* [PATCH 00/12] aio: remove AioContext lock
@ 2023-11-29 19:55 Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
` (11 more replies)
0 siblings, 12 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
This series removes the AioContext locking APIs from QEMU.
aio_context_acquire() and aio_context_release() are currently only needed to
support the locking discipline required by AIO_POLL_WHILE() (except for a stray
user that I converted in Patch 1). AIO_POLL_WHILE() doesn't really need the
AioContext lock anymore, so it's possible to remove the API. This is a nice
simplification because the AioContext locking rules were sometimes tricky or
underspecified, leading to many bugs of the years.
This patch series removes these APIs across the codebase and cleans up the
documentation/comments that refers to them.
Patch 1 is a AioContext lock user I forgot to convert in my earlier SCSI
conversion series.
Patch 2 removes tests for the AioContext lock because they will no longer be
needed when the lock is gone.
Patches 3-9 remove the AioContext lock. These can be reviewed by categorizing
the call sites into 1. places that take the lock because they call an API that
requires the lock (ultimately AIO_POLL_WHILE()) and 2. places that take the
lock to protect state. There should be no instances of case 2 left. If you see
one, you've found a bug in this patch series!
Patches 10-12 remove comments.
Based-on: 20231123194931.171598-1-stefanha@redhat.com ("[PATCH 0/4] scsi: eliminate AioContext lock")
Since SCSI needs to stop relying on the AioContext lock before we can remove
the lock.
Stefan Hajnoczi (12):
virtio-scsi: replace AioContext lock with tmf_bh_lock
tests: remove aio_context_acquire() tests
aio: make aio_context_acquire()/aio_context_release() a no-op
graph-lock: remove AioContext locking
block: remove AioContext locking
scsi: remove AioContext locking
aio-wait: draw equivalence between AIO_WAIT_WHILE() and
AIO_WAIT_WHILE_UNLOCKED()
aio: remove aio_context_acquire()/aio_context_release() API
docs: remove AioContext lock from IOThread docs
scsi: remove outdated AioContext lock comment
job: remove outdated AioContext locking comments
block: remove outdated AioContext locking comments
docs/devel/multiple-iothreads.txt | 45 ++--
include/block/aio-wait.h | 16 +-
include/block/aio.h | 17 --
include/block/block-common.h | 3 -
include/block/block-global-state.h | 9 +-
include/block/block-io.h | 12 +-
include/block/block_int-common.h | 2 -
include/block/graph-lock.h | 21 +-
include/block/snapshot.h | 2 -
include/hw/virtio/virtio-scsi.h | 17 +-
include/qemu/job.h | 20 --
block.c | 357 ++++-----------------------
block/backup.c | 4 +-
block/blklogwrites.c | 8 +-
block/blkverify.c | 4 +-
block/block-backend.c | 33 +--
block/commit.c | 16 +-
block/copy-before-write.c | 22 +-
block/export/export.c | 22 +-
block/export/vhost-user-blk-server.c | 4 -
block/graph-lock.c | 44 +---
block/io.c | 45 +---
block/mirror.c | 41 +--
block/monitor/bitmap-qmp-cmds.c | 20 +-
block/monitor/block-hmp-cmds.c | 29 ---
block/qapi-sysemu.c | 27 +-
block/qapi.c | 18 +-
block/qcow2.c | 4 +-
block/quorum.c | 8 +-
block/raw-format.c | 5 -
block/replication.c | 72 +-----
block/snapshot.c | 26 +-
block/stream.c | 12 +-
block/vmdk.c | 20 +-
block/write-threshold.c | 6 -
blockdev.c | 315 +++++------------------
blockjob.c | 30 +--
hw/block/dataplane/virtio-blk.c | 10 -
hw/block/dataplane/xen-block.c | 17 +-
hw/block/virtio-blk.c | 45 +---
hw/core/qdev-properties-system.c | 9 -
hw/scsi/scsi-bus.c | 2 -
hw/scsi/scsi-disk.c | 29 +--
hw/scsi/virtio-scsi.c | 80 +++---
job.c | 16 --
migration/block.c | 33 +--
migration/migration-hmp-cmds.c | 3 -
migration/savevm.c | 22 --
net/colo-compare.c | 2 -
qemu-img.c | 4 -
qemu-io.c | 10 +-
qemu-nbd.c | 2 -
replay/replay-debugging.c | 4 -
tests/unit/test-aio.c | 67 +----
tests/unit/test-bdrv-drain.c | 91 ++-----
tests/unit/test-bdrv-graph-mod.c | 26 +-
tests/unit/test-block-iothread.c | 31 ---
tests/unit/test-blockjob.c | 137 ----------
tests/unit/test-replication.c | 11 -
util/async.c | 14 --
util/vhost-user-server.c | 3 -
scripts/block-coroutine-wrapper.py | 7 +-
tests/qemu-iotests/202 | 2 +-
tests/qemu-iotests/203 | 3 +-
tests/tsan/suppressions.tsan | 1 -
65 files changed, 324 insertions(+), 1713 deletions(-)
--
2.42.0
^ permalink raw reply [flat|nested] 38+ messages in thread
* [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 15:25 ` Eric Blake
` (2 more replies)
2023-11-29 19:55 ` [PATCH 02/12] tests: remove aio_context_acquire() tests Stefan Hajnoczi
` (10 subsequent siblings)
11 siblings, 3 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Protect the Task Management Function BH state with a lock. The TMF BH
runs in the main loop thread. An IOThread might process a TMF at the
same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
must be protected by a lock.
Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
This avoids more locking to protect the virtqueue and SCSI layer state.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/hw/virtio/virtio-scsi.h | 3 +-
hw/scsi/virtio-scsi.c | 62 ++++++++++++++++++++++-----------
2 files changed, 43 insertions(+), 22 deletions(-)
diff --git a/include/hw/virtio/virtio-scsi.h b/include/hw/virtio/virtio-scsi.h
index 779568ab5d..da8cb928d9 100644
--- a/include/hw/virtio/virtio-scsi.h
+++ b/include/hw/virtio/virtio-scsi.h
@@ -85,8 +85,9 @@ struct VirtIOSCSI {
/*
* TMFs deferred to main loop BH. These fields are protected by
- * virtio_scsi_acquire().
+ * tmf_bh_lock.
*/
+ QemuMutex tmf_bh_lock;
QEMUBH *tmf_bh;
QTAILQ_HEAD(, VirtIOSCSIReq) tmf_bh_list;
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index 9c751bf296..4f8d35facc 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -123,6 +123,30 @@ static void virtio_scsi_complete_req(VirtIOSCSIReq *req)
virtio_scsi_free_req(req);
}
+static void virtio_scsi_complete_req_bh(void *opaque)
+{
+ VirtIOSCSIReq *req = opaque;
+
+ virtio_scsi_complete_req(req);
+}
+
+/*
+ * Called from virtio_scsi_do_one_tmf_bh() in main loop thread. The main loop
+ * thread cannot touch the virtqueue since that could race with an IOThread.
+ */
+static void virtio_scsi_complete_req_from_main_loop(VirtIOSCSIReq *req)
+{
+ VirtIOSCSI *s = req->dev;
+
+ if (!s->ctx || s->ctx == qemu_get_aio_context()) {
+ /* No need to schedule a BH when there is no IOThread */
+ virtio_scsi_complete_req(req);
+ } else {
+ /* Run request completion in the IOThread */
+ aio_wait_bh_oneshot(s->ctx, virtio_scsi_complete_req_bh, req);
+ }
+}
+
static void virtio_scsi_bad_req(VirtIOSCSIReq *req)
{
virtio_error(VIRTIO_DEVICE(req->dev), "wrong size for virtio-scsi headers");
@@ -338,10 +362,7 @@ static void virtio_scsi_do_one_tmf_bh(VirtIOSCSIReq *req)
out:
object_unref(OBJECT(d));
-
- virtio_scsi_acquire(s);
- virtio_scsi_complete_req(req);
- virtio_scsi_release(s);
+ virtio_scsi_complete_req_from_main_loop(req);
}
/* Some TMFs must be processed from the main loop thread */
@@ -354,18 +375,16 @@ static void virtio_scsi_do_tmf_bh(void *opaque)
GLOBAL_STATE_CODE();
- virtio_scsi_acquire(s);
+ WITH_QEMU_LOCK_GUARD(&s->tmf_bh_lock) {
+ QTAILQ_FOREACH_SAFE(req, &s->tmf_bh_list, next, tmp) {
+ QTAILQ_REMOVE(&s->tmf_bh_list, req, next);
+ QTAILQ_INSERT_TAIL(&reqs, req, next);
+ }
- QTAILQ_FOREACH_SAFE(req, &s->tmf_bh_list, next, tmp) {
- QTAILQ_REMOVE(&s->tmf_bh_list, req, next);
- QTAILQ_INSERT_TAIL(&reqs, req, next);
+ qemu_bh_delete(s->tmf_bh);
+ s->tmf_bh = NULL;
}
- qemu_bh_delete(s->tmf_bh);
- s->tmf_bh = NULL;
-
- virtio_scsi_release(s);
-
QTAILQ_FOREACH_SAFE(req, &reqs, next, tmp) {
QTAILQ_REMOVE(&reqs, req, next);
virtio_scsi_do_one_tmf_bh(req);
@@ -379,8 +398,7 @@ static void virtio_scsi_reset_tmf_bh(VirtIOSCSI *s)
GLOBAL_STATE_CODE();
- virtio_scsi_acquire(s);
-
+ /* Called after ioeventfd has been stopped, so tmf_bh_lock is not needed */
if (s->tmf_bh) {
qemu_bh_delete(s->tmf_bh);
s->tmf_bh = NULL;
@@ -393,19 +411,19 @@ static void virtio_scsi_reset_tmf_bh(VirtIOSCSI *s)
req->resp.tmf.response = VIRTIO_SCSI_S_TARGET_FAILURE;
virtio_scsi_complete_req(req);
}
-
- virtio_scsi_release(s);
}
static void virtio_scsi_defer_tmf_to_bh(VirtIOSCSIReq *req)
{
VirtIOSCSI *s = req->dev;
- QTAILQ_INSERT_TAIL(&s->tmf_bh_list, req, next);
+ WITH_QEMU_LOCK_GUARD(&s->tmf_bh_lock) {
+ QTAILQ_INSERT_TAIL(&s->tmf_bh_list, req, next);
- if (!s->tmf_bh) {
- s->tmf_bh = qemu_bh_new(virtio_scsi_do_tmf_bh, s);
- qemu_bh_schedule(s->tmf_bh);
+ if (!s->tmf_bh) {
+ s->tmf_bh = qemu_bh_new(virtio_scsi_do_tmf_bh, s);
+ qemu_bh_schedule(s->tmf_bh);
+ }
}
}
@@ -1235,6 +1253,7 @@ static void virtio_scsi_device_realize(DeviceState *dev, Error **errp)
Error *err = NULL;
QTAILQ_INIT(&s->tmf_bh_list);
+ qemu_mutex_init(&s->tmf_bh_lock);
virtio_scsi_common_realize(dev,
virtio_scsi_handle_ctrl,
@@ -1277,6 +1296,7 @@ static void virtio_scsi_device_unrealize(DeviceState *dev)
qbus_set_hotplug_handler(BUS(&s->bus), NULL);
virtio_scsi_common_unrealize(dev);
+ qemu_mutex_destroy(&s->tmf_bh_lock);
}
static Property virtio_scsi_properties[] = {
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 02/12] tests: remove aio_context_acquire() tests
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 15:29 ` Eric Blake
2023-12-04 12:54 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op Stefan Hajnoczi
` (9 subsequent siblings)
11 siblings, 2 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
The aio_context_acquire() API is being removed. Drop the test case that
calls the API.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
tests/unit/test-aio.c | 67 +------------------------------------------
1 file changed, 1 insertion(+), 66 deletions(-)
diff --git a/tests/unit/test-aio.c b/tests/unit/test-aio.c
index 337b6e4ea7..e77d86be87 100644
--- a/tests/unit/test-aio.c
+++ b/tests/unit/test-aio.c
@@ -100,76 +100,12 @@ static void event_ready_cb(EventNotifier *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 *nctx, EventNotifier *notifier,
EventNotifierHandler *handler)
{
aio_set_event_notifier(nctx, notifier, 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 };
@@ -879,7 +815,7 @@ static void test_worker_thread_co_enter(void)
qemu_thread_get_self(&this_thread);
co = qemu_coroutine_create(co_check_current_thread, &this_thread);
- qemu_thread_create(&worker_thread, "test_acquire_thread",
+ qemu_thread_create(&worker_thread, "test_aio_co_enter",
test_aio_co_enter,
co, QEMU_THREAD_JOINABLE);
@@ -899,7 +835,6 @@ int main(int argc, char **argv)
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);
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 02/12] tests: remove aio_context_acquire() tests Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 18:10 ` Eric Blake
2023-12-04 13:18 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 04/12] graph-lock: remove AioContext locking Stefan Hajnoczi
` (8 subsequent siblings)
11 siblings, 2 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
aio_context_acquire()/aio_context_release() has been replaced by
fine-grained locking to protect state shared by multiple threads. The
AioContext lock still plays the role of balancing locking in
AIO_WAIT_WHILE() and many functions in QEMU either require that the
AioContext lock is held or not held for this reason. In other words, the
AioContext lock is purely there for consistency with itself and serves
no real purpose anymore.
Stop actually acquiring/releasing the lock in
aio_context_acquire()/aio_context_release() so that subsequent patches
can remove callers across the codebase incrementally.
I have performed "make check" and qemu-iotests stress tests across
x86-64, ppc64le, and aarch64 to confirm that there are no failures as a
result of eliminating the lock.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
util/async.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/util/async.c b/util/async.c
index 8f90ddc304..04ee83d220 100644
--- a/util/async.c
+++ b/util/async.c
@@ -725,12 +725,12 @@ void aio_context_unref(AioContext *ctx)
void aio_context_acquire(AioContext *ctx)
{
- qemu_rec_mutex_lock(&ctx->lock);
+ /* TODO remove this function */
}
void aio_context_release(AioContext *ctx)
{
- qemu_rec_mutex_unlock(&ctx->lock);
+ /* TODO remove this function */
}
QEMU_DEFINE_STATIC_CO_TLS(AioContext *, my_aiocontext)
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 04/12] graph-lock: remove AioContext locking
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (2 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 19:33 ` Eric Blake
2023-12-04 13:21 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
` (7 subsequent siblings)
11 siblings, 2 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Stop acquiring/releasing the AioContext lock in
bdrv_graph_wrlock()/bdrv_graph_unlock() since the lock no longer has any
effect.
The distinction between bdrv_graph_wrunlock() and
bdrv_graph_wrunlock_ctx() becomes meaningless and they can be collapsed
into one function.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/block/graph-lock.h | 21 ++-----------
block.c | 50 +++++++++++++++---------------
block/backup.c | 4 +--
block/blklogwrites.c | 8 ++---
block/blkverify.c | 4 +--
block/block-backend.c | 11 +++----
block/commit.c | 16 +++++-----
block/graph-lock.c | 44 ++------------------------
block/mirror.c | 22 ++++++-------
block/qcow2.c | 4 +--
block/quorum.c | 8 ++---
block/replication.c | 14 ++++-----
block/snapshot.c | 4 +--
block/stream.c | 12 +++----
block/vmdk.c | 20 ++++++------
blockdev.c | 8 ++---
blockjob.c | 12 +++----
tests/unit/test-bdrv-drain.c | 40 ++++++++++++------------
tests/unit/test-bdrv-graph-mod.c | 20 ++++++------
scripts/block-coroutine-wrapper.py | 4 +--
20 files changed, 133 insertions(+), 193 deletions(-)
diff --git a/include/block/graph-lock.h b/include/block/graph-lock.h
index 22b5db1ed9..d7545e82d0 100644
--- a/include/block/graph-lock.h
+++ b/include/block/graph-lock.h
@@ -110,34 +110,17 @@ void unregister_aiocontext(AioContext *ctx);
*
* The wrlock can only be taken from the main loop, with BQL held, as only the
* main loop is allowed to modify the graph.
- *
- * If @bs is non-NULL, its AioContext is temporarily released.
- *
- * This function polls. Callers must not hold the lock of any AioContext other
- * than the current one and the one of @bs.
*/
void no_coroutine_fn TSA_ACQUIRE(graph_lock) TSA_NO_TSA
-bdrv_graph_wrlock(BlockDriverState *bs);
+bdrv_graph_wrlock(void);
/*
* bdrv_graph_wrunlock:
* Write finished, reset global has_writer to 0 and restart
* all readers that are waiting.
- *
- * If @bs is non-NULL, its AioContext is temporarily released.
*/
void no_coroutine_fn TSA_RELEASE(graph_lock) TSA_NO_TSA
-bdrv_graph_wrunlock(BlockDriverState *bs);
-
-/*
- * bdrv_graph_wrunlock_ctx:
- * Write finished, reset global has_writer to 0 and restart
- * all readers that are waiting.
- *
- * If @ctx is non-NULL, its lock is temporarily released.
- */
-void no_coroutine_fn TSA_RELEASE(graph_lock) TSA_NO_TSA
-bdrv_graph_wrunlock_ctx(AioContext *ctx);
+bdrv_graph_wrunlock(void);
/*
* bdrv_graph_co_rdlock:
diff --git a/block.c b/block.c
index bfb0861ec6..25e1ebc606 100644
--- a/block.c
+++ b/block.c
@@ -1708,12 +1708,12 @@ bdrv_open_driver(BlockDriverState *bs, BlockDriver *drv, const char *node_name,
open_failed:
bs->drv = NULL;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
if (bs->file != NULL) {
bdrv_unref_child(bs, bs->file);
assert(!bs->file);
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
g_free(bs->opaque);
bs->opaque = NULL;
@@ -3575,9 +3575,9 @@ int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd,
bdrv_ref(drain_bs);
bdrv_drained_begin(drain_bs);
- bdrv_graph_wrlock(backing_hd);
+ bdrv_graph_wrlock();
ret = bdrv_set_backing_hd_drained(bs, backing_hd, errp);
- bdrv_graph_wrunlock(backing_hd);
+ bdrv_graph_wrunlock();
bdrv_drained_end(drain_bs);
bdrv_unref(drain_bs);
@@ -3790,13 +3790,13 @@ BdrvChild *bdrv_open_child(const char *filename,
return NULL;
}
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
ctx = bdrv_get_aio_context(bs);
aio_context_acquire(ctx);
child = bdrv_attach_child(parent, bs, bdref_key, child_class, child_role,
errp);
aio_context_release(ctx);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
return child;
}
@@ -4650,9 +4650,9 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
aio_context_release(ctx);
}
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
tran_commit(tran);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
QTAILQ_FOREACH_REVERSE(bs_entry, bs_queue, entry) {
BlockDriverState *bs = bs_entry->state.bs;
@@ -4669,9 +4669,9 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
goto cleanup;
abort:
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
tran_abort(tran);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) {
if (bs_entry->prepared) {
@@ -4852,12 +4852,12 @@ bdrv_reopen_parse_file_or_backing(BDRVReopenState *reopen_state,
}
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(new_child_bs);
+ bdrv_graph_wrlock();
ret = bdrv_set_file_or_backing_noperm(bs, new_child_bs, is_backing,
tran, errp);
- bdrv_graph_wrunlock_ctx(ctx);
+ bdrv_graph_wrunlock();
if (old_ctx != ctx) {
aio_context_release(ctx);
@@ -5209,14 +5209,14 @@ static void bdrv_close(BlockDriverState *bs)
bs->drv = NULL;
}
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
QLIST_FOREACH_SAFE(child, &bs->children, next, next) {
bdrv_unref_child(bs, child);
}
assert(!bs->backing);
assert(!bs->file);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
g_free(bs->opaque);
bs->opaque = NULL;
@@ -5509,9 +5509,9 @@ int bdrv_drop_filter(BlockDriverState *bs, Error **errp)
bdrv_graph_rdunlock_main_loop();
bdrv_drained_begin(child_bs);
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
ret = bdrv_replace_node_common(bs, child_bs, true, true, errp);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(child_bs);
return ret;
@@ -5561,7 +5561,7 @@ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top,
aio_context_acquire(old_context);
new_context = NULL;
- bdrv_graph_wrlock(bs_top);
+ bdrv_graph_wrlock();
child = bdrv_attach_child_noperm(bs_new, bs_top, "backing",
&child_of_bds, bdrv_backing_role(bs_new),
@@ -5593,7 +5593,7 @@ out:
tran_finalize(tran, ret);
bdrv_refresh_limits(bs_top, NULL, NULL);
- bdrv_graph_wrunlock(bs_top);
+ bdrv_graph_wrunlock();
bdrv_drained_end(bs_top);
bdrv_drained_end(bs_new);
@@ -5620,7 +5620,7 @@ int bdrv_replace_child_bs(BdrvChild *child, BlockDriverState *new_bs,
bdrv_ref(old_bs);
bdrv_drained_begin(old_bs);
bdrv_drained_begin(new_bs);
- bdrv_graph_wrlock(new_bs);
+ bdrv_graph_wrlock();
bdrv_replace_child_tran(child, new_bs, tran);
@@ -5631,7 +5631,7 @@ int bdrv_replace_child_bs(BdrvChild *child, BlockDriverState *new_bs,
tran_finalize(tran, ret);
- bdrv_graph_wrunlock(new_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(old_bs);
bdrv_drained_end(new_bs);
bdrv_unref(old_bs);
@@ -5718,9 +5718,9 @@ BlockDriverState *bdrv_insert_node(BlockDriverState *bs, QDict *options,
bdrv_ref(bs);
bdrv_drained_begin(bs);
bdrv_drained_begin(new_node_bs);
- bdrv_graph_wrlock(new_node_bs);
+ bdrv_graph_wrlock();
ret = bdrv_replace_node(bs, new_node_bs, errp);
- bdrv_graph_wrunlock(new_node_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(new_node_bs);
bdrv_drained_end(bs);
bdrv_unref(bs);
@@ -5975,7 +5975,7 @@ int bdrv_drop_intermediate(BlockDriverState *top, BlockDriverState *base,
bdrv_ref(top);
bdrv_drained_begin(base);
- bdrv_graph_wrlock(base);
+ bdrv_graph_wrlock();
if (!top->drv || !base->drv) {
goto exit_wrlock;
@@ -6015,7 +6015,7 @@ int bdrv_drop_intermediate(BlockDriverState *top, BlockDriverState *base,
* That's a FIXME.
*/
bdrv_replace_node_common(top, base, false, false, &local_err);
- bdrv_graph_wrunlock(base);
+ bdrv_graph_wrunlock();
if (local_err) {
error_report_err(local_err);
@@ -6052,7 +6052,7 @@ int bdrv_drop_intermediate(BlockDriverState *top, BlockDriverState *base,
goto exit;
exit_wrlock:
- bdrv_graph_wrunlock(base);
+ bdrv_graph_wrunlock();
exit:
bdrv_drained_end(base);
bdrv_unref(top);
diff --git a/block/backup.c b/block/backup.c
index 8aae5836d7..ec29d6b810 100644
--- a/block/backup.c
+++ b/block/backup.c
@@ -496,10 +496,10 @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
block_copy_set_speed(bcs, speed);
/* Required permissions are taken by copy-before-write filter target */
- bdrv_graph_wrlock(target);
+ bdrv_graph_wrlock();
block_job_add_bdrv(&job->common, "target", target, 0, BLK_PERM_ALL,
&error_abort);
- bdrv_graph_wrunlock(target);
+ bdrv_graph_wrunlock();
return &job->common;
diff --git a/block/blklogwrites.c b/block/blklogwrites.c
index 3678f6cf42..7207b2e757 100644
--- a/block/blklogwrites.c
+++ b/block/blklogwrites.c
@@ -251,9 +251,9 @@ static int blk_log_writes_open(BlockDriverState *bs, QDict *options, int flags,
ret = 0;
fail_log:
if (ret < 0) {
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, s->log_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
s->log_file = NULL;
}
fail:
@@ -265,10 +265,10 @@ static void blk_log_writes_close(BlockDriverState *bs)
{
BDRVBlkLogWritesState *s = bs->opaque;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, s->log_file);
s->log_file = NULL;
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
}
static int64_t coroutine_fn GRAPH_RDLOCK
diff --git a/block/blkverify.c b/block/blkverify.c
index 9b17c46644..ec45d8335e 100644
--- a/block/blkverify.c
+++ b/block/blkverify.c
@@ -151,10 +151,10 @@ static void blkverify_close(BlockDriverState *bs)
{
BDRVBlkverifyState *s = bs->opaque;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, s->test_file);
s->test_file = NULL;
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
}
static int64_t coroutine_fn GRAPH_RDLOCK
diff --git a/block/block-backend.c b/block/block-backend.c
index ec21148806..abac4e0235 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -889,7 +889,6 @@ void blk_remove_bs(BlockBackend *blk)
{
ThrottleGroupMember *tgm = &blk->public.throttle_group_member;
BdrvChild *root;
- AioContext *ctx;
GLOBAL_STATE_CODE();
@@ -919,10 +918,9 @@ void blk_remove_bs(BlockBackend *blk)
root = blk->root;
blk->root = NULL;
- ctx = bdrv_get_aio_context(root->bs);
- bdrv_graph_wrlock(root->bs);
+ bdrv_graph_wrlock();
bdrv_root_unref_child(root);
- bdrv_graph_wrunlock_ctx(ctx);
+ bdrv_graph_wrunlock();
}
/*
@@ -933,16 +931,15 @@ void blk_remove_bs(BlockBackend *blk)
int blk_insert_bs(BlockBackend *blk, BlockDriverState *bs, Error **errp)
{
ThrottleGroupMember *tgm = &blk->public.throttle_group_member;
- AioContext *ctx = bdrv_get_aio_context(bs);
GLOBAL_STATE_CODE();
bdrv_ref(bs);
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
blk->root = bdrv_root_attach_child(bs, "root", &child_root,
BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY,
blk->perm, blk->shared_perm,
blk, errp);
- bdrv_graph_wrunlock_ctx(ctx);
+ bdrv_graph_wrunlock();
if (blk->root == NULL) {
return -EPERM;
}
diff --git a/block/commit.c b/block/commit.c
index 69cc75be0c..1dd7a65ffb 100644
--- a/block/commit.c
+++ b/block/commit.c
@@ -100,9 +100,9 @@ static void commit_abort(Job *job)
bdrv_graph_rdunlock_main_loop();
bdrv_drained_begin(commit_top_backing_bs);
- bdrv_graph_wrlock(commit_top_backing_bs);
+ bdrv_graph_wrlock();
bdrv_replace_node(s->commit_top_bs, commit_top_backing_bs, &error_abort);
- bdrv_graph_wrunlock(commit_top_backing_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(commit_top_backing_bs);
bdrv_unref(s->commit_top_bs);
@@ -339,7 +339,7 @@ void commit_start(const char *job_id, BlockDriverState *bs,
* this is the responsibility of the interface (i.e. whoever calls
* commit_start()).
*/
- bdrv_graph_wrlock(top);
+ bdrv_graph_wrlock();
s->base_overlay = bdrv_find_overlay(top, base);
assert(s->base_overlay);
@@ -370,19 +370,19 @@ void commit_start(const char *job_id, BlockDriverState *bs,
ret = block_job_add_bdrv(&s->common, "intermediate node", iter, 0,
iter_shared_perms, errp);
if (ret < 0) {
- bdrv_graph_wrunlock(top);
+ bdrv_graph_wrunlock();
goto fail;
}
}
if (bdrv_freeze_backing_chain(commit_top_bs, base, errp) < 0) {
- bdrv_graph_wrunlock(top);
+ bdrv_graph_wrunlock();
goto fail;
}
s->chain_frozen = true;
ret = block_job_add_bdrv(&s->common, "base", base, 0, BLK_PERM_ALL, errp);
- bdrv_graph_wrunlock(top);
+ bdrv_graph_wrunlock();
if (ret < 0) {
goto fail;
@@ -434,9 +434,9 @@ fail:
* otherwise this would fail because of lack of permissions. */
if (commit_top_bs) {
bdrv_drained_begin(top);
- bdrv_graph_wrlock(top);
+ bdrv_graph_wrlock();
bdrv_replace_node(commit_top_bs, top, &error_abort);
- bdrv_graph_wrunlock(top);
+ bdrv_graph_wrunlock();
bdrv_drained_end(top);
}
}
diff --git a/block/graph-lock.c b/block/graph-lock.c
index 079e878d9b..c81162b147 100644
--- a/block/graph-lock.c
+++ b/block/graph-lock.c
@@ -106,27 +106,12 @@ static uint32_t reader_count(void)
return rd;
}
-void no_coroutine_fn bdrv_graph_wrlock(BlockDriverState *bs)
+void no_coroutine_fn bdrv_graph_wrlock(void)
{
- AioContext *ctx = NULL;
-
GLOBAL_STATE_CODE();
assert(!qatomic_read(&has_writer));
assert(!qemu_in_coroutine());
- /*
- * Release only non-mainloop AioContext. The mainloop often relies on the
- * BQL and doesn't lock the main AioContext before doing things.
- */
- if (bs) {
- ctx = bdrv_get_aio_context(bs);
- if (ctx != qemu_get_aio_context()) {
- aio_context_release(ctx);
- } else {
- ctx = NULL;
- }
- }
-
/* Make sure that constantly arriving new I/O doesn't cause starvation */
bdrv_drain_all_begin_nopoll();
@@ -155,27 +140,13 @@ void no_coroutine_fn bdrv_graph_wrlock(BlockDriverState *bs)
} while (reader_count() >= 1);
bdrv_drain_all_end();
-
- if (ctx) {
- aio_context_acquire(bdrv_get_aio_context(bs));
- }
}
-void no_coroutine_fn bdrv_graph_wrunlock_ctx(AioContext *ctx)
+void no_coroutine_fn bdrv_graph_wrunlock(void)
{
GLOBAL_STATE_CODE();
assert(qatomic_read(&has_writer));
- /*
- * Release only non-mainloop AioContext. The mainloop often relies on the
- * BQL and doesn't lock the main AioContext before doing things.
- */
- if (ctx && ctx != qemu_get_aio_context()) {
- aio_context_release(ctx);
- } else {
- ctx = NULL;
- }
-
WITH_QEMU_LOCK_GUARD(&aio_context_list_lock) {
/*
* No need for memory barriers, this works in pair with
@@ -197,17 +168,6 @@ void no_coroutine_fn bdrv_graph_wrunlock_ctx(AioContext *ctx)
* progress.
*/
aio_bh_poll(qemu_get_aio_context());
-
- if (ctx) {
- aio_context_acquire(ctx);
- }
-}
-
-void no_coroutine_fn bdrv_graph_wrunlock(BlockDriverState *bs)
-{
- AioContext *ctx = bs ? bdrv_get_aio_context(bs) : NULL;
-
- bdrv_graph_wrunlock_ctx(ctx);
}
void coroutine_fn bdrv_graph_co_rdlock(void)
diff --git a/block/mirror.c b/block/mirror.c
index cd9d3ad4a8..51f9e2f17c 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -764,7 +764,7 @@ static int mirror_exit_common(Job *job)
* check for an op blocker on @to_replace, and we have our own
* there.
*/
- bdrv_graph_wrlock(target_bs);
+ bdrv_graph_wrlock();
if (bdrv_recurse_can_replace(src, to_replace)) {
bdrv_replace_node(to_replace, target_bs, &local_err);
} else {
@@ -773,7 +773,7 @@ static int mirror_exit_common(Job *job)
"would not lead to an abrupt change of visible data",
to_replace->node_name, target_bs->node_name);
}
- bdrv_graph_wrunlock(target_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(to_replace);
if (local_err) {
error_report_err(local_err);
@@ -796,9 +796,9 @@ static int mirror_exit_common(Job *job)
* valid.
*/
block_job_remove_all_bdrv(bjob);
- bdrv_graph_wrlock(mirror_top_bs);
+ bdrv_graph_wrlock();
bdrv_replace_node(mirror_top_bs, mirror_top_bs->backing->bs, &error_abort);
- bdrv_graph_wrunlock(mirror_top_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(target_bs);
bdrv_unref(target_bs);
@@ -1914,13 +1914,13 @@ static BlockJob *mirror_start_job(
*/
bdrv_disable_dirty_bitmap(s->dirty_bitmap);
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
ret = block_job_add_bdrv(&s->common, "source", bs, 0,
BLK_PERM_WRITE_UNCHANGED | BLK_PERM_WRITE |
BLK_PERM_CONSISTENT_READ,
errp);
if (ret < 0) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
goto fail;
}
@@ -1965,17 +1965,17 @@ static BlockJob *mirror_start_job(
ret = block_job_add_bdrv(&s->common, "intermediate node", iter, 0,
iter_shared_perms, errp);
if (ret < 0) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
goto fail;
}
}
if (bdrv_freeze_backing_chain(mirror_top_bs, target, errp) < 0) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
goto fail;
}
}
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
QTAILQ_INIT(&s->ops_in_flight);
@@ -2001,12 +2001,12 @@ fail:
bs_opaque->stop = true;
bdrv_drained_begin(bs);
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
assert(mirror_top_bs->backing->bs == bs);
bdrv_child_refresh_perms(mirror_top_bs, mirror_top_bs->backing,
&error_abort);
bdrv_replace_node(mirror_top_bs, bs, &error_abort);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(bs);
bdrv_unref(mirror_top_bs);
diff --git a/block/qcow2.c b/block/qcow2.c
index 13e032bd5e..9bee66fff5 100644
--- a/block/qcow2.c
+++ b/block/qcow2.c
@@ -2807,9 +2807,9 @@ qcow2_do_close(BlockDriverState *bs, bool close_data_file)
if (close_data_file && has_data_file(bs)) {
GLOBAL_STATE_CODE();
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, s->data_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
s->data_file = NULL;
bdrv_graph_rdlock_main_loop();
}
diff --git a/block/quorum.c b/block/quorum.c
index 505b8b3e18..db8fe891c4 100644
--- a/block/quorum.c
+++ b/block/quorum.c
@@ -1037,14 +1037,14 @@ static int quorum_open(BlockDriverState *bs, QDict *options, int flags,
close_exit:
/* cleanup on error */
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
for (i = 0; i < s->num_children; i++) {
if (!opened[i]) {
continue;
}
bdrv_unref_child(bs, s->children[i]);
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
g_free(s->children);
g_free(opened);
exit:
@@ -1057,11 +1057,11 @@ static void quorum_close(BlockDriverState *bs)
BDRVQuorumState *s = bs->opaque;
int i;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
for (i = 0; i < s->num_children; i++) {
bdrv_unref_child(bs, s->children[i]);
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
g_free(s->children);
}
diff --git a/block/replication.c b/block/replication.c
index 5ded5f1ca9..424b537ff7 100644
--- a/block/replication.c
+++ b/block/replication.c
@@ -560,7 +560,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
return;
}
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
bdrv_ref(hidden_disk->bs);
s->hidden_disk = bdrv_attach_child(bs, hidden_disk->bs, "hidden disk",
@@ -568,7 +568,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
&local_err);
if (local_err) {
error_propagate(errp, local_err);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
aio_context_release(aio_context);
return;
}
@@ -579,7 +579,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
BDRV_CHILD_DATA, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
aio_context_release(aio_context);
return;
}
@@ -592,7 +592,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (!top_bs || !bdrv_is_root_node(top_bs) ||
!check_top_bs(top_bs, bs)) {
error_setg(errp, "No top_bs or it is invalid");
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
reopen_backing_file(bs, false, NULL);
aio_context_release(aio_context);
return;
@@ -600,7 +600,7 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
bdrv_op_block_all(top_bs, s->blocker);
bdrv_op_unblock(top_bs, BLOCK_OP_TYPE_DATAPLANE, s->blocker);
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
s->backup_job = backup_job_create(
NULL, s->secondary_disk->bs, s->hidden_disk->bs,
@@ -691,12 +691,12 @@ static void replication_done(void *opaque, int ret)
if (ret == 0) {
s->stage = BLOCK_REPLICATION_DONE;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, s->secondary_disk);
s->secondary_disk = NULL;
bdrv_unref_child(bs, s->hidden_disk);
s->hidden_disk = NULL;
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
s->error = 0;
} else {
diff --git a/block/snapshot.c b/block/snapshot.c
index ec8cf4810b..e486d3e205 100644
--- a/block/snapshot.c
+++ b/block/snapshot.c
@@ -290,9 +290,9 @@ int bdrv_snapshot_goto(BlockDriverState *bs,
}
/* .bdrv_open() will re-attach it */
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, fallback);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
ret = bdrv_snapshot_goto(fallback_bs, snapshot_id, errp);
open_ret = drv->bdrv_open(bs, options, bs->open_flags, &local_err);
diff --git a/block/stream.c b/block/stream.c
index 01fe7c0f16..048c2d282f 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -99,9 +99,9 @@ static int stream_prepare(Job *job)
}
}
- bdrv_graph_wrlock(s->target_bs);
+ bdrv_graph_wrlock();
bdrv_set_backing_hd_drained(unfiltered_bs, base, &local_err);
- bdrv_graph_wrunlock(s->target_bs);
+ bdrv_graph_wrunlock();
/*
* This call will do I/O, so the graph can change again from here on.
@@ -366,10 +366,10 @@ void stream_start(const char *job_id, BlockDriverState *bs,
* already have our own plans. Also don't allow resize as the image size is
* queried only at the job start and then cached.
*/
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
if (block_job_add_bdrv(&s->common, "active node", bs, 0,
basic_flags | BLK_PERM_WRITE, errp)) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
goto fail;
}
@@ -389,11 +389,11 @@ void stream_start(const char *job_id, BlockDriverState *bs,
ret = block_job_add_bdrv(&s->common, "intermediate node", iter, 0,
basic_flags, errp);
if (ret < 0) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
goto fail;
}
}
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
s->base_overlay = base_overlay;
s->above_base = above_base;
diff --git a/block/vmdk.c b/block/vmdk.c
index d6971c7067..bf78e12383 100644
--- a/block/vmdk.c
+++ b/block/vmdk.c
@@ -272,7 +272,7 @@ static void vmdk_free_extents(BlockDriverState *bs)
BDRVVmdkState *s = bs->opaque;
VmdkExtent *e;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
for (i = 0; i < s->num_extents; i++) {
e = &s->extents[i];
g_free(e->l1_table);
@@ -283,7 +283,7 @@ static void vmdk_free_extents(BlockDriverState *bs)
bdrv_unref_child(bs, e->file);
}
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
g_free(s->extents);
}
@@ -1247,9 +1247,9 @@ vmdk_parse_extents(const char *desc, BlockDriverState *bs, QDict *options,
0, 0, 0, 0, 0, &extent, errp);
if (ret < 0) {
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, extent_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_graph_rdlock_main_loop();
goto out;
}
@@ -1266,9 +1266,9 @@ vmdk_parse_extents(const char *desc, BlockDriverState *bs, QDict *options,
g_free(buf);
if (ret) {
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, extent_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_graph_rdlock_main_loop();
goto out;
}
@@ -1277,9 +1277,9 @@ vmdk_parse_extents(const char *desc, BlockDriverState *bs, QDict *options,
ret = vmdk_open_se_sparse(bs, extent_file, bs->open_flags, errp);
if (ret) {
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, extent_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_graph_rdlock_main_loop();
goto out;
}
@@ -1287,9 +1287,9 @@ vmdk_parse_extents(const char *desc, BlockDriverState *bs, QDict *options,
} else {
error_setg(errp, "Unsupported extent type '%s'", type);
bdrv_graph_rdunlock_main_loop();
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(bs, extent_file);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_graph_rdlock_main_loop();
ret = -ENOTSUP;
goto out;
diff --git a/blockdev.c b/blockdev.c
index 4c1177e8db..db9cc96510 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -1611,9 +1611,9 @@ static void external_snapshot_abort(void *opaque)
}
bdrv_drained_begin(state->new_bs);
- bdrv_graph_wrlock(state->old_bs);
+ bdrv_graph_wrlock();
bdrv_replace_node(state->new_bs, state->old_bs, &error_abort);
- bdrv_graph_wrunlock(state->old_bs);
+ bdrv_graph_wrunlock();
bdrv_drained_end(state->new_bs);
bdrv_unref(state->old_bs); /* bdrv_replace_node() ref'ed old_bs */
@@ -3656,7 +3656,7 @@ void qmp_x_blockdev_change(const char *parent, const char *child,
BlockDriverState *parent_bs, *new_bs = NULL;
BdrvChild *p_child;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
parent_bs = bdrv_lookup_bs(parent, parent, errp);
if (!parent_bs) {
@@ -3692,7 +3692,7 @@ void qmp_x_blockdev_change(const char *parent, const char *child,
}
out:
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
}
BlockJobInfoList *qmp_query_block_jobs(Error **errp)
diff --git a/blockjob.c b/blockjob.c
index b7a29052b9..7310412313 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -199,7 +199,7 @@ void block_job_remove_all_bdrv(BlockJob *job)
* to process an already freed BdrvChild.
*/
aio_context_release(job->job.aio_context);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
aio_context_acquire(job->job.aio_context);
while (job->nodes) {
GSList *l = job->nodes;
@@ -212,7 +212,7 @@ void block_job_remove_all_bdrv(BlockJob *job)
g_slist_free_1(l);
}
- bdrv_graph_wrunlock_ctx(job->job.aio_context);
+ bdrv_graph_wrunlock();
}
bool block_job_has_bdrv(BlockJob *job, BlockDriverState *bs)
@@ -514,7 +514,7 @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,
int ret;
GLOBAL_STATE_CODE();
- bdrv_graph_wrlock(bs);
+ bdrv_graph_wrlock();
if (job_id == NULL && !(flags & JOB_INTERNAL)) {
job_id = bdrv_get_device_name(bs);
@@ -523,7 +523,7 @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,
job = job_create(job_id, &driver->job_driver, txn, bdrv_get_aio_context(bs),
flags, cb, opaque, errp);
if (job == NULL) {
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
return NULL;
}
@@ -563,11 +563,11 @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,
goto fail;
}
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
return job;
fail:
- bdrv_graph_wrunlock(bs);
+ bdrv_graph_wrunlock();
job_early_fail(&job->job);
return NULL;
}
diff --git a/tests/unit/test-bdrv-drain.c b/tests/unit/test-bdrv-drain.c
index 704d1a3f36..d9754dfebc 100644
--- a/tests/unit/test-bdrv-drain.c
+++ b/tests/unit/test-bdrv-drain.c
@@ -807,9 +807,9 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
tjob->bs = src;
job = &tjob->common;
- bdrv_graph_wrlock(target);
+ bdrv_graph_wrlock();
block_job_add_bdrv(job, "target", target, 0, BLK_PERM_ALL, &error_abort);
- bdrv_graph_wrunlock(target);
+ bdrv_graph_wrunlock();
switch (result) {
case TEST_JOB_SUCCESS:
@@ -991,11 +991,11 @@ static void bdrv_test_top_close(BlockDriverState *bs)
{
BdrvChild *c, *next_c;
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
QLIST_FOREACH_SAFE(c, &bs->children, next, next_c) {
bdrv_unref_child(bs, c);
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
}
static int coroutine_fn GRAPH_RDLOCK
@@ -1085,10 +1085,10 @@ static void do_test_delete_by_drain(bool detach_instead_of_delete,
null_bs = bdrv_open("null-co://", NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
&error_abort);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(bs, null_bs, "null-child", &child_of_bds,
BDRV_CHILD_DATA, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
/* This child will be the one to pass to requests through to, and
* it will stall until a drain occurs */
@@ -1096,21 +1096,21 @@ static void do_test_delete_by_drain(bool detach_instead_of_delete,
&error_abort);
child_bs->total_sectors = 65536 >> BDRV_SECTOR_BITS;
/* Takes our reference to child_bs */
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
tts->wait_child = bdrv_attach_child(bs, child_bs, "wait-child",
&child_of_bds,
BDRV_CHILD_DATA | BDRV_CHILD_PRIMARY,
&error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
/* This child is just there to be deleted
* (for detach_instead_of_delete == true) */
null_bs = bdrv_open("null-co://", NULL, NULL, BDRV_O_RDWR | BDRV_O_PROTOCOL,
&error_abort);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(bs, null_bs, "null-child", &child_of_bds, BDRV_CHILD_DATA,
&error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
blk = blk_new(qemu_get_aio_context(), BLK_PERM_ALL, BLK_PERM_ALL);
blk_insert_bs(blk, bs, &error_abort);
@@ -1193,14 +1193,14 @@ static void no_coroutine_fn detach_indirect_bh(void *opaque)
bdrv_dec_in_flight(data->child_b->bs);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_unref_child(data->parent_b, data->child_b);
bdrv_ref(data->c);
data->child_c = bdrv_attach_child(data->parent_b, data->c, "PB-C",
&child_of_bds, BDRV_CHILD_DATA,
&error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
}
static void coroutine_mixed_fn detach_by_parent_aio_cb(void *opaque, int ret)
@@ -1298,7 +1298,7 @@ static void TSA_NO_TSA test_detach_indirect(bool by_parent_cb)
/* Set child relationships */
bdrv_ref(b);
bdrv_ref(a);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
child_b = bdrv_attach_child(parent_b, b, "PB-B", &child_of_bds,
BDRV_CHILD_DATA, &error_abort);
child_a = bdrv_attach_child(parent_b, a, "PB-A", &child_of_bds,
@@ -1308,7 +1308,7 @@ static void TSA_NO_TSA test_detach_indirect(bool by_parent_cb)
bdrv_attach_child(parent_a, a, "PA-A",
by_parent_cb ? &child_of_bds : &detach_by_driver_cb_class,
BDRV_CHILD_DATA, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
g_assert_cmpint(parent_a->refcnt, ==, 1);
g_assert_cmpint(parent_b->refcnt, ==, 1);
@@ -1727,7 +1727,7 @@ static void test_drop_intermediate_poll(void)
* Establish the chain last, so the chain links are the first
* elements in the BDS.parents lists
*/
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
for (i = 0; i < 3; i++) {
if (i) {
/* Takes the reference to chain[i - 1] */
@@ -1735,7 +1735,7 @@ static void test_drop_intermediate_poll(void)
&chain_child_class, BDRV_CHILD_COW, &error_abort);
}
}
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
job = block_job_create("job", &test_simple_job_driver, NULL, job_node,
0, BLK_PERM_ALL, 0, 0, NULL, NULL, &error_abort);
@@ -1982,10 +1982,10 @@ static void do_test_replace_child_mid_drain(int old_drain_count,
new_child_bs->total_sectors = 1;
bdrv_ref(old_child_bs);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(parent_bs, old_child_bs, "child", &child_of_bds,
BDRV_CHILD_COW, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
parent_s->setup_completed = true;
for (i = 0; i < old_drain_count; i++) {
@@ -2016,9 +2016,9 @@ static void do_test_replace_child_mid_drain(int old_drain_count,
g_assert(parent_bs->quiesce_counter == old_drain_count);
bdrv_drained_begin(old_child_bs);
bdrv_drained_begin(new_child_bs);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_replace_node(old_child_bs, new_child_bs, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_drained_end(new_child_bs);
bdrv_drained_end(old_child_bs);
g_assert(parent_bs->quiesce_counter == new_drain_count);
diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c
index 074adcbb93..8ee6ef38d8 100644
--- a/tests/unit/test-bdrv-graph-mod.c
+++ b/tests/unit/test-bdrv-graph-mod.c
@@ -137,10 +137,10 @@ static void test_update_perm_tree(void)
blk_insert_bs(root, bs, &error_abort);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(filter, bs, "child", &child_of_bds,
BDRV_CHILD_DATA, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
aio_context_acquire(qemu_get_aio_context());
ret = bdrv_append(filter, bs, NULL);
@@ -206,11 +206,11 @@ static void test_should_update_child(void)
bdrv_set_backing_hd(target, bs, &error_abort);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
g_assert(target->backing->bs == bs);
bdrv_attach_child(filter, target, "target", &child_of_bds,
BDRV_CHILD_DATA, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
aio_context_acquire(qemu_get_aio_context());
bdrv_append(filter, bs, &error_abort);
aio_context_release(qemu_get_aio_context());
@@ -248,7 +248,7 @@ static void test_parallel_exclusive_write(void)
bdrv_ref(base);
bdrv_ref(fl1);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(top, fl1, "backing", &child_of_bds,
BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY,
&error_abort);
@@ -260,7 +260,7 @@ static void test_parallel_exclusive_write(void)
&error_abort);
bdrv_replace_node(fl1, fl2, &error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
bdrv_drained_end(fl2);
bdrv_drained_end(fl1);
@@ -367,7 +367,7 @@ static void test_parallel_perm_update(void)
*/
bdrv_ref(base);
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(top, ws, "file", &child_of_bds, BDRV_CHILD_DATA,
&error_abort);
c_fl1 = bdrv_attach_child(ws, fl1, "first", &child_of_bds,
@@ -380,7 +380,7 @@ static void test_parallel_perm_update(void)
bdrv_attach_child(fl2, base, "backing", &child_of_bds,
BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY,
&error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
/* Select fl1 as first child to be active */
s->selected = c_fl1;
@@ -434,11 +434,11 @@ static void test_append_greedy_filter(void)
BlockDriverState *base = no_perm_node("base");
BlockDriverState *fl = exclusive_writer_node("fl1");
- bdrv_graph_wrlock(NULL);
+ bdrv_graph_wrlock();
bdrv_attach_child(top, base, "backing", &child_of_bds,
BDRV_CHILD_FILTERED | BDRV_CHILD_PRIMARY,
&error_abort);
- bdrv_graph_wrunlock(NULL);
+ bdrv_graph_wrunlock();
aio_context_acquire(qemu_get_aio_context());
bdrv_append(fl, base, &error_abort);
diff --git a/scripts/block-coroutine-wrapper.py b/scripts/block-coroutine-wrapper.py
index a38e5833fb..38364fa557 100644
--- a/scripts/block-coroutine-wrapper.py
+++ b/scripts/block-coroutine-wrapper.py
@@ -261,8 +261,8 @@ def gen_no_co_wrapper(func: FuncDecl) -> str:
graph_lock=' bdrv_graph_rdlock_main_loop();'
graph_unlock=' bdrv_graph_rdunlock_main_loop();'
elif func.graph_wrlock:
- graph_lock=' bdrv_graph_wrlock(NULL);'
- graph_unlock=' bdrv_graph_wrunlock(NULL);'
+ graph_lock=' bdrv_graph_wrlock();'
+ graph_unlock=' bdrv_graph_wrunlock();'
return f"""\
/*
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 05/12] block: remove AioContext locking
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (3 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 04/12] graph-lock: remove AioContext locking Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 9:12 ` Paul Durrant
` (2 more replies)
2023-11-29 19:55 ` [PATCH 06/12] scsi: " Stefan Hajnoczi
` (6 subsequent siblings)
11 siblings, 3 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
This is the big patch that removes
aio_context_acquire()/aio_context_release() from the block layer and
affected block layer users.
There isn't a clean way to split this patch and the reviewers are likely
the same group of people, so I decided to do it in one patch.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/block/block-global-state.h | 9 +-
include/block/block-io.h | 3 +-
include/block/snapshot.h | 2 -
block.c | 234 +---------------------
block/block-backend.c | 14 --
block/copy-before-write.c | 22 +--
block/export/export.c | 22 +--
block/io.c | 45 +----
block/mirror.c | 19 --
block/monitor/bitmap-qmp-cmds.c | 20 +-
block/monitor/block-hmp-cmds.c | 29 ---
block/qapi-sysemu.c | 27 +--
block/qapi.c | 18 +-
block/raw-format.c | 5 -
block/replication.c | 58 +-----
block/snapshot.c | 22 +--
block/write-threshold.c | 6 -
blockdev.c | 307 +++++------------------------
blockjob.c | 18 --
hw/block/dataplane/virtio-blk.c | 10 -
hw/block/dataplane/xen-block.c | 17 +-
hw/block/virtio-blk.c | 45 +----
hw/core/qdev-properties-system.c | 9 -
job.c | 16 --
migration/block.c | 33 +---
migration/migration-hmp-cmds.c | 3 -
migration/savevm.c | 22 ---
net/colo-compare.c | 2 -
qemu-img.c | 4 -
qemu-io.c | 10 +-
qemu-nbd.c | 2 -
replay/replay-debugging.c | 4 -
tests/unit/test-bdrv-drain.c | 51 +----
tests/unit/test-bdrv-graph-mod.c | 6 -
tests/unit/test-block-iothread.c | 31 ---
tests/unit/test-blockjob.c | 137 -------------
tests/unit/test-replication.c | 11 --
util/async.c | 4 -
util/vhost-user-server.c | 3 -
scripts/block-coroutine-wrapper.py | 3 -
tests/tsan/suppressions.tsan | 1 -
41 files changed, 102 insertions(+), 1202 deletions(-)
diff --git a/include/block/block-global-state.h b/include/block/block-global-state.h
index 6b21fbc73f..0327f1c605 100644
--- a/include/block/block-global-state.h
+++ b/include/block/block-global-state.h
@@ -31,11 +31,10 @@
/*
* Global state (GS) API. These functions run under the BQL.
*
- * If a function modifies the graph, it also uses drain and/or
- * aio_context_acquire/release to be sure it has unique access.
- * aio_context locking is needed together with BQL because of
- * the thread-safe I/O API that concurrently runs and accesses
- * the graph without the BQL.
+ * If a function modifies the graph, it also uses the graph lock to be sure it
+ * has unique access. The graph lock is needed together with BQL because of the
+ * thread-safe I/O API that concurrently runs and accesses the graph without
+ * the BQL.
*
* It is important to note that not all of these functions are
* necessarily limited to running under the BQL, but they would
diff --git a/include/block/block-io.h b/include/block/block-io.h
index f8729ccc55..8eb39a858b 100644
--- a/include/block/block-io.h
+++ b/include/block/block-io.h
@@ -31,8 +31,7 @@
/*
* I/O API functions. These functions are thread-safe, and therefore
- * can run in any thread as long as the thread has called
- * aio_context_acquire/release().
+ * can run in any thread.
*
* These functions can only call functions from I/O and Common categories,
* but can be invoked by GS, "I/O or GS" and I/O APIs.
diff --git a/include/block/snapshot.h b/include/block/snapshot.h
index d49c5599d9..304cc6ea61 100644
--- a/include/block/snapshot.h
+++ b/include/block/snapshot.h
@@ -86,8 +86,6 @@ int bdrv_snapshot_load_tmp_by_id_or_name(BlockDriverState *bs,
/*
* Group operations. All block drivers are involved.
- * These functions will properly handle dataplane (take aio_context_acquire
- * when appropriate for appropriate block drivers
*/
bool bdrv_all_can_snapshot(bool has_devices, strList *devices,
diff --git a/block.c b/block.c
index 25e1ebc606..91ace5d2d5 100644
--- a/block.c
+++ b/block.c
@@ -1625,7 +1625,6 @@ static int no_coroutine_fn GRAPH_UNLOCKED
bdrv_open_driver(BlockDriverState *bs, BlockDriver *drv, const char *node_name,
QDict *options, int open_flags, Error **errp)
{
- AioContext *ctx;
Error *local_err = NULL;
int i, ret;
GLOBAL_STATE_CODE();
@@ -1673,21 +1672,15 @@ bdrv_open_driver(BlockDriverState *bs, BlockDriver *drv, const char *node_name,
bs->supported_read_flags |= BDRV_REQ_REGISTERED_BUF;
bs->supported_write_flags |= BDRV_REQ_REGISTERED_BUF;
- /* Get the context after .bdrv_open, it can change the context */
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
-
ret = bdrv_refresh_total_sectors(bs, bs->total_sectors);
if (ret < 0) {
error_setg_errno(errp, -ret, "Could not refresh total sector count");
- aio_context_release(ctx);
return ret;
}
bdrv_graph_rdlock_main_loop();
bdrv_refresh_limits(bs, NULL, &local_err);
bdrv_graph_rdunlock_main_loop();
- aio_context_release(ctx);
if (local_err) {
error_propagate(errp, local_err);
@@ -3062,7 +3055,7 @@ bdrv_attach_child_common(BlockDriverState *child_bs,
Transaction *tran, Error **errp)
{
BdrvChild *new_child;
- AioContext *parent_ctx, *new_child_ctx;
+ AioContext *parent_ctx;
AioContext *child_ctx = bdrv_get_aio_context(child_bs);
assert(child_class->get_parent_desc);
@@ -3114,12 +3107,6 @@ bdrv_attach_child_common(BlockDriverState *child_bs,
}
}
- new_child_ctx = bdrv_get_aio_context(child_bs);
- if (new_child_ctx != child_ctx) {
- aio_context_release(child_ctx);
- aio_context_acquire(new_child_ctx);
- }
-
bdrv_ref(child_bs);
/*
* Let every new BdrvChild start with a drained parent. Inserting the child
@@ -3149,11 +3136,6 @@ bdrv_attach_child_common(BlockDriverState *child_bs,
};
tran_add(tran, &bdrv_attach_child_common_drv, s);
- if (new_child_ctx != child_ctx) {
- aio_context_release(new_child_ctx);
- aio_context_acquire(child_ctx);
- }
-
return new_child;
}
@@ -3605,7 +3587,6 @@ int bdrv_open_backing_file(BlockDriverState *bs, QDict *parent_options,
int ret = 0;
bool implicit_backing = false;
BlockDriverState *backing_hd;
- AioContext *backing_hd_ctx;
QDict *options;
QDict *tmp_parent_options = NULL;
Error *local_err = NULL;
@@ -3691,11 +3672,8 @@ int bdrv_open_backing_file(BlockDriverState *bs, QDict *parent_options,
/* Hook up the backing file link; drop our reference, bs owns the
* backing_hd reference now */
- backing_hd_ctx = bdrv_get_aio_context(backing_hd);
- aio_context_acquire(backing_hd_ctx);
ret = bdrv_set_backing_hd(bs, backing_hd, errp);
bdrv_unref(backing_hd);
- aio_context_release(backing_hd_ctx);
if (ret < 0) {
goto free_exit;
@@ -3780,7 +3758,6 @@ BdrvChild *bdrv_open_child(const char *filename,
{
BlockDriverState *bs;
BdrvChild *child;
- AioContext *ctx;
GLOBAL_STATE_CODE();
@@ -3791,11 +3768,8 @@ BdrvChild *bdrv_open_child(const char *filename,
}
bdrv_graph_wrlock();
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
child = bdrv_attach_child(parent, bs, bdref_key, child_class, child_role,
errp);
- aio_context_release(ctx);
bdrv_graph_wrunlock();
return child;
@@ -3881,7 +3855,6 @@ static BlockDriverState *bdrv_append_temp_snapshot(BlockDriverState *bs,
int64_t total_size;
QemuOpts *opts = NULL;
BlockDriverState *bs_snapshot = NULL;
- AioContext *ctx = bdrv_get_aio_context(bs);
int ret;
GLOBAL_STATE_CODE();
@@ -3890,9 +3863,7 @@ static BlockDriverState *bdrv_append_temp_snapshot(BlockDriverState *bs,
instead of opening 'filename' directly */
/* Get the required size from the image */
- aio_context_acquire(ctx);
total_size = bdrv_getlength(bs);
- aio_context_release(ctx);
if (total_size < 0) {
error_setg_errno(errp, -total_size, "Could not get image size");
@@ -3927,10 +3898,7 @@ static BlockDriverState *bdrv_append_temp_snapshot(BlockDriverState *bs,
goto out;
}
- aio_context_acquire(ctx);
ret = bdrv_append(bs_snapshot, bs, errp);
- aio_context_release(ctx);
-
if (ret < 0) {
bs_snapshot = NULL;
goto out;
@@ -3974,7 +3942,6 @@ bdrv_open_inherit(const char *filename, const char *reference, QDict *options,
Error *local_err = NULL;
QDict *snapshot_options = NULL;
int snapshot_flags = 0;
- AioContext *ctx = qemu_get_aio_context();
assert(!child_class || !flags);
assert(!child_class == !parent);
@@ -4115,12 +4082,10 @@ bdrv_open_inherit(const char *filename, const char *reference, QDict *options,
/* Not requesting BLK_PERM_CONSISTENT_READ because we're only
* looking at the header to guess the image format. This works even
* in cases where a guest would not see a consistent state. */
- ctx = bdrv_get_aio_context(file_bs);
- aio_context_acquire(ctx);
+ AioContext *ctx = bdrv_get_aio_context(file_bs);
file = blk_new(ctx, 0, BLK_PERM_ALL);
blk_insert_bs(file, file_bs, &local_err);
bdrv_unref(file_bs);
- aio_context_release(ctx);
if (local_err) {
goto fail;
@@ -4167,13 +4132,8 @@ bdrv_open_inherit(const char *filename, const char *reference, QDict *options,
goto fail;
}
- /* The AioContext could have changed during bdrv_open_common() */
- ctx = bdrv_get_aio_context(bs);
-
if (file) {
- aio_context_acquire(ctx);
blk_unref(file);
- aio_context_release(ctx);
file = NULL;
}
@@ -4231,16 +4191,13 @@ bdrv_open_inherit(const char *filename, const char *reference, QDict *options,
* (snapshot_bs); thus, we have to drop the strong reference to bs
* (which we obtained by calling bdrv_new()). bs will not be deleted,
* though, because the overlay still has a reference to it. */
- aio_context_acquire(ctx);
bdrv_unref(bs);
- aio_context_release(ctx);
bs = snapshot_bs;
}
return bs;
fail:
- aio_context_acquire(ctx);
blk_unref(file);
qobject_unref(snapshot_options);
qobject_unref(bs->explicit_options);
@@ -4249,14 +4206,11 @@ fail:
bs->options = NULL;
bs->explicit_options = NULL;
bdrv_unref(bs);
- aio_context_release(ctx);
error_propagate(errp, local_err);
return NULL;
close_and_fail:
- aio_context_acquire(ctx);
bdrv_unref(bs);
- aio_context_release(ctx);
qobject_unref(snapshot_options);
qobject_unref(options);
error_propagate(errp, local_err);
@@ -4540,12 +4494,7 @@ void bdrv_reopen_queue_free(BlockReopenQueue *bs_queue)
if (bs_queue) {
BlockReopenQueueEntry *bs_entry, *next;
QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) {
- AioContext *ctx = bdrv_get_aio_context(bs_entry->state.bs);
-
- aio_context_acquire(ctx);
bdrv_drained_end(bs_entry->state.bs);
- aio_context_release(ctx);
-
qobject_unref(bs_entry->state.explicit_options);
qobject_unref(bs_entry->state.options);
g_free(bs_entry);
@@ -4577,7 +4526,6 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
{
int ret = -1;
BlockReopenQueueEntry *bs_entry, *next;
- AioContext *ctx;
Transaction *tran = tran_new();
g_autoptr(GSList) refresh_list = NULL;
@@ -4586,10 +4534,7 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
GLOBAL_STATE_CODE();
QTAILQ_FOREACH(bs_entry, bs_queue, entry) {
- ctx = bdrv_get_aio_context(bs_entry->state.bs);
- aio_context_acquire(ctx);
ret = bdrv_flush(bs_entry->state.bs);
- aio_context_release(ctx);
if (ret < 0) {
error_setg_errno(errp, -ret, "Error flushing drive");
goto abort;
@@ -4598,10 +4543,7 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
QTAILQ_FOREACH(bs_entry, bs_queue, entry) {
assert(bs_entry->state.bs->quiesce_counter > 0);
- ctx = bdrv_get_aio_context(bs_entry->state.bs);
- aio_context_acquire(ctx);
ret = bdrv_reopen_prepare(&bs_entry->state, bs_queue, tran, errp);
- aio_context_release(ctx);
if (ret < 0) {
goto abort;
}
@@ -4644,10 +4586,7 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
* to first element.
*/
QTAILQ_FOREACH_REVERSE(bs_entry, bs_queue, entry) {
- ctx = bdrv_get_aio_context(bs_entry->state.bs);
- aio_context_acquire(ctx);
bdrv_reopen_commit(&bs_entry->state);
- aio_context_release(ctx);
}
bdrv_graph_wrlock();
@@ -4658,10 +4597,7 @@ int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
BlockDriverState *bs = bs_entry->state.bs;
if (bs->drv->bdrv_reopen_commit_post) {
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
bs->drv->bdrv_reopen_commit_post(&bs_entry->state);
- aio_context_release(ctx);
}
}
@@ -4675,10 +4611,7 @@ abort:
QTAILQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) {
if (bs_entry->prepared) {
- ctx = bdrv_get_aio_context(bs_entry->state.bs);
- aio_context_acquire(ctx);
bdrv_reopen_abort(&bs_entry->state);
- aio_context_release(ctx);
}
}
@@ -4691,24 +4624,13 @@ cleanup:
int bdrv_reopen(BlockDriverState *bs, QDict *opts, bool keep_old_opts,
Error **errp)
{
- AioContext *ctx = bdrv_get_aio_context(bs);
BlockReopenQueue *queue;
- int ret;
GLOBAL_STATE_CODE();
queue = bdrv_reopen_queue(NULL, bs, opts, keep_old_opts);
- if (ctx != qemu_get_aio_context()) {
- aio_context_release(ctx);
- }
- ret = bdrv_reopen_multiple(queue, errp);
-
- if (ctx != qemu_get_aio_context()) {
- aio_context_acquire(ctx);
- }
-
- return ret;
+ return bdrv_reopen_multiple(queue, errp);
}
int bdrv_reopen_set_read_only(BlockDriverState *bs, bool read_only,
@@ -4760,7 +4682,6 @@ bdrv_reopen_parse_file_or_backing(BDRVReopenState *reopen_state,
const char *child_name = is_backing ? "backing" : "file";
QObject *value;
const char *str;
- AioContext *ctx, *old_ctx;
bool has_child;
int ret;
@@ -4844,13 +4765,6 @@ bdrv_reopen_parse_file_or_backing(BDRVReopenState *reopen_state,
bdrv_drained_begin(old_child_bs);
}
- old_ctx = bdrv_get_aio_context(bs);
- ctx = bdrv_get_aio_context(new_child_bs);
- if (old_ctx != ctx) {
- aio_context_release(old_ctx);
- aio_context_acquire(ctx);
- }
-
bdrv_graph_rdunlock_main_loop();
bdrv_graph_wrlock();
@@ -4859,11 +4773,6 @@ bdrv_reopen_parse_file_or_backing(BDRVReopenState *reopen_state,
bdrv_graph_wrunlock();
- if (old_ctx != ctx) {
- aio_context_release(ctx);
- aio_context_acquire(old_ctx);
- }
-
if (old_child_bs) {
bdrv_drained_end(old_child_bs);
bdrv_unref(old_child_bs);
@@ -5537,7 +5446,6 @@ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top,
int ret;
BdrvChild *child;
Transaction *tran = tran_new();
- AioContext *old_context, *new_context = NULL;
GLOBAL_STATE_CODE();
@@ -5545,21 +5453,8 @@ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top,
assert(!bs_new->backing);
bdrv_graph_rdunlock_main_loop();
- old_context = bdrv_get_aio_context(bs_top);
bdrv_drained_begin(bs_top);
-
- /*
- * bdrv_drained_begin() requires that only the AioContext of the drained
- * node is locked, and at this point it can still differ from the AioContext
- * of bs_top.
- */
- new_context = bdrv_get_aio_context(bs_new);
- aio_context_release(old_context);
- aio_context_acquire(new_context);
bdrv_drained_begin(bs_new);
- aio_context_release(new_context);
- aio_context_acquire(old_context);
- new_context = NULL;
bdrv_graph_wrlock();
@@ -5571,18 +5466,6 @@ int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top,
goto out;
}
- /*
- * bdrv_attach_child_noperm could change the AioContext of bs_top and
- * bs_new, but at least they are in the same AioContext now. This is the
- * AioContext that we need to lock for the rest of the function.
- */
- new_context = bdrv_get_aio_context(bs_top);
-
- if (old_context != new_context) {
- aio_context_release(old_context);
- aio_context_acquire(new_context);
- }
-
ret = bdrv_replace_node_noperm(bs_top, bs_new, true, tran, errp);
if (ret < 0) {
goto out;
@@ -5598,11 +5481,6 @@ out:
bdrv_drained_end(bs_top);
bdrv_drained_end(bs_new);
- if (new_context && old_context != new_context) {
- aio_context_release(new_context);
- aio_context_acquire(old_context);
- }
-
return ret;
}
@@ -5697,12 +5575,8 @@ BlockDriverState *bdrv_insert_node(BlockDriverState *bs, QDict *options,
GLOBAL_STATE_CODE();
- aio_context_release(ctx);
- aio_context_acquire(qemu_get_aio_context());
new_node_bs = bdrv_new_open_driver_opts(drv, node_name, options, flags,
errp);
- aio_context_release(qemu_get_aio_context());
- aio_context_acquire(ctx);
assert(bdrv_get_aio_context(bs) == ctx);
options = NULL; /* bdrv_new_open_driver() eats options */
@@ -7037,12 +6911,9 @@ void bdrv_activate_all(Error **errp)
GRAPH_RDLOCK_GUARD_MAINLOOP();
for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
int ret;
- aio_context_acquire(aio_context);
ret = bdrv_activate(bs, errp);
- aio_context_release(aio_context);
if (ret < 0) {
bdrv_next_cleanup(&it);
return;
@@ -7137,20 +7008,10 @@ int bdrv_inactivate_all(void)
BlockDriverState *bs = NULL;
BdrvNextIterator it;
int ret = 0;
- GSList *aio_ctxs = NULL, *ctx;
GLOBAL_STATE_CODE();
GRAPH_RDLOCK_GUARD_MAINLOOP();
- for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
-
- if (!g_slist_find(aio_ctxs, aio_context)) {
- aio_ctxs = g_slist_prepend(aio_ctxs, aio_context);
- aio_context_acquire(aio_context);
- }
- }
-
for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
/* Nodes with BDS parents are covered by recursion from the last
* parent that gets inactivated. Don't inactivate them a second
@@ -7161,17 +7022,10 @@ int bdrv_inactivate_all(void)
ret = bdrv_inactivate_recurse(bs);
if (ret < 0) {
bdrv_next_cleanup(&it);
- goto out;
+ break;
}
}
-out:
- for (ctx = aio_ctxs; ctx != NULL; ctx = ctx->next) {
- AioContext *aio_context = ctx->data;
- aio_context_release(aio_context);
- }
- g_slist_free(aio_ctxs);
-
return ret;
}
@@ -7257,11 +7111,8 @@ void bdrv_unref(BlockDriverState *bs)
static void bdrv_schedule_unref_bh(void *opaque)
{
BlockDriverState *bs = opaque;
- AioContext *ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
bdrv_unref(bs);
- aio_context_release(ctx);
}
/*
@@ -7398,8 +7249,6 @@ void bdrv_img_create(const char *filename, const char *fmt,
return;
}
- aio_context_acquire(qemu_get_aio_context());
-
/* Create parameter list */
create_opts = qemu_opts_append(create_opts, drv->create_opts);
create_opts = qemu_opts_append(create_opts, proto_drv->create_opts);
@@ -7549,7 +7398,6 @@ out:
qemu_opts_del(opts);
qemu_opts_free(create_opts);
error_propagate(errp, local_err);
- aio_context_release(qemu_get_aio_context());
}
AioContext *bdrv_get_aio_context(BlockDriverState *bs)
@@ -7585,29 +7433,12 @@ void coroutine_fn bdrv_co_leave(BlockDriverState *bs, AioContext *old_ctx)
void coroutine_fn bdrv_co_lock(BlockDriverState *bs)
{
- AioContext *ctx = bdrv_get_aio_context(bs);
-
- /* In the main thread, bs->aio_context won't change concurrently */
- assert(qemu_get_current_aio_context() == qemu_get_aio_context());
-
- /*
- * We're in coroutine context, so we already hold the lock of the main
- * loop AioContext. Don't lock it twice to avoid deadlocks.
- */
- assert(qemu_in_coroutine());
- if (ctx != qemu_get_aio_context()) {
- aio_context_acquire(ctx);
- }
+ /* TODO removed in next patch */
}
void coroutine_fn bdrv_co_unlock(BlockDriverState *bs)
{
- AioContext *ctx = bdrv_get_aio_context(bs);
-
- assert(qemu_in_coroutine());
- if (ctx != qemu_get_aio_context()) {
- aio_context_release(ctx);
- }
+ /* TODO removed in next patch */
}
static void bdrv_do_remove_aio_context_notifier(BdrvAioNotifier *ban)
@@ -7728,21 +7559,8 @@ static void bdrv_set_aio_context_commit(void *opaque)
BdrvStateSetAioContext *state = (BdrvStateSetAioContext *) opaque;
BlockDriverState *bs = (BlockDriverState *) state->bs;
AioContext *new_context = state->new_ctx;
- AioContext *old_context = bdrv_get_aio_context(bs);
- /*
- * Take the old AioContex when detaching it from bs.
- * At this point, new_context lock is already acquired, and we are now
- * also taking old_context. This is safe as long as bdrv_detach_aio_context
- * does not call AIO_POLL_WHILE().
- */
- if (old_context != qemu_get_aio_context()) {
- aio_context_acquire(old_context);
- }
bdrv_detach_aio_context(bs);
- if (old_context != qemu_get_aio_context()) {
- aio_context_release(old_context);
- }
bdrv_attach_aio_context(bs, new_context);
}
@@ -7827,7 +7645,6 @@ int bdrv_try_change_aio_context(BlockDriverState *bs, AioContext *ctx,
Transaction *tran;
GHashTable *visited;
int ret;
- AioContext *old_context = bdrv_get_aio_context(bs);
GLOBAL_STATE_CODE();
/*
@@ -7857,34 +7674,7 @@ int bdrv_try_change_aio_context(BlockDriverState *bs, AioContext *ctx,
return -EPERM;
}
- /*
- * Release old AioContext, it won't be needed anymore, as all
- * bdrv_drained_begin() have been called already.
- */
- if (qemu_get_aio_context() != old_context) {
- aio_context_release(old_context);
- }
-
- /*
- * Acquire new AioContext since bdrv_drained_end() is going to be called
- * after we switched all nodes in the new AioContext, and the function
- * assumes that the lock of the bs is always taken.
- */
- if (qemu_get_aio_context() != ctx) {
- aio_context_acquire(ctx);
- }
-
tran_commit(tran);
-
- if (qemu_get_aio_context() != ctx) {
- aio_context_release(ctx);
- }
-
- /* Re-acquire the old AioContext, since the caller takes and releases it. */
- if (qemu_get_aio_context() != old_context) {
- aio_context_acquire(old_context);
- }
-
return 0;
}
@@ -8006,7 +7796,6 @@ BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
const char *node_name, Error **errp)
{
BlockDriverState *to_replace_bs = bdrv_find_node(node_name);
- AioContext *aio_context;
GLOBAL_STATE_CODE();
@@ -8015,12 +7804,8 @@ BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
return NULL;
}
- aio_context = bdrv_get_aio_context(to_replace_bs);
- aio_context_acquire(aio_context);
-
if (bdrv_op_is_blocked(to_replace_bs, BLOCK_OP_TYPE_REPLACE, errp)) {
- to_replace_bs = NULL;
- goto out;
+ return NULL;
}
/* We don't want arbitrary node of the BDS chain to be replaced only the top
@@ -8033,12 +7818,9 @@ BlockDriverState *check_to_replace_node(BlockDriverState *parent_bs,
"because it cannot be guaranteed that doing so would not "
"lead to an abrupt change of visible data",
node_name, parent_bs->node_name);
- to_replace_bs = NULL;
- goto out;
+ return NULL;
}
-out:
- aio_context_release(aio_context);
return to_replace_bs;
}
diff --git a/block/block-backend.c b/block/block-backend.c
index abac4e0235..f412bed274 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -429,7 +429,6 @@ BlockBackend *blk_new_open(const char *filename, const char *reference,
{
BlockBackend *blk;
BlockDriverState *bs;
- AioContext *ctx;
uint64_t perm = 0;
uint64_t shared = BLK_PERM_ALL;
@@ -459,23 +458,18 @@ BlockBackend *blk_new_open(const char *filename, const char *reference,
shared = BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE_UNCHANGED;
}
- aio_context_acquire(qemu_get_aio_context());
bs = bdrv_open(filename, reference, options, flags, errp);
- aio_context_release(qemu_get_aio_context());
if (!bs) {
return NULL;
}
/* bdrv_open() could have moved bs to a different AioContext */
- ctx = bdrv_get_aio_context(bs);
blk = blk_new(bdrv_get_aio_context(bs), perm, shared);
blk->perm = perm;
blk->shared_perm = shared;
- aio_context_acquire(ctx);
blk_insert_bs(blk, bs, errp);
bdrv_unref(bs);
- aio_context_release(ctx);
if (!blk->root) {
blk_unref(blk);
@@ -577,13 +571,9 @@ void blk_remove_all_bs(void)
GLOBAL_STATE_CODE();
while ((blk = blk_all_next(blk)) != NULL) {
- AioContext *ctx = blk_get_aio_context(blk);
-
- aio_context_acquire(ctx);
if (blk->root) {
blk_remove_bs(blk);
}
- aio_context_release(ctx);
}
}
@@ -2736,20 +2726,16 @@ int blk_commit_all(void)
GRAPH_RDLOCK_GUARD_MAINLOOP();
while ((blk = blk_all_next(blk)) != NULL) {
- AioContext *aio_context = blk_get_aio_context(blk);
BlockDriverState *unfiltered_bs = bdrv_skip_filters(blk_bs(blk));
- aio_context_acquire(aio_context);
if (blk_is_inserted(blk) && bdrv_cow_child(unfiltered_bs)) {
int ret;
ret = bdrv_commit(unfiltered_bs);
if (ret < 0) {
- aio_context_release(aio_context);
return ret;
}
}
- aio_context_release(aio_context);
}
return 0;
}
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index 13972879b1..0842a1a6df 100644
--- a/block/copy-before-write.c
+++ b/block/copy-before-write.c
@@ -412,7 +412,6 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
int64_t cluster_size;
g_autoptr(BlockdevOptions) full_opts = NULL;
BlockdevOptionsCbw *opts;
- AioContext *ctx;
int ret;
full_opts = cbw_parse_options(options, errp);
@@ -435,15 +434,11 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
GRAPH_RDLOCK_GUARD_MAINLOOP();
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
-
if (opts->bitmap) {
bitmap = block_dirty_bitmap_lookup(opts->bitmap->node,
opts->bitmap->name, NULL, errp);
if (!bitmap) {
- ret = -EINVAL;
- goto out;
+ return -EINVAL;
}
}
s->on_cbw_error = opts->has_on_cbw_error ? opts->on_cbw_error :
@@ -461,24 +456,21 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
s->bcs = block_copy_state_new(bs->file, s->target, bitmap, errp);
if (!s->bcs) {
error_prepend(errp, "Cannot create block-copy-state: ");
- ret = -EINVAL;
- goto out;
+ return -EINVAL;
}
cluster_size = block_copy_cluster_size(s->bcs);
s->done_bitmap = bdrv_create_dirty_bitmap(bs, cluster_size, NULL, errp);
if (!s->done_bitmap) {
- ret = -EINVAL;
- goto out;
+ return -EINVAL;
}
bdrv_disable_dirty_bitmap(s->done_bitmap);
/* s->access_bitmap starts equal to bcs bitmap */
s->access_bitmap = bdrv_create_dirty_bitmap(bs, cluster_size, NULL, errp);
if (!s->access_bitmap) {
- ret = -EINVAL;
- goto out;
+ return -EINVAL;
}
bdrv_disable_dirty_bitmap(s->access_bitmap);
bdrv_dirty_bitmap_merge_internal(s->access_bitmap,
@@ -487,11 +479,7 @@ static int cbw_open(BlockDriverState *bs, QDict *options, int flags,
qemu_co_mutex_init(&s->lock);
QLIST_INIT(&s->frozen_read_reqs);
-
- ret = 0;
-out:
- aio_context_release(ctx);
- return ret;
+ return 0;
}
static void cbw_close(BlockDriverState *bs)
diff --git a/block/export/export.c b/block/export/export.c
index a8f274e526..6d51ae8ed7 100644
--- a/block/export/export.c
+++ b/block/export/export.c
@@ -114,7 +114,6 @@ BlockExport *blk_exp_add(BlockExportOptions *export, Error **errp)
}
ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
if (export->iothread) {
IOThread *iothread;
@@ -133,8 +132,6 @@ BlockExport *blk_exp_add(BlockExportOptions *export, Error **errp)
set_context_errp = fixed_iothread ? errp : NULL;
ret = bdrv_try_change_aio_context(bs, new_ctx, NULL, set_context_errp);
if (ret == 0) {
- aio_context_release(ctx);
- aio_context_acquire(new_ctx);
ctx = new_ctx;
} else if (fixed_iothread) {
goto fail;
@@ -191,8 +188,6 @@ BlockExport *blk_exp_add(BlockExportOptions *export, Error **errp)
assert(exp->blk != NULL);
QLIST_INSERT_HEAD(&block_exports, exp, next);
-
- aio_context_release(ctx);
return exp;
fail:
@@ -200,7 +195,6 @@ fail:
blk_set_dev_ops(blk, NULL, NULL);
blk_unref(blk);
}
- aio_context_release(ctx);
if (exp) {
g_free(exp->id);
g_free(exp);
@@ -218,9 +212,6 @@ void blk_exp_ref(BlockExport *exp)
static void blk_exp_delete_bh(void *opaque)
{
BlockExport *exp = opaque;
- AioContext *aio_context = exp->ctx;
-
- aio_context_acquire(aio_context);
assert(exp->refcount == 0);
QLIST_REMOVE(exp, next);
@@ -230,8 +221,6 @@ static void blk_exp_delete_bh(void *opaque)
qapi_event_send_block_export_deleted(exp->id);
g_free(exp->id);
g_free(exp);
-
- aio_context_release(aio_context);
}
void blk_exp_unref(BlockExport *exp)
@@ -249,22 +238,16 @@ void blk_exp_unref(BlockExport *exp)
* connections and other internally held references start to shut down. When
* the function returns, there may still be active references while the export
* is in the process of shutting down.
- *
- * Acquires exp->ctx internally. Callers must *not* hold the lock.
*/
void blk_exp_request_shutdown(BlockExport *exp)
{
- AioContext *aio_context = exp->ctx;
-
- aio_context_acquire(aio_context);
-
/*
* If the user doesn't own the export any more, it is already shutting
* down. We must not call .request_shutdown and decrease the refcount a
* second time.
*/
if (!exp->user_owned) {
- goto out;
+ return;
}
exp->drv->request_shutdown(exp);
@@ -272,9 +255,6 @@ void blk_exp_request_shutdown(BlockExport *exp)
assert(exp->user_owned);
exp->user_owned = false;
blk_exp_unref(exp);
-
-out:
- aio_context_release(aio_context);
}
/*
diff --git a/block/io.c b/block/io.c
index 7e62fabbf5..8fa7670571 100644
--- a/block/io.c
+++ b/block/io.c
@@ -294,8 +294,6 @@ static void bdrv_co_drain_bh_cb(void *opaque)
BlockDriverState *bs = data->bs;
if (bs) {
- AioContext *ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
bdrv_dec_in_flight(bs);
if (data->begin) {
bdrv_do_drained_begin(bs, data->parent, data->poll);
@@ -303,7 +301,6 @@ static void bdrv_co_drain_bh_cb(void *opaque)
assert(!data->poll);
bdrv_do_drained_end(bs, data->parent);
}
- aio_context_release(ctx);
} else {
assert(data->begin);
bdrv_drain_all_begin();
@@ -320,8 +317,6 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
{
BdrvCoDrainData data;
Coroutine *self = qemu_coroutine_self();
- AioContext *ctx = bdrv_get_aio_context(bs);
- AioContext *co_ctx = qemu_coroutine_get_aio_context(self);
/* Calling bdrv_drain() from a BH ensures the current coroutine yields and
* other coroutines run if they were queued by aio_co_enter(). */
@@ -340,17 +335,6 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
bdrv_inc_in_flight(bs);
}
- /*
- * Temporarily drop the lock across yield or we would get deadlocks.
- * bdrv_co_drain_bh_cb() reaquires the lock as needed.
- *
- * When we yield below, the lock for the current context will be
- * released, so if this is actually the lock that protects bs, don't drop
- * it a second time.
- */
- if (ctx != co_ctx) {
- aio_context_release(ctx);
- }
replay_bh_schedule_oneshot_event(qemu_get_aio_context(),
bdrv_co_drain_bh_cb, &data);
@@ -358,11 +342,6 @@ static void coroutine_fn bdrv_co_yield_to_drain(BlockDriverState *bs,
/* If we are resumed from some other event (such as an aio completion or a
* timer callback), it is a bug in the caller that should be fixed. */
assert(data.done);
-
- /* Reacquire the AioContext of bs if we dropped it */
- if (ctx != co_ctx) {
- aio_context_acquire(ctx);
- }
}
static void bdrv_do_drained_begin(BlockDriverState *bs, BdrvChild *parent,
@@ -478,13 +457,12 @@ static bool bdrv_drain_all_poll(void)
GLOBAL_STATE_CODE();
GRAPH_RDLOCK_GUARD_MAINLOOP();
- /* bdrv_drain_poll() can't make changes to the graph and we are holding the
- * main AioContext lock, so iterating bdrv_next_all_states() is safe. */
+ /*
+ * bdrv_drain_poll() can't make changes to the graph and we hold the BQL,
+ * so iterating bdrv_next_all_states() is safe.
+ */
while ((bs = bdrv_next_all_states(bs))) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
result |= bdrv_drain_poll(bs, NULL, true);
- aio_context_release(aio_context);
}
return result;
@@ -525,11 +503,7 @@ void bdrv_drain_all_begin_nopoll(void)
/* Quiesce all nodes, without polling in-flight requests yet. The graph
* cannot change during this loop. */
while ((bs = bdrv_next_all_states(bs))) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
-
- aio_context_acquire(aio_context);
bdrv_do_drained_begin(bs, NULL, false);
- aio_context_release(aio_context);
}
}
@@ -588,11 +562,7 @@ void bdrv_drain_all_end(void)
}
while ((bs = bdrv_next_all_states(bs))) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
-
- aio_context_acquire(aio_context);
bdrv_do_drained_end(bs, NULL);
- aio_context_release(aio_context);
}
assert(qemu_get_current_aio_context() == qemu_get_aio_context());
@@ -2368,15 +2338,10 @@ int bdrv_flush_all(void)
}
for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
- AioContext *aio_context = bdrv_get_aio_context(bs);
- int ret;
-
- aio_context_acquire(aio_context);
- ret = bdrv_flush(bs);
+ int ret = bdrv_flush(bs);
if (ret < 0 && !result) {
result = ret;
}
- aio_context_release(aio_context);
}
return result;
diff --git a/block/mirror.c b/block/mirror.c
index 51f9e2f17c..5145eb53e1 100644
--- a/block/mirror.c
+++ b/block/mirror.c
@@ -662,7 +662,6 @@ static int mirror_exit_common(Job *job)
MirrorBlockJob *s = container_of(job, MirrorBlockJob, common.job);
BlockJob *bjob = &s->common;
MirrorBDSOpaque *bs_opaque;
- AioContext *replace_aio_context = NULL;
BlockDriverState *src;
BlockDriverState *target_bs;
BlockDriverState *mirror_top_bs;
@@ -677,7 +676,6 @@ static int mirror_exit_common(Job *job)
}
s->prepared = true;
- aio_context_acquire(qemu_get_aio_context());
bdrv_graph_rdlock_main_loop();
mirror_top_bs = s->mirror_top_bs;
@@ -742,11 +740,6 @@ static int mirror_exit_common(Job *job)
}
bdrv_graph_rdunlock_main_loop();
- if (s->to_replace) {
- replace_aio_context = bdrv_get_aio_context(s->to_replace);
- aio_context_acquire(replace_aio_context);
- }
-
if (s->should_complete && !abort) {
BlockDriverState *to_replace = s->to_replace ?: src;
bool ro = bdrv_is_read_only(to_replace);
@@ -785,9 +778,6 @@ static int mirror_exit_common(Job *job)
error_free(s->replace_blocker);
bdrv_unref(s->to_replace);
}
- if (replace_aio_context) {
- aio_context_release(replace_aio_context);
- }
g_free(s->replaces);
/*
@@ -811,8 +801,6 @@ static int mirror_exit_common(Job *job)
bdrv_unref(mirror_top_bs);
bdrv_unref(src);
- aio_context_release(qemu_get_aio_context());
-
return ret;
}
@@ -1191,24 +1179,17 @@ static void mirror_complete(Job *job, Error **errp)
/* block all operations on to_replace bs */
if (s->replaces) {
- AioContext *replace_aio_context;
-
s->to_replace = bdrv_find_node(s->replaces);
if (!s->to_replace) {
error_setg(errp, "Node name '%s' not found", s->replaces);
return;
}
- replace_aio_context = bdrv_get_aio_context(s->to_replace);
- aio_context_acquire(replace_aio_context);
-
/* TODO Translate this into child freeze system. */
error_setg(&s->replace_blocker,
"block device is in use by block-job-complete");
bdrv_op_block_all(s->to_replace, s->replace_blocker);
bdrv_ref(s->to_replace);
-
- aio_context_release(replace_aio_context);
}
s->should_complete = true;
diff --git a/block/monitor/bitmap-qmp-cmds.c b/block/monitor/bitmap-qmp-cmds.c
index 70d01a3776..a738e7bbf7 100644
--- a/block/monitor/bitmap-qmp-cmds.c
+++ b/block/monitor/bitmap-qmp-cmds.c
@@ -95,7 +95,6 @@ void qmp_block_dirty_bitmap_add(const char *node, const char *name,
{
BlockDriverState *bs;
BdrvDirtyBitmap *bitmap;
- AioContext *aio_context;
if (!name || name[0] == '\0') {
error_setg(errp, "Bitmap name cannot be empty");
@@ -107,14 +106,11 @@ void qmp_block_dirty_bitmap_add(const char *node, const char *name,
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
if (has_granularity) {
if (granularity < 512 || !is_power_of_2(granularity)) {
error_setg(errp, "Granularity must be power of 2 "
"and at least 512");
- goto out;
+ return;
}
} else {
/* Default to cluster size, if available: */
@@ -132,12 +128,12 @@ void qmp_block_dirty_bitmap_add(const char *node, const char *name,
if (persistent &&
!bdrv_can_store_new_dirty_bitmap(bs, name, granularity, errp))
{
- goto out;
+ return;
}
bitmap = bdrv_create_dirty_bitmap(bs, granularity, name, errp);
if (bitmap == NULL) {
- goto out;
+ return;
}
if (disabled) {
@@ -145,9 +141,6 @@ void qmp_block_dirty_bitmap_add(const char *node, const char *name,
}
bdrv_dirty_bitmap_set_persistence(bitmap, persistent);
-
-out:
- aio_context_release(aio_context);
}
BdrvDirtyBitmap *block_dirty_bitmap_remove(const char *node, const char *name,
@@ -157,7 +150,6 @@ BdrvDirtyBitmap *block_dirty_bitmap_remove(const char *node, const char *name,
{
BlockDriverState *bs;
BdrvDirtyBitmap *bitmap;
- AioContext *aio_context;
GLOBAL_STATE_CODE();
@@ -166,19 +158,14 @@ BdrvDirtyBitmap *block_dirty_bitmap_remove(const char *node, const char *name,
return NULL;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
if (bdrv_dirty_bitmap_check(bitmap, BDRV_BITMAP_BUSY | BDRV_BITMAP_RO,
errp)) {
- aio_context_release(aio_context);
return NULL;
}
if (bdrv_dirty_bitmap_get_persistence(bitmap) &&
bdrv_remove_persistent_dirty_bitmap(bs, name, errp) < 0)
{
- aio_context_release(aio_context);
return NULL;
}
@@ -190,7 +177,6 @@ BdrvDirtyBitmap *block_dirty_bitmap_remove(const char *node, const char *name,
*bitmap_bs = bs;
}
- aio_context_release(aio_context);
return release ? NULL : bitmap;
}
diff --git a/block/monitor/block-hmp-cmds.c b/block/monitor/block-hmp-cmds.c
index c729cbf1eb..bdbb5cb141 100644
--- a/block/monitor/block-hmp-cmds.c
+++ b/block/monitor/block-hmp-cmds.c
@@ -141,7 +141,6 @@ void hmp_drive_del(Monitor *mon, const QDict *qdict)
const char *id = qdict_get_str(qdict, "id");
BlockBackend *blk;
BlockDriverState *bs;
- AioContext *aio_context;
Error *local_err = NULL;
GLOBAL_STATE_CODE();
@@ -168,14 +167,10 @@ void hmp_drive_del(Monitor *mon, const QDict *qdict)
return;
}
- aio_context = blk_get_aio_context(blk);
- aio_context_acquire(aio_context);
-
bs = blk_bs(blk);
if (bs) {
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_DRIVE_DEL, &local_err)) {
error_report_err(local_err);
- aio_context_release(aio_context);
return;
}
@@ -196,8 +191,6 @@ void hmp_drive_del(Monitor *mon, const QDict *qdict)
} else {
blk_unref(blk);
}
-
- aio_context_release(aio_context);
}
void hmp_commit(Monitor *mon, const QDict *qdict)
@@ -213,7 +206,6 @@ void hmp_commit(Monitor *mon, const QDict *qdict)
ret = blk_commit_all();
} else {
BlockDriverState *bs;
- AioContext *aio_context;
blk = blk_by_name(device);
if (!blk) {
@@ -222,18 +214,13 @@ void hmp_commit(Monitor *mon, const QDict *qdict)
}
bs = bdrv_skip_implicit_filters(blk_bs(blk));
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
if (!blk_is_available(blk)) {
error_report("Device '%s' has no medium", device);
- aio_context_release(aio_context);
return;
}
ret = bdrv_commit(bs);
-
- aio_context_release(aio_context);
}
if (ret < 0) {
error_report("'commit' error for '%s': %s", device, strerror(-ret));
@@ -560,7 +547,6 @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
BlockBackend *blk = NULL;
BlockDriverState *bs = NULL;
BlockBackend *local_blk = NULL;
- AioContext *ctx = NULL;
bool qdev = qdict_get_try_bool(qdict, "qdev", false);
const char *device = qdict_get_str(qdict, "device");
const char *command = qdict_get_str(qdict, "command");
@@ -582,9 +568,6 @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
}
}
- ctx = blk ? blk_get_aio_context(blk) : bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
-
if (bs) {
blk = local_blk = blk_new(bdrv_get_aio_context(bs), 0, BLK_PERM_ALL);
ret = blk_insert_bs(blk, bs, &err);
@@ -622,11 +605,6 @@ void hmp_qemu_io(Monitor *mon, const QDict *qdict)
fail:
blk_unref(local_blk);
-
- if (ctx) {
- aio_context_release(ctx);
- }
-
hmp_handle_error(mon, err);
}
@@ -882,7 +860,6 @@ void hmp_info_snapshots(Monitor *mon, const QDict *qdict)
int nb_sns, i;
int total;
int *global_snapshots;
- AioContext *aio_context;
typedef struct SnapshotEntry {
QEMUSnapshotInfo sn;
@@ -909,11 +886,8 @@ void hmp_info_snapshots(Monitor *mon, const QDict *qdict)
error_report_err(err);
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
nb_sns = bdrv_snapshot_list(bs, &sn_tab);
- aio_context_release(aio_context);
if (nb_sns < 0) {
monitor_printf(mon, "bdrv_snapshot_list: error %d\n", nb_sns);
@@ -924,9 +898,7 @@ void hmp_info_snapshots(Monitor *mon, const QDict *qdict)
int bs1_nb_sns = 0;
ImageEntry *ie;
SnapshotEntry *se;
- AioContext *ctx = bdrv_get_aio_context(bs1);
- aio_context_acquire(ctx);
if (bdrv_can_snapshot(bs1)) {
sn = NULL;
bs1_nb_sns = bdrv_snapshot_list(bs1, &sn);
@@ -944,7 +916,6 @@ void hmp_info_snapshots(Monitor *mon, const QDict *qdict)
}
g_free(sn);
}
- aio_context_release(ctx);
}
if (no_snapshot) {
diff --git a/block/qapi-sysemu.c b/block/qapi-sysemu.c
index 1618cd225a..e4282631d2 100644
--- a/block/qapi-sysemu.c
+++ b/block/qapi-sysemu.c
@@ -174,7 +174,6 @@ blockdev_remove_medium(const char *device, const char *id, Error **errp)
{
BlockBackend *blk;
BlockDriverState *bs;
- AioContext *aio_context;
bool has_attached_device;
GLOBAL_STATE_CODE();
@@ -204,13 +203,10 @@ blockdev_remove_medium(const char *device, const char *id, Error **errp)
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
bdrv_graph_rdlock_main_loop();
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_EJECT, errp)) {
bdrv_graph_rdunlock_main_loop();
- goto out;
+ return;
}
bdrv_graph_rdunlock_main_loop();
@@ -223,9 +219,6 @@ blockdev_remove_medium(const char *device, const char *id, Error **errp)
* value passed here (i.e. false). */
blk_dev_change_media_cb(blk, false, &error_abort);
}
-
-out:
- aio_context_release(aio_context);
}
void qmp_blockdev_remove_medium(const char *id, Error **errp)
@@ -237,7 +230,6 @@ static void qmp_blockdev_insert_anon_medium(BlockBackend *blk,
BlockDriverState *bs, Error **errp)
{
Error *local_err = NULL;
- AioContext *ctx;
bool has_device;
int ret;
@@ -259,11 +251,7 @@ static void qmp_blockdev_insert_anon_medium(BlockBackend *blk,
return;
}
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
ret = blk_insert_bs(blk, bs, errp);
- aio_context_release(ctx);
-
if (ret < 0) {
return;
}
@@ -374,9 +362,7 @@ void qmp_blockdev_change_medium(const char *device,
qdict_put_str(options, "driver", format);
}
- aio_context_acquire(qemu_get_aio_context());
medium_bs = bdrv_open(filename, NULL, options, bdrv_flags, errp);
- aio_context_release(qemu_get_aio_context());
if (!medium_bs) {
goto fail;
@@ -437,20 +423,16 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp)
ThrottleConfig cfg;
BlockDriverState *bs;
BlockBackend *blk;
- AioContext *aio_context;
blk = qmp_get_blk(arg->device, arg->id, errp);
if (!blk) {
return;
}
- aio_context = blk_get_aio_context(blk);
- aio_context_acquire(aio_context);
-
bs = blk_bs(blk);
if (!bs) {
error_setg(errp, "Device has no medium");
- goto out;
+ return;
}
throttle_config_init(&cfg);
@@ -505,7 +487,7 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp)
}
if (!throttle_is_valid(&cfg, errp)) {
- goto out;
+ return;
}
if (throttle_enabled(&cfg)) {
@@ -522,9 +504,6 @@ void qmp_block_set_io_throttle(BlockIOThrottle *arg, Error **errp)
/* If all throttling settings are set to 0, disable I/O limits */
blk_io_limits_disable(blk);
}
-
-out:
- aio_context_release(aio_context);
}
void qmp_block_latency_histogram_set(
diff --git a/block/qapi.c b/block/qapi.c
index 82a30b38fe..9e806fa230 100644
--- a/block/qapi.c
+++ b/block/qapi.c
@@ -234,13 +234,11 @@ bdrv_do_query_node_info(BlockDriverState *bs, BlockNodeInfo *info, Error **errp)
int ret;
Error *err = NULL;
- aio_context_acquire(bdrv_get_aio_context(bs));
-
size = bdrv_getlength(bs);
if (size < 0) {
error_setg_errno(errp, -size, "Can't get image size '%s'",
bs->exact_filename);
- goto out;
+ return;
}
bdrv_refresh_filename(bs);
@@ -265,7 +263,7 @@ bdrv_do_query_node_info(BlockDriverState *bs, BlockNodeInfo *info, Error **errp)
info->format_specific = bdrv_get_specific_info(bs, &err);
if (err) {
error_propagate(errp, err);
- goto out;
+ return;
}
backing_filename = bs->backing_file;
if (backing_filename[0] != '\0') {
@@ -300,11 +298,8 @@ bdrv_do_query_node_info(BlockDriverState *bs, BlockNodeInfo *info, Error **errp)
break;
default:
error_propagate(errp, err);
- goto out;
+ return;
}
-
-out:
- aio_context_release(bdrv_get_aio_context(bs));
}
/**
@@ -709,15 +704,10 @@ BlockStatsList *qmp_query_blockstats(bool has_query_nodes,
/* Just to be safe if query_nodes is not always initialized */
if (has_query_nodes && query_nodes) {
for (bs = bdrv_next_node(NULL); bs; bs = bdrv_next_node(bs)) {
- AioContext *ctx = bdrv_get_aio_context(bs);
-
- aio_context_acquire(ctx);
QAPI_LIST_APPEND(tail, bdrv_query_bds_stats(bs, false));
- aio_context_release(ctx);
}
} else {
for (blk = blk_all_next(NULL); blk; blk = blk_all_next(blk)) {
- AioContext *ctx = blk_get_aio_context(blk);
BlockStats *s;
char *qdev;
@@ -725,7 +715,6 @@ BlockStatsList *qmp_query_blockstats(bool has_query_nodes,
continue;
}
- aio_context_acquire(ctx);
s = bdrv_query_bds_stats(blk_bs(blk), true);
s->device = g_strdup(blk_name(blk));
@@ -737,7 +726,6 @@ BlockStatsList *qmp_query_blockstats(bool has_query_nodes,
}
bdrv_query_blk_stats(s->stats, blk);
- aio_context_release(ctx);
QAPI_LIST_APPEND(tail, s);
}
diff --git a/block/raw-format.c b/block/raw-format.c
index 1111dffd54..ac7e8495f6 100644
--- a/block/raw-format.c
+++ b/block/raw-format.c
@@ -470,7 +470,6 @@ static int raw_open(BlockDriverState *bs, QDict *options, int flags,
Error **errp)
{
BDRVRawState *s = bs->opaque;
- AioContext *ctx;
bool has_size;
uint64_t offset, size;
BdrvChildRole file_role;
@@ -522,11 +521,7 @@ static int raw_open(BlockDriverState *bs, QDict *options, int flags,
bs->file->bs->filename);
}
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
ret = raw_apply_options(bs, s, offset, has_size, size, errp);
- aio_context_release(ctx);
-
if (ret < 0) {
return ret;
}
diff --git a/block/replication.c b/block/replication.c
index 424b537ff7..ca6bd0a720 100644
--- a/block/replication.c
+++ b/block/replication.c
@@ -394,14 +394,7 @@ static void reopen_backing_file(BlockDriverState *bs, bool writable,
}
if (reopen_queue) {
- AioContext *ctx = bdrv_get_aio_context(bs);
- if (ctx != qemu_get_aio_context()) {
- aio_context_release(ctx);
- }
bdrv_reopen_multiple(reopen_queue, errp);
- if (ctx != qemu_get_aio_context()) {
- aio_context_acquire(ctx);
- }
}
}
@@ -462,14 +455,11 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
BlockDriverState *top_bs;
BdrvChild *active_disk, *hidden_disk, *secondary_disk;
int64_t active_length, hidden_length, disk_length;
- AioContext *aio_context;
Error *local_err = NULL;
BackupPerf perf = { .use_copy_range = true, .max_workers = 1 };
GLOBAL_STATE_CODE();
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
s = bs->opaque;
if (s->stage == BLOCK_REPLICATION_DONE ||
@@ -479,20 +469,17 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
* Ignore the request because the secondary side of replication
* doesn't have to do anything anymore.
*/
- aio_context_release(aio_context);
return;
}
if (s->stage != BLOCK_REPLICATION_NONE) {
error_setg(errp, "Block replication is running or done");
- aio_context_release(aio_context);
return;
}
if (s->mode != mode) {
error_setg(errp, "The parameter mode's value is invalid, needs %d,"
" but got %d", s->mode, mode);
- aio_context_release(aio_context);
return;
}
@@ -505,7 +492,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (!active_disk || !active_disk->bs || !active_disk->bs->backing) {
error_setg(errp, "Active disk doesn't have backing file");
bdrv_graph_rdunlock_main_loop();
- aio_context_release(aio_context);
return;
}
@@ -513,7 +499,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (!hidden_disk->bs || !hidden_disk->bs->backing) {
error_setg(errp, "Hidden disk doesn't have backing file");
bdrv_graph_rdunlock_main_loop();
- aio_context_release(aio_context);
return;
}
@@ -521,7 +506,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (!secondary_disk->bs || !bdrv_has_blk(secondary_disk->bs)) {
error_setg(errp, "The secondary disk doesn't have block backend");
bdrv_graph_rdunlock_main_loop();
- aio_context_release(aio_context);
return;
}
bdrv_graph_rdunlock_main_loop();
@@ -534,7 +518,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
active_length != hidden_length || hidden_length != disk_length) {
error_setg(errp, "Active disk, hidden disk, secondary disk's length"
" are not the same");
- aio_context_release(aio_context);
return;
}
@@ -546,7 +529,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
!hidden_disk->bs->drv->bdrv_make_empty) {
error_setg(errp,
"Active disk or hidden disk doesn't support make_empty");
- aio_context_release(aio_context);
bdrv_graph_rdunlock_main_loop();
return;
}
@@ -556,7 +538,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
reopen_backing_file(bs, true, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- aio_context_release(aio_context);
return;
}
@@ -569,7 +550,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (local_err) {
error_propagate(errp, local_err);
bdrv_graph_wrunlock();
- aio_context_release(aio_context);
return;
}
@@ -580,7 +560,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (local_err) {
error_propagate(errp, local_err);
bdrv_graph_wrunlock();
- aio_context_release(aio_context);
return;
}
@@ -594,7 +573,6 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
error_setg(errp, "No top_bs or it is invalid");
bdrv_graph_wrunlock();
reopen_backing_file(bs, false, NULL);
- aio_context_release(aio_context);
return;
}
bdrv_op_block_all(top_bs, s->blocker);
@@ -612,13 +590,11 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
if (local_err) {
error_propagate(errp, local_err);
backup_job_cleanup(bs);
- aio_context_release(aio_context);
return;
}
job_start(&s->backup_job->job);
break;
default:
- aio_context_release(aio_context);
abort();
}
@@ -629,18 +605,12 @@ static void replication_start(ReplicationState *rs, ReplicationMode mode,
}
s->error = 0;
- aio_context_release(aio_context);
}
static void replication_do_checkpoint(ReplicationState *rs, Error **errp)
{
BlockDriverState *bs = rs->opaque;
- BDRVReplicationState *s;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
- s = bs->opaque;
+ BDRVReplicationState *s = bs->opaque;
if (s->stage == BLOCK_REPLICATION_DONE ||
s->stage == BLOCK_REPLICATION_FAILOVER) {
@@ -649,38 +619,28 @@ static void replication_do_checkpoint(ReplicationState *rs, Error **errp)
* Ignore the request because the secondary side of replication
* doesn't have to do anything anymore.
*/
- aio_context_release(aio_context);
return;
}
if (s->mode == REPLICATION_MODE_SECONDARY) {
secondary_do_checkpoint(bs, errp);
}
- aio_context_release(aio_context);
}
static void replication_get_error(ReplicationState *rs, Error **errp)
{
BlockDriverState *bs = rs->opaque;
- BDRVReplicationState *s;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
- s = bs->opaque;
+ BDRVReplicationState *s = bs->opaque;
if (s->stage == BLOCK_REPLICATION_NONE) {
error_setg(errp, "Block replication is not running");
- aio_context_release(aio_context);
return;
}
if (s->error) {
error_setg(errp, "I/O error occurred");
- aio_context_release(aio_context);
return;
}
- aio_context_release(aio_context);
}
static void replication_done(void *opaque, int ret)
@@ -708,12 +668,7 @@ static void replication_done(void *opaque, int ret)
static void replication_stop(ReplicationState *rs, bool failover, Error **errp)
{
BlockDriverState *bs = rs->opaque;
- BDRVReplicationState *s;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
- s = bs->opaque;
+ BDRVReplicationState *s = bs->opaque;
if (s->stage == BLOCK_REPLICATION_DONE ||
s->stage == BLOCK_REPLICATION_FAILOVER) {
@@ -722,13 +677,11 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp)
* Ignore the request because the secondary side of replication
* doesn't have to do anything anymore.
*/
- aio_context_release(aio_context);
return;
}
if (s->stage != BLOCK_REPLICATION_RUNNING) {
error_setg(errp, "Block replication is not running");
- aio_context_release(aio_context);
return;
}
@@ -744,15 +697,12 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp)
* disk, secondary disk in backup_job_completed().
*/
if (s->backup_job) {
- aio_context_release(aio_context);
job_cancel_sync(&s->backup_job->job, true);
- aio_context_acquire(aio_context);
}
if (!failover) {
secondary_do_checkpoint(bs, errp);
s->stage = BLOCK_REPLICATION_DONE;
- aio_context_release(aio_context);
return;
}
@@ -765,10 +715,8 @@ static void replication_stop(ReplicationState *rs, bool failover, Error **errp)
bdrv_graph_rdunlock_main_loop();
break;
default:
- aio_context_release(aio_context);
abort();
}
- aio_context_release(aio_context);
}
static const char *const replication_strong_runtime_opts[] = {
diff --git a/block/snapshot.c b/block/snapshot.c
index e486d3e205..a28f2b039f 100644
--- a/block/snapshot.c
+++ b/block/snapshot.c
@@ -525,9 +525,7 @@ static bool GRAPH_RDLOCK bdrv_all_snapshots_includes_bs(BlockDriverState *bs)
return bdrv_has_blk(bs) || QLIST_EMPTY(&bs->parents);
}
-/* Group operations. All block drivers are involved.
- * These functions will properly handle dataplane (take aio_context_acquire
- * when appropriate for appropriate block drivers) */
+/* Group operations. All block drivers are involved. */
bool bdrv_all_can_snapshot(bool has_devices, strList *devices,
Error **errp)
@@ -545,14 +543,11 @@ bool bdrv_all_can_snapshot(bool has_devices, strList *devices,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
bool ok = true;
- aio_context_acquire(ctx);
if (devices || bdrv_all_snapshots_includes_bs(bs)) {
ok = bdrv_can_snapshot(bs);
}
- aio_context_release(ctx);
if (!ok) {
error_setg(errp, "Device '%s' is writable but does not support "
"snapshots", bdrv_get_device_or_node_name(bs));
@@ -582,18 +577,15 @@ int bdrv_all_delete_snapshot(const char *name,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
QEMUSnapshotInfo sn1, *snapshot = &sn1;
int ret = 0;
- aio_context_acquire(ctx);
if ((devices || bdrv_all_snapshots_includes_bs(bs)) &&
bdrv_snapshot_find(bs, snapshot, name) >= 0)
{
ret = bdrv_snapshot_delete(bs, snapshot->id_str,
snapshot->name, errp);
}
- aio_context_release(ctx);
if (ret < 0) {
error_prepend(errp, "Could not delete snapshot '%s' on '%s': ",
name, bdrv_get_device_or_node_name(bs));
@@ -628,17 +620,14 @@ int bdrv_all_goto_snapshot(const char *name,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
bool all_snapshots_includes_bs;
- aio_context_acquire(ctx);
bdrv_graph_rdlock_main_loop();
all_snapshots_includes_bs = bdrv_all_snapshots_includes_bs(bs);
bdrv_graph_rdunlock_main_loop();
ret = (devices || all_snapshots_includes_bs) ?
bdrv_snapshot_goto(bs, name, errp) : 0;
- aio_context_release(ctx);
if (ret < 0) {
bdrv_graph_rdlock_main_loop();
error_prepend(errp, "Could not load snapshot '%s' on '%s': ",
@@ -670,15 +659,12 @@ int bdrv_all_has_snapshot(const char *name,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
QEMUSnapshotInfo sn;
int ret = 0;
- aio_context_acquire(ctx);
if (devices || bdrv_all_snapshots_includes_bs(bs)) {
ret = bdrv_snapshot_find(bs, &sn, name);
}
- aio_context_release(ctx);
if (ret < 0) {
if (ret == -ENOENT) {
return 0;
@@ -715,10 +701,8 @@ int bdrv_all_create_snapshot(QEMUSnapshotInfo *sn,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
int ret = 0;
- aio_context_acquire(ctx);
if (bs == vm_state_bs) {
sn->vm_state_size = vm_state_size;
ret = bdrv_snapshot_create(bs, sn);
@@ -726,7 +710,6 @@ int bdrv_all_create_snapshot(QEMUSnapshotInfo *sn,
sn->vm_state_size = 0;
ret = bdrv_snapshot_create(bs, sn);
}
- aio_context_release(ctx);
if (ret < 0) {
error_setg(errp, "Could not create snapshot '%s' on '%s'",
sn->name, bdrv_get_device_or_node_name(bs));
@@ -757,13 +740,10 @@ BlockDriverState *bdrv_all_find_vmstate_bs(const char *vmstate_bs,
iterbdrvs = bdrvs;
while (iterbdrvs) {
BlockDriverState *bs = iterbdrvs->data;
- AioContext *ctx = bdrv_get_aio_context(bs);
bool found = false;
- aio_context_acquire(ctx);
found = (devices || bdrv_all_snapshots_includes_bs(bs)) &&
bdrv_can_snapshot(bs);
- aio_context_release(ctx);
if (vmstate_bs) {
if (g_str_equal(vmstate_bs,
diff --git a/block/write-threshold.c b/block/write-threshold.c
index 76d8885677..56fe88de81 100644
--- a/block/write-threshold.c
+++ b/block/write-threshold.c
@@ -33,7 +33,6 @@ void qmp_block_set_write_threshold(const char *node_name,
Error **errp)
{
BlockDriverState *bs;
- AioContext *aio_context;
bs = bdrv_find_node(node_name);
if (!bs) {
@@ -41,12 +40,7 @@ void qmp_block_set_write_threshold(const char *node_name,
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
bdrv_write_threshold_set(bs, threshold_bytes);
-
- aio_context_release(aio_context);
}
void bdrv_write_threshold_check_write(BlockDriverState *bs, int64_t offset,
diff --git a/blockdev.c b/blockdev.c
index db9cc96510..8a1b28f830 100644
--- a/blockdev.c
+++ b/blockdev.c
@@ -662,7 +662,6 @@ err_no_opts:
/* Takes the ownership of bs_opts */
BlockDriverState *bds_tree_init(QDict *bs_opts, Error **errp)
{
- BlockDriverState *bs;
int bdrv_flags = 0;
GLOBAL_STATE_CODE();
@@ -677,11 +676,7 @@ BlockDriverState *bds_tree_init(QDict *bs_opts, Error **errp)
bdrv_flags |= BDRV_O_INACTIVE;
}
- aio_context_acquire(qemu_get_aio_context());
- bs = bdrv_open(NULL, NULL, bs_opts, bdrv_flags, errp);
- aio_context_release(qemu_get_aio_context());
-
- return bs;
+ return bdrv_open(NULL, NULL, bs_opts, bdrv_flags, errp);
}
void blockdev_close_all_bdrv_states(void)
@@ -690,11 +685,7 @@ void blockdev_close_all_bdrv_states(void)
GLOBAL_STATE_CODE();
QTAILQ_FOREACH_SAFE(bs, &monitor_bdrv_states, monitor_list, next_bs) {
- AioContext *ctx = bdrv_get_aio_context(bs);
-
- aio_context_acquire(ctx);
bdrv_unref(bs);
- aio_context_release(ctx);
}
}
@@ -1048,7 +1039,6 @@ fail:
static BlockDriverState *qmp_get_root_bs(const char *name, Error **errp)
{
BlockDriverState *bs;
- AioContext *aio_context;
GRAPH_RDLOCK_GUARD_MAINLOOP();
@@ -1062,16 +1052,11 @@ static BlockDriverState *qmp_get_root_bs(const char *name, Error **errp)
return NULL;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
if (!bdrv_is_inserted(bs)) {
error_setg(errp, "Device has no medium");
bs = NULL;
}
- aio_context_release(aio_context);
-
return bs;
}
@@ -1141,7 +1126,6 @@ SnapshotInfo *qmp_blockdev_snapshot_delete_internal_sync(const char *device,
Error **errp)
{
BlockDriverState *bs;
- AioContext *aio_context;
QEMUSnapshotInfo sn;
Error *local_err = NULL;
SnapshotInfo *info = NULL;
@@ -1154,39 +1138,35 @@ SnapshotInfo *qmp_blockdev_snapshot_delete_internal_sync(const char *device,
if (!bs) {
return NULL;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
if (!id && !name) {
error_setg(errp, "Name or id must be provided");
- goto out_aio_context;
+ return NULL;
}
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_INTERNAL_SNAPSHOT_DELETE, errp)) {
- goto out_aio_context;
+ return NULL;
}
ret = bdrv_snapshot_find_by_id_and_name(bs, id, name, &sn, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- goto out_aio_context;
+ return NULL;
}
if (!ret) {
error_setg(errp,
"Snapshot with id '%s' and name '%s' does not exist on "
"device '%s'",
STR_OR_NULL(id), STR_OR_NULL(name), device);
- goto out_aio_context;
+ return NULL;
}
bdrv_snapshot_delete(bs, id, name, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- goto out_aio_context;
+ return NULL;
}
- aio_context_release(aio_context);
-
info = g_new0(SnapshotInfo, 1);
info->id = g_strdup(sn.id_str);
info->name = g_strdup(sn.name);
@@ -1201,10 +1181,6 @@ SnapshotInfo *qmp_blockdev_snapshot_delete_internal_sync(const char *device,
}
return info;
-
-out_aio_context:
- aio_context_release(aio_context);
- return NULL;
}
/* internal snapshot private data */
@@ -1232,7 +1208,6 @@ static void internal_snapshot_action(BlockdevSnapshotInternal *internal,
bool ret;
int64_t rt;
InternalSnapshotState *state = g_new0(InternalSnapshotState, 1);
- AioContext *aio_context;
int ret1;
GLOBAL_STATE_CODE();
@@ -1248,33 +1223,30 @@ static void internal_snapshot_action(BlockdevSnapshotInternal *internal,
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
state->bs = bs;
/* Paired with .clean() */
bdrv_drained_begin(bs);
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_INTERNAL_SNAPSHOT, errp)) {
- goto out;
+ return;
}
if (bdrv_is_read_only(bs)) {
error_setg(errp, "Device '%s' is read only", device);
- goto out;
+ return;
}
if (!bdrv_can_snapshot(bs)) {
error_setg(errp, "Block format '%s' used by device '%s' "
"does not support internal snapshots",
bs->drv->format_name, device);
- goto out;
+ return;
}
if (!strlen(name)) {
error_setg(errp, "Name is empty");
- goto out;
+ return;
}
/* check whether a snapshot with name exist */
@@ -1282,12 +1254,12 @@ static void internal_snapshot_action(BlockdevSnapshotInternal *internal,
&local_err);
if (local_err) {
error_propagate(errp, local_err);
- goto out;
+ return;
} else if (ret) {
error_setg(errp,
"Snapshot with name '%s' already exists on device '%s'",
name, device);
- goto out;
+ return;
}
/* 3. take the snapshot */
@@ -1308,14 +1280,11 @@ static void internal_snapshot_action(BlockdevSnapshotInternal *internal,
error_setg_errno(errp, -ret1,
"Failed to create snapshot '%s' on device '%s'",
name, device);
- goto out;
+ return;
}
/* 4. succeed, mark a snapshot is created */
state->created = true;
-
-out:
- aio_context_release(aio_context);
}
static void internal_snapshot_abort(void *opaque)
@@ -1323,7 +1292,6 @@ static void internal_snapshot_abort(void *opaque)
InternalSnapshotState *state = opaque;
BlockDriverState *bs = state->bs;
QEMUSnapshotInfo *sn = &state->sn;
- AioContext *aio_context;
Error *local_error = NULL;
GLOBAL_STATE_CODE();
@@ -1333,9 +1301,6 @@ static void internal_snapshot_abort(void *opaque)
return;
}
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
-
if (bdrv_snapshot_delete(bs, sn->id_str, sn->name, &local_error) < 0) {
error_reportf_err(local_error,
"Failed to delete snapshot with id '%s' and "
@@ -1343,25 +1308,17 @@ static void internal_snapshot_abort(void *opaque)
sn->id_str, sn->name,
bdrv_get_device_name(bs));
}
-
- aio_context_release(aio_context);
}
static void internal_snapshot_clean(void *opaque)
{
g_autofree InternalSnapshotState *state = opaque;
- AioContext *aio_context;
if (!state->bs) {
return;
}
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
-
bdrv_drained_end(state->bs);
-
- aio_context_release(aio_context);
}
/* external snapshot private data */
@@ -1395,7 +1352,6 @@ static void external_snapshot_action(TransactionAction *action,
/* File name of the new image (for 'blockdev-snapshot-sync') */
const char *new_image_file;
ExternalSnapshotState *state = g_new0(ExternalSnapshotState, 1);
- AioContext *aio_context;
uint64_t perm, shared;
/* TODO We'll eventually have to take a writer lock in this function */
@@ -1435,26 +1391,23 @@ static void external_snapshot_action(TransactionAction *action,
return;
}
- aio_context = bdrv_get_aio_context(state->old_bs);
- aio_context_acquire(aio_context);
-
/* Paired with .clean() */
bdrv_drained_begin(state->old_bs);
if (!bdrv_is_inserted(state->old_bs)) {
error_setg(errp, QERR_DEVICE_HAS_NO_MEDIUM, device);
- goto out;
+ return;
}
if (bdrv_op_is_blocked(state->old_bs,
BLOCK_OP_TYPE_EXTERNAL_SNAPSHOT, errp)) {
- goto out;
+ return;
}
if (!bdrv_is_read_only(state->old_bs)) {
if (bdrv_flush(state->old_bs)) {
error_setg(errp, QERR_IO_ERROR);
- goto out;
+ return;
}
}
@@ -1466,13 +1419,13 @@ static void external_snapshot_action(TransactionAction *action,
if (node_name && !snapshot_node_name) {
error_setg(errp, "New overlay node-name missing");
- goto out;
+ return;
}
if (snapshot_node_name &&
bdrv_lookup_bs(snapshot_node_name, snapshot_node_name, NULL)) {
error_setg(errp, "New overlay node-name already in use");
- goto out;
+ return;
}
flags = state->old_bs->open_flags;
@@ -1485,20 +1438,18 @@ static void external_snapshot_action(TransactionAction *action,
int64_t size = bdrv_getlength(state->old_bs);
if (size < 0) {
error_setg_errno(errp, -size, "bdrv_getlength failed");
- goto out;
+ return;
}
bdrv_refresh_filename(state->old_bs);
- aio_context_release(aio_context);
bdrv_img_create(new_image_file, format,
state->old_bs->filename,
state->old_bs->drv->format_name,
NULL, size, flags, false, &local_err);
- aio_context_acquire(aio_context);
if (local_err) {
error_propagate(errp, local_err);
- goto out;
+ return;
}
}
@@ -1508,20 +1459,15 @@ static void external_snapshot_action(TransactionAction *action,
}
qdict_put_str(options, "driver", format);
}
- aio_context_release(aio_context);
- aio_context_acquire(qemu_get_aio_context());
state->new_bs = bdrv_open(new_image_file, snapshot_ref, options, flags,
errp);
- aio_context_release(qemu_get_aio_context());
/* We will manually add the backing_hd field to the bs later */
if (!state->new_bs) {
return;
}
- aio_context_acquire(aio_context);
-
/*
* Allow attaching a backing file to an overlay that's already in use only
* if the parents don't assume that they are already seeing a valid image.
@@ -1530,41 +1476,34 @@ static void external_snapshot_action(TransactionAction *action,
bdrv_get_cumulative_perm(state->new_bs, &perm, &shared);
if (perm & BLK_PERM_CONSISTENT_READ) {
error_setg(errp, "The overlay is already in use");
- goto out;
+ return;
}
if (state->new_bs->drv->is_filter) {
error_setg(errp, "Filters cannot be used as overlays");
- goto out;
+ return;
}
if (bdrv_cow_child(state->new_bs)) {
error_setg(errp, "The overlay already has a backing image");
- goto out;
+ return;
}
if (!state->new_bs->drv->supports_backing) {
error_setg(errp, "The overlay does not support backing images");
- goto out;
+ return;
}
ret = bdrv_append(state->new_bs, state->old_bs, errp);
if (ret < 0) {
- goto out;
+ return;
}
state->overlay_appended = true;
-
-out:
- aio_context_release(aio_context);
}
static void external_snapshot_commit(void *opaque)
{
ExternalSnapshotState *state = opaque;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(state->old_bs);
- aio_context_acquire(aio_context);
/* We don't need (or want) to use the transactional
* bdrv_reopen_multiple() across all the entries at once, because we
@@ -1572,8 +1511,6 @@ static void external_snapshot_commit(void *opaque)
if (!qatomic_read(&state->old_bs->copy_on_read)) {
bdrv_reopen_set_read_only(state->old_bs, true, NULL);
}
-
- aio_context_release(aio_context);
}
static void external_snapshot_abort(void *opaque)
@@ -1586,7 +1523,6 @@ static void external_snapshot_abort(void *opaque)
int ret;
aio_context = bdrv_get_aio_context(state->old_bs);
- aio_context_acquire(aio_context);
bdrv_ref(state->old_bs); /* we can't let bdrv_set_backind_hd()
close state->old_bs; we need it */
@@ -1599,15 +1535,9 @@ static void external_snapshot_abort(void *opaque)
*/
tmp_context = bdrv_get_aio_context(state->old_bs);
if (aio_context != tmp_context) {
- aio_context_release(aio_context);
- aio_context_acquire(tmp_context);
-
ret = bdrv_try_change_aio_context(state->old_bs,
aio_context, NULL, NULL);
assert(ret == 0);
-
- aio_context_release(tmp_context);
- aio_context_acquire(aio_context);
}
bdrv_drained_begin(state->new_bs);
@@ -1617,8 +1547,6 @@ static void external_snapshot_abort(void *opaque)
bdrv_drained_end(state->new_bs);
bdrv_unref(state->old_bs); /* bdrv_replace_node() ref'ed old_bs */
-
- aio_context_release(aio_context);
}
}
}
@@ -1626,19 +1554,13 @@ static void external_snapshot_abort(void *opaque)
static void external_snapshot_clean(void *opaque)
{
g_autofree ExternalSnapshotState *state = opaque;
- AioContext *aio_context;
if (!state->old_bs) {
return;
}
- aio_context = bdrv_get_aio_context(state->old_bs);
- aio_context_acquire(aio_context);
-
bdrv_drained_end(state->old_bs);
bdrv_unref(state->new_bs);
-
- aio_context_release(aio_context);
}
typedef struct DriveBackupState {
@@ -1670,7 +1592,6 @@ static void drive_backup_action(DriveBackup *backup,
BlockDriverState *target_bs;
BlockDriverState *source = NULL;
AioContext *aio_context;
- AioContext *old_context;
const char *format;
QDict *options;
Error *local_err = NULL;
@@ -1698,7 +1619,6 @@ static void drive_backup_action(DriveBackup *backup,
}
aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
state->bs = bs;
/* Paired with .clean() */
@@ -1713,7 +1633,7 @@ static void drive_backup_action(DriveBackup *backup,
bdrv_graph_rdlock_main_loop();
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp)) {
bdrv_graph_rdunlock_main_loop();
- goto out;
+ return;
}
flags = bs->open_flags | BDRV_O_RDWR;
@@ -1744,7 +1664,7 @@ static void drive_backup_action(DriveBackup *backup,
size = bdrv_getlength(bs);
if (size < 0) {
error_setg_errno(errp, -size, "bdrv_getlength failed");
- goto out;
+ return;
}
if (backup->mode != NEW_IMAGE_MODE_EXISTING) {
@@ -1770,7 +1690,7 @@ static void drive_backup_action(DriveBackup *backup,
if (local_err) {
error_propagate(errp, local_err);
- goto out;
+ return;
}
options = qdict_new();
@@ -1779,30 +1699,18 @@ static void drive_backup_action(DriveBackup *backup,
if (format) {
qdict_put_str(options, "driver", format);
}
- aio_context_release(aio_context);
- aio_context_acquire(qemu_get_aio_context());
target_bs = bdrv_open(backup->target, NULL, options, flags, errp);
- aio_context_release(qemu_get_aio_context());
-
if (!target_bs) {
return;
}
- /* Honor bdrv_try_change_aio_context() context acquisition requirements. */
- old_context = bdrv_get_aio_context(target_bs);
- aio_context_acquire(old_context);
-
ret = bdrv_try_change_aio_context(target_bs, aio_context, NULL, errp);
if (ret < 0) {
bdrv_unref(target_bs);
- aio_context_release(old_context);
return;
}
- aio_context_release(old_context);
- aio_context_acquire(aio_context);
-
if (set_backing_hd) {
if (bdrv_set_backing_hd(target_bs, source, errp) < 0) {
goto unref;
@@ -1815,22 +1723,14 @@ static void drive_backup_action(DriveBackup *backup,
unref:
bdrv_unref(target_bs);
-out:
- aio_context_release(aio_context);
}
static void drive_backup_commit(void *opaque)
{
DriveBackupState *state = opaque;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
assert(state->job);
job_start(&state->job->job);
-
- aio_context_release(aio_context);
}
static void drive_backup_abort(void *opaque)
@@ -1845,18 +1745,12 @@ static void drive_backup_abort(void *opaque)
static void drive_backup_clean(void *opaque)
{
g_autofree DriveBackupState *state = opaque;
- AioContext *aio_context;
if (!state->bs) {
return;
}
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
-
bdrv_drained_end(state->bs);
-
- aio_context_release(aio_context);
}
typedef struct BlockdevBackupState {
@@ -1881,7 +1775,6 @@ static void blockdev_backup_action(BlockdevBackup *backup,
BlockDriverState *bs;
BlockDriverState *target_bs;
AioContext *aio_context;
- AioContext *old_context;
int ret;
tran_add(tran, &blockdev_backup_drv, state);
@@ -1898,17 +1791,12 @@ static void blockdev_backup_action(BlockdevBackup *backup,
/* Honor bdrv_try_change_aio_context() context acquisition requirements. */
aio_context = bdrv_get_aio_context(bs);
- old_context = bdrv_get_aio_context(target_bs);
- aio_context_acquire(old_context);
ret = bdrv_try_change_aio_context(target_bs, aio_context, NULL, errp);
if (ret < 0) {
- aio_context_release(old_context);
return;
}
- aio_context_release(old_context);
- aio_context_acquire(aio_context);
state->bs = bs;
/* Paired with .clean() */
@@ -1917,22 +1805,14 @@ static void blockdev_backup_action(BlockdevBackup *backup,
state->job = do_backup_common(qapi_BlockdevBackup_base(backup),
bs, target_bs, aio_context,
block_job_txn, errp);
-
- aio_context_release(aio_context);
}
static void blockdev_backup_commit(void *opaque)
{
BlockdevBackupState *state = opaque;
- AioContext *aio_context;
-
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
assert(state->job);
job_start(&state->job->job);
-
- aio_context_release(aio_context);
}
static void blockdev_backup_abort(void *opaque)
@@ -1947,18 +1827,12 @@ static void blockdev_backup_abort(void *opaque)
static void blockdev_backup_clean(void *opaque)
{
g_autofree BlockdevBackupState *state = opaque;
- AioContext *aio_context;
if (!state->bs) {
return;
}
- aio_context = bdrv_get_aio_context(state->bs);
- aio_context_acquire(aio_context);
-
bdrv_drained_end(state->bs);
-
- aio_context_release(aio_context);
}
typedef struct BlockDirtyBitmapState {
@@ -2453,7 +2327,6 @@ void qmp_block_stream(const char *job_id, const char *device,
}
aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
bdrv_graph_rdlock_main_loop();
if (base) {
@@ -2520,7 +2393,7 @@ void qmp_block_stream(const char *job_id, const char *device,
if (!base_bs && backing_file) {
error_setg(errp, "backing file specified, but streaming the "
"entire chain");
- goto out;
+ return;
}
if (has_auto_finalize && !auto_finalize) {
@@ -2535,18 +2408,14 @@ void qmp_block_stream(const char *job_id, const char *device,
filter_node_name, &local_err);
if (local_err) {
error_propagate(errp, local_err);
- goto out;
+ return;
}
trace_qmp_block_stream(bs);
-
-out:
- aio_context_release(aio_context);
return;
out_rdlock:
bdrv_graph_rdunlock_main_loop();
- aio_context_release(aio_context);
}
void qmp_block_commit(const char *job_id, const char *device,
@@ -2605,10 +2474,9 @@ void qmp_block_commit(const char *job_id, const char *device,
}
aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_COMMIT_SOURCE, errp)) {
- goto out;
+ return;
}
/* default top_bs is the active layer */
@@ -2616,16 +2484,16 @@ void qmp_block_commit(const char *job_id, const char *device,
if (top_node && top) {
error_setg(errp, "'top-node' and 'top' are mutually exclusive");
- goto out;
+ return;
} else if (top_node) {
top_bs = bdrv_lookup_bs(NULL, top_node, errp);
if (top_bs == NULL) {
- goto out;
+ return;
}
if (!bdrv_chain_contains(bs, top_bs)) {
error_setg(errp, "'%s' is not in this backing file chain",
top_node);
- goto out;
+ return;
}
} else if (top) {
/* This strcmp() is just a shortcut, there is no need to
@@ -2639,35 +2507,35 @@ void qmp_block_commit(const char *job_id, const char *device,
if (top_bs == NULL) {
error_setg(errp, "Top image file %s not found", top ? top : "NULL");
- goto out;
+ return;
}
assert(bdrv_get_aio_context(top_bs) == aio_context);
if (base_node && base) {
error_setg(errp, "'base-node' and 'base' are mutually exclusive");
- goto out;
+ return;
} else if (base_node) {
base_bs = bdrv_lookup_bs(NULL, base_node, errp);
if (base_bs == NULL) {
- goto out;
+ return;
}
if (!bdrv_chain_contains(top_bs, base_bs)) {
error_setg(errp, "'%s' is not in this backing file chain",
base_node);
- goto out;
+ return;
}
} else if (base) {
base_bs = bdrv_find_backing_image(top_bs, base);
if (base_bs == NULL) {
error_setg(errp, "Can't find '%s' in the backing chain", base);
- goto out;
+ return;
}
} else {
base_bs = bdrv_find_base(top_bs);
if (base_bs == NULL) {
error_setg(errp, "There is no backimg image");
- goto out;
+ return;
}
}
@@ -2677,14 +2545,14 @@ void qmp_block_commit(const char *job_id, const char *device,
iter = bdrv_filter_or_cow_bs(iter))
{
if (bdrv_op_is_blocked(iter, BLOCK_OP_TYPE_COMMIT_TARGET, errp)) {
- goto out;
+ return;
}
}
/* Do not allow attempts to commit an image into itself */
if (top_bs == base_bs) {
error_setg(errp, "cannot commit an image into itself");
- goto out;
+ return;
}
/*
@@ -2707,7 +2575,7 @@ void qmp_block_commit(const char *job_id, const char *device,
error_setg(errp, "'backing-file' specified, but 'top' has a "
"writer on it");
}
- goto out;
+ return;
}
if (!job_id) {
/*
@@ -2723,7 +2591,7 @@ void qmp_block_commit(const char *job_id, const char *device,
} else {
BlockDriverState *overlay_bs = bdrv_find_overlay(bs, top_bs);
if (bdrv_op_is_blocked(overlay_bs, BLOCK_OP_TYPE_COMMIT_TARGET, errp)) {
- goto out;
+ return;
}
commit_start(job_id, bs, base_bs, top_bs, job_flags,
speed, on_error, backing_file,
@@ -2731,11 +2599,8 @@ void qmp_block_commit(const char *job_id, const char *device,
}
if (local_err != NULL) {
error_propagate(errp, local_err);
- goto out;
+ return;
}
-
-out:
- aio_context_release(aio_context);
}
/* Common QMP interface for drive-backup and blockdev-backup */
@@ -2984,8 +2849,6 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
if (replaces) {
BlockDriverState *to_replace_bs;
- AioContext *aio_context;
- AioContext *replace_aio_context;
int64_t bs_size, replace_size;
bs_size = bdrv_getlength(bs);
@@ -2999,19 +2862,7 @@ static void blockdev_mirror_common(const char *job_id, BlockDriverState *bs,
return;
}
- aio_context = bdrv_get_aio_context(bs);
- replace_aio_context = bdrv_get_aio_context(to_replace_bs);
- /*
- * bdrv_getlength() is a co-wrapper and uses AIO_WAIT_WHILE. Be sure not
- * to acquire the same AioContext twice.
- */
- if (replace_aio_context != aio_context) {
- aio_context_acquire(replace_aio_context);
- }
replace_size = bdrv_getlength(to_replace_bs);
- if (replace_aio_context != aio_context) {
- aio_context_release(replace_aio_context);
- }
if (replace_size < 0) {
error_setg_errno(errp, -replace_size,
@@ -3040,7 +2891,6 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
BlockDriverState *bs;
BlockDriverState *target_backing_bs, *target_bs;
AioContext *aio_context;
- AioContext *old_context;
BlockMirrorBackingMode backing_mode;
Error *local_err = NULL;
QDict *options = NULL;
@@ -3063,7 +2913,6 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
}
aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
if (!arg->has_mode) {
arg->mode = NEW_IMAGE_MODE_ABSOLUTE_PATHS;
@@ -3087,14 +2936,14 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
size = bdrv_getlength(bs);
if (size < 0) {
error_setg_errno(errp, -size, "bdrv_getlength failed");
- goto out;
+ return;
}
if (arg->replaces) {
if (!arg->node_name) {
error_setg(errp, "a node-name must be provided when replacing a"
" named node of the graph");
- goto out;
+ return;
}
}
@@ -3142,7 +2991,7 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
if (local_err) {
error_propagate(errp, local_err);
- goto out;
+ return;
}
options = qdict_new();
@@ -3152,15 +3001,11 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
if (format) {
qdict_put_str(options, "driver", format);
}
- aio_context_release(aio_context);
/* Mirroring takes care of copy-on-write using the source's backing
* file.
*/
- aio_context_acquire(qemu_get_aio_context());
target_bs = bdrv_open(arg->target, NULL, options, flags, errp);
- aio_context_release(qemu_get_aio_context());
-
if (!target_bs) {
return;
}
@@ -3172,20 +3017,12 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
bdrv_graph_rdunlock_main_loop();
- /* Honor bdrv_try_change_aio_context() context acquisition requirements. */
- old_context = bdrv_get_aio_context(target_bs);
- aio_context_acquire(old_context);
-
ret = bdrv_try_change_aio_context(target_bs, aio_context, NULL, errp);
if (ret < 0) {
bdrv_unref(target_bs);
- aio_context_release(old_context);
return;
}
- aio_context_release(old_context);
- aio_context_acquire(aio_context);
-
blockdev_mirror_common(arg->job_id, bs, target_bs,
arg->replaces, arg->sync,
backing_mode, zero_target,
@@ -3201,8 +3038,6 @@ void qmp_drive_mirror(DriveMirror *arg, Error **errp)
arg->has_auto_dismiss, arg->auto_dismiss,
errp);
bdrv_unref(target_bs);
-out:
- aio_context_release(aio_context);
}
void qmp_blockdev_mirror(const char *job_id,
@@ -3225,7 +3060,6 @@ void qmp_blockdev_mirror(const char *job_id,
BlockDriverState *bs;
BlockDriverState *target_bs;
AioContext *aio_context;
- AioContext *old_context;
BlockMirrorBackingMode backing_mode = MIRROR_LEAVE_BACKING_CHAIN;
bool zero_target;
int ret;
@@ -3242,18 +3076,11 @@ void qmp_blockdev_mirror(const char *job_id,
zero_target = (sync == MIRROR_SYNC_MODE_FULL);
- /* Honor bdrv_try_change_aio_context() context acquisition requirements. */
- old_context = bdrv_get_aio_context(target_bs);
aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(old_context);
ret = bdrv_try_change_aio_context(target_bs, aio_context, NULL, errp);
-
- aio_context_release(old_context);
- aio_context_acquire(aio_context);
-
if (ret < 0) {
- goto out;
+ return;
}
blockdev_mirror_common(job_id, bs, target_bs,
@@ -3268,8 +3095,6 @@ void qmp_blockdev_mirror(const char *job_id,
has_auto_finalize, auto_finalize,
has_auto_dismiss, auto_dismiss,
errp);
-out:
- aio_context_release(aio_context);
}
/*
@@ -3432,7 +3257,6 @@ void qmp_change_backing_file(const char *device,
Error **errp)
{
BlockDriverState *bs = NULL;
- AioContext *aio_context;
BlockDriverState *image_bs = NULL;
Error *local_err = NULL;
bool ro;
@@ -3443,9 +3267,6 @@ void qmp_change_backing_file(const char *device,
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
-
bdrv_graph_rdlock_main_loop();
image_bs = bdrv_lookup_bs(NULL, image_node_name, &local_err);
@@ -3484,7 +3305,7 @@ void qmp_change_backing_file(const char *device,
if (ro) {
if (bdrv_reopen_set_read_only(image_bs, false, errp) != 0) {
- goto out;
+ return;
}
}
@@ -3502,14 +3323,10 @@ void qmp_change_backing_file(const char *device,
if (ro) {
bdrv_reopen_set_read_only(image_bs, true, errp);
}
-
-out:
- aio_context_release(aio_context);
return;
out_rdlock:
bdrv_graph_rdunlock_main_loop();
- aio_context_release(aio_context);
}
void qmp_blockdev_add(BlockdevOptions *options, Error **errp)
@@ -3549,7 +3366,6 @@ void qmp_blockdev_reopen(BlockdevOptionsList *reopen_list, Error **errp)
for (; reopen_list != NULL; reopen_list = reopen_list->next) {
BlockdevOptions *options = reopen_list->value;
BlockDriverState *bs;
- AioContext *ctx;
QObject *obj;
Visitor *v;
QDict *qdict;
@@ -3577,12 +3393,7 @@ void qmp_blockdev_reopen(BlockdevOptionsList *reopen_list, Error **errp)
qdict_flatten(qdict);
- ctx = bdrv_get_aio_context(bs);
- aio_context_acquire(ctx);
-
queue = bdrv_reopen_queue(queue, bs, qdict, false);
-
- aio_context_release(ctx);
}
/* Perform the reopen operation */
@@ -3595,7 +3406,6 @@ fail:
void qmp_blockdev_del(const char *node_name, Error **errp)
{
- AioContext *aio_context;
BlockDriverState *bs;
GLOBAL_STATE_CODE();
@@ -3610,30 +3420,25 @@ void qmp_blockdev_del(const char *node_name, Error **errp)
error_setg(errp, "Node %s is in use", node_name);
return;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_DRIVE_DEL, errp)) {
- goto out;
+ return;
}
if (!QTAILQ_IN_USE(bs, monitor_list)) {
error_setg(errp, "Node %s is not owned by the monitor",
bs->node_name);
- goto out;
+ return;
}
if (bs->refcnt > 1) {
error_setg(errp, "Block device %s is in use",
bdrv_get_device_or_node_name(bs));
- goto out;
+ return;
}
QTAILQ_REMOVE(&monitor_bdrv_states, bs, monitor_list);
bdrv_unref(bs);
-
-out:
- aio_context_release(aio_context);
}
static BdrvChild * GRAPH_RDLOCK
@@ -3723,7 +3528,6 @@ BlockJobInfoList *qmp_query_block_jobs(Error **errp)
void qmp_x_blockdev_set_iothread(const char *node_name, StrOrNull *iothread,
bool has_force, bool force, Error **errp)
{
- AioContext *old_context;
AioContext *new_context;
BlockDriverState *bs;
@@ -3755,12 +3559,7 @@ void qmp_x_blockdev_set_iothread(const char *node_name, StrOrNull *iothread,
new_context = qemu_get_aio_context();
}
- old_context = bdrv_get_aio_context(bs);
- aio_context_acquire(old_context);
-
bdrv_try_change_aio_context(bs, new_context, NULL, errp);
-
- aio_context_release(old_context);
}
QemuOptsList qemu_common_drive_opts = {
diff --git a/blockjob.c b/blockjob.c
index 7310412313..d5f29e14af 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -198,9 +198,7 @@ void block_job_remove_all_bdrv(BlockJob *job)
* one to make sure that such a concurrent access does not attempt
* to process an already freed BdrvChild.
*/
- aio_context_release(job->job.aio_context);
bdrv_graph_wrlock();
- aio_context_acquire(job->job.aio_context);
while (job->nodes) {
GSList *l = job->nodes;
BdrvChild *c = l->data;
@@ -234,28 +232,12 @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs,
uint64_t perm, uint64_t shared_perm, Error **errp)
{
BdrvChild *c;
- AioContext *ctx = bdrv_get_aio_context(bs);
- bool need_context_ops;
GLOBAL_STATE_CODE();
bdrv_ref(bs);
- need_context_ops = ctx != job->job.aio_context;
-
- if (need_context_ops) {
- if (job->job.aio_context != qemu_get_aio_context()) {
- aio_context_release(job->job.aio_context);
- }
- aio_context_acquire(ctx);
- }
c = bdrv_root_attach_child(bs, name, &child_job, 0, perm, shared_perm, job,
errp);
- if (need_context_ops) {
- aio_context_release(ctx);
- if (job->job.aio_context != qemu_get_aio_context()) {
- aio_context_acquire(job->job.aio_context);
- }
- }
if (c == NULL) {
return -EPERM;
}
diff --git a/hw/block/dataplane/virtio-blk.c b/hw/block/dataplane/virtio-blk.c
index f83bb0f116..7bbbd981ad 100644
--- a/hw/block/dataplane/virtio-blk.c
+++ b/hw/block/dataplane/virtio-blk.c
@@ -124,7 +124,6 @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
VirtIOBlockDataPlane *s = vblk->dataplane;
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vblk)));
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
- AioContext *old_context;
unsigned i;
unsigned nvqs = s->conf->num_queues;
Error *local_err = NULL;
@@ -178,10 +177,7 @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
trace_virtio_blk_data_plane_start(s);
- old_context = blk_get_aio_context(s->conf->conf.blk);
- aio_context_acquire(old_context);
r = blk_set_aio_context(s->conf->conf.blk, s->ctx, &local_err);
- aio_context_release(old_context);
if (r < 0) {
error_report_err(local_err);
goto fail_aio_context;
@@ -208,13 +204,11 @@ int virtio_blk_data_plane_start(VirtIODevice *vdev)
/* Get this show started by hooking up our callbacks */
if (!blk_in_drain(s->conf->conf.blk)) {
- aio_context_acquire(s->ctx);
for (i = 0; i < nvqs; i++) {
VirtQueue *vq = virtio_get_queue(s->vdev, i);
virtio_queue_aio_attach_host_notifier(vq, s->ctx);
}
- aio_context_release(s->ctx);
}
return 0;
@@ -314,8 +308,6 @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
*/
vblk->dataplane_started = false;
- aio_context_acquire(s->ctx);
-
/* Wait for virtio_blk_dma_restart_bh() and in flight I/O to complete */
blk_drain(s->conf->conf.blk);
@@ -325,8 +317,6 @@ void virtio_blk_data_plane_stop(VirtIODevice *vdev)
*/
blk_set_aio_context(s->conf->conf.blk, qemu_get_aio_context(), NULL);
- aio_context_release(s->ctx);
-
/* Clean up guest notifier (irq) */
k->set_guest_notifiers(qbus->parent, nvqs, false);
diff --git a/hw/block/dataplane/xen-block.c b/hw/block/dataplane/xen-block.c
index c4bb28c66f..98501e6885 100644
--- a/hw/block/dataplane/xen-block.c
+++ b/hw/block/dataplane/xen-block.c
@@ -260,8 +260,6 @@ static void xen_block_complete_aio(void *opaque, int ret)
XenBlockRequest *request = opaque;
XenBlockDataPlane *dataplane = request->dataplane;
- aio_context_acquire(dataplane->ctx);
-
if (ret != 0) {
error_report("%s I/O error",
request->req.operation == BLKIF_OP_READ ?
@@ -273,10 +271,10 @@ static void xen_block_complete_aio(void *opaque, int ret)
if (request->presync) {
request->presync = 0;
xen_block_do_aio(request);
- goto done;
+ return;
}
if (request->aio_inflight > 0) {
- goto done;
+ return;
}
switch (request->req.operation) {
@@ -318,9 +316,6 @@ static void xen_block_complete_aio(void *opaque, int ret)
if (dataplane->more_work) {
qemu_bh_schedule(dataplane->bh);
}
-
-done:
- aio_context_release(dataplane->ctx);
}
static bool xen_block_split_discard(XenBlockRequest *request,
@@ -601,9 +596,7 @@ static void xen_block_dataplane_bh(void *opaque)
{
XenBlockDataPlane *dataplane = opaque;
- aio_context_acquire(dataplane->ctx);
xen_block_handle_requests(dataplane);
- aio_context_release(dataplane->ctx);
}
static bool xen_block_dataplane_event(void *opaque)
@@ -703,10 +696,8 @@ void xen_block_dataplane_stop(XenBlockDataPlane *dataplane)
xen_block_dataplane_detach(dataplane);
}
- aio_context_acquire(dataplane->ctx);
/* Xen doesn't have multiple users for nodes, so this can't fail */
blk_set_aio_context(dataplane->blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(dataplane->ctx);
/*
* Now that the context has been moved onto the main thread, cancel
@@ -752,7 +743,6 @@ void xen_block_dataplane_start(XenBlockDataPlane *dataplane,
{
ERRP_GUARD();
XenDevice *xendev = dataplane->xendev;
- AioContext *old_context;
unsigned int ring_size;
unsigned int i;
@@ -836,11 +826,8 @@ void xen_block_dataplane_start(XenBlockDataPlane *dataplane,
goto stop;
}
- old_context = blk_get_aio_context(dataplane->blk);
- aio_context_acquire(old_context);
/* If other users keep the BlockBackend in the iothread, that's ok */
blk_set_aio_context(dataplane->blk, dataplane->ctx, NULL);
- aio_context_release(old_context);
if (!blk_in_drain(dataplane->blk)) {
xen_block_dataplane_attach(dataplane);
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index a1f8e15522..5e49c0625f 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -102,7 +102,6 @@ static void virtio_blk_rw_complete(void *opaque, int ret)
VirtIOBlock *s = next->dev;
VirtIODevice *vdev = VIRTIO_DEVICE(s);
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
while (next) {
VirtIOBlockReq *req = next;
next = req->mr_next;
@@ -135,7 +134,6 @@ static void virtio_blk_rw_complete(void *opaque, int ret)
block_acct_done(blk_get_stats(s->blk), &req->acct);
virtio_blk_free_request(req);
}
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
}
static void virtio_blk_flush_complete(void *opaque, int ret)
@@ -143,19 +141,15 @@ static void virtio_blk_flush_complete(void *opaque, int ret)
VirtIOBlockReq *req = opaque;
VirtIOBlock *s = req->dev;
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
if (ret) {
if (virtio_blk_handle_rw_error(req, -ret, 0, true)) {
- goto out;
+ return;
}
}
virtio_blk_req_complete(req, VIRTIO_BLK_S_OK);
block_acct_done(blk_get_stats(s->blk), &req->acct);
virtio_blk_free_request(req);
-
-out:
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
}
static void virtio_blk_discard_write_zeroes_complete(void *opaque, int ret)
@@ -165,10 +159,9 @@ static void virtio_blk_discard_write_zeroes_complete(void *opaque, int ret)
bool is_write_zeroes = (virtio_ldl_p(VIRTIO_DEVICE(s), &req->out.type) &
~VIRTIO_BLK_T_BARRIER) == VIRTIO_BLK_T_WRITE_ZEROES;
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
if (ret) {
if (virtio_blk_handle_rw_error(req, -ret, false, is_write_zeroes)) {
- goto out;
+ return;
}
}
@@ -177,9 +170,6 @@ static void virtio_blk_discard_write_zeroes_complete(void *opaque, int ret)
block_acct_done(blk_get_stats(s->blk), &req->acct);
}
virtio_blk_free_request(req);
-
-out:
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
}
#ifdef __linux__
@@ -226,10 +216,8 @@ static void virtio_blk_ioctl_complete(void *opaque, int status)
virtio_stl_p(vdev, &scsi->data_len, hdr->dxfer_len);
out:
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
virtio_blk_req_complete(req, status);
virtio_blk_free_request(req);
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
g_free(ioctl_req);
}
@@ -669,7 +657,6 @@ static void virtio_blk_zone_report_complete(void *opaque, int ret)
{
ZoneCmdData *data = opaque;
VirtIOBlockReq *req = data->req;
- VirtIOBlock *s = req->dev;
VirtIODevice *vdev = VIRTIO_DEVICE(req->dev);
struct iovec *in_iov = data->in_iov;
unsigned in_num = data->in_num;
@@ -760,10 +747,8 @@ static void virtio_blk_zone_report_complete(void *opaque, int ret)
}
out:
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
virtio_blk_req_complete(req, err_status);
virtio_blk_free_request(req);
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
g_free(data->zone_report_data.zones);
g_free(data);
}
@@ -826,10 +811,8 @@ static void virtio_blk_zone_mgmt_complete(void *opaque, int ret)
err_status = VIRTIO_BLK_S_ZONE_INVALID_CMD;
}
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
virtio_blk_req_complete(req, err_status);
virtio_blk_free_request(req);
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
}
static int virtio_blk_handle_zone_mgmt(VirtIOBlockReq *req, BlockZoneOp op)
@@ -879,7 +862,6 @@ static void virtio_blk_zone_append_complete(void *opaque, int ret)
{
ZoneCmdData *data = opaque;
VirtIOBlockReq *req = data->req;
- VirtIOBlock *s = req->dev;
VirtIODevice *vdev = VIRTIO_DEVICE(req->dev);
int64_t append_sector, n;
uint8_t err_status = VIRTIO_BLK_S_OK;
@@ -902,10 +884,8 @@ static void virtio_blk_zone_append_complete(void *opaque, int ret)
trace_virtio_blk_zone_append_complete(vdev, req, append_sector, ret);
out:
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
virtio_blk_req_complete(req, err_status);
virtio_blk_free_request(req);
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
g_free(data);
}
@@ -941,10 +921,8 @@ static int virtio_blk_handle_zone_append(VirtIOBlockReq *req,
return 0;
out:
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
virtio_blk_req_complete(req, err_status);
virtio_blk_free_request(req);
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
return err_status;
}
@@ -1134,7 +1112,6 @@ void virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
MultiReqBuffer mrb = {};
bool suppress_notifications = virtio_queue_get_notification(vq);
- aio_context_acquire(blk_get_aio_context(s->blk));
defer_call_begin();
do {
@@ -1160,7 +1137,6 @@ void virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
}
defer_call_end();
- aio_context_release(blk_get_aio_context(s->blk));
}
static void virtio_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
@@ -1188,7 +1164,6 @@ static void virtio_blk_dma_restart_bh(void *opaque)
s->rq = NULL;
- aio_context_acquire(blk_get_aio_context(s->conf.conf.blk));
while (req) {
VirtIOBlockReq *next = req->next;
if (virtio_blk_handle_request(req, &mrb)) {
@@ -1212,8 +1187,6 @@ static void virtio_blk_dma_restart_bh(void *opaque)
/* Paired with inc in virtio_blk_dma_restart_cb() */
blk_dec_in_flight(s->conf.conf.blk);
-
- aio_context_release(blk_get_aio_context(s->conf.conf.blk));
}
static void virtio_blk_dma_restart_cb(void *opaque, bool running,
@@ -1235,11 +1208,8 @@ static void virtio_blk_dma_restart_cb(void *opaque, bool running,
static void virtio_blk_reset(VirtIODevice *vdev)
{
VirtIOBlock *s = VIRTIO_BLK(vdev);
- AioContext *ctx;
VirtIOBlockReq *req;
- ctx = blk_get_aio_context(s->blk);
- aio_context_acquire(ctx);
blk_drain(s->blk);
/* We drop queued requests after blk_drain() because blk_drain() itself can
@@ -1251,8 +1221,6 @@ static void virtio_blk_reset(VirtIODevice *vdev)
virtio_blk_free_request(req);
}
- aio_context_release(ctx);
-
assert(!s->dataplane_started);
blk_set_enable_write_cache(s->blk, s->original_wce);
}
@@ -1268,10 +1236,6 @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
uint64_t capacity;
int64_t length;
int blk_size = conf->logical_block_size;
- AioContext *ctx;
-
- ctx = blk_get_aio_context(s->blk);
- aio_context_acquire(ctx);
blk_get_geometry(s->blk, &capacity);
memset(&blkcfg, 0, sizeof(blkcfg));
@@ -1295,7 +1259,6 @@ static void virtio_blk_update_config(VirtIODevice *vdev, uint8_t *config)
* per track (cylinder).
*/
length = blk_getlength(s->blk);
- aio_context_release(ctx);
if (length > 0 && length / conf->heads / conf->secs % blk_size) {
blkcfg.geometry.sectors = conf->secs & ~s->sector_mask;
} else {
@@ -1362,9 +1325,7 @@ static void virtio_blk_set_config(VirtIODevice *vdev, const uint8_t *config)
memcpy(&blkcfg, config, s->config_size);
- aio_context_acquire(blk_get_aio_context(s->blk));
blk_set_enable_write_cache(s->blk, blkcfg.wce != 0);
- aio_context_release(blk_get_aio_context(s->blk));
}
static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
@@ -1432,11 +1393,9 @@ static void virtio_blk_set_status(VirtIODevice *vdev, uint8_t status)
* s->blk would erroneously be placed in writethrough mode.
*/
if (!virtio_vdev_has_feature(vdev, VIRTIO_BLK_F_CONFIG_WCE)) {
- aio_context_acquire(blk_get_aio_context(s->blk));
blk_set_enable_write_cache(s->blk,
virtio_vdev_has_feature(vdev,
VIRTIO_BLK_F_WCE));
- aio_context_release(blk_get_aio_context(s->blk));
}
}
diff --git a/hw/core/qdev-properties-system.c b/hw/core/qdev-properties-system.c
index 1473ab3d5e..73cced4626 100644
--- a/hw/core/qdev-properties-system.c
+++ b/hw/core/qdev-properties-system.c
@@ -120,9 +120,7 @@ static void set_drive_helper(Object *obj, Visitor *v, const char *name,
"node");
}
- aio_context_acquire(ctx);
blk_replace_bs(blk, bs, errp);
- aio_context_release(ctx);
return;
}
@@ -148,10 +146,7 @@ static void set_drive_helper(Object *obj, Visitor *v, const char *name,
0, BLK_PERM_ALL);
blk_created = true;
- aio_context_acquire(ctx);
ret = blk_insert_bs(blk, bs, errp);
- aio_context_release(ctx);
-
if (ret < 0) {
goto fail;
}
@@ -207,12 +202,8 @@ static void release_drive(Object *obj, const char *name, void *opaque)
BlockBackend **ptr = object_field_prop_ptr(obj, prop);
if (*ptr) {
- AioContext *ctx = blk_get_aio_context(*ptr);
-
- aio_context_acquire(ctx);
blockdev_auto_del(*ptr);
blk_detach_dev(*ptr, dev);
- aio_context_release(ctx);
}
}
diff --git a/job.c b/job.c
index 99a2e54b54..660ce22c56 100644
--- a/job.c
+++ b/job.c
@@ -464,12 +464,8 @@ void job_unref_locked(Job *job)
assert(!job->txn);
if (job->driver->free) {
- AioContext *aio_context = job->aio_context;
job_unlock();
- /* FIXME: aiocontext lock is required because cb calls blk_unref */
- aio_context_acquire(aio_context);
job->driver->free(job);
- aio_context_release(aio_context);
job_lock();
}
@@ -840,12 +836,10 @@ static void job_clean(Job *job)
/*
* Called with job_mutex held, but releases it temporarily.
- * Takes AioContext lock internally to invoke a job->driver callback.
*/
static int job_finalize_single_locked(Job *job)
{
int job_ret;
- AioContext *ctx = job->aio_context;
assert(job_is_completed_locked(job));
@@ -854,7 +848,6 @@ static int job_finalize_single_locked(Job *job)
job_ret = job->ret;
job_unlock();
- aio_context_acquire(ctx);
if (!job_ret) {
job_commit(job);
@@ -867,7 +860,6 @@ static int job_finalize_single_locked(Job *job)
job->cb(job->opaque, job_ret);
}
- aio_context_release(ctx);
job_lock();
/* Emit events only if we actually started */
@@ -886,17 +878,13 @@ static int job_finalize_single_locked(Job *job)
/*
* Called with job_mutex held, but releases it temporarily.
- * Takes AioContext lock internally to invoke a job->driver callback.
*/
static void job_cancel_async_locked(Job *job, bool force)
{
- AioContext *ctx = job->aio_context;
GLOBAL_STATE_CODE();
if (job->driver->cancel) {
job_unlock();
- aio_context_acquire(ctx);
force = job->driver->cancel(job, force);
- aio_context_release(ctx);
job_lock();
} else {
/* No .cancel() means the job will behave as if force-cancelled */
@@ -931,7 +919,6 @@ static void job_cancel_async_locked(Job *job, bool force)
/*
* Called with job_mutex held, but releases it temporarily.
- * Takes AioContext lock internally to invoke a job->driver callback.
*/
static void job_completed_txn_abort_locked(Job *job)
{
@@ -979,15 +966,12 @@ static void job_completed_txn_abort_locked(Job *job)
static int job_prepare_locked(Job *job)
{
int ret;
- AioContext *ctx = job->aio_context;
GLOBAL_STATE_CODE();
if (job->ret == 0 && job->driver->prepare) {
job_unlock();
- aio_context_acquire(ctx);
ret = job->driver->prepare(job);
- aio_context_release(ctx);
job_lock();
job->ret = ret;
job_update_rc_locked(job);
diff --git a/migration/block.c b/migration/block.c
index a15f9bddcb..2bcfcbfdf6 100644
--- a/migration/block.c
+++ b/migration/block.c
@@ -66,7 +66,7 @@ typedef struct BlkMigDevState {
/* Protected by block migration lock. */
int64_t completed_sectors;
- /* During migration this is protected by iothread lock / AioContext.
+ /* During migration this is protected by bdrv_dirty_bitmap_lock().
* Allocation and free happen during setup and cleanup respectively.
*/
BdrvDirtyBitmap *dirty_bitmap;
@@ -101,7 +101,7 @@ typedef struct BlkMigState {
int prev_progress;
int bulk_completed;
- /* Lock must be taken _inside_ the iothread lock and any AioContexts. */
+ /* Lock must be taken _inside_ the iothread lock. */
QemuMutex lock;
} BlkMigState;
@@ -270,7 +270,6 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
if (bmds->shared_base) {
qemu_mutex_lock_iothread();
- aio_context_acquire(blk_get_aio_context(bb));
/* Skip unallocated sectors; intentionally treats failure or
* partial sector as an allocated sector */
while (cur_sector < total_sectors &&
@@ -281,7 +280,6 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
}
cur_sector += count >> BDRV_SECTOR_BITS;
}
- aio_context_release(blk_get_aio_context(bb));
qemu_mutex_unlock_iothread();
}
@@ -313,22 +311,10 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
block_mig_state.submitted++;
blk_mig_unlock();
- /* We do not know if bs is under the main thread (and thus does
- * not acquire the AioContext when doing AIO) or rather under
- * dataplane. Thus acquire both the iothread mutex and the
- * AioContext.
- *
- * This is ugly and will disappear when we make bdrv_* thread-safe,
- * without the need to acquire the AioContext.
- */
- qemu_mutex_lock_iothread();
- aio_context_acquire(blk_get_aio_context(bmds->blk));
bdrv_reset_dirty_bitmap(bmds->dirty_bitmap, cur_sector * BDRV_SECTOR_SIZE,
nr_sectors * BDRV_SECTOR_SIZE);
blk->aiocb = blk_aio_preadv(bb, cur_sector * BDRV_SECTOR_SIZE, &blk->qiov,
0, blk_mig_read_cb, blk);
- aio_context_release(blk_get_aio_context(bmds->blk));
- qemu_mutex_unlock_iothread();
bmds->cur_sector = cur_sector + nr_sectors;
return (bmds->cur_sector >= total_sectors);
@@ -512,7 +498,7 @@ static void blk_mig_reset_dirty_cursor(void)
}
}
-/* Called with iothread lock and AioContext taken. */
+/* Called with iothread lock taken. */
static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds,
int is_async)
@@ -606,9 +592,7 @@ static int blk_mig_save_dirty_block(QEMUFile *f, int is_async)
int ret = 1;
QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
- aio_context_acquire(blk_get_aio_context(bmds->blk));
ret = mig_save_device_dirty(f, bmds, is_async);
- aio_context_release(blk_get_aio_context(bmds->blk));
if (ret <= 0) {
break;
}
@@ -666,9 +650,9 @@ static int64_t get_remaining_dirty(void)
int64_t dirty = 0;
QSIMPLEQ_FOREACH(bmds, &block_mig_state.bmds_list, entry) {
- aio_context_acquire(blk_get_aio_context(bmds->blk));
+ bdrv_dirty_bitmap_lock(bmds->dirty_bitmap);
dirty += bdrv_get_dirty_count(bmds->dirty_bitmap);
- aio_context_release(blk_get_aio_context(bmds->blk));
+ bdrv_dirty_bitmap_unlock(bmds->dirty_bitmap);
}
return dirty;
@@ -681,7 +665,6 @@ static void block_migration_cleanup_bmds(void)
{
BlkMigDevState *bmds;
BlockDriverState *bs;
- AioContext *ctx;
unset_dirty_tracking();
@@ -693,13 +676,7 @@ static void block_migration_cleanup_bmds(void)
bdrv_op_unblock_all(bs, bmds->blocker);
}
error_free(bmds->blocker);
-
- /* Save ctx, because bmds->blk can disappear during blk_unref. */
- ctx = blk_get_aio_context(bmds->blk);
- aio_context_acquire(ctx);
blk_unref(bmds->blk);
- aio_context_release(ctx);
-
g_free(bmds->blk_name);
g_free(bmds->aio_bitmap);
g_free(bmds);
diff --git a/migration/migration-hmp-cmds.c b/migration/migration-hmp-cmds.c
index 86ae832176..99710c8ffb 100644
--- a/migration/migration-hmp-cmds.c
+++ b/migration/migration-hmp-cmds.c
@@ -852,14 +852,11 @@ static void vm_completion(ReadLineState *rs, const char *str)
for (bs = bdrv_first(&it); bs; bs = bdrv_next(&it)) {
SnapshotInfoList *snapshots, *snapshot;
- AioContext *ctx = bdrv_get_aio_context(bs);
bool ok = false;
- aio_context_acquire(ctx);
if (bdrv_can_snapshot(bs)) {
ok = bdrv_query_snapshot_info_list(bs, &snapshots, NULL) == 0;
}
- aio_context_release(ctx);
if (!ok) {
continue;
}
diff --git a/migration/savevm.c b/migration/savevm.c
index eec5503a42..1b9ab7b8ee 100644
--- a/migration/savevm.c
+++ b/migration/savevm.c
@@ -3049,7 +3049,6 @@ bool save_snapshot(const char *name, bool overwrite, const char *vmstate,
int saved_vm_running;
uint64_t vm_state_size;
g_autoptr(GDateTime) now = g_date_time_new_now_local();
- AioContext *aio_context;
GLOBAL_STATE_CODE();
@@ -3092,7 +3091,6 @@ bool save_snapshot(const char *name, bool overwrite, const char *vmstate,
if (bs == NULL) {
return false;
}
- aio_context = bdrv_get_aio_context(bs);
saved_vm_running = runstate_is_running();
@@ -3101,8 +3099,6 @@ bool save_snapshot(const char *name, bool overwrite, const char *vmstate,
bdrv_drain_all_begin();
- aio_context_acquire(aio_context);
-
memset(sn, 0, sizeof(*sn));
/* fill auxiliary fields */
@@ -3139,14 +3135,6 @@ bool save_snapshot(const char *name, bool overwrite, const char *vmstate,
goto the_end;
}
- /* The bdrv_all_create_snapshot() call that follows acquires the AioContext
- * for itself. BDRV_POLL_WHILE() does not support nested locking because
- * it only releases the lock once. Therefore synchronous I/O will deadlock
- * unless we release the AioContext before bdrv_all_create_snapshot().
- */
- aio_context_release(aio_context);
- aio_context = NULL;
-
ret = bdrv_all_create_snapshot(sn, bs, vm_state_size,
has_devices, devices, errp);
if (ret < 0) {
@@ -3157,10 +3145,6 @@ bool save_snapshot(const char *name, bool overwrite, const char *vmstate,
ret = 0;
the_end:
- if (aio_context) {
- aio_context_release(aio_context);
- }
-
bdrv_drain_all_end();
if (saved_vm_running) {
@@ -3258,7 +3242,6 @@ bool load_snapshot(const char *name, const char *vmstate,
QEMUSnapshotInfo sn;
QEMUFile *f;
int ret;
- AioContext *aio_context;
MigrationIncomingState *mis = migration_incoming_get_current();
if (!bdrv_all_can_snapshot(has_devices, devices, errp)) {
@@ -3278,12 +3261,9 @@ bool load_snapshot(const char *name, const char *vmstate,
if (!bs_vm_state) {
return false;
}
- aio_context = bdrv_get_aio_context(bs_vm_state);
/* Don't even try to load empty VM states */
- aio_context_acquire(aio_context);
ret = bdrv_snapshot_find(bs_vm_state, &sn, name);
- aio_context_release(aio_context);
if (ret < 0) {
return false;
} else if (sn.vm_state_size == 0) {
@@ -3320,10 +3300,8 @@ bool load_snapshot(const char *name, const char *vmstate,
ret = -EINVAL;
goto err_drain;
}
- aio_context_acquire(aio_context);
ret = qemu_loadvm_state(f);
migration_incoming_state_destroy();
- aio_context_release(aio_context);
bdrv_drain_all_end();
diff --git a/net/colo-compare.c b/net/colo-compare.c
index 7f9e6f89ce..f2dfc0ebdc 100644
--- a/net/colo-compare.c
+++ b/net/colo-compare.c
@@ -1439,12 +1439,10 @@ static void colo_compare_finalize(Object *obj)
qemu_bh_delete(s->event_bh);
AioContext *ctx = iothread_get_aio_context(s->iothread);
- aio_context_acquire(ctx);
AIO_WAIT_WHILE(ctx, !s->out_sendco.done);
if (s->notify_dev) {
AIO_WAIT_WHILE(ctx, !s->notify_sendco.done);
}
- aio_context_release(ctx);
/* Release all unhandled packets after compare thead exited */
g_queue_foreach(&s->conn_list, colo_flush_packets, s);
diff --git a/qemu-img.c b/qemu-img.c
index 5a77f67719..7668f86769 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -960,7 +960,6 @@ static int img_commit(int argc, char **argv)
Error *local_err = NULL;
CommonBlockJobCBInfo cbi;
bool image_opts = false;
- AioContext *aio_context;
int64_t rate_limit = 0;
fmt = NULL;
@@ -1078,12 +1077,9 @@ static int img_commit(int argc, char **argv)
.bs = bs,
};
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
commit_active_start("commit", bs, base_bs, JOB_DEFAULT, rate_limit,
BLOCKDEV_ON_ERROR_REPORT, NULL, common_block_job_cb,
&cbi, false, &local_err);
- aio_context_release(aio_context);
if (local_err) {
goto done;
}
diff --git a/qemu-io.c b/qemu-io.c
index 050c70835f..6cb1e00385 100644
--- a/qemu-io.c
+++ b/qemu-io.c
@@ -414,15 +414,7 @@ static void prep_fetchline(void *opaque)
static int do_qemuio_command(const char *cmd)
{
- int ret;
- AioContext *ctx =
- qemuio_blk ? blk_get_aio_context(qemuio_blk) : qemu_get_aio_context();
-
- aio_context_acquire(ctx);
- ret = qemuio_command(qemuio_blk, cmd);
- aio_context_release(ctx);
-
- return ret;
+ return qemuio_command(qemuio_blk, cmd);
}
static int command_loop(void)
diff --git a/qemu-nbd.c b/qemu-nbd.c
index 186e6468b1..bac0b5e3ec 100644
--- a/qemu-nbd.c
+++ b/qemu-nbd.c
@@ -1123,9 +1123,7 @@ int main(int argc, char **argv)
qdict_put_str(raw_opts, "file", bs->node_name);
qdict_put_int(raw_opts, "offset", dev_offset);
- aio_context_acquire(qemu_get_aio_context());
bs = bdrv_open(NULL, NULL, raw_opts, flags, &error_fatal);
- aio_context_release(qemu_get_aio_context());
blk_remove_bs(blk);
blk_insert_bs(blk, bs, &error_fatal);
diff --git a/replay/replay-debugging.c b/replay/replay-debugging.c
index 3e60549a4a..82c66fff26 100644
--- a/replay/replay-debugging.c
+++ b/replay/replay-debugging.c
@@ -144,7 +144,6 @@ static char *replay_find_nearest_snapshot(int64_t icount,
char *ret = NULL;
int rv;
int nb_sns, i;
- AioContext *aio_context;
*snapshot_icount = -1;
@@ -152,11 +151,8 @@ static char *replay_find_nearest_snapshot(int64_t icount,
if (!bs) {
goto fail;
}
- aio_context = bdrv_get_aio_context(bs);
- aio_context_acquire(aio_context);
nb_sns = bdrv_snapshot_list(bs, &sn_tab);
- aio_context_release(aio_context);
for (i = 0; i < nb_sns; i++) {
rv = bdrv_all_has_snapshot(sn_tab[i].name, false, NULL, NULL);
diff --git a/tests/unit/test-bdrv-drain.c b/tests/unit/test-bdrv-drain.c
index d9754dfebc..17830a69c1 100644
--- a/tests/unit/test-bdrv-drain.c
+++ b/tests/unit/test-bdrv-drain.c
@@ -179,13 +179,7 @@ static void do_drain_end(enum drain_type drain_type, BlockDriverState *bs)
static void do_drain_begin_unlocked(enum drain_type drain_type, BlockDriverState *bs)
{
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_acquire(bdrv_get_aio_context(bs));
- }
do_drain_begin(drain_type, bs);
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_release(bdrv_get_aio_context(bs));
- }
}
static BlockBackend * no_coroutine_fn test_setup(void)
@@ -209,13 +203,7 @@ static BlockBackend * no_coroutine_fn test_setup(void)
static void do_drain_end_unlocked(enum drain_type drain_type, BlockDriverState *bs)
{
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_acquire(bdrv_get_aio_context(bs));
- }
do_drain_end(drain_type, bs);
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_release(bdrv_get_aio_context(bs));
- }
}
/*
@@ -520,12 +508,8 @@ static void test_iothread_main_thread_bh(void *opaque)
{
struct test_iothread_data *data = opaque;
- /* Test that the AioContext is not yet locked in a random BH that is
- * executed during drain, otherwise this would deadlock. */
- aio_context_acquire(bdrv_get_aio_context(data->bs));
bdrv_flush(data->bs);
bdrv_dec_in_flight(data->bs); /* incremented by test_iothread_common() */
- aio_context_release(bdrv_get_aio_context(data->bs));
}
/*
@@ -567,7 +551,6 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
blk_set_disable_request_queuing(blk, true);
blk_set_aio_context(blk, ctx_a, &error_abort);
- aio_context_acquire(ctx_a);
s->bh_indirection_ctx = ctx_b;
@@ -582,8 +565,6 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
g_assert(acb != NULL);
g_assert_cmpint(aio_ret, ==, -EINPROGRESS);
- aio_context_release(ctx_a);
-
data = (struct test_iothread_data) {
.bs = bs,
.drain_type = drain_type,
@@ -592,10 +573,6 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
switch (drain_thread) {
case 0:
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_acquire(ctx_a);
- }
-
/*
* Increment in_flight so that do_drain_begin() waits for
* test_iothread_main_thread_bh(). This prevents the race between
@@ -613,20 +590,10 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
do_drain_begin(drain_type, bs);
g_assert_cmpint(bs->in_flight, ==, 0);
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_release(ctx_a);
- }
qemu_event_wait(&done_event);
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_acquire(ctx_a);
- }
g_assert_cmpint(aio_ret, ==, 0);
do_drain_end(drain_type, bs);
-
- if (drain_type != BDRV_DRAIN_ALL) {
- aio_context_release(ctx_a);
- }
break;
case 1:
co = qemu_coroutine_create(test_iothread_drain_co_entry, &data);
@@ -637,9 +604,7 @@ static void test_iothread_common(enum drain_type drain_type, int drain_thread)
g_assert_not_reached();
}
- aio_context_acquire(ctx_a);
blk_set_aio_context(blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(ctx_a);
bdrv_unref(bs);
blk_unref(blk);
@@ -757,7 +722,6 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
BlockJob *job;
TestBlockJob *tjob;
IOThread *iothread = NULL;
- AioContext *ctx;
int ret;
src = bdrv_new_open_driver(&bdrv_test, "source", BDRV_O_RDWR,
@@ -787,11 +751,11 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
}
if (use_iothread) {
+ AioContext *ctx;
+
iothread = iothread_new();
ctx = iothread_get_aio_context(iothread);
blk_set_aio_context(blk_src, ctx, &error_abort);
- } else {
- ctx = qemu_get_aio_context();
}
target = bdrv_new_open_driver(&bdrv_test, "target", BDRV_O_RDWR,
@@ -800,7 +764,6 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
blk_insert_bs(blk_target, target, &error_abort);
blk_set_allow_aio_context_change(blk_target, true);
- aio_context_acquire(ctx);
tjob = block_job_create("job0", &test_job_driver, NULL, src,
0, BLK_PERM_ALL,
0, 0, NULL, NULL, &error_abort);
@@ -821,7 +784,6 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
tjob->prepare_ret = -EIO;
break;
}
- aio_context_release(ctx);
job_start(&job->job);
@@ -912,12 +874,10 @@ static void test_blockjob_common_drain_node(enum drain_type drain_type,
}
g_assert_cmpint(ret, ==, (result == TEST_JOB_SUCCESS ? 0 : -EIO));
- aio_context_acquire(ctx);
if (use_iothread) {
blk_set_aio_context(blk_src, qemu_get_aio_context(), &error_abort);
assert(blk_get_aio_context(blk_target) == qemu_get_aio_context());
}
- aio_context_release(ctx);
blk_unref(blk_src);
blk_unref(blk_target);
@@ -1401,9 +1361,7 @@ static void test_append_to_drained(void)
g_assert_cmpint(base_s->drain_count, ==, 1);
g_assert_cmpint(base->in_flight, ==, 0);
- aio_context_acquire(qemu_get_aio_context());
bdrv_append(overlay, base, &error_abort);
- aio_context_release(qemu_get_aio_context());
g_assert_cmpint(base->in_flight, ==, 0);
g_assert_cmpint(overlay->in_flight, ==, 0);
@@ -1438,16 +1396,11 @@ static void test_set_aio_context(void)
bdrv_drained_begin(bs);
bdrv_try_change_aio_context(bs, ctx_a, NULL, &error_abort);
-
- aio_context_acquire(ctx_a);
bdrv_drained_end(bs);
bdrv_drained_begin(bs);
bdrv_try_change_aio_context(bs, ctx_b, NULL, &error_abort);
- aio_context_release(ctx_a);
- aio_context_acquire(ctx_b);
bdrv_try_change_aio_context(bs, qemu_get_aio_context(), NULL, &error_abort);
- aio_context_release(ctx_b);
bdrv_drained_end(bs);
bdrv_unref(bs);
diff --git a/tests/unit/test-bdrv-graph-mod.c b/tests/unit/test-bdrv-graph-mod.c
index 8ee6ef38d8..cafc023db4 100644
--- a/tests/unit/test-bdrv-graph-mod.c
+++ b/tests/unit/test-bdrv-graph-mod.c
@@ -142,10 +142,8 @@ static void test_update_perm_tree(void)
BDRV_CHILD_DATA, &error_abort);
bdrv_graph_wrunlock();
- aio_context_acquire(qemu_get_aio_context());
ret = bdrv_append(filter, bs, NULL);
g_assert_cmpint(ret, <, 0);
- aio_context_release(qemu_get_aio_context());
bdrv_unref(filter);
blk_unref(root);
@@ -211,9 +209,7 @@ static void test_should_update_child(void)
bdrv_attach_child(filter, target, "target", &child_of_bds,
BDRV_CHILD_DATA, &error_abort);
bdrv_graph_wrunlock();
- aio_context_acquire(qemu_get_aio_context());
bdrv_append(filter, bs, &error_abort);
- aio_context_release(qemu_get_aio_context());
bdrv_graph_rdlock_main_loop();
g_assert(target->backing->bs == bs);
@@ -440,9 +436,7 @@ static void test_append_greedy_filter(void)
&error_abort);
bdrv_graph_wrunlock();
- aio_context_acquire(qemu_get_aio_context());
bdrv_append(fl, base, &error_abort);
- aio_context_release(qemu_get_aio_context());
bdrv_unref(fl);
bdrv_unref(top);
}
diff --git a/tests/unit/test-block-iothread.c b/tests/unit/test-block-iothread.c
index 9b15d2768c..3766d5de6b 100644
--- a/tests/unit/test-block-iothread.c
+++ b/tests/unit/test-block-iothread.c
@@ -483,7 +483,6 @@ static void test_sync_op(const void *opaque)
bdrv_graph_rdunlock_main_loop();
blk_set_aio_context(blk, ctx, &error_abort);
- aio_context_acquire(ctx);
if (t->fn) {
t->fn(c);
}
@@ -491,7 +490,6 @@ static void test_sync_op(const void *opaque)
t->blkfn(blk);
}
blk_set_aio_context(blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(ctx);
bdrv_unref(bs);
blk_unref(blk);
@@ -576,9 +574,7 @@ static void test_attach_blockjob(void)
aio_poll(qemu_get_aio_context(), false);
}
- aio_context_acquire(ctx);
blk_set_aio_context(blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(ctx);
tjob->n = 0;
while (tjob->n == 0) {
@@ -595,9 +591,7 @@ static void test_attach_blockjob(void)
WITH_JOB_LOCK_GUARD() {
job_complete_sync_locked(&tjob->common.job, &error_abort);
}
- aio_context_acquire(ctx);
blk_set_aio_context(blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(ctx);
bdrv_unref(bs);
blk_unref(blk);
@@ -654,9 +648,7 @@ static void test_propagate_basic(void)
/* Switch the AioContext back */
main_ctx = qemu_get_aio_context();
- aio_context_acquire(ctx);
blk_set_aio_context(blk, main_ctx, &error_abort);
- aio_context_release(ctx);
g_assert(blk_get_aio_context(blk) == main_ctx);
g_assert(bdrv_get_aio_context(bs_a) == main_ctx);
g_assert(bdrv_get_aio_context(bs_verify) == main_ctx);
@@ -732,9 +724,7 @@ static void test_propagate_diamond(void)
/* Switch the AioContext back */
main_ctx = qemu_get_aio_context();
- aio_context_acquire(ctx);
blk_set_aio_context(blk, main_ctx, &error_abort);
- aio_context_release(ctx);
g_assert(blk_get_aio_context(blk) == main_ctx);
g_assert(bdrv_get_aio_context(bs_verify) == main_ctx);
g_assert(bdrv_get_aio_context(bs_a) == main_ctx);
@@ -764,13 +754,11 @@ static void test_propagate_mirror(void)
&error_abort);
/* Start a mirror job */
- aio_context_acquire(main_ctx);
mirror_start("job0", src, target, NULL, JOB_DEFAULT, 0, 0, 0,
MIRROR_SYNC_MODE_NONE, MIRROR_OPEN_BACKING_CHAIN, false,
BLOCKDEV_ON_ERROR_REPORT, BLOCKDEV_ON_ERROR_REPORT,
false, "filter_node", MIRROR_COPY_MODE_BACKGROUND,
&error_abort);
- aio_context_release(main_ctx);
WITH_JOB_LOCK_GUARD() {
job = job_get_locked("job0");
@@ -785,9 +773,7 @@ static void test_propagate_mirror(void)
g_assert(job->aio_context == ctx);
/* Change the AioContext of target */
- aio_context_acquire(ctx);
bdrv_try_change_aio_context(target, main_ctx, NULL, &error_abort);
- aio_context_release(ctx);
g_assert(bdrv_get_aio_context(src) == main_ctx);
g_assert(bdrv_get_aio_context(target) == main_ctx);
g_assert(bdrv_get_aio_context(filter) == main_ctx);
@@ -805,10 +791,8 @@ static void test_propagate_mirror(void)
g_assert(bdrv_get_aio_context(filter) == main_ctx);
/* ...unless we explicitly allow it */
- aio_context_acquire(ctx);
blk_set_allow_aio_context_change(blk, true);
bdrv_try_change_aio_context(target, ctx, NULL, &error_abort);
- aio_context_release(ctx);
g_assert(blk_get_aio_context(blk) == ctx);
g_assert(bdrv_get_aio_context(src) == ctx);
@@ -817,10 +801,8 @@ static void test_propagate_mirror(void)
job_cancel_sync_all();
- aio_context_acquire(ctx);
blk_set_aio_context(blk, main_ctx, &error_abort);
bdrv_try_change_aio_context(target, main_ctx, NULL, &error_abort);
- aio_context_release(ctx);
blk_unref(blk);
bdrv_unref(src);
@@ -836,7 +818,6 @@ static void test_attach_second_node(void)
BlockDriverState *bs, *filter;
QDict *options;
- aio_context_acquire(main_ctx);
blk = blk_new(ctx, BLK_PERM_ALL, BLK_PERM_ALL);
bs = bdrv_new_open_driver(&bdrv_test, "base", BDRV_O_RDWR, &error_abort);
blk_insert_bs(blk, bs, &error_abort);
@@ -846,15 +827,12 @@ static void test_attach_second_node(void)
qdict_put_str(options, "file", "base");
filter = bdrv_open(NULL, NULL, options, BDRV_O_RDWR, &error_abort);
- aio_context_release(main_ctx);
g_assert(blk_get_aio_context(blk) == ctx);
g_assert(bdrv_get_aio_context(bs) == ctx);
g_assert(bdrv_get_aio_context(filter) == ctx);
- aio_context_acquire(ctx);
blk_set_aio_context(blk, main_ctx, &error_abort);
- aio_context_release(ctx);
g_assert(blk_get_aio_context(blk) == main_ctx);
g_assert(bdrv_get_aio_context(bs) == main_ctx);
g_assert(bdrv_get_aio_context(filter) == main_ctx);
@@ -868,11 +846,9 @@ static void test_attach_preserve_blk_ctx(void)
{
IOThread *iothread = iothread_new();
AioContext *ctx = iothread_get_aio_context(iothread);
- AioContext *main_ctx = qemu_get_aio_context();
BlockBackend *blk;
BlockDriverState *bs;
- aio_context_acquire(main_ctx);
blk = blk_new(ctx, BLK_PERM_ALL, BLK_PERM_ALL);
bs = bdrv_new_open_driver(&bdrv_test, "base", BDRV_O_RDWR, &error_abort);
bs->total_sectors = 65536 / BDRV_SECTOR_SIZE;
@@ -881,25 +857,18 @@ static void test_attach_preserve_blk_ctx(void)
blk_insert_bs(blk, bs, &error_abort);
g_assert(blk_get_aio_context(blk) == ctx);
g_assert(bdrv_get_aio_context(bs) == ctx);
- aio_context_release(main_ctx);
/* Remove the node again */
- aio_context_acquire(ctx);
blk_remove_bs(blk);
- aio_context_release(ctx);
g_assert(blk_get_aio_context(blk) == ctx);
g_assert(bdrv_get_aio_context(bs) == qemu_get_aio_context());
/* Re-attach the node */
- aio_context_acquire(main_ctx);
blk_insert_bs(blk, bs, &error_abort);
- aio_context_release(main_ctx);
g_assert(blk_get_aio_context(blk) == ctx);
g_assert(bdrv_get_aio_context(bs) == ctx);
- aio_context_acquire(ctx);
blk_set_aio_context(blk, qemu_get_aio_context(), &error_abort);
- aio_context_release(ctx);
bdrv_unref(bs);
blk_unref(blk);
}
diff --git a/tests/unit/test-blockjob.c b/tests/unit/test-blockjob.c
index a130f6fefb..fe3e0d2d38 100644
--- a/tests/unit/test-blockjob.c
+++ b/tests/unit/test-blockjob.c
@@ -228,7 +228,6 @@ static void cancel_common(CancelJob *s)
BlockJob *job = &s->common;
BlockBackend *blk = s->blk;
JobStatus sts = job->job.status;
- AioContext *ctx = job->job.aio_context;
job_cancel_sync(&job->job, true);
WITH_JOB_LOCK_GUARD() {
@@ -240,9 +239,7 @@ static void cancel_common(CancelJob *s)
job_unref_locked(&job->job);
}
- aio_context_acquire(ctx);
destroy_blk(blk);
- aio_context_release(ctx);
}
@@ -391,132 +388,6 @@ static void test_cancel_concluded(void)
cancel_common(s);
}
-/* (See test_yielding_driver for the job description) */
-typedef struct YieldingJob {
- BlockJob common;
- bool should_complete;
-} YieldingJob;
-
-static void yielding_job_complete(Job *job, Error **errp)
-{
- YieldingJob *s = container_of(job, YieldingJob, common.job);
- s->should_complete = true;
- job_enter(job);
-}
-
-static int coroutine_fn yielding_job_run(Job *job, Error **errp)
-{
- YieldingJob *s = container_of(job, YieldingJob, common.job);
-
- job_transition_to_ready(job);
-
- while (!s->should_complete) {
- job_yield(job);
- }
-
- return 0;
-}
-
-/*
- * This job transitions immediately to the READY state, and then
- * yields until it is to complete.
- */
-static const BlockJobDriver test_yielding_driver = {
- .job_driver = {
- .instance_size = sizeof(YieldingJob),
- .free = block_job_free,
- .user_resume = block_job_user_resume,
- .run = yielding_job_run,
- .complete = yielding_job_complete,
- },
-};
-
-/*
- * Test that job_complete_locked() works even on jobs that are in a paused
- * state (i.e., STANDBY).
- *
- * To do this, run YieldingJob in an IO thread, get it into the READY
- * state, then have a drained section. Before ending the section,
- * acquire the context so the job will not be entered and will thus
- * remain on STANDBY.
- *
- * job_complete_locked() should still work without error.
- *
- * Note that on the QMP interface, it is impossible to lock an IO
- * thread before a drained section ends. In practice, the
- * bdrv_drain_all_end() and the aio_context_acquire() will be
- * reversed. However, that makes for worse reproducibility here:
- * Sometimes, the job would no longer be in STANDBY then but already
- * be started. We cannot prevent that, because the IO thread runs
- * concurrently. We can only prevent it by taking the lock before
- * ending the drained section, so we do that.
- *
- * (You can reverse the order of operations and most of the time the
- * test will pass, but sometimes the assert(status == STANDBY) will
- * fail.)
- */
-static void test_complete_in_standby(void)
-{
- BlockBackend *blk;
- IOThread *iothread;
- AioContext *ctx;
- Job *job;
- BlockJob *bjob;
-
- /* Create a test drive, move it to an IO thread */
- blk = create_blk(NULL);
- iothread = iothread_new();
-
- ctx = iothread_get_aio_context(iothread);
- blk_set_aio_context(blk, ctx, &error_abort);
-
- /* Create our test job */
- bjob = mk_job(blk, "job", &test_yielding_driver, true,
- JOB_MANUAL_FINALIZE | JOB_MANUAL_DISMISS);
- job = &bjob->job;
- assert_job_status_is(job, JOB_STATUS_CREATED);
-
- /* Wait for the job to become READY */
- job_start(job);
- /*
- * Here we are waiting for the status to change, so don't bother
- * protecting the read every time.
- */
- AIO_WAIT_WHILE_UNLOCKED(ctx, job->status != JOB_STATUS_READY);
-
- /* Begin the drained section, pausing the job */
- bdrv_drain_all_begin();
- assert_job_status_is(job, JOB_STATUS_STANDBY);
-
- /* Lock the IO thread to prevent the job from being run */
- aio_context_acquire(ctx);
- /* This will schedule the job to resume it */
- bdrv_drain_all_end();
- aio_context_release(ctx);
-
- WITH_JOB_LOCK_GUARD() {
- /* But the job cannot run, so it will remain on standby */
- assert(job->status == JOB_STATUS_STANDBY);
-
- /* Even though the job is on standby, this should work */
- job_complete_locked(job, &error_abort);
-
- /* The test is done now, clean up. */
- job_finish_sync_locked(job, NULL, &error_abort);
- assert(job->status == JOB_STATUS_PENDING);
-
- job_finalize_locked(job, &error_abort);
- assert(job->status == JOB_STATUS_CONCLUDED);
-
- job_dismiss_locked(&job, &error_abort);
- }
-
- aio_context_acquire(ctx);
- destroy_blk(blk);
- aio_context_release(ctx);
- iothread_join(iothread);
-}
-
int main(int argc, char **argv)
{
qemu_init_main_loop(&error_abort);
@@ -531,13 +402,5 @@ int main(int argc, char **argv)
g_test_add_func("/blockjob/cancel/standby", test_cancel_standby);
g_test_add_func("/blockjob/cancel/pending", test_cancel_pending);
g_test_add_func("/blockjob/cancel/concluded", test_cancel_concluded);
-
- /*
- * This test is flaky and sometimes fails in CI and otherwise:
- * don't run unless user opts in via environment variable.
- */
- if (getenv("QEMU_TEST_FLAKY_TESTS")) {
- g_test_add_func("/blockjob/complete_in_standby", test_complete_in_standby);
- }
return g_test_run();
}
diff --git a/tests/unit/test-replication.c b/tests/unit/test-replication.c
index afff908d77..5d2003b8ce 100644
--- a/tests/unit/test-replication.c
+++ b/tests/unit/test-replication.c
@@ -199,17 +199,13 @@ static BlockBackend *start_primary(void)
static void teardown_primary(void)
{
BlockBackend *blk;
- AioContext *ctx;
/* remove P_ID */
blk = blk_by_name(P_ID);
assert(blk);
- ctx = blk_get_aio_context(blk);
- aio_context_acquire(ctx);
monitor_remove_blk(blk);
blk_unref(blk);
- aio_context_release(ctx);
}
static void test_primary_read(void)
@@ -345,27 +341,20 @@ static void teardown_secondary(void)
{
/* only need to destroy two BBs */
BlockBackend *blk;
- AioContext *ctx;
/* remove S_LOCAL_DISK_ID */
blk = blk_by_name(S_LOCAL_DISK_ID);
assert(blk);
- ctx = blk_get_aio_context(blk);
- aio_context_acquire(ctx);
monitor_remove_blk(blk);
blk_unref(blk);
- aio_context_release(ctx);
/* remove S_ID */
blk = blk_by_name(S_ID);
assert(blk);
- ctx = blk_get_aio_context(blk);
- aio_context_acquire(ctx);
monitor_remove_blk(blk);
blk_unref(blk);
- aio_context_release(ctx);
}
static void test_secondary_read(void)
diff --git a/util/async.c b/util/async.c
index 04ee83d220..dfd44ef612 100644
--- a/util/async.c
+++ b/util/async.c
@@ -562,12 +562,10 @@ static void co_schedule_bh_cb(void *opaque)
Coroutine *co = QSLIST_FIRST(&straight);
QSLIST_REMOVE_HEAD(&straight, co_scheduled_next);
trace_aio_co_schedule_bh_cb(ctx, co);
- aio_context_acquire(ctx);
/* Protected by write barrier in qemu_aio_coroutine_enter */
qatomic_set(&co->scheduled, NULL);
qemu_aio_coroutine_enter(ctx, co);
- aio_context_release(ctx);
}
}
@@ -707,9 +705,7 @@ void aio_co_enter(AioContext *ctx, Coroutine *co)
assert(self != co);
QSIMPLEQ_INSERT_TAIL(&self->co_queue_wakeup, co, co_queue_next);
} else {
- aio_context_acquire(ctx);
qemu_aio_coroutine_enter(ctx, co);
- aio_context_release(ctx);
}
}
diff --git a/util/vhost-user-server.c b/util/vhost-user-server.c
index a9a48fffb8..3bfb1ad3ec 100644
--- a/util/vhost-user-server.c
+++ b/util/vhost-user-server.c
@@ -360,10 +360,7 @@ static void vu_accept(QIONetListener *listener, QIOChannelSocket *sioc,
qio_channel_set_follow_coroutine_ctx(server->ioc, true);
- /* Attaching the AioContext starts the vu_client_trip coroutine */
- aio_context_acquire(server->ctx);
vhost_user_server_attach_aio_context(server, server->ctx);
- aio_context_release(server->ctx);
}
/* server->ctx acquired by caller */
diff --git a/scripts/block-coroutine-wrapper.py b/scripts/block-coroutine-wrapper.py
index 38364fa557..c9c09fcacd 100644
--- a/scripts/block-coroutine-wrapper.py
+++ b/scripts/block-coroutine-wrapper.py
@@ -278,12 +278,9 @@ def gen_no_co_wrapper(func: FuncDecl) -> str:
static void {name}_bh(void *opaque)
{{
{struct_name} *s = opaque;
- AioContext *ctx = {func.gen_ctx('s->')};
{graph_lock}
- aio_context_acquire(ctx);
{func.get_result}{name}({ func.gen_list('s->{name}') });
- aio_context_release(ctx);
{graph_unlock}
aio_co_wake(s->co);
diff --git a/tests/tsan/suppressions.tsan b/tests/tsan/suppressions.tsan
index d9a002a2ef..b3ef59c27c 100644
--- a/tests/tsan/suppressions.tsan
+++ b/tests/tsan/suppressions.tsan
@@ -4,7 +4,6 @@
# TSan reports a double lock on RECURSIVE mutexes.
# Since the recursive lock is intentional, we choose to ignore it.
-mutex:aio_context_acquire
mutex:pthread_mutex_lock
# TSan reports a race between pthread_mutex_init() and
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 06/12] scsi: remove AioContext locking
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (4 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 21:36 ` Eric Blake
2023-12-04 12:23 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED() Stefan Hajnoczi
` (5 subsequent siblings)
11 siblings, 2 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
The AioContext lock no longer has any effect. Remove it.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/hw/virtio/virtio-scsi.h | 14 --------------
hw/scsi/scsi-bus.c | 2 --
hw/scsi/scsi-disk.c | 28 ++++------------------------
hw/scsi/virtio-scsi.c | 18 ------------------
4 files changed, 4 insertions(+), 58 deletions(-)
diff --git a/include/hw/virtio/virtio-scsi.h b/include/hw/virtio/virtio-scsi.h
index da8cb928d9..7f0573b1bf 100644
--- a/include/hw/virtio/virtio-scsi.h
+++ b/include/hw/virtio/virtio-scsi.h
@@ -101,20 +101,6 @@ struct VirtIOSCSI {
uint32_t host_features;
};
-static inline void virtio_scsi_acquire(VirtIOSCSI *s)
-{
- if (s->ctx) {
- aio_context_acquire(s->ctx);
- }
-}
-
-static inline void virtio_scsi_release(VirtIOSCSI *s)
-{
- if (s->ctx) {
- aio_context_release(s->ctx);
- }
-}
-
void virtio_scsi_common_realize(DeviceState *dev,
VirtIOHandleOutput ctrl,
VirtIOHandleOutput evt,
diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c
index b8bfde9565..0031164cc3 100644
--- a/hw/scsi/scsi-bus.c
+++ b/hw/scsi/scsi-bus.c
@@ -1725,9 +1725,7 @@ void scsi_device_purge_requests(SCSIDevice *sdev, SCSISense sense)
{
scsi_device_for_each_req_async(sdev, scsi_device_purge_one_req, NULL);
- aio_context_acquire(blk_get_aio_context(sdev->conf.blk));
blk_drain(sdev->conf.blk);
- aio_context_release(blk_get_aio_context(sdev->conf.blk));
scsi_device_set_ua(sdev, sense);
}
diff --git a/hw/scsi/scsi-disk.c b/hw/scsi/scsi-disk.c
index 2c1bbb3530..f1bd5f5c6e 100644
--- a/hw/scsi/scsi-disk.c
+++ b/hw/scsi/scsi-disk.c
@@ -2325,14 +2325,10 @@ static void scsi_disk_reset(DeviceState *dev)
{
SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev.qdev, dev);
uint64_t nb_sectors;
- AioContext *ctx;
scsi_device_purge_requests(&s->qdev, SENSE_CODE(RESET));
- ctx = blk_get_aio_context(s->qdev.conf.blk);
- aio_context_acquire(ctx);
blk_get_geometry(s->qdev.conf.blk, &nb_sectors);
- aio_context_release(ctx);
nb_sectors /= s->qdev.blocksize / BDRV_SECTOR_SIZE;
if (nb_sectors) {
@@ -2531,13 +2527,11 @@ static void scsi_unrealize(SCSIDevice *dev)
static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
{
SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, dev);
- AioContext *ctx = NULL;
+
/* can happen for devices without drive. The error message for missing
* backend will be issued in scsi_realize
*/
if (s->qdev.conf.blk) {
- ctx = blk_get_aio_context(s->qdev.conf.blk);
- aio_context_acquire(ctx);
if (!blkconf_blocksizes(&s->qdev.conf, errp)) {
goto out;
}
@@ -2549,15 +2543,11 @@ static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
}
scsi_realize(&s->qdev, errp);
out:
- if (ctx) {
- aio_context_release(ctx);
- }
}
static void scsi_cd_realize(SCSIDevice *dev, Error **errp)
{
SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, dev);
- AioContext *ctx;
int ret;
uint32_t blocksize = 2048;
@@ -2573,8 +2563,6 @@ static void scsi_cd_realize(SCSIDevice *dev, Error **errp)
blocksize = dev->conf.physical_block_size;
}
- ctx = blk_get_aio_context(dev->conf.blk);
- aio_context_acquire(ctx);
s->qdev.blocksize = blocksize;
s->qdev.type = TYPE_ROM;
s->features |= 1 << SCSI_DISK_F_REMOVABLE;
@@ -2582,7 +2570,6 @@ static void scsi_cd_realize(SCSIDevice *dev, Error **errp)
s->product = g_strdup("QEMU CD-ROM");
}
scsi_realize(&s->qdev, errp);
- aio_context_release(ctx);
}
@@ -2713,7 +2700,6 @@ static int get_device_type(SCSIDiskState *s)
static void scsi_block_realize(SCSIDevice *dev, Error **errp)
{
SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, dev);
- AioContext *ctx;
int sg_version;
int rc;
@@ -2728,9 +2714,6 @@ static void scsi_block_realize(SCSIDevice *dev, Error **errp)
"be removed in a future version");
}
- ctx = blk_get_aio_context(s->qdev.conf.blk);
- aio_context_acquire(ctx);
-
/* check we are using a driver managing SG_IO (version 3 and after) */
rc = blk_ioctl(s->qdev.conf.blk, SG_GET_VERSION_NUM, &sg_version);
if (rc < 0) {
@@ -2738,18 +2721,18 @@ static void scsi_block_realize(SCSIDevice *dev, Error **errp)
if (rc != -EPERM) {
error_append_hint(errp, "Is this a SCSI device?\n");
}
- goto out;
+ return;
}
if (sg_version < 30000) {
error_setg(errp, "scsi generic interface too old");
- goto out;
+ return;
}
/* get device type from INQUIRY data */
rc = get_device_type(s);
if (rc < 0) {
error_setg(errp, "INQUIRY failed");
- goto out;
+ return;
}
/* Make a guess for the block size, we'll fix it when the guest sends.
@@ -2769,9 +2752,6 @@ static void scsi_block_realize(SCSIDevice *dev, Error **errp)
scsi_realize(&s->qdev, errp);
scsi_generic_read_device_inquiry(&s->qdev);
-
-out:
- aio_context_release(ctx);
}
typedef struct SCSIBlockReq {
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index 4f8d35facc..ca365a70e9 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -642,9 +642,7 @@ static void virtio_scsi_handle_ctrl(VirtIODevice *vdev, VirtQueue *vq)
return;
}
- virtio_scsi_acquire(s);
virtio_scsi_handle_ctrl_vq(s, vq);
- virtio_scsi_release(s);
}
static void virtio_scsi_complete_cmd_req(VirtIOSCSIReq *req)
@@ -882,9 +880,7 @@ static void virtio_scsi_handle_cmd(VirtIODevice *vdev, VirtQueue *vq)
return;
}
- virtio_scsi_acquire(s);
virtio_scsi_handle_cmd_vq(s, vq);
- virtio_scsi_release(s);
}
static void virtio_scsi_get_config(VirtIODevice *vdev,
@@ -1031,9 +1027,7 @@ static void virtio_scsi_handle_event(VirtIODevice *vdev, VirtQueue *vq)
return;
}
- virtio_scsi_acquire(s);
virtio_scsi_handle_event_vq(s, vq);
- virtio_scsi_release(s);
}
static void virtio_scsi_change(SCSIBus *bus, SCSIDevice *dev, SCSISense sense)
@@ -1052,9 +1046,7 @@ static void virtio_scsi_change(SCSIBus *bus, SCSIDevice *dev, SCSISense sense)
},
};
- virtio_scsi_acquire(s);
virtio_scsi_push_event(s, &info);
- virtio_scsi_release(s);
}
}
@@ -1071,17 +1063,13 @@ static void virtio_scsi_hotplug(HotplugHandler *hotplug_dev, DeviceState *dev,
VirtIODevice *vdev = VIRTIO_DEVICE(hotplug_dev);
VirtIOSCSI *s = VIRTIO_SCSI(vdev);
SCSIDevice *sd = SCSI_DEVICE(dev);
- AioContext *old_context;
int ret;
if (s->ctx && !s->dataplane_fenced) {
if (blk_op_is_blocked(sd->conf.blk, BLOCK_OP_TYPE_DATAPLANE, errp)) {
return;
}
- old_context = blk_get_aio_context(sd->conf.blk);
- aio_context_acquire(old_context);
ret = blk_set_aio_context(sd->conf.blk, s->ctx, errp);
- aio_context_release(old_context);
if (ret < 0) {
return;
}
@@ -1097,10 +1085,8 @@ static void virtio_scsi_hotplug(HotplugHandler *hotplug_dev, DeviceState *dev,
},
};
- virtio_scsi_acquire(s);
virtio_scsi_push_event(s, &info);
scsi_bus_set_ua(&s->bus, SENSE_CODE(REPORTED_LUNS_CHANGED));
- virtio_scsi_release(s);
}
}
@@ -1122,17 +1108,13 @@ static void virtio_scsi_hotunplug(HotplugHandler *hotplug_dev, DeviceState *dev,
qdev_simple_device_unplug_cb(hotplug_dev, dev, errp);
if (s->ctx) {
- virtio_scsi_acquire(s);
/* If other users keep the BlockBackend in the iothread, that's ok */
blk_set_aio_context(sd->conf.blk, qemu_get_aio_context(), NULL);
- virtio_scsi_release(s);
}
if (virtio_vdev_has_feature(vdev, VIRTIO_SCSI_F_HOTPLUG)) {
- virtio_scsi_acquire(s);
virtio_scsi_push_event(s, &info);
scsi_bus_set_ua(&s->bus, SENSE_CODE(REPORTED_LUNS_CHANGED));
- virtio_scsi_release(s);
}
}
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED()
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (5 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 06/12] scsi: " Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 21:39 ` Eric Blake
2023-11-29 19:55 ` [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API Stefan Hajnoczi
` (4 subsequent siblings)
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Now that the AioContext lock no longer exists, AIO_WAIT_WHILE() and
AIO_WAIT_WHILE_UNLOCKED() are equivalent.
A future patch will get rid of AIO_WAIT_WHILE_UNLOCKED().
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/block/aio-wait.h | 16 ++++------------
1 file changed, 4 insertions(+), 12 deletions(-)
diff --git a/include/block/aio-wait.h b/include/block/aio-wait.h
index 5449b6d742..157f105916 100644
--- a/include/block/aio-wait.h
+++ b/include/block/aio-wait.h
@@ -63,9 +63,6 @@ extern AioWait global_aio_wait;
* @ctx: the aio context, or NULL if multiple aio contexts (for which the
* caller does not hold a lock) are involved in the polling condition.
* @cond: wait while this conditional expression is true
- * @unlock: whether to unlock and then lock again @ctx. This applies
- * only when waiting for another AioContext from the main loop.
- * Otherwise it's ignored.
*
* Wait while a condition is true. Use this to implement synchronous
* operations that require event loop activity.
@@ -78,7 +75,7 @@ extern AioWait global_aio_wait;
* wait on conditions between two IOThreads since that could lead to deadlock,
* go via the main loop instead.
*/
-#define AIO_WAIT_WHILE_INTERNAL(ctx, cond, unlock) ({ \
+#define AIO_WAIT_WHILE_INTERNAL(ctx, cond) ({ \
bool waited_ = false; \
AioWait *wait_ = &global_aio_wait; \
AioContext *ctx_ = (ctx); \
@@ -95,13 +92,7 @@ extern AioWait global_aio_wait;
assert(qemu_get_current_aio_context() == \
qemu_get_aio_context()); \
while ((cond)) { \
- if (unlock && ctx_) { \
- aio_context_release(ctx_); \
- } \
aio_poll(qemu_get_aio_context(), true); \
- if (unlock && ctx_) { \
- aio_context_acquire(ctx_); \
- } \
waited_ = true; \
} \
} \
@@ -109,10 +100,11 @@ extern AioWait global_aio_wait;
waited_; })
#define AIO_WAIT_WHILE(ctx, cond) \
- AIO_WAIT_WHILE_INTERNAL(ctx, cond, true)
+ AIO_WAIT_WHILE_INTERNAL(ctx, cond)
+/* TODO replace this with AIO_WAIT_WHILE() in a future patch */
#define AIO_WAIT_WHILE_UNLOCKED(ctx, cond) \
- AIO_WAIT_WHILE_INTERNAL(ctx, cond, false)
+ AIO_WAIT_WHILE_INTERNAL(ctx, cond)
/**
* aio_wait_kick:
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (6 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED() Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 21:42 ` Eric Blake
2023-11-29 19:55 ` [PATCH 09/12] docs: remove AioContext lock from IOThread docs Stefan Hajnoczi
` (3 subsequent siblings)
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Delete these functions because nothing calls these functions anymore.
I introduced these APIs in commit 98563fc3ec44 ("aio: add
aio_context_acquire() and aio_context_release()") in 2014. It's with a
sigh of relief that I delete these APIs almost 10 years later.
Thanks to Paolo Bonzini's vision for multi-queue QEMU, we got an
understanding of where the code needed to go in order to remove the
limitations that the original dataplane and the IOThread/AioContext
approach that followed it.
Emanuele Giuseppe Esposito had the splendid determination to convert
large parts of the codebase so that they no longer needed the AioContext
lock. This was a painstaking process, both in the actual code changes
required and the iterations of code review that Emanuele eeked out of
Kevin and me over many months.
Kevin Wolf tackled multitudes of graph locking conversions to protect
in-flight I/O from run-time changes to the block graph as well as the
clang Thread Safety Analysis annotations that allow the compiler to
check whether the graph lock is being used correctly.
And me, well, I'm just here to add some pizzazz to the QEMU multi-queue
block layer :). Thank you to everyone who helped with this effort,
including Eric Blake, code reviewer extraordinaire, and others who I've
forgotten to mention.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/block/aio.h | 17 -----------------
util/async.c | 10 ----------
2 files changed, 27 deletions(-)
diff --git a/include/block/aio.h b/include/block/aio.h
index f08b358077..af05512a7d 100644
--- a/include/block/aio.h
+++ b/include/block/aio.h
@@ -278,23 +278,6 @@ void aio_context_ref(AioContext *ctx);
*/
void aio_context_unref(AioContext *ctx);
-/* Take ownership of the AioContext. If the AioContext will be shared between
- * threads, and a thread does not want to be interrupted, it will have to
- * take ownership around calls to aio_poll(). Otherwise, aio_poll()
- * automatically takes care of calling aio_context_acquire and
- * aio_context_release.
- *
- * Note that this is separate from bdrv_drained_begin/bdrv_drained_end. A
- * thread still has to call those to avoid being interrupted by the guest.
- *
- * Bottom halves, timers and callbacks can be created or removed without
- * acquiring the AioContext.
- */
-void aio_context_acquire(AioContext *ctx);
-
-/* Relinquish ownership of the AioContext. */
-void aio_context_release(AioContext *ctx);
-
/**
* aio_bh_schedule_oneshot_full: Allocate a new bottom half structure that will
* run only once and as soon as possible.
diff --git a/util/async.c b/util/async.c
index dfd44ef612..460529057c 100644
--- a/util/async.c
+++ b/util/async.c
@@ -719,16 +719,6 @@ void aio_context_unref(AioContext *ctx)
g_source_unref(&ctx->source);
}
-void aio_context_acquire(AioContext *ctx)
-{
- /* TODO remove this function */
-}
-
-void aio_context_release(AioContext *ctx)
-{
- /* TODO remove this function */
-}
-
QEMU_DEFINE_STATIC_CO_TLS(AioContext *, my_aiocontext)
AioContext *qemu_get_current_aio_context(void)
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 09/12] docs: remove AioContext lock from IOThread docs
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (7 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 22:38 ` Eric Blake
2023-11-29 19:55 ` [PATCH 10/12] scsi: remove outdated AioContext lock comment Stefan Hajnoczi
` (2 subsequent siblings)
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Encourage the use of locking primitives and stop mentioning the
AioContext lock since it is being removed.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
docs/devel/multiple-iothreads.txt | 45 +++++++++++--------------------
1 file changed, 15 insertions(+), 30 deletions(-)
diff --git a/docs/devel/multiple-iothreads.txt b/docs/devel/multiple-iothreads.txt
index a3e949f6b3..4865196bde 100644
--- a/docs/devel/multiple-iothreads.txt
+++ b/docs/devel/multiple-iothreads.txt
@@ -88,27 +88,18 @@ loop, depending on which AioContext instance the caller passes in.
How to synchronize with an IOThread
-----------------------------------
-AioContext is not thread-safe so some rules must be followed when using file
-descriptors, event notifiers, timers, or BHs across threads:
+Variables that can be accessed by multiple threads require some form of
+synchronization such as qemu_mutex_lock(), rcu_read_lock(), etc.
-1. AioContext functions can always be called safely. They handle their
-own locking internally.
-
-2. Other threads wishing to access the AioContext must use
-aio_context_acquire()/aio_context_release() for mutual exclusion. Once the
-context is acquired no other thread can access it or run event loop iterations
-in this AioContext.
-
-Legacy code sometimes nests aio_context_acquire()/aio_context_release() calls.
-Do not use nesting anymore, it is incompatible with the BDRV_POLL_WHILE() macro
-used in the block layer and can lead to hangs.
-
-There is currently no lock ordering rule if a thread needs to acquire multiple
-AioContexts simultaneously. Therefore, it is only safe for code holding the
-QEMU global mutex to acquire other AioContexts.
+AioContext functions like aio_set_fd_handler(), aio_set_event_notifier(),
+aio_bh_new(), and aio_timer_new() are thread-safe. They can be used to trigger
+activity in an IOThread.
Side note: the best way to schedule a function call across threads is to call
-aio_bh_schedule_oneshot(). No acquire/release or locking is needed.
+aio_bh_schedule_oneshot().
+
+The main loop thread can wait synchronously for a condition using
+AIO_WAIT_WHILE().
AioContext and the block layer
------------------------------
@@ -124,22 +115,16 @@ Block layer code must therefore expect to run in an IOThread and avoid using
old APIs that implicitly use the main loop. See the "How to program for
IOThreads" above for information on how to do that.
-If main loop code such as a QMP function wishes to access a BlockDriverState
-it must first call aio_context_acquire(bdrv_get_aio_context(bs)) to ensure
-that callbacks in the IOThread do not run in parallel.
-
Code running in the monitor typically needs to ensure that past
requests from the guest are completed. When a block device is running
in an IOThread, the IOThread can also process requests from the guest
(via ioeventfd). To achieve both objects, wrap the code between
bdrv_drained_begin() and bdrv_drained_end(), thus creating a "drained
-section". The functions must be called between aio_context_acquire()
-and aio_context_release(). You can freely release and re-acquire the
-AioContext within a drained section.
+section".
-Long-running jobs (usually in the form of coroutines) are best scheduled in
-the BlockDriverState's AioContext to avoid the need to acquire/release around
-each bdrv_*() call. The functions bdrv_add/remove_aio_context_notifier,
-or alternatively blk_add/remove_aio_context_notifier if you use BlockBackends,
-can be used to get a notification whenever bdrv_try_change_aio_context() moves a
+Long-running jobs (usually in the form of coroutines) are often scheduled in
+the BlockDriverState's AioContext. The functions
+bdrv_add/remove_aio_context_notifier, or alternatively
+blk_add/remove_aio_context_notifier if you use BlockBackends, can be used to
+get a notification whenever bdrv_try_change_aio_context() moves a
BlockDriverState to a different AioContext.
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 10/12] scsi: remove outdated AioContext lock comment
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (8 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 09/12] docs: remove AioContext lock from IOThread docs Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-11-30 22:39 ` Eric Blake
2023-11-29 19:55 ` [PATCH 11/12] job: remove outdated AioContext locking comments Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 12/12] block: " Stefan Hajnoczi
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
The SCSI subsystem no longer uses the AioContext lock. Request
processing runs exclusively in the BlockBackend's AioContext since
"scsi: only access SCSIDevice->requests from one thread" and hence the
lock is unnecessary.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/scsi/scsi-disk.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/hw/scsi/scsi-disk.c b/hw/scsi/scsi-disk.c
index f1bd5f5c6e..ef0d21d737 100644
--- a/hw/scsi/scsi-disk.c
+++ b/hw/scsi/scsi-disk.c
@@ -351,7 +351,6 @@ done:
scsi_req_unref(&r->req);
}
-/* Called with AioContext lock held */
static void scsi_dma_complete(void *opaque, int ret)
{
SCSIDiskReq *r = (SCSIDiskReq *)opaque;
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 11/12] job: remove outdated AioContext locking comments
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (9 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 10/12] scsi: remove outdated AioContext lock comment Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-12-01 17:50 ` Eric Blake
2023-11-29 19:55 ` [PATCH 12/12] block: " Stefan Hajnoczi
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
The AioContext lock no longer exists.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/qemu/job.h | 20 --------------------
1 file changed, 20 deletions(-)
diff --git a/include/qemu/job.h b/include/qemu/job.h
index e502787dd8..9ea98b5927 100644
--- a/include/qemu/job.h
+++ b/include/qemu/job.h
@@ -67,8 +67,6 @@ typedef struct Job {
/**
* The completion function that will be called when the job completes.
- * Called with AioContext lock held, since many callback implementations
- * use bdrv_* functions that require to hold the lock.
*/
BlockCompletionFunc *cb;
@@ -264,9 +262,6 @@ struct JobDriver {
*
* This callback will not be invoked if the job has already failed.
* If it fails, abort and then clean will be called.
- *
- * Called with AioContext lock held, since many callbacs implementations
- * use bdrv_* functions that require to hold the lock.
*/
int (*prepare)(Job *job);
@@ -277,9 +272,6 @@ struct JobDriver {
*
* All jobs will complete with a call to either .commit() or .abort() but
* never both.
- *
- * Called with AioContext lock held, since many callback implementations
- * use bdrv_* functions that require to hold the lock.
*/
void (*commit)(Job *job);
@@ -290,9 +282,6 @@ struct JobDriver {
*
* All jobs will complete with a call to either .commit() or .abort() but
* never both.
- *
- * Called with AioContext lock held, since many callback implementations
- * use bdrv_* functions that require to hold the lock.
*/
void (*abort)(Job *job);
@@ -301,9 +290,6 @@ struct JobDriver {
* .commit() or .abort(). Regardless of which callback is invoked after
* completion, .clean() will always be called, even if the job does not
* belong to a transaction group.
- *
- * Called with AioContext lock held, since many callbacs implementations
- * use bdrv_* functions that require to hold the lock.
*/
void (*clean)(Job *job);
@@ -318,17 +304,12 @@ struct JobDriver {
* READY).
* (If the callback is NULL, the job is assumed to terminate
* without I/O.)
- *
- * Called with AioContext lock held, since many callback implementations
- * use bdrv_* functions that require to hold the lock.
*/
bool (*cancel)(Job *job, bool force);
/**
* Called when the job is freed.
- * Called with AioContext lock held, since many callback implementations
- * use bdrv_* functions that require to hold the lock.
*/
void (*free)(Job *job);
};
@@ -424,7 +405,6 @@ void job_ref_locked(Job *job);
* Release a reference that was previously acquired with job_ref_locked() or
* job_create(). If it's the last reference to the object, it will be freed.
*
- * Takes AioContext lock internally to invoke a job->driver callback.
* Called with job lock held.
*/
void job_unref_locked(Job *job);
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* [PATCH 12/12] block: remove outdated AioContext locking comments
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
` (10 preceding siblings ...)
2023-11-29 19:55 ` [PATCH 11/12] job: remove outdated AioContext locking comments Stefan Hajnoczi
@ 2023-11-29 19:55 ` Stefan Hajnoczi
2023-12-01 18:41 ` Eric Blake
11 siblings, 1 reply; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-11-29 19:55 UTC (permalink / raw)
To: qemu-devel
Cc: Hanna Reitz, Paul Durrant, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Stefan Hajnoczi,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
The AioContext lock no longer exists.
There is one noteworthy change:
- * More specifically, these functions use BDRV_POLL_WHILE(bs), which
- * requires the caller to be either in the main thread and hold
- * the BlockdriverState (bs) AioContext lock, or directly in the
- * home thread that runs the bs AioContext. Calling them from
- * another thread in another AioContext would cause deadlocks.
+ * More specifically, these functions use BDRV_POLL_WHILE(bs), which requires
+ * the caller to be either in the main thread or directly in the home thread
+ * that runs the bs AioContext. Calling them from another thread in another
+ * AioContext would cause deadlocks.
I am not sure whether deadlocks are still possible. Maybe they have just
moved to the fine-grained locks that have replaced the AioContext. Since
I am not sure if the deadlocks are gone, I have kept the substance
unchanged and just removed mention of the AioContext.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
include/block/block-common.h | 3 --
include/block/block-io.h | 9 ++--
include/block/block_int-common.h | 2 -
block.c | 73 ++++++----------------------
block/block-backend.c | 8 ---
block/export/vhost-user-blk-server.c | 4 --
tests/qemu-iotests/202 | 2 +-
tests/qemu-iotests/203 | 3 +-
8 files changed, 22 insertions(+), 82 deletions(-)
diff --git a/include/block/block-common.h b/include/block/block-common.h
index d7599564db..a846023a09 100644
--- a/include/block/block-common.h
+++ b/include/block/block-common.h
@@ -70,9 +70,6 @@
* automatically takes the graph rdlock when calling the wrapped function. In
* the same way, no_co_wrapper_bdrv_wrlock functions automatically take the
* graph wrlock.
- *
- * If the first parameter of the function is a BlockDriverState, BdrvChild or
- * BlockBackend pointer, the AioContext lock for it is taken in the wrapper.
*/
#define no_co_wrapper
#define no_co_wrapper_bdrv_rdlock
diff --git a/include/block/block-io.h b/include/block/block-io.h
index 8eb39a858b..b49e0537dd 100644
--- a/include/block/block-io.h
+++ b/include/block/block-io.h
@@ -332,11 +332,10 @@ bdrv_co_copy_range(BdrvChild *src, int64_t src_offset,
* "I/O or GS" API functions. These functions can run without
* the BQL, but only in one specific iothread/main loop.
*
- * More specifically, these functions use BDRV_POLL_WHILE(bs), which
- * requires the caller to be either in the main thread and hold
- * the BlockdriverState (bs) AioContext lock, or directly in the
- * home thread that runs the bs AioContext. Calling them from
- * another thread in another AioContext would cause deadlocks.
+ * More specifically, these functions use BDRV_POLL_WHILE(bs), which requires
+ * the caller to be either in the main thread or directly in the home thread
+ * that runs the bs AioContext. Calling them from another thread in another
+ * AioContext would cause deadlocks.
*
* Therefore, these functions are not proper I/O, because they
* can't run in *any* iothreads, but only in a specific one.
diff --git a/include/block/block_int-common.h b/include/block/block_int-common.h
index 4e31d161c5..151279d481 100644
--- a/include/block/block_int-common.h
+++ b/include/block/block_int-common.h
@@ -1192,8 +1192,6 @@ struct BlockDriverState {
/* The error object in use for blocking operations on backing_hd */
Error *backing_blocker;
- /* Protected by AioContext lock */
-
/*
* If we are reading a disk image, give its size in sectors.
* Generally read-only; it is written to by load_snapshot and
diff --git a/block.c b/block.c
index 91ace5d2d5..e773584dfd 100644
--- a/block.c
+++ b/block.c
@@ -1616,11 +1616,6 @@ out:
g_free(gen_node_name);
}
-/*
- * The caller must always hold @bs AioContext lock, because this function calls
- * bdrv_refresh_total_sectors() which polls when called from non-coroutine
- * context.
- */
static int no_coroutine_fn GRAPH_UNLOCKED
bdrv_open_driver(BlockDriverState *bs, BlockDriver *drv, const char *node_name,
QDict *options, int open_flags, Error **errp)
@@ -2901,7 +2896,7 @@ uint64_t bdrv_qapi_perm_to_blk_perm(BlockPermission qapi_perm)
* Replaces the node that a BdrvChild points to without updating permissions.
*
* If @new_bs is non-NULL, the parent of @child must already be drained through
- * @child and the caller must hold the AioContext lock for @new_bs.
+ * @child.
*/
static void GRAPH_WRLOCK
bdrv_replace_child_noperm(BdrvChild *child, BlockDriverState *new_bs)
@@ -3041,9 +3036,8 @@ static TransactionActionDrv bdrv_attach_child_common_drv = {
*
* Returns new created child.
*
- * The caller must hold the AioContext lock for @child_bs. Both @parent_bs and
- * @child_bs can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * Both @parent_bs and @child_bs can move to a different AioContext in this
+ * function.
*/
static BdrvChild * GRAPH_WRLOCK
bdrv_attach_child_common(BlockDriverState *child_bs,
@@ -3142,9 +3136,8 @@ bdrv_attach_child_common(BlockDriverState *child_bs,
/*
* Function doesn't update permissions, caller is responsible for this.
*
- * The caller must hold the AioContext lock for @child_bs. Both @parent_bs and
- * @child_bs can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * Both @parent_bs and @child_bs can move to a different AioContext in this
+ * function.
*
* After calling this function, the transaction @tran may only be completed
* while holding a writer lock for the graph.
@@ -3184,9 +3177,6 @@ bdrv_attach_child_noperm(BlockDriverState *parent_bs,
*
* On failure NULL is returned, errp is set and the reference to
* child_bs is also dropped.
- *
- * The caller must hold the AioContext lock @child_bs, but not that of @ctx
- * (unless @child_bs is already in @ctx).
*/
BdrvChild *bdrv_root_attach_child(BlockDriverState *child_bs,
const char *child_name,
@@ -3226,9 +3216,6 @@ out:
*
* On failure NULL is returned, errp is set and the reference to
* child_bs is also dropped.
- *
- * If @parent_bs and @child_bs are in different AioContexts, the caller must
- * hold the AioContext lock for @child_bs, but not for @parent_bs.
*/
BdrvChild *bdrv_attach_child(BlockDriverState *parent_bs,
BlockDriverState *child_bs,
@@ -3418,9 +3405,8 @@ static BdrvChildRole bdrv_backing_role(BlockDriverState *bs)
*
* Function doesn't update permissions, caller is responsible for this.
*
- * The caller must hold the AioContext lock for @child_bs. Both @parent_bs and
- * @child_bs can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * Both @parent_bs and @child_bs can move to a different AioContext in this
+ * function.
*
* After calling this function, the transaction @tran may only be completed
* while holding a writer lock for the graph.
@@ -3513,9 +3499,8 @@ out:
}
/*
- * The caller must hold the AioContext lock for @backing_hd. Both @bs and
- * @backing_hd can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * Both @bs and @backing_hd can move to a different AioContext in this
+ * function.
*
* If a backing child is already present (i.e. we're detaching a node), that
* child node must be drained.
@@ -3574,8 +3559,6 @@ int bdrv_set_backing_hd(BlockDriverState *bs, BlockDriverState *backing_hd,
* itself, all options starting with "${bdref_key}." are considered part of the
* BlockdevRef.
*
- * The caller must hold the main AioContext lock.
- *
* TODO Can this be unified with bdrv_open_image()?
*/
int bdrv_open_backing_file(BlockDriverState *bs, QDict *parent_options,
@@ -3745,9 +3728,7 @@ done:
*
* The BlockdevRef will be removed from the options QDict.
*
- * The caller must hold the lock of the main AioContext and no other AioContext.
- * @parent can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * @parent can move to a different AioContext in this function.
*/
BdrvChild *bdrv_open_child(const char *filename,
QDict *options, const char *bdref_key,
@@ -3778,9 +3759,7 @@ BdrvChild *bdrv_open_child(const char *filename,
/*
* Wrapper on bdrv_open_child() for most popular case: open primary child of bs.
*
- * The caller must hold the lock of the main AioContext and no other AioContext.
- * @parent can move to a different AioContext in this function. Callers must
- * make sure that their AioContext locking is still correct after this.
+ * @parent can move to a different AioContext in this function.
*/
int bdrv_open_file_child(const char *filename,
QDict *options, const char *bdref_key,
@@ -3923,8 +3902,6 @@ out:
* The reference parameter may be used to specify an existing block device which
* should be opened. If specified, neither options nor a filename may be given,
* nor can an existing BDS be reused (that is, *pbs has to be NULL).
- *
- * The caller must always hold the main AioContext lock.
*/
static BlockDriverState * no_coroutine_fn
bdrv_open_inherit(const char *filename, const char *reference, QDict *options,
@@ -4217,7 +4194,6 @@ close_and_fail:
return NULL;
}
-/* The caller must always hold the main AioContext lock. */
BlockDriverState *bdrv_open(const char *filename, const char *reference,
QDict *options, int flags, Error **errp)
{
@@ -4665,10 +4641,7 @@ int bdrv_reopen_set_read_only(BlockDriverState *bs, bool read_only,
*
* Return 0 on success, otherwise return < 0 and set @errp.
*
- * The caller must hold the AioContext lock of @reopen_state->bs.
* @reopen_state->bs can move to a different AioContext in this function.
- * Callers must make sure that their AioContext locking is still correct after
- * this.
*/
static int GRAPH_UNLOCKED
bdrv_reopen_parse_file_or_backing(BDRVReopenState *reopen_state,
@@ -4801,8 +4774,6 @@ out_rdlock:
* It is the responsibility of the caller to then call the abort() or
* commit() for any other BDS that have been left in a prepare() state
*
- * The caller must hold the AioContext lock of @reopen_state->bs.
- *
* After calling this function, the transaction @change_child_tran may only be
* completed while holding a writer lock for the graph.
*/
@@ -5437,8 +5408,6 @@ int bdrv_drop_filter(BlockDriverState *bs, Error **errp)
* child.
*
* This function does not create any image files.
- *
- * The caller must hold the AioContext lock for @bs_top.
*/
int bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top,
Error **errp)
@@ -5545,9 +5514,8 @@ static void bdrv_delete(BlockDriverState *bs)
* after the call (even on failure), so if the caller intends to reuse the
* dictionary, it needs to use qobject_ref() before calling bdrv_open.
*
- * The caller holds the AioContext lock for @bs. It must make sure that @bs
- * stays in the same AioContext, i.e. @options must not refer to nodes in a
- * different AioContext.
+ * The caller must make sure that @bs stays in the same AioContext, i.e.
+ * @options must not refer to nodes in a different AioContext.
*/
BlockDriverState *bdrv_insert_node(BlockDriverState *bs, QDict *options,
int flags, Error **errp)
@@ -7575,10 +7543,6 @@ static TransactionActionDrv set_aio_context = {
*
* Must be called from the main AioContext.
*
- * The caller must own the AioContext lock for the old AioContext of bs, but it
- * must not own the AioContext lock for new_context (unless new_context is the
- * same as the current context of bs).
- *
* @visited will accumulate all visited BdrvChild objects. The caller is
* responsible for freeing the list afterwards.
*/
@@ -7631,13 +7595,6 @@ static bool bdrv_change_aio_context(BlockDriverState *bs, AioContext *ctx,
*
* If ignore_child is not NULL, that child (and its subgraph) will not
* be touched.
- *
- * This function still requires the caller to take the bs current
- * AioContext lock, otherwise draining will fail since AIO_WAIT_WHILE
- * assumes the lock is always held if bs is in another AioContext.
- * For the same reason, it temporarily also holds the new AioContext, since
- * bdrv_drained_end calls BDRV_POLL_WHILE that assumes the lock is taken too.
- * Therefore the new AioContext lock must not be taken by the caller.
*/
int bdrv_try_change_aio_context(BlockDriverState *bs, AioContext *ctx,
BdrvChild *ignore_child, Error **errp)
@@ -7663,8 +7620,8 @@ int bdrv_try_change_aio_context(BlockDriverState *bs, AioContext *ctx,
/*
* Linear phase: go through all callbacks collected in the transaction.
- * Run all callbacks collected in the recursion to switch all nodes
- * AioContext lock (transaction commit), or undo all changes done in the
+ * Run all callbacks collected in the recursion to switch every node's
+ * AioContext (transaction commit), or undo all changes done in the
* recursion (transaction abort).
*/
diff --git a/block/block-backend.c b/block/block-backend.c
index f412bed274..209eb07528 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -390,8 +390,6 @@ BlockBackend *blk_new(AioContext *ctx, uint64_t perm, uint64_t shared_perm)
* Both sets of permissions can be changed later using blk_set_perm().
*
* Return the new BlockBackend on success, null on failure.
- *
- * Callers must hold the AioContext lock of @bs.
*/
BlockBackend *blk_new_with_bs(BlockDriverState *bs, uint64_t perm,
uint64_t shared_perm, Error **errp)
@@ -416,8 +414,6 @@ BlockBackend *blk_new_with_bs(BlockDriverState *bs, uint64_t perm,
* Just as with bdrv_open(), after having called this function the reference to
* @options belongs to the block layer (even on failure).
*
- * Called without holding an AioContext lock.
- *
* TODO: Remove @filename and @flags; it should be possible to specify a whole
* BDS tree just by specifying the @options QDict (or @reference,
* alternatively). At the time of adding this function, this is not possible,
@@ -872,8 +868,6 @@ BlockBackend *blk_by_public(BlockBackendPublic *public)
/*
* Disassociates the currently associated BlockDriverState from @blk.
- *
- * The caller must hold the AioContext lock for the BlockBackend.
*/
void blk_remove_bs(BlockBackend *blk)
{
@@ -915,8 +909,6 @@ void blk_remove_bs(BlockBackend *blk)
/*
* Associates a new BlockDriverState with @blk.
- *
- * Callers must hold the AioContext lock of @bs.
*/
int blk_insert_bs(BlockBackend *blk, BlockDriverState *bs, Error **errp)
{
diff --git a/block/export/vhost-user-blk-server.c b/block/export/vhost-user-blk-server.c
index 16f48388d3..50c358e8cd 100644
--- a/block/export/vhost-user-blk-server.c
+++ b/block/export/vhost-user-blk-server.c
@@ -278,7 +278,6 @@ static void vu_blk_exp_resize(void *opaque)
vu_config_change_msg(&vexp->vu_server.vu_dev);
}
-/* Called with vexp->export.ctx acquired */
static void vu_blk_drained_begin(void *opaque)
{
VuBlkExport *vexp = opaque;
@@ -287,7 +286,6 @@ static void vu_blk_drained_begin(void *opaque)
vhost_user_server_detach_aio_context(&vexp->vu_server);
}
-/* Called with vexp->export.blk AioContext acquired */
static void vu_blk_drained_end(void *opaque)
{
VuBlkExport *vexp = opaque;
@@ -300,8 +298,6 @@ static void vu_blk_drained_end(void *opaque)
* Ensures that bdrv_drained_begin() waits until in-flight requests complete
* and the server->co_trip coroutine has terminated. It will be restarted in
* vhost_user_server_attach_aio_context().
- *
- * Called with vexp->export.ctx acquired.
*/
static bool vu_blk_drained_poll(void *opaque)
{
diff --git a/tests/qemu-iotests/202 b/tests/qemu-iotests/202
index b784dcd791..13304242e5 100755
--- a/tests/qemu-iotests/202
+++ b/tests/qemu-iotests/202
@@ -21,7 +21,7 @@
# Check that QMP 'transaction' blockdev-snapshot-sync with multiple drives on a
# single IOThread completes successfully. This particular command triggered a
# hang due to recursive AioContext locking and BDRV_POLL_WHILE(). Protect
-# against regressions.
+# against regressions even though the AioContext lock no longer exists.
import iotests
diff --git a/tests/qemu-iotests/203 b/tests/qemu-iotests/203
index ab80fd0e44..1ba878522b 100755
--- a/tests/qemu-iotests/203
+++ b/tests/qemu-iotests/203
@@ -21,7 +21,8 @@
# Check that QMP 'migrate' with multiple drives on a single IOThread completes
# successfully. This particular command triggered a hang in the source QEMU
# process due to recursive AioContext locking in bdrv_invalidate_all() and
-# BDRV_POLL_WHILE().
+# BDRV_POLL_WHILE(). Protect against regressions even though the AioContext
+# lock no longer exists.
import iotests
--
2.42.0
^ permalink raw reply related [flat|nested] 38+ messages in thread
* Re: [PATCH 05/12] block: remove AioContext locking
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
@ 2023-11-30 9:12 ` Paul Durrant
2023-11-30 21:31 ` Eric Blake
2023-12-04 14:33 ` Kevin Wolf
2 siblings, 0 replies; 38+ messages in thread
From: Paul Durrant @ 2023-11-30 9:12 UTC (permalink / raw)
To: Stefan Hajnoczi, qemu-devel
Cc: Hanna Reitz, Paolo Bonzini, Alberto Garcia,
Emanuele Giuseppe Esposito, John Snow, Kevin Wolf, Eric Blake,
Wen Congyang, qemu-block, xen-devel, Coiby Xu, Eduardo Habkost,
Xie Changlong, Ari Sundholm, Li Zhijian, Cleber Rosa,
Juan Quintela, Michael S. Tsirkin, Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On 29/11/2023 19:55, Stefan Hajnoczi wrote:
> This is the big patch that removes
> aio_context_acquire()/aio_context_release() from the block layer and
> affected block layer users.
>
> There isn't a clean way to split this patch and the reviewers are likely
> the same group of people, so I decided to do it in one patch.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/block/block-global-state.h | 9 +-
> include/block/block-io.h | 3 +-
> include/block/snapshot.h | 2 -
> block.c | 234 +---------------------
> block/block-backend.c | 14 --
> block/copy-before-write.c | 22 +--
> block/export/export.c | 22 +--
> block/io.c | 45 +----
> block/mirror.c | 19 --
> block/monitor/bitmap-qmp-cmds.c | 20 +-
> block/monitor/block-hmp-cmds.c | 29 ---
> block/qapi-sysemu.c | 27 +--
> block/qapi.c | 18 +-
> block/raw-format.c | 5 -
> block/replication.c | 58 +-----
> block/snapshot.c | 22 +--
> block/write-threshold.c | 6 -
> blockdev.c | 307 +++++------------------------
> blockjob.c | 18 --
> hw/block/dataplane/virtio-blk.c | 10 -
> hw/block/dataplane/xen-block.c | 17 +-
> hw/block/virtio-blk.c | 45 +----
> hw/core/qdev-properties-system.c | 9 -
> job.c | 16 --
> migration/block.c | 33 +---
> migration/migration-hmp-cmds.c | 3 -
> migration/savevm.c | 22 ---
> net/colo-compare.c | 2 -
> qemu-img.c | 4 -
> qemu-io.c | 10 +-
> qemu-nbd.c | 2 -
> replay/replay-debugging.c | 4 -
> tests/unit/test-bdrv-drain.c | 51 +----
> tests/unit/test-bdrv-graph-mod.c | 6 -
> tests/unit/test-block-iothread.c | 31 ---
> tests/unit/test-blockjob.c | 137 -------------
> tests/unit/test-replication.c | 11 --
> util/async.c | 4 -
> util/vhost-user-server.c | 3 -
> scripts/block-coroutine-wrapper.py | 3 -
> tests/tsan/suppressions.tsan | 1 -
> 41 files changed, 102 insertions(+), 1202 deletions(-)
>
Reviewed-by: Paul Durrant <paul@xen.org>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
@ 2023-11-30 15:25 ` Eric Blake
2023-12-04 14:49 ` Stefan Hajnoczi
2023-12-04 12:35 ` Kevin Wolf
2023-12-04 12:46 ` Kevin Wolf
2 siblings, 1 reply; 38+ messages in thread
From: Eric Blake @ 2023-11-30 15:25 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:42PM -0500, Stefan Hajnoczi wrote:
> Protect the Task Management Function BH state with a lock. The TMF BH
> runs in the main loop thread. An IOThread might process a TMF at the
> same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
> must be protected by a lock.
>
> Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
> This avoids more locking to protect the virtqueue and SCSI layer state.
Are we trying to get this into 8.2?
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/hw/virtio/virtio-scsi.h | 3 +-
> hw/scsi/virtio-scsi.c | 62 ++++++++++++++++++++++-----------
> 2 files changed, 43 insertions(+), 22 deletions(-)
>
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 02/12] tests: remove aio_context_acquire() tests
2023-11-29 19:55 ` [PATCH 02/12] tests: remove aio_context_acquire() tests Stefan Hajnoczi
@ 2023-11-30 15:29 ` Eric Blake
2023-12-04 12:54 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 15:29 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:43PM -0500, Stefan Hajnoczi wrote:
> The aio_context_acquire() API is being removed. Drop the test case that
> calls the API.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> tests/unit/test-aio.c | 67 +------------------------------------------
> 1 file changed, 1 insertion(+), 66 deletions(-)
The rest of this series should not hold up 8.2.
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op
2023-11-29 19:55 ` [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op Stefan Hajnoczi
@ 2023-11-30 18:10 ` Eric Blake
2023-12-04 13:18 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 18:10 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:44PM -0500, Stefan Hajnoczi wrote:
> aio_context_acquire()/aio_context_release() has been replaced by
> fine-grained locking to protect state shared by multiple threads. The
> AioContext lock still plays the role of balancing locking in
> AIO_WAIT_WHILE() and many functions in QEMU either require that the
> AioContext lock is held or not held for this reason. In other words, the
> AioContext lock is purely there for consistency with itself and serves
> no real purpose anymore.
>
> Stop actually acquiring/releasing the lock in
> aio_context_acquire()/aio_context_release() so that subsequent patches
> can remove callers across the codebase incrementally.
>
> I have performed "make check" and qemu-iotests stress tests across
> x86-64, ppc64le, and aarch64 to confirm that there are no failures as a
> result of eliminating the lock.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> util/async.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/util/async.c b/util/async.c
> index 8f90ddc304..04ee83d220 100644
> --- a/util/async.c
> +++ b/util/async.c
> @@ -725,12 +725,12 @@ void aio_context_unref(AioContext *ctx)
>
> void aio_context_acquire(AioContext *ctx)
> {
> - qemu_rec_mutex_lock(&ctx->lock);
> + /* TODO remove this function */
> }
>
> void aio_context_release(AioContext *ctx)
> {
> - qemu_rec_mutex_unlock(&ctx->lock);
> + /* TODO remove this function */
> }
>
> QEMU_DEFINE_STATIC_CO_TLS(AioContext *, my_aiocontext)
> --
> 2.42.0
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 04/12] graph-lock: remove AioContext locking
2023-11-29 19:55 ` [PATCH 04/12] graph-lock: remove AioContext locking Stefan Hajnoczi
@ 2023-11-30 19:33 ` Eric Blake
2023-12-04 13:21 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 19:33 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:45PM -0500, Stefan Hajnoczi wrote:
> Stop acquiring/releasing the AioContext lock in
> bdrv_graph_wrlock()/bdrv_graph_unlock() since the lock no longer has any
> effect.
>
> The distinction between bdrv_graph_wrunlock() and
> bdrv_graph_wrunlock_ctx() becomes meaningless and they can be collapsed
> into one function.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 05/12] block: remove AioContext locking
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
2023-11-30 9:12 ` Paul Durrant
@ 2023-11-30 21:31 ` Eric Blake
2023-12-04 15:28 ` Stefan Hajnoczi
2023-12-04 14:33 ` Kevin Wolf
2 siblings, 1 reply; 38+ messages in thread
From: Eric Blake @ 2023-11-30 21:31 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:46PM -0500, Stefan Hajnoczi wrote:
> This is the big patch that removes
> aio_context_acquire()/aio_context_release() from the block layer and
> affected block layer users.
>
> There isn't a clean way to split this patch and the reviewers are likely
> the same group of people, so I decided to do it in one patch.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> +++ b/block.c
> @@ -7585,29 +7433,12 @@ void coroutine_fn bdrv_co_leave(BlockDriverState *bs, AioContext *old_ctx)
>
> void coroutine_fn bdrv_co_lock(BlockDriverState *bs)
> {
> - AioContext *ctx = bdrv_get_aio_context(bs);
> -
> - /* In the main thread, bs->aio_context won't change concurrently */
> - assert(qemu_get_current_aio_context() == qemu_get_aio_context());
> -
> - /*
> - * We're in coroutine context, so we already hold the lock of the main
> - * loop AioContext. Don't lock it twice to avoid deadlocks.
> - */
> - assert(qemu_in_coroutine());
Is this assertion worth keeping in the short term?...
> - if (ctx != qemu_get_aio_context()) {
> - aio_context_acquire(ctx);
> - }
> + /* TODO removed in next patch */
> }
...I guess I'll see in the next patch.
>
> void coroutine_fn bdrv_co_unlock(BlockDriverState *bs)
> {
> - AioContext *ctx = bdrv_get_aio_context(bs);
> -
> - assert(qemu_in_coroutine());
> - if (ctx != qemu_get_aio_context()) {
> - aio_context_release(ctx);
> - }
> + /* TODO removed in next patch */
> }
Same comment.
> +++ b/blockdev.c
> @@ -1395,7 +1352,6 @@ static void external_snapshot_action(TransactionAction *action,
> /* File name of the new image (for 'blockdev-snapshot-sync') */
> const char *new_image_file;
> ExternalSnapshotState *state = g_new0(ExternalSnapshotState, 1);
> - AioContext *aio_context;
> uint64_t perm, shared;
>
> /* TODO We'll eventually have to take a writer lock in this function */
I'm guessing removal of the locking gets us one step closer to
implementing this TODO at a later time? Or is it now a stale comment?
Either way, it doesn't affect this patch.
> +++ b/migration/block.c
> @@ -270,7 +270,6 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
>
> if (bmds->shared_base) {
> qemu_mutex_lock_iothread();
> - aio_context_acquire(blk_get_aio_context(bb));
...
> @@ -313,22 +311,10 @@ static int mig_save_device_bulk(QEMUFile *f, BlkMigDevState *bmds)
> block_mig_state.submitted++;
> blk_mig_unlock();
>
> - /* We do not know if bs is under the main thread (and thus does
> - * not acquire the AioContext when doing AIO) or rather under
> - * dataplane. Thus acquire both the iothread mutex and the
> - * AioContext.
> - *
> - * This is ugly and will disappear when we make bdrv_* thread-safe,
> - * without the need to acquire the AioContext.
> - */
> - qemu_mutex_lock_iothread();
> - aio_context_acquire(blk_get_aio_context(bmds->blk));
Will conflict, but with trivial resolution, with your other thread
renaming things to qemu_bql_lock().
> +++ b/tests/unit/test-blockjob.c
> -static void test_complete_in_standby(void)
> -{
> @@ -531,13 +402,5 @@ int main(int argc, char **argv)
> g_test_add_func("/blockjob/cancel/standby", test_cancel_standby);
> g_test_add_func("/blockjob/cancel/pending", test_cancel_pending);
> g_test_add_func("/blockjob/cancel/concluded", test_cancel_concluded);
> -
> - /*
> - * This test is flaky and sometimes fails in CI and otherwise:
> - * don't run unless user opts in via environment variable.
> - */
> - if (getenv("QEMU_TEST_FLAKY_TESTS")) {
> - g_test_add_func("/blockjob/complete_in_standby", test_complete_in_standby);
> - }
Looks like you ripped out this entire test, because it is no longer
viable. I might have mentioned it in the commit message, or squashed
the removal of this test into the earlier 02/12 patch.
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 06/12] scsi: remove AioContext locking
2023-11-29 19:55 ` [PATCH 06/12] scsi: " Stefan Hajnoczi
@ 2023-11-30 21:36 ` Eric Blake
2023-12-04 12:23 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 21:36 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:47PM -0500, Stefan Hajnoczi wrote:
> The AioContext lock no longer has any effect. Remove it.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/hw/virtio/virtio-scsi.h | 14 --------------
> hw/scsi/scsi-bus.c | 2 --
> hw/scsi/scsi-disk.c | 28 ++++------------------------
> hw/scsi/virtio-scsi.c | 18 ------------------
> 4 files changed, 4 insertions(+), 58 deletions(-)
>
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED()
2023-11-29 19:55 ` [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED() Stefan Hajnoczi
@ 2023-11-30 21:39 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 21:39 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:48PM -0500, Stefan Hajnoczi wrote:
> Now that the AioContext lock no longer exists, AIO_WAIT_WHILE() and
> AIO_WAIT_WHILE_UNLOCKED() are equivalent.
>
> A future patch will get rid of AIO_WAIT_WHILE_UNLOCKED().
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/block/aio-wait.h | 16 ++++------------
> 1 file changed, 4 insertions(+), 12 deletions(-)
>
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API
2023-11-29 19:55 ` [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API Stefan Hajnoczi
@ 2023-11-30 21:42 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 21:42 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:49PM -0500, Stefan Hajnoczi wrote:
> Delete these functions because nothing calls these functions anymore.
>
> I introduced these APIs in commit 98563fc3ec44 ("aio: add
> aio_context_acquire() and aio_context_release()") in 2014. It's with a
> sigh of relief that I delete these APIs almost 10 years later.
>
> Thanks to Paolo Bonzini's vision for multi-queue QEMU, we got an
> understanding of where the code needed to go in order to remove the
> limitations that the original dataplane and the IOThread/AioContext
> approach that followed it.
>
> Emanuele Giuseppe Esposito had the splendid determination to convert
> large parts of the codebase so that they no longer needed the AioContext
> lock. This was a painstaking process, both in the actual code changes
> required and the iterations of code review that Emanuele eeked out of
s/eeked/eked/
> Kevin and me over many months.
>
> Kevin Wolf tackled multitudes of graph locking conversions to protect
> in-flight I/O from run-time changes to the block graph as well as the
> clang Thread Safety Analysis annotations that allow the compiler to
> check whether the graph lock is being used correctly.
>
> And me, well, I'm just here to add some pizzazz to the QEMU multi-queue
> block layer :). Thank you to everyone who helped with this effort,
> including Eric Blake, code reviewer extraordinaire, and others who I've
> forgotten to mention.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/block/aio.h | 17 -----------------
> util/async.c | 10 ----------
> 2 files changed, 27 deletions(-)
>
Yay!
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 09/12] docs: remove AioContext lock from IOThread docs
2023-11-29 19:55 ` [PATCH 09/12] docs: remove AioContext lock from IOThread docs Stefan Hajnoczi
@ 2023-11-30 22:38 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 22:38 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:50PM -0500, Stefan Hajnoczi wrote:
> Encourage the use of locking primitives and stop mentioning the
> AioContext lock since it is being removed.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> docs/devel/multiple-iothreads.txt | 45 +++++++++++--------------------
> 1 file changed, 15 insertions(+), 30 deletions(-)
>
> diff --git a/docs/devel/multiple-iothreads.txt b/docs/devel/multiple-iothreads.txt
> index a3e949f6b3..4865196bde 100644
> --- a/docs/devel/multiple-iothreads.txt
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 10/12] scsi: remove outdated AioContext lock comment
2023-11-29 19:55 ` [PATCH 10/12] scsi: remove outdated AioContext lock comment Stefan Hajnoczi
@ 2023-11-30 22:39 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-11-30 22:39 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:51PM -0500, Stefan Hajnoczi wrote:
> The SCSI subsystem no longer uses the AioContext lock. Request
> processing runs exclusively in the BlockBackend's AioContext since
> "scsi: only access SCSIDevice->requests from one thread" and hence the
> lock is unnecessary.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> hw/scsi/scsi-disk.c | 1 -
> 1 file changed, 1 deletion(-)
>
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 11/12] job: remove outdated AioContext locking comments
2023-11-29 19:55 ` [PATCH 11/12] job: remove outdated AioContext locking comments Stefan Hajnoczi
@ 2023-12-01 17:50 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-12-01 17:50 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:52PM -0500, Stefan Hajnoczi wrote:
> The AioContext lock no longer exists.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/qemu/job.h | 20 --------------------
> 1 file changed, 20 deletions(-)
>
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 12/12] block: remove outdated AioContext locking comments
2023-11-29 19:55 ` [PATCH 12/12] block: " Stefan Hajnoczi
@ 2023-12-01 18:41 ` Eric Blake
0 siblings, 0 replies; 38+ messages in thread
From: Eric Blake @ 2023-12-01 18:41 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
On Wed, Nov 29, 2023 at 02:55:53PM -0500, Stefan Hajnoczi wrote:
> The AioContext lock no longer exists.
>
> There is one noteworthy change:
>
> - * More specifically, these functions use BDRV_POLL_WHILE(bs), which
> - * requires the caller to be either in the main thread and hold
> - * the BlockdriverState (bs) AioContext lock, or directly in the
> - * home thread that runs the bs AioContext. Calling them from
> - * another thread in another AioContext would cause deadlocks.
> + * More specifically, these functions use BDRV_POLL_WHILE(bs), which requires
> + * the caller to be either in the main thread or directly in the home thread
> + * that runs the bs AioContext. Calling them from another thread in another
> + * AioContext would cause deadlocks.
>
> I am not sure whether deadlocks are still possible. Maybe they have just
> moved to the fine-grained locks that have replaced the AioContext. Since
> I am not sure if the deadlocks are gone, I have kept the substance
> unchanged and just removed mention of the AioContext.
I'd rather text that may be overly conservative than an omission that
could lead to a bug; so I'm okay with your action there.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
Reviewed-by: Eric Blake <eblake@redhat.com>
--
Eric Blake, Principal Software Engineer
Red Hat, Inc.
Virtualization: qemu.org | libguestfs.org
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 06/12] scsi: remove AioContext locking
2023-11-29 19:55 ` [PATCH 06/12] scsi: " Stefan Hajnoczi
2023-11-30 21:36 ` Eric Blake
@ 2023-12-04 12:23 ` Kevin Wolf
2023-12-04 15:28 ` Stefan Hajnoczi
1 sibling, 1 reply; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 12:23 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> The AioContext lock no longer has any effect. Remove it.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> ---
> include/hw/virtio/virtio-scsi.h | 14 --------------
> hw/scsi/scsi-bus.c | 2 --
> hw/scsi/scsi-disk.c | 28 ++++------------------------
> hw/scsi/virtio-scsi.c | 18 ------------------
> 4 files changed, 4 insertions(+), 58 deletions(-)
> @@ -2531,13 +2527,11 @@ static void scsi_unrealize(SCSIDevice *dev)
> static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
> {
> SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, dev);
> - AioContext *ctx = NULL;
> +
> /* can happen for devices without drive. The error message for missing
> * backend will be issued in scsi_realize
> */
> if (s->qdev.conf.blk) {
> - ctx = blk_get_aio_context(s->qdev.conf.blk);
> - aio_context_acquire(ctx);
> if (!blkconf_blocksizes(&s->qdev.conf, errp)) {
> goto out;
> }
> @@ -2549,15 +2543,11 @@ static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
> }
> scsi_realize(&s->qdev, errp);
> out:
> - if (ctx) {
> - aio_context_release(ctx);
> - }
> }
This doesn't build for me:
../hw/scsi/scsi-disk.c:2545:1: error: label at end of compound statement is a C2x extension [-Werror,-Wc2x-extensions]
}
^
1 error generated.
Kevin
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
2023-11-30 15:25 ` Eric Blake
@ 2023-12-04 12:35 ` Kevin Wolf
2023-12-04 12:46 ` Kevin Wolf
2 siblings, 0 replies; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 12:35 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> Protect the Task Management Function BH state with a lock. The TMF BH
> runs in the main loop thread. An IOThread might process a TMF at the
> same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
> must be protected by a lock.
>
> Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
> This avoids more locking to protect the virtqueue and SCSI layer state.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
2023-11-30 15:25 ` Eric Blake
2023-12-04 12:35 ` Kevin Wolf
@ 2023-12-04 12:46 ` Kevin Wolf
2023-12-04 14:51 ` Stefan Hajnoczi
2 siblings, 1 reply; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 12:46 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> Protect the Task Management Function BH state with a lock. The TMF BH
> runs in the main loop thread. An IOThread might process a TMF at the
> same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
> must be protected by a lock.
>
> Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
> This avoids more locking to protect the virtqueue and SCSI layer state.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The second part reminds me that the implicit protection of the virtqueue
and SCSI data structures by having all accesses in a single thread is
hard to review and I think we wanted to put some assertions there to
check that we're really running in the right thread. I don't think we
have done that so far, so I suppose after this patch would be the place
in the series to add them, before we remove the protection by the
AioContext lock?
Kevin
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 02/12] tests: remove aio_context_acquire() tests
2023-11-29 19:55 ` [PATCH 02/12] tests: remove aio_context_acquire() tests Stefan Hajnoczi
2023-11-30 15:29 ` Eric Blake
@ 2023-12-04 12:54 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 12:54 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> The aio_context_acquire() API is being removed. Drop the test case that
> calls the API.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op
2023-11-29 19:55 ` [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op Stefan Hajnoczi
2023-11-30 18:10 ` Eric Blake
@ 2023-12-04 13:18 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 13:18 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> aio_context_acquire()/aio_context_release() has been replaced by
> fine-grained locking to protect state shared by multiple threads. The
> AioContext lock still plays the role of balancing locking in
> AIO_WAIT_WHILE() and many functions in QEMU either require that the
> AioContext lock is held or not held for this reason. In other words, the
> AioContext lock is purely there for consistency with itself and serves
> no real purpose anymore.
>
> Stop actually acquiring/releasing the lock in
> aio_context_acquire()/aio_context_release() so that subsequent patches
> can remove callers across the codebase incrementally.
>
> I have performed "make check" and qemu-iotests stress tests across
> x86-64, ppc64le, and aarch64 to confirm that there are no failures as a
> result of eliminating the lock.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
YOLO.
Acked-by: Kevin Wolf <kwolf@redhat.com>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 04/12] graph-lock: remove AioContext locking
2023-11-29 19:55 ` [PATCH 04/12] graph-lock: remove AioContext locking Stefan Hajnoczi
2023-11-30 19:33 ` Eric Blake
@ 2023-12-04 13:21 ` Kevin Wolf
1 sibling, 0 replies; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 13:21 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> Stop acquiring/releasing the AioContext lock in
> bdrv_graph_wrlock()/bdrv_graph_unlock() since the lock no longer has any
> effect.
>
> The distinction between bdrv_graph_wrunlock() and
> bdrv_graph_wrunlock_ctx() becomes meaningless and they can be collapsed
> into one function.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 05/12] block: remove AioContext locking
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
2023-11-30 9:12 ` Paul Durrant
2023-11-30 21:31 ` Eric Blake
@ 2023-12-04 14:33 ` Kevin Wolf
2023-12-04 15:17 ` Stefan Hajnoczi
2 siblings, 1 reply; 38+ messages in thread
From: Kevin Wolf @ 2023-12-04 14:33 UTC (permalink / raw)
To: Stefan Hajnoczi
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> This is the big patch that removes
> aio_context_acquire()/aio_context_release() from the block layer and
> affected block layer users.
>
> There isn't a clean way to split this patch and the reviewers are likely
> the same group of people, so I decided to do it in one patch.
>
> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> @@ -7585,29 +7433,12 @@ void coroutine_fn bdrv_co_leave(BlockDriverState *bs, AioContext *old_ctx)
>
> void coroutine_fn bdrv_co_lock(BlockDriverState *bs)
> {
> - AioContext *ctx = bdrv_get_aio_context(bs);
> -
> - /* In the main thread, bs->aio_context won't change concurrently */
> - assert(qemu_get_current_aio_context() == qemu_get_aio_context());
> -
> - /*
> - * We're in coroutine context, so we already hold the lock of the main
> - * loop AioContext. Don't lock it twice to avoid deadlocks.
> - */
> - assert(qemu_in_coroutine());
> - if (ctx != qemu_get_aio_context()) {
> - aio_context_acquire(ctx);
> - }
> + /* TODO removed in next patch */
> }
It's still there at the end of the series.
> void coroutine_fn bdrv_co_unlock(BlockDriverState *bs)
> {
> - AioContext *ctx = bdrv_get_aio_context(bs);
> -
> - assert(qemu_in_coroutine());
> - if (ctx != qemu_get_aio_context()) {
> - aio_context_release(ctx);
> - }
> + /* TODO removed in next patch */
> }
This one, too.
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-11-30 15:25 ` Eric Blake
@ 2023-12-04 14:49 ` Stefan Hajnoczi
0 siblings, 0 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-12-04 14:49 UTC (permalink / raw)
To: Eric Blake
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
[-- Attachment #1: Type: text/plain, Size: 1115 bytes --]
On Thu, Nov 30, 2023 at 09:25:52AM -0600, Eric Blake wrote:
> On Wed, Nov 29, 2023 at 02:55:42PM -0500, Stefan Hajnoczi wrote:
> > Protect the Task Management Function BH state with a lock. The TMF BH
> > runs in the main loop thread. An IOThread might process a TMF at the
> > same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
> > must be protected by a lock.
> >
> > Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
> > This avoids more locking to protect the virtqueue and SCSI layer state.
>
> Are we trying to get this into 8.2?
No. 8.2 still has the AioContext lock is therefore safe without this
patch.
Stefan
>
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> > ---
> > include/hw/virtio/virtio-scsi.h | 3 +-
> > hw/scsi/virtio-scsi.c | 62 ++++++++++++++++++++++-----------
> > 2 files changed, 43 insertions(+), 22 deletions(-)
> >
>
> Reviewed-by: Eric Blake <eblake@redhat.com>
>
> --
> Eric Blake, Principal Software Engineer
> Red Hat, Inc.
> Virtualization: qemu.org | libguestfs.org
>
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock
2023-12-04 12:46 ` Kevin Wolf
@ 2023-12-04 14:51 ` Stefan Hajnoczi
0 siblings, 0 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-12-04 14:51 UTC (permalink / raw)
To: Kevin Wolf
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
[-- Attachment #1: Type: text/plain, Size: 1161 bytes --]
On Mon, Dec 04, 2023 at 01:46:13PM +0100, Kevin Wolf wrote:
> Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> > Protect the Task Management Function BH state with a lock. The TMF BH
> > runs in the main loop thread. An IOThread might process a TMF at the
> > same time as the TMF BH is running. Therefore tmf_bh_list and tmf_bh
> > must be protected by a lock.
> >
> > Run TMF request completion in the IOThread using aio_wait_bh_oneshot().
> > This avoids more locking to protect the virtqueue and SCSI layer state.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
>
> The second part reminds me that the implicit protection of the virtqueue
> and SCSI data structures by having all accesses in a single thread is
> hard to review and I think we wanted to put some assertions there to
> check that we're really running in the right thread. I don't think we
> have done that so far, so I suppose after this patch would be the place
> in the series to add them, before we remove the protection by the
> AioContext lock?
Thanks for reminding me. I will add assertions in the next revision of
this series.
Stefan
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 05/12] block: remove AioContext locking
2023-12-04 14:33 ` Kevin Wolf
@ 2023-12-04 15:17 ` Stefan Hajnoczi
0 siblings, 0 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-12-04 15:17 UTC (permalink / raw)
To: Kevin Wolf
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
[-- Attachment #1: Type: text/plain, Size: 1330 bytes --]
On Mon, Dec 04, 2023 at 03:33:57PM +0100, Kevin Wolf wrote:
> Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> > This is the big patch that removes
> > aio_context_acquire()/aio_context_release() from the block layer and
> > affected block layer users.
> >
> > There isn't a clean way to split this patch and the reviewers are likely
> > the same group of people, so I decided to do it in one patch.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
>
> > @@ -7585,29 +7433,12 @@ void coroutine_fn bdrv_co_leave(BlockDriverState *bs, AioContext *old_ctx)
> >
> > void coroutine_fn bdrv_co_lock(BlockDriverState *bs)
> > {
> > - AioContext *ctx = bdrv_get_aio_context(bs);
> > -
> > - /* In the main thread, bs->aio_context won't change concurrently */
> > - assert(qemu_get_current_aio_context() == qemu_get_aio_context());
> > -
> > - /*
> > - * We're in coroutine context, so we already hold the lock of the main
> > - * loop AioContext. Don't lock it twice to avoid deadlocks.
> > - */
> > - assert(qemu_in_coroutine());
> > - if (ctx != qemu_get_aio_context()) {
> > - aio_context_acquire(ctx);
> > - }
> > + /* TODO removed in next patch */
> > }
>
> It's still there at the end of the series.
Will fix in v2. Thanks!
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 05/12] block: remove AioContext locking
2023-11-30 21:31 ` Eric Blake
@ 2023-12-04 15:28 ` Stefan Hajnoczi
0 siblings, 0 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-12-04 15:28 UTC (permalink / raw)
To: Eric Blake
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Kevin Wolf, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
[-- Attachment #1: Type: text/plain, Size: 3743 bytes --]
On Thu, Nov 30, 2023 at 03:31:37PM -0600, Eric Blake wrote:
> On Wed, Nov 29, 2023 at 02:55:46PM -0500, Stefan Hajnoczi wrote:
> > This is the big patch that removes
> > aio_context_acquire()/aio_context_release() from the block layer and
> > affected block layer users.
> >
> > There isn't a clean way to split this patch and the reviewers are likely
> > the same group of people, so I decided to do it in one patch.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> > ---
>
> > +++ b/block.c
> > @@ -7585,29 +7433,12 @@ void coroutine_fn bdrv_co_leave(BlockDriverState *bs, AioContext *old_ctx)
> >
> > void coroutine_fn bdrv_co_lock(BlockDriverState *bs)
> > {
> > - AioContext *ctx = bdrv_get_aio_context(bs);
> > -
> > - /* In the main thread, bs->aio_context won't change concurrently */
> > - assert(qemu_get_current_aio_context() == qemu_get_aio_context());
> > -
> > - /*
> > - * We're in coroutine context, so we already hold the lock of the main
> > - * loop AioContext. Don't lock it twice to avoid deadlocks.
> > - */
> > - assert(qemu_in_coroutine());
>
> Is this assertion worth keeping in the short term?...
Probably not because coroutine vs non-coroutine functions don't change
in this patch series, so it's unlikely that this will break.
>
> > - if (ctx != qemu_get_aio_context()) {
> > - aio_context_acquire(ctx);
> > - }
> > + /* TODO removed in next patch */
> > }
>
> ...I guess I'll see in the next patch.
>
> >
> > void coroutine_fn bdrv_co_unlock(BlockDriverState *bs)
> > {
> > - AioContext *ctx = bdrv_get_aio_context(bs);
> > -
> > - assert(qemu_in_coroutine());
> > - if (ctx != qemu_get_aio_context()) {
> > - aio_context_release(ctx);
> > - }
> > + /* TODO removed in next patch */
> > }
>
> Same comment.
>
> > +++ b/blockdev.c
> > @@ -1395,7 +1352,6 @@ static void external_snapshot_action(TransactionAction *action,
> > /* File name of the new image (for 'blockdev-snapshot-sync') */
> > const char *new_image_file;
> > ExternalSnapshotState *state = g_new0(ExternalSnapshotState, 1);
> > - AioContext *aio_context;
> > uint64_t perm, shared;
> >
> > /* TODO We'll eventually have to take a writer lock in this function */
>
> I'm guessing removal of the locking gets us one step closer to
> implementing this TODO at a later time? Or is it now a stale comment?
> Either way, it doesn't affect this patch.
I'm not sure. Kevin can answer questions about the graph lock.
> > +++ b/tests/unit/test-blockjob.c
>
> > -static void test_complete_in_standby(void)
> > -{
>
> > @@ -531,13 +402,5 @@ int main(int argc, char **argv)
> > g_test_add_func("/blockjob/cancel/standby", test_cancel_standby);
> > g_test_add_func("/blockjob/cancel/pending", test_cancel_pending);
> > g_test_add_func("/blockjob/cancel/concluded", test_cancel_concluded);
> > -
> > - /*
> > - * This test is flaky and sometimes fails in CI and otherwise:
> > - * don't run unless user opts in via environment variable.
> > - */
> > - if (getenv("QEMU_TEST_FLAKY_TESTS")) {
> > - g_test_add_func("/blockjob/complete_in_standby", test_complete_in_standby);
> > - }
>
> Looks like you ripped out this entire test, because it is no longer
> viable. I might have mentioned it in the commit message, or squashed
> the removal of this test into the earlier 02/12 patch.
I have sent a separate patch to remove this test and once it's merged
this hunk will disappear this patch series:
https://lore.kernel.org/qemu-devel/20231127170210.422728-1-stefanha@redhat.com/
Stefan
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 38+ messages in thread
* Re: [PATCH 06/12] scsi: remove AioContext locking
2023-12-04 12:23 ` Kevin Wolf
@ 2023-12-04 15:28 ` Stefan Hajnoczi
0 siblings, 0 replies; 38+ messages in thread
From: Stefan Hajnoczi @ 2023-12-04 15:28 UTC (permalink / raw)
To: Kevin Wolf
Cc: qemu-devel, Hanna Reitz, Paul Durrant, Paolo Bonzini,
Alberto Garcia, Emanuele Giuseppe Esposito, John Snow,
Eric Blake, Wen Congyang, qemu-block, xen-devel, Coiby Xu,
Eduardo Habkost, Xie Changlong, Ari Sundholm, Li Zhijian,
Cleber Rosa, Juan Quintela, Michael S. Tsirkin,
Daniel P. Berrangé,
Jason Wang, Vladimir Sementsov-Ogievskiy, Zhang Chen, Peter Xu,
Anthony Perard, Stefano Stabellini, Leonardo Bras,
Pavel Dovgalyuk, Fam Zheng, Fabiano Rosas
[-- Attachment #1: Type: text/plain, Size: 1653 bytes --]
On Mon, Dec 04, 2023 at 01:23:09PM +0100, Kevin Wolf wrote:
> Am 29.11.2023 um 20:55 hat Stefan Hajnoczi geschrieben:
> > The AioContext lock no longer has any effect. Remove it.
> >
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> > ---
> > include/hw/virtio/virtio-scsi.h | 14 --------------
> > hw/scsi/scsi-bus.c | 2 --
> > hw/scsi/scsi-disk.c | 28 ++++------------------------
> > hw/scsi/virtio-scsi.c | 18 ------------------
> > 4 files changed, 4 insertions(+), 58 deletions(-)
>
> > @@ -2531,13 +2527,11 @@ static void scsi_unrealize(SCSIDevice *dev)
> > static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
> > {
> > SCSIDiskState *s = DO_UPCAST(SCSIDiskState, qdev, dev);
> > - AioContext *ctx = NULL;
> > +
> > /* can happen for devices without drive. The error message for missing
> > * backend will be issued in scsi_realize
> > */
> > if (s->qdev.conf.blk) {
> > - ctx = blk_get_aio_context(s->qdev.conf.blk);
> > - aio_context_acquire(ctx);
> > if (!blkconf_blocksizes(&s->qdev.conf, errp)) {
> > goto out;
> > }
> > @@ -2549,15 +2543,11 @@ static void scsi_hd_realize(SCSIDevice *dev, Error **errp)
> > }
> > scsi_realize(&s->qdev, errp);
> > out:
> > - if (ctx) {
> > - aio_context_release(ctx);
> > - }
> > }
>
> This doesn't build for me:
>
> ../hw/scsi/scsi-disk.c:2545:1: error: label at end of compound statement is a C2x extension [-Werror,-Wc2x-extensions]
> }
> ^
> 1 error generated.
Will fix in v2. Thanks!
Stefan
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 38+ messages in thread
end of thread, other threads:[~2023-12-04 15:29 UTC | newest]
Thread overview: 38+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-11-29 19:55 [PATCH 00/12] aio: remove AioContext lock Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 01/12] virtio-scsi: replace AioContext lock with tmf_bh_lock Stefan Hajnoczi
2023-11-30 15:25 ` Eric Blake
2023-12-04 14:49 ` Stefan Hajnoczi
2023-12-04 12:35 ` Kevin Wolf
2023-12-04 12:46 ` Kevin Wolf
2023-12-04 14:51 ` Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 02/12] tests: remove aio_context_acquire() tests Stefan Hajnoczi
2023-11-30 15:29 ` Eric Blake
2023-12-04 12:54 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 03/12] aio: make aio_context_acquire()/aio_context_release() a no-op Stefan Hajnoczi
2023-11-30 18:10 ` Eric Blake
2023-12-04 13:18 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 04/12] graph-lock: remove AioContext locking Stefan Hajnoczi
2023-11-30 19:33 ` Eric Blake
2023-12-04 13:21 ` Kevin Wolf
2023-11-29 19:55 ` [PATCH 05/12] block: " Stefan Hajnoczi
2023-11-30 9:12 ` Paul Durrant
2023-11-30 21:31 ` Eric Blake
2023-12-04 15:28 ` Stefan Hajnoczi
2023-12-04 14:33 ` Kevin Wolf
2023-12-04 15:17 ` Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 06/12] scsi: " Stefan Hajnoczi
2023-11-30 21:36 ` Eric Blake
2023-12-04 12:23 ` Kevin Wolf
2023-12-04 15:28 ` Stefan Hajnoczi
2023-11-29 19:55 ` [PATCH 07/12] aio-wait: draw equivalence between AIO_WAIT_WHILE() and AIO_WAIT_WHILE_UNLOCKED() Stefan Hajnoczi
2023-11-30 21:39 ` Eric Blake
2023-11-29 19:55 ` [PATCH 08/12] aio: remove aio_context_acquire()/aio_context_release() API Stefan Hajnoczi
2023-11-30 21:42 ` Eric Blake
2023-11-29 19:55 ` [PATCH 09/12] docs: remove AioContext lock from IOThread docs Stefan Hajnoczi
2023-11-30 22:38 ` Eric Blake
2023-11-29 19:55 ` [PATCH 10/12] scsi: remove outdated AioContext lock comment Stefan Hajnoczi
2023-11-30 22:39 ` Eric Blake
2023-11-29 19:55 ` [PATCH 11/12] job: remove outdated AioContext locking comments Stefan Hajnoczi
2023-12-01 17:50 ` Eric Blake
2023-11-29 19:55 ` [PATCH 12/12] block: " Stefan Hajnoczi
2023-12-01 18:41 ` Eric Blake
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.