From 29117d8ea323429138020a8de63119f15da01d83 Mon Sep 17 00:00:00 2001

From: Kevin Wolf <kwolf@redhat.com>

Date: Fri, 14 Sep 2018 10:55:25 +0200

Subject: [PATCH 34/49] blockjob: Wake up BDS when job becomes idle

RH-Author: Kevin Wolf <kwolf@redhat.com>

Message-id: <20180914105540.18077-28-kwolf@redhat.com>

Patchwork-id: 82180

O-Subject: [RHV-7.6 qemu-kvm-rhev PATCH 27/42] blockjob: Wake up BDS when job becomes idle

Bugzilla: 1601212

RH-Acked-by: John Snow <jsnow@redhat.com>

RH-Acked-by: Max Reitz <mreitz@redhat.com>

RH-Acked-by: Fam Zheng <famz@redhat.com>

In the context of draining a BDS, the .drained_poll callback of block

jobs is called. If this returns true (i.e. there is still some activity

pending), the drain operation may call aio_poll() with blocking=true to

wait for completion.

As soon as the pending activity is completed and the job finally arrives

in a quiescent state (i.e. its coroutine either yields with busy=false

or terminates), the block job must notify the aio_poll() loop to wake

up, otherwise we get a deadlock if both are running in different

threads.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>

Reviewed-by: Fam Zheng <famz@redhat.com>

Reviewed-by: Max Reitz <mreitz@redhat.com>

Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>

---

 blockjob.c               | 18 ++++++++++++++++++

 include/block/blockjob.h | 13 +++++++++++++

 include/qemu/job.h       |  3 +++

 job.c                    |  7 +++++++

 4 files changed, 41 insertions(+)

diff --git a/blockjob.c b/blockjob.c

index be5903a..8d27e8e 100644

--- a/blockjob.c

+++ b/blockjob.c

@@ -221,6 +221,22 @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs,

     return 0;

 }

+void block_job_wakeup_all_bdrv(BlockJob *job)

+{

+    GSList *l;

+

+    for (l = job->nodes; l; l = l->next) {

+        BdrvChild *c = l->data;

+        bdrv_wakeup(c->bs);

+    }

+}

+

+static void block_job_on_idle(Notifier *n, void *opaque)

+{

+    BlockJob *job = opaque;

+    block_job_wakeup_all_bdrv(job);

+}

+

 bool block_job_is_internal(BlockJob *job)

 {

     return (job->job.id == NULL);

@@ -419,6 +435,7 @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,

     job->finalize_completed_notifier.notify = block_job_event_completed;

     job->pending_notifier.notify = block_job_event_pending;

     job->ready_notifier.notify = block_job_event_ready;

+    job->idle_notifier.notify = block_job_on_idle;

     notifier_list_add(&job->job.on_finalize_cancelled,

                       &job->finalize_cancelled_notifier);

@@ -426,6 +443,7 @@ void *block_job_create(const char *job_id, const BlockJobDriver *driver,

                       &job->finalize_completed_notifier);

     notifier_list_add(&job->job.on_pending, &job->pending_notifier);

     notifier_list_add(&job->job.on_ready, &job->ready_notifier);

+    notifier_list_add(&job->job.on_idle, &job->idle_notifier);

     error_setg(&job->blocker, "block device is in use by block job: %s",

                job_type_str(&job->job));

diff --git a/include/block/blockjob.h b/include/block/blockjob.h

index 32c00b7..2290bbb 100644

--- a/include/block/blockjob.h

+++ b/include/block/blockjob.h

@@ -70,6 +70,9 @@ typedef struct BlockJob {

     /** Called when the job transitions to READY */

     Notifier ready_notifier;

+    /** Called when the job coroutine yields or terminates */

+    Notifier idle_notifier;

+

     /** BlockDriverStates that are involved in this block job */

     GSList *nodes;

 } BlockJob;

@@ -119,6 +122,16 @@ int block_job_add_bdrv(BlockJob *job, const char *name, BlockDriverState *bs,

 void block_job_remove_all_bdrv(BlockJob *job);

 /**

+ * block_job_wakeup_all_bdrv:

+ * @job: The block job

+ *

+ * Calls bdrv_wakeup() for all BlockDriverStates that have been added to the

+ * job. This function is to be called whenever child_job_drained_poll() would

+ * go from true to false to notify waiting drain requests.

+ */

+void block_job_wakeup_all_bdrv(BlockJob *job);

+

+/**

  * block_job_set_speed:

  * @job: The job to set the speed for.

  * @speed: The new value

diff --git a/include/qemu/job.h b/include/qemu/job.h

index fdaa06f..407d549 100644

--- a/include/qemu/job.h

+++ b/include/qemu/job.h

@@ -156,6 +156,9 @@ typedef struct Job {

     /** Notifiers called when the job transitions to READY */

     NotifierList on_ready;

+    /** Notifiers called when the job coroutine yields or terminates */

+    NotifierList on_idle;

+

     /** Element of the list of jobs */

     QLIST_ENTRY(Job) job_list;

diff --git a/job.c b/job.c

index db53163..5a0ccc7 100644

--- a/job.c

+++ b/job.c

@@ -397,6 +397,11 @@ static void job_event_ready(Job *job)

     notifier_list_notify(&job->on_ready, job);

 }

+static void job_event_idle(Job *job)

+{

+    notifier_list_notify(&job->on_idle, job);

+}

+

 void job_enter_cond(Job *job, bool(*fn)(Job *job))

 {

     if (!job_started(job)) {

@@ -442,6 +447,7 @@ static void coroutine_fn job_do_yield(Job *job, uint64_t ns)

         timer_mod(&job->sleep_timer, ns);

     }

     job->busy = false;

+    job_event_idle(job);

     job_unlock();

     qemu_coroutine_yield();

@@ -860,6 +866,7 @@ static void coroutine_fn job_co_entry(void *opaque)

     assert(job && job->driver && job->driver->run);

     job_pause_point(job);

     job->ret = job->driver->run(job, &job->err);

+    job_event_idle(job);

     job->deferred_to_main_loop = true;

     aio_bh_schedule_oneshot(qemu_get_aio_context(), job_exit, job);

 }

-- 

1.8.3.1