[PATCH] drm/qxl: Remove deprecated create_singlethread_workqueue

[PATCH] drm/qxl: Remove deprecated create_singlethread_workqueue

[Bhaktipriya Shridhar]

System workqueues have been able to handle high level of concurrency
for a long time now and there's no reason to use dedicated workqueues
just to gain concurrency. Since the workqueue in the QXL graphics device
driver is involved in freeing and processing the release ring
(workitem &qdev->gc_workqxl, maps to gc_work which calls
qxl_garbage_collect) and is not being used on a memory reclaim path,
dedicated gc_queue has been replaced with the use of system_wq.

Unlike a dedicated per-cpu workqueue created with create_workqueue(),
system_wq allows multiple work items to overlap executions even on
the same CPU; however, a per-cpu workqueue doesn't have any CPU
locality or global ordering guarantees unless the target CPU is
explicitly specified and thus the increase of local concurrency
shouldn't make any difference.

flush_work() has been called in qxl_device_fini() to ensure that there
are no pending tasks while disconnecting the driver.

Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>
---
 drivers/gpu/drm/qxl/qxl_cmd.c | 2 +-
 drivers/gpu/drm/qxl/qxl_drv.h | 1 -
 drivers/gpu/drm/qxl/qxl_kms.c | 6 +-----
 3 files changed, 2 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/qxl/qxl_cmd.c b/drivers/gpu/drm/qxl/qxl_cmd.c
index b5d4b41..04270f5 100644
--- a/drivers/gpu/drm/qxl/qxl_cmd.c
+++ b/drivers/gpu/drm/qxl/qxl_cmd.c
@@ -203,7 +203,7 @@ qxl_push_cursor_ring_release(struct qxl_device *qdev, struct qxl_release *releas
 bool qxl_queue_garbage_collect(struct qxl_device *qdev, bool flush)
 {
 	if (!qxl_check_idle(qdev->release_ring)) {
-		queue_work(qdev->gc_queue, &qdev->gc_work);
+		schedule_work(&qdev->gc_work);
 		if (flush)
 			flush_work(&qdev->gc_work);
 		return true;
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 3ad6604..8e633ca 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -321,7 +321,6 @@ struct qxl_device {
 	struct qxl_bo *current_release_bo[3];
 	int current_release_bo_offset[3];

-	struct workqueue_struct *gc_queue;
 	struct work_struct gc_work;

 	struct drm_property *hotplug_mode_update_property;
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index 2319800..144627c 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -258,7 +258,6 @@ static int qxl_device_init(struct qxl_device *qdev,
 		 (unsigned long)qdev->surfaceram_size);


-	qdev->gc_queue = create_singlethread_workqueue("qxl_gc");
 	INIT_WORK(&qdev->gc_work, qxl_gc_work);

 	return 0;
@@ -270,10 +269,7 @@ static void qxl_device_fini(struct qxl_device *qdev)
 		qxl_bo_unref(&qdev->current_release_bo[0]);
 	if (qdev->current_release_bo[1])
 		qxl_bo_unref(&qdev->current_release_bo[1]);
-	flush_workqueue(qdev->gc_queue);
-	destroy_workqueue(qdev->gc_queue);
-	qdev->gc_queue = NULL;
-
+	flush_work(&qdev->gc_work);
 	qxl_ring_free(qdev->command_ring);
 	qxl_ring_free(qdev->cursor_ring);
 	qxl_ring_free(qdev->release_ring);
--
2.1.4

Re: [PATCH] drm/qxl: Remove deprecated create_singlethread_workqueue

[Tejun Heo]

On Sat, Jul 02, 2016 at 04:32:09PM +0530, Bhaktipriya Shridhar wrote:
> System workqueues have been able to handle high level of concurrency
> for a long time now and there's no reason to use dedicated workqueues
> just to gain concurrency. Since the workqueue in the QXL graphics device
> driver is involved in freeing and processing the release ring
> (workitem &qdev->gc_workqxl, maps to gc_work which calls
> qxl_garbage_collect) and is not being used on a memory reclaim path,
> dedicated gc_queue has been replaced with the use of system_wq.
> 
> Unlike a dedicated per-cpu workqueue created with create_workqueue(),
> system_wq allows multiple work items to overlap executions even on
> the same CPU; however, a per-cpu workqueue doesn't have any CPU
> locality or global ordering guarantees unless the target CPU is
> explicitly specified and thus the increase of local concurrency
> shouldn't make any difference.
> 
> flush_work() has been called in qxl_device_fini() to ensure that there
> are no pending tasks while disconnecting the driver.
> 
> Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>

Acked-by: Tejun Heo <tj@kernel.org>

Thanks.

-- 
tejun

Re: [PATCH] drm/qxl: Remove deprecated create_singlethread_workqueue

[Daniel Vetter]

On Sat, Jul 02, 2016 at 09:44:17AM -0400, Tejun Heo wrote:
> On Sat, Jul 02, 2016 at 04:32:09PM +0530, Bhaktipriya Shridhar wrote:
> > System workqueues have been able to handle high level of concurrency
> > for a long time now and there's no reason to use dedicated workqueues
> > just to gain concurrency. Since the workqueue in the QXL graphics device
> > driver is involved in freeing and processing the release ring
> > (workitem &qdev->gc_workqxl, maps to gc_work which calls
> > qxl_garbage_collect) and is not being used on a memory reclaim path,
> > dedicated gc_queue has been replaced with the use of system_wq.
> > 
> > Unlike a dedicated per-cpu workqueue created with create_workqueue(),
> > system_wq allows multiple work items to overlap executions even on
> > the same CPU; however, a per-cpu workqueue doesn't have any CPU
> > locality or global ordering guarantees unless the target CPU is
> > explicitly specified and thus the increase of local concurrency
> > shouldn't make any difference.
> > 
> > flush_work() has been called in qxl_device_fini() to ensure that there
> > are no pending tasks while disconnecting the driver.
> > 
> > Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>
> 
> Acked-by: Tejun Heo <tj@kernel.org>

Applied to drm-misc, thanks.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch