) {
+ char *next_ent = ent_end + 1;
+ /*
+ * Limit string with with the zero byte for
+ * formatted output. Anyway, don't need this \n
+ * anymore.
+ */
+ *ent_end = '\0';
+ test_message(" %s", ent);
+ ent = next_ent;
+ }
+ test_message(" ...");
+ test_clear_diag_data();
+}
+
+static jmp_buf test_run_env;
+
+TEST_NORET void _test_exit(int status)
+{
+ longjmp(test_run_env, status);
+}
+
+static int test_run(const struct test_unit *test, size_t test_number,
+ void *test_state)
+{
+ int status = TEST_EXIT_SUCCESS;
+ /*
+ * Run unit test. Diagnostic in case of failure setup by
+ * helpers assert macros defined in the header.
+ */
+ int jmp_status;
+ if ((jmp_status = setjmp(test_run_env)) == 0) {
+ if (test->f(test_state) != TEST_EXIT_SUCCESS)
+ status = TEST_EXIT_FAILURE;
+ } else {
+ status = jmp_status - TEST_JMP_STATUS_SHIFT;
+ }
+ const char *result = status == TEST_EXIT_SUCCESS ? "ok" : "not ok";
+
+ /*
+ * Format suffix of the test message for SKIP or TODO
+ * directives.
+ */
+#define SUFFIX_SZ 1024
+ char suffix[SUFFIX_SZ] = {0};
+ if (skip_reason) {
+ snprintf(suffix, SUFFIX_SZ, SKIP_DIRECTIVE "%s", skip_reason);
+ skip_reason = NULL;
+ } else if (todo_reason) {
+ /* Prevent count this test as failed. */
+ status = TEST_EXIT_SUCCESS;
+ snprintf(suffix, SUFFIX_SZ, TODO_DIRECTIVE "%s", todo_reason);
+ todo_reason = NULL;
+ }
+#undef SUFFIX_SZ
+
+ test_message("%s %lu - %s%s", result, test_number, test->name,
+ suffix);
+
+ if (status && test_diagnostic_is_set())
+ test_diagnostic();
+ return status;
+}
+
+int _test_run_group(const char *group_name, const struct test_unit *tests,
+ size_t n_tests, void *test_state)
+{
+ test_start_comment(group_name);
+
+ level++;
+ test_print_tap_version();
+
+ test_plan(n_tests);
+
+ size_t n_failed = 0;
+
+ size_t i;
+ for (i = 0; i < n_tests; i++) {
+ size_t test_number = i + 1;
+ /* Return 1 on failure, 0 on success. */
+ n_failed += test_run(&tests[i], test_number, test_state);
+ }
+
+ test_finish(n_tests, n_failed);
+
+ level--;
+ return n_failed > 0 ? TEST_EXIT_FAILURE : TEST_EXIT_SUCCESS;
+}
diff --git a/test/tarantool-c-tests/test.h b/test/tarantool-c-tests/test.h
new file mode 100644
index 00000000..695c5b4d
--- /dev/null
+++ b/test/tarantool-c-tests/test.h
@@ -0,0 +1,251 @@
+#ifndef TEST_H
+#define TEST_H
+
+#include <stdio.h>
+#include <stdlib.h>
+
+/*
+ * Test module, based on TAP 14 specification [1].
+ * [1]: https://testanything.org/tap-version-14-specification.html
+ * Version 13 is set for better compatibility on old machines.
+ *
+ * TODO:
+ * * Helpers assert macros:
+ * - assert_uint_equal if needed
+ * - assert_uint_not_equal if needed
+ * - assert_str_equal if needed
+ * - assert_str_not_equal if needed
+ * - assert_memory_equal if needed
+ * - assert_memory_not_equal if needed
+ * * Pragmas.
+ */
+
+#define TAP_VERSION 13
+
+#define TEST_EXIT_SUCCESS 0
+#define TEST_EXIT_FAILURE 1
+
+#define TEST_JMP_STATUS_SHIFT 2
+#define TEST_LJMP_EXIT_SUCCESS (TEST_EXIT_SUCCESS + TEST_JMP_STATUS_SHIFT)
+#define TEST_LJMP_EXIT_FAILURE (TEST_EXIT_FAILURE + TEST_JMP_STATUS_SHIFT)
+
+#define TEST_NORET __attribute__((noreturn))
+
+typedef int (*test_func)(void *test_state);
+struct test_unit {
+ const char *name;
+ test_func f;
+};
+
+/* Initialize `test_unit` structure. */
+#define test_unit_new(f) {#f, f}
+
+#define lengthof(arr) (sizeof(arr) / sizeof((arr)[0]))
+
+/*
+ * __func__ is the name for a test group, "main" for the parent
+ * test.
+ */
+#define test_run_group(t_arr, t_state) \
+ _test_run_group(__func__, t_arr, lengthof(t_arr), t_state)
+
+#define SKIP_DIRECTIVE " # SKIP "
+#define TODO_DIRECTIVE " # TODO "
+
+/*
+ * XXX: May be implemented as well via
+ * `_test_run_group(__func, NULL, 0, NULL)` and
+ * `test_set_skip_reason` with additional changes in the former.
+ * But the current approach is easier to maintain, as far as we
+ * don't want to interfere different entities.
+ */
+#define skip_all(reason) do { \
+ _test_print_skip_all(__func__, reason); \
+ return TEST_EXIT_SUCCESS; \
+} while (0)
+
+#define skip(reason) do { \
+ test_set_skip_reason(reason); \
+ return TEST_EXIT_SUCCESS; \
+} while (0)
+
+#define todo(reason) do { \
+ test_set_todo_reason(reason); \
+ return TEST_EXIT_FAILURE; \
+} while (0)
+
+#define bail_out(reason) do { \
+ /* \
+ * For backwards compatibility with TAP13 Harnesses, \
+ * Producers _should_ emit a "Bail out!" line at the root \
+ * indentation level whenever a Subtest bails out [1]. \
+ */ \
+ printf("Bail out! %s\n", reason); \
+ exit(TEST_EXIT_FAILURE); \
+} while (0)
+
+/* `fmt` should always be a format string here. */
+#define test_comment(fmt, ...) test_message("# " fmt, __VA_ARGS__)
+
+/*
+ * This is a set of useful assert macros like the standard C
+ * libary's assert(3) macro.
+ *
+ * On an assertion failure an assert macro will save the
+ * diagnostic to the special buffer, to be reported via YAML
+ * Diagnostic block and finish a test function with
+ * `return TEST_EXIT_FAILURE`.
+ *
+ * Due to limitations of the C language `assert_true()` and
+ * `assert_false()` macros can only display the expression that
+ * caused the assertion failure. Type specific assert macros,
+ * `assert_{type}_equal()` and `assert_{type}_not_equal()`, save
+ * the data that caused the assertion failure which increases data
+ * visibility aiding debugging of failing test cases.
+ */
+
+#define LOCATION_FMT "location:\t%s:%d\n"
+#define ASSERT_NAME_FMT(name) "failed_assertion:\t" #name "\n"
+
+#define assert_true(cond) do { \
+ if (!(cond)) { \
+ test_save_diag_data(LOCATION_FMT \
+ "condition_failed:\t'" #cond "'\n", \
+ __FILE__, __LINE__); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_false(cond) assert_true(!(cond))
+
+#define assert_ptr_equal(got, expected) do { \
+ if ((got) != (expected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_ptr_equal) \
+ "got: %p\n" \
+ "expected: %p\n", \
+ __FILE__, __LINE__, (got), (expected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_ptr_not_equal(got, unexpected) do { \
+ if ((got) == (unexpected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_ptr_not_equal) \
+ "got: %p\n" \
+ "unexpected: %p\n", \
+ __FILE__, __LINE__, (got), (unexpected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_int_equal(got, expected) do { \
+ if ((got) != (expected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_int_equal) \
+ "got: %d\n" \
+ "expected: %d\n", \
+ __FILE__, __LINE__, (got), (expected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_int_not_equal(got, unexpected) do { \
+ if ((got) == (unexpected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_int_not_equal) \
+ "got: %d\n" \
+ "unexpected: %d\n", \
+ __FILE__, __LINE__, (got), (unexpected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_sizet_equal(got, expected) do { \
+ if ((got) != (expected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_sizet_equal) \
+ "got: %lu\n" \
+ "expected: %lu\n", \
+ __FILE__, __LINE__, (got), (expected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+#define assert_sizet_not_equal(got, unexpected) do { \
+ if ((got) == (unexpected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_sizet_not_equal) \
+ "got: %lu\n" \
+ "unexpected: %lu\n", \
+ __FILE__, __LINE__, (got), (unexpected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+/* Check that doubles are __exactly__ the same. */
+#define assert_double_equal(got, expected) do { \
+ if ((got) != (expected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_double_equal) \
+ "got: %lf\n" \
+ "expected: %lf\n", \
+ __FILE__, __LINE__, (got), (expected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+/* Check that doubles are not __exactly__ the same. */
+#define assert_double_not_equal(got, unexpected) do { \
+ if ((got) == (unexpected)) { \
+ test_save_diag_data( \
+ LOCATION_FMT \
+ ASSERT_NAME_FMT(assert_double_not_equal) \
+ "got: %lf\n" \
+ "unexpected: %lf\n", \
+ __FILE__, __LINE__, (got), (unexpected) \
+ ); \
+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
+ } \
+} while (0)
+
+/* API declaration. */
+
+/*
+ * Print formatted message with the corresponding indent.
+ * If you want to leave a comment, use `test_comment()` instead.
+ */
+void test_message(const char *fmt, ...);
+
+/* Need for `skip_all()`, please, don't use it. */
+void _test_print_skip_all(const char *group_name, const char *reason);
+/* End test via `longjmp()`, please, don't use it. */
+TEST_NORET void _test_exit(int status);
+
+void test_set_skip_reason(const char *reason);
+void test_set_todo_reason(const char *reason);
+/*
+ * Save formatted diagnostic data. Each entry separated with \n.
+ */
+void test_save_diag_data(const char *fmt, ...);
+
+/* Internal, it is better to use `test_run_group()` instead. */
+int _test_run_group(const char *group_name, const struct test_unit *tests,
+ size_t n_tests, void *test_state);
+
+#endif /* TEST_H */
--
2.34.1
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [Tarantool-patches] [PATCH v1 luajit 2/5] test: introduce module for C tests
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 2/5] test: introduce module for C tests Sergey Kaplun via Tarantool-patches
@ 2023-03-20 15:17 ` Maxim Kokryashkin via Tarantool-patches
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-20 15:17 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 21377 bytes --]
Hi!
Thanks for the patch!
Please consider my comments below.
>Среда, 15 марта 2023, 19:14 +03:00 от Sergey Kaplun <skaplun@tarantool.org>:
>
>We need an instrument to write tests in plain C for LuaJIT, to be able:
>* easily test LuaC API
>* test patches without usage plain Lua
Typo: s/usage/usage of/
>* write unit tests
>* startup LuaJIT with custom memory allocator, to test some GC issues
>* maybe, in future, use custom hashing function to test a behavior
> of LuaJIT tables
>and so on.
>
>The <test.c> module serves to achieve these goals without too fancy
>features.
>
>It's functionality inspired by cmoka API [1], but only TAP14 [2]
>protocol is supported (Version of TAP set to 13 to be compatible with
>old TAP13 harnesses).
>
>The group of unit tests is declared like the following:
>
>| void *t_state = NULL;
>| const struct test_unit tgroup[] = {
>| test_unit_new(test_base),
>| test_unit_new(test_subtest),
>| };
>| return test_run_group(tgroup, t_state);
>
>`test_run_group()` runs the whole group of tests, returns
>`TEST_EXIT_SUCCESS` or `TEST_EXIT_FAILURE`.
>
>If a similar group is declared inside unit test, this group will be
>considered as a subtest.
>
>This library provides an API similar to glibc (3) `assert()` to use
>inside unit tests. `assert_[true,false]()` are useful for condition
>checks and `assert_{type}_[not_,]_equal()` are useful for value
>comparisons. If some assertion fails diagnostic is set, all test
>considered as failing and finished via `longjmp()`, so these assertions
>can be used inside custom subroutines.
>
>Also, this module provides ability to skip one test or all tests, mark
>test as todo, bail out all tests. `skip()`, `skip_all()` and `todo()`
>macros are implemented via an early return to be used only in the test
>body to make skipping clear. `skip_all()` may be used both for the
>parent test and for a subtest.
>
>As a part of this commit, tarantool-c-tests directory is created with
>the corresponding CMakeLists.txt file to build this test library.
>Tests to be rewritten in C with this library in the next commit and
>placed as unit tests are:
>* misclib-getmetrics-capi.test.lua
>* misclib-sysprof-capi.test.lua
>
>For now the tarantool-c-tests target just build the test library without
>new tests to run.
>
>[1]: https://github.com/clibs/cmocka
>[2]: https://testanything.org/tap-version-14-specification.html
>
>Part of tarantool/tarantool#7900
>---
>
>I left some notes about this test module and I'll be happy to read your
>thoughts about them.
>
>* Should we cast to `(void *)` in `assert_ptr_[not_]equal()`? Or it will
> be better to notice user about bad type comparisons?
I believe we should not cast it to the (void *) and notice user. It is the C
language, so if we can prevent some potentially incorrect behavior in
compile time, we should do that.
>* How often should we flush stdout?
It seems to be often enough for now. I believe, it is better
to get used to that module and add additional `flush()`’es
a little bit down the road, if we’ll need that.
>* Obviously we can use `_test_run_group(__func__, NULL, 0, NULL)` with
> `test_set_skip_reason()` set to implement `skip_all()` functionality.
> Nevertheless, I decided to reimpliment it's logic separately to be
> more easily maintained in the future.
If we’ll change the signature to the one I proposed below, the
`_test_run_group(__func__, NULL, 0, NULL)` approach is not
possible anymore, so there is nothing wrong with reimplementation.
>
>
> test/CMakeLists.txt | 2 +
> test/tarantool-c-tests/CMakeLists.txt | 43 +++++
> test/tarantool-c-tests/test.c | 251 ++++++++++++++++++++++++++
> test/tarantool-c-tests/test.h | 251 ++++++++++++++++++++++++++
> 4 files changed, 547 insertions(+)
> create mode 100644 test/tarantool-c-tests/CMakeLists.txt
> create mode 100644 test/tarantool-c-tests/test.c
> create mode 100644 test/tarantool-c-tests/test.h
>
>diff --git a/test/CMakeLists.txt b/test/CMakeLists.txt
>index a8262b12..47296a22 100644
>--- a/test/CMakeLists.txt
>+++ b/test/CMakeLists.txt
>@@ -48,12 +48,14 @@ separate_arguments(LUAJIT_TEST_COMMAND)
> add_subdirectory(LuaJIT-tests)
> add_subdirectory(PUC-Rio-Lua-5.1-tests)
> add_subdirectory(lua-Harness-tests)
>+add_subdirectory(tarantool-c-tests)
> add_subdirectory(tarantool-tests)
>
> add_custom_target(${PROJECT_NAME}-test DEPENDS
> LuaJIT-tests
> PUC-Rio-Lua-5.1-tests
> lua-Harness-tests
>+ tarantool-c-tests
> tarantool-tests
> )
>
>diff --git a/test/tarantool-c-tests/CMakeLists.txt b/test/tarantool-c-tests/CMakeLists.txt
>new file mode 100644
>index 00000000..5ebea441
>--- /dev/null
>+++ b/test/tarantool-c-tests/CMakeLists.txt
>@@ -0,0 +1,43 @@
>+find_program(PROVE prove)
>+if(NOT PROVE)
>+ message(WARNING "`prove' is not found, so tarantool-c-tests target is not generated")
>+ return()
>+endif()
There is the same check in the test/tarantool-tests/CMakeLists.txt. Maybe
we should move it to the higher-level CMake so the lower-level CMakeLists
inherit it.
>+
>+set(C_TEST_SUFFIX .c_test)
>+set(C_TEST_FLAGS --failures --shuffle)
>+
>+if(CMAKE_VERBOSE_MAKEFILE)
>+ list(APPEND C_TEST_FLAGS --verbose)
>+endif()
>+
>+# Build libtest.
>+
>+set(TEST_LIB_NAME "test")
>+add_library(libtest STATIC EXCLUDE_FROM_ALL ${CMAKE_CURRENT_SOURCE_DIR}/test.c)
>+target_include_directories(libtest PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})
>+set_target_properties(libtest PROPERTIES
>+ COMPILE_FLAGS "-Wall -Wextra"
>+ OUTPUT_NAME "${TEST_LIB_NAME}"
>+ LIBRARY_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
>+)
>+
>+# XXX: For now, just build libtest. The tests to be depended on
>+# will be added at the next commit.
Typo: s/at the next/in the next/
>+add_custom_target(tarantool-c-tests
>+ DEPENDS libluajit libtest
>+)
>+
>+# XXX: For now, run 0 tests. Just verify that libtest was build.
Typo: s/was build/was built/
>+add_custom_command(TARGET tarantool-c-tests
>+ COMMENT "Running Tarantool C tests"
>+ COMMAND
>+ ${PROVE}
>+ ${CMAKE_CURRENT_BINARY_DIR}
>+ --ext ${C_TEST_SUFFIX}
>+ --jobs ${CMAKE_BUILD_PARALLEL_LEVEL}
>+ ${C_TEST_FLAGS}
>+ WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
>+)
>+
>+# vim: ft=cmake expandtab shiftwidth=2: tabstop=2:
That change is not necessary.
>diff --git a/test/tarantool-c-tests/test.c b/test/tarantool-c-tests/test.c
>new file mode 100644
>index 00000000..dc63cf3f
>--- /dev/null
>+++ b/test/tarantool-c-tests/test.c
>@@ -0,0 +1,251 @@
>+#include "test.h"
>+
>+/*
>+ * Test module, based on TAP 14 specification [1].
>+ * [1]: https://testanything.org/tap-version-14-specification.html
>+ */
>+
>+/* Need for `PATH_MAX` in diagnostic definition. */
>+#include <limits.h>
>+#include <setjmp.h>
>+#include <stdarg.h>
>+/* Need for `strchr()` in diagnostic parsing. */
`strchr()` is not safe, despite the fact it searches till `\0`.
We should at least replace it with `memchr()`, which has
the explicit constraint for buffer length.
>+#include <string.h>
>+
>+/*
>+ * Test level: 0 for the parent test, >0 for any subtests.
>+ */
>+static int level = -1;
>+
>+/*
>+ * The last diagnostic data to be used in the YAML Diagnostic
>+ * block.
>+ *
>+ * Contains filename, line number and failed expression or assert
>+ * name and "got" and "expected" fields. All entries are separated
>+ * by \n.
>+ * The longest field is filename here, so PATH_MAX * 3 as
>+ * the diagnostic string length should be enough.
>+ *
>+ * The first \0 means the end of diagnostic data.
>+ *
>+ * As far as `strchr()` searches until \0, all previous entries
>+ * are suppressed by the last one. If the first byte is \0 --
>+ * diagnostic is empty.
>+ */
>+#define TEST_DIAG_DATA_MAX (PATH_MAX * 3)
>+char test_diag_buf[TEST_DIAG_DATA_MAX] = {0};
>+
>+const char *skip_reason = NULL;
>+const char *todo_reason = NULL;
>+
>+/* Indent for the TAP. 4 spaces is default for subtest. */
>+static void indent(void)
>+{
>+ int i;
>+ for (i = 0; i < level; i++)
>+ printf(" ");
>+}
>+
>+void test_message(const char *fmt, ...)
>+{
>+ va_list ap;
>+ indent();
>+ va_start(ap, fmt);
>+ vprintf(fmt, ap);
>+ printf("\n");
>+ va_end(ap);
>+}
>+
>+static void test_print_tap_version(void)
>+{
>+ /*
>+ * Since several TAP13 parsers in popular usage treat
>+ * a repeated Version declaration as an error, even if the
>+ * Version is indented, Subtests _should not_ include a
>+ * Version, if TAP13 Harness compatibility is
>+ * desirable [1].
>+ */
>+ if (level == 0)
>+ test_message("TAP version %d", TAP_VERSION);
>+}
>+
>+static void test_start_comment(const char *t_name)
>+{
>+ if (level > -1)
>+ /*
>+ * Inform about starting subtest, easier for
>+ * humans to read.
>+ * Subtest with a name must be terminated by a
>+ * Test Point with a matching Description [1].
>+ */
>+ test_comment("Subtest: %s", t_name);
>+}
>+
>+void _test_print_skip_all(const char *group_name, const char *reason)
>+{
>+ test_start_comment(group_name);
>+ /*
>+ * XXX: This test isn't started yet, so set indent level
>+ * manually.
>+ */
>+ level++;
>+ test_print_tap_version();
>+ /*
>+ * XXX: `SKIP_DIRECTIVE` is not necessary here according
>+ * to the TAP14 specification [1], but some harnesses may
>+ * fail to parse the output without it.
>+ */
>+ test_message("1..0" SKIP_DIRECTIVE "%s", reason);
>+ level--;
>+}
>+
>+/* Just inform TAP parser how many tests we want to run. */
>+static void test_plan(size_t planned)
>+{
>+ test_message("1..%lu", planned);
>+}
>+
>+/* Human-readable output how many tests/subtests are failed. */
>+static void test_finish(size_t planned, size_t failed)
>+{
>+ const char *t_type = level == 0 ? "tests" : "subtests";
>+ if (failed > 0)
>+ test_comment("Looks like you failed %lu %s out of %lu",
>+ failed, t_type, planned);
Side note: «Looks like» is a bit misleading, it seems like we are
not sure whether the tests failed or not. I propose to rephrase it
in a more strict fashion: «Failed %lu out of %lu».
>+ fflush(stdout);
>+}
>+
>+void test_set_skip_reason(const char *reason)
>+{
>+ skip_reason = reason;
>+}
>+
>+void test_set_todo_reason(const char *reason)
>+{
>+ todo_reason = reason;
>+}
>+
>+void test_save_diag_data(const char *fmt, ...)
>+{
>+ va_list ap;
>+ va_start(ap, fmt);
>+ vsnprintf(test_diag_buf, TEST_DIAG_DATA_MAX, fmt, ap);
>+ va_end(ap);
>+}
>+
>+static void test_clear_diag_data(void)
>+{
>+ /*
>+ * Limit buffer with zero byte to show that there is no
>+ * any entry.
>+ */
>+ test_diag_buf[0] = '\0';
>+}
>+
>+static int test_diagnostic_is_set(void)
>+{
>+ return test_diag_buf[0] != '\0';
>+}
>+
>+/*
>+ * Parse the last diagnostic data entry and print it in YAML
>+ * format with the corresponding additional half-indent in TAP
>+ * (2 spaces).
>+ * Clear diagnostic message to be sure that it's printed once.
>+ * XXX: \n separators are changed to \0 during parsing and
>+ * printing output for convenience in usage.
>+ */
>+static void test_diagnostic(void)
>+{
>+ test_message(" ---");
>+ char *ent = test_diag_buf;
>+ char *ent_end = NULL;
>+ while ((ent_end = strchr(ent, '\n')) != NULL) {
>+ char *next_ent = ent_end + 1;
>+ /*
>+ * Limit string with with the zero byte for
>+ * formatted output. Anyway, don't need this \n
>+ * anymore.
>+ */
>+ *ent_end = '\0';
>+ test_message(" %s", ent);
>+ ent = next_ent;
>+ }
>+ test_message(" ...");
>+ test_clear_diag_data();
>+}
>+
>+static jmp_buf test_run_env;
>+
>+TEST_NORET void _test_exit(int status)
>+{
>+ longjmp(test_run_env, status);
>+}
>+
>+static int test_run(const struct test_unit *test, size_t test_number,
>+ void *test_state)
>+{
>+ int status = TEST_EXIT_SUCCESS;
>+ /*
>+ * Run unit test. Diagnostic in case of failure setup by
>+ * helpers assert macros defined in the header.
>+ */
>+ int jmp_status;
>+ if ((jmp_status = setjmp(test_run_env)) == 0) {
>+ if (test->f(test_state) != TEST_EXIT_SUCCESS)
>+ status = TEST_EXIT_FAILURE;
>+ } else {
>+ status = jmp_status - TEST_JMP_STATUS_SHIFT;
>+ }
>+ const char *result = status == TEST_EXIT_SUCCESS ? "ok" : "not ok";
>+
>+ /*
>+ * Format suffix of the test message for SKIP or TODO
>+ * directives.
>+ */
>+#define SUFFIX_SZ 1024
>+ char suffix[SUFFIX_SZ] = {0};
>+ if (skip_reason) {
>+ snprintf(suffix, SUFFIX_SZ, SKIP_DIRECTIVE "%s", skip_reason);
>+ skip_reason = NULL;
>+ } else if (todo_reason) {
>+ /* Prevent count this test as failed. */
>+ status = TEST_EXIT_SUCCESS;
>+ snprintf(suffix, SUFFIX_SZ, TODO_DIRECTIVE "%s", todo_reason);
>+ todo_reason = NULL;
>+ }
>+#undef SUFFIX_SZ
>+
>+ test_message("%s %lu - %s%s", result, test_number, test->name,
>+ suffix);
>+
>+ if (status && test_diagnostic_is_set())
>+ test_diagnostic();
>+ return status;
>+}
>+
>+int _test_run_group(const char *group_name, const struct test_unit *tests,
>+ size_t n_tests, void *test_state)
Strictly saying, the <type>* and <type>[] are different, and since that testing
facility is dependent on the `sizeof` behavior on <type>[], I think that argument
type should be changed to `const struct test_unit[]`.
>+{
>+ test_start_comment(group_name);
>+
>+ level++;
>+ test_print_tap_version();
>+
>+ test_plan(n_tests);
>+
>+ size_t n_failed = 0;
>+
>+ size_t i;
>+ for (i = 0; i < n_tests; i++) {
>+ size_t test_number = i + 1;
>+ /* Return 1 on failure, 0 on success. */
>+ n_failed += test_run(&tests[i], test_number, test_state);
>+ }
>+
>+ test_finish(n_tests, n_failed);
>+
>+ level--;
>+ return n_failed > 0 ? TEST_EXIT_FAILURE : TEST_EXIT_SUCCESS;
>+}
>diff --git a/test/tarantool-c-tests/test.h b/test/tarantool-c-tests/test.h
>new file mode 100644
>index 00000000..695c5b4d
>--- /dev/null
>+++ b/test/tarantool-c-tests/test.h
>@@ -0,0 +1,251 @@
>+#ifndef TEST_H
>+#define TEST_H
>+
>+#include <stdio.h>
>+#include <stdlib.h>
>+
>+/*
>+ * Test module, based on TAP 14 specification [1].
>+ * [1]: https://testanything.org/tap-version-14-specification.html
>+ * Version 13 is set for better compatibility on old machines.
>+ *
>+ * TODO:
>+ * * Helpers assert macros:
>+ * - assert_uint_equal if needed
>+ * - assert_uint_not_equal if needed
>+ * - assert_str_equal if needed
>+ * - assert_str_not_equal if needed
>+ * - assert_memory_equal if needed
>+ * - assert_memory_not_equal if needed
>+ * * Pragmas.
>+ */
>+
>+#define TAP_VERSION 13
>+
>+#define TEST_EXIT_SUCCESS 0
>+#define TEST_EXIT_FAILURE 1
>+
>+#define TEST_JMP_STATUS_SHIFT 2
>+#define TEST_LJMP_EXIT_SUCCESS (TEST_EXIT_SUCCESS + TEST_JMP_STATUS_SHIFT)
>+#define TEST_LJMP_EXIT_FAILURE (TEST_EXIT_FAILURE + TEST_JMP_STATUS_SHIFT)
>+
>+#define TEST_NORET __attribute__((noreturn))
>+
>+typedef int (*test_func)(void *test_state);
>+struct test_unit {
>+ const char *name;
>+ test_func f;
>+};
>+
>+/* Initialize `test_unit` structure. */
>+#define test_unit_new(f) {#f, f}
>+
>+#define lengthof(arr) (sizeof(arr) / sizeof((arr)[0]))
See the comment with my concerns about the <type> and <type>[]
above.
>+
>+/*
>+ * __func__ is the name for a test group, "main" for the parent
>+ * test.
>+ */
>+#define test_run_group(t_arr, t_state) \
>+ _test_run_group(__func__, t_arr, lengthof(t_arr), t_state)
Is there any reason for it to be a macro and not a function wrapper?
I believe it is better to use the functions when possible, since they are
easier to support and debug.
>+
>+#define SKIP_DIRECTIVE " # SKIP "
>+#define TODO_DIRECTIVE " # TODO "
>+
>+/*
>+ * XXX: May be implemented as well via
>+ * `_test_run_group(__func, NULL, 0, NULL)` and
>+ * `test_set_skip_reason` with additional changes in the former.
>+ * But the current approach is easier to maintain, as far as we
>+ * don't want to interfere different entities.
>+ */
>+#define skip_all(reason) do { \
>+ _test_print_skip_all(__func__, reason); \
>+ return TEST_EXIT_SUCCESS; \
>+} while (0)
Again, I propose to replace it with a conventional function to
make the return point explicit in the test implementation, so it
looks like `return skip_all(<reason>)`, instead of just `skip_all`.
Same for the skipcond and todo facilities below.
>+
>+#define skip(reason) do { \
>+ test_set_skip_reason(reason); \
>+ return TEST_EXIT_SUCCESS; \
>+} while (0)
>+
>+#define todo(reason) do { \
>+ test_set_todo_reason(reason); \
>+ return TEST_EXIT_FAILURE; \
>+} while (0)
>+
>+#define bail_out(reason) do { \
>+ /* \
>+ * For backwards compatibility with TAP13 Harnesses, \
>+ * Producers _should_ emit a "Bail out!" line at the root \
>+ * indentation level whenever a Subtest bails out [1]. \
>+ */ \
>+ printf("Bail out! %s\n", reason); \
>+ exit(TEST_EXIT_FAILURE); \
>+} while (0)
>+
>+/* `fmt` should always be a format string here. */
>+#define test_comment(fmt, ...) test_message("# " fmt, __VA_ARGS__)
>+
>+/*
>+ * This is a set of useful assert macros like the standard C
>+ * libary's assert(3) macro.
>+ *
>+ * On an assertion failure an assert macro will save the
>+ * diagnostic to the special buffer, to be reported via YAML
>+ * Diagnostic block and finish a test function with
>+ * `return TEST_EXIT_FAILURE`.
>+ *
>+ * Due to limitations of the C language `assert_true()` and
>+ * `assert_false()` macros can only display the expression that
>+ * caused the assertion failure. Type specific assert macros,
>+ * `assert_{type}_equal()` and `assert_{type}_not_equal()`, save
>+ * the data that caused the assertion failure which increases data
>+ * visibility aiding debugging of failing test cases.
>+ */
>+
>+#define LOCATION_FMT "location:\t%s:%d\n"
>+#define ASSERT_NAME_FMT(name) "failed_assertion:\t" #name "\n"
>+
>+#define assert_true(cond) do { \
>+ if (!(cond)) { \
>+ test_save_diag_data(LOCATION_FMT \
>+ "condition_failed:\t'" #cond "'\n", \
>+ __FILE__, __LINE__); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_false(cond) assert_true(!(cond))
>+
>+#define assert_ptr_equal(got, expected) do { \
>+ if ((got) != (expected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_ptr_equal) \
>+ "got: %p\n" \
>+ "expected: %p\n", \
>+ __FILE__, __LINE__, (got), (expected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_ptr_not_equal(got, unexpected) do { \
>+ if ((got) == (unexpected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_ptr_not_equal) \
>+ "got: %p\n" \
>+ "unexpected: %p\n", \
>+ __FILE__, __LINE__, (got), (unexpected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_int_equal(got, expected) do { \
>+ if ((got) != (expected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_int_equal) \
>+ "got: %d\n" \
>+ "expected: %d\n", \
>+ __FILE__, __LINE__, (got), (expected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_int_not_equal(got, unexpected) do { \
>+ if ((got) == (unexpected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_int_not_equal) \
>+ "got: %d\n" \
>+ "unexpected: %d\n", \
>+ __FILE__, __LINE__, (got), (unexpected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_sizet_equal(got, expected) do { \
>+ if ((got) != (expected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_sizet_equal) \
>+ "got: %lu\n" \
>+ "expected: %lu\n", \
>+ __FILE__, __LINE__, (got), (expected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+#define assert_sizet_not_equal(got, unexpected) do { \
>+ if ((got) == (unexpected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_sizet_not_equal) \
>+ "got: %lu\n" \
>+ "unexpected: %lu\n", \
>+ __FILE__, __LINE__, (got), (unexpected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+/* Check that doubles are __exactly__ the same. */
>+#define assert_double_equal(got, expected) do { \
>+ if ((got) != (expected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_double_equal) \
>+ "got: %lf\n" \
>+ "expected: %lf\n", \
>+ __FILE__, __LINE__, (got), (expected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+/* Check that doubles are not __exactly__ the same. */
>+#define assert_double_not_equal(got, unexpected) do { \
>+ if ((got) == (unexpected)) { \
>+ test_save_diag_data( \
>+ LOCATION_FMT \
>+ ASSERT_NAME_FMT(assert_double_not_equal) \
>+ "got: %lf\n" \
>+ "unexpected: %lf\n", \
>+ __FILE__, __LINE__, (got), (unexpected) \
>+ ); \
>+ _test_exit(TEST_LJMP_EXIT_FAILURE); \
>+ } \
>+} while (0)
>+
>+/* API declaration. */
>+
>+/*
>+ * Print formatted message with the corresponding indent.
>+ * If you want to leave a comment, use `test_comment()` instead.
>+ */
>+void test_message(const char *fmt, ...);
>+
>+/* Need for `skip_all()`, please, don't use it. */
>+void _test_print_skip_all(const char *group_name, const char *reason);
>+/* End test via `longjmp()`, please, don't use it. */
>+TEST_NORET void _test_exit(int status);
>+
>+void test_set_skip_reason(const char *reason);
>+void test_set_todo_reason(const char *reason);
>+/*
>+ * Save formatted diagnostic data. Each entry separated with \n.
>+ */
>+void test_save_diag_data(const char *fmt, ...);
>+
>+/* Internal, it is better to use `test_run_group()` instead. */
>+int _test_run_group(const char *group_name, const struct test_unit *tests,
>+ size_t n_tests, void *test_state);
>+
>+#endif /* TEST_H */
>--
>2.34.1
--
Best regards,
Maxim Kokryashkin
[-- Attachment #2: Type: text/html, Size: 26600 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* [Tarantool-patches] [PATCH v1 luajit 3/5] test: introduce utils.h helper for C tests
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 1/5] test: fix setting of {DY}LD_LIBRARY_PATH variables Sergey Kaplun via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 2/5] test: introduce module for C tests Sergey Kaplun via Tarantool-patches
@ 2023-03-15 16:11 ` Sergey Kaplun via Tarantool-patches
2023-03-20 15:21 ` Maxim Kokryashkin via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 4/5] test: rewrite misclib-getmetrics-capi test in C Sergey Kaplun via Tarantool-patches
` (2 subsequent siblings)
5 siblings, 1 reply; 12+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2023-03-15 16:11 UTC (permalink / raw)
To: Igor Munkin, Maxim Kokryashkin; +Cc: tarantool-patches
This header contains generic init and close for tests and helpers for
loading auxiliary Lua script with functions to run inside a C test.
It will be properly used in the next commit.
Part of tarantool/tarantool#7900
---
Those introduced utils macros assume that we have the table on the top
of the Lua stack. So, after each test the stack should be stayed
"untouched". But if we push helper function or some value of the Lua
stack and them assertion fails, we exit from function without stack
cleaning. So we should try to avoid such situation whenever appropriate,
but don't be fanatical at this point -- if some assertion fails, we are
in troubles anyway.
Don't use __FILE__ here, because it may contain relative path to the
file depending on CMake version or -fmacro-prefix-map= flag.
test/tarantool-c-tests/utils.h | 63 ++++++++++++++++++++++++++++++++++
1 file changed, 63 insertions(+)
create mode 100644 test/tarantool-c-tests/utils.h
diff --git a/test/tarantool-c-tests/utils.h b/test/tarantool-c-tests/utils.h
new file mode 100644
index 00000000..cf668006
--- /dev/null
+++ b/test/tarantool-c-tests/utils.h
@@ -0,0 +1,63 @@
+#include <limits.h>
+#include <string.h>
+
+#include "lauxlib.h"
+#include "lua.h"
+#include "luajit.h"
+#include "lualib.h"
+
+#include "test.h"
+
+/* Generic init for our tests. */
+static lua_State *utils_lua_init(void)
+{
+ lua_State *L = luaL_newstate();
+ if (!L)
+ bail_out("Can't init Lua state");
+ /* Stop collector during library initialization. */
+ lua_gc(L, LUA_GCSTOP, 0);
+ luaL_openlibs(L);
+ lua_gc(L, LUA_GCRESTART, -1);
+ return L;
+}
+
+/* Generic close for our tests. */
+static void utils_lua_close(lua_State *L)
+{
+ lua_close(L);
+}
+
+/*
+ * Load the pair to the test file <filename-script.lua>.
+ * Each file should return the table with functions (named the
+ * same as the corresponding unit tests) to call in unit tests.
+ * Push the table with those functions on the Lua stack.
+ */
+#define utils_load_aux_script(L) do { \
+ /* \
+ * Format script name. \
+ * `__ABS_FILENAME__` is set via CMake. \
+ */ \
+ char script[PATH_MAX] = __ABS_FILENAME__; \
+ char *file_suffix = strstr(script, ".test.c"); \
+ strcpy(file_suffix, "-script.lua"); \
+ \
+ if (luaL_dofile((L), script) != LUA_OK) { \
+ test_comment("Can't load %s: '%s'", script, \
+ lua_tostring((L), -1)); \
+ bail_out("Can't load auxiliary script"); \
+ } \
+ \
+ if (!lua_istable((L), -1)) \
+ bail_out("Returned value from script is not a table"); \
+} while (0)
+
+/*
+ * Accept a table on top of the Lua stack which containing the
+ * function named as the unit test we currently executing.
+ */
+#define utils_get_aux_lfunc(L) do { \
+ lua_getfield((L), -1, __func__); \
+ if (!lua_isfunction((L), -1)) \
+ bail_out("Can't get auxiliary test function"); \
+} while (0)
--
2.34.1
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [Tarantool-patches] [PATCH v1 luajit 3/5] test: introduce utils.h helper for C tests
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 3/5] test: introduce utils.h helper " Sergey Kaplun via Tarantool-patches
@ 2023-03-20 15:21 ` Maxim Kokryashkin via Tarantool-patches
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-20 15:21 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 3209 bytes --]
Hi!
Thanks for the patch!
It, generally, LGTM, but with concerns about the skipcond-like
implementation, I mentioned in review for the previous commit.
--
Best regards,
Maxim Kokryashkin
>Среда, 15 марта 2023, 19:14 +03:00 от Sergey Kaplun <skaplun@tarantool.org>:
>
>This header contains generic init and close for tests and helpers for
>loading auxiliary Lua script with functions to run inside a C test.
>
>It will be properly used in the next commit.
>
>Part of tarantool/tarantool#7900
>---
>
>Those introduced utils macros assume that we have the table on the top
>of the Lua stack. So, after each test the stack should be stayed
>"untouched". But if we push helper function or some value of the Lua
>stack and them assertion fails, we exit from function without stack
>cleaning. So we should try to avoid such situation whenever appropriate,
>but don't be fanatical at this point -- if some assertion fails, we are
>in troubles anyway.
>
>Don't use __FILE__ here, because it may contain relative path to the
>file depending on CMake version or -fmacro-prefix-map= flag.
>
> test/tarantool-c-tests/utils.h | 63 ++++++++++++++++++++++++++++++++++
> 1 file changed, 63 insertions(+)
> create mode 100644 test/tarantool-c-tests/utils.h
>
>diff --git a/test/tarantool-c-tests/utils.h b/test/tarantool-c-tests/utils.h
>new file mode 100644
>index 00000000..cf668006
>--- /dev/null
>+++ b/test/tarantool-c-tests/utils.h
>@@ -0,0 +1,63 @@
>+#include <limits.h>
>+#include <string.h>
>+
>+#include "lauxlib.h"
>+#include "lua.h"
>+#include "luajit.h"
>+#include "lualib.h"
>+
>+#include "test.h"
>+
>+/* Generic init for our tests. */
>+static lua_State *utils_lua_init(void)
>+{
>+ lua_State *L = luaL_newstate();
>+ if (!L)
>+ bail_out("Can't init Lua state");
>+ /* Stop collector during library initialization. */
>+ lua_gc(L, LUA_GCSTOP, 0);
>+ luaL_openlibs(L);
>+ lua_gc(L, LUA_GCRESTART, -1);
>+ return L;
>+}
>+
>+/* Generic close for our tests. */
>+static void utils_lua_close(lua_State *L)
>+{
>+ lua_close(L);
>+}
>+
>+/*
>+ * Load the pair to the test file <filename-script.lua>.
>+ * Each file should return the table with functions (named the
>+ * same as the corresponding unit tests) to call in unit tests.
>+ * Push the table with those functions on the Lua stack.
>+ */
>+#define utils_load_aux_script(L) do { \
>+ /* \
>+ * Format script name. \
>+ * `__ABS_FILENAME__` is set via CMake. \
>+ */ \
>+ char script[PATH_MAX] = __ABS_FILENAME__; \
>+ char *file_suffix = strstr(script, ".test.c"); \
>+ strcpy(file_suffix, "-script.lua"); \
>+ \
>+ if (luaL_dofile((L), script) != LUA_OK) { \
>+ test_comment("Can't load %s: '%s'", script, \
>+ lua_tostring((L), -1)); \
>+ bail_out("Can't load auxiliary script"); \
>+ } \
>+ \
>+ if (!lua_istable((L), -1)) \
>+ bail_out("Returned value from script is not a table"); \
>+} while (0)
>+
>+/*
>+ * Accept a table on top of the Lua stack which containing the
>+ * function named as the unit test we currently executing.
>+ */
>+#define utils_get_aux_lfunc(L) do { \
>+ lua_getfield((L), -1, __func__); \
>+ if (!lua_isfunction((L), -1)) \
>+ bail_out("Can't get auxiliary test function"); \
>+} while (0)
>--
>2.34.1
[-- Attachment #2: Type: text/html, Size: 3863 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* [Tarantool-patches] [PATCH v1 luajit 4/5] test: rewrite misclib-getmetrics-capi test in C
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
` (2 preceding siblings ...)
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 3/5] test: introduce utils.h helper " Sergey Kaplun via Tarantool-patches
@ 2023-03-15 16:11 ` Sergey Kaplun via Tarantool-patches
2023-03-22 0:07 ` Maxim Kokryashkin via Tarantool-patches
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 5/5] test: rewrite misclib-sysprof-capi " Sergey Kaplun via Tarantool-patches
2023-03-20 13:50 ` [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Maxim Kokryashkin via Tarantool-patches
5 siblings, 1 reply; 12+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2023-03-15 16:11 UTC (permalink / raw)
To: Igor Munkin, Maxim Kokryashkin; +Cc: tarantool-patches
This patch rewrites the aforementioned test with usage libtest recently
introduced. Since we still stand in need of a Lua helper script for
generation of traces, the original test file is reworked as a standalone
script, which returns the table with helper functions. Each helper
function is named the same as the test where it will be used. Now single
quotes are used according our Lua code style.
In C part all asserts from glibc are replaced with the corresponding
assert_{cond} from the libtest. Now tests return `TEST_EXIT_SUCCESS` at
the finish. Also, stack check for the amount of return results from
helper function is slightly changed, since there is one more stack slot
in use (table with these functions). `snap_restores()` C function
duplicates 4 times for each subtest. Common helper
`check_snap_restores()` is used for each of them. Each error throwing is
replaced with `bail_out()` call.
NB: `lua_pop()` to clear the Lua stack after call should be done before
any possible assertion, which exit from test leaving the stack
uncleaned.
All skipconds use macros defined in <lj_arch.h>, so it is included in
the test. As far as this test initializes LuaJIT VM manually, there is
no need to check `jit.status()` result; check `LJ_HASJIT` is enough.
Now, only JIT-related tests are skipped, when compiled without JIT.
Nevertheless, all tests are skipped for *BSD arches.
Also, this patch sets the new CMake variable named `LUAJIT_LIBRARY`
equals to `LUAJIT_LIB` in `PARENT_SCOPE` to be used in tarantool-c-tests
linking.
Also, .c_test suffix is added to the <.gitignore>.
Part of tarantool/tarantool#7900
---
.gitignore | 1 +
src/CMakeLists.txt | 2 +
test/tarantool-c-tests/CMakeLists.txt | 32 +-
.../misclib-getmetrics-capi-script.lua} | 82 ++---
.../misclib-getmetrics-capi.test.c | 341 ++++++++++++++++++
test/tarantool-tests/CMakeLists.txt | 1 -
.../misclib-getmetrics-capi/CMakeLists.txt | 1 -
.../misclib-getmetrics-capi/testgetmetrics.c | 270 --------------
8 files changed, 409 insertions(+), 321 deletions(-)
rename test/{tarantool-tests/misclib-getmetrics-capi.test.lua => tarantool-c-tests/misclib-getmetrics-capi-script.lua} (68%)
create mode 100644 test/tarantool-c-tests/misclib-getmetrics-capi.test.c
delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
diff --git a/.gitignore b/.gitignore
index b7908aee..dc5ea5fc 100644
--- a/.gitignore
+++ b/.gitignore
@@ -24,3 +24,4 @@ install_manifest.txt
luajit-parse-memprof
luajit-parse-sysprof
luajit.pc
+*.c_test
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index dffc0a4d..58d83330 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -325,6 +325,8 @@ if(NOT BUILDMODE STREQUAL "dynamic")
set(LUAJIT_BIN luajit_static)
set(LUAJIT_LIB libluajit_static)
endif()
+# Need for test linking, so PARENT_SCOPE option is used.
+set(LUAJIT_LIBRARY ${LUAJIT_LIB} PARENT_SCOPE)
set(LIBLUAJIT_DEPS ${LIBLUAJIT_STATIC_DEPS} ${LIBLUAJIT_SHARED_DEPS})
add_executable(${LUAJIT_BIN} EXCLUDE_FROM_ALL ${CLI_SOURCES})
diff --git a/test/tarantool-c-tests/CMakeLists.txt b/test/tarantool-c-tests/CMakeLists.txt
index 5ebea441..96beef1a 100644
--- a/test/tarantool-c-tests/CMakeLists.txt
+++ b/test/tarantool-c-tests/CMakeLists.txt
@@ -22,13 +22,37 @@ set_target_properties(libtest PROPERTIES
LIBRARY_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
)
-# XXX: For now, just build libtest. The tests to be depended on
-# will be added at the next commit.
+# TARGET_C_FLAGS is required here to be sure that headers like
+# lj_arch.h in compiled test is consistent with the LuaJIT library
+# to link.
+AppendFlags(TESTS_C_FLAGS ${TARGET_C_FLAGS})
+
+set(CTEST_SRC_SUFFIX ".test.c")
+file(GLOB tests "${CMAKE_CURRENT_SOURCE_DIR}/*${CTEST_SRC_SUFFIX}")
+foreach(test_source ${tests})
+ string(REGEX REPLACE ".+/([^/]+)${CTEST_SRC_SUFFIX}" "\\1" exe ${test_source})
+ add_executable(${exe} EXCLUDE_FROM_ALL ${test_source})
+ target_include_directories(${exe} PRIVATE
+ ${CMAKE_CURRENT_SOURCE_DIR}
+ ${LUAJIT_SOURCE_DIR}
+ )
+ set_target_properties(${exe} PROPERTIES
+ # `__FILE__` macro may not represnt absolute path to the
+ # source file, depening on cmake version or
+ # -fmacro-prefix-map flag value.
+ # Use custom macro.
+ COMPILE_FLAGS "${TESTS_C_FLAGS} -D__ABS_FILENAME__='\"${test_source}\"'"
+ OUTPUT_NAME "${exe}${C_TEST_SUFFIX}"
+ RUNTIME_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
+ )
+ target_link_libraries(${exe} libtest ${LUAJIT_LIBRARY})
+ LIST(APPEND TESTS_COMPILED ${exe})
+endforeach()
+
add_custom_target(tarantool-c-tests
- DEPENDS libluajit libtest
+ DEPENDS libluajit libtest ${TESTS_COMPILED}
)
-# XXX: For now, run 0 tests. Just verify that libtest was build.
add_custom_command(TARGET tarantool-c-tests
COMMENT "Running Tarantool C tests"
COMMAND
diff --git a/test/tarantool-tests/misclib-getmetrics-capi.test.lua b/test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
similarity index 68%
rename from test/tarantool-tests/misclib-getmetrics-capi.test.lua
rename to test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
index 654e5545..8fab485a 100644
--- a/test/tarantool-tests/misclib-getmetrics-capi.test.lua
+++ b/test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
@@ -1,32 +1,25 @@
-local tap = require('tap')
-local test = tap.test("clib-misc-getmetrics"):skipcond({
- ['Test requires JIT enabled'] = not jit.status(),
- ['Disabled on *BSD due to #4819'] = jit.os == 'BSD',
-})
+local ffi = require('ffi')
-test:plan(11)
+-- Auxiliary script to provide Lua functions to be used in tests
+-- for `getmetrics()` C API inside the test
+-- <misclib-getmetrics-capi.test.c>.
+local M = {}
-local path = arg[0]:gsub('%.test%.lua', '')
-local suffix = package.cpath:match('?.(%a+);')
-package.cpath = ('%s/?.%s;'):format(path, suffix)..package.cpath
+-- XXX: Max nins is limited by max IRRef, that equals to
+-- REF_DROP - REF_BIAS. Unfortunately, these constants are not
+-- provided to Lua space, so we ought to make some math:
+-- * REF_DROP = 0xffff
+-- * REF_BIAS = 0x8000
+local MAXNINS = 0xffff - 0x8000
-local MAXNINS = require('utils').const.maxnins
local jit_opt_default = {
3, -- level
- "hotloop=56",
- "hotexit=10",
- "minstitch=0",
+ 'hotloop=56',
+ 'hotexit=10',
+ 'minstitch=0',
}
-local testgetmetrics = require("testgetmetrics")
-
-test:ok(testgetmetrics.base())
-test:ok(testgetmetrics.gc_allocated_freed())
-test:ok(testgetmetrics.gc_steps())
-
-test:ok(testgetmetrics.objcount(function(iterations)
- local ffi = require("ffi")
-
+M.objcount = function(iterations)
jit.opt.start(0)
local placeholder = {
@@ -51,35 +44,34 @@ test:ok(testgetmetrics.objcount(function(iterations)
for _ = 1, iterations do
-- Check counting of VLA/VLS/aligned cdata.
- table.insert(placeholder.cdata, ffi.new("char[?]", 4))
+ table.insert(placeholder.cdata, ffi.new('char[?]', 4))
end
for _ = 1, iterations do
-- Check counting of non-VLA/VLS/aligned cdata.
- table.insert(placeholder.cdata, ffi.new("uint64_t", _))
+ table.insert(placeholder.cdata, ffi.new('uint64_t', _))
end
placeholder = nil -- luacheck: no unused
-- Restore default jit settings.
jit.opt.start(unpack(jit_opt_default))
-end))
+end
-test:ok(testgetmetrics.objcount_cdata_decrement(function()
+M.objcount_cdata_decrement = function()
-- gc_cdatanum decrement test.
-- See https://github.com/tarantool/tarantool/issues/5820.
- local ffi = require("ffi")
local function nop() end
- ffi.gc(ffi.cast("void *", 0), nop)
+ ffi.gc(ffi.cast('void *', 0), nop)
-- Does not collect the cdata, but resurrects the object and
-- removes LJ_GC_CDATA_FIN flag.
collectgarbage()
-- Collects the cdata.
collectgarbage()
-end))
+end
-- Compiled loop with a direct exit to the interpreter.
-test:ok(testgetmetrics.snap_restores(function()
- jit.opt.start(0, "hotloop=1")
+M.snap_restores_direct_exit = function()
+ jit.opt.start(0, 'hotloop=1')
local sum = 0
for i = 1, 20 do
@@ -91,11 +83,11 @@ test:ok(testgetmetrics.snap_restores(function()
-- A single snapshot restoration happened on loop finish.
return 1
-end))
+end
-- Compiled loop with a side exit which does not get compiled.
-test:ok(testgetmetrics.snap_restores(function()
- jit.opt.start(0, "hotloop=1", "hotexit=2", ("minstitch=%d"):format(MAXNINS))
+M.snap_restores_side_exit_not_compiled = function()
+ jit.opt.start(0, 'hotloop=1', 'hotexit=2', ('minstitch=%d'):format(MAXNINS))
local function foo(i)
-- math.fmod is not yet compiled!
@@ -112,13 +104,13 @@ test:ok(testgetmetrics.snap_restores(function()
-- Side exits from the root trace could not get compiled.
return 5
-end))
+end
-- Compiled loop with a side exit which gets compiled.
-test:ok(testgetmetrics.snap_restores(function()
+M.snap_restores_side_exit_compiled = function()
-- Optimization level is important here as `loop` optimization
-- may unroll the loop body and insert +1 side exit.
- jit.opt.start(0, "hotloop=1", "hotexit=5")
+ jit.opt.start(0, 'hotloop=1', 'hotexit=5')
local function foo(i)
return i <= 10 and i or tostring(i)
@@ -136,13 +128,13 @@ test:ok(testgetmetrics.snap_restores(function()
-- and compiled
-- 1 side exit on loop end
return 6
-end))
+end
-- Compiled scalar trace with a direct exit to the interpreter.
-test:ok(testgetmetrics.snap_restores(function()
+M.snap_restores_direct_exit_scalar = function()
-- For calls it will be 2 * hotloop (see lj_dispatch.{c,h}
-- and hotcall@vm_*.dasc).
- jit.opt.start(3, "hotloop=2", "hotexit=3")
+ jit.opt.start(3, 'hotloop=2', 'hotexit=3')
local function foo(i)
return i <= 15 and i or tostring(i)
@@ -167,15 +159,15 @@ test:ok(testgetmetrics.snap_restores(function()
jit.opt.start(unpack(jit_opt_default))
return 2
-end))
+end
-test:ok(testgetmetrics.strhash())
-
-test:ok(testgetmetrics.tracenum_base(function()
+M.tracenum_base = function()
local sum = 0
for i = 1, 200 do
sum = sum + i
end
-- Compiled only 1 loop as new trace.
return 1
-end))
+end
+
+return M
diff --git a/test/tarantool-c-tests/misclib-getmetrics-capi.test.c b/test/tarantool-c-tests/misclib-getmetrics-capi.test.c
new file mode 100644
index 00000000..a1bafbe6
--- /dev/null
+++ b/test/tarantool-c-tests/misclib-getmetrics-capi.test.c
@@ -0,0 +1,341 @@
+#include "lua.h"
+#include "luajit.h"
+#include "lauxlib.h"
+#include "lmisclib.h"
+
+#include "test.h"
+#include "utils.h"
+
+/* Need for skipcond for BSD and JIT. */
+#include "lj_arch.h"
+
+static int base(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Metrics metrics;
+ luaM_metrics(L, &metrics);
+
+ /*
+ * Just check structure format, not values that fields
+ * contain.
+ */
+ (void)metrics.strhash_hit;
+ (void)metrics.strhash_miss;
+
+ (void)metrics.gc_strnum;
+ (void)metrics.gc_tabnum;
+ (void)metrics.gc_udatanum;
+ (void)metrics.gc_cdatanum;
+
+ (void)metrics.gc_total;
+ (void)metrics.gc_freed;
+ (void)metrics.gc_allocated;
+
+ (void)metrics.gc_steps_pause;
+ (void)metrics.gc_steps_propagate;
+ (void)metrics.gc_steps_atomic;
+ (void)metrics.gc_steps_sweepstring;
+ (void)metrics.gc_steps_sweep;
+ (void)metrics.gc_steps_finalize;
+
+ (void)metrics.jit_snap_restore;
+ (void)metrics.jit_trace_abort;
+ (void)metrics.jit_mcode_size;
+ (void)metrics.jit_trace_num;
+
+ return TEST_EXIT_SUCCESS;
+}
+
+static int gc_allocated_freed(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Metrics oldm, newm;
+ /* Force up garbage collect all dead objects. */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+
+ luaM_metrics(L, &oldm);
+ /* Simple garbage generation. */
+ if (luaL_dostring(L, "local i = 0 for j = 1, 10 do i = i + j end"))
+ bail_out("failed to translate Lua code snippet");
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ luaM_metrics(L, &newm);
+ assert_true(newm.gc_allocated - oldm.gc_allocated > 0);
+ assert_true(newm.gc_freed - oldm.gc_freed > 0);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+static int gc_steps(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Metrics oldm, newm;
+ /*
+ * Some garbage has already happened before the next line,
+ * i.e. during frontend processing Lua test chunk.
+ * Let's put a full garbage collection cycle on top
+ * of that, and confirm that non-null values are reported
+ * (we are not yet interested in actual numbers):
+ */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+
+ luaM_metrics(L, &oldm);
+ assert_true(oldm.gc_steps_pause > 0);
+ assert_true(oldm.gc_steps_propagate > 0);
+ assert_true(oldm.gc_steps_atomic > 0);
+ assert_true(oldm.gc_steps_sweepstring > 0);
+ assert_true(oldm.gc_steps_sweep > 0);
+ /* Nothing to finalize, skipped. */
+ assert_true(oldm.gc_steps_finalize == 0);
+
+ /*
+ * As long as we don't create new Lua objects
+ * consequent call should return the same values:
+ */
+ luaM_metrics(L, &newm);
+ assert_sizet_equal(newm.gc_steps_pause, oldm.gc_steps_pause);
+ assert_sizet_equal(newm.gc_steps_propagate, oldm.gc_steps_propagate);
+ assert_sizet_equal(newm.gc_steps_atomic, oldm.gc_steps_atomic);
+ assert_sizet_equal(newm.gc_steps_sweepstring,
+ oldm.gc_steps_sweepstring);
+ assert_sizet_equal(newm.gc_steps_sweep, oldm.gc_steps_sweep);
+ /* Nothing to finalize, skipped. */
+ assert_true(newm.gc_steps_finalize == 0);
+ oldm = newm;
+
+ /*
+ * Now the last phase: run full GC once and make sure that
+ * everything is being reported as expected:
+ */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ luaM_metrics(L, &newm);
+ assert_true(newm.gc_steps_pause - oldm.gc_steps_pause == 1);
+ assert_true(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 1);
+ assert_true(newm.gc_steps_atomic - oldm.gc_steps_atomic == 1);
+ assert_true(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 1);
+ assert_true(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 1);
+ /* Nothing to finalize, skipped. */
+ assert_true(newm.gc_steps_finalize == 0);
+ oldm = newm;
+
+ /*
+ * Now let's run three GC cycles to ensure that
+ * increment was not a lucky coincidence.
+ */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ luaM_metrics(L, &newm);
+ assert_true(newm.gc_steps_pause - oldm.gc_steps_pause == 3);
+ assert_true(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 3);
+ assert_true(newm.gc_steps_atomic - oldm.gc_steps_atomic == 3);
+ assert_true(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 3);
+ assert_true(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 3);
+ /* Nothing to finalize, skipped. */
+ assert_true(newm.gc_steps_finalize == 0);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+static int objcount(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Metrics oldm, newm;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+
+ utils_get_aux_lfunc(L);
+
+ /* Force up garbage collect all dead objects. */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+
+ luaM_metrics(L, &oldm);
+ /* Generate garbage. Argument is iterations amount. */
+ lua_pushnumber(L, 1000);
+ lua_call(L, 1, 0);
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ luaM_metrics(L, &newm);
+ assert_sizet_equal(newm.gc_strnum, oldm.gc_strnum);
+ assert_sizet_equal(newm.gc_tabnum, oldm.gc_tabnum);
+ assert_sizet_equal(newm.gc_udatanum, oldm.gc_udatanum);
+ assert_sizet_equal(newm.gc_cdatanum, oldm.gc_cdatanum);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+static int objcount_cdata_decrement(void *test_state)
+{
+ lua_State *L = test_state;
+ /*
+ * cdata decrement test.
+ * See https://github.com/tarantool/tarantool/issues/5820.
+ */
+ struct luam_Metrics oldm, newm;
+ utils_get_aux_lfunc(L);
+
+ /* Force up garbage collect all dead objects. */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+
+ luaM_metrics(L, &oldm);
+ /*
+ * The function generates and collects cdata with
+ * LJ_GC_CDATA_FIN flag.
+ */
+ lua_call(L, 0, 0);
+ luaM_metrics(L, &newm);
+ assert_sizet_equal(newm.gc_cdatanum, oldm.gc_cdatanum);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+/*
+ * Get function to call to generate the corresponding snapshot
+ * restores on top of the Lua stack. Function returns the amount
+ * of snapshot restorations expected.
+ * Clear stack after call.
+ */
+static void check_snap_restores(lua_State *L)
+{
+ struct luam_Metrics oldm, newm;
+ luaM_metrics(L, &oldm);
+ /* Generate snapshots. */
+ lua_call(L, 0, 1);
+ int n = lua_gettop(L);
+ /*
+ * The first value is the table with functions,
+ * the second is number of snapshot restores.
+ */
+ if (n != 2 || !lua_isnumber(L, -1))
+ bail_out("incorrect return value: 1 number is required");
+ size_t snap_restores = lua_tonumber(L, -1);
+ luaM_metrics(L, &newm);
+ /*
+ * Remove `snap_restores` from stack.
+ * Must be done before potiential assert and exit from
+ * the test.
+ */
+ lua_pop(L, 1);
+ assert_true(newm.jit_snap_restore - oldm.jit_snap_restore
+ == snap_restores);
+}
+
+static int snap_restores_direct_exit(void *test_state)
+{
+ lua_State *L = test_state;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+ utils_get_aux_lfunc(L);
+ check_snap_restores(L);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int snap_restores_direct_exit_scalar(void *test_state)
+{
+ lua_State *L = test_state;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+ utils_get_aux_lfunc(L);
+ check_snap_restores(L);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int snap_restores_side_exit_compiled(void *test_state)
+{
+ lua_State *L = test_state;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+ utils_get_aux_lfunc(L);
+ check_snap_restores(L);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int snap_restores_side_exit_not_compiled(void *test_state)
+{
+ lua_State *L = test_state;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+ utils_get_aux_lfunc(L);
+ check_snap_restores(L);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int strhash(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Metrics oldm, newm;
+ lua_pushstring(L, "strhash_hit");
+ luaM_metrics(L, &oldm);
+ lua_pushstring(L, "strhash_hit");
+ lua_pushstring(L, "new_str");
+ luaM_metrics(L, &newm);
+ /* Remove pushed strings. */
+ lua_pop(L, 3);
+ assert_true(newm.strhash_hit - oldm.strhash_hit == 1);
+ assert_true(newm.strhash_miss - oldm.strhash_miss == 1);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int tracenum_base(void *test_state)
+{
+ lua_State *L = test_state;
+ if (!LJ_HASJIT)
+ skip("Test requires JIT enabled");
+ struct luam_Metrics metrics;
+ utils_get_aux_lfunc(L);
+
+ luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
+ /* Force up garbage collect all dead objects. */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+
+ luaM_metrics(L, &metrics);
+ assert_true(metrics.jit_trace_num == 0);
+
+ /* Generate traces. */
+ lua_call(L, 0, 1);
+ int n = lua_gettop(L);
+ /*
+ * The first value is the table with functions,
+ * the second is the amount of traces.
+ */
+ if (n != 2 || !lua_isnumber(L, -1))
+ bail_out("incorrect return value: 1 number is required");
+ size_t jit_trace_num = lua_tonumber(L, -1);
+ luaM_metrics(L, &metrics);
+ /* Remove `jit_trace_num` from Lua stack. */
+ lua_pop(L, 1);
+
+ assert_sizet_equal(metrics.jit_trace_num, jit_trace_num);
+
+ luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
+ /* Force up garbage collect all dead objects. */
+ lua_gc(L, LUA_GCCOLLECT, 0);
+ luaM_metrics(L, &metrics);
+ assert_true(metrics.jit_trace_num == 0);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+int main(void)
+{
+ if (LUAJIT_OS == LUAJIT_OS_BSD)
+ skip_all("Disabled on *BSD due to #4819");
+
+ lua_State *L = utils_lua_init();
+
+ utils_load_aux_script(L);
+ const struct test_unit tgroup[] = {
+ test_unit_new(base),
+ test_unit_new(gc_allocated_freed),
+ test_unit_new(gc_steps),
+ test_unit_new(objcount),
+ test_unit_new(objcount_cdata_decrement),
+ test_unit_new(snap_restores_direct_exit),
+ test_unit_new(snap_restores_direct_exit_scalar),
+ test_unit_new(snap_restores_side_exit_compiled),
+ test_unit_new(snap_restores_side_exit_not_compiled),
+ test_unit_new(strhash),
+ test_unit_new(tracenum_base)
+ };
+ const int test_result = test_run_group(tgroup, L);
+ utils_lua_close(L);
+ return test_result;
+}
diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
index 38d6ae49..b4ce39d3 100644
--- a/test/tarantool-tests/CMakeLists.txt
+++ b/test/tarantool-tests/CMakeLists.txt
@@ -66,7 +66,6 @@ add_subdirectory(lj-416-xor-before-jcc)
add_subdirectory(lj-601-fix-gc-finderrfunc)
add_subdirectory(lj-727-lightuserdata-itern)
add_subdirectory(lj-flush-on-trace)
-add_subdirectory(misclib-getmetrics-capi)
add_subdirectory(misclib-sysprof-capi)
# The part of the memory profiler toolchain is located in tools
diff --git a/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt b/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
deleted file mode 100644
index 60eb5bbb..00000000
--- a/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
+++ /dev/null
@@ -1 +0,0 @@
-BuildTestCLib(testgetmetrics testgetmetrics.c)
diff --git a/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c b/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
deleted file mode 100644
index 67776338..00000000
--- a/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
+++ /dev/null
@@ -1,270 +0,0 @@
-#include <lua.h>
-#include <luajit.h>
-#include <lauxlib.h>
-
-#include <lmisclib.h>
-
-#undef NDEBUG
-#include <assert.h>
-
-static int base(lua_State *L)
-{
- struct luam_Metrics metrics;
- luaM_metrics(L, &metrics);
-
- /* Just check structure format, not values that fields contain. */
- (void)metrics.strhash_hit;
- (void)metrics.strhash_miss;
-
- (void)metrics.gc_strnum;
- (void)metrics.gc_tabnum;
- (void)metrics.gc_udatanum;
- (void)metrics.gc_cdatanum;
-
- (void)metrics.gc_total;
- (void)metrics.gc_freed;
- (void)metrics.gc_allocated;
-
- (void)metrics.gc_steps_pause;
- (void)metrics.gc_steps_propagate;
- (void)metrics.gc_steps_atomic;
- (void)metrics.gc_steps_sweepstring;
- (void)metrics.gc_steps_sweep;
- (void)metrics.gc_steps_finalize;
-
- (void)metrics.jit_snap_restore;
- (void)metrics.jit_trace_abort;
- (void)metrics.jit_mcode_size;
- (void)metrics.jit_trace_num;
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int gc_allocated_freed(lua_State *L)
-{
- struct luam_Metrics oldm, newm;
- /* Force up garbage collect all dead objects. */
- lua_gc(L, LUA_GCCOLLECT, 0);
-
- luaM_metrics(L, &oldm);
- /* Simple garbage generation. */
- if (luaL_dostring(L, "local i = 0 for j = 1, 10 do i = i + j end"))
- luaL_error(L, "failed to translate Lua code snippet");
- lua_gc(L, LUA_GCCOLLECT, 0);
- luaM_metrics(L, &newm);
- assert(newm.gc_allocated - oldm.gc_allocated > 0);
- assert(newm.gc_freed - oldm.gc_freed > 0);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int gc_steps(lua_State *L)
-{
- struct luam_Metrics oldm, newm;
- /*
- * Some garbage has already happened before the next line,
- * i.e. during frontend processing Lua test chunk.
- * Let's put a full garbage collection cycle on top
- * of that, and confirm that non-null values are reported
- * (we are not yet interested in actual numbers):
- */
- lua_gc(L, LUA_GCCOLLECT, 0);
-
- luaM_metrics(L, &oldm);
- assert(oldm.gc_steps_pause > 0);
- assert(oldm.gc_steps_propagate > 0);
- assert(oldm.gc_steps_atomic > 0);
- assert(oldm.gc_steps_sweepstring > 0);
- assert(oldm.gc_steps_sweep > 0);
- /* Nothing to finalize, skipped. */
- assert(oldm.gc_steps_finalize == 0);
-
- /*
- * As long as we don't create new Lua objects
- * consequent call should return the same values:
- */
- luaM_metrics(L, &newm);
- assert(newm.gc_steps_pause - oldm.gc_steps_pause == 0);
- assert(newm.gc_steps_propagate - oldm.gc_steps_propagate == 0);
- assert(newm.gc_steps_atomic - oldm.gc_steps_atomic == 0);
- assert(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring == 0);
- assert(newm.gc_steps_sweep - oldm.gc_steps_sweep == 0);
- /* Nothing to finalize, skipped. */
- assert(newm.gc_steps_finalize == 0);
- oldm = newm;
-
- /*
- * Now the last phase: run full GC once and make sure that
- * everything is being reported as expected:
- */
- lua_gc(L, LUA_GCCOLLECT, 0);
- luaM_metrics(L, &newm);
- assert(newm.gc_steps_pause - oldm.gc_steps_pause == 1);
- assert(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 1);
- assert(newm.gc_steps_atomic - oldm.gc_steps_atomic == 1);
- assert(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 1);
- assert(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 1);
- /* Nothing to finalize, skipped. */
- assert(newm.gc_steps_finalize == 0);
- oldm = newm;
-
- /*
- * Now let's run three GC cycles to ensure that
- * increment was not a lucky coincidence.
- */
- lua_gc(L, LUA_GCCOLLECT, 0);
- lua_gc(L, LUA_GCCOLLECT, 0);
- lua_gc(L, LUA_GCCOLLECT, 0);
- luaM_metrics(L, &newm);
- assert(newm.gc_steps_pause - oldm.gc_steps_pause == 3);
- assert(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 3);
- assert(newm.gc_steps_atomic - oldm.gc_steps_atomic == 3);
- assert(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 3);
- assert(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 3);
- /* Nothing to finalize, skipped. */
- assert(newm.gc_steps_finalize == 0);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int objcount(lua_State *L)
-{
- struct luam_Metrics oldm, newm;
- int n = lua_gettop(L);
- if (n != 1 || !lua_isfunction(L, 1))
- luaL_error(L, "incorrect argument: 1 function is required");
-
- /* Force up garbage collect all dead objects. */
- lua_gc(L, LUA_GCCOLLECT, 0);
-
- luaM_metrics(L, &oldm);
- /* Generate garbage. Argument is iterations amount. */
- lua_pushnumber(L, 1000);
- lua_call(L, 1, 0);
- lua_gc(L, LUA_GCCOLLECT, 0);
- luaM_metrics(L, &newm);
- assert(newm.gc_strnum - oldm.gc_strnum == 0);
- assert(newm.gc_tabnum - oldm.gc_tabnum == 0);
- assert(newm.gc_udatanum - oldm.gc_udatanum == 0);
- assert(newm.gc_cdatanum - oldm.gc_cdatanum == 0);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int objcount_cdata_decrement(lua_State *L)
-{
- /*
- * cdata decrement test.
- * See https://github.com/tarantool/tarantool/issues/5820.
- */
- struct luam_Metrics oldm, newm;
- int n = lua_gettop(L);
- if (n != 1 || !lua_isfunction(L, 1))
- luaL_error(L, "incorrect argument: 1 function is required");
-
- /* Force up garbage collect all dead objects. */
- lua_gc(L, LUA_GCCOLLECT, 0);
-
- luaM_metrics(L, &oldm);
- /*
- * The function generates and collects cdata with
- * LJ_GC_CDATA_FIN flag.
- */
- lua_call(L, 0, 0);
- luaM_metrics(L, &newm);
- assert(newm.gc_cdatanum - oldm.gc_cdatanum == 0);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int snap_restores(lua_State *L)
-{
- struct luam_Metrics oldm, newm;
- int n = lua_gettop(L);
- if (n != 1 || !lua_isfunction(L, 1))
- luaL_error(L, "incorrect arguments: 1 function is required");
-
- luaM_metrics(L, &oldm);
- /* Generate snapshots. */
- lua_call(L, 0, 1);
- n = lua_gettop(L);
- if (n != 1 || !lua_isnumber(L, 1))
- luaL_error(L, "incorrect return value: 1 number is required");
- size_t snap_restores = lua_tonumber(L, 1);
- luaM_metrics(L, &newm);
- assert(newm.jit_snap_restore - oldm.jit_snap_restore == snap_restores);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int strhash(lua_State *L)
-{
- struct luam_Metrics oldm, newm;
- lua_pushstring(L, "strhash_hit");
- luaM_metrics(L, &oldm);
- lua_pushstring(L, "strhash_hit");
- lua_pushstring(L, "new_str");
- luaM_metrics(L, &newm);
- assert(newm.strhash_hit - oldm.strhash_hit == 1);
- assert(newm.strhash_miss - oldm.strhash_miss == 1);
- lua_pop(L, 3);
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int tracenum_base(lua_State *L)
-{
- struct luam_Metrics metrics;
- int n = lua_gettop(L);
- if (n != 1 || !lua_isfunction(L, 1))
- luaL_error(L, "incorrect arguments: 1 function is required");
-
- luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
- /* Force up garbage collect all dead objects. */
- lua_gc(L, LUA_GCCOLLECT, 0);
-
- luaM_metrics(L, &metrics);
- assert(metrics.jit_trace_num == 0);
-
- /* Generate traces. */
- lua_call(L, 0, 1);
- n = lua_gettop(L);
- if (n != 1 || !lua_isnumber(L, 1))
- luaL_error(L, "incorrect return value: 1 number is required");
- size_t jit_trace_num = lua_tonumber(L, 1);
- luaM_metrics(L, &metrics);
- assert(metrics.jit_trace_num == jit_trace_num);
-
- luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
- /* Force up garbage collect all dead objects. */
- lua_gc(L, LUA_GCCOLLECT, 0);
- luaM_metrics(L, &metrics);
- assert(metrics.jit_trace_num == 0);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static const struct luaL_Reg testgetmetrics[] = {
- {"base", base},
- {"gc_allocated_freed", gc_allocated_freed},
- {"gc_steps", gc_steps},
- {"objcount", objcount},
- {"objcount_cdata_decrement", objcount_cdata_decrement},
- {"snap_restores", snap_restores},
- {"strhash", strhash},
- {"tracenum_base", tracenum_base},
- {NULL, NULL}
-};
-
-LUA_API int luaopen_testgetmetrics(lua_State *L)
-{
- luaL_register(L, "testgetmetrics", testgetmetrics);
- return 1;
-}
--
2.34.1
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [Tarantool-patches] [PATCH v1 luajit 4/5] test: rewrite misclib-getmetrics-capi test in C
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 4/5] test: rewrite misclib-getmetrics-capi test in C Sergey Kaplun via Tarantool-patches
@ 2023-03-22 0:07 ` Maxim Kokryashkin via Tarantool-patches
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-22 0:07 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 23917 bytes --]
Hi!
Thanks for the patch!
Please consider my comments below.
>
>>This patch rewrites the aforementioned test with usage libtest recently
>Typo: s/with usage libtest/with the usage of libtest,/
>>introduced. Since we still stand in need of a Lua helper script for
>>generation of traces, the original test file is reworked as a standalone
>>script, which returns the table with helper functions. Each helper
>>function is named the same as the test where it will be used. Now single
>>quotes are used according our Lua code style.
>Typo: s/according our/according to our/
>>
>>In C part all asserts from glibc are replaced with the corresponding
>>assert_{cond} from the libtest. Now tests return `TEST_EXIT_SUCCESS` at
>>the finish. Also, stack check for the amount of return results from
>Typo s:/stack check/the stack check/
>>helper function is slightly changed, since there is one more stack slot
>Typo: s/helper/the helper/
>>in use (table with these functions). `snap_restores()` C function
>>duplicates 4 times for each subtest. Common helper
>>`check_snap_restores()` is used for each of them. Each error throwing is
>>replaced with `bail_out()` call.
>>
>>NB: `lua_pop()` to clear the Lua stack after call should be done before
>Typo: s/after call/after a call/
>>any possible assertion, which exit from test leaving the stack
>Typo: s/which exit from test/which would exit from the test/
>>uncleaned.
>>
>>All skipconds use macros defined in <lj_arch.h>, so it is included in
>>the test. As far as this test initializes LuaJIT VM manually, there is
>Typo: s/LuaJIT/the LuaJIT
>>no need to check `jit.status()` result; check `LJ_HASJIT` is enough.
>Typo: s/check/checking/
>>Now, only JIT-related tests are skipped, when compiled without JIT.
>>Nevertheless, all tests are skipped for *BSD arches.
>>
>>Also, this patch sets the new CMake variable named `LUAJIT_LIBRARY`
>>equals to `LUAJIT_LIB` in `PARENT_SCOPE` to be used in tarantool-c-tests
>Typo: s/equals/equal/
>>linking.
>>
>>Also, .c_test suffix is added to the <.gitignore>.
>>
>>Part of tarantool/tarantool#7900
>>---
>> .gitignore | 1 +
>> src/CMakeLists.txt | 2 +
>> test/tarantool-c-tests/CMakeLists.txt | 32 +-
>> .../misclib-getmetrics-capi-script.lua} | 82 ++---
>> .../misclib-getmetrics-capi.test.c | 341 ++++++++++++++++++
>> test/tarantool-tests/CMakeLists.txt | 1 -
>> .../misclib-getmetrics-capi/CMakeLists.txt | 1 -
>> .../misclib-getmetrics-capi/testgetmetrics.c | 270 --------------
>> 8 files changed, 409 insertions(+), 321 deletions(-)
>> rename test/{tarantool-tests/misclib-getmetrics-capi.test.lua => tarantool-c-tests/misclib-getmetrics-capi-script.lua} (68%)
>> create mode 100644 test/tarantool-c-tests/misclib-getmetrics-capi.test.c
>> delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
>> delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
>>
>>diff --git a/.gitignore b/.gitignore
>>index b7908aee..dc5ea5fc 100644
>>--- a/.gitignore
>>+++ b/.gitignore
>>@@ -24,3 +24,4 @@ install_manifest.txt
>> luajit-parse-memprof
>> luajit-parse-sysprof
>> luajit.pc
>>+*.c_test
>>diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
>>index dffc0a4d..58d83330 100644
>>--- a/src/CMakeLists.txt
>>+++ b/src/CMakeLists.txt
>>@@ -325,6 +325,8 @@ if(NOT BUILDMODE STREQUAL "dynamic")
>> set(LUAJIT_BIN luajit_static)
>> set(LUAJIT_LIB libluajit_static)
>> endif()
>>+# Need for test linking, so PARENT_SCOPE option is used.
>Typo: s/Need for test/Needed for the test/
>Typo: s/so PARENT_SCOPE/so the PARENT_SCOPE/
>>+set(LUAJIT_LIBRARY ${LUAJIT_LIB} PARENT_SCOPE)
>> set(LIBLUAJIT_DEPS ${LIBLUAJIT_STATIC_DEPS} ${LIBLUAJIT_SHARED_DEPS})
>>
>> add_executable(${LUAJIT_BIN} EXCLUDE_FROM_ALL ${CLI_SOURCES})
>>diff --git a/test/tarantool-c-tests/CMakeLists.txt b/test/tarantool-c-tests/CMakeLists.txt
>>index 5ebea441..96beef1a 100644
>>--- a/test/tarantool-c-tests/CMakeLists.txt
>>+++ b/test/tarantool-c-tests/CMakeLists.txt
>>@@ -22,13 +22,37 @@ set_target_properties(libtest PROPERTIES
>> LIBRARY_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
>> )
>>
>>-# XXX: For now, just build libtest. The tests to be depended on
>>-# will be added at the next commit.
>>+# TARGET_C_FLAGS is required here to be sure that headers like
>>+# lj_arch.h in compiled test is consistent with the LuaJIT library
>Typo: s/is consistent/are consistent/
>>+# to link.
>>+AppendFlags(TESTS_C_FLAGS ${TARGET_C_FLAGS})
>>+
>>+set(CTEST_SRC_SUFFIX ".test.c")
>>+file(GLOB tests "${CMAKE_CURRENT_SOURCE_DIR}/*${CTEST_SRC_SUFFIX}")
>>+foreach(test_source ${tests})
>>+ string(REGEX REPLACE ".+/([^/]+)${CTEST_SRC_SUFFIX}" "\\1" exe ${test_source})
>>+ add_executable(${exe} EXCLUDE_FROM_ALL ${test_source})
>>+ target_include_directories(${exe} PRIVATE
>>+ ${CMAKE_CURRENT_SOURCE_DIR}
>>+ ${LUAJIT_SOURCE_DIR}
>>+ )
>>+ set_target_properties(${exe} PROPERTIES
>>+ # `__FILE__` macro may not represnt absolute path to the
>>+ # source file, depening on cmake version or
>>+ # -fmacro-prefix-map flag value.
>>+ # Use custom macro.
>>+ COMPILE_FLAGS "${TESTS_C_FLAGS} -D__ABS_FILENAME__='\"${test_source}\"'"
>>+ OUTPUT_NAME "${exe}${C_TEST_SUFFIX}"
>>+ RUNTIME_OUTPUT_DIRECTORY "${CMAKE_CURRENT_BINARY_DIR}"
>>+ )
>>+ target_link_libraries(${exe} libtest ${LUAJIT_LIBRARY})
>>+ LIST(APPEND TESTS_COMPILED ${exe})
>>+endforeach()
>>+
>> add_custom_target(tarantool-c-tests
>>- DEPENDS libluajit libtest
>>+ DEPENDS libluajit libtest ${TESTS_COMPILED}
>> )
>>
>>-# XXX: For now, run 0 tests. Just verify that libtest was build.
>> add_custom_command(TARGET tarantool-c-tests
>> COMMENT "Running Tarantool C tests"
>> COMMAND
>>diff --git a/test/tarantool-tests/misclib-getmetrics-capi.test.lua b/test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
>>similarity index 68%
>>rename from test/tarantool-tests/misclib-getmetrics-capi.test.lua
>>rename to test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
>>index 654e5545..8fab485a 100644
>>--- a/test/tarantool-tests/misclib-getmetrics-capi.test.lua
>>+++ b/test/tarantool-c-tests/misclib-getmetrics-capi-script.lua
>>@@ -1,32 +1,25 @@
>>-local tap = require('tap')
>>-local test = tap.test("clib-misc-getmetrics"):skipcond({
>>- ['Test requires JIT enabled'] = not jit.status(),
>>- ['Disabled on *BSD due to #4819'] = jit.os == 'BSD',
>>-})
>>+local ffi = require('ffi')
>>
>>-test:plan(11)
>>+-- Auxiliary script to provide Lua functions to be used in tests
>>+-- for `getmetrics()` C API inside the test
>>+-- <misclib-getmetrics-capi.test.c>.
>>+local M = {}
>>
>>-local path = arg[0]:gsub('%.test%.lua', '')
>>-local suffix = package.cpath:match('?.(%a+);')
>>-package.cpath = ('%s/?.%s;'):format(path, suffix)..package.cpath
>>+-- XXX: Max nins is limited by max IRRef, that equals to
>>+-- REF_DROP - REF_BIAS. Unfortunately, these constants are not
>>+-- provided to Lua space, so we ought to make some math:
>Maybe, it’s better to say `exposed` instead of `provided`. Feel free to ignore.
>>+-- * REF_DROP = 0xffff
>>+-- * REF_BIAS = 0x8000
>Path to file with their definition would be useful.
>>+local MAXNINS = 0xffff - 0x8000
>>
>>-local MAXNINS = require('utils').const.maxnins
>> local jit_opt_default = {
>> 3, -- level
>>- "hotloop=56",
>>- "hotexit=10",
>>- "minstitch=0",
>>+ 'hotloop=56',
>>+ 'hotexit=10',
>>+ 'minstitch=0',
>> }
>>
>>-local testgetmetrics = require("testgetmetrics")
>>-
>>-test:ok(testgetmetrics.base())
>>-test:ok(testgetmetrics.gc_allocated_freed())
>>-test:ok(testgetmetrics.gc_steps())
>>-
>>-test:ok(testgetmetrics.objcount(function(iterations)
>>- local ffi = require("ffi")
>>-
>>+M.objcount = function(iterations)
>> jit.opt.start(0)
>>
>> local placeholder = {
>>@@ -51,35 +44,34 @@ test:ok(testgetmetrics.objcount(function(iterations)
>>
>> for _ = 1, iterations do
>> -- Check counting of VLA/VLS/aligned cdata.
>>- table.insert(placeholder.cdata, ffi.new("char[?]", 4))
>>+ table.insert(placeholder.cdata, ffi.new('char[?]', 4))
>> end
>>
>> for _ = 1, iterations do
>> -- Check counting of non-VLA/VLS/aligned cdata.
>>- table.insert(placeholder.cdata, ffi.new("uint64_t", _))
>>+ table.insert(placeholder.cdata, ffi.new('uint64_t', _))
>> end
>>
>> placeholder = nil -- luacheck: no unused
>> -- Restore default jit settings.
>> jit.opt.start(unpack(jit_opt_default))
>>-end))
>>+end
>>
>>-test:ok(testgetmetrics.objcount_cdata_decrement(function()
>>+M.objcount_cdata_decrement = function()
>> -- gc_cdatanum decrement test.
>> -- See https://github.com/tarantool/tarantool/issues/5820 .
>>- local ffi = require("ffi")
>> local function nop() end
>>- ffi.gc(ffi.cast("void *", 0), nop)
>>+ ffi.gc(ffi.cast('void *', 0), nop)
>> -- Does not collect the cdata, but resurrects the object and
>> -- removes LJ_GC_CDATA_FIN flag.
>> collectgarbage()
>> -- Collects the cdata.
>> collectgarbage()
>>-end))
>>+end
>>
>> -- Compiled loop with a direct exit to the interpreter.
>>-test:ok(testgetmetrics.snap_restores(function()
>>- jit.opt.start(0, "hotloop=1")
>>+M.snap_restores_direct_exit = function()
>>+ jit.opt.start(0, 'hotloop=1')
>>
>> local sum = 0
>> for i = 1, 20 do
>>@@ -91,11 +83,11 @@ test:ok(testgetmetrics.snap_restores(function()
>>
>> -- A single snapshot restoration happened on loop finish.
>> return 1
>>-end))
>>+end
>>
>> -- Compiled loop with a side exit which does not get compiled.
>>-test:ok(testgetmetrics.snap_restores(function()
>>- jit.opt.start(0, "hotloop=1", "hotexit=2", ("minstitch=%d"):format(MAXNINS))
>>+M.snap_restores_side_exit_not_compiled = function()
>>+ jit.opt.start(0, 'hotloop=1', 'hotexit=2', ('minstitch=%d'):format(MAXNINS))
>>
>> local function foo(i)
>> -- math.fmod is not yet compiled!
>>@@ -112,13 +104,13 @@ test:ok(testgetmetrics.snap_restores(function()
>>
>> -- Side exits from the root trace could not get compiled.
>> return 5
>>-end))
>>+end
>>
>> -- Compiled loop with a side exit which gets compiled.
>>-test:ok(testgetmetrics.snap_restores(function()
>>+M.snap_restores_side_exit_compiled = function()
>> -- Optimization level is important here as `loop` optimization
>> -- may unroll the loop body and insert +1 side exit.
>>- jit.opt.start(0, "hotloop=1", "hotexit=5")
>>+ jit.opt.start(0, 'hotloop=1', 'hotexit=5')
>>
>> local function foo(i)
>> return i <= 10 and i or tostring(i)
>>@@ -136,13 +128,13 @@ test:ok(testgetmetrics.snap_restores(function()
>> -- and compiled
>> -- 1 side exit on loop end
>> return 6
>>-end))
>>+end
>>
>> -- Compiled scalar trace with a direct exit to the interpreter.
>>-test:ok(testgetmetrics.snap_restores(function()
>>+M.snap_restores_direct_exit_scalar = function()
>> -- For calls it will be 2 * hotloop (see lj_dispatch.{c,h}
>> -- and hotcall@vm_*.dasc).
>>- jit.opt.start(3, "hotloop=2", "hotexit=3")
>>+ jit.opt.start(3, 'hotloop=2', 'hotexit=3')
>>
>> local function foo(i)
>> return i <= 15 and i or tostring(i)
>>@@ -167,15 +159,15 @@ test:ok(testgetmetrics.snap_restores(function()
>> jit.opt.start(unpack(jit_opt_default))
>>
>> return 2
>>-end))
>>+end
>>
>>-test:ok(testgetmetrics.strhash())
>>-
>>-test:ok(testgetmetrics.tracenum_base(function()
>>+M.tracenum_base = function()
>> local sum = 0
>> for i = 1, 200 do
>> sum = sum + i
>> end
>> -- Compiled only 1 loop as new trace.
>> return 1
>>-end))
>>+end
>>+
>>+return M
>>diff --git a/test/tarantool-c-tests/misclib-getmetrics-capi.test.c b/test/tarantool-c-tests/misclib-getmetrics-capi.test.c
>>new file mode 100644
>>index 00000000..a1bafbe6
>>--- /dev/null
>>+++ b/test/tarantool-c-tests/misclib-getmetrics-capi.test.c
>>@@ -0,0 +1,341 @@
>>+#include "lua.h"
>>+#include "luajit.h"
>>+#include "lauxlib.h"
>>+#include "lmisclib.h"
>>+
>>+#include "test.h"
>>+#include "utils.h"
>>+
>>+/* Need for skipcond for BSD and JIT. */
>>+#include "lj_arch.h"
>Still don’t like the fact that we need to do that.
>>+
>>+static int base(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ struct luam_Metrics metrics;
>>+ luaM_metrics(L, &metrics);
>>+
>>+ /*
>>+ * Just check structure format, not values that fields
>>+ * contain.
>>+ */
>>+ (void)metrics.strhash_hit;
>>+ (void)metrics.strhash_miss;
>>+
>>+ (void)metrics.gc_strnum;
>>+ (void)metrics.gc_tabnum;
>>+ (void)metrics.gc_udatanum;
>>+ (void)metrics.gc_cdatanum;
>>+
>>+ (void)metrics.gc_total;
>>+ (void)metrics.gc_freed;
>>+ (void)metrics.gc_allocated;
>>+
>>+ (void)metrics.gc_steps_pause;
>>+ (void)metrics.gc_steps_propagate;
>>+ (void)metrics.gc_steps_atomic;
>>+ (void)metrics.gc_steps_sweepstring;
>>+ (void)metrics.gc_steps_sweep;
>>+ (void)metrics.gc_steps_finalize;
>>+
>>+ (void)metrics.jit_snap_restore;
>>+ (void)metrics.jit_trace_abort;
>>+ (void)metrics.jit_mcode_size;
>>+ (void)metrics.jit_trace_num;
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int gc_allocated_freed(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ struct luam_Metrics oldm, newm;
>>+ /* Force up garbage collect all dead objects. */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+
>>+ luaM_metrics(L, &oldm);
>>+ /* Simple garbage generation. */
>>+ if (luaL_dostring(L, "local i = 0 for j = 1, 10 do i = i + j end"))
>>+ bail_out("failed to translate Lua code snippet");
>Why `bail_out` and not an assertion? Here and below.
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ luaM_metrics(L, &newm);
>>+ assert_true(newm.gc_allocated - oldm.gc_allocated > 0);
>>+ assert_true(newm.gc_freed - oldm.gc_freed > 0);
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int gc_steps(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ struct luam_Metrics oldm, newm;
>>+ /*
>>+ * Some garbage has already happened before the next line,
>>+ * i.e. during frontend processing Lua test chunk.
>>+ * Let's put a full garbage collection cycle on top
>>+ * of that, and confirm that non-null values are reported
>>+ * (we are not yet interested in actual numbers):
>>+ */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+
>>+ luaM_metrics(L, &oldm);
>>+ assert_true(oldm.gc_steps_pause > 0);
>>+ assert_true(oldm.gc_steps_propagate > 0);
>>+ assert_true(oldm.gc_steps_atomic > 0);
>>+ assert_true(oldm.gc_steps_sweepstring > 0);
>>+ assert_true(oldm.gc_steps_sweep > 0);
>>+ /* Nothing to finalize, skipped. */
>>+ assert_true(oldm.gc_steps_finalize == 0);
>>+
>>+ /*
>>+ * As long as we don't create new Lua objects
>>+ * consequent call should return the same values:
>>+ */
>>+ luaM_metrics(L, &newm);
>>+ assert_sizet_equal(newm.gc_steps_pause, oldm.gc_steps_pause);
>>+ assert_sizet_equal(newm.gc_steps_propagate, oldm.gc_steps_propagate);
>>+ assert_sizet_equal(newm.gc_steps_atomic, oldm.gc_steps_atomic);
>>+ assert_sizet_equal(newm.gc_steps_sweepstring,
>>+ oldm.gc_steps_sweepstring);
>>+ assert_sizet_equal(newm.gc_steps_sweep, oldm.gc_steps_sweep);
>>+ /* Nothing to finalize, skipped. */
>>+ assert_true(newm.gc_steps_finalize == 0);
>>+ oldm = newm;
>>+
>>+ /*
>>+ * Now the last phase: run full GC once and make sure that
>>+ * everything is being reported as expected:
>>+ */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ luaM_metrics(L, &newm);
>>+ assert_true(newm.gc_steps_pause - oldm.gc_steps_pause == 1);
>>+ assert_true(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 1);
>>+ assert_true(newm.gc_steps_atomic - oldm.gc_steps_atomic == 1);
>>+ assert_true(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 1);
>>+ assert_true(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 1);
>>+ /* Nothing to finalize, skipped. */
>>+ assert_true(newm.gc_steps_finalize == 0);
>>+ oldm = newm;
>>+
>>+ /*
>>+ * Now let's run three GC cycles to ensure that
>>+ * increment was not a lucky coincidence.
>>+ */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ luaM_metrics(L, &newm);
>>+ assert_true(newm.gc_steps_pause - oldm.gc_steps_pause == 3);
>>+ assert_true(newm.gc_steps_propagate - oldm.gc_steps_propagate >= 3);
>>+ assert_true(newm.gc_steps_atomic - oldm.gc_steps_atomic == 3);
>>+ assert_true(newm.gc_steps_sweepstring - oldm.gc_steps_sweepstring >= 3);
>>+ assert_true(newm.gc_steps_sweep - oldm.gc_steps_sweep >= 3);
>>+ /* Nothing to finalize, skipped. */
>>+ assert_true(newm.gc_steps_finalize == 0);
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int objcount(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ struct luam_Metrics oldm, newm;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+
>>+ utils_get_aux_lfunc(L);
>>+
>>+ /* Force up garbage collect all dead objects. */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+
>>+ luaM_metrics(L, &oldm);
>>+ /* Generate garbage. Argument is iterations amount. */
>>+ lua_pushnumber(L, 1000);
>>+ lua_call(L, 1, 0);
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ luaM_metrics(L, &newm);
>>+ assert_sizet_equal(newm.gc_strnum, oldm.gc_strnum);
>>+ assert_sizet_equal(newm.gc_tabnum, oldm.gc_tabnum);
>>+ assert_sizet_equal(newm.gc_udatanum, oldm.gc_udatanum);
>>+ assert_sizet_equal(newm.gc_cdatanum, oldm.gc_cdatanum);
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int objcount_cdata_decrement(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ /*
>>+ * cdata decrement test.
>>+ * See https://github.com/tarantool/tarantool/issues/5820 .
>>+ */
>>+ struct luam_Metrics oldm, newm;
>>+ utils_get_aux_lfunc(L);
>>+
>>+ /* Force up garbage collect all dead objects. */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+
>>+ luaM_metrics(L, &oldm);
>>+ /*
>>+ * The function generates and collects cdata with
>>+ * LJ_GC_CDATA_FIN flag.
>>+ */
>>+ lua_call(L, 0, 0);
>>+ luaM_metrics(L, &newm);
>>+ assert_sizet_equal(newm.gc_cdatanum, oldm.gc_cdatanum);
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+/*
>>+ * Get function to call to generate the corresponding snapshot
>>+ * restores on top of the Lua stack. Function returns the amount
>>+ * of snapshot restorations expected.
>>+ * Clear stack after call.
>>+ */
>>+static void check_snap_restores(lua_State *L)
>>+{
>>+ struct luam_Metrics oldm, newm;
>>+ luaM_metrics(L, &oldm);
>>+ /* Generate snapshots. */
>>+ lua_call(L, 0, 1);
>>+ int n = lua_gettop(L);
>>+ /*
>>+ * The first value is the table with functions,
>>+ * the second is number of snapshot restores.
>>+ */
>>+ if (n != 2 || !lua_isnumber(L, -1))
>>+ bail_out("incorrect return value: 1 number is required");
>>+ size_t snap_restores = lua_tonumber(L, -1);
>>+ luaM_metrics(L, &newm);
>>+ /*
>>+ * Remove `snap_restores` from stack.
>>+ * Must be done before potiential assert and exit from
>>+ * the test.
>>+ */
>>+ lua_pop(L, 1);
>>+ assert_true(newm.jit_snap_restore - oldm.jit_snap_restore
>>+ == snap_restores);
>>+}
>>+
>>+static int snap_restores_direct_exit(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+ utils_get_aux_lfunc(L);
>>+ check_snap_restores(L);
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int snap_restores_direct_exit_scalar(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+ utils_get_aux_lfunc(L);
>>+ check_snap_restores(L);
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int snap_restores_side_exit_compiled(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+ utils_get_aux_lfunc(L);
>>+ check_snap_restores(L);
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int snap_restores_side_exit_not_compiled(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+ utils_get_aux_lfunc(L);
>>+ check_snap_restores(L);
>>+ return TEST_EXIT_SUCCESS;
>>+}
>Maybe we should extend the testing utility, so it is possible to join test sets like
>those into a single group and skip them all at once?
>It would be much easier to maintain them for sure.
>>+
>>+static int strhash(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ struct luam_Metrics oldm, newm;
>>+ lua_pushstring(L, "strhash_hit");
>>+ luaM_metrics(L, &oldm);
>>+ lua_pushstring(L, "strhash_hit");
>>+ lua_pushstring(L, "new_str");
>>+ luaM_metrics(L, &newm);
>>+ /* Remove pushed strings. */
>>+ lua_pop(L, 3);
>>+ assert_true(newm.strhash_hit - oldm.strhash_hit == 1);
>>+ assert_true(newm.strhash_miss - oldm.strhash_miss == 1);
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+static int tracenum_base(void *test_state)
>>+{
>>+ lua_State *L = test_state;
>>+ if (!LJ_HASJIT)
>>+ skip("Test requires JIT enabled");
>>+ struct luam_Metrics metrics;
>>+ utils_get_aux_lfunc(L);
>>+
>>+ luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
>>+ /* Force up garbage collect all dead objects. */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+
>>+ luaM_metrics(L, &metrics);
>>+ assert_true(metrics.jit_trace_num == 0);
>>+
>>+ /* Generate traces. */
>>+ lua_call(L, 0, 1);
>>+ int n = lua_gettop(L);
>>+ /*
>>+ * The first value is the table with functions,
>>+ * the second is the amount of traces.
>>+ */
>>+ if (n != 2 || !lua_isnumber(L, -1))
>>+ bail_out("incorrect return value: 1 number is required");
>>+ size_t jit_trace_num = lua_tonumber(L, -1);
>>+ luaM_metrics(L, &metrics);
>>+ /* Remove `jit_trace_num` from Lua stack. */
>>+ lua_pop(L, 1);
>>+
>>+ assert_sizet_equal(metrics.jit_trace_num, jit_trace_num);
>>+
>>+ luaJIT_setmode(L, 0, LUAJIT_MODE_FLUSH);
>>+ /* Force up garbage collect all dead objects. */
>>+ lua_gc(L, LUA_GCCOLLECT, 0);
>>+ luaM_metrics(L, &metrics);
>>+ assert_true(metrics.jit_trace_num == 0);
>>+
>>+ return TEST_EXIT_SUCCESS;
>>+}
>>+
>>+int main(void)
>>+{
>>+ if (LUAJIT_OS == LUAJIT_OS_BSD)
>>+ skip_all("Disabled on *BSD due to #4819");
>>+
>>+ lua_State *L = utils_lua_init();
>>+
>>+ utils_load_aux_script(L);
>>+ const struct test_unit tgroup[] = {
>>+ test_unit_new(base),
>>+ test_unit_new(gc_allocated_freed),
>>+ test_unit_new(gc_steps),
>>+ test_unit_new(objcount),
>>+ test_unit_new(objcount_cdata_decrement),
>>+ test_unit_new(snap_restores_direct_exit),
>>+ test_unit_new(snap_restores_direct_exit_scalar),
>>+ test_unit_new(snap_restores_side_exit_compiled),
>>+ test_unit_new(snap_restores_side_exit_not_compiled),
>>+ test_unit_new(strhash),
>>+ test_unit_new(tracenum_base)
>I think that we need to make an effort and try move all skipconds into
>one place. The main functions seems like a good place for them. It
>would greatly increase readabilty and maintainability.
>>+ };
>>+ const int test_result = test_run_group(tgroup, L);
>>+ utils_lua_close(L);
>>+ return test_result;
>>+}
>>diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
>>index 38d6ae49..b4ce39d3 100644
>>--- a/test/tarantool-tests/CMakeLists.txt
>>+++ b/test/tarantool-tests/CMakeLists.txt
>>@@ -66,7 +66,6 @@ add_subdirectory(lj-416-xor-before-jcc)
>><snipped>
>>diff --git a/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt b/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
>>deleted file mode 100644
>>index 60eb5bbb..00000000
>>--- a/test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
>>+++ /dev/null
>>@@ -1 +0,0 @@
>>-BuildTestCLib(testgetmetrics testgetmetrics.c)
>>diff --git a/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c b/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
>>deleted file mode 100644
>>index 67776338..00000000
>>--- a/test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
>>+++ /dev/null
>>@@ -1,270 +0,0 @@
>><snipped>
>>--
>>2.34.1
>--
>Best regards,
>Maxim Kokryashkin
>
[-- Attachment #2: Type: text/html, Size: 29881 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* [Tarantool-patches] [PATCH v1 luajit 5/5] test: rewrite misclib-sysprof-capi test in C
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
` (3 preceding siblings ...)
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 4/5] test: rewrite misclib-getmetrics-capi test in C Sergey Kaplun via Tarantool-patches
@ 2023-03-15 16:11 ` Sergey Kaplun via Tarantool-patches
2023-03-20 16:24 ` Maxim Kokryashkin via Tarantool-patches
2023-03-20 13:50 ` [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Maxim Kokryashkin via Tarantool-patches
5 siblings, 1 reply; 12+ messages in thread
From: Sergey Kaplun via Tarantool-patches @ 2023-03-15 16:11 UTC (permalink / raw)
To: Igor Munkin, Maxim Kokryashkin; +Cc: tarantool-patches
This patch rewrites the aforementioned test with usage libtest recently
introduced. The approach is similar to the previous patch. `c_payload()`
function to profile now is set up globally on script loading. The
payload for enabled and disabled JIT is the same. As far as for sysprof
testing is necessary to replace backtrace with libunwind first, the
payload tests are marked with `todo()`.
Nevertheless, glibc `assert()` still necessary to check correctness of
profile writer, so it is still used there.
Relates to tarantool/tarantool#781
Part of tarantool/tarantool#7900
---
.../misclib-sysprof-capi-script.lua | 35 ++
.../misclib-sysprof-capi.test.c | 317 ++++++++++++++++++
test/tarantool-tests/CMakeLists.txt | 1 -
.../misclib-sysprof-capi.test.lua | 54 ---
.../misclib-sysprof-capi/CMakeLists.txt | 1 -
.../misclib-sysprof-capi/testsysprof.c | 260 --------------
6 files changed, 352 insertions(+), 316 deletions(-)
create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi-script.lua
create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi.test.c
delete mode 100644 test/tarantool-tests/misclib-sysprof-capi.test.lua
delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
diff --git a/test/tarantool-c-tests/misclib-sysprof-capi-script.lua b/test/tarantool-c-tests/misclib-sysprof-capi-script.lua
new file mode 100644
index 00000000..dd8387db
--- /dev/null
+++ b/test/tarantool-c-tests/misclib-sysprof-capi-script.lua
@@ -0,0 +1,35 @@
+local M = {}
+
+-- luacheck: no global
+assert(c_payload, 'c_payload global function should be set via script loader')
+
+local function lua_payload(n)
+ if n <= 1 then
+ return n
+ end
+ return lua_payload(n - 1) + lua_payload(n - 2)
+end
+
+local function payload()
+ local n_iterations = 500000
+
+ local co = coroutine.create(function()
+ for i = 1, n_iterations do
+ if i % 2 == 0 then
+ c_payload(10)
+ else
+ lua_payload(10)
+ end
+ coroutine.yield()
+ end
+ end)
+
+ for _ = 1, n_iterations do
+ coroutine.resume(co)
+ end
+end
+
+M.profile_func_jiton = payload
+M.profile_func_jitoff = payload
+
+return M
diff --git a/test/tarantool-c-tests/misclib-sysprof-capi.test.c b/test/tarantool-c-tests/misclib-sysprof-capi.test.c
new file mode 100644
index 00000000..4c54877e
--- /dev/null
+++ b/test/tarantool-c-tests/misclib-sysprof-capi.test.c
@@ -0,0 +1,317 @@
+#include "lauxlib.h"
+#include "lmisclib.h"
+#include "lua.h"
+#include "luajit.h"
+
+#include "test.h"
+#include "utils.h"
+
+#include <errno.h>
+#include <fcntl.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+/* XXX: Still need normal assert inside writer functions. */
+#undef NDEBUG
+#include <assert.h>
+
+/* Need for skipcond for OS and ARCH. */
+#include "lj_arch.h"
+
+#define UNUSED(x) ((void)(x))
+
+/* --- utils -------------------------------------------------- */
+
+#define SYSPROF_INTERVAL_DEFAULT 100
+
+/*
+ * Yep, 8Mb. Tuned in order not to bother the platform with too
+ * often flushes.
+ */
+#define STREAM_BUFFER_SIZE (8 * 1024 * 1024)
+
+/*
+ * Structure given as ctx to sysprof writer and on_stop callback.
+ */
+struct sysprof_ctx {
+ /* Output file descriptor for data. */
+ int fd;
+ /* Buffer for data. */
+ uint8_t buf[STREAM_BUFFER_SIZE];
+};
+
+/*
+ * Default buffer writer function.
+ * Just call fwrite to the corresponding FILE.
+ */
+static size_t buffer_writer_default(const void **buf_addr, size_t len,
+ void *opt)
+{
+ struct sysprof_ctx *ctx = opt;
+ int fd = ctx->fd;
+ const void * const buf_start = *buf_addr;
+ const void *data = *buf_addr;
+ size_t write_total = 0;
+
+ assert(len <= STREAM_BUFFER_SIZE);
+
+ for (;;) {
+ const size_t written = write(fd, data, len - write_total);
+
+ if (written == 0) {
+ /* Re-tries write in case of EINTR. */
+ if (errno != EINTR) {
+ /*
+ * Will be freed as whole chunk
+ * later.
+ */
+ *buf_addr = NULL;
+ return write_total;
+ }
+ errno = 0;
+ continue;
+ }
+
+ write_total += written;
+ assert(write_total <= len);
+
+ if (write_total == len)
+ break;
+
+ data = (uint8_t *)data + (ptrdiff_t)written;
+ }
+
+ *buf_addr = buf_start;
+ return write_total;
+}
+
+/*
+ * Default on stop callback. Just close the corresponding stream.
+ */
+static int on_stop_cb_default(void *opt, uint8_t *buf)
+{
+ UNUSED(buf);
+ struct sysprof_ctx *ctx = opt;
+ int fd = ctx->fd;
+ free(ctx);
+ return close(fd);
+}
+
+static int stream_init(struct luam_Sysprof_Options *opt)
+{
+ struct sysprof_ctx *ctx = calloc(1, sizeof(struct sysprof_ctx));
+ if (NULL == ctx)
+ return PROFILE_ERRIO;
+
+ ctx->fd = open("/dev/null", O_WRONLY | O_CREAT, 0644);
+ if (-1 == ctx->fd) {
+ free(ctx);
+ return PROFILE_ERRIO;
+ }
+
+ opt->ctx = ctx;
+ opt->buf = ctx->buf;
+ opt->len = STREAM_BUFFER_SIZE;
+
+ return PROFILE_SUCCESS;
+}
+
+/* --- C Payload ---------------------------------------------- */
+
+static double fib(double n)
+{
+ if (n <= 1)
+ return n;
+ return fib(n - 1) + fib(n - 2);
+}
+
+static int c_payload(lua_State *L)
+{
+ fib(luaL_checknumber(L, 1));
+ lua_pushboolean(L, 1);
+ return 1;
+}
+
+/* --- sysprof C API tests ------------------------------------ */
+
+static int base(void *test_state)
+{
+ UNUSED(test_state);
+ struct luam_Sysprof_Options opt = {};
+ struct luam_Sysprof_Counters cnt = {};
+
+ (void)opt.interval;
+ (void)opt.mode;
+ (void)opt.ctx;
+ (void)opt.buf;
+ (void)opt.len;
+
+ luaM_sysprof_report(&cnt);
+
+ (void)cnt.samples;
+ (void)cnt.vmst_interp;
+ (void)cnt.vmst_lfunc;
+ (void)cnt.vmst_ffunc;
+ (void)cnt.vmst_cfunc;
+ (void)cnt.vmst_gc;
+ (void)cnt.vmst_exit;
+ (void)cnt.vmst_record;
+ (void)cnt.vmst_opt;
+ (void)cnt.vmst_asm;
+ (void)cnt.vmst_trace;
+
+ return TEST_EXIT_SUCCESS;
+}
+
+static int validation(void *test_state)
+{
+ lua_State *L = test_state;
+ struct luam_Sysprof_Options opt = {};
+ int status = PROFILE_SUCCESS;
+
+ /* Unknown mode. */
+ opt.mode = 0x40;
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_ERRUSE);
+
+ /* Buffer not configured. */
+ opt.mode = LUAM_SYSPROF_CALLGRAPH;
+ opt.buf = NULL;
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_ERRUSE);
+
+ /* Bad interval. */
+ opt.mode = LUAM_SYSPROF_DEFAULT;
+ opt.interval = 0;
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_ERRUSE);
+
+ /* Check if profiling started. */
+ opt.mode = LUAM_SYSPROF_DEFAULT;
+ opt.interval = SYSPROF_INTERVAL_DEFAULT;
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_SUCCESS);
+
+ /* Already running. */
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_ERRRUN);
+
+ /* Profiler stopping. */
+ status = luaM_sysprof_stop(L);
+ assert_true(status == PROFILE_SUCCESS);
+
+ /* Stopping profiler which is not running. */
+ status = luaM_sysprof_stop(L);
+ assert_true(status == PROFILE_ERRRUN);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+/* Get function to profile on top, call it. */
+static int check_profile_func(lua_State *L)
+{
+ struct luam_Sysprof_Options opt = {};
+ struct luam_Sysprof_Counters cnt = {};
+ int status = PROFILE_ERRUSE;
+ /*
+ * Since all the other modes functionality is the
+ * subset of CALLGRAPH mode, run this mode to test
+ * the profiler's behavior.
+ */
+ opt.mode = LUAM_SYSPROF_CALLGRAPH;
+ opt.interval = SYSPROF_INTERVAL_DEFAULT;
+ stream_init(&opt);
+
+ /*
+ * XXX: Payload function on top will not be removed if any
+ * of those assertions fail. So, the next call to the
+ * `utils_get_aux_lfunc()` will fail. It's OK, since
+ * we are already in trouble, just keep it in mind.
+ */
+ assert_true(luaM_sysprof_set_writer(buffer_writer_default)
+ == PROFILE_SUCCESS);
+ assert_true(luaM_sysprof_set_on_stop(on_stop_cb_default)
+ == PROFILE_SUCCESS);
+ assert_true(luaM_sysprof_set_backtracer(NULL) == PROFILE_SUCCESS);
+
+ status = luaM_sysprof_start(L, &opt);
+ assert_true(status == PROFILE_SUCCESS);
+
+ /* Run payload. */
+ if (lua_pcall(L, 0, 0, 0) != LUA_OK) {
+ test_comment("error running payload: %s", lua_tostring(L, -1));
+ bail_out("error running sysprof test payload");
+ }
+
+ status = luaM_sysprof_stop(L);
+ assert_true(status == PROFILE_SUCCESS);
+
+ status = luaM_sysprof_report(&cnt);
+ assert_true(status == PROFILE_SUCCESS);
+
+ assert_true(cnt.samples > 1);
+ assert_true(cnt.samples == cnt.vmst_asm +
+ cnt.vmst_cfunc +
+ cnt.vmst_exit +
+ cnt.vmst_ffunc +
+ cnt.vmst_gc +
+ cnt.vmst_interp +
+ cnt.vmst_lfunc +
+ cnt.vmst_opt +
+ cnt.vmst_record +
+ cnt.vmst_trace);
+
+ return TEST_EXIT_SUCCESS;
+}
+
+/*
+ * FIXME: The following two tests are disabled because sometimes
+ * `backtrace` dynamically loads a platform-specific unwinder,
+ * which is not signal-safe.
+ */
+
+static int profile_func_jitoff(void *test_state)
+{
+ todo("Need to replace backtrace with libunwind first");
+ lua_State *L = test_state;
+ utils_get_aux_lfunc(L);
+ (void)luaJIT_setmode(L, 0, LUAJIT_MODE_ENGINE | LUAJIT_MODE_OFF);
+ (void)luaJIT_setmode(L, 0, LUAJIT_MODE_ENGINE | LUAJIT_MODE_FLUSH);
+ check_profile_func(L);
+ (void)luaJIT_setmode(L, 0, LUAJIT_MODE_ENGINE | LUAJIT_MODE_ON);
+ return TEST_EXIT_SUCCESS;
+}
+
+static int profile_func_jiton(void *test_state)
+{
+ todo("Need to replace backtrace with libunwind first");
+ lua_State *L = test_state;
+ utils_get_aux_lfunc(L);
+ check_profile_func(L);
+ return TEST_EXIT_SUCCESS;
+}
+
+int main(void)
+{
+ if (LUAJIT_OS != LUAJIT_OS_LINUX)
+ skip_all("Sysprof is implemented for Linux only");
+ if (LUAJIT_TARGET != LUAJIT_ARCH_X86
+ && LUAJIT_TARGET != LUAJIT_ARCH_X64)
+ skip_all("Sysprof is implemented for x86_64 only");
+
+ lua_State *L = utils_lua_init();
+
+ lua_pushcfunction(L, c_payload);
+ lua_setfield(L, LUA_GLOBALSINDEX, "c_payload");
+ utils_load_aux_script(L);
+
+ const struct test_unit tgroup[] = {
+ test_unit_new(base),
+ test_unit_new(validation),
+ test_unit_new(profile_func_jitoff),
+ test_unit_new(profile_func_jiton)
+ };
+ const int test_result = test_run_group(tgroup, L);
+ utils_lua_close(L);
+ return test_result;
+}
diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
index b4ce39d3..b1c7207f 100644
--- a/test/tarantool-tests/CMakeLists.txt
+++ b/test/tarantool-tests/CMakeLists.txt
@@ -66,7 +66,6 @@ add_subdirectory(lj-416-xor-before-jcc)
add_subdirectory(lj-601-fix-gc-finderrfunc)
add_subdirectory(lj-727-lightuserdata-itern)
add_subdirectory(lj-flush-on-trace)
-add_subdirectory(misclib-sysprof-capi)
# The part of the memory profiler toolchain is located in tools
# directory, jit, profiler, and bytecode toolchains are located
diff --git a/test/tarantool-tests/misclib-sysprof-capi.test.lua b/test/tarantool-tests/misclib-sysprof-capi.test.lua
deleted file mode 100644
index 4395bce3..00000000
--- a/test/tarantool-tests/misclib-sysprof-capi.test.lua
+++ /dev/null
@@ -1,54 +0,0 @@
-local tap = require("tap")
-local test = tap.test("clib-misc-sysprof"):skipcond({
- ["Sysprof is implemented for x86_64 only"] = jit.arch ~= "x86" and
- jit.arch ~= "x64",
- ["Sysprof is implemented for Linux only"] = jit.os ~= "Linux",
-})
-
-test:plan(2)
-
-local testsysprof = require("testsysprof")
-
-jit.off()
-
-test:ok(testsysprof.base())
-test:ok(testsysprof.validation())
-
--- FIXME: The following two tests are disabled because sometimes
--- `backtrace` dynamically loads a platform-specific unwinder, which is
--- not signal-safe.
---[[
-local function lua_payload(n)
- if n <= 1 then
- return n
- end
- return lua_payload(n - 1) + lua_payload(n - 2)
-end
-
-local function payload()
- local n_iterations = 500000
-
- local co = coroutine.create(function ()
- for i = 1, n_iterations do
- if i % 2 == 0 then
- testsysprof.c_payload(10)
- else
- lua_payload(10)
- end
- coroutine.yield()
- end
- end)
-
- for _ = 1, n_iterations do
- coroutine.resume(co)
- end
-end
-
-test:ok(testsysprof.profile_func(payload))
-
-jit.on()
-jit.flush()
-
-test:ok(testsysprof.profile_func(payload))
---]]
-os.exit(test:check() and 0 or 1)
diff --git a/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt b/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
deleted file mode 100644
index d9fb1a1a..00000000
--- a/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
+++ /dev/null
@@ -1 +0,0 @@
-BuildTestCLib(testsysprof testsysprof.c)
diff --git a/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c b/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
deleted file mode 100644
index d7a3e355..00000000
--- a/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
+++ /dev/null
@@ -1,260 +0,0 @@
-#include <lua.h>
-#include <luajit.h>
-#include <lauxlib.h>
-
-#include <lmisclib.h>
-
-#include <fcntl.h>
-#include <stdlib.h>
-#include <unistd.h>
-#include <errno.h>
-
-#undef NDEBUG
-#include <assert.h>
-
-#define UNUSED(x) ((void)(x))
-
-/* --- utils -------------------------------------------------------------- */
-
-#define SYSPROF_INTERVAL_DEFAULT 100
-
-/*
- ** Yep, 8Mb. Tuned in order not to bother the platform with too often flushes.
- */
-#define STREAM_BUFFER_SIZE (8 * 1024 * 1024)
-
-/* Structure given as ctx to sysprof writer and on_stop callback. */
-struct sysprof_ctx {
- /* Output file descriptor for data. */
- int fd;
- /* Buffer for data. */
- uint8_t buf[STREAM_BUFFER_SIZE];
-};
-
-/*
- ** Default buffer writer function.
- ** Just call fwrite to the corresponding FILE.
- */
-static size_t buffer_writer_default(const void **buf_addr, size_t len,
- void *opt)
-{
- struct sysprof_ctx *ctx = opt;
- int fd = ctx->fd;
- const void * const buf_start = *buf_addr;
- const void *data = *buf_addr;
- size_t write_total = 0;
-
- assert(len <= STREAM_BUFFER_SIZE);
-
- for (;;) {
- const size_t written = write(fd, data, len - write_total);
-
- if (written == 0) {
- /* Re-tries write in case of EINTR. */
- if (errno != EINTR) {
- /* Will be freed as whole chunk later. */
- *buf_addr = NULL;
- return write_total;
- }
- errno = 0;
- continue;
- }
-
- write_total += written;
- assert(write_total <= len);
-
- if (write_total == len)
- break;
-
- data = (uint8_t *)data + (ptrdiff_t)written;
- }
-
- *buf_addr = buf_start;
- return write_total;
-}
-
-/* Default on stop callback. Just close the corresponding stream. */
-static int on_stop_cb_default(void *opt, uint8_t *buf)
-{
- UNUSED(buf);
- struct sysprof_ctx *ctx = opt;
- int fd = ctx->fd;
- free(ctx);
- return close(fd);
-}
-
-static int stream_init(struct luam_Sysprof_Options *opt)
-{
- struct sysprof_ctx *ctx = calloc(1, sizeof(struct sysprof_ctx));
- if (NULL == ctx)
- return PROFILE_ERRIO;
-
- ctx->fd = open("/dev/null", O_WRONLY | O_CREAT, 0644);
- if (-1 == ctx->fd) {
- free(ctx);
- return PROFILE_ERRIO;
- }
-
- opt->ctx = ctx;
- opt->buf = ctx->buf;
- opt->len = STREAM_BUFFER_SIZE;
-
- return PROFILE_SUCCESS;
-}
-
-/* --- Payload ------------------------------------------------------------ */
-
-static double fib(double n)
-{
- if (n <= 1)
- return n;
- return fib(n - 1) + fib(n - 2);
-}
-
-static int c_payload(lua_State *L)
-{
- fib(luaL_checknumber(L, 1));
- lua_pushboolean(L, 1);
- return 1;
-}
-
-/* --- sysprof C API tests ------------------------------------------------ */
-
-static int base(lua_State *L)
-{
- struct luam_Sysprof_Options opt = {};
- struct luam_Sysprof_Counters cnt = {};
-
- (void)opt.interval;
- (void)opt.mode;
- (void)opt.ctx;
- (void)opt.buf;
- (void)opt.len;
-
- luaM_sysprof_report(&cnt);
-
- (void)cnt.samples;
- (void)cnt.vmst_interp;
- (void)cnt.vmst_lfunc;
- (void)cnt.vmst_ffunc;
- (void)cnt.vmst_cfunc;
- (void)cnt.vmst_gc;
- (void)cnt.vmst_exit;
- (void)cnt.vmst_record;
- (void)cnt.vmst_opt;
- (void)cnt.vmst_asm;
- (void)cnt.vmst_trace;
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int validation(lua_State *L)
-{
- struct luam_Sysprof_Options opt = {};
- int status = PROFILE_SUCCESS;
-
- /* Unknown mode. */
- opt.mode = 0x40;
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_ERRUSE == status);
-
- /* Buffer not configured. */
- opt.mode = LUAM_SYSPROF_CALLGRAPH;
- opt.buf = NULL;
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_ERRUSE == status);
-
- /* Bad interval. */
- opt.mode = LUAM_SYSPROF_DEFAULT;
- opt.interval = 0;
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_ERRUSE == status);
-
- /* Check if profiling started. */
- opt.mode = LUAM_SYSPROF_DEFAULT;
- opt.interval = SYSPROF_INTERVAL_DEFAULT;
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_SUCCESS == status);
-
- /* Already running. */
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_ERRRUN == status);
-
- /* Profiler stopping. */
- status = luaM_sysprof_stop(L);
- assert(PROFILE_SUCCESS == status);
-
- /* Stopping profiler which is not running. */
- status = luaM_sysprof_stop(L);
- assert(PROFILE_ERRRUN == status);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static int profile_func(lua_State *L)
-{
- struct luam_Sysprof_Options opt = {};
- struct luam_Sysprof_Counters cnt = {};
- int status = PROFILE_ERRUSE;
-
- int n = lua_gettop(L);
- if (n != 1 || !lua_isfunction(L, 1))
- luaL_error(L, "incorrect argument: 1 function is required");
-
- /*
- ** Since all the other modes functionality is the
- ** subset of CALLGRAPH mode, run this mode to test
- ** the profiler's behavior.
- */
- opt.mode = LUAM_SYSPROF_CALLGRAPH;
- opt.interval = SYSPROF_INTERVAL_DEFAULT;
- stream_init(&opt);
-
- assert(luaM_sysprof_set_writer(buffer_writer_default) == PROFILE_SUCCESS);
- assert(luaM_sysprof_set_on_stop(on_stop_cb_default) == PROFILE_SUCCESS);
- assert(luaM_sysprof_set_backtracer(NULL) == PROFILE_SUCCESS);
-
- status = luaM_sysprof_start(L, &opt);
- assert(PROFILE_SUCCESS == status);
-
- /* Run payload. */
- if (lua_pcall(L, 0, 0, 0) != LUA_OK)
- luaL_error(L, "error running payload: %s", lua_tostring(L, -1));
-
- status = luaM_sysprof_stop(L);
- assert(PROFILE_SUCCESS == status);
-
- status = luaM_sysprof_report(&cnt);
- assert(PROFILE_SUCCESS == status);
-
- assert(cnt.samples > 1);
- assert(cnt.samples == cnt.vmst_asm +
- cnt.vmst_cfunc +
- cnt.vmst_exit +
- cnt.vmst_ffunc +
- cnt.vmst_gc +
- cnt.vmst_interp +
- cnt.vmst_lfunc +
- cnt.vmst_opt +
- cnt.vmst_record +
- cnt.vmst_trace);
-
- lua_pushboolean(L, 1);
- return 1;
-}
-
-static const struct luaL_Reg testsysprof[] = {
- {"base", base},
- {"c_payload", c_payload},
- {"profile_func", profile_func},
- {"validation", validation},
- {NULL, NULL}
-};
-
-LUA_API int luaopen_testsysprof(lua_State *L)
-{
- luaL_register(L, "testsysprof", testsysprof);
- return 1;
-}
--
2.34.1
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [Tarantool-patches] [PATCH v1 luajit 5/5] test: rewrite misclib-sysprof-capi test in C
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 5/5] test: rewrite misclib-sysprof-capi " Sergey Kaplun via Tarantool-patches
@ 2023-03-20 16:24 ` Maxim Kokryashkin via Tarantool-patches
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-20 16:24 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 4742 bytes --]
Hi!
Thanks for the patch!
Please consider my comments below.
>Среда, 15 марта 2023, 19:14 +03:00 от Sergey Kaplun <skaplun@tarantool.org>:
>
>This patch rewrites the aforementioned test with usage libtest recently
Typo: s/with usage libtest/with the usage of libtest/
>introduced. The approach is similar to the previous patch. `c_payload()`
>function to profile now is set up globally on script loading. The
Typo: s/now is set up/is now set up/
>payload for enabled and disabled JIT is the same. As far as for sysprof
>testing is necessary to replace backtrace with libunwind first, the
>payload tests are marked with `todo()`.
>
>Nevertheless, glibc `assert()` still necessary to check correctness of
>profile writer, so it is still used there.
Typo: s/to check correctness of profile/to check the correctness of the profile/
>
>Relates to tarantool/tarantool#781
>Part of tarantool/tarantool#7900
>---
> .../misclib-sysprof-capi-script.lua | 35 ++
> .../misclib-sysprof-capi.test.c | 317 ++++++++++++++++++
> test/tarantool-tests/CMakeLists.txt | 1 -
> .../misclib-sysprof-capi.test.lua | 54 ---
> .../misclib-sysprof-capi/CMakeLists.txt | 1 -
> .../misclib-sysprof-capi/testsysprof.c | 260 --------------
> 6 files changed, 352 insertions(+), 316 deletions(-)
> create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi-script.lua
> create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi.test.c
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi.test.lua
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
>
>diff --git a/test/tarantool-c-tests/misclib-sysprof-capi-script.lua b/test/tarantool-c-tests/misclib-sysprof-capi-script.lua
>new file mode 100644
>index 00000000..dd8387db
>--- /dev/null
>+++ b/test/tarantool-c-tests/misclib-sysprof-capi-script.lua
>@@ -0,0 +1,35 @@
>+local M = {}
>+
>+-- luacheck: no global
>+assert(c_payload, 'c_payload global function should be set via script loader')
>+
>+local function lua_payload(n)
>+ if n <= 1 then
>+ return n
>+ end
>+ return lua_payload(n - 1) + lua_payload(n - 2)
>+end
>+
>+local function payload()
>+ local n_iterations = 500000
>+
>+ local co = coroutine.create(function()
>+ for i = 1, n_iterations do
>+ if i % 2 == 0 then
>+ c_payload(10)
>+ else
>+ lua_payload(10)
>+ end
>+ coroutine.yield()
>+ end
>+ end)
>+
>+ for _ = 1, n_iterations do
>+ coroutine.resume(co)
>+ end
>+end
>+
>+M.profile_func_jiton = payload
>+M.profile_func_jitoff = payload
>+
>+return M
>diff --git a/test/tarantool-c-tests/misclib-sysprof-capi.test.c b/test/tarantool-c-tests/misclib-sysprof-capi.test.c
>new file mode 100644
>index 00000000..4c54877e
>--- /dev/null
>+++ b/test/tarantool-c-tests/misclib-sysprof-capi.test.c
>@@ -0,0 +1,317 @@
>+#include "lauxlib.h"
>+#include "lmisclib.h"
>+#include "lua.h"
>+#include "luajit.h"
>+
>+#include "test.h"
>+#include "utils.h"
>+
>+#include <errno.h>
>+#include <fcntl.h>
>+#include <stdlib.h>
>+#include <unistd.h>
>+
>+/* XXX: Still need normal assert inside writer functions. */
>+#undef NDEBUG
>+#include <assert.h>
>+
>+/* Need for skipcond for OS and ARCH. */
>+#include "lj_arch.h"
Side note: I don’t like the approach with private headers, but
I couldn’t find any better way to check that. Maybe it is a good
idea to implement a public C API function to get the information
about OS and ARCH, since it is a really common to check them?
<snipped>
>
>diff --git a/test/tarantool-tests/CMakeLists.txt b/test/tarantool-tests/CMakeLists.txt
>index b4ce39d3..b1c7207f 100644
>--- a/test/tarantool-tests/CMakeLists.txt
>+++ b/test/tarantool-tests/CMakeLists.txt
<snipped>
>
> # The part of the memory profiler toolchain is located in tools
> # directory, jit, profiler, and bytecode toolchains are located
>diff --git a/test/tarantool-tests/misclib-sysprof-capi.test.lua b/test/tarantool-tests/misclib-sysprof-capi.test.lua
>deleted file mode 100644
>index 4395bce3..00000000
>--- a/test/tarantool-tests/misclib-sysprof-capi.test.lua
>+++ /dev/null
<snipped>
>diff --git a/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt b/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
>deleted file mode 100644
>index d9fb1a1a..00000000
>--- a/test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
>+++ /dev/null
>@@ -1 +0,0 @@
<snipped>
>diff --git a/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c b/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
>deleted file mode 100644
>index d7a3e355..00000000
>--- a/test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
>+++ /dev/null
>@@ -1,260 +0,0 @@
<snipped>
---
Best regards,
Maxim Kokryashkin
[-- Attachment #2: Type: text/html, Size: 6748 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests
2023-03-15 16:11 [Tarantool-patches] [PATCH v1 luajit 0/5] reworking C tests Sergey Kaplun via Tarantool-patches
` (4 preceding siblings ...)
2023-03-15 16:11 ` [Tarantool-patches] [PATCH v1 luajit 5/5] test: rewrite misclib-sysprof-capi " Sergey Kaplun via Tarantool-patches
@ 2023-03-20 13:50 ` Maxim Kokryashkin via Tarantool-patches
5 siblings, 0 replies; 12+ messages in thread
From: Maxim Kokryashkin via Tarantool-patches @ 2023-03-20 13:50 UTC (permalink / raw)
To: Sergey Kaplun; +Cc: tarantool-patches
[-- Attachment #1: Type: text/plain, Size: 3078 bytes --]
Hi!
Are you sure that issues tarantool/tarantool#781 is related?
--
Best regards,
Maxim Kokryashkin
>Среда, 15 марта 2023, 19:14 +03:00 от Sergey Kaplun <skaplun@tarantool.org>:
>
>The whole idea of the patch-set introduce module for LuaJIT C tests. It
>also, can be used for unit tests.
>* The first patch is the prerequisite for the patch-set. It fixes
> LD_LIBRARY_PATH definition.
>* The 2nd and 3d patches provides an API and helper for writing the tests.
>* The last 2 patches rewrite existing tests that should be written in C in
> the proper way.
>
>I'll be glad to hear some suggestions and ideas to improving the test
>module:).
>
>Branch: https://github.com/tarantool/luajit/tree/skaplun/gh-noticket-tarantool-c-tests
>PR: https://github.com/tarantool/tarantool/pull/8444
>Related Issue:
>* https://github.com/tarantool/tarantool/issues/7900
>* https://github.com/tarantool/tarantool/issues/781
>
>Sergey Kaplun (5):
> test: fix setting of {DY}LD_LIBRARY_PATH variables
> test: introduce module for C tests
> test: introduce utils.h helper for C tests
> test: rewrite misclib-getmetrics-capi test in C
> test: rewrite misclib-sysprof-capi test in C
>
> .gitignore | 1 +
> src/CMakeLists.txt | 2 +
> test/CMakeLists.txt | 2 +
> test/tarantool-c-tests/CMakeLists.txt | 67 ++++
> .../misclib-getmetrics-capi-script.lua} | 82 ++---
> .../misclib-getmetrics-capi.test.c | 341 ++++++++++++++++++
> .../misclib-sysprof-capi-script.lua | 35 ++
> .../misclib-sysprof-capi.test.c | 317 ++++++++++++++++
> test/tarantool-c-tests/test.c | 251 +++++++++++++
> test/tarantool-c-tests/test.h | 251 +++++++++++++
> test/tarantool-c-tests/utils.h | 63 ++++
> test/tarantool-tests/CMakeLists.txt | 11 +-
> .../misclib-getmetrics-capi/CMakeLists.txt | 1 -
> .../misclib-getmetrics-capi/testgetmetrics.c | 270 --------------
> .../misclib-sysprof-capi.test.lua | 54 ---
> .../misclib-sysprof-capi/CMakeLists.txt | 1 -
> .../misclib-sysprof-capi/testsysprof.c | 260 -------------
> 17 files changed, 1374 insertions(+), 635 deletions(-)
> create mode 100644 test/tarantool-c-tests/CMakeLists.txt
> rename test/{tarantool-tests/misclib-getmetrics-capi.test.lua => tarantool-c-tests/misclib-getmetrics-capi-script.lua} (68%)
> create mode 100644 test/tarantool-c-tests/misclib-getmetrics-capi.test.c
> create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi-script.lua
> create mode 100644 test/tarantool-c-tests/misclib-sysprof-capi.test.c
> create mode 100644 test/tarantool-c-tests/test.c
> create mode 100644 test/tarantool-c-tests/test.h
> create mode 100644 test/tarantool-c-tests/utils.h
> delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/CMakeLists.txt
> delete mode 100644 test/tarantool-tests/misclib-getmetrics-capi/testgetmetrics.c
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi.test.lua
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/CMakeLists.txt
> delete mode 100644 test/tarantool-tests/misclib-sysprof-capi/testsysprof.c
>
>--
>2.34.1
[-- Attachment #2: Type: text/html, Size: 4159 bytes --]
^ permalink raw reply [flat|nested] 12+ messages in thread