Skip to contentSkip to conversation

Add driver for GE B850v3 LVDS/DP++ Bridge

PATCH Superseded Peter Senna Tschudin 12 participants 64 messages May 30, 2016 – Sep 29, 2016

Subsystems (5)

v1 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

cover Add driver for GE B850v3 LVDS/DP++ Bridge

The series adds a driver that creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3. There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However the physical bridges are automatically configured by the input video signal, and the driver has no access to the video processing pipeline. The driver is only needed to read EDID from the STDP2690 and to handle HPD events from the STDP4028. The driver communicates with both bridges over i2c. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

The patches from the series:
[1/5] Change the imx-ldb driver to allow attaching a bridge and not only a LVDS
panel.

[2/5] Configure the mapping between IPUs and external displays on the dts file
of the B850v3. Needed so the GPU can drive two Full-HD monitors.

[3/5] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[4/5] Add the driver, make changes to MAINTAINERS, Kconfig and Makefile

[5/5] Make the changes to the B850v3 dts file to enalbe the GE B850v3
LVDS/DP++ Bridge.

Peter Senna Tschudin (5):
drm/imx-ldb: Add support to drm-bridge dts/imx6q-b850v3: Configure IPU assignment order Documentation/devicetree/bindings: b850v3_lvds_dp drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      |  38 ++
 MAINTAINERS                                        |   8 +
 arch/arm/boot/dts/imx6q-b850v3.dts                 |  36 ++
 drivers/gpu/drm/bridge/Kconfig                     |   8 +
 drivers/gpu/drm/bridge/Makefile                    |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c         | 397 +++++++++++++++++++++
 drivers/gpu/drm/imx/imx-ldb.c                      |  75 ++--
 7 files changed, 542 insertions(+), 21 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
Signature · 1 line
2.5.5
#1 lore raw

PATCH 1/5 drm/imx-ldb: Add support to drm-bridge

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

---
 drivers/gpu/drm/imx/imx-ldb.c | 75 +++++++++++++++++++++++++++++++------------
 1 file changed, 54 insertions(+), 21 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +54 -21
index a58eee5..7233a81 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *ext_bridge;
+
 	struct device_node *child;
 	int chno;
 	void *edid;
@@ -295,6 +299,10 @@ static void imx_ldb_encoder_mode_set(struct drm_encoder *encoder,
 	}
 }
 
+static void imx_ldb_encoder_enable(struct drm_encoder *encoder)
+{
+}
+
 static void imx_ldb_encoder_disable(struct drm_encoder *encoder)
 {
 	struct imx_ldb_channel *imx_ldb_ch = enc_to_imx_ldb_ch(encoder);
@@ -373,6 +381,7 @@ static const struct drm_encoder_helper_funcs imx_ldb_encoder_helper_funcs = {
 	.prepare = imx_ldb_encoder_prepare,
 	.commit = imx_ldb_encoder_commit,
 	.mode_set = imx_ldb_encoder_mode_set,
+	.enable = imx_ldb_encoder_enable,
 	.disable = imx_ldb_encoder_disable,
 };
 
@@ -417,16 +426,28 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, &imx_ldb_ch->encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
-
-	if (imx_ldb_ch->panel)
+	if (imx_ldb_ch->panel) {
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
 		drm_panel_attach(imx_ldb_ch->panel, &imx_ldb_ch->connector);
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
-			&imx_ldb_ch->encoder);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				&imx_ldb_ch->encoder);
+	}
+
+	if (imx_ldb_ch->ext_bridge) {
+		imx_ldb_ch->ext_bridge->encoder = &imx_ldb_ch->encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->ext_bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->ext_bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	}
 
 	return 0;
 }
@@ -583,23 +604,35 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 			endpoint = of_get_child_by_name(port, "endpoint");
 			if (endpoint) {
 				remote = of_graph_get_remote_port_parent(endpoint);
-				if (remote)
-					channel->panel = of_drm_find_panel(remote);
-				else
-					return -EPROBE_DEFER;
-				if (!channel->panel) {
-					dev_err(dev, "panel not found: %s\n",
-						remote->full_name);
-					return -EPROBE_DEFER;
+				if (remote) {
+					/* Only one of these two will succeed */
+					channel->panel =
+						of_drm_find_panel(remote);
+
+					channel->ext_bridge =
+						of_drm_find_bridge(remote);
+
+					/*
+					 * If the bridge is compiled as a
+					 * module, it may take some time until
+					 * the bridge driver is available.
+					 * Defer until the bridge driver is
+					 * ready.
+					 */
+					if (!channel->panel &&
+							!channel->ext_bridge)
+						return -EPROBE_DEFER;
 				}
 			}
 		}
 
-		edidp = of_get_property(child, "edid", &channel->edid_len);
-		if (edidp) {
-			channel->edid = kmemdup(edidp, channel->edid_len,
-						GFP_KERNEL);
-		} else if (!channel->panel) {
+		if (channel->panel) {
+			edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+			if (edidp)
+				channel->edid = kmemdup(edidp,
+						channel->edid_len, GFP_KERNEL);
+		} else {
 			ret = of_get_drm_display_mode(child, &channel->mode, 0);
 			if (!ret)
 				channel->mode_valid = 1;
Signature · 1 line
2.5.5
#2 lore raw

Re: [PATCH 1/5] drm/imx-ldb: Add support to drm-bridge

Am Montag, den 30.05.2016, 18:39 +0200 schrieb Peter Senna Tschudin:

Quoted 13 lines

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

---
 drivers/gpu/drm/imx/imx-ldb.c | 75 +++++++++++++++++++++++++++++++------------
 1 file changed, 54 insertions(+), 21 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +0 -0
index a58eee5..7233a81 100644

drivers/gpu/drm/imx/imx-ldb.c struct imx_ldb_channel { · line 64

Quoted 8 lines
drivers/gpu/drm/imx/imx-ldb.c +4 -0
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *ext_bridge;
+

Just bridge should be clear enough.

drivers/gpu/drm/imx/imx-ldb.c static void imx_ldb_encoder_mode_set(struct drm_encoder *encoder, · line 305

Quoted 11 lines
drivers/gpu/drm/imx/imx-ldb.c +4 -0
 	struct device_node *child;
 	int chno;
 	void *edid;
@@ -295,6 +299,10 @@ static void imx_ldb_encoder_mode_set(struct drm_encoder *encoder,
 	}
 }
 
+static void imx_ldb_encoder_enable(struct drm_encoder *encoder)
+{
+}
+

Why?

Note that Liu Ying's atomic series touches this also, and after "drm/imx: atomic phase 3 step 3: Legacy callback fixups" the enable callback exists. Depending on how either series proceeds, one will have to be rebased onto the other.

Quoted 12 lines
drivers/gpu/drm/imx/imx-ldb.c +1 -0
 static void imx_ldb_encoder_disable(struct drm_encoder *encoder)
 {
 	struct imx_ldb_channel *imx_ldb_ch = enc_to_imx_ldb_ch(encoder);
@@ -373,6 +381,7 @@ static const struct drm_encoder_helper_funcs imx_ldb_encoder_helper_funcs = {
 	.prepare = imx_ldb_encoder_prepare,
 	.commit = imx_ldb_encoder_commit,
 	.mode_set = imx_ldb_encoder_mode_set,
+	.enable = imx_ldb_encoder_enable,
 	.disable = imx_ldb_encoder_disable,
 };
 

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_register(struct drm_device *drm, · line 429

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +1 -6
@@ -417,16 +426,28 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, &imx_ldb_ch->encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
-
-	if (imx_ldb_ch->panel)
+	if (imx_ldb_ch->panel) {
	if (!imx_ldb_ch->bridge) {
Quoted 27 lines
drivers/gpu/drm/imx/imx-ldb.c +19 -2
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
 		drm_panel_attach(imx_ldb_ch->panel, &imx_ldb_ch->connector);
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
-			&imx_ldb_ch->encoder);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				&imx_ldb_ch->encoder);
+	}
+
+	if (imx_ldb_ch->ext_bridge) {
+		imx_ldb_ch->ext_bridge->encoder = &imx_ldb_ch->encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->ext_bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->ext_bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	}
 
 	return 0;
 }

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_bind(struct device *dev, struct device *master, void *data) · line 592

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +0 -7
@@ -583,23 +604,35 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 			endpoint = of_get_child_by_name(port, "endpoint");
 			if (endpoint) {
 				remote = of_graph_get_remote_port_parent(endpoint);
-				if (remote)
-					channel->panel = of_drm_find_panel(remote);
-				else
-					return -EPROBE_DEFER;
-				if (!channel->panel) {
-					dev_err(dev, "panel not found: %s\n",
-						remote->full_name);

Let's keep this message, at least as dev_dbg.

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +9 -1
-					return -EPROBE_DEFER;
+				if (remote) {
+					/* Only one of these two will succeed */
+					channel->panel =
+						of_drm_find_panel(remote);
+
+					channel->ext_bridge =
+						of_drm_find_bridge(remote);
+
+					/*

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_bind(struct device *dev, struct device *master, void *data) · line 625

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +9 -0
+					 * If the bridge is compiled as a
+					 * module, it may take some time until
+					 * the bridge driver is available.
+					 * Defer until the bridge driver is
+					 * ready.
+					 */
+					if (!channel->panel &&
+							!channel->ext_bridge)
+						return -EPROBE_DEFER;
 				}

What happens if (!remote)?

Quoted 5 lines
drivers/gpu/drm/imx/imx-ldb.c +0 -2
 			}
 		}
 
-		edidp = of_get_property(child, "edid", &channel->edid_len);
-		if (edidp) {

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_bind(struct device *dev, struct device *master, void *data) · line 635

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +7 -3
-			channel->edid = kmemdup(edidp, channel->edid_len,
-						GFP_KERNEL);
-		} else if (!channel->panel) {
+		if (channel->panel) {
+			edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+			if (edidp)
+				channel->edid = kmemdup(edidp,
+						channel->edid_len, GFP_KERNEL);
+		} else {

Any reason to disallow overriding EDID from DT if there is no panel? I'd leave this part as is.

regards
Philipp

#3 lore raw

PATCH 2/5 arm/dts/imx6q-b850v3: Configure IPU assignment order

In reply to Peter Senna Tschudin

Configure the IPU assignment order to assign one IPU per external display. A single IPU can drive multiple external displays but there are resolution restrictions. After this patch the GPU is capalbe of driving two Full-HD monitors.

---
 arch/arm/boot/dts/imx6q-b850v3.dts | 5 +++++
 1 file changed, 5 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +5 -0
index 167f744..88a70de 100644
@@ -51,6 +51,11 @@
 	chosen {
 		stdout-path = &uart3;
 	};
+
+	display-subsystem {
+		compatible = "fsl,imx-display-subsystem";
+		ports = <&ipu1_di0>, <&ipu2_di0>, <&ipu1_di1>, <&ipu2_di1>;
+	};
 };
 
 &clks {
Signature · 1 line
2.5.5
reviewer
#4 lore raw

Re: [PATCH 2/5] arm/dts/imx6q-b850v3: Configure IPU assignment order

Quoted 3 lines · Peter Senna Tschudin

Configure the IPU assignment order to assign one IPU per external display. A single IPU can drive multiple external displays but there are resolution restrictions. After this patch the GPU is capalbe of driving two

I think you meant "the IPU is capable"

#5 lore raw

Re: [PATCH 2/5] arm/dts/imx6q-b850v3: Configure IPU assignment order

In reply to Peter Senna Tschudin

Am Montag, den 30.05.2016, 18:39 +0200 schrieb Peter Senna Tschudin:

Quoted 4 lines

Configure the IPU assignment order to assign one IPU per external display. A single IPU can drive multiple external displays but there are resolution restrictions. After this patch the GPU is capalbe of driving two Full-HD monitors.

It's also capable to do it before this patch, if you use the first and third crtc. You are just reordering the crtcs.
Unfortunately the IPU has combined limitations across multiple crtcs in one IPU, which currently can't be communicated to userspace.

regards
Philipp

#6 lore raw

PATCH 3/5 Documentation/devicetree/bindings: Add b850v3_lvds_dp

In reply to Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

---
 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +38 -0
new file mode 100644
index 0000000..32e123a
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : should link to the gpio used as interrupt
+    source on the host.
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the vide signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3_dp_bridge {
+		compatible = "ge,b850v3_lvds_dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port@0 {
+			reg = <0>;
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#7 lore raw

Re: [PATCH 3/5] Documentation/devicetree/bindings: Add b850v3_lvds_dp

Hi Peter,

Am Montag, den 30.05.2016, 18:39 +0200 schrieb Peter Senna Tschudin:

Quoted 15 lines

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

---
 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +0 -0
new file mode 100644
index 0000000..32e123a

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 4

Quoted 4 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +4 -0
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".

In the cover mail you write that this is a combination of the STDP4028 DP transmitter and STDP2690 DP/DP++ converter. So shouldn't this binding comprise two device tree nodes, one with compatible "st,stdp4028", one with "st,stdp2690", then?
Is the stdp2690 connected to the stdp4028's i2c master?

regards
Philipp

#8 lore raw

Re: [PATCH 3/5] Documentation/devicetree/bindings: Add b850v3_lvds_dp

Hi Philipp,

Thank you very much for the review! I'll send V2 soon, doing my best to avoid collisions with other patches that are under review.

Quoted 13 lines · Philipp Zabel

Hi Peter,

Am Montag, den 30.05.2016, 18:39 +0200 schrieb Peter Senna Tschudin:

Quoted 9 lines

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

---
 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Quoted 16 lines

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 4

Quoted 10 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +4 -0
new file mode 100644
index 0000000..32e123a
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".

In the cover mail you write that this is a combination of the STDP4028 DP transmitter and STDP2690 DP/DP++ converter. So shouldn't this binding comprise two device tree nodes, one with compatible "st,stdp4028", one with "st,stdp2690", then?
Is the stdp2690 connected to the stdp4028's i2c master?

The hardware and firmware made it complicated for this binding to comprise two device tree nodes. The hardware and firmware are designed to configure both bridges based on the LVDS signal. This part works very well, but it leave the driver powerless to control the video processing pipeline. The bridges sort of behave as a single bridge, the sort of part is the need of interacting with both bridges: EDID from the STDP2690 and HPD events from the STDP4028. So the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts.

I would find it nicer to be author of two drivers instead of one, but as the behavior handled by the driver is specific of the B850v3, and do not reflect the standard behavior of each bridge, I find it unlikely that the driver could be useful to handle STDP4028 and STPD2690 with stock firmware, or in other designs. On the other hand, I made sure to use the chip name on the defines to maximize chances of reuse.

Peter

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

maintainer
#9 lore raw

Re: [PATCH 3/5] Documentation/devicetree/bindings: Add b850v3_lvds_dp

In reply to Peter Senna Tschudin
Quoted 15 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

---
 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +0 -0
new file mode 100644
index 0000000..32e123a

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 10

Quoted 10 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +10 -0
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : should link to the gpio used as interrupt
+    source on the host.
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information

This should be known based on the bridge chip you are using.

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 20

Quoted 10 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +10 -0
+  - port : should describe the vide signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3_dp_bridge {

Don't use '_' in node or property names or compatible strings.

Quoted 21 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +18 -0
+		compatible = "ge,b850v3_lvds_dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port@0 {
+			reg = <0>;
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#10 lore raw

PATCH 4/5 drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

This driver creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3(imx6q-b850v3.dts). There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However the physical bridges are automatically configured by the input video signal, and the driver has no access to the video processing pipeline. The driver is only needed to read EDID from the STDP2690 and to handle HPD events from the STDP4028. The driver communicates with both bridges over i2c. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

---
 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |   8 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 397 +++++++++++++++++++++++++++++
 4 files changed, 414 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 3273ffa..7bb5e89 100644
@@ -5009,6 +5009,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Martin Donnelly <[email protected]>
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +8 -0
index 8f7423f..e9c32fc 100644
@@ -52,4 +52,12 @@ config DRM_PARADE_PS8622
 
 source "drivers/gpu/drm/bridge/analogix/Kconfig"
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "LVDS/DP bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+	  Driver for GE B850v3 DP2 LVDS to DP+ Bridge
+
 endmenu
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..ba2e355 100644
@@ -6,3 +6,4 @@ obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +397 -0

397 lines not on this page · raw message

Signature · 1 line
2.5.5
#11 lore raw

Re: [PATCH 4/5] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Peter,

Just some comments that I received when I submitted my bridge patches and that I think can help you ...

2016-05-30 18:39 GMT+02:00 Peter Senna Tschudin <[email protected]>:

Quoted 46 lines

This driver creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3(imx6q-b850v3.dts). There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However the physical bridges are automatically configured by the input video signal, and the driver has no access to the video processing pipeline. The driver is only needed to read EDID from the STDP2690 and to handle HPD events from the STDP4028. The driver communicates with both bridges over i2c. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

---
 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |   8 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 397 +++++++++++++++++++++++++++++
 4 files changed, 414 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 3273ffa..7bb5e89 100644
@@ -5009,6 +5009,14 @@ W:       https://linuxtv.org
 S:     Maintained
 F:     drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:     Martin Donnelly <[email protected]>
+M:     Peter Senna Tschudin <[email protected]>
+M:     Martyn Welch <[email protected]>
+S:     Maintained
+F:     drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:     Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:     Haavard Skinnemoen <[email protected]>
 S:     Supported
drivers/gpu/drm/bridge/Kconfig +0 -0
index 8f7423f..e9c32fc 100644
@@ -52,4 +52,12 @@ config DRM_PARADE_PS8622
 

drivers/gpu/drm/bridge/Kconfig config DRM_PARADE_PS8622 · line 62

Quoted 10 lines
drivers/gpu/drm/bridge/Kconfig +8 -0
 source "drivers/gpu/drm/bridge/analogix/Kconfig"
 
+config DRM_GE_B850V3_LVDS_DP
+       tristate "LVDS/DP bridge"
+       depends on OF
+       select DRM_KMS_HELPER
+       select DRM_PANEL
+       ---help---
+         Driver for GE B850v3 DP2 LVDS to DP+ Bridge
+

I think the maintainer prefers the entries ordered alphabetically (by vendor, then name), so your config should go before "config DRM_NXP_PTN3460"

drivers/gpu/drm/bridge/Makefile obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o · line 9

Quoted 10 lines
drivers/gpu/drm/bridge/Kconfig +0 -0
 endmenu
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..ba2e355 100644
@@ -6,3 +6,4 @@ obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o

Same here, this needs to be sorted by vendor, then name as well. Also I think is preferred use hyphens like the others drivers.

Hmm, I see now that CONFIG_DRM_ANALOGIX_DP is not sorted, but I guess the preferred is be sorted.

Quoted 84 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +78 -0
new file mode 100644
index 0000000..37a4e7a
@@ -0,0 +1,397 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+
+ *
+ */
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_edid.h"
+#include "drmP.h"
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+               (STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+               (STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+               (STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+       struct drm_connector connector;
+       struct drm_bridge bridge;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c;
+       struct i2c_client *edid_i2c;
+       struct edid *edid;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+               bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+       return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 88

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+static inline struct ge_b850v3_lvds_dp *
+               connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+       return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+static void ge_b850v3_lvds_dp_pre_enable(struct drm_bridge *bridge)
+{
+}
+

You don't need to keep empty callbacks for the (pre/post) enable/disable drm_bridge ops anymore so you can remove this function and the below.

Quoted 71 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +71 -0
+static void ge_b850v3_lvds_dp_enable(struct drm_bridge *bridge)
+{
+}
+
+static void ge_b850v3_lvds_dp_disable(struct drm_bridge *bridge)
+{
+}
+
+static void ge_b850v3_lvds_dp_post_disable(struct drm_bridge *bridge)
+{
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+       struct i2c_adapter *adapter = client->adapter;
+       unsigned char start = 0x00;
+       unsigned int total_size;
+       u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+       struct i2c_msg msgs[] = {
+               {
+                       .addr   = client->addr,
+                       .flags  = 0,
+                       .len    = 1,
+                       .buf    = &start,
+               }, {
+                       .addr   = client->addr,
+                       .flags  = I2C_M_RD,
+                       .len    = EDID_LENGTH,
+                       .buf    = block,
+               }
+       };
+
+       if (!block)
+               return NULL;
+
+       if (i2c_transfer(adapter, msgs, 2) != 2) {
+               DRM_ERROR("Unable to read EDID.\n");
+               goto err;
+       }
+
+       if (!drm_edid_block_valid(block, 0, false, NULL)) {
+               DRM_ERROR("Invalid EDID block\n");
+               goto err;
+       }
+
+       total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+       if (total_size > EDID_LENGTH) {
+               kfree(block);
+               block = kmalloc(total_size, GFP_KERNEL);
+               if (!block)
+                       return NULL;
+
+               /* Yes, read the entire buffer, and do not skip the first
+                * EDID_LENGTH bytes.
+                */
+               start = 0x00;
+               msgs[1].len = total_size;
+               msgs[1].buf = block;
+
+               if (i2c_transfer(adapter, msgs, 2) != 2) {
+                       DRM_ERROR("Unable to read EDID extension blocks.\n");
+                       goto err;
+               }
+       }
+
+       return block;
+
+err:
+       kfree(block);
+       return NULL;

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 169

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       struct i2c_client *client;
+       u8 *block;
+       int num_modes;
+
+       ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);

Move this assignation to ptn_bridge declaration

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 172

Quoted 3 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +3 -0
+       client = ptn_bridge->edid_i2c;
+
+       block = stdp2690_get_edid(client);

I think you can remove the client variable and pass ptn_bridge->edid_i2c directly to the stdp2690_get_edid function.

Note that get_modes is called several times and can be called from userspace, you are allocating memory on every call to get_modes that I think is not released or I'm missing something?

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 174

Quoted 2 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +2 -0
+       ptn_bridge->edid = (struct edid *) block;
+

What if block is NULL here ?

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 178

Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+       drm_mode_connector_update_edid_property(connector, ptn_bridge->edid);
+       num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+
+       return num_modes;

nit: return drm_add_edid_modes(connector, ptn_bridge->edid); ? And remove num_modes ?

Quoted 32 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +32 -0
+}
+
+static struct drm_encoder
+*ge_b850v3_lvds_dp_best_encoder(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               connector_to_ge_b850v3_lvds_dp(connector);
+
+       return ptn_bridge->bridge.encoder;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+       .get_modes = ge_b850v3_lvds_dp_get_modes,
+       .best_encoder = ge_b850v3_lvds_dp_best_encoder,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+               struct drm_connector *connector, bool force)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+                       connector_to_ge_b850v3_lvds_dp(connector);
+       struct i2c_client *ge_b850v3_lvds_dp_i2c =
+                       ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       s32 link_state;
+
+       link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_STS_REG);
+
+       if (link_state == STDP4028_CON_STATE_CONNECTED)
+               return connector_status_connected;
+

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 220

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+       if (link_state == 0)
+               return connector_status_disconnected;
+
+       return connector_status_unknown;
+}
+
+static void ge_b850v3_lvds_dp_connector_force(struct drm_connector *connector)
+{
+}
+

Guess you can remove this empty callback.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 231

Quoted 11 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +11 -0
+static void ge_b850v3_lvds_dp_connector_destroy(struct drm_connector *connector)
+{
+       drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+       .dpms = drm_helper_connector_dpms,
+       .fill_modes = drm_helper_probe_single_connector_modes,
+       .detect = ge_b850v3_lvds_dp_detect,
+       .destroy = ge_b850v3_lvds_dp_connector_destroy,
+       .force = ge_b850v3_lvds_dp_connector_force,

Remove .force

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 247

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       if (ptn_bridge->connector.dev)
+               drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+       return IRQ_HANDLED;
+

Remove the empty line.

Quoted 44 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +44 -0
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge
+                       = bridge_to_ge_b850v3_lvds_dp(bridge);
+       struct drm_connector *connector = &ptn_bridge->connector;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       int ret;
+
+       if (!bridge->encoder) {
+               DRM_ERROR("Parent encoder object not found");
+               return -ENODEV;
+       }
+
+       connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+       drm_connector_helper_add(connector,
+                       &ge_b850v3_lvds_dp_connector_helper_funcs);
+
+       ret = drm_connector_init(bridge->dev, connector,
+                       &ge_b850v3_lvds_dp_connector_funcs,
+                       DRM_MODE_CONNECTOR_DisplayPort);
+       if (ret) {
+               DRM_ERROR("Failed to initialize connector with drm\n");
+               return ret;
+       }
+
+       ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+       if (ret)
+               return ret;
+
+       drm_bridge_enable(bridge);
+       if (ge_b850v3_lvds_dp_i2c->irq) {
+               drm_helper_hpd_irq_event(connector->dev);
+
+               ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+                               ge_b850v3_lvds_dp_i2c->irq, NULL,
+                               ge_b850v3_lvds_dp_irq_handler,
+                               IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+                               "ge_b850v3_lvds_dp", ptn_bridge);
+               if (ret)
+                       return ret;

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 301

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+       }
+
+       return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+       .pre_enable = ge_b850v3_lvds_dp_pre_enable,
+       .enable = ge_b850v3_lvds_dp_enable,
+       .disable = ge_b850v3_lvds_dp_disable,
+       .post_disable = ge_b850v3_lvds_dp_post_disable,

Remove the above empty callbacks.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 310

Quoted 9 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +9 -0
+       .attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+                               const struct i2c_device_id *id)
+{
+       struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+

Remove this empty line between these two blocks.

Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+       int ret;
+       u32 edid_i2c_reg;
+
+       ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 324

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+       if (!ptn_bridge)
+               return -ENOMEM;
+
+       ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+       ptn_bridge->bridge.driver_private = ptn_bridge;
+       i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+       ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+       if (ret) {
+               dev_err(dev, "edid-reg not specified, aborting...\n");

Use DRM_ERROR?

Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+               return -ENODEV;
+       }
+
+       ptn_bridge->edid_i2c = devm_kzalloc(dev,

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 338

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+                       sizeof(struct i2c_client), GFP_KERNEL);
+
+       if (!ptn_bridge->edid_i2c)
+               return -ENOMEM;
+
+       memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+                       sizeof(struct i2c_client));
+
+       ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+

Remove empty line.

Quoted 60 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +59 -1
+
+       /* Configures the bridge to re-enable interrupts after each ack */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+       /* Clear pending interrupts since power up. */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+       ptn_bridge->bridge.of_node = dev->of_node;
+       ret = drm_bridge_add(&ptn_bridge->bridge);
+       if (ret) {
+               DRM_ERROR("Failed to add bridge\n");
+               return ret;
+       }
+
+       return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+       drm_bridge_remove(&ptn_bridge->bridge);
+
+       return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+       {"b850v3_lvds_dp", 0},
+       {},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+       { .compatible = "ge,b850v3_lvds_dp" },
+       {},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+       .id_table       = ge_b850v3_lvds_dp_i2c_table,
+       .probe          = ge_b850v3_lvds_dp_probe,
+       .remove         = ge_b850v3_lvds_dp_remove,
+       .driver         = {
+               .name           = "ge,b850v3_lvds_dp",
+               .of_match_table = ge_b850v3_lvds_dp_match,
+       },
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <[email protected]>");
+MODULE_AUTHOR("Martyn Welch <[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ bridge)");
+MODULE_LICENSE("GPL v2");
--
Quoted 2 lines

2.5.5

Best regards,
Enric

#12 lore raw

PATCH 5/5 arm/dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Configure the GE B850v3 to use the LVDS/DP++ bridge.

---
 arch/arm/boot/dts/imx6q-b850v3.dts | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +31 -0
index 88a70de..d366384 100644
@@ -77,6 +77,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -147,3 +154,27 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3_lvds_dp_bridge {
+		compatible = "ge,b850v3_lvds_dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port@0 {
+			reg = <0>;
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
reviewer
#13 lore raw

Re: [PATCH 5/5] arm/dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

arch/arm/boot/dts/imx6q-b850v3.dts line 167

Quoted 10 lines · Peter Senna Tschudin

+&mux2_i2c2 {

arch/arm/boot/dts/imx6q-b850v3.dts +9 -0
+       status = "okay";
+       clock-frequency = <100000>;
+
+       b850v3_lvds_dp_bridge {
+               compatible = "ge,b850v3_lvds_dp";
+               #address-cells = <1>;
+               #size-cells = <0>;
+
+               reg = <0x73>;

You use 'reg = <0x73>' here, but no @73.

#14 lore raw

v2 cover Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

The series adds a driver that creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

The patches from the series:
[1/5] Change the imx-ldb driver to allow attaching a bridge and not only a LVDS
panel.

[2/5] Configure the mapping between IPUs and external displays on the dts file
of the B850v3. Needed if connect two Full-HD monitors.

[3/5] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[4/5] Add the driver, make changes to MAINTAINERS, Kconfig and Makefile

[5/5] Make the changes to the B850v3 dts file to enable the GE B850v3
LVDS/DP++ Bridge.

Peter Senna Tschudin (5):
drm/imx-ldb: Add support to drm-bridge dts/imx6q-b850v3: Configure IPU assignment order Documentation/devicetree/bindings: b850v3_lvds_dp drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      |  38 ++
 MAINTAINERS                                        |   8 +
 arch/arm/boot/dts/imx6q-b850v3.dts                 |  36 ++
 drivers/gpu/drm/bridge/Kconfig                     |  11 +
 drivers/gpu/drm/bridge/Makefile                    |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c         | 392 +++++++++++++++++++++
 drivers/gpu/drm/imx/imx-ldb.c                      | 121 ++++---
 7 files changed, 565 insertions(+), 42 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#15 lore raw

PATCH v2 1/5 drm/imx-ldb: Add support to drm-bridge

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>

---
Changes from V1:
- Reanmed ext_bridge to bridge
- Removed empty entry point imx_ldb_encoder_enable()
- Adapted the code to apply to the latest linux next: next-20160609

 drivers/gpu/drm/imx/imx-ldb.c | 121 +++++++++++++++++++++++++++---------------
 1 file changed, 79 insertions(+), 42 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +79 -42
index beff793..55e94cc 100644
@@ -58,7 +58,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -422,16 +426,27 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, &imx_ldb_ch->encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
-
-	if (imx_ldb_ch->panel)
+	if (imx_ldb_ch->panel) {
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
 		drm_panel_attach(imx_ldb_ch->panel, &imx_ldb_ch->connector);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				&imx_ldb_ch->encoder);
+	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
-			&imx_ldb_ch->encoder);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = &imx_ldb_ch->encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	}
 
 	return 0;
 }
@@ -501,6 +516,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -508,7 +562,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -558,7 +611,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 
 		ret = of_property_read_u32(child, "reg", &i);
@@ -590,46 +642,31 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
+
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			/*
+			 * If the bridge is compiled as a module, it may take
+			 * some time until the bridge driver is available.
+			 * Defer until the bridge driver is ready.
+			 */
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)
-					channel->mode_valid = 1;
-			}
+		/* Only does panel ddc work if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		channel->bus_format = of_get_bus_format(dev, child);
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#16 lore raw

PATCH v2 2/5 dts/imx6q-b850v3: Configure IPU assignment order

In reply to Peter Senna Tschudin

As the IPU has combined limitations across multiple crtcs, and as that can't be communicated to userspace at the moment, reorder the crtcs to allow support to two Full-HD monitors by avoiding assigning two monitors to a single IPU.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V1:
- New commit message

 arch/arm/boot/dts/imx6q-b850v3.dts | 5 +++++
 1 file changed, 5 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +5 -0
index 167f744..88a70de 100644
@@ -51,6 +51,11 @@
 	chosen {
 		stdout-path = &uart3;
 	};
+
+	display-subsystem {
+		compatible = "fsl,imx-display-subsystem";
+		ports = <&ipu1_di0>, <&ipu2_di0>, <&ipu1_di1>, <&ipu2_di1>;
+	};
 };
 
 &clks {
Signature · 1 line
2.5.5
#17 lore raw

PATCH v2 3/5 Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +38 -0
new file mode 100644
index 0000000..46bbea9
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : should link to the gpio used as interrupt
+    source on the host.
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the vide signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port@0 {
+			reg = <0>;
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
maintainer Acked-by on v2 3/5
#18 lore raw

Re: [PATCH V2 3/5] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 8 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Quoted 8 lines

---
Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
#19 lore raw

Re: [PATCH V2 3/5] Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Hello Peter,

Quoted 23 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3_lvds_dp.txt      | 38 ++++++++++++++++++++++
 1 file changed, 38 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +0 -0
new file mode 100644
index 0000000..46bbea9

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 4

Quoted 4 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +4 -0
@@ -0,0 +1,38 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3_lvds_dp".

It seems you forgot to replace '_' by '-' (you did in the example though)

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 7

Quoted 3 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +3 -0
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : should link to the gpio used as interrupt
+    source on the host.

Is the interrupt parent always a GPIO controller since that is what this description says.

Shouldn't be instead something like instead?

interrupt-parent: phandle of the interrupt controller that services interrupts to the device

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 11

Quoted 4 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +4 -0
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the vide signal connection between the host

s/vide/video

Quoted 11 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +11 -0
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+       status = "okay";
+       clock-frequency = <100000>;
+
+       b850v3-lvds-dp-bridge@73  {
+               compatible = "ge,b850v3-lvds-dp";
+               #address-cells = <1>;

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt line 32

Quoted 10 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +10 -0
+               #size-cells = <0>;
+
+               reg = <0x73>;
+               interrupt-parent = <&gpio2>;
+               interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+               edid-reg = <0x72>;
+
+               port@0 {
+                       reg = <0>;

AFAIU a unit-address and reg property for ports are only needed if you have more than one port according to Documentation/devicetree/bindings/graph.txt and Documentation/devicetree/bindings/media/video-interfaces.txt.

Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt

Quoted 7 lines
Documentation/devicetree/bindings/ge/b850v3_lvds_dp.txt new file +6 -1
+                       b850v3_dp_bridge_in: endpoint {
+                               remote-endpoint = <&lvds0_out>;
+                       };
+               };
+       };
+};
--

Best regards,
Javier
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#20 lore raw

PATCH v2 4/5 drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 392 +++++++++++++++++++++++++++++
 4 files changed, 412 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 2ce5e91..2dd3d7f 100644
@@ -5010,6 +5010,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Martin Donnelly <[email protected]>
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index 8f7423f..93dae5bd 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..47ea6c1 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +392 -0

392 lines not on this page · raw message

Signature · 1 line
2.5.5
#21 lore raw

Re: [PATCH V2 4/5] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Peter,

Only a few comments ;)

2016-06-09 18:25 GMT+02:00 Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>:

Quoted 177 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 392 +++++++++++++++++++++++++++++
 4 files changed, 412 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 2ce5e91..2dd3d7f 100644
@@ -5010,6 +5010,14 @@ W:       https://linuxtv.org
 S:     Maintained
 F:     drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:     Martin Donnelly <[email protected]>
+M:     Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:     Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:     Maintained
+F:     drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:     Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:     Haavard Skinnemoen <[email protected]>
 S:     Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index 8f7423f..93dae5bd 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
          Designware HDMI block.  This is used in conjunction with
          the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+       tristate "GE B850v3 LVDS to DP++ display bridge"
+       depends on OF
+       select DRM_KMS_HELPER
+       select DRM_PANEL
+       ---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
        tristate "NXP PTN3460 DP/LVDS bridge"
        depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..47ea6c1 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +79 -0
new file mode 100644
index 0000000..c73cd77
@@ -0,0 +1,392 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+
+ *
+ */
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_edid.h"
+#include "drmP.h"
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+               (STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+               (STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+               (STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+       struct drm_connector connector;
+       struct drm_bridge bridge;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c;
+       struct i2c_client *edid_i2c;
+       struct edid *edid;
+       struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+               bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+       return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 89

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+static inline struct ge_b850v3_lvds_dp *
+               connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+       return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+static void ge_b850v3_lvds_dp_enable(struct drm_bridge *bridge)
+{
+}
+

You can remove this function, see
http://lxr.free-electrons.com/source/drivers/gpu/drm/drm_bridge.c#L277

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 93

Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+static void ge_b850v3_lvds_dp_disable(struct drm_bridge *bridge)
+{
+}
+

And this one, see
http://lxr.free-electrons.com/source/drivers/gpu/drm/drm_bridge.c#L182

Quoted 193 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +193 -0
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+       struct i2c_adapter *adapter = client->adapter;
+       unsigned char start = 0x00;
+       unsigned int total_size;
+       u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+       struct i2c_msg msgs[] = {
+               {
+                       .addr   = client->addr,
+                       .flags  = 0,
+                       .len    = 1,
+                       .buf    = &start,
+               }, {
+                       .addr   = client->addr,
+                       .flags  = I2C_M_RD,
+                       .len    = EDID_LENGTH,
+                       .buf    = block,
+               }
+       };
+
+       if (!block)
+               return NULL;
+
+       if (i2c_transfer(adapter, msgs, 2) != 2) {
+               DRM_ERROR("Unable to read EDID.\n");
+               goto err;
+       }
+
+       if (!drm_edid_block_valid(block, 0, false, NULL)) {
+               DRM_ERROR("Invalid EDID block\n");
+               goto err;
+       }
+
+       total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+       if (total_size > EDID_LENGTH) {
+               kfree(block);
+               block = kmalloc(total_size, GFP_KERNEL);
+               if (!block)
+                       return NULL;
+
+               /* Yes, read the entire buffer, and do not skip the first
+                * EDID_LENGTH bytes.
+                */
+               start = 0x00;
+               msgs[1].len = total_size;
+               msgs[1].buf = block;
+
+               if (i2c_transfer(adapter, msgs, 2) != 2) {
+                       DRM_ERROR("Unable to read EDID extension blocks.\n");
+                       goto err;
+               }
+       }
+
+       return block;
+
+err:
+       kfree(block);
+       return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       struct i2c_client *client;
+       int num_modes = 0;
+
+       ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+       client = ptn_bridge->edid_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       kfree(ptn_bridge->edid);
+       ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+       if (ptn_bridge->edid) {
+               drm_mode_connector_update_edid_property(connector,
+                               ptn_bridge->edid);
+               num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+       }
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       return num_modes;
+}
+
+static struct drm_encoder
+*ge_b850v3_lvds_dp_best_encoder(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               connector_to_ge_b850v3_lvds_dp(connector);
+
+       return ptn_bridge->bridge.encoder;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+       .get_modes = ge_b850v3_lvds_dp_get_modes,
+       .best_encoder = ge_b850v3_lvds_dp_best_encoder,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+               struct drm_connector *connector, bool force)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+                       connector_to_ge_b850v3_lvds_dp(connector);
+       struct i2c_client *ge_b850v3_lvds_dp_i2c =
+                       ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       s32 link_state;
+
+       link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_STS_REG);
+
+       if (link_state == STDP4028_CON_STATE_CONNECTED)
+               return connector_status_connected;
+
+       if (link_state == 0)
+               return connector_status_disconnected;
+
+       return connector_status_unknown;
+}
+
+static void ge_b850v3_lvds_dp_connector_destroy(struct drm_connector *connector)
+{
+       drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+       .dpms = drm_helper_connector_dpms,
+       .fill_modes = drm_helper_probe_single_connector_modes,
+       .detect = ge_b850v3_lvds_dp_detect,
+       .destroy = ge_b850v3_lvds_dp_connector_destroy,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       if (ptn_bridge->connector.dev)
+               drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+       return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge
+                       = bridge_to_ge_b850v3_lvds_dp(bridge);
+       struct drm_connector *connector = &ptn_bridge->connector;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       int ret;
+
+       if (!bridge->encoder) {
+               DRM_ERROR("Parent encoder object not found");
+               return -ENODEV;
+       }
+
+       connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+       drm_connector_helper_add(connector,
+                       &ge_b850v3_lvds_dp_connector_helper_funcs);
+
+       ret = drm_connector_init(bridge->dev, connector,
+                       &ge_b850v3_lvds_dp_connector_funcs,
+                       DRM_MODE_CONNECTOR_DisplayPort);
+       if (ret) {
+               DRM_ERROR("Failed to initialize connector with drm\n");
+               return ret;
+       }
+
+       ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+       if (ret)
+               return ret;
+
+       drm_bridge_enable(bridge);
+       if (ge_b850v3_lvds_dp_i2c->irq) {
+               drm_helper_hpd_irq_event(connector->dev);
+
+               ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+                               ge_b850v3_lvds_dp_i2c->irq, NULL,
+                               ge_b850v3_lvds_dp_irq_handler,
+                               IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+                               "ge-b850v3-lvds-dp", ptn_bridge);

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 296

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+               if (ret)
+                       return ret;
+       }
+
+       return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+       .enable = ge_b850v3_lvds_dp_enable,
+       .disable = ge_b850v3_lvds_dp_disable,

Remove the above empty callbacks.

Quoted 55 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +55 -0
+       .attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+                               const struct i2c_device_id *id)
+{
+       struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       int ret;
+       u32 edid_i2c_reg;
+
+       ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+       if (!ptn_bridge)
+               return -ENOMEM;
+
+       mutex_init(&ptn_bridge->lock);
+
+       ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+       ptn_bridge->bridge.driver_private = ptn_bridge;
+       i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+       ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+       if (ret) {
+               dev_err(dev, "edid-reg not specified, aborting...\n");
+               return -ENODEV;
+       }
+
+       ptn_bridge->edid_i2c = devm_kzalloc(dev,
+                       sizeof(struct i2c_client), GFP_KERNEL);
+
+       if (!ptn_bridge->edid_i2c)
+               return -ENOMEM;
+
+       memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+                       sizeof(struct i2c_client));
+
+       ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+       /* Configures the bridge to re-enable interrupts after each ack */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+       /* Clear pending interrupts since power up. */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+       ptn_bridge->bridge.of_node = dev->of_node;
+       ret = drm_bridge_add(&ptn_bridge->bridge);
+       if (ret) {
+               DRM_ERROR("Failed to add bridge\n");
+               return ret;
+       }

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 361

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+
+       return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+       drm_bridge_remove(&ptn_bridge->bridge);

Guess you need to free ptn_bridge->edid here.

Quoted 32 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +31 -1
+
+       return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+       {"b850v3-lvds-dp", 0},
+       {},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+       { .compatible = "ge,b850v3-lvds-dp" },
+       {},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+       .id_table       = ge_b850v3_lvds_dp_i2c_table,
+       .probe          = ge_b850v3_lvds_dp_probe,
+       .remove         = ge_b850v3_lvds_dp_remove,
+       .driver         = {
+               .name           = "ge,b850v3-lvds-dp",
+               .of_match_table = ge_b850v3_lvds_dp_match,
+       },
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>");
+MODULE_AUTHOR("Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ bridge)");
+MODULE_LICENSE("GPL v2");
--
Quoted 2 lines

2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#22 lore raw

Re: [PATCH V2 4/5] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Enric,

Quoted 3 lines · Enric Balletbo Serra

Hi Peter,

Only a few comments ;)

Thanks a lot for the review!

Quoted 186 lines

2016-06-09 18:25 GMT+02:00 Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>:

Quoted 184 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and

firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 392 +++++++++++++++++++++++++++++
 4 files changed, 412 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 2ce5e91..2dd3d7f 100644
@@ -5010,6 +5010,14 @@ W:       https://linuxtv.org
 S:     Maintained
 F:     drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:     Martin Donnelly <[email protected]>
+M:     Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:     Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:     Maintained
+F:     drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:     Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:     Haavard Skinnemoen <[email protected]>
 S:     Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index 8f7423f..93dae5bd 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
          Designware HDMI block.  This is used in conjunction with
 
          the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+       tristate "GE B850v3 LVDS to DP++ display bridge"
+       depends on OF
+       select DRM_KMS_HELPER
+       select DRM_PANEL
+       ---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
        tristate "NXP PTN3460 DP/LVDS bridge"

depends on OF

drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..47ea6c1 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +84 -0
new file mode 100644
index 0000000..c73cd77
@@ -0,0 +1,392 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+
+ *
+ */
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_edid.h"
+#include "drmP.h"
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+               (STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+               (STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+               (STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+       struct drm_connector connector;
+       struct drm_bridge bridge;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c;
+       struct i2c_client *edid_i2c;
+       struct edid *edid;
+       struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+               bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+       return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+               connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+       return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 89

Quoted 5 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +5 -0
+
+static void ge_b850v3_lvds_dp_enable(struct drm_bridge *bridge)
+{
+}
+

You can remove this function, see
http://lxr.free-electrons.com/source/drivers/gpu/drm/drm_bridge.c#L277

You are right, I removed the empty callbacks, but they returned due me doing something wrong when rebasing. V3 will be out soon.

Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 93

Quoted 5 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+static void ge_b850v3_lvds_dp_disable(struct drm_bridge *bridge)
+{
+}
+

And this one, see
http://lxr.free-electrons.com/source/drivers/gpu/drm/drm_bridge.c#L182

Same. This should not be here.

Quoted 199 lines
Quoted 199 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +197 -0
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+       struct i2c_adapter *adapter = client->adapter;
+       unsigned char start = 0x00;
+       unsigned int total_size;
+       u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+       struct i2c_msg msgs[] = {
+               {
+                       .addr   = client->addr,
+                       .flags  = 0,
+                       .len    = 1,
+                       .buf    = &start,
+               }, {
+                       .addr   = client->addr,
+                       .flags  = I2C_M_RD,
+                       .len    = EDID_LENGTH,
+                       .buf    = block,
+               }
+       };
+
+       if (!block)
+               return NULL;
+
+       if (i2c_transfer(adapter, msgs, 2) != 2) {
+               DRM_ERROR("Unable to read EDID.\n");
+               goto err;
+       }
+
+       if (!drm_edid_block_valid(block, 0, false, NULL)) {
+               DRM_ERROR("Invalid EDID block\n");
+               goto err;
+       }
+
+       total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+       if (total_size > EDID_LENGTH) {
+               kfree(block);
+               block = kmalloc(total_size, GFP_KERNEL);
+               if (!block)
+                       return NULL;
+
+               /* Yes, read the entire buffer, and do not skip the first
+                * EDID_LENGTH bytes.
+                */
+               start = 0x00;
+               msgs[1].len = total_size;
+               msgs[1].buf = block;
+
+               if (i2c_transfer(adapter, msgs, 2) != 2) {
+                       DRM_ERROR("Unable to read EDID extension blocks.\n");
+                       goto err;
+               }
+       }
+
+       return block;
+
+err:
+       kfree(block);
+       return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       struct i2c_client *client;
+       int num_modes = 0;
+
+       ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+       client = ptn_bridge->edid_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       kfree(ptn_bridge->edid);
+       ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+       if (ptn_bridge->edid) {
+               drm_mode_connector_update_edid_property(connector,
+                               ptn_bridge->edid);
+               num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+       }
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       return num_modes;
+}
+
+static struct drm_encoder
+*ge_b850v3_lvds_dp_best_encoder(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               connector_to_ge_b850v3_lvds_dp(connector);
+
+       return ptn_bridge->bridge.encoder;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+       .get_modes = ge_b850v3_lvds_dp_get_modes,
+       .best_encoder = ge_b850v3_lvds_dp_best_encoder,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+               struct drm_connector *connector, bool force)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+                       connector_to_ge_b850v3_lvds_dp(connector);
+       struct i2c_client *ge_b850v3_lvds_dp_i2c =
+                       ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       s32 link_state;
+
+       link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_STS_REG);
+
+       if (link_state == STDP4028_CON_STATE_CONNECTED)
+               return connector_status_connected;
+
+       if (link_state == 0)
+               return connector_status_disconnected;
+
+       return connector_status_unknown;
+}
+
+static void ge_b850v3_lvds_dp_connector_destroy(struct drm_connector *connector)
+{
+       drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+       .dpms = drm_helper_connector_dpms,
+       .fill_modes = drm_helper_probe_single_connector_modes,
+       .detect = ge_b850v3_lvds_dp_detect,
+       .destroy = ge_b850v3_lvds_dp_connector_destroy,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
 
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       if (ptn_bridge->connector.dev)
+               drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+       return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge
+                       = bridge_to_ge_b850v3_lvds_dp(bridge);
+       struct drm_connector *connector = &ptn_bridge->connector;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       int ret;
+
+       if (!bridge->encoder) {
+               DRM_ERROR("Parent encoder object not found");
+               return -ENODEV;
+       }
+
+       connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+       drm_connector_helper_add(connector,
+                       &ge_b850v3_lvds_dp_connector_helper_funcs);
+
+       ret = drm_connector_init(bridge->dev, connector,
+                       &ge_b850v3_lvds_dp_connector_funcs,
+                       DRM_MODE_CONNECTOR_DisplayPort);
+       if (ret) {
+               DRM_ERROR("Failed to initialize connector with drm\n");
+               return ret;
+       }
+
+       ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+       if (ret)
+               return ret;
+
+       drm_bridge_enable(bridge);
+       if (ge_b850v3_lvds_dp_i2c->irq) {
+               drm_helper_hpd_irq_event(connector->dev);
+
+               ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+                               ge_b850v3_lvds_dp_i2c->irq, NULL,
+                               ge_b850v3_lvds_dp_irq_handler,
+                               IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+                               "ge-b850v3-lvds-dp", ptn_bridge);
+               if (ret)
+                       return ret;
+       }
+
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 296

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+       return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+       .enable = ge_b850v3_lvds_dp_enable,
+       .disable = ge_b850v3_lvds_dp_disable,

Remove the above empty callbacks.

Here too.

Quoted 60 lines
Quoted 60 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +59 -0
+       .attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+                               const struct i2c_device_id *id)
+{
+       struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       int ret;
+       u32 edid_i2c_reg;
+
+       ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+       if (!ptn_bridge)
+               return -ENOMEM;
+
+       mutex_init(&ptn_bridge->lock);
+
+       ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+       ptn_bridge->bridge.driver_private = ptn_bridge;
+       i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+       ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+       if (ret) {
+               dev_err(dev, "edid-reg not specified, aborting...\n");
+               return -ENODEV;
+       }
+
+       ptn_bridge->edid_i2c = devm_kzalloc(dev,
+                       sizeof(struct i2c_client), GFP_KERNEL);
+
+       if (!ptn_bridge->edid_i2c)
+               return -ENOMEM;
+
+       memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+                       sizeof(struct i2c_client));
+
+       ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+       /* Configures the bridge to re-enable interrupts after each ack */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+       /* Clear pending interrupts since power up. */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+       ptn_bridge->bridge.of_node = dev->of_node;
+       ret = drm_bridge_add(&ptn_bridge->bridge);
+       if (ret) {
+               DRM_ERROR("Failed to add bridge\n");
+               return ret;
+       }
+
+       return 0;
+}
+
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 361

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+       drm_bridge_remove(&ptn_bridge->bridge);

Guess you need to free ptn_bridge->edid here.

Thanks a lot! I'll fix it when sending V3.

Thank you for the review!

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#23 lore raw

Re: [PATCH V2 4/5] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin
Quoted 23 lines · Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>
Quoted 8 lines

---
Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

Two comments below.

Quoted 149 lines
 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 392 +++++++++++++++++++++++++++++
 4 files changed, 412 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index 2ce5e91..2dd3d7f 100644
@@ -5010,6 +5010,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Martin Donnelly <[email protected]>
+M:	Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:	Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index 8f7423f..93dae5bd 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..47ea6c1 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +82 -0
new file mode 100644
index 0000000..c73cd77
@@ -0,0 +1,392 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+
+ *
+ */
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_edid.h"
+#include "drmP.h"
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 92

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+static void ge_b850v3_lvds_dp_enable(struct drm_bridge *bridge)
+{
+}
+
+static void ge_b850v3_lvds_dp_disable(struct drm_bridge *bridge)
+{
+}

Please remove the above two dummy funcs, they should be unecessary. If they're not that would be an issue in there core - we want to avoid boilerplat like this.

Quoted 85 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +85 -0
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 187

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+}
+
+static struct drm_encoder
+*ge_b850v3_lvds_dp_best_encoder(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		connector_to_ge_b850v3_lvds_dp(connector);
+
+	return ptn_bridge->bridge.encoder;
+}

Please remove your best_encoder callback, that just became unecessary in drm-next, thanks to some work from Boris.
-Daniel

Quoted 212 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +205 -0
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.best_encoder = ge_b850v3_lvds_dp_best_encoder,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static void ge_b850v3_lvds_dp_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = ge_b850v3_lvds_dp_connector_destroy,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.enable = ge_b850v3_lvds_dp_enable,
+	.disable = ge_b850v3_lvds_dp_disable,
+	.attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+				const struct i2c_device_id *id)
+{
+	struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	int ret;
+	u32 edid_i2c_reg;
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->lock);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+	if (ret) {
+		dev_err(dev, "edid-reg not specified, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->edid_i2c = devm_kzalloc(dev,
+			sizeof(struct i2c_client), GFP_KERNEL);
+
+	if (!ptn_bridge->edid_i2c)
+		return -ENOMEM;
+
+	memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+			sizeof(struct i2c_client));
+
+	ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+	/* Configures the bridge to re-enable interrupts after each ack */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	/* Clear pending interrupts since power up. */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	ret = drm_bridge_add(&ptn_bridge->bridge);
+	if (ret) {
+		DRM_ERROR("Failed to add bridge\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+	drm_bridge_remove(&ptn_bridge->bridge);
+
+	return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+	{"b850v3-lvds-dp", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+	{ .compatible = "ge,b850v3-lvds-dp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+	.id_table	= ge_b850v3_lvds_dp_i2c_table,
+	.probe		= ge_b850v3_lvds_dp_probe,
+	.remove		= ge_b850v3_lvds_dp_remove,
+	.driver		= {
+		.name		= "ge,b850v3-lvds-dp",
+		.of_match_table = ge_b850v3_lvds_dp_match,
+	},
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>");
+MODULE_AUTHOR("Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ bridge)");
+MODULE_LICENSE("GPL v2");
Signature · 6 lines
2.5.5

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#24 lore raw

Re: [PATCH V2 4/5] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin
Quoted 13 lines · Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Quoted 1 line
   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Are these two chips always expected to be used together? I don't think it's right to pair up two encoder chips into one driver just for one board.

Is one device @0x72 and other @0x73? Or is only one of them an i2c slave?

What's preventing us to create these as two different bridge drivers? The drm framework allows us to daisy chain encoder bridges. The only problem I see is that we don't have a clear-cut way to tell the bridge driver whether we want it to create a connector for us or not. Because, it looks like both can potentially create connectors. This isn't a big problem either if we have DT. We just need to check whether our output port is connected to another bridge or a connector.

Thanks,
Archit

Quoted 477 lines
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

  MAINTAINERS                                |   8 +
  drivers/gpu/drm/bridge/Kconfig             |  11 +
  drivers/gpu/drm/bridge/Makefile            |   1 +
  drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 392 +++++++++++++++++++++++++++++
  4 files changed, 412 insertions(+)

create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c

MAINTAINERS +8 -0
index 2ce5e91..2dd3d7f 100644
@@ -5010,6 +5010,14 @@ W:	https://linuxtv.org
  S:	Maintained
  F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Martin Donnelly <[email protected]>
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
  GENERIC GPIO I2C DRIVER
  M:	Haavard Skinnemoen <[email protected]>
  S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index 8f7423f..93dae5bd 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
  	  Designware HDMI block.  This is used in conjunction with
  	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
  config DRM_NXP_PTN3460
  	tristate "NXP PTN3460 DP/LVDS bridge"
  	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index 96b13b3..47ea6c1 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
  obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
  obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
  obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
  obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
  obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
  obj-$(CONFIG_DRM_ANALOGIX_DP) += analogix/
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +392 -0
new file mode 100644
index 0000000..c73cd77
@@ -0,0 +1,392 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+
+ *
+ */
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_edid.h"
+#include "drmP.h"
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+static void ge_b850v3_lvds_dp_enable(struct drm_bridge *bridge)
+{
+}
+
+static void ge_b850v3_lvds_dp_disable(struct drm_bridge *bridge)
+{
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+static struct drm_encoder
+*ge_b850v3_lvds_dp_best_encoder(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		connector_to_ge_b850v3_lvds_dp(connector);
+
+	return ptn_bridge->bridge.encoder;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.best_encoder = ge_b850v3_lvds_dp_best_encoder,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static void ge_b850v3_lvds_dp_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = ge_b850v3_lvds_dp_connector_destroy,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.enable = ge_b850v3_lvds_dp_enable,
+	.disable = ge_b850v3_lvds_dp_disable,
+	.attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+				const struct i2c_device_id *id)
+{
+	struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	int ret;
+	u32 edid_i2c_reg;
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->lock);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+	if (ret) {
+		dev_err(dev, "edid-reg not specified, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->edid_i2c = devm_kzalloc(dev,
+			sizeof(struct i2c_client), GFP_KERNEL);
+
+	if (!ptn_bridge->edid_i2c)
+		return -ENOMEM;
+
+	memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+			sizeof(struct i2c_client));
+
+	ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+	/* Configures the bridge to re-enable interrupts after each ack */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	/* Clear pending interrupts since power up. */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	ret = drm_bridge_add(&ptn_bridge->bridge);
+	if (ret) {
+		DRM_ERROR("Failed to add bridge\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+	drm_bridge_remove(&ptn_bridge->bridge);
+
+	return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+	{"b850v3-lvds-dp", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+	{ .compatible = "ge,b850v3-lvds-dp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+	.id_table	= ge_b850v3_lvds_dp_i2c_table,
+	.probe		= ge_b850v3_lvds_dp_probe,
+	.remove		= ge_b850v3_lvds_dp_remove,
+	.driver		= {
+		.name		= "ge,b850v3-lvds-dp",
+		.of_match_table = ge_b850v3_lvds_dp_match,
+	},
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <[email protected]>");
+MODULE_AUTHOR("Martyn Welch <[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ bridge)");
+MODULE_LICENSE("GPL v2");
Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
#25 lore raw

PATCH v2 5/5 dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Configures the GE B850v3 LVDS/DP++ bridge on the dts file.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to b850v3-lvds-dp-bridge

 arch/arm/boot/dts/imx6q-b850v3.dts | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +31 -0
index 88a70de..a57dd5b 100644
@@ -77,6 +77,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -147,3 +154,27 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73 {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port@0 {
+			reg = <0>;
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#26 lore raw

v3 cover Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

The series adds a driver that creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

The patches from the series:
[1/5] Change the imx-ldb driver to allow attaching a bridge and not only a LVDS
panel.

[2/5] Configure the mapping between IPUs and external displays on the dts file
of the B850v3. Needed to support two simultaneos Full-HD monitors.

[3/5] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[4/5] Add the driver, make changes to MAINTAINERS, Kconfig and Makefile

[5/5] Make the changes to the B850v3 dts file to enable the GE B850v3
LVDS/DP++ Bridge.

Peter Senna Tschudin (5):
drm/imx-ldb: Add support to drm-bridge dts/imx6q-b850v3: Configure IPU assignment order Documentation/devicetree/bindings: b850v3_lvds_dp drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      |  37 ++
 MAINTAINERS                                        |   8 +
 arch/arm/boot/dts/imx6q-b850v3.dts                 |  35 ++
 drivers/gpu/drm/bridge/Kconfig                     |  11 +
 drivers/gpu/drm/bridge/Makefile                    |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c         | 397 +++++++++++++++++++++
 drivers/gpu/drm/imx/imx-ldb.c                      | 111 +++---
 7 files changed, 559 insertions(+), 41 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#27 lore raw

PATCH v3 1/5 drm/imx-ldb: Add support to drm-bridge

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>

---
Changes from V2:
- Updated to be aplied on top of Liu Ying changes that made imx-ldb atomic.
- Tested on next-20160729.

Changes from V1:
- Reanmed ext_bridge to bridge
- Removed empty entry point imx_ldb_encoder_enable()
- Adapted the code to apply to the latest linux next: next-20160609

 drivers/gpu/drm/imx/imx-ldb.c | 111 ++++++++++++++++++++++++++----------------
 1 file changed, 70 insertions(+), 41 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +70 -41
index 5d2831d..ebe9abd 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -469,19 +473,28 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
-
 	if (imx_ldb_ch->panel) {
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
 		ret = drm_panel_attach(imx_ldb_ch->panel,
-				       &imx_ldb_ch->connector);
+				&imx_ldb_ch->connector);
 		if (ret)
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	}
 
 	return 0;
 }
@@ -551,6 +564,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -558,7 +610,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -608,7 +659,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 		int bus_format;
 
@@ -641,46 +691,25 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)
-					channel->mode_valid = 1;
-			}
+		/* panel ddc only if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		bus_format = of_get_bus_format(dev, child);
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#28 lore raw

Re: [PATCH V3 1/5] drm/imx-ldb: Add support to drm-bridge

Am Sonntag, den 31.07.2016, 21:55 +0200 schrieb Peter Senna Tschudin:

Quoted 20 lines

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>

---
Changes from V2:
- Updated to be aplied on top of Liu Ying changes that made imx-ldb atomic.
- Tested on next-20160729.

[...]

Quoted 5 lines
drivers/gpu/drm/imx/imx-ldb.c +0 -1
@@ -469,19 +473,28 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_register(struct drm_device *drm, · line 481

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +5 -4
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
-
 	if (imx_ldb_ch->panel) {
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);

This is still not right. We want to add the connector whenever there is no bridge that brings its own, not only when there is a panel. For historical reasons, the ldb driver can also work without a panel.

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_register(struct drm_device *drm, · line 483

Quoted 3 lines
drivers/gpu/drm/imx/imx-ldb.c +1 -1
 		ret = drm_panel_attach(imx_ldb_ch->panel,
-				       &imx_ldb_ch->connector);
+				&imx_ldb_ch->connector);

What is the purpose of this change?

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_register(struct drm_device *drm, · line 484

Quoted 5 lines
drivers/gpu/drm/imx/imx-ldb.c +0 -1
 		if (ret)
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);

Where is this gone?

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_register(struct drm_device *drm, · line 500

Quoted 13 lines
drivers/gpu/drm/imx/imx-ldb.c +10 -0
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	}
 
 	return 0;
 }

regards
Philipp

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#29 lore raw

Re: [PATCH V3 1/5] drm/imx-ldb: Add support to drm-bridge

Hi Philipp,

Thank you for the review. I'm preparing V4, what about this:

---
 drivers/gpu/drm/imx/imx-ldb.c | 118 ++++++++++++++++++++++++++++--------------
 1 file changed, 78 insertions(+), 40 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +78 -40
index b03919e..4a33077 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -466,10 +470,30 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	} else {
+		/*
+		 * We want to add the connector whenever there is no bridge
+		 * that brings its own, not only when there is a panel. For
+		 * historical reasons, the ldb driver can also work without
+		 * a panel.
+		 */
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				encoder);
+	}
 
 	if (imx_ldb_ch->panel) {
 		ret = drm_panel_attach(imx_ldb_ch->panel,
@@ -478,8 +502,6 @@ static int imx_ldb_register(struct drm_device *drm,
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
-
 	return 0;
 }
 
@@ -548,6 +570,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -555,7 +616,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -605,7 +665,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 		int bus_format;
 
@@ -638,46 +697,25 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)
-					channel->mode_valid = 1;
-			}
+		/* panel ddc only if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		bus_format = of_get_bus_format(dev, child);
Signature · 1 line
2.5.5
#30 lore raw

PATCH v3 2/5 dts/imx6q-b850v3: Configure IPU assignment order

In reply to Peter Senna Tschudin

As the IPU has combined limitations across multiple crtcs, and as that can't be communicated to userspace at the moment, reorder the crtcs to allow support to two Full-HD monitors by avoiding assigning two monitors to a single IPU.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V2.

Changes from V1:
- New commit message

 arch/arm/boot/dts/imx6q-b850v3.dts | 5 +++++
 1 file changed, 5 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +5 -0
index 167f744..88a70de 100644
@@ -51,6 +51,11 @@
 	chosen {
 		stdout-path = &uart3;
 	};
+
+	display-subsystem {
+		compatible = "fsl,imx-display-subsystem";
+		ports = <&ipu1_di0>, <&ipu2_di0>, <&ipu1_di1>, <&ipu2_di1>;
+	};
 };
 
 &clks {
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#31 lore raw

Re: [PATCH V3 2/5] dts/imx6q-b850v3: Configure IPU assignment order

Am Sonntag, den 31.07.2016, 21:55 +0200 schrieb Peter Senna Tschudin:

Quoted 10 lines

As the IPU has combined limitations across multiple crtcs, and as that can't be communicated to userspace at the moment, reorder the crtcs to allow support to two Full-HD monitors by avoiding assigning two monitors to a single IPU.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

NACK. This is a userspace issue. Changing the assignment order of the CRTCs just shifts the failure to a userspace that want to use CRTC 0 and 2 now.

imx-drm just got atomic support and with the atomic check it should be possible to inform userspace in a reasonable way about such issues.

Regards,
Lucas

Quoted 15 lines

---
Unchanged from V2.

Changes from V1:
- New commit message

 arch/arm/boot/dts/imx6q-b850v3.dts | 5 +++++
 1 file changed, 5 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +0 -0
index 167f744..88a70de 100644
@@ -51,6 +51,11 @@
 	chosen {

arch/arm/boot/dts/imx6q-b850v3.dts line 61

Quoted 10 lines
arch/arm/boot/dts/imx6q-b850v3.dts +5 -0
 		stdout-path = &uart3;
 	};
+
+	display-subsystem {
+		compatible = "fsl,imx-display-subsystem";
+		ports = <&ipu1_di0>, <&ipu2_di0>, <&ipu1_di1>, <&ipu2_di1>;
+	};
 };
 
 &clks {

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#32 lore raw

Re: [PATCH V3 2/5] dts/imx6q-b850v3: Configure IPU assignment order

Hi Lucas,

Thank you for the prompt review.

Quoted 12 lines · Lucas Stach

Am Sonntag, den 31.07.2016, 21:55 +0200 schrieb Peter Senna Tschudin:

Quoted 10 lines

As the IPU has combined limitations across multiple crtcs, and as that can't be communicated to userspace at the moment, reorder the crtcs to allow support to two Full-HD monitors by avoiding assigning two monitors to a single IPU.

Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Quoted 3 lines

NACK. This is a userspace issue. Changing the assignment order of the CRTCs just shifts the failure to a userspace that want to use CRTC 0 and 2 now.

Err, yeah user space issue... But how the kernel is currently telling user space about what exactly went wrong and how user space might fix it? How Weston(our user space) is going to know that reshuffling crtcs is going to lead to success; how could it? I guess some platform-specific code in user space is needed for this to work...

Quoted 3 lines

imx-drm just got atomic support and with the atomic check it should be possible to inform userspace in a reasonable way about such issues.

Should be possible, but I guess it isn't, and wont be until a considerable effort is put on both kernel and user space. Or am I missing something? What do you propose?

I got inspiration from: arch/arm/boot/dts/imx6q.dtsi
...

        display-subsystem {
                compatible = "fsl,imx-display-subsystem";
                ports = <&ipu1_di0>, <&ipu1_di1>, <&ipu2_di0>, <&ipu2_di1>;
        };

...

This is there for more than 2 years now, and I get that the idea here is not ordering, but just declaring.

However even if this patch is not the perfect solution, it allows us to stay close to upstream now without creating problems(does it create any issue?).

Can you reconsider or propose a concrete solution that is not more complex than our entire driver?

Thanks a lot!

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#33 lore raw

Re: [PATCH V3 2/5] dts/imx6q-b850v3: Configure IPU assignment order

Quoted 22 lines · Peter Senna Tschudin

Hi Lucas,

Thank you for the prompt review.

Quoted 12 lines · Lucas Stach

Am Sonntag, den 31.07.2016, 21:55 +0200 schrieb Peter Senna Tschudin:

Quoted 10 lines

As the IPU has combined limitations across multiple crtcs, and as that can't be communicated to userspace at the moment, reorder the crtcs to allow support to two Full-HD monitors by avoiding assigning two monitors to a single IPU.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Quoted 3 lines

NACK. This is a userspace issue. Changing the assignment order of the CRTCs just shifts the failure to a userspace that want to use CRTC 0 and 2 now.

Quoted 5 lines

Err, yeah user space issue... But how the kernel is currently telling user space about what exactly went wrong and how user space might fix it? How Weston(our user space) is going to know that reshuffling crtcs is going to lead to success; how could it? I guess some platform-specific code in user space is needed for this to work...

atomic with the TEST_ONLY flag. This is what userspace should do:
1. submit atomic TEST_ONLY request with 1 screen on first crtc
2. If reject, move to another crtc or if those are all tried, reduce mode (this should never happen for the 1st screen, kernel /should/ filter out should impossible modes. But for 2nd/3rd screen combined modes might not all work).
3. Once you have a successfuly config for the 1st screen, add 2nd screen.
4. goto 2 with that 2nd screen.
5. Once all the screens have a mode/crtc they can be used on, do the real atomic request without TEST_ONLY.

Like Lucas said, no need to fumble around with ordering of CRTCs. The only thing we do in the driver is move the preferred output (if there is any) to the front of the _connector_ list, e.g. for built-in panels.

No need at all for platform specific code.

Cheers, Daniel

Quoted 27 lines
Quoted 4 lines

imx-drm just got atomic support and with the atomic check it should be possible to inform userspace in a reasonable way about such issues.

Should be possible, but I guess it isn't, and wont be until a considerable effort is put on both kernel and user space. Or am I missing something? What do you propose?

I got inspiration from: arch/arm/boot/dts/imx6q.dtsi
...

        display-subsystem {
                compatible = "fsl,imx-display-subsystem";
                ports = <&ipu1_di0>, <&ipu1_di1>, <&ipu2_di0>, <&ipu2_di1>;
        };

...

This is there for more than 2 years now, and I get that the idea here is not ordering, but just declaring.

However even if this patch is not the perfect solution, it allows us to stay close to upstream now without creating problems(does it create any issue?).

Can you reconsider or propose a concrete solution that is not more complex than our entire driver?

Thanks a lot!

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

Signature · 3 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
#34 lore raw

PATCH v3 3/5 Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V2.

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -0
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
maintainer Acked-by on v3 3/5
#35 lore raw

Re: [PATCH V3 3/5] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 12 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V2.

Quoted 7 lines

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
#36 lore raw

PATCH v3 4/5 drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic.

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 397 +++++++++++++++++++++++++++++
 4 files changed, 417 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index aaf36c0..ec52e17 100644
@@ -5139,6 +5139,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +397 -0

397 lines not on this page · raw message

Signature · 1 line
2.5.5
#37 lore raw

PATCH v3 5/5 dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Configures the GE B850v3 LVDS/DP++ bridge on the dts file.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V2.

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to b850v3-lvds-dp-bridge

 arch/arm/boot/dts/imx6q-b850v3.dts | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +30 -0
index 88a70de..10dfc3b 100644
@@ -77,6 +77,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -147,3 +154,26 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73 {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#38 lore raw

v4 cover Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

The series adds a driver that creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

The patches from the series:
[1/4] Change the imx-ldb driver to allow attaching a bridge and not only a LVDS
panel.

[2/4] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[3/4] Add the driver, make changes to MAINTAINERS, Kconfig and Makefile

[4/4] Make the changes to the B850v3 dts file to enable the GE B850v3
LVDS/DP++ Bridge.

Changes from V3:
- Removed the patch that was configuring the mapping between IPUs and external displays on the dts file

Peter Senna Tschudin (4):
drm/imx-ldb: Add support to drm-bridge Documentation/devicetree/bindings: b850v3_lvds_dp drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      |  37 ++
 MAINTAINERS                                        |   8 +
 arch/arm/boot/dts/imx6q-b850v3.dts                 |  30 ++
 drivers/gpu/drm/bridge/Kconfig                     |  11 +
 drivers/gpu/drm/bridge/Makefile                    |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c         | 397 +++++++++++++++++++++
 drivers/gpu/drm/imx/imx-ldb.c                      | 118 +++---
 7 files changed, 562 insertions(+), 40 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#39 lore raw

PATCH v4 1/4 drm/imx-ldb: Add support to drm-bridge

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>

---
Changes from V3:
- The connector is created when there is no bridge instead of only when there is a pannel
- Tested on next-20160804

Changes from V2:
- Updated to be aplied on top of Liu Ying changes that made imx-ldb atomic
- Tested on next-20160729

Changes from V1:
- Reanmed ext_bridge to bridge
- Removed empty entry point imx_ldb_encoder_enable()
- Adapted the code to apply to the latest linux next: next-20160609

 drivers/gpu/drm/imx/imx-ldb.c | 118 ++++++++++++++++++++++++++++--------------
 1 file changed, 78 insertions(+), 40 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +78 -40
index b03919e..4a33077 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -466,10 +470,30 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	} else {
+		/*
+		 * We want to add the connector whenever there is no bridge
+		 * that brings its own, not only when there is a panel. For
+		 * historical reasons, the ldb driver can also work without
+		 * a panel.
+		 */
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				encoder);
+	}
 
 	if (imx_ldb_ch->panel) {
 		ret = drm_panel_attach(imx_ldb_ch->panel,
@@ -478,8 +502,6 @@ static int imx_ldb_register(struct drm_device *drm,
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
-
 	return 0;
 }
 
@@ -548,6 +570,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -555,7 +616,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -605,7 +665,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 		int bus_format;
 
@@ -638,46 +697,25 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)
-					channel->mode_valid = 1;
-			}
+		/* panel ddc only if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		bus_format = of_get_bus_format(dev, child);
Signature · 1 line
2.5.5
Acked-by on v4 1/4
#40 lore raw

Re: [PATCH V4 1/4] drm/imx-ldb: Add support to drm-bridge

Quoted 8 lines · Peter Senna Tschudin

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Quoted 8 lines
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>
Quoted 189 lines

---
Changes from V3:
- The connector is created when there is no bridge instead of only when there is a pannel
- Tested on next-20160804

Changes from V2:
- Updated to be aplied on top of Liu Ying changes that made imx-ldb atomic
- Tested on next-20160729

Changes from V1:
- Reanmed ext_bridge to bridge
- Removed empty entry point imx_ldb_encoder_enable()
- Adapted the code to apply to the latest linux next: next-20160609

 drivers/gpu/drm/imx/imx-ldb.c | 118 ++++++++++++++++++++++++++++--------------
 1 file changed, 78 insertions(+), 40 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +73 -38
index b03919e..4a33077 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -466,10 +470,30 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	} else {
+		/*
+		 * We want to add the connector whenever there is no bridge
+		 * that brings its own, not only when there is a panel. For
+		 * historical reasons, the ldb driver can also work without
+		 * a panel.
+		 */
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				encoder);
+	}
 
 	if (imx_ldb_ch->panel) {
 		ret = drm_panel_attach(imx_ldb_ch->panel,
@@ -478,8 +502,6 @@ static int imx_ldb_register(struct drm_device *drm,
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
-
 	return 0;
 }
 
@@ -548,6 +570,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -555,7 +616,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -605,7 +665,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 		int bus_format;
 
@@ -638,46 +697,25 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)

drivers/gpu/drm/imx/imx-ldb.c static int imx_ldb_bind(struct device *dev, struct device *master, void *data) · line 721

Quoted 10 lines
drivers/gpu/drm/imx/imx-ldb.c +5 -2
-					channel->mode_valid = 1;
-			}
+		/* panel ddc only if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		bus_format = of_get_bus_format(dev, child);

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#41 lore raw

PATCH v4 2/4 Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V3:
- 2/4 instead of 3/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -0
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#42 lore raw

Re: [PATCH V4 2/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Peter,

2016-08-05 0:36 GMT+02:00 Peter Senna Tschudin <[email protected]>:

Quoted 9 lines

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

I think your previous version was acked by Rob and this version has no changes from previous so would be good add here the Acked-by: Rob Herring <[email protected]>

Quoted 60 lines

---
Changes from V3:
- 2/4 instead of 3/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -1
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+       status = "okay";
+       clock-frequency = <100000>;
+
+       b850v3-lvds-dp-bridge@73  {
+               compatible = "ge,b850v3-lvds-dp";
+               #address-cells = <1>;
+               #size-cells = <0>;
+
+               reg = <0x73>;
+               interrupt-parent = <&gpio2>;
+               interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+               edid-reg = <0x72>;
+
+               port {
+                       b850v3_dp_bridge_in: endpoint {
+                               remote-endpoint = <&lvds0_out>;
+                       };
+               };
+       };
+};
--

2.5.5

Acked-by on v4 2/4
#43 lore raw

Re: [PATCH V4 2/4] Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin
Quoted 9 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Quoted 58 lines

---
Changes from V3:
- 2/4 instead of 3/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -0
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
#44 lore raw

PATCH v4 3/4 drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 397 +++++++++++++++++++++++++++++
 4 files changed, 417 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index a5b6569..f51123c 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +397 -0

397 lines not on this page · raw message

Signature · 1 line
2.5.5
Reviewed-by Enric Balletbo i Serra on v4 3/4
#45 lore raw

Re: [PATCH V4 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Peter,

Only one comment below and one nitpick

2016-08-05 0:37 GMT+02:00 Peter Senna Tschudin <[email protected]>:

Quoted 437 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 397 +++++++++++++++++++++++++++++
 4 files changed, 417 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index a5b6569..f51123c 100644
@@ -5142,6 +5142,14 @@ W:       https://linuxtv.org
 S:     Maintained
 F:     drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:     Peter Senna Tschudin <[email protected]>
+M:     Martin Donnelly <[email protected]>
+M:     Martyn Welch <[email protected]>
+S:     Maintained
+F:     drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:     Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:     Haavard Skinnemoen <[email protected]>
 S:     Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
          Designware HDMI block.  This is used in conjunction with
          the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+       tristate "GE B850v3 LVDS to DP++ display bridge"
+       depends on OF
+       select DRM_KMS_HELPER
+       select DRM_PANEL
+       ---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
        tristate "NXP PTN3460 DP/LVDS bridge"
        depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +330 -0
new file mode 100644
index 0000000..eee8eac
@@ -0,0 +1,397 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+               (STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+               (STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+               (STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+       struct drm_connector connector;
+       struct drm_bridge bridge;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c;
+       struct i2c_client *edid_i2c;
+       struct edid *edid;
+       struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+               bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+       return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+               connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+       return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+       struct i2c_adapter *adapter = client->adapter;
+       unsigned char start = 0x00;
+       unsigned int total_size;
+       u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+       struct i2c_msg msgs[] = {
+               {
+                       .addr   = client->addr,
+                       .flags  = 0,
+                       .len    = 1,
+                       .buf    = &start,
+               }, {
+                       .addr   = client->addr,
+                       .flags  = I2C_M_RD,
+                       .len    = EDID_LENGTH,
+                       .buf    = block,
+               }
+       };
+
+       if (!block)
+               return NULL;
+
+       if (i2c_transfer(adapter, msgs, 2) != 2) {
+               DRM_ERROR("Unable to read EDID.\n");
+               goto err;
+       }
+
+       if (!drm_edid_block_valid(block, 0, false, NULL)) {
+               DRM_ERROR("Invalid EDID block\n");
+               goto err;
+       }
+
+       total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+       if (total_size > EDID_LENGTH) {
+               kfree(block);
+               block = kmalloc(total_size, GFP_KERNEL);
+               if (!block)
+                       return NULL;
+
+               /* Yes, read the entire buffer, and do not skip the first
+                * EDID_LENGTH bytes.
+                */
+               start = 0x00;
+               msgs[1].len = total_size;
+               msgs[1].buf = block;
+
+               if (i2c_transfer(adapter, msgs, 2) != 2) {
+                       DRM_ERROR("Unable to read EDID extension blocks.\n");
+                       goto err;
+               }
+       }
+
+       return block;
+
+err:
+       kfree(block);
+       return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       struct i2c_client *client;
+       int num_modes = 0;
+
+       ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+       client = ptn_bridge->edid_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       kfree(ptn_bridge->edid);
+       ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+       if (ptn_bridge->edid) {
+               drm_mode_connector_update_edid_property(connector,
+                               ptn_bridge->edid);
+               num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+       }
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+               struct drm_connector *connector, struct drm_display_mode *mode)
+{
+       if (mode->clock > MAX_PIXEL_CLOCK) {
+               DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+                               mode->name);
+               return MODE_CLOCK_HIGH;
+       }
+
+       return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+       .get_modes = ge_b850v3_lvds_dp_get_modes,
+       .mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+               struct drm_connector *connector, bool force)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+                       connector_to_ge_b850v3_lvds_dp(connector);
+       struct i2c_client *ge_b850v3_lvds_dp_i2c =
+                       ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       s32 link_state;
+
+       link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_STS_REG);
+
+       if (link_state == STDP4028_CON_STATE_CONNECTED)
+               return connector_status_connected;
+
+       if (link_state == 0)
+               return connector_status_disconnected;
+
+       return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+       .dpms = drm_atomic_helper_connector_dpms,
+       .fill_modes = drm_helper_probe_single_connector_modes,
+       .detect = ge_b850v3_lvds_dp_detect,
+       .destroy = drm_connector_cleanup,
+       .reset = drm_atomic_helper_connector_reset,
+       .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+       .atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+       mutex_lock(&ptn_bridge->lock);
+
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       mutex_unlock(&ptn_bridge->lock);
+
+       if (ptn_bridge->connector.dev)
+               drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+       return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge
+                       = bridge_to_ge_b850v3_lvds_dp(bridge);
+       struct drm_connector *connector = &ptn_bridge->connector;
+       struct i2c_client *ge_b850v3_lvds_dp_i2c
+                       = ptn_bridge->ge_b850v3_lvds_dp_i2c;
+       int ret;
+
+       if (!bridge->encoder) {
+               DRM_ERROR("Parent encoder object not found");
+               return -ENODEV;
+       }
+
+       connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+       drm_connector_helper_add(connector,
+                       &ge_b850v3_lvds_dp_connector_helper_funcs);
+
+       ret = drm_connector_init(bridge->dev, connector,
+                       &ge_b850v3_lvds_dp_connector_funcs,
+                       DRM_MODE_CONNECTOR_DisplayPort);
+       if (ret) {
+               DRM_ERROR("Failed to initialize connector with drm\n");
+               return ret;
+       }
+
+       drm_connector_register(connector);
+       ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+       if (ret)
+               return ret;
+
+       drm_bridge_enable(bridge);
+       if (ge_b850v3_lvds_dp_i2c->irq) {
+               drm_helper_hpd_irq_event(connector->dev);
+
+               ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+                               ge_b850v3_lvds_dp_i2c->irq, NULL,
+                               ge_b850v3_lvds_dp_irq_handler,
+                               IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+                               "ge-b850v3-lvds-dp", ptn_bridge);
+               if (ret)
+                       return ret;
+       }
+
+       return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+       .attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+                               const struct i2c_device_id *id)
+{
+       struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+       struct ge_b850v3_lvds_dp *ptn_bridge;
+       int ret;
+       u32 edid_i2c_reg;
+
+       ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+       if (!ptn_bridge)
+               return -ENOMEM;
+
+       mutex_init(&ptn_bridge->lock);
+
+       ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+       ptn_bridge->bridge.driver_private = ptn_bridge;
+       i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+       ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+       if (ret) {
+               dev_err(dev, "edid-reg not specified, aborting...\n");
+               return -ENODEV;
+       }
+
+       ptn_bridge->edid_i2c = devm_kzalloc(dev,
+                       sizeof(struct i2c_client), GFP_KERNEL);
+
+       if (!ptn_bridge->edid_i2c)

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 340

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+               return -ENOMEM;
+
+       memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+                       sizeof(struct i2c_client));
+
+       ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+       /* Configures the bridge to re-enable interrupts after each ack */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);

I'd check here and below the return value. Guess this is the first place where the comunication with the chip takes place, if something is wrong with the i2c communication, i.e the dt configures a wrong address, you should fail.

Quoted 23 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +23 -0
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+       /* Clear pending interrupts since power up. */
+       i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+                       STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+       ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+       ptn_bridge->bridge.of_node = dev->of_node;
+       ret = drm_bridge_add(&ptn_bridge->bridge);
+       if (ret) {
+               DRM_ERROR("Failed to add bridge\n");
+               return ret;
+       }
+
+       return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+       struct ge_b850v3_lvds_dp *ptn_bridge =
+               i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 373

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+       drm_bridge_remove(&ptn_bridge->bridge);
+
+       kfree(ptn_bridge->edid);
+
+       return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+       {"b850v3-lvds-dp", 0},
+       {},

nit: Use { /* sentinel */ }

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 379

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+       { .compatible = "ge,b850v3-lvds-dp" },
+       {},

nit: Use { /* sentinel */ }

Quoted 19 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +18 -1
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+       .id_table       = ge_b850v3_lvds_dp_i2c_table,
+       .probe          = ge_b850v3_lvds_dp_probe,
+       .remove         = ge_b850v3_lvds_dp_remove,
+       .driver         = {
+               .name           = "ge,b850v3-lvds-dp",
+               .of_match_table = ge_b850v3_lvds_dp_match,
+       },
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <[email protected]>");
+MODULE_AUTHOR("Martyn Welch <[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ display bridge)");
+MODULE_LICENSE("GPL v2");
--
Quoted 2 lines

2.5.5

After solving the i2c thing,

#46 lore raw

PATCH v4 4/4 dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Configures the GE B850v3 LVDS/DP++ bridge on the dts file.

Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Changes from V3:
- 4/4 instead of 5/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to b850v3-lvds-dp-bridge

 arch/arm/boot/dts/imx6q-b850v3.dts | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +30 -0
index 167f744..8db3bf2 100644
@@ -72,6 +72,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -142,3 +149,26 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73 {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#47 lore raw

v5 cover Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

The series adds a driver that creates a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

The patches from the series:
[1/4] Change the imx-ldb driver to allow attaching a bridge and not only a LVDS
panel.

[2/4] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[3/4] Add the driver, make changes to MAINTAINERS, Kconfig and Makefile

[4/4] Make the changes to the B850v3 dts file to enable the GE B850v3
LVDS/DP++ Bridge.

Changes from V3:
- Removed the patch that was configuring the mapping between IPUs and external displays on the dts file

Peter Senna Tschudin (4):
drm/imx-ldb: Add support to drm-bridge Documentation/devicetree/bindings: b850v3_lvds_dp drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      |  37 ++
 MAINTAINERS                                        |   8 +
 arch/arm/boot/dts/imx6q-b850v3.dts                 |  30 ++
 drivers/gpu/drm/bridge/Kconfig                     |  11 +
 drivers/gpu/drm/bridge/Makefile                    |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c         | 405 +++++++++++++++++++++
 drivers/gpu/drm/imx/imx-ldb.c                      | 118 ++++--
 7 files changed, 570 insertions(+), 40 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
Signature · 6 lines
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#48 lore raw

PATCH v5 1/4 drm/imx-ldb: Add support to drm-bridge

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>

---
Unchanged from V4

Changes from V3:
- The connector is created when there is no bridge instead of only when there is a pannel
- Tested on next-20160804

Changes from V2:
- Updated to be aplied on top of Liu Ying changes that made imx-ldb atomic
- Tested on next-20160729

Changes from V1:
- Reanmed ext_bridge to bridge
- Removed empty entry point imx_ldb_encoder_enable()
- Adapted the code to apply to the latest linux next: next-20160609

 drivers/gpu/drm/imx/imx-ldb.c | 118 ++++++++++++++++++++++++++++--------------
 1 file changed, 78 insertions(+), 40 deletions(-)
drivers/gpu/drm/imx/imx-ldb.c +78 -40
index b03919e..4a33077 100644
@@ -57,7 +57,11 @@ struct imx_ldb_channel {
 	struct imx_ldb *ldb;
 	struct drm_connector connector;
 	struct drm_encoder encoder;
+
+	/* Defines what is connected to the ldb, only one at a time */
 	struct drm_panel *panel;
+	struct drm_bridge *bridge;
+
 	struct device_node *child;
 	struct i2c_adapter *ddc;
 	int chno;
@@ -466,10 +470,30 @@ static int imx_ldb_register(struct drm_device *drm,
 	drm_encoder_init(drm, encoder, &imx_ldb_encoder_funcs,
 			 DRM_MODE_ENCODER_LVDS, NULL);
 
-	drm_connector_helper_add(&imx_ldb_ch->connector,
-			&imx_ldb_connector_helper_funcs);
-	drm_connector_init(drm, &imx_ldb_ch->connector,
-			   &imx_ldb_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
+	if (imx_ldb_ch->bridge) {
+		imx_ldb_ch->bridge->encoder = encoder;
+
+		imx_ldb_ch->encoder.bridge = imx_ldb_ch->bridge;
+		ret = drm_bridge_attach(drm, imx_ldb_ch->bridge);
+		if (ret) {
+			DRM_ERROR("Failed to initialize bridge with drm\n");
+			return ret;
+		}
+	} else {
+		/*
+		 * We want to add the connector whenever there is no bridge
+		 * that brings its own, not only when there is a panel. For
+		 * historical reasons, the ldb driver can also work without
+		 * a panel.
+		 */
+		drm_connector_helper_add(&imx_ldb_ch->connector,
+				&imx_ldb_connector_helper_funcs);
+		drm_connector_init(drm, &imx_ldb_ch->connector,
+				&imx_ldb_connector_funcs,
+				DRM_MODE_CONNECTOR_LVDS);
+		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
+				encoder);
+	}
 
 	if (imx_ldb_ch->panel) {
 		ret = drm_panel_attach(imx_ldb_ch->panel,
@@ -478,8 +502,6 @@ static int imx_ldb_register(struct drm_device *drm,
 			return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
-
 	return 0;
 }
 
@@ -548,6 +570,45 @@ static const struct of_device_id imx_ldb_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, imx_ldb_dt_ids);
 
+static int imx_ldb_panel_ddc(struct device *dev,
+		struct imx_ldb_channel *channel, struct device_node *child)
+{
+	struct device_node *ddc_node;
+	const u8 *edidp;
+	int ret;
+
+	ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
+	if (ddc_node) {
+		channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
+		of_node_put(ddc_node);
+		if (!channel->ddc) {
+			dev_warn(dev, "failed to get ddc i2c adapter\n");
+			return -EPROBE_DEFER;
+		}
+	}
+
+	if (!channel->ddc) {
+		/* if no DDC available, fallback to hardcoded EDID */
+		dev_dbg(dev, "no ddc available\n");
+
+		edidp = of_get_property(child, "edid",
+					&channel->edid_len);
+		if (edidp) {
+			channel->edid = kmemdup(edidp,
+						channel->edid_len,
+						GFP_KERNEL);
+		} else if (!channel->panel) {
+			/* fallback to display-timings node */
+			ret = of_get_drm_display_mode(child,
+						      &channel->mode,
+						      OF_USE_NATIVE_MODE);
+			if (!ret)
+				channel->mode_valid = 1;
+		}
+	}
+	return 0;
+}
+
 static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 {
 	struct drm_device *drm = data;
@@ -555,7 +616,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 	const struct of_device_id *of_id =
 			of_match_device(imx_ldb_dt_ids, dev);
 	struct device_node *child;
-	const u8 *edidp;
 	struct imx_ldb *imx_ldb;
 	int dual;
 	int ret;
@@ -605,7 +665,6 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 	for_each_child_of_node(np, child) {
 		struct imx_ldb_channel *channel;
-		struct device_node *ddc_node;
 		struct device_node *ep;
 		int bus_format;
 
@@ -638,46 +697,25 @@ static int imx_ldb_bind(struct device *dev, struct device *master, void *data)
 
 			remote = of_graph_get_remote_port_parent(ep);
 			of_node_put(ep);
-			if (remote)
+			if (remote) {
 				channel->panel = of_drm_find_panel(remote);
-			else
+				channel->bridge = of_drm_find_bridge(remote);
+			} else
 				return -EPROBE_DEFER;
 			of_node_put(remote);
-			if (!channel->panel) {
-				dev_err(dev, "panel not found: %s\n",
-					remote->full_name);
-				return -EPROBE_DEFER;
-			}
-		}
 
-		ddc_node = of_parse_phandle(child, "ddc-i2c-bus", 0);
-		if (ddc_node) {
-			channel->ddc = of_find_i2c_adapter_by_node(ddc_node);
-			of_node_put(ddc_node);
-			if (!channel->ddc) {
-				dev_warn(dev, "failed to get ddc i2c adapter\n");
+			if (!channel->panel && !channel->bridge) {
+				dev_err(dev, "panel/bridge not found: %s\n",
+					remote->full_name);
 				return -EPROBE_DEFER;
 			}
 		}
 
-		if (!channel->ddc) {
-			/* if no DDC available, fallback to hardcoded EDID */
-			dev_dbg(dev, "no ddc available\n");
-
-			edidp = of_get_property(child, "edid",
-						&channel->edid_len);
-			if (edidp) {
-				channel->edid = kmemdup(edidp,
-							channel->edid_len,
-							GFP_KERNEL);
-			} else if (!channel->panel) {
-				/* fallback to display-timings node */
-				ret = of_get_drm_display_mode(child,
-							      &channel->mode,
-							      OF_USE_NATIVE_MODE);
-				if (!ret)
-					channel->mode_valid = 1;
-			}
+		/* panel ddc only if there is no bridge */
+		if (!channel->bridge) {
+			ret = imx_ldb_panel_ddc(dev, channel, child);
+			if (ret)
+				return ret;
 		}
 
 		bus_format = of_get_bus_format(dev, child);
Signature · 1 line
2.5.5
#49 lore raw

Re: [PATCH V5 1/4] drm/imx-ldb: Add support to drm-bridge

Am Dienstag, den 09.08.2016, 18:41 +0200 schrieb Peter Senna Tschudin:

Quoted 18 lines

Add support to attach a drm_bridge to imx-ldb in addition to existing support to attach a LVDS panel.

This patch does a simple code refactoring by moving code from for_each_child_of_node iterator to a new function named imx_ldb_panel_ddc(). This was necessary to allow the panel ddc code to run only when the imx_ldb is not attached to a bridge.

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc David Airlie <[email protected]>
Cc Thierry Reding <[email protected]>
Cc Thierry Reding <[email protected]>
Quoted 2 lines

---
Unchanged from V4

I've rebased and applied patch 1.

thanks
Philipp

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#50 lore raw

PATCH v5 2/4 Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V4

Changes from V3:
- 2/4 instead of 3/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -0
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#51 lore raw

Re: [PATCH V5 2/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Patch 1/4 is already on linux-next, but what about this one? Ping?

Quoted 75 lines · Peter Senna Tschudin

Devicetree bindings documentation for the GE B850v3 LVDS/DP++ display bridge.

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V4

Changes from V3:
- 2/4 instead of 3/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to the example

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 37 ++++++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +37 -0
new file mode 100644
index 0000000..f05c3e9
@@ -0,0 +1,37 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the address used to ack the interrupts.
+  - interrupt-parent : phandle of the interrupt controller that services
+    interrupts to the device
+  - interrupts : one interrupt should be described here, as in
+    <0 IRQ_TYPE_LEVEL_HIGH>.
+  - edid-reg : should contain the address used to read edid information
+  - port : should describe the video signal connection between the host
+    and the bridge.
+
+Example:
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73  {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#52 lore raw

PATCH v5 3/4 drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +405 -0

405 lines not on this page · raw message

Signature · 1 line
2.5.5
#53 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi,

Quoted 13 lines · Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Quoted 2 lines
   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

I'd commented on an earlier revision (v2) of this patch, but hadn't got a response on it. Pasting the query again:

Are these two chips always expected to be used together? I don't think it's right to pair up two encoder chips into one driver just for one board.

Is one device @0x72 and other @0x73? Or is only one of them an i2c slave?

What's preventing us to create these as two different bridge drivers? The drm framework allows us to daisy chain encoder bridges. The only problem I see is that we don't have a clear-cut way to tell the bridge driver whether we want it to create a connector for us or not. Because, it looks like both can potentially create connectors. This isn't a big problem either if we have DT. We just need to check whether our output port is connected to another bridge or a connector.

Thanks,
Archit

Quoted 507 lines
Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <daniel.vetter-/[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

  MAINTAINERS                                |   8 +
  drivers/gpu/drm/bridge/Kconfig             |  11 +
  drivers/gpu/drm/bridge/Makefile            |   1 +
  drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
  4 files changed, 425 insertions(+)

create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c

MAINTAINERS +8 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
  S:	Maintained
  F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
  GENERIC GPIO I2C DRIVER
  M:	Haavard Skinnemoen <[email protected]>
  S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
  	  Designware HDMI block.  This is used in conjunction with
  	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
  config DRM_NXP_PTN3460
  	tristate "NXP PTN3460 DP/LVDS bridge"
  	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
  obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
  obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
  obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
  obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
  obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
  obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +405 -0
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_register(connector);
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+				const struct i2c_device_id *id)
+{
+	struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	int ret;
+	u32 edid_i2c_reg;
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->lock);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+	if (ret) {
+		dev_err(dev, "edid-reg not specified, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->edid_i2c = devm_kzalloc(dev,
+			sizeof(struct i2c_client), GFP_KERNEL);
+
+	if (!ptn_bridge->edid_i2c)
+		return -ENOMEM;
+
+	memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+			sizeof(struct i2c_client));
+
+	ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+	/*
+	 * Configures the bridge to re-enable interrupts after each ack. As
+	 * this is the first communication with the chip, fail on error.
+	 */
+	ret = i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+	if (ret) {
+		dev_err(dev, "i2c communication failed, aborting...\n");
+		return ret;
+	}
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	/* Clear pending interrupts since power up. */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	ret = drm_bridge_add(&ptn_bridge->bridge);
+	if (ret) {
+		DRM_ERROR("Failed to add bridge\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+	drm_bridge_remove(&ptn_bridge->bridge);
+
+	kfree(ptn_bridge->edid);
+
+	return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+	{"b850v3-lvds-dp", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+	{ .compatible = "ge,b850v3-lvds-dp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+	.id_table	= ge_b850v3_lvds_dp_i2c_table,
+	.probe		= ge_b850v3_lvds_dp_probe,
+	.remove		= ge_b850v3_lvds_dp_remove,
+	.driver		= {
+		.name		= "b850v3-lvds-dp",
+		.of_match_table = ge_b850v3_lvds_dp_match,
+	},
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>");
+MODULE_AUTHOR("Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ display bridge)");
+MODULE_LICENSE("GPL v2");
Signature · 6 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#54 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Patch 1/4 is already on linux-next, but what about this one? Ping?

Quoted 529 lines · Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device

tree nodes, as the design goal is to configure both bridges based on

the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and

the driver is only needed for telling the host about EDID / HPD, and

for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <daniel.vetter-/[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +405 -0
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_register(connector);
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+				const struct i2c_device_id *id)
+{
+	struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	int ret;
+	u32 edid_i2c_reg;
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->lock);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+	if (ret) {
+		dev_err(dev, "edid-reg not specified, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->edid_i2c = devm_kzalloc(dev,
+			sizeof(struct i2c_client), GFP_KERNEL);
+
+	if (!ptn_bridge->edid_i2c)
+		return -ENOMEM;
+
+	memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+			sizeof(struct i2c_client));
+
+	ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+	/*
+	 * Configures the bridge to re-enable interrupts after each ack. As
+	 * this is the first communication with the chip, fail on error.
+	 */
+	ret = i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+	if (ret) {
+		dev_err(dev, "i2c communication failed, aborting...\n");
+		return ret;
+	}
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	/* Clear pending interrupts since power up. */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	ret = drm_bridge_add(&ptn_bridge->bridge);
+	if (ret) {
+		DRM_ERROR("Failed to add bridge\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+	drm_bridge_remove(&ptn_bridge->bridge);
+
+	kfree(ptn_bridge->edid);
+
+	return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+	{"b850v3-lvds-dp", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+	{ .compatible = "ge,b850v3-lvds-dp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+	.id_table	= ge_b850v3_lvds_dp_i2c_table,
+	.probe		= ge_b850v3_lvds_dp_probe,
+	.remove		= ge_b850v3_lvds_dp_remove,
+	.driver		= {
+		.name		= "b850v3-lvds-dp",
+		.of_match_table = ge_b850v3_lvds_dp_match,
+	},
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>");
+MODULE_AUTHOR("Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ display bridge)");
+MODULE_LICENSE("GPL v2");
Signature · 1 line
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#55 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Peter,

Quoted 1 line · Peter Senna Tschudin

Patch 1/4 is already on linux-next, but what about this one? Ping?

I'd posted some queries a couple of times which you didn't answer to. Could you please respond to them before we try to get this merged?

Archit

Quoted 533 lines · Peter Senna Tschudin
Quoted 525 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <daniel.vetter-/[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +8 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +405 -1
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_register(connector);
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.attach = ge_b850v3_lvds_dp_attach,
+};
+
+static int ge_b850v3_lvds_dp_probe(struct i2c_client *ge_b850v3_lvds_dp_i2c,
+				const struct i2c_device_id *id)
+{
+	struct device *dev = &ge_b850v3_lvds_dp_i2c->dev;
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	int ret;
+	u32 edid_i2c_reg;
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->lock);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ret = of_property_read_u32(dev->of_node, "edid-reg", &edid_i2c_reg);
+	if (ret) {
+		dev_err(dev, "edid-reg not specified, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->edid_i2c = devm_kzalloc(dev,
+			sizeof(struct i2c_client), GFP_KERNEL);
+
+	if (!ptn_bridge->edid_i2c)
+		return -ENOMEM;
+
+	memcpy(ptn_bridge->edid_i2c, ge_b850v3_lvds_dp_i2c,
+			sizeof(struct i2c_client));
+
+	ptn_bridge->edid_i2c->addr = (unsigned short) edid_i2c_reg;
+
+	/*
+	 * Configures the bridge to re-enable interrupts after each ack. As
+	 * this is the first communication with the chip, fail on error.
+	 */
+	ret = i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_IRQ_OUT_CONF_REG, STDP4028_DPTX_DP_IRQ_EN);
+	if (ret) {
+		dev_err(dev, "i2c communication failed, aborting...\n");
+		return ret;
+	}
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	/* Clear pending interrupts since power up. */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	ret = drm_bridge_add(&ptn_bridge->bridge);
+	if (ret) {
+		DRM_ERROR("Failed to add bridge\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ge_b850v3_lvds_dp_remove(struct i2c_client *ge_b850v3_lvds_dp_i2c)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+		i2c_get_clientdata(ge_b850v3_lvds_dp_i2c);
+
+	drm_bridge_remove(&ptn_bridge->bridge);
+
+	kfree(ptn_bridge->edid);
+
+	return 0;
+}
+
+static const struct i2c_device_id ge_b850v3_lvds_dp_i2c_table[] = {
+	{"b850v3-lvds-dp", 0},
+	{},
+};
+MODULE_DEVICE_TABLE(i2c, ge_b850v3_lvds_dp_i2c_table);
+
+static const struct of_device_id ge_b850v3_lvds_dp_match[] = {
+	{ .compatible = "ge,b850v3-lvds-dp" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ge_b850v3_lvds_dp_match);
+
+static struct i2c_driver ge_b850v3_lvds_dp_driver = {
+	.id_table	= ge_b850v3_lvds_dp_i2c_table,
+	.probe		= ge_b850v3_lvds_dp_probe,
+	.remove		= ge_b850v3_lvds_dp_remove,
+	.driver		= {
+		.name		= "b850v3-lvds-dp",
+		.of_match_table = ge_b850v3_lvds_dp_match,
+	},
+};
+module_i2c_driver(ge_b850v3_lvds_dp_driver);
+
+MODULE_AUTHOR("Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>");
+MODULE_AUTHOR("Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>");
+MODULE_DESCRIPTION("GE LVDS to DP++ display bridge)");
+MODULE_LICENSE("GPL v2");
--

2.5.5

Signature · 6 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#56 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Archit,

Quoted 8 lines · Archit Taneja

Hi Peter,

Quoted 1 line · Peter Senna Tschudin

Patch 1/4 is already on linux-next, but what about this one? Ping?

I'd posted some queries a couple of times which you didn't answer to. Could you please respond to them before we try to get this merged?

Your queries were already answered by similar questions. The commit messages and cover letter also addresses the design decisions of the code. But basically the driver usefulness to other scenarios is severely limited by the firmware used by both chips. And when using the firmware that goes with this specific hardware, then yes, the two chips are always expected to work together. But the main point is that with the custom firmware each chip do not behave as independent bridges anymore.

On the other side, I was careful to use meaningful names for the registers, so a future implementation based on same chips can take the basics from this work, at least as a starting point.

Thanks,

Peter

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#57 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi,

Quoted 6 lines · Peter Senna Tschudin

Hi Archit,

Quoted 2 lines · Archit Taneja

Hi Peter,

Quoted 8 lines
Quoted 6 lines · Peter Senna Tschudin
Quoted 1 line

Patch 1/4 is already on linux-next, but what about this one? Ping?

I'd posted some queries a couple of times which you didn't answer to. Could you please respond to them before we try to get this merged?

Your queries were already answered by similar questions. The commit messages and cover letter also addresses the design decisions of the code. But basically the driver usefulness to other scenarios is severely limited by the firmware used by both chips. And when using the firmware that goes with this specific hardware, then yes, the two chips are always expected to work together. But the main point is that with the custom firmware each chip do not behave as independent bridges anymore.

Thanks for the reply.

It wasn't entirely clear from the commit message that a custom firmware was exclusively used on this board to program these chips in order to get these 2 working together.

I browsed the earlier versions of the patch and saw you explained the same thing to someone else. Sorry about that, I missed reading that before.

Could you please specify this explicitly in the commit message? Perhaps, also mention that there is an external microcontroller with a custom firmware that manages most of the video operations. For the sake of completeness, could you also mention the part name of the controller that's running this firmware?

Also, in the comments in the beginning of the driver:

"However the physical bridges are automatically configured by the input video signal, and the driver has no access to the video processing pipeline."

Is the automatic configuration done by the firmware, or is it a feature of the chips itself?

Quoted 2 lines

On the other side, I was careful to use meaningful names for the registers, so a future implementation based on same chips can take the basics from this work, at least as a starting point.

Thanks, that would be handy for later.

I had some comments on the code. I'll share those in another reply.

Archit

Quoted 5 lines

Thanks,

Peter

Signature · 6 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#58 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Hi,

Some comments.

Quoted 66 lines · Peter Senna Tschudin

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +1 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE

MAINTAINERS W: https://linuxtv.org · line 5155

Quoted 10 lines
MAINTAINERS +7 -0
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported

Could you move the MAINTAINERS change to a different patch? It would make it easier to integrate separately.

Quoted 308 lines
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +268 -0
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 278

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_register(connector);

Connectors shouldn't be registered in the bridge driver, they should be registered by the kms driver after drm_dev_register is called. The drm_connector_register_all() API is used for that.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 283

Quoted 5 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +5 -0
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);

This drm_bridge_enable() doesn't seem to serve any purpose here. It also doesn't seem to make much sense to call drm_bridge_() funcs within a bridge op itself.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 291

Quoted 8 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +8 -0
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);

Is there a reason why we register the interrupt handler here and not in probe?

Thanks,
Archit

Signature · 2 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
#59 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 3 lines · Archit Taneja

Hi,

Some comments.

Thank you for the review!

Quoted 76 lines · Peter Senna Tschudin
Quoted 73 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and

firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +7 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
Quoted 7 lines

MAINTAINERS W: https://linuxtv.org · line 5155

Quoted 4 lines
MAINTAINERS +1 -0
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported

Could you move the MAINTAINERS change to a different patch? It would make it easier to integrate separately.

If needed, yes sure.

Quoted 323 lines
Quoted 310 lines
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +268 -0
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
 
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 278

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_register(connector);

Connectors shouldn't be registered in the bridge driver, they should

Quoted 2 lines

be registered by the kms driver after drm_dev_register is called. The drm_connector_register_all() API is used for that.

Hmm, I got this from:

drivers/gpu/drm/bridge/nxp-ptn3460.c: drm_connector_register(&ptn_bridge->connector);
drivers/gpu/drm/bridge/parade-ps8622.c: drm_connector_register(&ps8622->connector);
drivers/gpu/drm/bridge/analogix-anx78xx.c: err = drm_connector_register(&anx78xx->connector);

Quoted 10 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 283

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +5 -0
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);

This drm_bridge_enable() doesn't seem to serve any purpose here. It also doesn't seem to make much sense to call drm_bridge_() funcs within a bridge op itself.

Yes, removing this line has no effect. I'll send V6. Thank you!

Quoted 12 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 291

Quoted 9 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +8 -0
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);

Is there a reason why we register the interrupt handler here and not in probe?

Yes, drm_bridge_funcs.attach() is called when drm decides it is time to initialize the bridge, which means the drm is already running. i2c_driver.probe() is likely to be called way before drm initializes, and if for some reason the bridge is not attached, there is no need for the interrupt handler. In this case the interrupt handler is mostly used for events related to display plugging/unplugging.

#60 lore raw

Re: [PATCH V5 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 417 lines · Peter Senna Tschudin
Quoted 3 lines · Archit Taneja

Hi,

Some comments.

Thank you for the review!

Quoted 67 lines · Peter Senna Tschudin
Quoted 65 lines

Add a driver that create a drm_bridge and a drm_connector for the LVDS to DP++ display bridge of the GE B850v3.

There are two physical bridges on the video signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). The hardware and firmware made it complicated for this binding to comprise two device tree nodes, as the design goal is to configure both bridges based on the LVDS signal, which leave the driver powerless to control the video processing pipeline. The two bridges behaves as a single bridge, and the driver is only needed for telling the host about EDID / HPD, and for giving the host powers to ack interrupts. The video signal pipeline is as follows:

Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Daniel Vetter <daniel.vetter-/[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>
CC David Airlie <[email protected]>
CC Thierry Reding <[email protected]>
CC Thierry Reding <[email protected]>

---
Changes from V4:
- Check the output of the first call to i2c_smbus_write_word_data() and return it's error code for failing gracefully on i2c issues
- Renamed the i2c_driver.name from "ge,b850v3-lvds-dp" to "b850v3-lvds-dp" to remove the comma from the driver name

Changes from V3:
- 3/4 instead of 4/5
- Tested on next-20160804

Changes from V2:
- Made it atomic to be applied on next-20160729 on top of Liu Ying changes that made imx-ldb atomic

Changes from V1:
- New commit message
- Removed 3 empty entry points
- Removed memory leak from ge_b850v3_lvds_dp_get_modes()
- Added a lock for mode setting
- Removed a few blank lines
- Changed the order at Makefile and Kconfig

 MAINTAINERS                                |   8 +
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 405 +++++++++++++++++++++++++++++
 4 files changed, 425 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
MAINTAINERS +0 -0
index a306795..e8d106a 100644
@@ -5142,6 +5142,14 @@ W:	https://linuxtv.org
 S:	Maintained
 F:	drivers/media/radio/radio-gemtek*
 
Quoted 14 lines

MAINTAINERS W: https://linuxtv.org · line 5155

Quoted 11 lines
MAINTAINERS +8 -0
+GENERAL ELECTRIC B850V3 LVDS/DP++ BRIDGE
+M:	Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]>
+M:	Martin Donnelly <[email protected]>
+M:	Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
+S:	Maintained
+F:	drivers/gpu/drm/bridge/ge_b850v3_dp2.c
+F:	Documentation/devicetree/bindings/ge/b850v3_dp2_bridge.txt
+
 GENERIC GPIO I2C DRIVER
 M:	Haavard Skinnemoen <[email protected]>
 S:	Supported

Could you move the MAINTAINERS change to a different patch? It would make it easier to integrate separately.

If needed, yes sure.

Quoted 315 lines
Quoted 315 lines
drivers/gpu/drm/bridge/Kconfig +11 -0
index b590e67..b4b70fb 100644
@@ -32,6 +32,17 @@ config DRM_DW_HDMI_AHB_AUDIO
 	  Designware HDMI block.  This is used in conjunction with
 	  the i.MX6 HDMI driver.
 
+config DRM_GE_B850V3_LVDS_DP
+	tristate "GE B850v3 LVDS to DP++ display bridge"
+	depends on OF
+	select DRM_KMS_HELPER
+	select DRM_PANEL
+	---help---
+          This is a driver for the display bridge of
+          GE B850v3 that convert dual channel LVDS
+          to DP++. This is used with the i.MX6 imx-ldb
+          driver.
+
 config DRM_NXP_PTN3460
 	tristate "NXP PTN3460 DP/LVDS bridge"
 	depends on OF
drivers/gpu/drm/bridge/Makefile +1 -0
index efdb07e..b9606f3 100644
@@ -3,6 +3,7 @@ ccflags-y := -Iinclude/drm
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
+obj-$(CONFIG_DRM_GE_B850V3_LVDS_DP) += ge_b850v3_lvds_dp.o
 obj-$(CONFIG_DRM_NXP_PTN3460) += nxp-ptn3460.o
 obj-$(CONFIG_DRM_PARADE_PS8622) += parade-ps8622.o
 obj-$(CONFIG_DRM_SII902X) += sii902x.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +274 -0
new file mode 100644
index 0000000..81e9279
@@ -0,0 +1,405 @@
+/*
+ * Driver for GE B850v3 DP display bridge
+
+ * Copyright (c) 2016, Collabora Ltd.
+ * Copyright (c) 2016, General Electric Company
+
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+ * This driver creates a drm_bridge and a drm_connector for the LVDS to DP++
+ * display bridge of the GE B850v3. There are two physical bridges on the video
+ * signal pipeline: a STDP4028(LVDS to DP) and a STDP2690(DP to DP++). However
+ * the physical bridges are automatically configured by the input video signal,
+ * and the driver has no access to the video processing pipeline. The driver is
+ * only needed to read EDID from the STDP2690 and to handle HPD events from the
+ * STDP4028. The driver communicates with both bridges over i2c. The video
+ * signal pipeline is as follows:
+ *
+ *   Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output
+ *
+ */
+
+#include <linux/gpio.h>
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drmP.h>
+
+/*
+ * 220Mhz is a limitation of the host, as the bridge is capable of up to
+ * 330Mhz. See section 9.2.1.2.4 of the i.MX 6Dual/6Quad Applications
+ * Processor Reference Manual for more information about the 220Mhz limit.
+ * The imx-ldb driver will warn about clocks over 170Mhz, but it seem to work
+ * fine.
+ */
+#define MAX_PIXEL_CLOCK 220000
+
+#define EDID_EXT_BLOCK_CNT 0x7E
+
+#define STDP4028_IRQ_OUT_CONF_REG 0x02
+#define STDP4028_DPTX_IRQ_EN_REG 0x3C
+#define STDP4028_DPTX_IRQ_STS_REG 0x3D
+#define STDP4028_DPTX_STS_REG 0x3E
+
+#define STDP4028_DPTX_DP_IRQ_EN 0x1000
+
+#define STDP4028_DPTX_HOTPLUG_IRQ_EN 0x0400
+#define STDP4028_DPTX_LINK_CH_IRQ_EN 0x2000
+#define STDP4028_DPTX_IRQ_CONFIG \
+		(STDP4028_DPTX_LINK_CH_IRQ_EN | STDP4028_DPTX_HOTPLUG_IRQ_EN)
+
+#define STDP4028_DPTX_HOTPLUG_STS 0x0200
+#define STDP4028_DPTX_LINK_STS 0x1000
+#define STDP4028_CON_STATE_CONNECTED \
+		(STDP4028_DPTX_HOTPLUG_STS | STDP4028_DPTX_LINK_STS)
+
+#define STDP4028_DPTX_HOTPLUG_CH_STS 0x0400
+#define STDP4028_DPTX_LINK_CH_STS 0x2000
+#define STDP4028_DPTX_IRQ_CLEAR \
+		(STDP4028_DPTX_LINK_CH_STS | STDP4028_DPTX_HOTPLUG_CH_STS)
+
+struct ge_b850v3_lvds_dp {
+	struct drm_connector connector;
+	struct drm_bridge bridge;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c;
+	struct i2c_client *edid_i2c;
+	struct edid *edid;
+	struct mutex lock;
+};
+
+static inline struct ge_b850v3_lvds_dp *
+		bridge_to_ge_b850v3_lvds_dp(struct drm_bridge *bridge)
+{
+	return container_of(bridge, struct ge_b850v3_lvds_dp, bridge);
+}
+
+static inline struct ge_b850v3_lvds_dp *
+		connector_to_ge_b850v3_lvds_dp(struct drm_connector *connector)
+{
+	return container_of(connector, struct ge_b850v3_lvds_dp, connector);
+}
+
+u8 *stdp2690_get_edid(struct i2c_client *client)
+{
+	struct i2c_adapter *adapter = client->adapter;
+	unsigned char start = 0x00;
+	unsigned int total_size;
+	u8 *block = kmalloc(EDID_LENGTH, GFP_KERNEL);
+
+	struct i2c_msg msgs[] = {
+		{
+			.addr	= client->addr,
+			.flags	= 0,
+			.len	= 1,
+			.buf	= &start,
+		}, {
+			.addr	= client->addr,
+			.flags	= I2C_M_RD,
+			.len	= EDID_LENGTH,
+			.buf	= block,
+		}
+	};
+
+	if (!block)
+		return NULL;
+
+	if (i2c_transfer(adapter, msgs, 2) != 2) {
+		DRM_ERROR("Unable to read EDID.\n");
+		goto err;
+	}
+
+	if (!drm_edid_block_valid(block, 0, false, NULL)) {
+		DRM_ERROR("Invalid EDID block\n");
+		goto err;
+	}
+
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */
+		start = 0x00;
+		msgs[1].len = total_size;
+		msgs[1].buf = block;
+
+		if (i2c_transfer(adapter, msgs, 2) != 2) {
+			DRM_ERROR("Unable to read EDID extension blocks.\n");
+			goto err;
+		}
+	}
+
+	return block;
+
+err:
+	kfree(block);
+	return NULL;
+}
+
+static int ge_b850v3_lvds_dp_get_modes(struct drm_connector *connector)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge;
+	struct i2c_client *client;
+	int num_modes = 0;
+
+	ptn_bridge = connector_to_ge_b850v3_lvds_dp(connector);
+	client = ptn_bridge->edid_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	kfree(ptn_bridge->edid);
+	ptn_bridge->edid = (struct edid *) stdp2690_get_edid(client);
+
+	if (ptn_bridge->edid) {
+		drm_mode_connector_update_edid_property(connector,
+				ptn_bridge->edid);
+		num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+	}
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	if (mode->clock > MAX_PIXEL_CLOCK) {
+		DRM_INFO("The pixel clock for the mode %s is too high, and not supported.",
+				mode->name);
+		return MODE_CLOCK_HIGH;
+	}
+
+	return MODE_OK;
+}
+
+static const struct
+drm_connector_helper_funcs ge_b850v3_lvds_dp_connector_helper_funcs = {
+	.get_modes = ge_b850v3_lvds_dp_get_modes,
+	.mode_valid = ge_b850v3_lvds_dp_mode_valid,
+};
+
+static enum drm_connector_status ge_b850v3_lvds_dp_detect(
+		struct drm_connector *connector, bool force)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge =
+			connector_to_ge_b850v3_lvds_dp(connector);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c =
+			ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	s32 link_state;
+
+	link_state = i2c_smbus_read_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_STS_REG);
+
+	if (link_state == STDP4028_CON_STATE_CONNECTED)
+		return connector_status_connected;
+
+	if (link_state == 0)
+		return connector_status_disconnected;
+
+	return connector_status_unknown;
+}
+
+static const struct drm_connector_funcs ge_b850v3_lvds_dp_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ge_b850v3_lvds_dp_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static irqreturn_t ge_b850v3_lvds_dp_irq_handler(int irq, void *dev_id)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge = dev_id;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	mutex_lock(&ptn_bridge->lock);
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->lock);
+
+	if (ptn_bridge->connector.dev)
+		drm_kms_helper_hotplug_event(ptn_bridge->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static int ge_b850v3_lvds_dp_attach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct drm_connector *connector = &ptn_bridge->connector;
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+	int ret;
+
+	if (!bridge->encoder) {
+		DRM_ERROR("Parent encoder object not found");
+		return -ENODEV;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	drm_connector_helper_add(connector,
+			&ge_b850v3_lvds_dp_connector_helper_funcs);
+
+	ret = drm_connector_init(bridge->dev, connector,
+			&ge_b850v3_lvds_dp_connector_funcs,
+			DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 278

Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+		return ret;
+	}
+
+	drm_connector_register(connector);

Connectors shouldn't be registered in the bridge driver, they should be registered by the kms driver after drm_dev_register is called. The drm_connector_register_all() API is used for that.

Quoted 5 lines

Hmm, I got this from:

drivers/gpu/drm/bridge/nxp-ptn3460.c: drm_connector_register(&ptn_bridge->connector);
drivers/gpu/drm/bridge/parade-ps8622.c: drm_connector_register(&ps8622->connector);
drivers/gpu/drm/bridge/analogix-anx78xx.c: err = drm_connector_register(&anx78xx->connector);

Yeah, we need to get rid of these too, eventually.

The connectors should be registered only after the drm device has been registered. Relying on the drm core to do that ensures that. See drm_modeset_register_all() for more info.

Also, btw, the imx driver shouldn't use the load() drm_driver ops anymore. You would want to change that too. See the comments in drm_dev_register() in drm_drv.c for more details.

Quoted 28 lines
Quoted 12 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 283

Quoted 8 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +5 -0
+	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	if (ret)
+		return ret;
+
+	drm_bridge_enable(bridge);

This drm_bridge_enable() doesn't seem to serve any purpose here. It also doesn't seem to make much sense to call drm_bridge_() funcs within a bridge op itself.

Yes, removing this line has no effect. I'll send V6. Thank you!

Quoted 12 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 291

Quoted 9 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +8 -0
+	if (ge_b850v3_lvds_dp_i2c->irq) {
+		drm_helper_hpd_irq_event(connector->dev);
+
+		ret = devm_request_threaded_irq(&ge_b850v3_lvds_dp_i2c->dev,
+				ge_b850v3_lvds_dp_i2c->irq, NULL,
+				ge_b850v3_lvds_dp_irq_handler,
+				IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"ge-b850v3-lvds-dp", ptn_bridge);

Is there a reason why we register the interrupt handler here and not in probe?

Quoted 1 line

Yes, drm_bridge_funcs.attach() is called when drm decides it is time to initialize the bridge, which means the drm is already running. i2c_driver.probe() is likely to be called way before drm initializes, and if for some reason the bridge is not attached, there is no need for the interrupt handler. In this case the interrupt handler is mostly used for events related to display plugging/unplugging.

The i2c device can probe before drm, this should be fixed by making sure the device's interrupts are masked before we request the handler, and unmasking it when we know it is ready for use.

If a bridge device exists and isn't attached, then we should question why it probed in the first place if it wasn't to be used.

Besides, in the current scenario, consider the case where the imx driver's call to drm_bridge_attach() succeeds, but then later fails at some point. There would be no one to remove this interrupt handler. The devm_ api attaches the lifetime of the handler with the i2c device, not the drm device.

Also, the call to drm_helper_hpd_irq_event(drm_dev) should go too. These should be called only when the drm device is registered.

Thanks,
Archit

Signature · 6 lines
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#61 lore raw

PATCH v5 4/4 dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Configures the GE B850v3 LVDS/DP++ bridge on the dts file.

Cc Martyn Welch <[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V4

Changes from V3:
- 4/4 instead of 5/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to b850v3-lvds-dp-bridge

 arch/arm/boot/dts/imx6q-b850v3.dts | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +30 -0
index 167f744..8db3bf2 100644
@@ -72,6 +72,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -142,3 +149,26 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73 {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
#62 lore raw

Re: [PATCH V5 4/4] dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

Patch 1/4 is already on linux-next, but what about this one? Ping?

Quoted 73 lines · Peter Senna Tschudin

Configures the GE B850v3 LVDS/DP++ bridge on the dts file.

Cc Martyn Welch <martyn.welch-ZGY8ohtN/[email protected]>
Cc Martin Donnelly <[email protected]>
Cc Javier Martinez Canillas <[email protected]>
Cc Enric Balletbo i Serra <enric.balletbo-ZGY8ohtN/[email protected]>
Cc Philipp Zabel <[email protected]>
Cc Rob Herring <[email protected]>
Cc Fabio Estevam <[email protected]>

---
Unchanged from V4

Changes from V3:
- 4/4 instead of 5/5

Unchanged from V2

Changes from V1:
- Replaced '_' by '-' in node names or compatible strings
- Added missing @73 to b850v3-lvds-dp-bridge

 arch/arm/boot/dts/imx6q-b850v3.dts | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +30 -0
index 167f744..8db3bf2 100644
@@ -72,6 +72,13 @@
 		fsl,data-mapping = "spwg";
 		fsl,data-width = <24>;
 		status = "okay";
+
+		port@4 {
+			reg = <4>;
+			lvds0_out: endpoint {
+				remote-endpoint = <&b850v3_lvds_dp_bridge_in>;
+			};
+		};
 	};
 };
 
@@ -142,3 +149,26 @@
 		reg = <0x4a>;
 	};
 };
+
+&mux2_i2c2 {
+	status = "okay";
+	clock-frequency = <100000>;
+
+	b850v3-lvds-dp-bridge@73 {
+		compatible = "ge,b850v3-lvds-dp";
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		reg = <0x73>;
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		edid-reg = <0x72>;
+
+		port {
+			b850v3_lvds_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

#63 lore raw

Re: [PATCH V5 4/4] dts/imx6q-b850v3: Use GE B850v3 LVDS/DP++ Bridge

Quoted 1 line · Peter Senna Tschudin

Patch 1/4 is already on linux-next, but what about this one? Ping?

Ping me after driver part (patch #3) lands on mainline.

Shawn
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html