media: rc: drivers should produce alternate pulse and space timing events

Report an error if this is not the case or any problem with the generated raw events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c +3 -3

 			list_for_each_entry(handler, &ir_raw_handler_list, list)
-				if (raw->dev->enabled_protocols &
+				if (dev->enabled_protocols &
 				    handler->protocols || !handler->protocols)
-					handler->decode(raw->dev, ev);
-			ir_lirc_raw_event(raw->dev, ev);
+					handler->decode(dev, ev);
+			ir_lirc_raw_event(dev, ev);
 			raw->prev_ev = ev;
 		}
 		mutex_unlock(&ir_raw_handler_lock);

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001

Quoted 50 lines · Jerome Brunet

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Quoted 1 line

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

Maybe it should be dev_warn(). The fact that it is not dev_warn_once() means that we now know this occurs regularly.

Would you mind testing the following patch please?

Thanks

Sean

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 19 Jun 2018 13:50:36 +0100
Subject: [PATCH] media: rc: meson: rc rc0: two consecutive events of type
space

The meson generates one edge per interrupt. The duration is encoded in 12 bits of 10 microseconds, so it can only encoding a maximum of 40 milliseconds. As a result, it can produce multiple space events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/meson-ir.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

drivers/media/rc/meson-ir.c +0 -0

index f449b35d25e7..9747426719b2 100644

Quoted 51 lines · Sean Young

Quoted 50 lines · Jerome Brunet

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Quoted 4 lines

Quoted 1 line

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

Maybe it should be dev_warn(). The fact that it is not dev_warn_once() means that we now know this occurs regularly.

It seems weird to report this over and over again.

Quoted 34 lines

Would you mind testing the following patch please?

Thanks

Sean

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 19 Jun 2018 13:50:36 +0100
Subject: [PATCH] media: rc: meson: rc rc0: two consecutive events of type
space

The meson generates one edge per interrupt. The duration is encoded in 12 bits of 10 microseconds, so it can only encoding a maximum of 40 milliseconds. As a result, it can produce multiple space events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/meson-ir.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

drivers/media/rc/meson-ir.c +2 -1

index f449b35d25e7..9747426719b2 100644
@@ -97,7 +97,8 @@ static irqreturn_t meson_ir_irq(int irqno, void *dev_id)
 	status = readl_relaxed(ir->reg + IR_DEC_STATUS);
 	rawir.pulse = !!(status & STATUS_IR_DEC_IN);
 
-	ir_raw_event_store_with_timeout(ir->rc, &rawir);
+	if (ir_raw_event_store_with_filter(ir->rc, &rawir))
+		ir_raw_event_handle(ir->rc);
 

drivers/media/rc/meson-ir.c static irqreturn_t meson_ir_irq(int irqno, void *dev_id) · line 104

Quoted 2 lines

drivers/media/rc/meson-ir.c +0 -0

 	spin_unlock(&ir->lock);
 

Solve the problem on meson. Thx
Feel free to add this submitting the patch

Quoted 93 lines · Jerome Brunet

Quoted 51 lines · Sean Young

Quoted 50 lines · Jerome Brunet

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Quoted 4 lines

Quoted 1 line

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

Maybe it should be dev_warn(). The fact that it is not dev_warn_once() means that we now know this occurs regularly.

It seems weird to report this over and over again.

Quoted 34 lines

Would you mind testing the following patch please?

Thanks

Sean

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 19 Jun 2018 13:50:36 +0100
Subject: [PATCH] media: rc: meson: rc rc0: two consecutive events of type
space

The meson generates one edge per interrupt. The duration is encoded in 12 bits of 10 microseconds, so it can only encoding a maximum of 40 milliseconds. As a result, it can produce multiple space events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/meson-ir.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

drivers/media/rc/meson-ir.c +2 -1

index f449b35d25e7..9747426719b2 100644
@@ -97,7 +97,8 @@ static irqreturn_t meson_ir_irq(int irqno, void *dev_id)
 	status = readl_relaxed(ir->reg + IR_DEC_STATUS);
 	rawir.pulse = !!(status & STATUS_IR_DEC_IN);
 
-	ir_raw_event_store_with_timeout(ir->rc, &rawir);
+	if (ir_raw_event_store_with_filter(ir->rc, &rawir))
+		ir_raw_event_handle(ir->rc);
 

Quoted 5 lines

drivers/media/rc/meson-ir.c static irqreturn_t meson_ir_irq(int irqno, void *dev_id) · line 104

Quoted 2 lines

drivers/media/rc/meson-ir.c +0 -0

 	spin_unlock(&ir->lock);
 

Solve the problem on meson. Thx
Feel free to add this submitting the patch

Actually this patch is broken. ir_raw_event_store_with_timeout() generates IR timeouts, but ir_raw_event_store_with_filter() does not.

So this will need some rethinking. I think we need a ir_raw_event_store_with_timeout_with_filter() but that is getting silly now.

Maybe filter should be a boolean property of an rc device and happen transparently. I'll write something like that when I get some time.

Sean

Quoted 106 lines · Sean Young

Quoted 93 lines · Jerome Brunet

Quoted 51 lines · Sean Young

Quoted 50 lines · Jerome Brunet

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Quoted 4 lines

Quoted 1 line

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

Maybe it should be dev_warn(). The fact that it is not dev_warn_once() means that we now know this occurs regularly.

It seems weird to report this over and over again.

Quoted 34 lines

Would you mind testing the following patch please?

Thanks

Sean

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 19 Jun 2018 13:50:36 +0100
Subject: [PATCH] media: rc: meson: rc rc0: two consecutive events of type
space

The meson generates one edge per interrupt. The duration is encoded in 12 bits of 10 microseconds, so it can only encoding a maximum of 40 milliseconds. As a result, it can produce multiple space events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/meson-ir.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

drivers/media/rc/meson-ir.c +2 -1

index f449b35d25e7..9747426719b2 100644
@@ -97,7 +97,8 @@ static irqreturn_t meson_ir_irq(int irqno, void *dev_id)
 	status = readl_relaxed(ir->reg + IR_DEC_STATUS);
 	rawir.pulse = !!(status & STATUS_IR_DEC_IN);
 
-	ir_raw_event_store_with_timeout(ir->rc, &rawir);
+	if (ir_raw_event_store_with_filter(ir->rc, &rawir))
+		ir_raw_event_handle(ir->rc);
 

Quoted 5 lines

drivers/media/rc/meson-ir.c static irqreturn_t meson_ir_irq(int irqno, void *dev_id) · line 104

Quoted 2 lines

drivers/media/rc/meson-ir.c +0 -0

 	spin_unlock(&ir->lock);
 

Solve the problem on meson. Thx
Feel free to add this submitting the patch

Actually this patch is broken. ir_raw_event_store_with_timeout() generates IR timeouts, but ir_raw_event_store_with_filter() does not.

So this will need some rethinking. I think we need a ir_raw_event_store_with_timeout_with_filter() but that is getting silly now.

Quoted 3 lines

Maybe filter should be a boolean property of an rc device and happen transparently. I'll write something like that when I get some time.

would you still object to use dev_warn_once() in the meantime ? Console flooding is quite annoying.

From 2f61a8b6e850e34def282f1cbd1bfee0f5ab66c5 Mon Sep 17 00:00:00 2001

Quoted 107 lines · Jerome Brunet

Quoted 106 lines · Sean Young

Quoted 93 lines · Jerome Brunet

Quoted 51 lines · Sean Young

Quoted 50 lines · Jerome Brunet

Quoted 2 lines · Sean Young

Report an error if this is not the case or any problem with the generated raw events.

Hi,

Since the inclusion of this patch, every 3 to 15 seconds, I get the following message:

"rc rc0: two consecutive events of type space"

on the console of amlogic s400 platform (arch/arm64/boot/dts/amlogic/meson-axg- s400.dts). I don't know much about ir protocol and surely there is something worth investigating in the related driver, but ...

Quoted 21 lines

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +2 -1

index 2e50104ae138..49c56da9bc67 100644
@@ -22,16 +22,27 @@ static int ir_raw_event_thread(void *data)
 {
 	struct ir_raw_event ev;
 	struct ir_raw_handler *handler;
-	struct ir_raw_event_ctrl *raw = (struct ir_raw_event_ctrl *)data;
+	struct ir_raw_event_ctrl *raw = data;
+	struct rc_dev *dev = raw->dev;
 
 	while (1) {
 		mutex_lock(&ir_raw_handler_lock);
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {

drivers/media/rc/rc-ir-raw.c static int ir_raw_event_thread(void *data) · line 40

Quoted 10 lines

drivers/media/rc/rc-ir-raw.c +10 -0

+			if (is_timing_event(ev)) {
+				if (ev.duration == 0)
+					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+				if (is_timing_event(raw->prev_ev) &&
+				    !is_transition(&ev, &raw->prev_ev))
+					dev_err(&dev->dev, "two consecutive events of type %s",
+						TO_STR(ev.pulse));
+				if (raw->prev_ev.reset && ev.pulse == 0)
+					dev_err(&dev->dev, "timing event after reset should be pulse");
+			}

... considering that we continue the processing as if nothing happened, is it really an error ?

Quoted 4 lines

Quoted 1 line

Could we consider something less invasive ? like dev_dbg() or dev_warn_once() ?

Maybe it should be dev_warn(). The fact that it is not dev_warn_once() means that we now know this occurs regularly.

It seems weird to report this over and over again.

Quoted 34 lines

Would you mind testing the following patch please?

Thanks

Sean

From 6a44fbe4738d230b9cf378777e7e9a93e5fda726 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 19 Jun 2018 13:50:36 +0100
Subject: [PATCH] media: rc: meson: rc rc0: two consecutive events of type
space

The meson generates one edge per interrupt. The duration is encoded in 12 bits of 10 microseconds, so it can only encoding a maximum of 40 milliseconds. As a result, it can produce multiple space events.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/meson-ir.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

drivers/media/rc/meson-ir.c +2 -1

index f449b35d25e7..9747426719b2 100644
@@ -97,7 +97,8 @@ static irqreturn_t meson_ir_irq(int irqno, void *dev_id)
 	status = readl_relaxed(ir->reg + IR_DEC_STATUS);
 	rawir.pulse = !!(status & STATUS_IR_DEC_IN);
 
-	ir_raw_event_store_with_timeout(ir->rc, &rawir);
+	if (ir_raw_event_store_with_filter(ir->rc, &rawir))
+		ir_raw_event_handle(ir->rc);
 

Quoted 5 lines

drivers/media/rc/meson-ir.c static irqreturn_t meson_ir_irq(int irqno, void *dev_id) · line 104

Quoted 2 lines

drivers/media/rc/meson-ir.c +0 -0

 	spin_unlock(&ir->lock);
 

Solve the problem on meson. Thx
Feel free to add this submitting the patch

Actually this patch is broken. ir_raw_event_store_with_timeout() generates IR timeouts, but ir_raw_event_store_with_filter() does not.

So this will need some rethinking. I think we need a ir_raw_event_store_with_timeout_with_filter() but that is getting silly now.

Quoted 6 lines

Quoted 3 lines

Maybe filter should be a boolean property of an rc device and happen transparently. I'll write something like that when I get some time.

would you still object to use dev_warn_once() in the meantime ? Console flooding is quite annoying.

Yes, as much as I dislike it, I think that is the right solution for now.

Thanks,

Sean

From 2f61a8b6e850e34def282f1cbd1bfee0f5ab66c5 Mon Sep 17 00:00:00 2001
From: Sean Young <[email protected]>
Date: Tue, 26 Jun 2018 16:03:18 +0100
Subject: [PATCH] media: rc: be less noisy when driver misbehaves

Since commit 48231f289e52 ("media: rc: drivers should produce alternate pulse and space timing events"), on meson-ir we are regularly producing errors. Reduce to warning level and only warn once to avoid flooding the log.

A proper fix for meson-ir is going to be too large for v4.18.

Signed-off-by Sean Young <[email protected]>

---
 drivers/media/rc/rc-ir-raw.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

drivers/media/rc/rc-ir-raw.c +4 -4

index 2e0066b1a31c..e7948908e78c 100644
@@ -30,13 +30,13 @@ static int ir_raw_event_thread(void *data)
 		while (kfifo_out(&raw->kfifo, &ev, 1)) {
 			if (is_timing_event(ev)) {
 				if (ev.duration == 0)
-					dev_err(&dev->dev, "nonsensical timing event of duration 0");
+					dev_warn_once(&dev->dev, "nonsensical timing event of duration 0");
 				if (is_timing_event(raw->prev_ev) &&
 				    !is_transition(&ev, &raw->prev_ev))
-					dev_err(&dev->dev, "two consecutive events of type %s",
-						TO_STR(ev.pulse));
+					dev_warn_once(&dev->dev, "two consecutive events of type %s",
+						      TO_STR(ev.pulse));
 				if (raw->prev_ev.reset && ev.pulse == 0)
-					dev_err(&dev->dev, "timing event after reset should be pulse");
+					dev_warn_once(&dev->dev, "timing event after reset should be pulse");
 			}
 			list_for_each_entry(handler, &ir_raw_handler_list, list)
 				if (dev->enabled_protocols &