blob: a48714961c175b7504ae0b076f04d85b85ccc8dc [file] [log] [blame]
/* Unit tests for GCond
* Copyright (C) 2011 Red Hat, Inc
* Author: Matthias Clasen
*
* This work is provided "as is"; redistribution and modification
* in whole or in part, in any medium, physical or electronic is
* permitted without restriction.
*
* This work is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
*
* In no event shall the authors or contributors be liable for any
* direct, indirect, incidental, special, exemplary, or consequential
* damages (including, but not limited to, procurement of substitute
* goods or services; loss of use, data, or profits; or business
* interruption) however caused and on any theory of liability, whether
* in contract, strict liability, or tort (including negligence or
* otherwise) arising in any way out of the use of this software, even
* if advised of the possibility of such damage.
*/
/* We are testing some deprecated APIs here */
#define GLIB_DISABLE_DEPRECATION_WARNINGS
#include <glib.h>
static GCond cond;
static GMutex mutex;
static volatile gint next;
static void
push_value (gint value)
{
g_mutex_lock (&mutex);
while (next != 0)
g_cond_wait (&cond, &mutex);
next = value;
if (g_test_verbose ())
g_printerr ("Thread %p producing next value: %d\n", g_thread_self (), value);
if (value % 10 == 0)
g_cond_broadcast (&cond);
else
g_cond_signal (&cond);
g_mutex_unlock (&mutex);
}
static gint
pop_value (void)
{
gint value;
g_mutex_lock (&mutex);
while (next == 0)
{
if (g_test_verbose ())
g_printerr ("Thread %p waiting for cond\n", g_thread_self ());
g_cond_wait (&cond, &mutex);
}
value = next;
next = 0;
g_cond_broadcast (&cond);
if (g_test_verbose ())
g_printerr ("Thread %p consuming value %d\n", g_thread_self (), value);
g_mutex_unlock (&mutex);
return value;
}
static gpointer
produce_values (gpointer data)
{
gint total;
gint i;
total = 0;
for (i = 1; i < 100; i++)
{
total += i;
push_value (i);
}
push_value (-1);
push_value (-1);
if (g_test_verbose ())
g_printerr ("Thread %p produced %d altogether\n", g_thread_self (), total);
return GINT_TO_POINTER (total);
}
static gpointer
consume_values (gpointer data)
{
gint accum = 0;
gint value;
while (TRUE)
{
value = pop_value ();
if (value == -1)
break;
accum += value;
}
if (g_test_verbose ())
g_printerr ("Thread %p accumulated %d\n", g_thread_self (), accum);
return GINT_TO_POINTER (accum);
}
static GThread *producer, *consumer1, *consumer2;
static void
test_cond1 (void)
{
gint total, acc1, acc2;
producer = g_thread_create (produce_values, NULL, TRUE, NULL);
consumer1 = g_thread_create (consume_values, NULL, TRUE, NULL);
consumer2 = g_thread_create (consume_values, NULL, TRUE, NULL);
total = GPOINTER_TO_INT (g_thread_join (producer));
acc1 = GPOINTER_TO_INT (g_thread_join (consumer1));
acc2 = GPOINTER_TO_INT (g_thread_join (consumer2));
g_assert_cmpint (total, ==, acc1 + acc2);
}
typedef struct
{
GMutex mutex;
GCond cond;
gint limit;
gint count;
} Barrier;
static void
barrier_init (Barrier *barrier,
gint limit)
{
g_mutex_init (&barrier->mutex);
g_cond_init (&barrier->cond);
barrier->limit = limit;
barrier->count = limit;
}
static gint
barrier_wait (Barrier *barrier)
{
gint ret;
g_mutex_lock (&barrier->mutex);
barrier->count--;
if (barrier->count == 0)
{
ret = -1;
barrier->count = barrier->limit;
g_cond_broadcast (&barrier->cond);
}
else
{
ret = 0;
while (barrier->count != barrier->limit)
g_cond_wait (&barrier->cond, &barrier->mutex);
}
g_mutex_unlock (&barrier->mutex);
return ret;
}
static void
barrier_clear (Barrier *barrier)
{
g_mutex_clear (&barrier->mutex);
g_cond_clear (&barrier->cond);
}
static Barrier b;
static gint check;
static gpointer
cond2_func (gpointer data)
{
gint value = GPOINTER_TO_INT (data);
gint ret;
g_atomic_int_inc (&check);
if (g_test_verbose ())
g_printerr ("thread %d starting, check %d\n", value, g_atomic_int_get (&check));
g_usleep (10000 * value);
g_atomic_int_inc (&check);
if (g_test_verbose ())
g_printerr ("thread %d reaching barrier, check %d\n", value, g_atomic_int_get (&check));
ret = barrier_wait (&b);
g_assert_cmpint (g_atomic_int_get (&check), ==, 10);
if (g_test_verbose ())
g_printerr ("thread %d leaving barrier (%d), check %d\n", value, ret, g_atomic_int_get (&check));
return NULL;
}
/* this test demonstrates how to use a condition variable
* to implement a barrier
*/
static void
test_cond2 (void)
{
gint i;
GThread *threads[5];
g_atomic_int_set (&check, 0);
barrier_init (&b, 5);
for (i = 0; i < 5; i++)
threads[i] = g_thread_create (cond2_func, GINT_TO_POINTER (i), TRUE, NULL);
for (i = 0; i < 5; i++)
g_thread_join (threads[i]);
g_assert_cmpint (g_atomic_int_get (&check), ==, 10);
barrier_clear (&b);
}
static void
test_wait_until (void)
{
gint64 until;
GMutex lock;
GCond cond;
/* This test will make sure we don't wait too much or too little.
*
* We check the 'too long' with a timeout of 60 seconds.
*
* We check the 'too short' by verifying a guarantee of the API: we
* should not wake up until the specified time has passed.
*/
g_mutex_init (&lock);
g_cond_init (&cond);
until = g_get_monotonic_time () + G_TIME_SPAN_SECOND;
/* Could still have spurious wakeups, so we must loop... */
g_mutex_lock (&lock);
while (g_cond_wait_until (&cond, &lock, until))
;
g_mutex_unlock (&lock);
/* Make sure it's after the until time */
g_assert_cmpint (until, <=, g_get_monotonic_time ());
/* Make sure it returns FALSE on timeout */
until = g_get_monotonic_time () + G_TIME_SPAN_SECOND / 50;
g_mutex_lock (&lock);
g_assert (g_cond_wait_until (&cond, &lock, until) == FALSE);
g_mutex_unlock (&lock);
g_mutex_clear (&lock);
g_cond_clear (&cond);
}
#ifdef __linux__
#include <pthread.h>
#include <signal.h>
#include <unistd.h>
static pthread_t main_thread;
static void *
mutex_holder (void *data)
{
GMutex *lock = data;
g_mutex_lock (lock);
/* Let the lock become contended */
g_usleep (G_TIME_SPAN_SECOND);
/* Interrupt the wait on the other thread */
pthread_kill (main_thread, SIGHUP);
/* If we don't sleep here, then the g_mutex_unlock() below will clear
* the mutex, causing the interrupted futex call in the other thread
* to return success (which is not what we want).
*
* The other thread needs to have time to wake up and see that the
* lock is still contended.
*/
g_usleep (G_TIME_SPAN_SECOND / 10);
g_mutex_unlock (lock);
return NULL;
}
static void
signal_handler (int sig)
{
}
static void
test_wait_until_errno (void)
{
gboolean result;
GMutex lock;
GCond cond;
struct sigaction act = { };
/* important: no SA_RESTART (we want EINTR) */
act.sa_handler = signal_handler;
g_test_summary ("Check proper handling of errno in g_cond_wait_until with a contended mutex");
g_test_bug_base ("https://gitlab.gnome.org/GNOME/glib/");
g_test_bug ("merge_requests/957");
g_mutex_init (&lock);
g_cond_init (&cond);
main_thread = pthread_self ();
sigaction (SIGHUP, &act, NULL);
g_mutex_lock (&lock);
/* We create an annoying worker thread that will do two things:
*
* 1) hold the lock that we want to reacquire after returning from
* the condition variable wait
*
* 2) send us a signal to cause our wait on the contended lock to
* return EINTR, clobbering the errno return from the condition
* variable
*/
g_thread_unref (g_thread_new ("mutex-holder", mutex_holder, &lock));
result = g_cond_wait_until (&cond, &lock,
g_get_monotonic_time () + G_TIME_SPAN_SECOND / 50);
/* Even after all that disruption, we should still successfully return
* 'timed out'.
*/
g_assert_false (result);
g_mutex_unlock (&lock);
g_cond_clear (&cond);
g_mutex_clear (&lock);
}
#else
static void
test_wait_until_errno (void)
{
g_test_skip ("We only test this on Linux");
}
#endif
int
main (int argc, char *argv[])
{
g_test_init (&argc, &argv, NULL);
g_test_add_func ("/thread/cond1", test_cond1);
g_test_add_func ("/thread/cond2", test_cond2);
g_test_add_func ("/thread/cond/wait-until", test_wait_until);
g_test_add_func ("/thread/cond/wait-until/contended-and-interrupted", test_wait_until_errno);
return g_test_run ();
}