qemu-e2k/tests/unit/test-aio.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

959 lines
28 KiB
C
Raw Normal View History

/*
* AioContext tests
*
* Copyright Red Hat, Inc. 2012
*
* Authors:
* Paolo Bonzini <pbonzini@redhat.com>
*
* This work is licensed under the terms of the GNU LGPL, version 2 or later.
* See the COPYING.LIB file in the top-level directory.
*/
#include "qemu/osdep.h"
#include "block/aio.h"
2016-03-14 09:01:28 +01:00
#include "qapi/error.h"
#include "qemu/timer.h"
#include "qemu/sockets.h"
#include "qemu/error-report.h"
#include "qemu/coroutine-core.h"
#include "qemu/main-loop.h"
static AioContext *ctx;
typedef struct {
EventNotifier e;
int n;
int active;
bool auto_set;
} EventNotifierTestData;
/* Wait until event notifier becomes inactive */
static void wait_until_inactive(EventNotifierTestData *data)
{
while (data->active > 0) {
aio_poll(ctx, true);
}
}
/* Simple callbacks for testing. */
typedef struct {
QEMUBH *bh;
int n;
int max;
} BHTestData;
typedef struct {
QEMUTimer timer;
QEMUClockType clock_type;
int n;
int max;
int64_t ns;
AioContext *ctx;
} TimerTestData;
static void bh_test_cb(void *opaque)
{
BHTestData *data = opaque;
if (++data->n < data->max) {
qemu_bh_schedule(data->bh);
}
}
static void timer_test_cb(void *opaque)
{
TimerTestData *data = opaque;
if (++data->n < data->max) {
timer_mod(&data->timer,
qemu_clock_get_ns(data->clock_type) + data->ns);
}
}
static void dummy_io_handler_read(EventNotifier *e)
{
}
static void bh_delete_cb(void *opaque)
{
BHTestData *data = opaque;
if (++data->n < data->max) {
qemu_bh_schedule(data->bh);
} else {
qemu_bh_delete(data->bh);
data->bh = NULL;
}
}
static void event_ready_cb(EventNotifier *e)
{
EventNotifierTestData *data = container_of(e, EventNotifierTestData, e);
g_assert(event_notifier_test_and_clear(e));
data->n++;
if (data->active > 0) {
data->active--;
}
if (data->auto_set && data->active) {
event_notifier_set(e);
}
}
/* Tests using aio_*. */
typedef struct {
QemuMutex start_lock;
EventNotifier notifier;
bool thread_acquired;
} AcquireTestData;
static void *test_acquire_thread(void *opaque)
{
AcquireTestData *data = opaque;
/* Wait for other thread to let us start */
qemu_mutex_lock(&data->start_lock);
qemu_mutex_unlock(&data->start_lock);
/* event_notifier_set might be called either before or after
* the main thread's call to poll(). The test case's outcome
* should be the same in either case.
*/
event_notifier_set(&data->notifier);
aio_context_acquire(ctx);
aio_context_release(ctx);
data->thread_acquired = true; /* success, we got here */
return NULL;
}
static void set_event_notifier(AioContext *ctx, EventNotifier *notifier,
EventNotifierHandler *handler)
{
aio-posix: split poll check from ready handler Adaptive polling measures the execution time of the polling check plus handlers called when a polled event becomes ready. Handlers can take a significant amount of time, making it look like polling was running for a long time when in fact the event handler was running for a long time. For example, on Linux the io_submit(2) syscall invoked when a virtio-blk device's virtqueue becomes ready can take 10s of microseconds. This can exceed the default polling interval (32 microseconds) and cause adaptive polling to stop polling. By excluding the handler's execution time from the polling check we make the adaptive polling calculation more accurate. As a result, the event loop now stays in polling mode where previously it would have fallen back to file descriptor monitoring. The following data was collected with virtio-blk num-queues=2 event_idx=off using an IOThread. Before: 168k IOPS, IOThread syscalls: 9837.115 ( 0.020 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 16, iocbpp: 0x7fcb9f937db0) = 16 9837.158 ( 0.002 ms): IO iothread1/620155 write(fd: 103, buf: 0x556a2ef71b88, count: 8) = 8 9837.161 ( 0.001 ms): IO iothread1/620155 write(fd: 104, buf: 0x556a2ef71b88, count: 8) = 8 9837.163 ( 0.001 ms): IO iothread1/620155 ppoll(ufds: 0x7fcb90002800, nfds: 4, tsp: 0x7fcb9f1342d0, sigsetsize: 8) = 3 9837.164 ( 0.001 ms): IO iothread1/620155 read(fd: 107, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.174 ( 0.001 ms): IO iothread1/620155 read(fd: 105, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.176 ( 0.001 ms): IO iothread1/620155 read(fd: 106, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.209 ( 0.035 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 32, iocbpp: 0x7fca7d0cebe0) = 32 174k IOPS (+3.6%), IOThread syscalls: 9809.566 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0cdd62be0) = 32 9809.625 ( 0.001 ms): IO iothread1/623061 write(fd: 103, buf: 0x5647cfba5f58, count: 8) = 8 9809.627 ( 0.002 ms): IO iothread1/623061 write(fd: 104, buf: 0x5647cfba5f58, count: 8) = 8 9809.663 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0d0388b50) = 32 Notice that ppoll(2) and eventfd read(2) syscalls are eliminated because the IOThread stays in polling mode instead of falling back to file descriptor monitoring. As usual, polling is not implemented on Windows so this patch ignores the new io_poll_read() callback in aio-win32.c. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Message-id: 20211207132336.36627-2-stefanha@redhat.com [Fixed up aio_set_event_notifier() calls in tests/unit/test-fdmon-epoll.c added after this series was queued. --Stefan] Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2021-12-07 14:23:31 +01:00
aio_set_event_notifier(ctx, notifier, false, handler, NULL, NULL);
}
static void dummy_notifier_read(EventNotifier *n)
{
event_notifier_test_and_clear(n);
}
static void test_acquire(void)
{
QemuThread thread;
AcquireTestData data;
/* Dummy event notifier ensures aio_poll() will block */
event_notifier_init(&data.notifier, false);
set_event_notifier(ctx, &data.notifier, dummy_notifier_read);
g_assert(!aio_poll(ctx, false)); /* consume aio_notify() */
qemu_mutex_init(&data.start_lock);
qemu_mutex_lock(&data.start_lock);
data.thread_acquired = false;
qemu_thread_create(&thread, "test_acquire_thread",
test_acquire_thread,
&data, QEMU_THREAD_JOINABLE);
/* Block in aio_poll(), let other thread kick us and acquire context */
aio_context_acquire(ctx);
qemu_mutex_unlock(&data.start_lock); /* let the thread run */
g_assert(aio_poll(ctx, true));
g_assert(!data.thread_acquired);
aio_context_release(ctx);
qemu_thread_join(&thread);
set_event_notifier(ctx, &data.notifier, NULL);
event_notifier_cleanup(&data.notifier);
g_assert(data.thread_acquired);
}
static void test_bh_schedule(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(aio_poll(ctx, true));
g_assert_cmpint(data.n, ==, 1);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
qemu_bh_delete(data.bh);
}
static void test_bh_schedule10(void)
{
BHTestData data = { .n = 0, .max = 10 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
g_assert(aio_poll(ctx, true));
g_assert_cmpint(data.n, ==, 2);
AioContext: do not rely on aio_poll(ctx, true) result to end a loop Currently, whenever aio_poll(ctx, true) has completed all pending work it returns true *and* the next call to aio_poll(ctx, true) will not block. This invariant has its roots in qemu_aio_flush()'s implementation as "while (qemu_aio_wait()) {}". However, qemu_aio_flush() does not exist anymore and bdrv_drain_all() is implemented differently; and this invariant is complicated to maintain and subtly different from the return value of GMainLoop's g_main_context_iteration. All calls to aio_poll(ctx, true) except one are guarded by a while() loop checking for a request to be incomplete, or a BlockDriverState to be idle. The one remaining call (in iothread.c) uses this to delay the aio_context_release/acquire pair until the AioContext is quiescent, however: - we can do the same just by using non-blocking aio_poll, similar to how vl.c invokes main_loop_wait - it is buggy, because it does not ensure that the AioContext is released between an aio_notify and the next time the iothread goes to sleep. This leads to hangs when stopping the dataplane thread. In the end, these semantics are a bad match for the current users of AioContext. So modify that one exception in iothread.c, which also fixes the hangs, as well as the testcase so that it use the same idiom as the actual QEMU code. Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-07-09 10:49:46 +02:00
while (data.n < 10) {
aio_poll(ctx, true);
}
g_assert_cmpint(data.n, ==, 10);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 10);
qemu_bh_delete(data.bh);
}
static void test_bh_cancel(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
qemu_bh_cancel(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
qemu_bh_delete(data.bh);
}
static void test_bh_delete(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
qemu_bh_delete(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
}
static void test_bh_delete_from_cb(void)
{
BHTestData data1 = { .n = 0, .max = 1 };
data1.bh = aio_bh_new(ctx, bh_delete_cb, &data1);
qemu_bh_schedule(data1.bh);
g_assert_cmpint(data1.n, ==, 0);
AioContext: do not rely on aio_poll(ctx, true) result to end a loop Currently, whenever aio_poll(ctx, true) has completed all pending work it returns true *and* the next call to aio_poll(ctx, true) will not block. This invariant has its roots in qemu_aio_flush()'s implementation as "while (qemu_aio_wait()) {}". However, qemu_aio_flush() does not exist anymore and bdrv_drain_all() is implemented differently; and this invariant is complicated to maintain and subtly different from the return value of GMainLoop's g_main_context_iteration. All calls to aio_poll(ctx, true) except one are guarded by a while() loop checking for a request to be incomplete, or a BlockDriverState to be idle. The one remaining call (in iothread.c) uses this to delay the aio_context_release/acquire pair until the AioContext is quiescent, however: - we can do the same just by using non-blocking aio_poll, similar to how vl.c invokes main_loop_wait - it is buggy, because it does not ensure that the AioContext is released between an aio_notify and the next time the iothread goes to sleep. This leads to hangs when stopping the dataplane thread. In the end, these semantics are a bad match for the current users of AioContext. So modify that one exception in iothread.c, which also fixes the hangs, as well as the testcase so that it use the same idiom as the actual QEMU code. Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-07-09 10:49:46 +02:00
while (data1.n < data1.max) {
aio_poll(ctx, true);
}
g_assert_cmpint(data1.n, ==, data1.max);
g_assert(data1.bh == NULL);
g_assert(!aio_poll(ctx, false));
}
static void test_bh_delete_from_cb_many(void)
{
BHTestData data1 = { .n = 0, .max = 1 };
BHTestData data2 = { .n = 0, .max = 3 };
BHTestData data3 = { .n = 0, .max = 2 };
BHTestData data4 = { .n = 0, .max = 4 };
data1.bh = aio_bh_new(ctx, bh_delete_cb, &data1);
data2.bh = aio_bh_new(ctx, bh_delete_cb, &data2);
data3.bh = aio_bh_new(ctx, bh_delete_cb, &data3);
data4.bh = aio_bh_new(ctx, bh_delete_cb, &data4);
qemu_bh_schedule(data1.bh);
qemu_bh_schedule(data2.bh);
qemu_bh_schedule(data3.bh);
qemu_bh_schedule(data4.bh);
g_assert_cmpint(data1.n, ==, 0);
g_assert_cmpint(data2.n, ==, 0);
g_assert_cmpint(data3.n, ==, 0);
g_assert_cmpint(data4.n, ==, 0);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data1.n, ==, 1);
g_assert_cmpint(data2.n, ==, 1);
g_assert_cmpint(data3.n, ==, 1);
g_assert_cmpint(data4.n, ==, 1);
g_assert(data1.bh == NULL);
AioContext: do not rely on aio_poll(ctx, true) result to end a loop Currently, whenever aio_poll(ctx, true) has completed all pending work it returns true *and* the next call to aio_poll(ctx, true) will not block. This invariant has its roots in qemu_aio_flush()'s implementation as "while (qemu_aio_wait()) {}". However, qemu_aio_flush() does not exist anymore and bdrv_drain_all() is implemented differently; and this invariant is complicated to maintain and subtly different from the return value of GMainLoop's g_main_context_iteration. All calls to aio_poll(ctx, true) except one are guarded by a while() loop checking for a request to be incomplete, or a BlockDriverState to be idle. The one remaining call (in iothread.c) uses this to delay the aio_context_release/acquire pair until the AioContext is quiescent, however: - we can do the same just by using non-blocking aio_poll, similar to how vl.c invokes main_loop_wait - it is buggy, because it does not ensure that the AioContext is released between an aio_notify and the next time the iothread goes to sleep. This leads to hangs when stopping the dataplane thread. In the end, these semantics are a bad match for the current users of AioContext. So modify that one exception in iothread.c, which also fixes the hangs, as well as the testcase so that it use the same idiom as the actual QEMU code. Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-07-09 10:49:46 +02:00
while (data1.n < data1.max ||
data2.n < data2.max ||
data3.n < data3.max ||
data4.n < data4.max) {
aio_poll(ctx, true);
}
g_assert_cmpint(data1.n, ==, data1.max);
g_assert_cmpint(data2.n, ==, data2.max);
g_assert_cmpint(data3.n, ==, data3.max);
g_assert_cmpint(data4.n, ==, data4.max);
g_assert(data1.bh == NULL);
g_assert(data2.bh == NULL);
g_assert(data3.bh == NULL);
g_assert(data4.bh == NULL);
}
static void test_bh_flush(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
AioContext: do not rely on aio_poll(ctx, true) result to end a loop Currently, whenever aio_poll(ctx, true) has completed all pending work it returns true *and* the next call to aio_poll(ctx, true) will not block. This invariant has its roots in qemu_aio_flush()'s implementation as "while (qemu_aio_wait()) {}". However, qemu_aio_flush() does not exist anymore and bdrv_drain_all() is implemented differently; and this invariant is complicated to maintain and subtly different from the return value of GMainLoop's g_main_context_iteration. All calls to aio_poll(ctx, true) except one are guarded by a while() loop checking for a request to be incomplete, or a BlockDriverState to be idle. The one remaining call (in iothread.c) uses this to delay the aio_context_release/acquire pair until the AioContext is quiescent, however: - we can do the same just by using non-blocking aio_poll, similar to how vl.c invokes main_loop_wait - it is buggy, because it does not ensure that the AioContext is released between an aio_notify and the next time the iothread goes to sleep. This leads to hangs when stopping the dataplane thread. In the end, these semantics are a bad match for the current users of AioContext. So modify that one exception in iothread.c, which also fixes the hangs, as well as the testcase so that it use the same idiom as the actual QEMU code. Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2014-07-09 10:49:46 +02:00
g_assert(aio_poll(ctx, true));
g_assert_cmpint(data.n, ==, 1);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
qemu_bh_delete(data.bh);
}
static void test_set_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 0 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
set_event_notifier(ctx, &data.e, NULL);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
event_notifier_cleanup(&data.e);
}
static void test_wait_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 1 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
g_assert_cmpint(data.active, ==, 1);
event_notifier_set(&data.e);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 0);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 0);
set_event_notifier(ctx, &data.e, NULL);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
event_notifier_cleanup(&data.e);
}
static void test_flush_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 10, .auto_set = true };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
g_assert_cmpint(data.active, ==, 10);
event_notifier_set(&data.e);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 9);
g_assert(aio_poll(ctx, false));
wait_until_inactive(&data);
g_assert_cmpint(data.n, ==, 10);
g_assert_cmpint(data.active, ==, 0);
g_assert(!aio_poll(ctx, false));
set_event_notifier(ctx, &data.e, NULL);
g_assert(!aio_poll(ctx, false));
event_notifier_cleanup(&data.e);
}
static void test_aio_external_client(void)
{
int i, j;
for (i = 1; i < 3; i++) {
EventNotifierTestData data = { .n = 0, .active = 10, .auto_set = true };
event_notifier_init(&data.e, false);
aio-posix: split poll check from ready handler Adaptive polling measures the execution time of the polling check plus handlers called when a polled event becomes ready. Handlers can take a significant amount of time, making it look like polling was running for a long time when in fact the event handler was running for a long time. For example, on Linux the io_submit(2) syscall invoked when a virtio-blk device's virtqueue becomes ready can take 10s of microseconds. This can exceed the default polling interval (32 microseconds) and cause adaptive polling to stop polling. By excluding the handler's execution time from the polling check we make the adaptive polling calculation more accurate. As a result, the event loop now stays in polling mode where previously it would have fallen back to file descriptor monitoring. The following data was collected with virtio-blk num-queues=2 event_idx=off using an IOThread. Before: 168k IOPS, IOThread syscalls: 9837.115 ( 0.020 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 16, iocbpp: 0x7fcb9f937db0) = 16 9837.158 ( 0.002 ms): IO iothread1/620155 write(fd: 103, buf: 0x556a2ef71b88, count: 8) = 8 9837.161 ( 0.001 ms): IO iothread1/620155 write(fd: 104, buf: 0x556a2ef71b88, count: 8) = 8 9837.163 ( 0.001 ms): IO iothread1/620155 ppoll(ufds: 0x7fcb90002800, nfds: 4, tsp: 0x7fcb9f1342d0, sigsetsize: 8) = 3 9837.164 ( 0.001 ms): IO iothread1/620155 read(fd: 107, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.174 ( 0.001 ms): IO iothread1/620155 read(fd: 105, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.176 ( 0.001 ms): IO iothread1/620155 read(fd: 106, buf: 0x7fcb9f939cc0, count: 512) = 8 9837.209 ( 0.035 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 32, iocbpp: 0x7fca7d0cebe0) = 32 174k IOPS (+3.6%), IOThread syscalls: 9809.566 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0cdd62be0) = 32 9809.625 ( 0.001 ms): IO iothread1/623061 write(fd: 103, buf: 0x5647cfba5f58, count: 8) = 8 9809.627 ( 0.002 ms): IO iothread1/623061 write(fd: 104, buf: 0x5647cfba5f58, count: 8) = 8 9809.663 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0d0388b50) = 32 Notice that ppoll(2) and eventfd read(2) syscalls are eliminated because the IOThread stays in polling mode instead of falling back to file descriptor monitoring. As usual, polling is not implemented on Windows so this patch ignores the new io_poll_read() callback in aio-win32.c. Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Stefano Garzarella <sgarzare@redhat.com> Message-id: 20211207132336.36627-2-stefanha@redhat.com [Fixed up aio_set_event_notifier() calls in tests/unit/test-fdmon-epoll.c added after this series was queued. --Stefan] Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2021-12-07 14:23:31 +01:00
aio_set_event_notifier(ctx, &data.e, true, event_ready_cb, NULL, NULL);
event_notifier_set(&data.e);
for (j = 0; j < i; j++) {
aio_disable_external(ctx);
}
for (j = 0; j < i; j++) {
assert(!aio_poll(ctx, false));
assert(event_notifier_test_and_clear(&data.e));
event_notifier_set(&data.e);
aio_enable_external(ctx);
}
assert(aio_poll(ctx, false));
set_event_notifier(ctx, &data.e, NULL);
event_notifier_cleanup(&data.e);
}
}
static void test_wait_event_notifier_noflush(void)
{
EventNotifierTestData data = { .n = 0 };
EventNotifierTestData dummy = { .n = 0, .active = 1 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
/* Until there is an active descriptor, aio_poll may or may not call
* event_ready_cb. Still, it must not block. */
event_notifier_set(&data.e);
aio: stop using .io_flush() Now that aio_poll() users check their termination condition themselves, it is no longer necessary to call .io_flush() handlers. The behavior of aio_poll() changes as follows: 1. .io_flush() is no longer invoked and file descriptors are *always* monitored. Previously returning 0 from .io_flush() would skip this file descriptor. Due to this change it is essential to check that requests are pending before calling qemu_aio_wait(). Failure to do so means we block, for example, waiting for an idle iSCSI socket to become readable when there are no requests. Currently all qemu_aio_wait()/aio_poll() callers check before calling. 2. aio_poll() now returns true if progress was made (BH or fd handlers executed) and false otherwise. Previously it would return true whenever 'busy', which means that .io_flush() returned true. The 'busy' concept no longer exists so just progress is returned. Due to this change we need to update tests/test-aio.c which asserts aio_poll() return values. Note that QEMU doesn't actually rely on these return values so only tests/test-aio.c cares. Note that ctx->notifier, the EventNotifier fd used for aio_notify(), is now handled as a special case. This is a little ugly but maintains aio_poll() semantics, i.e. aio_notify() does not count as 'progress' and aio_poll() avoids blocking when the user has not set any fd handlers yet. Patches after this remove .io_flush() handler code until we can finally drop the io_flush arguments to aio_set_fd_handler() and friends. Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-04-11 16:56:50 +02:00
g_assert(aio_poll(ctx, true));
data.n = 0;
/* An active event notifier forces aio_poll to look at EventNotifiers. */
event_notifier_init(&dummy.e, false);
set_event_notifier(ctx, &dummy.e, event_ready_cb);
event_notifier_set(&data.e);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
aio: stop using .io_flush() Now that aio_poll() users check their termination condition themselves, it is no longer necessary to call .io_flush() handlers. The behavior of aio_poll() changes as follows: 1. .io_flush() is no longer invoked and file descriptors are *always* monitored. Previously returning 0 from .io_flush() would skip this file descriptor. Due to this change it is essential to check that requests are pending before calling qemu_aio_wait(). Failure to do so means we block, for example, waiting for an idle iSCSI socket to become readable when there are no requests. Currently all qemu_aio_wait()/aio_poll() callers check before calling. 2. aio_poll() now returns true if progress was made (BH or fd handlers executed) and false otherwise. Previously it would return true whenever 'busy', which means that .io_flush() returned true. The 'busy' concept no longer exists so just progress is returned. Due to this change we need to update tests/test-aio.c which asserts aio_poll() return values. Note that QEMU doesn't actually rely on these return values so only tests/test-aio.c cares. Note that ctx->notifier, the EventNotifier fd used for aio_notify(), is now handled as a special case. This is a little ugly but maintains aio_poll() semantics, i.e. aio_notify() does not count as 'progress' and aio_poll() avoids blocking when the user has not set any fd handlers yet. Patches after this remove .io_flush() handler code until we can finally drop the io_flush arguments to aio_set_fd_handler() and friends. Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-04-11 16:56:50 +02:00
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
event_notifier_set(&data.e);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 2);
aio: stop using .io_flush() Now that aio_poll() users check their termination condition themselves, it is no longer necessary to call .io_flush() handlers. The behavior of aio_poll() changes as follows: 1. .io_flush() is no longer invoked and file descriptors are *always* monitored. Previously returning 0 from .io_flush() would skip this file descriptor. Due to this change it is essential to check that requests are pending before calling qemu_aio_wait(). Failure to do so means we block, for example, waiting for an idle iSCSI socket to become readable when there are no requests. Currently all qemu_aio_wait()/aio_poll() callers check before calling. 2. aio_poll() now returns true if progress was made (BH or fd handlers executed) and false otherwise. Previously it would return true whenever 'busy', which means that .io_flush() returned true. The 'busy' concept no longer exists so just progress is returned. Due to this change we need to update tests/test-aio.c which asserts aio_poll() return values. Note that QEMU doesn't actually rely on these return values so only tests/test-aio.c cares. Note that ctx->notifier, the EventNotifier fd used for aio_notify(), is now handled as a special case. This is a little ugly but maintains aio_poll() semantics, i.e. aio_notify() does not count as 'progress' and aio_poll() avoids blocking when the user has not set any fd handlers yet. Patches after this remove .io_flush() handler code until we can finally drop the io_flush arguments to aio_set_fd_handler() and friends. Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-04-11 16:56:50 +02:00
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 2);
event_notifier_set(&dummy.e);
wait_until_inactive(&dummy);
g_assert_cmpint(data.n, ==, 2);
g_assert_cmpint(dummy.n, ==, 1);
g_assert_cmpint(dummy.active, ==, 0);
set_event_notifier(ctx, &dummy.e, NULL);
event_notifier_cleanup(&dummy.e);
set_event_notifier(ctx, &data.e, NULL);
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 2);
event_notifier_cleanup(&data.e);
}
static void test_timer_schedule(void)
{
TimerTestData data = { .n = 0, .ctx = ctx, .ns = SCALE_MS * 750LL,
.max = 2,
.clock_type = QEMU_CLOCK_REALTIME };
EventNotifier e;
/* aio_poll will not block to wait for timers to complete unless it has
* an fd to wait on. Fixing this breaks other tests. So create a dummy one.
*/
event_notifier_init(&e, false);
set_event_notifier(ctx, &e, dummy_io_handler_read);
aio_poll(ctx, false);
aio_timer_init(ctx, &data.timer, data.clock_type,
SCALE_NS, timer_test_cb, &data);
timer_mod(&data.timer,
qemu_clock_get_ns(data.clock_type) +
data.ns);
g_assert_cmpint(data.n, ==, 0);
/* timer_mod may well cause an event notifer to have gone off,
* so clear that
*/
do {} while (aio_poll(ctx, false));
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 0);
g_usleep(1 * G_USEC_PER_SEC);
g_assert_cmpint(data.n, ==, 0);
g_assert(aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
/* timer_mod called by our callback */
do {} while (aio_poll(ctx, false));
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 1);
g_assert(aio_poll(ctx, true));
g_assert_cmpint(data.n, ==, 2);
/* As max is now 2, an event notifier should not have gone off */
g_assert(!aio_poll(ctx, false));
g_assert_cmpint(data.n, ==, 2);
set_event_notifier(ctx, &e, NULL);
event_notifier_cleanup(&e);
timer_del(&data.timer);
}
/* Now the same tests, using the context as a GSource. They are
* very similar to the ones above, with g_main_context_iteration
* replacing aio_poll. However:
* - sometimes both the AioContext and the glib main loop wake
* themselves up. Hence, some "g_assert(!aio_poll(ctx, false));"
* are replaced by "while (g_main_context_iteration(NULL, false));".
* - there is no exact replacement for a blocking wait.
* "while (g_main_context_iteration(NULL, true)" seems to work,
* but it is not documented _why_ it works. For these tests a
* non-blocking loop like "while (g_main_context_iteration(NULL, false)"
* works well, and that's what I am using.
*/
static void test_source_flush(void)
{
g_assert(!g_main_context_iteration(NULL, false));
aio_notify(ctx);
while (g_main_context_iteration(NULL, false));
g_assert(!g_main_context_iteration(NULL, false));
}
static void test_source_bh_schedule(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(g_main_context_iteration(NULL, true));
g_assert_cmpint(data.n, ==, 1);
g_assert(!g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
qemu_bh_delete(data.bh);
}
static void test_source_bh_schedule10(void)
{
BHTestData data = { .n = 0, .max = 10 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
g_assert(g_main_context_iteration(NULL, true));
g_assert_cmpint(data.n, ==, 2);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 10);
g_assert(!g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 10);
qemu_bh_delete(data.bh);
}
static void test_source_bh_cancel(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
qemu_bh_cancel(data.bh);
g_assert_cmpint(data.n, ==, 0);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
qemu_bh_delete(data.bh);
}
static void test_source_bh_delete(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
qemu_bh_delete(data.bh);
g_assert_cmpint(data.n, ==, 0);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
}
static void test_source_bh_delete_from_cb(void)
{
BHTestData data1 = { .n = 0, .max = 1 };
data1.bh = aio_bh_new(ctx, bh_delete_cb, &data1);
qemu_bh_schedule(data1.bh);
g_assert_cmpint(data1.n, ==, 0);
g_main_context_iteration(NULL, true);
g_assert_cmpint(data1.n, ==, data1.max);
g_assert(data1.bh == NULL);
assert(g_main_context_iteration(NULL, false));
assert(!g_main_context_iteration(NULL, false));
}
static void test_source_bh_delete_from_cb_many(void)
{
BHTestData data1 = { .n = 0, .max = 1 };
BHTestData data2 = { .n = 0, .max = 3 };
BHTestData data3 = { .n = 0, .max = 2 };
BHTestData data4 = { .n = 0, .max = 4 };
data1.bh = aio_bh_new(ctx, bh_delete_cb, &data1);
data2.bh = aio_bh_new(ctx, bh_delete_cb, &data2);
data3.bh = aio_bh_new(ctx, bh_delete_cb, &data3);
data4.bh = aio_bh_new(ctx, bh_delete_cb, &data4);
qemu_bh_schedule(data1.bh);
qemu_bh_schedule(data2.bh);
qemu_bh_schedule(data3.bh);
qemu_bh_schedule(data4.bh);
g_assert_cmpint(data1.n, ==, 0);
g_assert_cmpint(data2.n, ==, 0);
g_assert_cmpint(data3.n, ==, 0);
g_assert_cmpint(data4.n, ==, 0);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data1.n, ==, 1);
g_assert_cmpint(data2.n, ==, 1);
g_assert_cmpint(data3.n, ==, 1);
g_assert_cmpint(data4.n, ==, 1);
g_assert(data1.bh == NULL);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data1.n, ==, data1.max);
g_assert_cmpint(data2.n, ==, data2.max);
g_assert_cmpint(data3.n, ==, data3.max);
g_assert_cmpint(data4.n, ==, data4.max);
g_assert(data1.bh == NULL);
g_assert(data2.bh == NULL);
g_assert(data3.bh == NULL);
g_assert(data4.bh == NULL);
}
static void test_source_bh_flush(void)
{
BHTestData data = { .n = 0 };
data.bh = aio_bh_new(ctx, bh_test_cb, &data);
qemu_bh_schedule(data.bh);
g_assert_cmpint(data.n, ==, 0);
g_assert(g_main_context_iteration(NULL, true));
g_assert_cmpint(data.n, ==, 1);
g_assert(!g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
qemu_bh_delete(data.bh);
}
static void test_source_set_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 0 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
set_event_notifier(ctx, &data.e, NULL);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
event_notifier_cleanup(&data.e);
}
static void test_source_wait_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 1 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
g_assert_cmpint(data.active, ==, 1);
event_notifier_set(&data.e);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 0);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 0);
set_event_notifier(ctx, &data.e, NULL);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
event_notifier_cleanup(&data.e);
}
static void test_source_flush_event_notifier(void)
{
EventNotifierTestData data = { .n = 0, .active = 10, .auto_set = true };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
g_assert_cmpint(data.active, ==, 10);
event_notifier_set(&data.e);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
g_assert_cmpint(data.active, ==, 9);
g_assert(g_main_context_iteration(NULL, false));
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 10);
g_assert_cmpint(data.active, ==, 0);
g_assert(!g_main_context_iteration(NULL, false));
set_event_notifier(ctx, &data.e, NULL);
while (g_main_context_iteration(NULL, false));
event_notifier_cleanup(&data.e);
}
static void test_source_wait_event_notifier_noflush(void)
{
EventNotifierTestData data = { .n = 0 };
EventNotifierTestData dummy = { .n = 0, .active = 1 };
event_notifier_init(&data.e, false);
set_event_notifier(ctx, &data.e, event_ready_cb);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 0);
/* Until there is an active descriptor, glib may or may not call
* event_ready_cb. Still, it must not block. */
event_notifier_set(&data.e);
g_main_context_iteration(NULL, true);
data.n = 0;
/* An active event notifier forces aio_poll to look at EventNotifiers. */
event_notifier_init(&dummy.e, false);
set_event_notifier(ctx, &dummy.e, event_ready_cb);
event_notifier_set(&data.e);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
g_assert(!g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 1);
event_notifier_set(&data.e);
g_assert(g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 2);
g_assert(!g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 2);
event_notifier_set(&dummy.e);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 2);
g_assert_cmpint(dummy.n, ==, 1);
g_assert_cmpint(dummy.active, ==, 0);
set_event_notifier(ctx, &dummy.e, NULL);
event_notifier_cleanup(&dummy.e);
set_event_notifier(ctx, &data.e, NULL);
while (g_main_context_iteration(NULL, false));
g_assert_cmpint(data.n, ==, 2);
event_notifier_cleanup(&data.e);
}
static void test_source_timer_schedule(void)
{
TimerTestData data = { .n = 0, .ctx = ctx, .ns = SCALE_MS * 750LL,
.max = 2,
.clock_type = QEMU_CLOCK_REALTIME };
EventNotifier e;
int64_t expiry;
/* aio_poll will not block to wait for timers to complete unless it has
* an fd to wait on. Fixing this breaks other tests. So create a dummy one.
*/
event_notifier_init(&e, false);
set_event_notifier(ctx, &e, dummy_io_handler_read);
do {} while (g_main_context_iteration(NULL, false));
aio_timer_init(ctx, &data.timer, data.clock_type,
SCALE_NS, timer_test_cb, &data);
expiry = qemu_clock_get_ns(data.clock_type) +
data.ns;
timer_mod(&data.timer, expiry);
g_assert_cmpint(data.n, ==, 0);
g_usleep(1 * G_USEC_PER_SEC);
g_assert_cmpint(data.n, ==, 0);
g_assert(g_main_context_iteration(NULL, true));
g_assert_cmpint(data.n, ==, 1);
expiry += data.ns;
while (data.n < 2) {
g_main_context_iteration(NULL, true);
}
g_assert_cmpint(data.n, ==, 2);
g_assert(qemu_clock_get_ns(data.clock_type) > expiry);
set_event_notifier(ctx, &e, NULL);
event_notifier_cleanup(&e);
timer_del(&data.timer);
}
/*
* Check that aio_co_enter() can chain many times
*
* Two coroutines should be able to invoke each other via aio_co_enter() many
* times without hitting a limit like stack exhaustion. In other words, the
* calls should be chained instead of nested.
*/
typedef struct {
Coroutine *other;
unsigned i;
unsigned max;
} ChainData;
static void coroutine_fn chain(void *opaque)
{
ChainData *data = opaque;
for (data->i = 0; data->i < data->max; data->i++) {
/* Queue up the other coroutine... */
aio_co_enter(ctx, data->other);
/* ...and give control to it */
qemu_coroutine_yield();
}
}
static void test_queue_chaining(void)
{
/* This number of iterations hit stack exhaustion in the past: */
ChainData data_a = { .max = 25000 };
ChainData data_b = { .max = 25000 };
data_b.other = qemu_coroutine_create(chain, &data_a);
data_a.other = qemu_coroutine_create(chain, &data_b);
qemu_coroutine_enter(data_b.other);
g_assert_cmpint(data_a.i, ==, data_a.max);
g_assert_cmpint(data_b.i, ==, data_b.max - 1);
/* Allow the second coroutine to terminate */
qemu_coroutine_enter(data_a.other);
g_assert_cmpint(data_b.i, ==, data_b.max);
}
static void co_check_current_thread(void *opaque)
{
QemuThread *main_thread = opaque;
assert(qemu_thread_is_self(main_thread));
}
static void *test_aio_co_enter(void *co)
{
/*
* qemu_get_current_aio_context() should not to be the main thread
* AioContext, because this is a worker thread that has not taken
* the BQL. So aio_co_enter will schedule the coroutine in the
* main thread AioContext.
*/
aio_co_enter(qemu_get_aio_context(), co);
return NULL;
}
static void test_worker_thread_co_enter(void)
{
QemuThread this_thread, worker_thread;
Coroutine *co;
qemu_thread_get_self(&this_thread);
co = qemu_coroutine_create(co_check_current_thread, &this_thread);
qemu_thread_create(&worker_thread, "test_acquire_thread",
test_aio_co_enter,
co, QEMU_THREAD_JOINABLE);
/* Test aio_co_enter from a worker thread. */
qemu_thread_join(&worker_thread);
g_assert(aio_poll(ctx, true));
g_assert(!aio_poll(ctx, false));
}
/* End of tests. */
int main(int argc, char **argv)
{
qemu_init_main_loop(&error_fatal);
ctx = qemu_get_aio_context();
while (g_main_context_iteration(NULL, false));
g_test_init(&argc, &argv, NULL);
g_test_add_func("/aio/acquire", test_acquire);
g_test_add_func("/aio/bh/schedule", test_bh_schedule);
g_test_add_func("/aio/bh/schedule10", test_bh_schedule10);
g_test_add_func("/aio/bh/cancel", test_bh_cancel);
g_test_add_func("/aio/bh/delete", test_bh_delete);
g_test_add_func("/aio/bh/callback-delete/one", test_bh_delete_from_cb);
g_test_add_func("/aio/bh/callback-delete/many", test_bh_delete_from_cb_many);
g_test_add_func("/aio/bh/flush", test_bh_flush);
g_test_add_func("/aio/event/add-remove", test_set_event_notifier);
g_test_add_func("/aio/event/wait", test_wait_event_notifier);
g_test_add_func("/aio/event/wait/no-flush-cb", test_wait_event_notifier_noflush);
g_test_add_func("/aio/event/flush", test_flush_event_notifier);
g_test_add_func("/aio/external-client", test_aio_external_client);
g_test_add_func("/aio/timer/schedule", test_timer_schedule);
g_test_add_func("/aio/coroutine/queue-chaining", test_queue_chaining);
g_test_add_func("/aio/coroutine/worker-thread-co-enter", test_worker_thread_co_enter);
g_test_add_func("/aio-gsource/flush", test_source_flush);
g_test_add_func("/aio-gsource/bh/schedule", test_source_bh_schedule);
g_test_add_func("/aio-gsource/bh/schedule10", test_source_bh_schedule10);
g_test_add_func("/aio-gsource/bh/cancel", test_source_bh_cancel);
g_test_add_func("/aio-gsource/bh/delete", test_source_bh_delete);
g_test_add_func("/aio-gsource/bh/callback-delete/one", test_source_bh_delete_from_cb);
g_test_add_func("/aio-gsource/bh/callback-delete/many", test_source_bh_delete_from_cb_many);
g_test_add_func("/aio-gsource/bh/flush", test_source_bh_flush);
g_test_add_func("/aio-gsource/event/add-remove", test_source_set_event_notifier);
g_test_add_func("/aio-gsource/event/wait", test_source_wait_event_notifier);
g_test_add_func("/aio-gsource/event/wait/no-flush-cb", test_source_wait_event_notifier_noflush);
g_test_add_func("/aio-gsource/event/flush", test_source_flush_event_notifier);
g_test_add_func("/aio-gsource/timer/schedule", test_source_timer_schedule);
return g_test_run();
}