c/libuv
1
0
Fork 0
mirror of https://github.com/libuv/libuv synced 2025-03-28 21:13:16 +00:00
Code Issues Packages Projects Releases Wiki Activity
libuv/test/test-ref.c

446 lines
11 KiB
C
Raw Normal View History

Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
/* Copyright Joyent, Inc. and other Node contributors. All rights reserved.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to
* deal in the Software without restriction, including without limitation the
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
* sell copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
* IN THE SOFTWARE.
*/
#include "uv.h"
#include "task.h"
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
#include <stdlib.h>
#include <string.h>
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
static uv_write_t write_req;
static uv_shutdown_t shutdown_req;
static uv_connect_t connect_req;
static char buffer[32767];
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
static int req_cb_called;
static int connect_cb_called;
static int write_cb_called;
static int shutdown_cb_called;
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
static int close_cb_called;
static void close_cb(uv_handle_t* handle) {
close_cb_called++;
}
static void do_close(void* handle) {
close_cb_called = 0;
uv_close((uv_handle_t*)handle, close_cb);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(close_cb_called);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, close_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
}
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
static void fail_cb(void) {
FATAL("fail_cb should not have been called");
}
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
test, bench: ANSI-fy function prototypes Replace `void f()` with `void f(void)`; the former means "a function that takes any number of arguments, including none" while the latter is what is actually intended: a function taking no arguments. The first form also isn't strictly conforming ANSI/ISO C.
2013-01-06 20:01:22 +01:00
static void fail_cb2(void) {
prepare/idle/check: don't allow NULL callback
2012-11-30 00:32:04 +01:00
ASSERT(0 && "fail_cb2 should not have been called");
}
build: ubsan fixes (#4254) MSVC does not actually support ubsan. There is a long-standing ticket requesting this: https://developercommunity.visualstudio.com/t/add-support-for-ubsan/840750 There are no known compilers that currently accept the `/fsanitize=undefined` spelling. clang-cl accepts `-fsanitize...`, same as regular clang. Also passes no-sanitizer-recover so that tests actually fail. Fix various ubsan-detected errors, including: * win: fix req-inl.h ubsan failure Don't use CONTAINING_RECORD macro from WinSDK, as it doesn't use the right trick which avoids member access on null pointer. Fixes: ``` src/win/req-inl.h:86:10: runtime error: member access within null pointer of type 'uv_req_t' (aka 'struct uv_req_s') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior D:/a/libuv/libuv/src/win/req-inl.h:86:10 ``` * test: fix ubsan failure on udp_ref3 Don't call functions through different function type. Fixes: ``` src/win/udp.c:537:5: runtime error: call to function req_cb through pointer to incorrect function type 'void (*)(struct uv_udp_send_s *, int)' test\test-ref.c:66: note: req_cb defined here SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/udp.c:537:5 in ``` * win: fix process-stdio.c ubsan failure When accessing HANDLEs within the stdio buffer, use memcpy / memset in order to respect alignment. Fixes: ``` src/win/process-stdio.c:197:5: runtime error: store to misaligned address 0x0230ee72d107 for type 'HANDLE' (aka 'void *'), which requires 8 byte alignment 0x0230ee72d107: note: pointer points here 00 00 cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd fd fd fd fd ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/process-stdio.c:197:5 in ``` * win: fix getaddrinfo.c ubsan failure Reworks buffer alignment handling to respect requirements. Fixes: ``` src/win/getaddrinfo.c:157:23: runtime error: member access within misaligned address 0x0290e4c6a17c for type 'struct addrinfo', which requires 8 byte alignment 0x0290e4c6a17c: note: pointer points here 00 00 00 00 cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/getaddrinfo.c:157:23 in ``` * win: fix pipe.c ubsan failure Changes "random" representation from pointer to number. Fixes: ``` src/win/pipe.c:234:11: runtime error: applying non-zero offset to non-null pointer 0xffffffffffffffff produced null pointer SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/pipe.c:234:11 in ``` * unix: fix stream.c ubsan failure Avoids performing pointer arithmetic on null pointer. Fixes: ``` src/unix/stream.c:701:15: runtime error: applying zero offset to null pointer SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /Users/runner/work/libuv/libuv/src/unix/stream.c:701:15 in ```
2024-08-05 18:15:53 -03:00
static void req_cb(uv_udp_send_t* req, int status) {
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
req_cb_called++;
}
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
static void shutdown_cb(uv_shutdown_t* req, int status) {
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_PTR_EQ(req, &shutdown_req);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
shutdown_cb_called++;
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
}
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
static void write_cb(uv_write_t* req, int status) {
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_PTR_EQ(req, &write_req);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_shutdown(&shutdown_req, req->handle, shutdown_cb);
write_cb_called++;
}
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
static void connect_and_write(uv_connect_t* req, int status) {
uv_buf_t buf = uv_buf_init(buffer, sizeof buffer);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_PTR_EQ(req, &connect_req);
ASSERT_OK(status);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_write(&write_req, req->handle, &buf, 1, write_cb);
connect_cb_called++;
}
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
static void connect_and_shutdown(uv_connect_t* req, int status) {
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_PTR_EQ(req, &connect_req);
ASSERT_OK(status);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_shutdown(&shutdown_req, req->handle, shutdown_cb);
connect_cb_called++;
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
}
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
TEST_IMPL(ref) {
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
return 0;
}
TEST_IMPL(idle_ref) {
uv_idle_t h;
unix: multiplicity
2011-08-31 04:18:29 +02:00
uv_idle_init(uv_default_loop(), &h);
test, bench: ANSI-fy function prototypes Replace `void f()` with `void f(void)`; the former means "a function that takes any number of arguments, including none" while the latter is what is actually intended: a function taking no arguments. The first form also isn't strictly conforming ANSI/ISO C.
2013-01-06 20:01:22 +01:00
uv_idle_start(&h, (uv_idle_cb) fail_cb2);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
return 0;
}
TEST_IMPL(async_ref) {
uv_async_t h;
unix: multiplicity
2011-08-31 04:18:29 +02:00
uv_async_init(uv_default_loop(), &h, NULL);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
return 0;
}
TEST_IMPL(prepare_ref) {
uv_prepare_t h;
unix: multiplicity
2011-08-31 04:18:29 +02:00
uv_prepare_init(uv_default_loop(), &h);
test, bench: ANSI-fy function prototypes Replace `void f()` with `void f(void)`; the former means "a function that takes any number of arguments, including none" while the latter is what is actually intended: a function taking no arguments. The first form also isn't strictly conforming ANSI/ISO C.
2013-01-06 20:01:22 +01:00
uv_prepare_start(&h, (uv_prepare_cb) fail_cb2);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
return 0;
}
TEST_IMPL(check_ref) {
uv_check_t h;
unix: multiplicity
2011-08-31 04:18:29 +02:00
uv_check_init(uv_default_loop(), &h);
test, bench: ANSI-fy function prototypes Replace `void f()` with `void f(void)`; the former means "a function that takes any number of arguments, including none" while the latter is what is actually intended: a function taking no arguments. The first form also isn't strictly conforming ANSI/ISO C.
2013-01-06 20:01:22 +01:00
uv_check_start(&h, (uv_check_cb) fail_cb2);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Move loop reference tests to their own file
2011-07-21 14:51:56 +02:00
return 0;
}
Test that loop refs going down to zero in a prepare callback does not hang the event loop
2011-07-21 14:53:40 +02:00
unix, windows: removed unused status parameter async, timer, prepare, idle and check handles don't need the status parameter.
2014-03-11 00:34:40 +01:00
static void prepare_cb(uv_prepare_t* h) {
test: move to ASSERT_NULL and ASSERT_NOT_NULL test macros Moving to new style test macros will make debugging easier in case of test failure and improve redability. This commit will replace all ASSERT macros matching the statement: `ASSERT(identifier (== or !=) value);` to: `ASSERT_(NOT_)NULL(identifier);` Refs: https://github.com/libuv/libuv/issues/2974 PR-URL: https://github.com/libuv/libuv/pull/3081 Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
2020-12-29 13:56:10 +05:30
ASSERT_NOT_NULL(h);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)h);
Test that loop refs going down to zero in a prepare callback does not hang the event loop
2011-07-21 14:53:40 +02:00
}
TEST_IMPL(unref_in_prepare_cb) {
uv_prepare_t h;
unix: multiplicity
2011-08-31 04:18:29 +02:00
uv_prepare_init(uv_default_loop(), &h);
Test that loop refs going down to zero in a prepare callback does not hang the event loop
2011-07-21 14:53:40 +02:00
uv_prepare_start(&h, prepare_cb);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Test that loop refs going down to zero in a prepare callback does not hang the event loop
2011-07-21 14:53:40 +02:00
return 0;
}
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
TEST_IMPL(timer_ref) {
uv_timer_t h;
uv_timer_init(uv_default_loop(), &h);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(timer_ref2) {
uv_timer_t h;
uv_timer_init(uv_default_loop(), &h);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_timer_start(&h, (uv_timer_cb)fail_cb, 42, 42);
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(fs_event_ref) {
test: factor out fsevents skip explanation Factor out a dedicated test macro for use on platforms that do not support fsevents instead of duplicating the os390 platform condition and explanation text. PR-URL: https://github.com/libuv/libuv/pull/1312 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Santiago Gimeno <santiago.gimeno@gmail.com>
2017-02-20 22:03:11 -05:00
#if defined(NO_FS_EVENTS)
RETURN_SKIP(NO_FS_EVENTS);
zos: add support for new platform - zos: disable test cases not applicable - zos: build options - zos: semaphore implementation - zos: use compare and swap builtins - zos: struct rusage not the same as other platforms - zos: backlog<=0 produces undefined behaviour Will redefine backlog in the following way * if backlog == 0, set it to 1 * if backlog < 0, set it to SOMAXCONN - zos: define IMAXBEL as empty flag and implement uv__tty_make_raw - zos: use udp multicast operations from aix - zos: ESC in ebcdic - zos: use LIBPATH for dynamic linker path - zos: uv_udp_set_ttl only works for ipv6 - zos: increase pthread stack size by factor of 4 - zos: return ENODEV instead of ENXIO errors for setsockopt - zos: use uv_cond_init the same way as aix - test: enable oob test for zos - zos: return EINVAL for zos error code EOPNOTSUPP PR-URL: https://github.com/libuv/libuv/pull/937 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Saúl Ibarra Corretgé <saghul@gmail.com>
2016-07-05 18:37:45 -04:00
#endif
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_fs_event_t h;
unix, windows: add uv_fs_event_start/stop functions Make uv_fs_event behave like other handles, that is, it's inactive after init, and it's active between start and stop.
2013-09-24 00:08:31 +02:00
uv_fs_event_init(uv_default_loop(), &h);
uv_fs_event_start(&h, (uv_fs_event_cb)fail_cb, ".", 0);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
unix, windows: add stat() based file watcher Monitors a file path for changes. Supersedes ev_stat.
2012-06-14 07:30:08 +02:00
TEST_IMPL(fs_poll_ref) {
uv_fs_poll_t h;
uv_fs_poll_init(uv_default_loop(), &h);
uv_fs_poll_start(&h, NULL, ".", 999);
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
unix, windows: add stat() based file watcher Monitors a file path for changes. Supersedes ev_stat.
2012-06-14 07:30:08 +02:00
return 0;
}
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
TEST_IMPL(tcp_ref) {
uv_tcp_t h;
uv_tcp_init(uv_default_loop(), &h);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(tcp_ref2) {
uv_tcp_t h;
uv_tcp_init(uv_default_loop(), &h);
uv_listen((uv_stream_t*)&h, 128, (uv_connection_cb)fail_cb);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
windows: closing handles should always keep the loop alive This makes the tcp-ref2 and udp-ref2 tests pass again. Also adds another reference count test.
2012-10-25 15:10:16 +02:00
TEST_IMPL(tcp_ref2b) {
uv_tcp_t h;
uv_tcp_init(uv_default_loop(), &h);
uv_listen((uv_stream_t*)&h, 128, (uv_connection_cb)fail_cb);
uv_unref((uv_handle_t*)&h);
test: simplify tcp_ref2b Said test doesn't need its own close callback, it can piggyback on the common close callback.
2013-01-06 22:27:40 +01:00
uv_close((uv_handle_t*)&h, close_cb);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, close_cb_called);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
windows: closing handles should always keep the loop alive This makes the tcp-ref2 and udp-ref2 tests pass again. Also adds another reference count test.
2012-10-25 15:10:16 +02:00
return 0;
}
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
TEST_IMPL(tcp_ref3) {
include: uv_ip[46]_addr now takes sockaddr_in* Passing or returning structs as values makes life hard for people that work with libuv through a foreign function interface. Switch to a pointer-based approach. Fixes #684.
2013-08-31 03:00:34 +02:00
struct sockaddr_in addr;
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_tcp_t h;
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(uv_ip4_addr("127.0.0.1", TEST_PORT, &addr));
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_tcp_init(uv_default_loop(), &h);
include: merge uv_tcp_connect and uv_tcp_connect6 Merge uv_tcp_connect6() into uv_tcp_connect(). uv_tcp_connect() now takes a const struct sockaddr*.
2013-09-03 23:12:07 +02:00
uv_tcp_connect(&connect_req,
&h,
(const struct sockaddr*) &addr,
connect_and_shutdown);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, connect_cb_called);
ASSERT_EQ(1, shutdown_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
TEST_IMPL(tcp_ref4) {
include: uv_ip[46]_addr now takes sockaddr_in* Passing or returning structs as values makes life hard for people that work with libuv through a foreign function interface. Switch to a pointer-based approach. Fixes #684.
2013-08-31 03:00:34 +02:00
struct sockaddr_in addr;
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
uv_tcp_t h;
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(uv_ip4_addr("127.0.0.1", TEST_PORT, &addr));
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
uv_tcp_init(uv_default_loop(), &h);
include: merge uv_tcp_connect and uv_tcp_connect6 Merge uv_tcp_connect6() into uv_tcp_connect(). uv_tcp_connect() now takes a const struct sockaddr*.
2013-09-03 23:12:07 +02:00
uv_tcp_connect(&connect_req,
&h,
(const struct sockaddr*) &addr,
connect_and_write);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, connect_cb_called);
ASSERT_EQ(1, write_cb_called);
ASSERT_EQ(1, shutdown_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
return 0;
}
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
TEST_IMPL(udp_ref) {
uv_udp_t h;
uv_udp_init(uv_default_loop(), &h);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(udp_ref2) {
include: uv_ip[46]_addr now takes sockaddr_in* Passing or returning structs as values makes life hard for people that work with libuv through a foreign function interface. Switch to a pointer-based approach. Fixes #684.
2013-08-31 03:00:34 +02:00
struct sockaddr_in addr;
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_udp_t h;
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(uv_ip4_addr("127.0.0.1", TEST_PORT, &addr));
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_udp_init(uv_default_loop(), &h);
include: merge uv_udp_bind and uv_udp_bind6 Merge uv_udp_bind6() into uv_udp_bind(). uv_udp_bind() now takes a const struct sockaddr*.
2013-09-03 23:49:26 +02:00
uv_udp_bind(&h, (const struct sockaddr*) &addr, 0);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_udp_recv_start(&h, (uv_alloc_cb)fail_cb, (uv_udp_recv_cb)fail_cb);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(udp_ref3) {
include: uv_ip[46]_addr now takes sockaddr_in* Passing or returning structs as values makes life hard for people that work with libuv through a foreign function interface. Switch to a pointer-based approach. Fixes #684.
2013-08-31 03:00:34 +02:00
struct sockaddr_in addr;
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_buf_t buf = uv_buf_init("PING", 4);
uv_udp_send_t req;
uv_udp_t h;
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(uv_ip4_addr("127.0.0.1", TEST_PORT, &addr));
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
uv_udp_init(uv_default_loop(), &h);
include: merge uv_udp_send and uv_udp_send6 Merge uv_udp_send6() into uv_udp_send(). uv_udp_send() now takes a const struct sockaddr*.
2013-09-04 01:48:50 +02:00
uv_udp_send(&req,
&h,
&buf,
1,
(const struct sockaddr*) &addr,
build: ubsan fixes (#4254) MSVC does not actually support ubsan. There is a long-standing ticket requesting this: https://developercommunity.visualstudio.com/t/add-support-for-ubsan/840750 There are no known compilers that currently accept the `/fsanitize=undefined` spelling. clang-cl accepts `-fsanitize...`, same as regular clang. Also passes no-sanitizer-recover so that tests actually fail. Fix various ubsan-detected errors, including: * win: fix req-inl.h ubsan failure Don't use CONTAINING_RECORD macro from WinSDK, as it doesn't use the right trick which avoids member access on null pointer. Fixes: ``` src/win/req-inl.h:86:10: runtime error: member access within null pointer of type 'uv_req_t' (aka 'struct uv_req_s') SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior D:/a/libuv/libuv/src/win/req-inl.h:86:10 ``` * test: fix ubsan failure on udp_ref3 Don't call functions through different function type. Fixes: ``` src/win/udp.c:537:5: runtime error: call to function req_cb through pointer to incorrect function type 'void (*)(struct uv_udp_send_s *, int)' test\test-ref.c:66: note: req_cb defined here SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/udp.c:537:5 in ``` * win: fix process-stdio.c ubsan failure When accessing HANDLEs within the stdio buffer, use memcpy / memset in order to respect alignment. Fixes: ``` src/win/process-stdio.c:197:5: runtime error: store to misaligned address 0x0230ee72d107 for type 'HANDLE' (aka 'void *'), which requires 8 byte alignment 0x0230ee72d107: note: pointer points here 00 00 cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd fd fd fd fd ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/process-stdio.c:197:5 in ``` * win: fix getaddrinfo.c ubsan failure Reworks buffer alignment handling to respect requirements. Fixes: ``` src/win/getaddrinfo.c:157:23: runtime error: member access within misaligned address 0x0290e4c6a17c for type 'struct addrinfo', which requires 8 byte alignment 0x0290e4c6a17c: note: pointer points here 00 00 00 00 cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd cd ^ SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/getaddrinfo.c:157:23 in ``` * win: fix pipe.c ubsan failure Changes "random" representation from pointer to number. Fixes: ``` src/win/pipe.c:234:11: runtime error: applying non-zero offset to non-null pointer 0xffffffffffffffff produced null pointer SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior src/win/pipe.c:234:11 in ``` * unix: fix stream.c ubsan failure Avoids performing pointer arithmetic on null pointer. Fixes: ``` src/unix/stream.c:701:15: runtime error: applying zero offset to null pointer SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /Users/runner/work/libuv/libuv/src/unix/stream.c:701:15 in ```
2024-08-05 18:15:53 -03:00
req_cb);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, req_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(pipe_ref) {
uv_pipe_t h;
uv_pipe_init(uv_default_loop(), &h, 0);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(pipe_ref2) {
uv_pipe_t h;
uv_pipe_init(uv_default_loop(), &h, 0);
uv_listen((uv_stream_t*)&h, 128, (uv_connection_cb)fail_cb);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
TEST_IMPL(pipe_ref3) {
uv_pipe_t h;
uv_pipe_init(uv_default_loop(), &h, 0);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_pipe_connect(&connect_req, &h, TEST_PIPENAME, connect_and_shutdown);
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, connect_cb_called);
ASSERT_EQ(1, shutdown_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
TEST_IMPL(pipe_ref4) {
uv_pipe_t h;
uv_pipe_init(uv_default_loop(), &h, 0);
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_pipe_connect(&connect_req, &h, TEST_PIPENAME, connect_and_write);
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, connect_cb_called);
ASSERT_EQ(1, write_cb_called);
ASSERT_EQ(1, shutdown_cb_called);
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
Tests: verify that uv_write and uv_shutdown ref the event loop
2012-03-09 04:26:51 +01:00
return 0;
}
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
TEST_IMPL(process_ref) {
/* spawn_helper4 blocks indefinitely. */
char *argv[] = { NULL, "spawn_helper4", NULL };
uv_process_options_t options;
size_t exepath_size;
char exepath[256];
uv_process_t h;
int r;
memset(&options, 0, sizeof(options));
exepath_size = sizeof(exepath);
r = uv_exepath(exepath, &exepath_size);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(r);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
argv[0] = exepath;
options.file = exepath;
options.args = argv;
options.exit_cb = NULL;
include: uv_spawn takes const uv_process_options_t* Passing or returning structs as values makes life hard for people that work with libuv through a foreign function interface. Switch to a pointer-based approach. Fixes #684.
2013-09-01 08:20:15 +02:00
r = uv_spawn(uv_default_loop(), &h, &options);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(r);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
unix, windows: rework reference counting scheme This commit changes how the event loop determines if it needs to stay alive. Previously, an internal counter was increased whenever a handle got created and decreased again when the handle was closed. While conceptually simple, it turned out hard to work with: you often want to keep the event loop alive only if the handle is actually doing something. Stopped or inactive handles were a frequent source of hanging event loops. That's why this commit changes the reference counting scheme to a model where a handle only references the event loop when it's active. 'Active' means different things for different handle types, e.g.: * timers: ticking * sockets: reading, writing or listening * processes: always active (for now, subject to change) * idle, check, prepare: only active when started This commit also changes how the uv_ref() and uv_unref() functions work: they now operate on the level of individual handles, not the whole event loop. The Windows implementation was done by Bert Belder.
2012-05-17 06:28:46 +02:00
uv_unref((uv_handle_t*)&h);
unix, windows: rename uv_run2 to uv_run This changes the prototype of uv_run() from: int uv_run(uv_loop_t* loop); To: int uv_run(uv_loop_t* loop, uv_run_mode mode); Where `mode` is UV_RUN_DEFAULT, UV_RUN_ONCE or UV_RUN_NOWAIT. Fixes #683.
2013-01-16 23:20:03 +01:00
uv_run(uv_default_loop(), UV_RUN_DEFAULT);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
r = uv_process_kill(&h, /* SIGTERM */ 15);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(r);
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
test: close handle in ref tests Everything that gets reported by valgrind now is an actual memory leak.
2012-10-01 14:07:10 +02:00
do_close(&h);
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
test: add lots of refcount tests
2012-01-14 00:11:27 +01:00
return 0;
}
unix, windows: add uv_has_ref() function
2013-04-17 00:16:42 +02:00
TEST_IMPL(has_ref) {
uv_idle_t h;
uv_idle_init(uv_default_loop(), &h);
uv_ref((uv_handle_t*)&h);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_EQ(1, uv_has_ref((uv_handle_t*)&h));
unix, windows: add uv_has_ref() function
2013-04-17 00:16:42 +02:00
uv_unref((uv_handle_t*)&h);
test: switch to new-style ASSERT_EQ macros (#4159) Switch from old-style ASSERT macro to new-style ASSERT_EQ,... macros. Using new-style macros makes it easier to debug test failures Fixes: https://github.com/libuv/libuv/issues/2974
2023-10-06 19:50:15 +02:00
ASSERT_OK(uv_has_ref((uv_handle_t*)&h));
test: silence more valgrind warnings (#3917) Pass the loop to MAKE_VALGRIND_HAPPY() so it's explicit on which loop needs to be cleaned up. Since it asserts on uv_loop_close(), need to remove a couple of those that were being done before the call. Cleanup where loop was assigned, so the entire test either uses loop or uv_default_loop(). Not both. Also take care of any reqs that may have been left uncleaned.
2023-03-12 07:59:00 -06:00
MAKE_VALGRIND_HAPPY(uv_default_loop());
unix, windows: add uv_has_ref() function
2013-04-17 00:16:42 +02:00
return 0;
}
Reference in New Issue Copy Permalink