All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series
@ 2016-08-17  1:11 Sean Paul
  2016-08-17  1:11 ` [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it Sean Paul
                   ` (6 more replies)
  0 siblings, 7 replies; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

This is a follow-on set to Yakir's original PSR set here:
        https://lkml.org/lkml/2016/7/24/34
and applies to the for-next branch at:
	https://cgit.freedesktop.org/~seanpaul/dogwood

There are a few issues with the code that needed to be
shored up.
 (1) The use of mutexes instead of spinlocks caused issues calling the
     psr functions from vblank_enable/disable.
 (2) The proliferation of workers due to (1)
 (3) A bunch of races due to (2)
 (4) vblank is not enabled unless an event is requested, this breaks
     a lot of things, but most noticeable was cursor.

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood
	  instead of random on-list patches (some of which had drifted)
	- Added the "small fixes" patch to catch some nits


Sean Paul (6):
  drm/rockchip: Convert psr_list_mutex to spinlock and use it
  drm/rockchip: Don't use a delayed worker for psr state changes
  drm/rockchip: Use a spinlock to protect psr state
  drm/rockchip: A couple small fixes to psr
  drm/rockchip: Improve analogix-dp psr handling
  drm/rockchip: Enable vblank without event

 drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++--
 drivers/gpu/drm/rockchip/rockchip_drm_drv.c     |  2 +-
 drivers/gpu/drm/rockchip/rockchip_drm_drv.h     |  2 +-
 drivers/gpu/drm/rockchip/rockchip_drm_psr.c     | 90 ++++++++++++-------------
 drivers/gpu/drm/rockchip/rockchip_drm_vop.c     | 15 +++--
 5 files changed, 69 insertions(+), 59 deletions(-)

-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  2:29   ` Yakir Yang
  2016-08-17  1:11 ` [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes Sean Paul
                   ` (5 subsequent siblings)
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

This patch converts the psr_list_mutex to a spinlock and locks
all access to psr_list to avoid races (however unlikely they
were).

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next

 drivers/gpu/drm/rockchip/rockchip_drm_drv.c |  2 +-
 drivers/gpu/drm/rockchip/rockchip_drm_drv.h |  2 +-
 drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 25 ++++++++++++++++++-------
 3 files changed, 20 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
index b43fe5d9..76eaf1d 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
@@ -157,7 +157,7 @@ static int rockchip_drm_bind(struct device *dev)
 	drm_dev->dev_private = private;
 
 	INIT_LIST_HEAD(&private->psr_list);
-	mutex_init(&private->psr_list_mutex);
+	spin_lock_init(&private->psr_list_lock);
 
 	drm_mode_config_init(drm_dev);
 
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
index 9c34c9e..5c69845 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
@@ -63,7 +63,7 @@ struct rockchip_drm_private {
 	struct drm_atomic_state *state;
 
 	struct list_head psr_list;
-	struct mutex psr_list_mutex;
+	spinlock_t psr_list_lock;
 };
 
 int rockchip_register_crtc_funcs(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
index a6d3bd25..bd25273 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
@@ -45,12 +45,18 @@ static struct psr_drv *find_psr_by_crtc(struct drm_crtc *crtc)
 {
 	struct rockchip_drm_private *drm_drv = crtc->dev->dev_private;
 	struct psr_drv *psr;
+	unsigned long flags;
 
-	list_for_each_entry(psr, &drm_drv->psr_list, list)
+	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
+	list_for_each_entry(psr, &drm_drv->psr_list, list) {
 		if (psr->encoder->crtc == crtc)
-			return psr;
+			goto out;
+	}
+	psr = ERR_PTR(-ENODEV);
 
-	return ERR_PTR(-ENODEV);
+out:
+	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
+	return psr;
 }
 
 static void psr_state_work(struct work_struct *work)
@@ -173,7 +179,9 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
 {
 	struct rockchip_drm_private *drm_drv = dev->dev_private;
 	struct psr_drv *psr;
+	unsigned long flags;
 
+	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
 	list_for_each_entry(psr, &drm_drv->psr_list, list) {
 		if (psr->request_state == PSR_DISABLE)
 			continue;
@@ -183,6 +191,7 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
 
 		psr_set_state(psr, PSR_FLUSH);
 	}
+	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
 }
 EXPORT_SYMBOL(rockchip_drm_psr_flush);
 
@@ -199,6 +208,7 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
 {
 	struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
 	struct psr_drv *psr;
+	unsigned long flags;
 
 	if (!encoder || !psr_set)
 		return -EINVAL;
@@ -215,9 +225,9 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
 	psr->encoder = encoder;
 	psr->set = psr_set;
 
-	mutex_lock(&drm_drv->psr_list_mutex);
+	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
 	list_add_tail(&psr->list, &drm_drv->psr_list);
-	mutex_unlock(&drm_drv->psr_list_mutex);
+	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
 
 	return 0;
 }
@@ -235,8 +245,9 @@ void rockchip_drm_psr_unregister(struct drm_encoder *encoder)
 {
 	struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
 	struct psr_drv *psr, *n;
+	unsigned long flags;
 
-	mutex_lock(&drm_drv->psr_list_mutex);
+	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
 	list_for_each_entry_safe(psr, n, &drm_drv->psr_list, list) {
 		if (psr->encoder == encoder) {
 			del_timer(&psr->flush_timer);
@@ -244,6 +255,6 @@ void rockchip_drm_psr_unregister(struct drm_encoder *encoder)
 			kfree(psr);
 		}
 	}
-	mutex_unlock(&drm_drv->psr_list_mutex);
+	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
 }
 EXPORT_SYMBOL(rockchip_drm_psr_unregister);
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
  2016-08-17  1:11 ` [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  2:31   ` Yakir Yang
  2016-08-17  1:11 ` [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state Sean Paul
                   ` (4 subsequent siblings)
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

The delayed worker isn't needed and is racey. Remove it and do
the state change in line.

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next

 drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 38 ++++++++---------------------
 1 file changed, 10 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
index bd25273..4c645d7 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
@@ -19,7 +19,6 @@
 #include "rockchip_drm_psr.h"
 
 #define PSR_FLUSH_TIMEOUT	msecs_to_jiffies(3000) /* 3 seconds */
-#define PSR_SET_DELAY_TIME	msecs_to_jiffies(10)
 
 enum psr_state {
 	PSR_FLUSH,
@@ -31,11 +30,8 @@ struct psr_drv {
 	struct list_head	list;
 	struct drm_encoder	*encoder;
 
-	enum psr_state		request_state;
 	enum psr_state		state;
 
-	struct delayed_work	state_work;
-
 	struct timer_list	flush_timer;
 
 	void (*set)(struct drm_encoder *encoder, bool enable);
@@ -59,11 +55,8 @@ out:
 	return psr;
 }
 
-static void psr_state_work(struct work_struct *work)
+static void psr_set_state(struct psr_drv *psr, enum psr_state state)
 {
-	struct psr_drv *psr = container_of(work, typeof(*psr), state_work.work);
-	enum psr_state request_state = psr->request_state;
-
 	/*
 	 * Allowed finite state machine:
 	 *
@@ -75,24 +68,22 @@ static void psr_state_work(struct work_struct *work)
 	 */
 
 	/* Forbid no state change */
-	if (request_state == psr->state)
+	if (state == psr->state)
 		return;
 
 	/* Forbid DISABLE change to FLUSH */
-	if (request_state == PSR_FLUSH && psr->state == PSR_DISABLE)
+	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
 		return;
 
+	/* Only wrote in this work, no need lock protection */
+	psr->state = state;
+
 	/* Allow but no need hardware change, just need assign the state */
-	if (request_state == PSR_DISABLE && psr->state == PSR_FLUSH) {
-		psr->state = request_state;
+	if (state == PSR_DISABLE && psr->state == PSR_FLUSH)
 		return;
-	}
-
-	/* Only wrote in this work, no need lock protection */
-	psr->state = request_state;
 
 	/* Refact to hardware state change */
-	switch (request_state) {
+	switch (psr->state) {
 	case PSR_ENABLE:
 		psr->set(psr->encoder, true);
 		break;
@@ -104,13 +95,6 @@ static void psr_state_work(struct work_struct *work)
 	}
 }
 
-static void psr_set_state(struct psr_drv *psr, enum psr_state state)
-{
-	psr->request_state = state;
-
-	schedule_delayed_work(&psr->state_work, PSR_SET_DELAY_TIME);
-}
-
 static void psr_flush_handler(unsigned long data)
 {
 	struct psr_drv *psr = (struct psr_drv *)data;
@@ -119,7 +103,7 @@ static void psr_flush_handler(unsigned long data)
 		return;
 
 	/* State changed between flush time, then keep it */
-	if (psr->request_state != PSR_FLUSH)
+	if (psr->state != PSR_FLUSH)
 		return;
 
 	psr_set_state(psr, PSR_ENABLE);
@@ -183,7 +167,7 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
 
 	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
 	list_for_each_entry(psr, &drm_drv->psr_list, list) {
-		if (psr->request_state == PSR_DISABLE)
+		if (psr->state == PSR_DISABLE)
 			continue;
 
 		mod_timer(&psr->flush_timer,
@@ -219,8 +203,6 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
 
 	setup_timer(&psr->flush_timer, psr_flush_handler, (unsigned long)psr);
 
-	INIT_DELAYED_WORK(&psr->state_work, psr_state_work);
-
 	psr->state = PSR_DISABLE;
 	psr->encoder = encoder;
 	psr->set = psr_set;
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
  2016-08-17  1:11 ` [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it Sean Paul
  2016-08-17  1:11 ` [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  2:33   ` Yakir Yang
  2016-08-17  1:11 ` [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr Sean Paul
                   ` (3 subsequent siblings)
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

The handling of psr state is racey, shore that up with
a per-psr driver lock.

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next

 drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
index 4c645d7..5bd54f2 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
@@ -30,6 +30,7 @@ struct psr_drv {
 	struct list_head	list;
 	struct drm_encoder	*encoder;
 
+	spinlock_t		lock;
 	enum psr_state		state;
 
 	struct timer_list	flush_timer;
@@ -55,7 +56,7 @@ out:
 	return psr;
 }
 
-static void psr_set_state(struct psr_drv *psr, enum psr_state state)
+static void psr_set_state_locked(struct psr_drv *psr, enum psr_state state)
 {
 	/*
 	 * Allowed finite state machine:
@@ -75,7 +76,6 @@ static void psr_set_state(struct psr_drv *psr, enum psr_state state)
 	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
 		return;
 
-	/* Only wrote in this work, no need lock protection */
 	psr->state = state;
 
 	/* Allow but no need hardware change, just need assign the state */
@@ -95,18 +95,28 @@ static void psr_set_state(struct psr_drv *psr, enum psr_state state)
 	}
 }
 
+static void psr_set_state(struct psr_drv *psr, enum psr_state state)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&psr->lock, flags);
+	psr_set_state_locked(psr, state);
+	spin_unlock_irqrestore(&psr->lock, flags);
+}
+
 static void psr_flush_handler(unsigned long data)
 {
 	struct psr_drv *psr = (struct psr_drv *)data;
+	unsigned long flags;
 
 	if (!psr)
 		return;
 
 	/* State changed between flush time, then keep it */
-	if (psr->state != PSR_FLUSH)
-		return;
-
-	psr_set_state(psr, PSR_ENABLE);
+	spin_lock_irqsave(&psr->lock, flags);
+	if (psr->state == PSR_FLUSH)
+		psr_set_state_locked(psr, PSR_ENABLE);
+	spin_unlock_irqrestore(&psr->lock, flags);
 }
 
 /**
@@ -167,9 +177,6 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
 
 	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
 	list_for_each_entry(psr, &drm_drv->psr_list, list) {
-		if (psr->state == PSR_DISABLE)
-			continue;
-
 		mod_timer(&psr->flush_timer,
 			  round_jiffies_up(jiffies + PSR_FLUSH_TIMEOUT));
 
@@ -202,6 +209,7 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
 		return -ENOMEM;
 
 	setup_timer(&psr->flush_timer, psr_flush_handler, (unsigned long)psr);
+	spin_lock_init(&psr->lock);
 
 	psr->state = PSR_DISABLE;
 	psr->encoder = encoder;
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
                   ` (2 preceding siblings ...)
  2016-08-17  1:11 ` [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  2:34   ` Yakir Yang
  2016-08-17  1:11 ` [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling Sean Paul
                   ` (2 subsequent siblings)
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

A few things that need tidying up, no functional changes.

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Introduced

 drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
index 5bd54f2..c6ac5d0 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
@@ -62,27 +62,25 @@ static void psr_set_state_locked(struct psr_drv *psr, enum psr_state state)
 	 * Allowed finite state machine:
 	 *
 	 *   PSR_ENABLE  < = = = = = >  PSR_FLUSH
-	  *      | ^                        |
-	  *      | |                        |
-	  *      v |                        |
+	 *       | ^                        |
+	 *       | |                        |
+	 *       v |                        |
 	 *   PSR_DISABLE < - - - - - - - - -
 	 */
-
-	/* Forbid no state change */
 	if (state == psr->state)
 		return;
 
-	/* Forbid DISABLE change to FLUSH */
+	/* Requesting a flush when disabled is a noop */
 	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
 		return;
 
 	psr->state = state;
 
-	/* Allow but no need hardware change, just need assign the state */
+	/* Already disabled in flush, change the state, but not the hardware */
 	if (state == PSR_DISABLE && psr->state == PSR_FLUSH)
 		return;
 
-	/* Refact to hardware state change */
+	/* Actually commit the state change to hardware */
 	switch (psr->state) {
 	case PSR_ENABLE:
 		psr->set(psr->encoder, true);
@@ -109,10 +107,7 @@ static void psr_flush_handler(unsigned long data)
 	struct psr_drv *psr = (struct psr_drv *)data;
 	unsigned long flags;
 
-	if (!psr)
-		return;
-
-	/* State changed between flush time, then keep it */
+	/* If the state has changed since we initiated the flush, do nothing */
 	spin_lock_irqsave(&psr->lock, flags);
 	if (psr->state == PSR_FLUSH)
 		psr_set_state_locked(psr, PSR_ENABLE);
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
                   ` (3 preceding siblings ...)
  2016-08-17  1:11 ` [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  2:35   ` Yakir Yang
  2016-08-17  1:11 ` [PATCH v2 6/6] drm/rockchip: Enable vblank without event Sean Paul
  2016-08-17  2:41 ` [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Yakir Yang
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

Remove the delayed worker, opting instead for the non-delayed
variety. Also introduce a lock to ensure we don't have races
with the worker and psr_state. Finally, cancel and wait for
the worker to finish when disabling the bridge.

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next

 drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
index d6d0751..439b933 100644
--- a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
@@ -42,7 +42,6 @@
 
 #define HIWORD_UPDATE(val, mask)	(val | (mask) << 16)
 
-#define PSR_SET_DELAY_TIME		msecs_to_jiffies(10)
 #define PSR_WAIT_LINE_FLAG_TIMEOUT_MS	100
 
 #define to_dp(nm)	container_of(nm, struct rockchip_dp_device, nm)
@@ -72,7 +71,8 @@ struct rockchip_dp_device {
 	struct regmap            *grf;
 	struct reset_control     *rst;
 
-	struct delayed_work      psr_work;
+	struct work_struct	 psr_work;
+	spinlock_t		 psr_lock;
 	unsigned int             psr_state;
 
 	const struct rockchip_dp_chip_data *data;
@@ -83,25 +83,29 @@ struct rockchip_dp_device {
 static void analogix_dp_psr_set(struct drm_encoder *encoder, bool enabled)
 {
 	struct rockchip_dp_device *dp = to_dp(encoder);
+	unsigned long flags;
 
 	dev_dbg(dp->dev, "%s PSR...\n", enabled ? "Entry" : "Exit");
 
+	spin_lock_irqsave(&dp->psr_lock, flags);
 	if (enabled)
 		dp->psr_state = EDP_VSC_PSR_STATE_ACTIVE;
 	else
 		dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
 
-	schedule_delayed_work(&dp->psr_work, PSR_SET_DELAY_TIME);
+	schedule_work(&dp->psr_work);
+	spin_unlock_irqrestore(&dp->psr_lock, flags);
 }
 
 static void analogix_dp_psr_work(struct work_struct *work)
 {
 	struct rockchip_dp_device *dp =
-				container_of(work, typeof(*dp), psr_work.work);
+				container_of(work, typeof(*dp), psr_work);
 	struct drm_crtc *crtc = dp->encoder.crtc;
 	int psr_state = dp->psr_state;
 	int vact_end;
 	int ret;
+	unsigned long flags;
 
 	if (!crtc)
 		return;
@@ -115,10 +119,12 @@ static void analogix_dp_psr_work(struct work_struct *work)
 		return;
 	}
 
+	spin_lock_irqsave(&dp->psr_lock, flags);
 	if (psr_state == EDP_VSC_PSR_STATE_ACTIVE)
 		analogix_dp_enable_psr(dp->dev);
 	else
 		analogix_dp_disable_psr(dp->dev);
+	spin_unlock_irqrestore(&dp->psr_lock, flags);
 }
 
 static int rockchip_dp_pre_init(struct rockchip_dp_device *dp)
@@ -135,6 +141,8 @@ static int rockchip_dp_poweron(struct analogix_dp_plat_data *plat_data)
 	struct rockchip_dp_device *dp = to_dp(plat_data);
 	int ret;
 
+	cancel_work_sync(&dp->psr_work);
+
 	ret = clk_prepare_enable(dp->pclk);
 	if (ret < 0) {
 		dev_err(dp->dev, "failed to enable pclk %d\n", ret);
@@ -390,8 +398,9 @@ static int rockchip_dp_bind(struct device *dev, struct device *master,
 	dp->plat_data.power_off = rockchip_dp_powerdown;
 	dp->plat_data.get_modes = rockchip_dp_get_modes;
 
+	spin_lock_init(&dp->psr_lock);
 	dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
-	INIT_DELAYED_WORK(&dp->psr_work, analogix_dp_psr_work);
+	INIT_WORK(&dp->psr_work, analogix_dp_psr_work);
 
 	rockchip_drm_psr_register(&dp->encoder, analogix_dp_psr_set);
 
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* [PATCH v2 6/6] drm/rockchip: Enable vblank without event
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
                   ` (4 preceding siblings ...)
  2016-08-17  1:11 ` [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling Sean Paul
@ 2016-08-17  1:11 ` Sean Paul
  2016-08-17  3:27   ` Yakir Yang
  2016-08-17  2:41 ` [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Yakir Yang
  6 siblings, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  1:11 UTC (permalink / raw)
  To: mark.yao, dri-devel, ykk; +Cc: marcheu, dianders, linux-rockchip

vblank should be enabled regardless of whether an event
is expected back. This is especially important for a cursor
plane.

Signed-off-by: Sean Paul <seanpaul@chromium.org>
---

Changes in v2:
	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next

 drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
index d1e0e06..1787084 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -112,6 +112,7 @@ struct vop {
 	struct device *dev;
 	struct drm_device *drm_dev;
 	bool is_enabled;
+	bool vblank_active;
 
 	/* mutex vsync_ work */
 	struct mutex vsync_mutex;
@@ -1107,10 +1108,11 @@ static void vop_crtc_atomic_begin(struct drm_crtc *crtc,
 	struct vop *vop = to_vop(crtc);
 
 	spin_lock_irq(&crtc->dev->event_lock);
-	if (crtc->state->event) {
-		WARN_ON(drm_crtc_vblank_get(crtc) != 0);
-		WARN_ON(vop->event);
+	vop->vblank_active = true;
+	WARN_ON(drm_crtc_vblank_get(crtc) != 0);
+	WARN_ON(vop->event);
 
+	if (crtc->state->event) {
 		vop->event = crtc->state->event;
 		crtc->state->event = NULL;
 	}
@@ -1197,12 +1199,14 @@ static void vop_handle_vblank(struct vop *vop)
 
 	spin_lock_irqsave(&drm->event_lock, flags);
 	if (vop->event) {
-
 		drm_crtc_send_vblank_event(crtc, vop->event);
-		drm_crtc_vblank_put(crtc);
 		vop->event = NULL;
 
 	}
+	if (vop->vblank_active) {
+		vop->vblank_active = false;
+		drm_crtc_vblank_put(crtc);
+	}
 	spin_unlock_irqrestore(&drm->event_lock, flags);
 
 	if (!completion_done(&vop->wait_update_complete))
@@ -1472,6 +1476,7 @@ static int vop_initial(struct vop *vop)
 	clk_disable(vop->aclk);
 
 	vop->is_enabled = false;
+	vop->vblank_active = false;
 
 	return 0;
 
-- 
2.8.0.rc3.226.g39d4020

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply related	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it
  2016-08-17  1:11 ` [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it Sean Paul
@ 2016-08-17  2:29   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:29 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip


On 08/17/2016 09:11 AM, Sean Paul wrote:
> This patch converts the psr_list_mutex to a spinlock and locks
> all access to psr_list to avoid races (however unlikely they
> were).
>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Yakir Yang <ykk@rock-chips.com>
> ---
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next
>
>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c |  2 +-
>   drivers/gpu/drm/rockchip/rockchip_drm_drv.h |  2 +-
>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 25 ++++++++++++++++++-------
>   3 files changed, 20 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
> index b43fe5d9..76eaf1d 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.c
> @@ -157,7 +157,7 @@ static int rockchip_drm_bind(struct device *dev)
>   	drm_dev->dev_private = private;
>   
>   	INIT_LIST_HEAD(&private->psr_list);
> -	mutex_init(&private->psr_list_mutex);
> +	spin_lock_init(&private->psr_list_lock);
>   
>   	drm_mode_config_init(drm_dev);
>   
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
> index 9c34c9e..5c69845 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
> @@ -63,7 +63,7 @@ struct rockchip_drm_private {
>   	struct drm_atomic_state *state;
>   
>   	struct list_head psr_list;
> -	struct mutex psr_list_mutex;
> +	spinlock_t psr_list_lock;
>   };
>   
>   int rockchip_register_crtc_funcs(struct drm_crtc *crtc,
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> index a6d3bd25..bd25273 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> @@ -45,12 +45,18 @@ static struct psr_drv *find_psr_by_crtc(struct drm_crtc *crtc)
>   {
>   	struct rockchip_drm_private *drm_drv = crtc->dev->dev_private;
>   	struct psr_drv *psr;
> +	unsigned long flags;
>   
> -	list_for_each_entry(psr, &drm_drv->psr_list, list)
> +	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
> +	list_for_each_entry(psr, &drm_drv->psr_list, list) {
>   		if (psr->encoder->crtc == crtc)
> -			return psr;
> +			goto out;
> +	}
> +	psr = ERR_PTR(-ENODEV);
>   
> -	return ERR_PTR(-ENODEV);
> +out:
> +	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
> +	return psr;
>   }
>   
>   static void psr_state_work(struct work_struct *work)
> @@ -173,7 +179,9 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
>   {
>   	struct rockchip_drm_private *drm_drv = dev->dev_private;
>   	struct psr_drv *psr;
> +	unsigned long flags;
>   
> +	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
>   	list_for_each_entry(psr, &drm_drv->psr_list, list) {
>   		if (psr->request_state == PSR_DISABLE)
>   			continue;
> @@ -183,6 +191,7 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
>   
>   		psr_set_state(psr, PSR_FLUSH);
>   	}
> +	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
>   }
>   EXPORT_SYMBOL(rockchip_drm_psr_flush);
>   
> @@ -199,6 +208,7 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
>   {
>   	struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
>   	struct psr_drv *psr;
> +	unsigned long flags;
>   
>   	if (!encoder || !psr_set)
>   		return -EINVAL;
> @@ -215,9 +225,9 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
>   	psr->encoder = encoder;
>   	psr->set = psr_set;
>   
> -	mutex_lock(&drm_drv->psr_list_mutex);
> +	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
>   	list_add_tail(&psr->list, &drm_drv->psr_list);
> -	mutex_unlock(&drm_drv->psr_list_mutex);
> +	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
>   
>   	return 0;
>   }
> @@ -235,8 +245,9 @@ void rockchip_drm_psr_unregister(struct drm_encoder *encoder)
>   {
>   	struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
>   	struct psr_drv *psr, *n;
> +	unsigned long flags;
>   
> -	mutex_lock(&drm_drv->psr_list_mutex);
> +	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
>   	list_for_each_entry_safe(psr, n, &drm_drv->psr_list, list) {
>   		if (psr->encoder == encoder) {
>   			del_timer(&psr->flush_timer);
> @@ -244,6 +255,6 @@ void rockchip_drm_psr_unregister(struct drm_encoder *encoder)
>   			kfree(psr);
>   		}
>   	}
> -	mutex_unlock(&drm_drv->psr_list_mutex);
> +	spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
>   }
>   EXPORT_SYMBOL(rockchip_drm_psr_unregister);


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes
  2016-08-17  1:11 ` [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes Sean Paul
@ 2016-08-17  2:31   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:31 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip


On 08/17/2016 09:11 AM, Sean Paul wrote:
> The delayed worker isn't needed and is racey. Remove it and do
> the state change in line.
>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Yakir Yang <ykk@rock-chips.com>
Tested-by: Yakir Yang <ykk@rock-chips.com>
> ---
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next
>
>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 38 ++++++++---------------------
>   1 file changed, 10 insertions(+), 28 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> index bd25273..4c645d7 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> @@ -19,7 +19,6 @@
>   #include "rockchip_drm_psr.h"
>   
>   #define PSR_FLUSH_TIMEOUT	msecs_to_jiffies(3000) /* 3 seconds */
> -#define PSR_SET_DELAY_TIME	msecs_to_jiffies(10)
>   
>   enum psr_state {
>   	PSR_FLUSH,
> @@ -31,11 +30,8 @@ struct psr_drv {
>   	struct list_head	list;
>   	struct drm_encoder	*encoder;
>   
> -	enum psr_state		request_state;
>   	enum psr_state		state;
>   
> -	struct delayed_work	state_work;
> -
>   	struct timer_list	flush_timer;
>   
>   	void (*set)(struct drm_encoder *encoder, bool enable);
> @@ -59,11 +55,8 @@ out:
>   	return psr;
>   }
>   
> -static void psr_state_work(struct work_struct *work)
> +static void psr_set_state(struct psr_drv *psr, enum psr_state state)
>   {
> -	struct psr_drv *psr = container_of(work, typeof(*psr), state_work.work);
> -	enum psr_state request_state = psr->request_state;
> -
>   	/*
>   	 * Allowed finite state machine:
>   	 *
> @@ -75,24 +68,22 @@ static void psr_state_work(struct work_struct *work)
>   	 */
>   
>   	/* Forbid no state change */
> -	if (request_state == psr->state)
> +	if (state == psr->state)
>   		return;
>   
>   	/* Forbid DISABLE change to FLUSH */
> -	if (request_state == PSR_FLUSH && psr->state == PSR_DISABLE)
> +	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
>   		return;
>   
> +	/* Only wrote in this work, no need lock protection */
> +	psr->state = state;
> +
>   	/* Allow but no need hardware change, just need assign the state */
> -	if (request_state == PSR_DISABLE && psr->state == PSR_FLUSH) {
> -		psr->state = request_state;
> +	if (state == PSR_DISABLE && psr->state == PSR_FLUSH)
>   		return;
> -	}
> -
> -	/* Only wrote in this work, no need lock protection */
> -	psr->state = request_state;
>   
>   	/* Refact to hardware state change */
> -	switch (request_state) {
> +	switch (psr->state) {
>   	case PSR_ENABLE:
>   		psr->set(psr->encoder, true);
>   		break;
> @@ -104,13 +95,6 @@ static void psr_state_work(struct work_struct *work)
>   	}
>   }
>   
> -static void psr_set_state(struct psr_drv *psr, enum psr_state state)
> -{
> -	psr->request_state = state;
> -
> -	schedule_delayed_work(&psr->state_work, PSR_SET_DELAY_TIME);
> -}
> -
>   static void psr_flush_handler(unsigned long data)
>   {
>   	struct psr_drv *psr = (struct psr_drv *)data;
> @@ -119,7 +103,7 @@ static void psr_flush_handler(unsigned long data)
>   		return;
>   
>   	/* State changed between flush time, then keep it */
> -	if (psr->request_state != PSR_FLUSH)
> +	if (psr->state != PSR_FLUSH)
>   		return;
>   
>   	psr_set_state(psr, PSR_ENABLE);
> @@ -183,7 +167,7 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
>   
>   	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
>   	list_for_each_entry(psr, &drm_drv->psr_list, list) {
> -		if (psr->request_state == PSR_DISABLE)
> +		if (psr->state == PSR_DISABLE)
>   			continue;
>   
>   		mod_timer(&psr->flush_timer,
> @@ -219,8 +203,6 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
>   
>   	setup_timer(&psr->flush_timer, psr_flush_handler, (unsigned long)psr);
>   
> -	INIT_DELAYED_WORK(&psr->state_work, psr_state_work);
> -
>   	psr->state = PSR_DISABLE;
>   	psr->encoder = encoder;
>   	psr->set = psr_set;


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state
  2016-08-17  1:11 ` [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state Sean Paul
@ 2016-08-17  2:33   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:33 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip


On 08/17/2016 09:11 AM, Sean Paul wrote:
> The handling of psr state is racey, shore that up with
> a per-psr driver lock.
>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Yakir Yang <ykk@rock-chips.com>
> ---
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next
>
>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 26 +++++++++++++++++---------
>   1 file changed, 17 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> index 4c645d7..5bd54f2 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> @@ -30,6 +30,7 @@ struct psr_drv {
>   	struct list_head	list;
>   	struct drm_encoder	*encoder;
>   
> +	spinlock_t		lock;
>   	enum psr_state		state;
>   
>   	struct timer_list	flush_timer;
> @@ -55,7 +56,7 @@ out:
>   	return psr;
>   }
>   
> -static void psr_set_state(struct psr_drv *psr, enum psr_state state)
> +static void psr_set_state_locked(struct psr_drv *psr, enum psr_state state)
>   {
>   	/*
>   	 * Allowed finite state machine:
> @@ -75,7 +76,6 @@ static void psr_set_state(struct psr_drv *psr, enum psr_state state)
>   	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
>   		return;
>   
> -	/* Only wrote in this work, no need lock protection */
>   	psr->state = state;
>   
>   	/* Allow but no need hardware change, just need assign the state */
> @@ -95,18 +95,28 @@ static void psr_set_state(struct psr_drv *psr, enum psr_state state)
>   	}
>   }
>   
> +static void psr_set_state(struct psr_drv *psr, enum psr_state state)
> +{
> +	unsigned long flags;
> +
> +	spin_lock_irqsave(&psr->lock, flags);
> +	psr_set_state_locked(psr, state);
> +	spin_unlock_irqrestore(&psr->lock, flags);
> +}
> +
>   static void psr_flush_handler(unsigned long data)
>   {
>   	struct psr_drv *psr = (struct psr_drv *)data;
> +	unsigned long flags;
>   
>   	if (!psr)
>   		return;
>   
>   	/* State changed between flush time, then keep it */
> -	if (psr->state != PSR_FLUSH)
> -		return;
> -
> -	psr_set_state(psr, PSR_ENABLE);
> +	spin_lock_irqsave(&psr->lock, flags);
> +	if (psr->state == PSR_FLUSH)
> +		psr_set_state_locked(psr, PSR_ENABLE);
> +	spin_unlock_irqrestore(&psr->lock, flags);
>   }
>   
>   /**
> @@ -167,9 +177,6 @@ void rockchip_drm_psr_flush(struct drm_device *dev)
>   
>   	spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
>   	list_for_each_entry(psr, &drm_drv->psr_list, list) {
> -		if (psr->state == PSR_DISABLE)
> -			continue;
> -
>   		mod_timer(&psr->flush_timer,
>   			  round_jiffies_up(jiffies + PSR_FLUSH_TIMEOUT));
>   
> @@ -202,6 +209,7 @@ int rockchip_drm_psr_register(struct drm_encoder *encoder,
>   		return -ENOMEM;
>   
>   	setup_timer(&psr->flush_timer, psr_flush_handler, (unsigned long)psr);
> +	spin_lock_init(&psr->lock);
>   
>   	psr->state = PSR_DISABLE;
>   	psr->encoder = encoder;


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr
  2016-08-17  1:11 ` [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr Sean Paul
@ 2016-08-17  2:34   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:34 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip


On 08/17/2016 09:11 AM, Sean Paul wrote:
> A few things that need tidying up, no functional changes.
>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Yakir Yang <ykk@rock-chips.com>
> ---
>
> Changes in v2:
> 	- Introduced
>
>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c | 19 +++++++------------
>   1 file changed, 7 insertions(+), 12 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> index 5bd54f2..c6ac5d0 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_psr.c
> @@ -62,27 +62,25 @@ static void psr_set_state_locked(struct psr_drv *psr, enum psr_state state)
>   	 * Allowed finite state machine:
>   	 *
>   	 *   PSR_ENABLE  < = = = = = >  PSR_FLUSH
> -	  *      | ^                        |
> -	  *      | |                        |
> -	  *      v |                        |
> +	 *       | ^                        |
> +	 *       | |                        |
> +	 *       v |                        |
>   	 *   PSR_DISABLE < - - - - - - - - -
>   	 */
> -
> -	/* Forbid no state change */
>   	if (state == psr->state)
>   		return;
>   
> -	/* Forbid DISABLE change to FLUSH */
> +	/* Requesting a flush when disabled is a noop */
>   	if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
>   		return;
>   
>   	psr->state = state;
>   
> -	/* Allow but no need hardware change, just need assign the state */
> +	/* Already disabled in flush, change the state, but not the hardware */
>   	if (state == PSR_DISABLE && psr->state == PSR_FLUSH)
>   		return;
>   
> -	/* Refact to hardware state change */
> +	/* Actually commit the state change to hardware */
>   	switch (psr->state) {
>   	case PSR_ENABLE:
>   		psr->set(psr->encoder, true);
> @@ -109,10 +107,7 @@ static void psr_flush_handler(unsigned long data)
>   	struct psr_drv *psr = (struct psr_drv *)data;
>   	unsigned long flags;
>   
> -	if (!psr)
> -		return;
> -
> -	/* State changed between flush time, then keep it */
> +	/* If the state has changed since we initiated the flush, do nothing */
>   	spin_lock_irqsave(&psr->lock, flags);
>   	if (psr->state == PSR_FLUSH)
>   		psr_set_state_locked(psr, PSR_ENABLE);


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling
  2016-08-17  1:11 ` [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling Sean Paul
@ 2016-08-17  2:35   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:35 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip


On 08/17/2016 09:11 AM, Sean Paul wrote:
> Remove the delayed worker, opting instead for the non-delayed
> variety. Also introduce a lock to ensure we don't have races
> with the worker and psr_state. Finally, cancel and wait for
> the worker to finish when disabling the bridge.
>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Yakir Yang <ykk@rock-chips.com>
> ---
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next
>
>   drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++++++++++++-----
>   1 file changed, 14 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
> index d6d0751..439b933 100644
> --- a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
> +++ b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
> @@ -42,7 +42,6 @@
>   
>   #define HIWORD_UPDATE(val, mask)	(val | (mask) << 16)
>   
> -#define PSR_SET_DELAY_TIME		msecs_to_jiffies(10)
>   #define PSR_WAIT_LINE_FLAG_TIMEOUT_MS	100
>   
>   #define to_dp(nm)	container_of(nm, struct rockchip_dp_device, nm)
> @@ -72,7 +71,8 @@ struct rockchip_dp_device {
>   	struct regmap            *grf;
>   	struct reset_control     *rst;
>   
> -	struct delayed_work      psr_work;
> +	struct work_struct	 psr_work;
> +	spinlock_t		 psr_lock;
>   	unsigned int             psr_state;
>   
>   	const struct rockchip_dp_chip_data *data;
> @@ -83,25 +83,29 @@ struct rockchip_dp_device {
>   static void analogix_dp_psr_set(struct drm_encoder *encoder, bool enabled)
>   {
>   	struct rockchip_dp_device *dp = to_dp(encoder);
> +	unsigned long flags;
>   
>   	dev_dbg(dp->dev, "%s PSR...\n", enabled ? "Entry" : "Exit");
>   
> +	spin_lock_irqsave(&dp->psr_lock, flags);
>   	if (enabled)
>   		dp->psr_state = EDP_VSC_PSR_STATE_ACTIVE;
>   	else
>   		dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
>   
> -	schedule_delayed_work(&dp->psr_work, PSR_SET_DELAY_TIME);
> +	schedule_work(&dp->psr_work);
> +	spin_unlock_irqrestore(&dp->psr_lock, flags);
>   }
>   
>   static void analogix_dp_psr_work(struct work_struct *work)
>   {
>   	struct rockchip_dp_device *dp =
> -				container_of(work, typeof(*dp), psr_work.work);
> +				container_of(work, typeof(*dp), psr_work);
>   	struct drm_crtc *crtc = dp->encoder.crtc;
>   	int psr_state = dp->psr_state;
>   	int vact_end;
>   	int ret;
> +	unsigned long flags;
>   
>   	if (!crtc)
>   		return;
> @@ -115,10 +119,12 @@ static void analogix_dp_psr_work(struct work_struct *work)
>   		return;
>   	}
>   
> +	spin_lock_irqsave(&dp->psr_lock, flags);
>   	if (psr_state == EDP_VSC_PSR_STATE_ACTIVE)
>   		analogix_dp_enable_psr(dp->dev);
>   	else
>   		analogix_dp_disable_psr(dp->dev);
> +	spin_unlock_irqrestore(&dp->psr_lock, flags);
>   }
>   
>   static int rockchip_dp_pre_init(struct rockchip_dp_device *dp)
> @@ -135,6 +141,8 @@ static int rockchip_dp_poweron(struct analogix_dp_plat_data *plat_data)
>   	struct rockchip_dp_device *dp = to_dp(plat_data);
>   	int ret;
>   
> +	cancel_work_sync(&dp->psr_work);
> +
>   	ret = clk_prepare_enable(dp->pclk);
>   	if (ret < 0) {
>   		dev_err(dp->dev, "failed to enable pclk %d\n", ret);
> @@ -390,8 +398,9 @@ static int rockchip_dp_bind(struct device *dev, struct device *master,
>   	dp->plat_data.power_off = rockchip_dp_powerdown;
>   	dp->plat_data.get_modes = rockchip_dp_get_modes;
>   
> +	spin_lock_init(&dp->psr_lock);
>   	dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
> -	INIT_DELAYED_WORK(&dp->psr_work, analogix_dp_psr_work);
> +	INIT_WORK(&dp->psr_work, analogix_dp_psr_work);
>   
>   	rockchip_drm_psr_register(&dp->encoder, analogix_dp_psr_set);
>   


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series
  2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
                   ` (5 preceding siblings ...)
  2016-08-17  1:11 ` [PATCH v2 6/6] drm/rockchip: Enable vblank without event Sean Paul
@ 2016-08-17  2:41 ` Yakir Yang
  2016-08-17  2:45   ` Sean Paul
  2016-08-17  2:45   ` Yakir Yang
  6 siblings, 2 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:41 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip

Sean,

Thanks a lot for your good fixes. I have reviewed most of them, and all 
looks good to me.

But I got a question for merging things. My PSR patch set still under 
reviewing, haven't been picked up Mark or other maintainers. Feel a 
little bit embarrassed, how could we handle this situation ?

- Yakir

On 08/17/2016 09:11 AM, Sean Paul wrote:
> This is a follow-on set to Yakir's original PSR set here:
>          https://lkml.org/lkml/2016/7/24/34
> and applies to the for-next branch at:
> 	https://cgit.freedesktop.org/~seanpaul/dogwood
>
> There are a few issues with the code that needed to be
> shored up.
>   (1) The use of mutexes instead of spinlocks caused issues calling the
>       psr functions from vblank_enable/disable.
>   (2) The proliferation of workers due to (1)
>   (3) A bunch of races due to (2)
>   (4) vblank is not enabled unless an event is requested, this breaks
>       a lot of things, but most noticeable was cursor.
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood
> 	  instead of random on-list patches (some of which had drifted)
> 	- Added the "small fixes" patch to catch some nits
>
>
> Sean Paul (6):
>    drm/rockchip: Convert psr_list_mutex to spinlock and use it
>    drm/rockchip: Don't use a delayed worker for psr state changes
>    drm/rockchip: Use a spinlock to protect psr state
>    drm/rockchip: A couple small fixes to psr
>    drm/rockchip: Improve analogix-dp psr handling
>    drm/rockchip: Enable vblank without event
>
>   drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++--
>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c     |  2 +-
>   drivers/gpu/drm/rockchip/rockchip_drm_drv.h     |  2 +-
>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c     | 90 ++++++++++++-------------
>   drivers/gpu/drm/rockchip/rockchip_drm_vop.c     | 15 +++--
>   5 files changed, 69 insertions(+), 59 deletions(-)
>


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series
  2016-08-17  2:41 ` [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Yakir Yang
@ 2016-08-17  2:45   ` Sean Paul
  2016-08-17  3:20     ` Yakir Yang
  2016-08-17  2:45   ` Yakir Yang
  1 sibling, 1 reply; 17+ messages in thread
From: Sean Paul @ 2016-08-17  2:45 UTC (permalink / raw)
  To: Yakir Yang; +Cc: linux-rockchip, dri-devel, Doug Anderson, marcheu


[-- Attachment #1.1: Type: text/plain, Size: 2201 bytes --]

On Aug 16, 2016 7:41 PM, "Yakir Yang" <ykk@rock-chips.com> wrote:
>
> Sean,
>
> Thanks a lot for your good fixes. I have reviewed most of them, and all
looks good to me.
>
> But I got a question for merging things. My PSR patch set still under
reviewing, haven't been picked up Mark or other maintainers.

I've picked them up in my tree. I'll send a pull request to Dave once all
of the dependencies have been reviewed (marked NEEDS REVIEW).

Sean

> Feel a little bit embarrassed, how could we handle this situation ?
>
> - Yakir
>
>
> On 08/17/2016 09:11 AM, Sean Paul wrote:
>>
>> This is a follow-on set to Yakir's original PSR set here:
>>          https://lkml.org/lkml/2016/7/24/34
>> and applies to the for-next branch at:
>>         https://cgit.freedesktop.org/~seanpaul/dogwood
>>
>> There are a few issues with the code that needed to be
>> shored up.
>>   (1) The use of mutexes instead of spinlocks caused issues calling the
>>       psr functions from vblank_enable/disable.
>>   (2) The proliferation of workers due to (1)
>>   (3) A bunch of races due to (2)
>>   (4) vblank is not enabled unless an event is requested, this breaks
>>       a lot of things, but most noticeable was cursor.
>>
>> Changes in v2:
>>         - Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood
>>           instead of random on-list patches (some of which had drifted)
>>         - Added the "small fixes" patch to catch some nits
>>
>>
>> Sean Paul (6):
>>    drm/rockchip: Convert psr_list_mutex to spinlock and use it
>>    drm/rockchip: Don't use a delayed worker for psr state changes
>>    drm/rockchip: Use a spinlock to protect psr state
>>    drm/rockchip: A couple small fixes to psr
>>    drm/rockchip: Improve analogix-dp psr handling
>>    drm/rockchip: Enable vblank without event
>>
>>   drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++--
>>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c     |  2 +-
>>   drivers/gpu/drm/rockchip/rockchip_drm_drv.h     |  2 +-
>>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c     | 90
++++++++++++-------------
>>   drivers/gpu/drm/rockchip/rockchip_drm_vop.c     | 15 +++--
>>   5 files changed, 69 insertions(+), 59 deletions(-)
>>
>
>

[-- Attachment #1.2: Type: text/html, Size: 3122 bytes --]

[-- Attachment #2: Type: text/plain, Size: 160 bytes --]

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series
  2016-08-17  2:41 ` [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Yakir Yang
  2016-08-17  2:45   ` Sean Paul
@ 2016-08-17  2:45   ` Yakir Yang
  1 sibling, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  2:45 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip

Sean,

On 08/17/2016 10:41 AM, Yakir Yang wrote:
> Sean,
>
> Thanks a lot for your good fixes. I have reviewed most of them, and 
> all looks good to me.
>
> But I got a question for merging things. My PSR patch set still under 
> reviewing, haven't been picked up Mark or other maintainers. Feel a 
> little bit embarrassed, how could we handle this situation ?
>
> - Yakir
>
> On 08/17/2016 09:11 AM, Sean Paul wrote:
>> This is a follow-on set to Yakir's original PSR set here:
>>          https://lkml.org/lkml/2016/7/24/34
>> and applies to the for-next branch at:
>>     https://cgit.freedesktop.org/~seanpaul/dogwood

Oops, sorry for missing this comment, do you mean my PSR patch already 
have been site on your tree  :-D

- Yakir
>>
>> There are a few issues with the code that needed to be
>> shored up.
>>   (1) The use of mutexes instead of spinlocks caused issues calling the
>>       psr functions from vblank_enable/disable.
>>   (2) The proliferation of workers due to (1)
>>   (3) A bunch of races due to (2)
>>   (4) vblank is not enabled unless an event is requested, this breaks
>>       a lot of things, but most noticeable was cursor.
>>
>> Changes in v2:
>>     - Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood
>>       instead of random on-list patches (some of which had drifted)
>>     - Added the "small fixes" patch to catch some nits
>>
>>
>> Sean Paul (6):
>>    drm/rockchip: Convert psr_list_mutex to spinlock and use it
>>    drm/rockchip: Don't use a delayed worker for psr state changes
>>    drm/rockchip: Use a spinlock to protect psr state
>>    drm/rockchip: A couple small fixes to psr
>>    drm/rockchip: Improve analogix-dp psr handling
>>    drm/rockchip: Enable vblank without event
>>
>>   drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++--
>>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c     |  2 +-
>>   drivers/gpu/drm/rockchip/rockchip_drm_drv.h     |  2 +-
>>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c     | 90 
>> ++++++++++++-------------
>>   drivers/gpu/drm/rockchip/rockchip_drm_vop.c     | 15 +++--
>>   5 files changed, 69 insertions(+), 59 deletions(-)
>>
>


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series
  2016-08-17  2:45   ` Sean Paul
@ 2016-08-17  3:20     ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  3:20 UTC (permalink / raw)
  To: Sean Paul; +Cc: linux-rockchip, dri-devel, Doug Anderson, marcheu


[-- Attachment #1.1: Type: text/plain, Size: 2525 bytes --]

Sean,


On 08/17/2016 10:45 AM, Sean Paul wrote:
>
> On Aug 16, 2016 7:41 PM, "Yakir Yang" <ykk@rock-chips.com 
> <mailto:ykk@rock-chips.com>> wrote:
> >
> > Sean,
> >
> > Thanks a lot for your good fixes. I have reviewed most of them, and 
> all looks good to me.
> >
> > But I got a question for merging things. My PSR patch set still 
> under reviewing, haven't been picked up Mark or other maintainers.
>
> I've picked them up in my tree. I'll send a pull request to Dave once 
> all of the dependencies have been reviewed (marked NEEDS REVIEW).
>

Got it, thanks.

- Yakir

> Sean
>
> > Feel a little bit embarrassed, how could we handle this situation ?
> >
> > - Yakir
> >
> >
> > On 08/17/2016 09:11 AM, Sean Paul wrote:
> >>
> >> This is a follow-on set to Yakir's original PSR set here:
> >> https://lkml.org/lkml/2016/7/24/34
> >> and applies to the for-next branch at:
> >> https://cgit.freedesktop.org/~seanpaul/dogwood 
> <https://cgit.freedesktop.org/%7Eseanpaul/dogwood>
> >>
> >> There are a few issues with the code that needed to be
> >> shored up.
> >>   (1) The use of mutexes instead of spinlocks caused issues calling the
> >>       psr functions from vblank_enable/disable.
> >>   (2) The proliferation of workers due to (1)
> >>   (3) A bunch of races due to (2)
> >>   (4) vblank is not enabled unless an event is requested, this breaks
> >>       a lot of things, but most noticeable was cursor.
> >>
> >> Changes in v2:
> >>         - Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood 
> <https://cgit.freedesktop.org/%7Eseanpaul/dogwood>
> >>           instead of random on-list patches (some of which had drifted)
> >>         - Added the "small fixes" patch to catch some nits
> >>
> >>
> >> Sean Paul (6):
> >>    drm/rockchip: Convert psr_list_mutex to spinlock and use it
> >>    drm/rockchip: Don't use a delayed worker for psr state changes
> >>    drm/rockchip: Use a spinlock to protect psr state
> >>    drm/rockchip: A couple small fixes to psr
> >>    drm/rockchip: Improve analogix-dp psr handling
> >>    drm/rockchip: Enable vblank without event
> >>
> >>   drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 19 ++++--
> >>   drivers/gpu/drm/rockchip/rockchip_drm_drv.c     |  2 +-
> >>   drivers/gpu/drm/rockchip/rockchip_drm_drv.h     |  2 +-
> >>   drivers/gpu/drm/rockchip/rockchip_drm_psr.c     | 90 
> ++++++++++++-------------
> >>   drivers/gpu/drm/rockchip/rockchip_drm_vop.c     | 15 +++--
> >>   5 files changed, 69 insertions(+), 59 deletions(-)
> >>
> >
> >
>


[-- Attachment #1.2: Type: text/html, Size: 4356 bytes --]

[-- Attachment #2: Type: text/plain, Size: 160 bytes --]

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
* Re: [PATCH v2 6/6] drm/rockchip: Enable vblank without event
  2016-08-17  1:11 ` [PATCH v2 6/6] drm/rockchip: Enable vblank without event Sean Paul
@ 2016-08-17  3:27   ` Yakir Yang
  0 siblings, 0 replies; 17+ messages in thread
From: Yakir Yang @ 2016-08-17  3:27 UTC (permalink / raw)
  To: Sean Paul, mark.yao, dri-devel; +Cc: marcheu, dianders, linux-rockchip

Sean,

On 08/17/2016 09:11 AM, Sean Paul wrote:
> vblank should be enabled regardless of whether an event
> is expected back. This is especially important for a cursor
> plane.
Yep, I also found that sometimes vblank haven't been enabled when I move 
the mouse lightly, that would cause eDP panel wound't exit from PSR 
active state, and then nothing would be updated on panel. After apply 
this patch, things work rightly now, thanks for fixing.

Reviewed-by: Yakir Yang <ykk@rock-chips.com>
Tested-by: Yakir Yang <ykk@rock-chip.com>
> Signed-off-by: Sean Paul <seanpaul@chromium.org>
> ---
>
> Changes in v2:
> 	- Rebased on https://cgit.freedesktop.org/~seanpaul/dogwood/log/?h=for-next
>
>   drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 15 ++++++++++-----
>   1 file changed, 10 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> index d1e0e06..1787084 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> @@ -112,6 +112,7 @@ struct vop {
>   	struct device *dev;
>   	struct drm_device *drm_dev;
>   	bool is_enabled;
> +	bool vblank_active;
>   
>   	/* mutex vsync_ work */
>   	struct mutex vsync_mutex;
> @@ -1107,10 +1108,11 @@ static void vop_crtc_atomic_begin(struct drm_crtc *crtc,
>   	struct vop *vop = to_vop(crtc);
>   
>   	spin_lock_irq(&crtc->dev->event_lock);
> -	if (crtc->state->event) {
> -		WARN_ON(drm_crtc_vblank_get(crtc) != 0);
> -		WARN_ON(vop->event);
> +	vop->vblank_active = true;
> +	WARN_ON(drm_crtc_vblank_get(crtc) != 0);
> +	WARN_ON(vop->event);
>   
> +	if (crtc->state->event) {
>   		vop->event = crtc->state->event;
>   		crtc->state->event = NULL;
>   	}
> @@ -1197,12 +1199,14 @@ static void vop_handle_vblank(struct vop *vop)
>   
>   	spin_lock_irqsave(&drm->event_lock, flags);
>   	if (vop->event) {
> -
>   		drm_crtc_send_vblank_event(crtc, vop->event);
> -		drm_crtc_vblank_put(crtc);
>   		vop->event = NULL;
>   
>   	}
> +	if (vop->vblank_active) {
> +		vop->vblank_active = false;
> +		drm_crtc_vblank_put(crtc);
> +	}
>   	spin_unlock_irqrestore(&drm->event_lock, flags);
>   
>   	if (!completion_done(&vop->wait_update_complete))
> @@ -1472,6 +1476,7 @@ static int vop_initial(struct vop *vop)
>   	clk_disable(vop->aclk);
>   
>   	vop->is_enabled = false;
> +	vop->vblank_active = false;
>   
>   	return 0;
>   


_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

^ permalink raw reply	[flat|nested] 17+ messages in thread
end of thread, other threads:[~2016-08-17  3:27 UTC | newest]

Thread overview: 17+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-08-17  1:11 [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Sean Paul
2016-08-17  1:11 ` [PATCH v2 1/6] drm/rockchip: Convert psr_list_mutex to spinlock and use it Sean Paul
2016-08-17  2:29   ` Yakir Yang
2016-08-17  1:11 ` [PATCH v2 2/6] drm/rockchip: Don't use a delayed worker for psr state changes Sean Paul
2016-08-17  2:31   ` Yakir Yang
2016-08-17  1:11 ` [PATCH v2 3/6] drm/rockchip: Use a spinlock to protect psr state Sean Paul
2016-08-17  2:33   ` Yakir Yang
2016-08-17  1:11 ` [PATCH v2 4/6] drm/rockchip: A couple small fixes to psr Sean Paul
2016-08-17  2:34   ` Yakir Yang
2016-08-17  1:11 ` [PATCH v2 5/6] drm/rockchip: Improve analogix-dp psr handling Sean Paul
2016-08-17  2:35   ` Yakir Yang
2016-08-17  1:11 ` [PATCH v2 6/6] drm/rockchip: Enable vblank without event Sean Paul
2016-08-17  3:27   ` Yakir Yang
2016-08-17  2:41 ` [PATCH v2 0/6] drm/rockchip: Some patches to update the PSR series Yakir Yang
2016-08-17  2:45   ` Sean Paul
2016-08-17  3:20     ` Yakir Yang
2016-08-17  2:45   ` Yakir Yang

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.