Skip to contentSkip to conversation

Add driver for GE B850v3 LVDS/DP++ Bridge

PATCH In review Peter Senna Tschudin 8 participants 33 messages Jan 1, 2017 – Feb 6, 2017

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

v7 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++). 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

This series depends on commit dc80d7038883 ("drm/imx-ldb: Add support to drm-bridge") which is already on linux-next.

The patches from the series:
[1/4] Devicetree documentation for the GE B850v3 LVDS/DP++ Bridge

[2/4] Update the MAINTAINERS file

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

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

Tested on next20161224.

Changes from V6:
- Removed check for pixel clock as the bridge supports up to 330Mhz while the host is limited to 264 MHz in very specific conditions.
- Added a second mutex to avoid concurrency issues while acking interrupts from threaded interrupt handlers.
- Renamed the edid mutex to be more descriptive.
- Added a .detach() function that disables the interrupts.
- Adopted i2c_new_secondary_device() and updated the dts and documentation to match the new method.
- Removed useless test to drm_bridge_add()
- Did some refactoring around devm_request_threaded_irq()
- Added a missing call to i2c_unregister_device() on the i2c_driver.remove() function.

Changes from V5:
- Change to MAINTAINERS in a separate patch
- Reworked interrupt handler initialization
- Removed useless calls to: drm_connector_register(), drm_helper_hpd_irq_event(), and drm_bridge_enable()

Changes from V4:
- 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:
- Removed the patch that was configuring the mapping between IPUs and external displays on the dts file

Peter Senna Tschudin (4):
Documentation/devicetree/bindings: b850v3_lvds_dp
MAINTAINERS: Add entry for GE B850v3 LVDS/DP++ Bridge 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      |  39 +++
 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         | 384 +++++++++++++++++++++
 6 files changed, 473 insertions(+)
 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
#1 lore raw

PATCH v7 1/4 Documentation/devicetree/bindings: b850v3_lvds_dp

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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++++++++++
 1 file changed, 39 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +39 -0
new file mode 100644
index 0000000..1bc6ebf
@@ -0,0 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".
+  - reg : should contain the main address which is used to ack the
+    interrupts and address for edid.
+  - reg-names : comma separeted list of register names. Valid values
+    are "main", and "edid".
+  - 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>.
+  - 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 0x72>;
+		reg-names = "main", "edid";
+
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 1 line
2.5.5
maintainer
#2 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 11 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]>
Quoted 6 lines

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit
message.

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt line 4

Quoted 13 lines
 1 file changed, 39 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +4 -0
new file mode 100644
index 0000000..1bc6ebf
@@ -0,0 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

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 main address which is used to ack the
+    interrupts and address for edid.
+  - reg-names : comma separeted list of register names. Valid values

s/separeted/separated/

Quoted 35 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +32 -0
+    are "main", and "edid".
+  - 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>.
+  - 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 0x72>;
+		reg-names = "main", "edid";
+
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		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

#3 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Rob,

Thank you for the review.

Quoted 17 lines · Rob Herring
Quoted 16 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit
message.

Quoted 4 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 4 lines
Quoted 4 lines
 1 file changed, 39 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Quoted 12 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..1bc6ebf
@@ -0,0 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Quoted 6 lines

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt line 7

Quoted 4 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +3 -0
+  - reg : should contain the main address which is used to ack the
+    interrupts and address for edid.
+  - reg-names : comma separeted list of register names. Valid values

s/separeted/separated/

argh, sorry for this. Will fix it.

Quoted 28 lines
Quoted 28 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +27 -0
+    are "main", and "edid".
+  - 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>.
+  - 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 0x72>;
+		reg-names = "main", "edid";
+
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
Quoted 8 lines
Quoted 8 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +5 -0
+				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

maintainer
#4 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 47 lines · Peter Senna Tschudin

Hi Rob,

Thank you for the review.

Quoted 17 lines · Rob Herring
Quoted 16 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit
message.

Quoted 4 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 4 lines
Quoted 4 lines
 1 file changed, 39 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
Quoted 12 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..1bc6ebf
@@ -0,0 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

Quoted 1 line

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

Rob
--
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

#5 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 50 lines · Rob Herring
Quoted 48 lines · Peter Senna Tschudin

Hi Rob,

Thank you for the review.

Quoted 18 lines · Rob Herring
Quoted 17 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit
message.

Quoted 4 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 16 lines
Quoted 10 lines
 1 file changed, 39 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..1bc6ebf

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 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

Quoted 5 lines
Quoted 1 line

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

-- // --
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
-- // --

--
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

#6 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Peter,

Quoted 62 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 43 lines · Rob Herring
Quoted 42 lines · Peter Senna Tschudin
Quoted 17 lines · Rob Herring
Quoted 16 lines · 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

Quoted 3 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 17 lines
Quoted 11 lines
 1 file changed, 39 insertions(+)
 create mode 100644

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt

(quoted diff) +0 -0

b/Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file mode 100644
index 0000000..1bc6ebf

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 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

Quoted 5 lines
Quoted 2 lines

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

-- // --
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
-- // --

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

Signature · 3 lines
Regards,

Laurent Pinchart
#7 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 1 line · Laurent Pinchart

Hi Peter,

Laurent!

Quoted 64 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 62 lines · Rob Herring
Quoted 43 lines · Peter Senna Tschudin
Quoted 42 lines · Rob Herring
Quoted 17 lines · Peter Senna Tschudin
Quoted 16 lines

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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

Quoted 3 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 ++++++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 17 lines
Quoted 11 lines
 1 file changed, 39 insertions(+)
 create mode 100644

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt

(quoted diff) +0 -0

b/Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file mode 100644
index 0000000..1bc6ebf

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 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

Quoted 5 lines
Quoted 2 lines

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

-- // --
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 7 lines
Quoted 2 lines

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

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other. Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

Quoted 6 lines
Signature · 4 lines
-- 
Regards,

Laurent Pinchart
#8 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Peter,

Quoted 72 lines · On Monday 16 Jan 2017 09:37:11 Peter Senna Tschudin wrote:
Quoted 71 lines · Laurent Pinchart
Quoted 54 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 45 lines · Rob Herring
Quoted 44 lines · Peter Senna Tschudin
Quoted 17 lines · Rob Herring
Quoted 16 lines · 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

Quoted 4 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 +++++++++++

Generally, bindings are not organized by vendor. Put in bindings/display/bridge/... instead.

Will change that.

Quoted 18 lines
Quoted 12 lines
 1 file changed, 39 insertions(+)
 create mode 100644

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt

diff --git a/Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt b/Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file mode 100644
index 0000000..1bc6ebf

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 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Isn't '-lvds-dp' redundant? The part# should be enough.

Quoted 5 lines
Quoted 2 lines

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

Quoted 13 lines

-- // --
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
-- // --

You forgot to prefix your patch series with [HACK] ;-)

Quoted 7 lines
Quoted 2 lines

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other.

Which firmware are you talking about ? Firmware running on the bridges, or somewhere else ?

Quoted 3 lines

Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

You didn't, we did :-) I've been telling for quite some time now that we must decouple bridges from connectors, and this is another example of why we have such a need. Bridges should expose additional functions needed to implement connector operations, and the connector should be instantiated by the display driver with the help of bridge operations. You could then create a connector that relies on one bridge to read the EDID and on the other bridge to handle HPD.

Signature · 3 lines
Regards,

Laurent Pinchart
#9 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 25 lines · Laurent Pinchart

Hi Peter,

Quoted 22 lines · On Monday 16 Jan 2017 09:37:11 Peter Senna Tschudin wrote:
Quoted 21 lines · Laurent Pinchart
Quoted 20 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 19 lines · Rob Herring
Quoted 18 lines · Peter Senna Tschudin
Quoted 17 lines · Rob Herring
Quoted 16 lines · 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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

[...]

Quoted 31 lines
Quoted 31 lines
Quoted 30 lines
Quoted 12 lines
Quoted 9 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 +++++++++++

Isn't '-lvds-dp' redundant? The part# should be enough.

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

Quoted 13 lines

-- // --
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
-- // --

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

Quoted 7 lines
Quoted 4 lines

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other.

Which firmware are you talking about ? Firmware running on the bridges, or somewhere else ?

Each bridge has it's own external flash containing a binary firmware. The goal of the firmware is to configure the output end based on the input end. This is part of what makes handling EDID and HPD challenging.

Quoted 5 lines
Quoted 4 lines

Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

Quoted 7 lines

You didn't, we did :-) I've been telling for quite some time now that we must decouple bridges from connectors, and this is another example of why we have such a need. Bridges should expose additional functions needed to implement connector operations, and the connector should be instantiated by the display driver with the help of bridge operations. You could then create a connector that relies on one bridge to read the EDID and on the other bridge to handle HPD.

Ah thanks. So for now the single DT node approach is acceptable, right? The problem is that even if the driver is getting better on each iteration, the single DT node for two chips issue comes back often and I believe is _the_ issue preventing the driver from getting upstream. V1 was sent ~ 8 months ago...

Can I have some blessing on the single DT node approach for now? I'm one of the 3 proposed maintainers for the driver, and I'm willing to maintain the driver on the long run, as is the same with the other two proposed maintainers. So when the time to split the node in two comes, we will be around, and willing to do it ourselves.

Thank you!

Peter

Quoted 6 lines
Signature · 4 lines
-- 
Regards,

Laurent Pinchart
#10 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Peter,

Quoted 8 lines · On Thursday 19 Jan 2017 09:12:14 Peter Senna Tschudin wrote:
Quoted 7 lines · Laurent Pinchart
Quoted 6 lines · On Monday 16 Jan 2017 09:37:11 Peter Senna Tschudin wrote:
Quoted 5 lines · Laurent Pinchart
Quoted 4 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 3 lines · Rob Herring
Quoted 2 lines · Peter Senna Tschudin
Quoted 1 line · Rob Herring

On Sun, Jan 01, 2017 at 09:24:29PM +0100, Peter Senna Tschudin

Quoted 76 lines · wrote:
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines

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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

[...]

Quoted 34 lines
Quoted 34 lines
Quoted 33 lines
Quoted 23 lines
Quoted 9 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 +++++++++

Isn't '-lvds-dp' redundant? The part# should be enough.

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

-- // --
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

Quoted 5 lines

is as follows:
Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

-- // --

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

Quoted 7 lines
Quoted 4 lines

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other.

Which firmware are you talking about ? Firmware running on the bridges, or somewhere else ?

Each bridge has it's own external flash containing a binary firmware. The goal of the firmware is to configure the output end based on the input end. This is part of what makes handling EDID and HPD challenging.

Quoted 11 lines
Quoted 3 lines

Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

You didn't, we did :-) I've been telling for quite some time now that we must decouple bridges from connectors, and this is another example of why we have such a need. Bridges should expose additional functions needed to implement connector operations, and the connector should be instantiated by the display driver with the help of bridge operations. You could then create a connector that relies on one bridge to read the EDID and on the other bridge to handle HPD.

Quoted 7 lines

Ah thanks. So for now the single DT node approach is acceptable, right? The problem is that even if the driver is getting better on each iteration, the single DT node for two chips issue comes back often and I believe is _the_ issue preventing the driver from getting upstream. V1 was sent ~ 8 months ago...

Can I have some blessing on the single DT node approach for now?

With the "DT as an ABI" approach, I'm afraid not. Temporary hacks are acceptable on the driver side, but you need two nodes in DT.

Quoted 4 lines

I'm one of the 3 proposed maintainers for the driver, and I'm willing to maintain the driver on the long run, as is the same with the other two proposed maintainers. So when the time to split the node in two comes, we will be around, and willing to do it ourselves.

How about putting that team of 3 maintainers to work on fixing the problem in the bridge API ? :-)

Signature · 3 lines
Regards,

Laurent Pinchart
#11 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Quoted 94 lines · Laurent Pinchart

Hi Peter,

Quoted 8 lines · On Thursday 19 Jan 2017 09:12:14 Peter Senna Tschudin wrote:
Quoted 7 lines · Laurent Pinchart
Quoted 6 lines · On Monday 16 Jan 2017 09:37:11 Peter Senna Tschudin wrote:
Quoted 5 lines · Laurent Pinchart
Quoted 4 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 3 lines · Rob Herring
Quoted 2 lines · Peter Senna Tschudin
Quoted 1 line · Rob Herring

On Sun, Jan 01, 2017 at 09:24:29PM +0100, Peter Senna Tschudin

Quoted 82 lines · wrote:
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines

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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

[...]

Quoted 34 lines
Quoted 34 lines
Quoted 33 lines
Quoted 23 lines
Quoted 9 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 +++++++++

Isn't '-lvds-dp' redundant? The part# should be enough.

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of the bridge chip.

From the cover letter:

-- // --
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

Quoted 5 lines

is as follows:
Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

-- // --

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

Quoted 7 lines
Quoted 4 lines

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other.

Which firmware are you talking about ? Firmware running on the bridges, or somewhere else ?

Each bridge has it's own external flash containing a binary firmware. The goal of the firmware is to configure the output end based on the input end. This is part of what makes handling EDID and HPD challenging.

Quoted 11 lines
Quoted 3 lines

Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

You didn't, we did :-) I've been telling for quite some time now that we must decouple bridges from connectors, and this is another example of why we have such a need. Bridges should expose additional functions needed to implement connector operations, and the connector should be instantiated by the display driver with the help of bridge operations. You could then create a connector that relies on one bridge to read the EDID and on the other bridge to handle HPD.

Ah thanks. So for now the single DT node approach is acceptable, right? The problem is that even if the driver is getting better on each iteration, the single DT node for two chips issue comes back often and I believe is _the_ issue preventing the driver from getting upstream. V1 was sent ~ 8 months ago...

Quoted 4 lines
Quoted 1 line

Can I have some blessing on the single DT node approach for now?

With the "DT as an ABI" approach, I'm afraid not. Temporary hacks are acceptable on the driver side, but you need two nodes in DT.

So can I make two node DT "in the right way" and work around current connectors vs. bridge limitations on the driver side? This seems to be doable.

Then I could fix bridge API, with my own driver and update API clients affected by the change...

Quoted 8 lines
Quoted 5 lines

I'm one of the 3 proposed maintainers for the driver, and I'm willing to maintain the driver on the long run, as is the same with the other two proposed maintainers. So when the time to split the node in two comes, we will be around, and willing to do it ourselves.

How about putting that team of 3 maintainers to work on fixing the problem in the bridge API ? :-)

Guess you would be a good lawyer! My point was not exactly that we could work in parallel. Point was that there is redundancy in case one or two of us loose interest. But nice try! :-)

Chances of having resources to fix bridge API and clients were better 6 months ago, but let me see what I can get. Last blocking issue was the migration to atomic, now this. I'm going to need to answer what the next blocking issue is going to be.

Actually in these ~8 months one bit of the required changes was
accepted: dc80d7038883, but this was generic and not related to our specific use case.

Thanks!

Peter

Quoted 6 lines
Signature · 4 lines
-- 
Regards,

Laurent Pinchart
#12 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

Hi Peter,

Quoted 99 lines · On Thursday 19 Jan 2017 10:25:32 Peter Senna Tschudin wrote:
Quoted 93 lines · Laurent Pinchart
Quoted 8 lines · On Thursday 19 Jan 2017 09:12:14 Peter Senna Tschudin wrote:
Quoted 7 lines · Laurent Pinchart
Quoted 6 lines · On Monday 16 Jan 2017 09:37:11 Peter Senna Tschudin wrote:
Quoted 5 lines · Laurent Pinchart
Quoted 4 lines · On Saturday 07 Jan 2017 01:29:52 Peter Senna Tschudin wrote:
Quoted 3 lines · Rob Herring
Quoted 2 lines · Peter Senna Tschudin
Quoted 1 line · Rob Herring

On Sun, Jan 01, 2017 at 09:24:29PM +0100, Peter Senna Tschudin

Quoted 83 lines · wrote:
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 11 lines

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]>

---

Quoted 3 lines

There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

[...]

Quoted 34 lines
Quoted 34 lines
Quoted 33 lines
Quoted 23 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line

.../devicetree/bindings/ge/b850v3-lvds-dp.txt | 39 +++++++++

Isn't '-lvds-dp' redundant? The part# should be enough.

b850v3 is the name of the product, this is why the proposed name. What about, b850v3-dp2 dp2 indicating the second DP output?

Humm, b850v3 is the board name? This node should be the name of

Quoted 1 line

the bridge chip.

From the cover letter:

-- // --
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

Quoted 5 lines

is as follows:
Host -> LVDS|--(STDP4028)--|DP -> DP|--(STDP2690)--|DP++ -> Video output

-- // --

You forgot to prefix your patch series with [HACK] ;-)

How about fixing the issues that make the two DT nodes solution difficult ? What are they ?

Quoted 7 lines
Quoted 4 lines

The Firmware and the hardware design. Both bridges, with stock firmware, are fully capable of providig EDID information and handling interrupts. But on this specific design, with this specific firmware, I need to read EDID from one bridge, and handle interrupts on the other.

Which firmware are you talking about ? Firmware running on the bridges, or somewhere else ?

Each bridge has it's own external flash containing a binary firmware. The goal of the firmware is to configure the output end based on the input end. This is part of what makes handling EDID and HPD challenging.

Quoted 5 lines
Quoted 4 lines

Back when I was starting the development I could not come up with a proper way to split EDID and interrupts between two bridges in a way that would result in a fully functional connector. Did I miss something?

Quoted 7 lines

You didn't, we did :-) I've been telling for quite some time now that we must decouple bridges from connectors, and this is another example of why we have such a need. Bridges should expose additional functions needed to implement connector operations, and the connector should be instantiated by the display driver with the help of bridge operations. You could then create a connector that relies on one bridge to read the EDID and on the other bridge to handle HPD.

Ah thanks. So for now the single DT node approach is acceptable, right? The problem is that even if the driver is getting better on each iteration, the single DT node for two chips issue comes back often and I believe is _the_ issue preventing the driver from getting upstream. V1 was sent ~ 8 months ago...

Quoted 4 lines
Quoted 1 line

Can I have some blessing on the single DT node approach for now?

With the "DT as an ABI" approach, I'm afraid not. Temporary hacks are acceptable on the driver side, but you need two nodes in DT.

Quoted 6 lines

So can I make two node DT "in the right way" and work around current connectors vs. bridge limitations on the driver side? This seems to be doable.

Then I could fix bridge API, with my own driver and update API clients affected by the change...

I'm willing to discuss that as long as the DT bindings are correct, yes.

Quoted 20 lines
Quoted 7 lines
Quoted 4 lines

I'm one of the 3 proposed maintainers for the driver, and I'm willing to maintain the driver on the long run, as is the same with the other two proposed maintainers. So when the time to split the node in two comes, we will be around, and willing to do it ourselves.

How about putting that team of 3 maintainers to work on fixing the problem in the bridge API ? :-)

Guess you would be a good lawyer! My point was not exactly that we could work in parallel. Point was that there is redundancy in case one or two of us loose interest. But nice try! :-)

Chances of having resources to fix bridge API and clients were better 6 months ago, but let me see what I can get. Last blocking issue was the migration to atomic, now this. I'm going to need to answer what the next blocking issue is going to be.

Actually in these ~8 months one bit of the required changes was
accepted: dc80d7038883, but this was generic and not related to our specific use case.

Signature · 3 lines
Regards,

Laurent Pinchart
#13 lore raw

Re: [PATCH V7 1/4] Documentation/devicetree/bindings: b850v3_lvds_dp

In reply to Peter Senna Tschudin

Hi Peter,

Thank you for the patch.

Quoted 31 lines · On Sunday 01 Jan 2017 21:24:29 Peter Senna Tschudin wrote:

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]>

---
There was an Acked-by from Rob Herring <[email protected]> for V6, but I changed the bindings to use i2c_new_secondary_device() so I removed it from the commit message.

 .../devicetree/bindings/ge/b850v3-lvds-dp.txt      | 39 +++++++++++++++++++
 1 file changed, 39 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt
(quoted diff) +0 -0

b/Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file mode 100644
index 0000000..1bc6ebf

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +4 -0
@@ -0,0 +1,39 @@
+Driver for GE B850v3 LVDS/DP++ display bridge
+
+Required properties:
+  - compatible : should be "ge,b850v3-lvds-dp".

Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt line 14

Quoted 10 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +10 -0
+  - reg : should contain the main address which is used to ack the
+    interrupts and address for edid.
+  - reg-names : comma separeted list of register names. Valid values
+    are "main", and "edid".
+  - 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>.
+  - port : should describe the video signal connection between the host
+    and the bridge.

A bridge has, by definition, (at least) one input and (at least) one output. You should thus have (at least) two ports.

Quoted 24 lines
Documentation/devicetree/bindings/ge/b850v3-lvds-dp.txt new file +24 -0
+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 0x72>;
+		reg-names = "main", "edid";
+
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		port {
+			b850v3_dp_bridge_in: endpoint {
+				remote-endpoint = <&lvds0_out>;
+			};
+		};
+	};
+};
Signature · 3 lines
Regards,

Laurent Pinchart
#14 lore raw

PATCH v7 2/4 MAINTAINERS: Add entry for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Update the MAINTAINERS file for the GE B850v3 LVDS/DP++ Bridge.

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]>
CC Archit Taneja <[email protected]>
---
 MAINTAINERS | 8 ++++++++
 1 file changed, 8 insertions(+)
MAINTAINERS +8 -0
index fdd9d5e3..1d3f17a 100644
@@ -5371,6 +5371,14 @@ 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
Signature · 1 line
2.5.5
#15 lore raw

PATCH v7 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]>
CC Archit Taneja <[email protected]>
---
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 384 +++++++++++++++++++++++++++++
 3 files changed, 396 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
drivers/gpu/drm/bridge/Kconfig +11 -0
index eb8688e..e3e1f3b 100644
@@ -48,6 +48,17 @@ config DRM_DW_HDMI_I2S_AUDIO
 	  Support the I2S Audio interface which is part of the Synopsis
 	  Designware HDMI block.
 
+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 2e83a785..886d0fd 100644
@@ -5,6 +5,7 @@ obj-$(CONFIG_DRM_DUMB_VGA_DAC) += dumb-vga-dac.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
 obj-$(CONFIG_DRM_DW_HDMI_I2S_AUDIO) += dw-hdmi-i2s-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_SIL_SII8620) += sil-sii8620.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +384 -0

384 lines not on this page · raw message

Signature · 1 line
2.5.5
#16 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

On 01 January, 2017 21:24 CET, Peter Senna Tschudin <peter.senna-ZGY8ohtN/[email protected]> wrote:

[ ... ]

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 293

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+static void ge_b850v3_lvds_dp_detach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	/* Disable interrupts */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, ~STDP4028_DPTX_IRQ_CONFIG);

~STDP4028_DPTX_IRQ_CONFIG? Argh! Will fix this and resend after reviews...

[ ... ]

--
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

#17 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Peter Senna Tschudin

Hi,

Some comments below.

Quoted 75 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]>
CC Archit Taneja <[email protected]>
---
 drivers/gpu/drm/bridge/Kconfig             |  11 +
 drivers/gpu/drm/bridge/Makefile            |   1 +
 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c | 384 +++++++++++++++++++++++++++++
 3 files changed, 396 insertions(+)
 create mode 100644 drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c
drivers/gpu/drm/bridge/Kconfig +11 -0
index eb8688e..e3e1f3b 100644
@@ -48,6 +48,17 @@ config DRM_DW_HDMI_I2S_AUDIO
 	  Support the I2S Audio interface which is part of the Synopsis
 	  Designware HDMI block.
 
+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 2e83a785..886d0fd 100644
@@ -5,6 +5,7 @@ obj-$(CONFIG_DRM_DUMB_VGA_DAC) += dumb-vga-dac.o
 obj-$(CONFIG_DRM_DW_HDMI) += dw-hdmi.o
 obj-$(CONFIG_DRM_DW_HDMI_AHB_AUDIO) += dw-hdmi-ahb-audio.o
 obj-$(CONFIG_DRM_DW_HDMI_I2S_AUDIO) += dw-hdmi-i2s-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_SIL_SII8620) += sil-sii8620.o
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +0 -0
new file mode 100644
index 0000000..4574f6e

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 2

Quoted 2 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +2 -0
@@ -0,0 +1,384 @@
+/*
+ * Driver for GE B850v3 DP display bridge

Mentioning LVDS to DP++ here would be nice.

Quoted 122 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +122 -0
+
+ * 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>
+
+#define DEFAULT_EDID_REG 0x72
+#define DEFAULT_EDID_REG_NAME "edid"
+
+#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 edid_mutex;
+	struct mutex irq_reg_mutex;
+};
+
+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;
+	}
+

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+	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.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 142

Quoted 8 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +8 -0
+		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;
+		}

We should ideally check if the extension blocks are valid too.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 154

Quoted 12 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +12 -0
+	}
+
+	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;

Why are the bridge pointers named ptn_bridge? I'm guessing it's because you used nxp-ptn3460 bridge driver as reference. You should use something relevant to your device.

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 161

Quoted 7 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +7 -0
+	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->edid_mutex);

Do we really need this mutex? All the paths that call a connector's get_modes hold the drm device's dev->mode_config.mutex lock anyway.

Quoted 56 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +56 -0
+
+	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->edid_mutex);
+
+	return num_modes;
+}
+
+
+static enum drm_mode_status ge_b850v3_lvds_dp_mode_valid(
+		struct drm_connector *connector, struct drm_display_mode *mode)
+{
+	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,

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 227

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+	.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->irq_reg_mutex);

Do we need this mutex? The handler is registered with the IRQF_ONESHOT flag, so we won't get another interrupt until this handler returns.

Quoted 34 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +34 -0
+
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_STS_REG, STDP4028_DPTX_IRQ_CLEAR);
+
+	mutex_unlock(&ptn_bridge->irq_reg_mutex);
+
+	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);

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 271

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+	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_helper_hpd_irq_event(connector->dev);

This call doesn't serve any purpose for a connector until it is registered. You can drop this.

Quoted 34 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +34 -0
+
+	/* 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);
+
+	/* Enable interrupts */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, STDP4028_DPTX_IRQ_CONFIG);
+
+	return 0;
+}
+
+static void ge_b850v3_lvds_dp_detach(struct drm_bridge *bridge)
+{
+	struct ge_b850v3_lvds_dp *ptn_bridge
+			= bridge_to_ge_b850v3_lvds_dp(bridge);
+	struct i2c_client *ge_b850v3_lvds_dp_i2c
+			= ptn_bridge->ge_b850v3_lvds_dp_i2c;
+
+	/* Disable interrupts */
+	i2c_smbus_write_word_data(ge_b850v3_lvds_dp_i2c,
+			STDP4028_DPTX_IRQ_EN_REG, ~STDP4028_DPTX_IRQ_CONFIG);
+}
+
+static const struct drm_bridge_funcs ge_b850v3_lvds_dp_funcs = {
+	.attach = ge_b850v3_lvds_dp_attach,
+	.detach = ge_b850v3_lvds_dp_detach,
+};
+
+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;

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 315

Quoted 10 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +10 -0
+
+	ptn_bridge = devm_kzalloc(dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge)
+		return -ENOMEM;
+
+	mutex_init(&ptn_bridge->edid_mutex);
+	mutex_init(&ptn_bridge->irq_reg_mutex);
+
+	ptn_bridge->ge_b850v3_lvds_dp_i2c = ge_b850v3_lvds_dp_i2c;
+	ptn_bridge->bridge.driver_private = ptn_bridge;

bridge->driver_private isn't used by the driver anywhere. It's probably better to drop it until it is used.

Quoted 70 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +69 -0
+	i2c_set_clientdata(ge_b850v3_lvds_dp_i2c, ptn_bridge);
+
+	ptn_bridge->edid_i2c = i2c_new_secondary_device(ge_b850v3_lvds_dp_i2c,
+			DEFAULT_EDID_REG_NAME, DEFAULT_EDID_REG);
+
+	if (!ptn_bridge->edid_i2c) {
+		dev_err(dev, "Error registering edid i2c_client, aborting...\n");
+		return -ENODEV;
+	}
+
+	ptn_bridge->bridge.funcs = &ge_b850v3_lvds_dp_funcs;
+	ptn_bridge->bridge.of_node = dev->of_node;
+	drm_bridge_add(&ptn_bridge->bridge);
+
+	/* 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);
+
+	if (!ge_b850v3_lvds_dp_i2c->irq)
+		return 0;
+
+	return 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);
+}
+
+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);
+
+	i2c_unregister_device(ptn_bridge->edid_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
#18 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

[...]

I fixed all your other suggestions. Thank you!

#19 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 19 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

BR,
Jani.

Quoted 8 lines

[...]

I fixed all your other suggestions. Thank you!
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Signature · 1 line
Jani Nikula, Intel Open Source Technology Center
#20 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 23 lines · Jani Nikula
Quoted 19 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add in the future.

Thanks,
Archit

Quoted 14 lines

BR,
Jani.

Quoted 8 lines

[...]

I fixed all your other suggestions. Thank you!
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

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

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Hi Archit,

Quoted 43 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 20 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Quoted 5 lines

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 11 lines

Thanks,
Archit

Quoted 6 lines

BR,
Jani.

Quoted 1 line
Quoted 12 lines
Quoted 8 lines
Quoted 7 lines

[...]

I fixed all your other suggestions. Thank you!
_______________________________________________
dri-devel mailing list [email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Signature · 2 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

#22 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 20 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c line 134

Quoted 6 lines
drivers/gpu/drm/bridge/ge_b850v3_lvds_dp.c new file +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?
-Daniel

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
#23 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 60 lines · Daniel Vetter
Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 20 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
(quoted diff) +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

Review

Quoted 6 lines
(quoted diff) +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

Yes, sure!

Quoted 5 lines

-Daniel

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

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 63 lines · Peter Senna Tschudin
Quoted 60 lines · Daniel Vetter
Quoted 58 lines · Peter Senna Tschudin

Hi Archit,

Quoted 43 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 20 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 4 lines
Quoted 4 lines
(quoted diff) +4 -0
+	total_size = (block[EDID_EXT_BLOCK_CNT] + 1) * EDID_LENGTH;
+	if (total_size > EDID_LENGTH) {
+		kfree(block);
+		block = kmalloc(total_size, GFP_KERNEL);
Quoted 8 lines

Review

Quoted 6 lines
(quoted diff) +6 -0
+		if (!block)
+			return NULL;
+
+		/* Yes, read the entire buffer, and do not skip the first
+		 * EDID_LENGTH bytes.
+		 */

Is this the reason why you aren't using drm_do_get_edid()?

Quoted 2 lines

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Quoted 5 lines

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines
Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

The edid thing was only a suggestion. As Daniel said, it's okay to work on it post merge.

Please do fix the minor comments I mentioned in the latest patch set. I'll pull in the first 3 patches once Rob H gives an Ack on the DT bindings patch. The 4th patch needs to go through the SoC maintainer.

Thanks,
Archit

Quoted 14 lines

Yes, sure!

Quoted 5 lines

-Daniel
--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch

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
#25 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Daniel Vetter
Quoted 60 lines · Daniel Vetter
Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 17 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 11 lines
Quoted 10 lines
(quoted diff) +10 -0
+      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.
+               */
Quoted 5 lines
Quoted 1 line

Is this the reason why you aren't using drm_do_get_edid()?

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

Wasn't there some entry level for people ? It's not like Peter is thick or anything, just that he has not reviewed or replied (afaict) to any patches, even on ones where he was explicitly cc'd. Although he's got dozens of contributions elsewhere, there's only a single patch of his in DRM.

-Emil

#26 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 65 lines · Emil Velikov
Quoted 60 lines · Daniel Vetter
Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 17 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 11 lines
Quoted 10 lines
(quoted diff) +10 -0
+      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.
+               */
Quoted 5 lines
Quoted 1 line

Is this the reason why you aren't using drm_do_get_edid()?

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

Quoted 6 lines

Wasn't there some entry level for people ? It's not like Peter is thick or anything, just that he has not reviewed or replied (afaict)

to any patches, even on ones where he was explicitly cc'd. Although he's got dozens of contributions elsewhere, there's only a single patch of his in DRM.

"Any" is an exaggeration. See:

c6c1f9bc798bee7cf
...

Even if my involvement with drm started with my driver last year, I took my personal time to report this issue, and to test multiple iterations of the fix. Then my "single patch in DRM" involved some work to get it right, and it is something, that made imx-ldb better by adding a feature it did not support before.

You made me curious. Can you point to the explicit cases where Peter "has not reviewed or replied to any patches, even on ones where he was explicitly cc'd"? And can you list case by case your take on why I should have done differently?

Also it is not clear to me what the problem really is here. I'm assuming is not me, but if I was qualified from your perspective then would be no problem, right?

#27 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 79 lines · Peter Senna Tschudin
Quoted 65 lines · Emil Velikov
Quoted 60 lines · Daniel Vetter
Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 17 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 11 lines
Quoted 10 lines
(quoted diff) +10 -0
+      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.
+               */
Quoted 5 lines
Quoted 1 line

Is this the reason why you aren't using drm_do_get_edid()?

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

Quoted 6 lines

Wasn't there some entry level for people ? It's not like Peter is thick or anything, just that he has not reviewed or replied (afaict)

to any patches, even on ones where he was explicitly cc'd. Although he's got dozens of contributions elsewhere, there's only a single patch of his in DRM.

"Any" is an exaggeration. See:

c6c1f9bc798bee7cf
...

Quoted 8 lines

Even if my involvement with drm started with my driver last year, I took my personal time to report this issue, and to test multiple iterations of the fix. Then my "single patch in DRM" involved some work to get it right, and it is something, that made imx-ldb better by adding a feature it did not support before.

You made me curious. Can you point to the explicit cases where Peter "has not reviewed or replied to any patches, even on ones where he was explicitly cc'd"? And can you list case by case your take on why I should have done differently?

Also it is not clear to me what the problem really is here. I'm assuming is not me, but if I was qualified from your perspective then would be no problem, right?

Ok seems like my wording came out, rather off.

Peter, as mentioned on IRC my messages is not aimed to be picking on you in the slightest way. Pardon if it came as such.

It had two main points:
- Daniel, gents - is drm-misc aimed at devs with limited (no?)
review/commit history in the area and/or the kernel in general ? In this case, Peter have quite noticeable experience [in kernel development] with little-to no in DRM.
Should one draw a line in the sand somewhere ?

- You patch has been on a long [bit rocky road] for a while, with
devs giving you [what seems like] a partial reviews.
Have you considered reviewing others' patches in exchange for a [more complete one] of this one ? According to git log people have poked you a handful of times, but seemingly you were busy.

Note that neither of the above is supposed to belittle individuals or the group maintainer model of drm-misc, but to point out what may not seem obvious.

Regards,
Emil
--
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 V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 5 lines · Emil Velikov

- Daniel, gents - is drm-misc aimed at devs with limited (no?)
review/commit history in the area and/or the kernel in general ? In this case, Peter have quite noticeable experience [in kernel development] with little-to no in DRM.
Should one draw a line in the sand somewhere ?

We're fairly lenient with accepting new drivers already, and for drivers in drm-misc we require some checks and peer review to make sure new folks learn faster. But it's an expirement, I fully expect some fallout and then some learnign and fine tuning from that, and maybe we even need to crank requirements back up to a much higher level of experience.

But for drivers I'm honestly not too concerned: As long as there's some process in place to foster learning&collaboration (which I think will be better with this) I really don't see much harm in merging non-perfect driver code.

Quoted 5 lines

- You patch has been on a long [bit rocky road] for a while, with
devs giving you [what seems like] a partial reviews.
Have you considered reviewing others' patches in exchange for a [more complete one] of this one ? According to git log people have poked you a handful of times, but seemingly you were busy.

Just a clarification: When I review drivers I don't do a fairly cursor look, hunting for areas where some refactoring and align with drm best practices would be good. And I think that's perfectly fine and enough to get a driver landed, but we're definitely not 100% consistent on this within drm-misc. Probably will take some time to figure this out. -Daniel

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
#29 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 13 lines · Daniel Vetter
Quoted 5 lines · Emil Velikov

- Daniel, gents - is drm-misc aimed at devs with limited (no?)
review/commit history in the area and/or the kernel in general ? In this case, Peter have quite noticeable experience [in kernel development] with little-to no in DRM.
Should one draw a line in the sand somewhere ?

We're fairly lenient with accepting new drivers already, and for drivers in drm-misc we require some checks and peer review to make sure new folks learn faster. But it's an expirement, I fully expect some fallout and then some learnign and fine tuning from that, and maybe we even need to crank requirements back up to a much higher level of experience.

Quoted 5 lines

But for drivers I'm honestly not too concerned: As long as there's some process in place to foster learning&collaboration (which I think will be better with this) I really don't see much harm in merging non-perfect driver code.

Thank you Daniel, this is very well said. Being lenient to begin with and adjusting where/if needed is the more welcoming approach, indeed. Doubt there'll be (m)any cases of (ab|mis)use but even if so, nobody can see the future.

Quoted 11 lines
Quoted 5 lines

- You patch has been on a long [bit rocky road] for a while, with
devs giving you [what seems like] a partial reviews.
Have you considered reviewing others' patches in exchange for a [more complete one] of this one ? According to git log people have poked you a handful of times, but seemingly you were busy.

Just a clarification: When I review drivers I don't do a fairly cursor look, hunting for areas where some refactoring and align with drm best practices would be good. And I think that's perfectly fine and enough to get a driver landed, but we're definitely not 100% consistent on this within drm-misc. Probably will take some time to figure this out.

Seems like I've missed "Peter, your patch..." above, silly me.

Thanks for answering my [what may seem silly] questions ;-) Emil
--
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

#30 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

Quoted 19 lines · Emil Velikov
Quoted 13 lines · Daniel Vetter
Quoted 5 lines · Emil Velikov

- Daniel, gents - is drm-misc aimed at devs with limited (no?)
review/commit history in the area and/or the kernel in general ? In this case, Peter have quite noticeable experience [in kernel development] with little-to no in DRM.
Should one draw a line in the sand somewhere ?

We're fairly lenient with accepting new drivers already, and for drivers in drm-misc we require some checks and peer review to make sure new folks learn faster. But it's an expirement, I fully expect some fallout and then some learnign and fine tuning from that, and maybe we even need to crank requirements back up to a much higher level of experience.

Quoted 5 lines

But for drivers I'm honestly not too concerned: As long as there's some process in place to foster learning&collaboration (which I think will be better with this) I really don't see much harm in merging non-perfect driver code.

Quoted 4 lines

Thank you Daniel, this is very well said. Being lenient to begin with and adjusting where/if needed is the more welcoming approach, indeed. Doubt there'll be (m)any cases of (ab|mis)use but even if so, nobody can see the future.

Yeah, the big idea behind drm-misc is also to make drm more welcoming for new folks, and I think starting with a small driver or a few driver patches for hw you have lying around for hacking is a great way. And every once in a while we can "trick" one of these newbies into doing some cool core refactoring or cleanup, and maybe even to stick around for a while.

All for world domination and other evil purposes ofc :-) -Daniel

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
#31 lore raw

Re: [PATCH V7 3/4] drm/bridge: Add driver for GE B850v3 LVDS/DP++ Bridge

In reply to Daniel Vetter
Quoted 60 lines · Daniel Vetter
Quoted 51 lines · Peter Senna Tschudin

Hi Archit,

Quoted 46 lines · Archit Taneja
Quoted 24 lines · Jani Nikula
Quoted 17 lines · Peter Senna Tschudin

On Thu, Jan 05, 2017 at 01:18:47PM +0530, Archit Taneja wrote:
Hi Archit,

Thank you for the comments!

[...]

Quoted 11 lines
Quoted 10 lines
(quoted diff) +10 -0
+      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.
+               */
Quoted 5 lines
Quoted 1 line

Is this the reason why you aren't using drm_do_get_edid()?

Yes, for some hw specific reason, it is necessary to read the entire EDID buffer starting from 0, not block by block.

Quoted 6 lines

Hrmh, I'm planning on moving the edid override and firmware edid mechanisms at the drm_do_get_edid() level to be able to truly and transparently use a different edid. Currently, they're only used for modes, really, and lead to some info retrieved from overrides, some from the real edid. This kind of hacks will bypass the override/firmware edid mechanisms then too. :(

It seems like there is a HW issue which prevents them from reading EDID from an offset. So, I'm not sure if it is a hack or a HW limitation.

One way around this would be to hide the HW requirement in the get_edid_block func pointer passed to drm_do_get_edid(). This would, however, result in more i2c reads (equal to # of extension blocks) than what the patch currently does.

Peter, if you think doing extra EDID reads isn't too costly on your platform, you could consider using drm_do_get_edid(). If not, I guess you'll miss out on the additional functionality Jani is going to add

Quoted 7 lines
Quoted 1 line

in the future.

My concern is that for almost one year now, every time I fix something one or two new requests are made. I'm happy to fix the driver, but I want a list of the changes that are required to get it upstream, before I make more changes. Can we agree on exactly what is preventing this driver to get upstream? Then I'll fix it.

Quoted 3 lines

I think addressing this edid reading question post-merge is perfectly fine. Aside, want to keep maintaing your stuff as part of the drm-misc group, with the drivers-in-misc experiment?

Wasn't there some entry level for people ? It's not like Peter is thick or anything, just that he has not reviewed or replied (afaict) to any patches, even on ones where he was explicitly cc'd. Although he's got dozens of contributions elsewhere, there's only a single patch of his in DRM.

-Emil
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#32 lore raw

PATCH v7 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 <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]>
---
 arch/arm/boot/dts/imx6q-b850v3.dts | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/arm/boot/dts/imx6q-b850v3.dts +30 -0
index 167f744..ce0ca3a 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 0x72>;
+		reg-names = "main", "edid";
+
+		interrupt-parent = <&gpio2>;
+		interrupts = <0 IRQ_TYPE_LEVEL_HIGH>;
+
+		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