[PATCH v2 0/6] R-Car DU DPAD support for D3 and E3

[PATCH v2 0/6] R-Car DU DPAD support for D3 and E3

[Laurent Pinchart]

Hello,

This series adds support for the DPAD0 output for the D3 and E3 SoCs. On
the Draak and Ebisu boards, DPAD0 is used for the VGA output.

Patches 1/6 and 2/6 prepare the grounds by successfully probing LVDS
encoders that have no connected output. This is required in order to
provide a dot clock to the DPAD output, as on the D3 and E3 SoCs the dot
clock is provided by the LVDS encoders.

Patch 3/6 then adds an API to the LVDS encoder driver to control the
LVDS output clock independently of the LVDS encoder itself, and patch
4/6 makes use of that API to control the clock from the DU driver.

Patches 5/6 and 6/6 finally enable the LVDS1 encoders in the Ebisu and
Draak boards DTs, required to operate the HDMI (out of LVDS0) and VGA
(using the LVDS1 dot clock) together.

Changes compared to v1 are minor and include compilation fixes (in patch
3/6), typo fixes and additional comments in the DT files.

The patches have been tested on Draak only as I don't have access to an
Ebisu board, but they should work equally well on both boards.

For your convenience the patches are available from

        git://linuxtv.org/pinchartl/media.git drm/du/d3e3

Laurent Pinchart (6):
  drm: rcar-du: Simplify encoder registration
  drm: rcar-du: lvds: Don't fail probe if output is not connected on
    D3/E3
  drm: rcar-du: lvds: Add API to enable/disable clock output
  drm: rcar-du: Turn LVDS clock output on/off for DPAD0 output on D3/E3
  arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder
  arm64: dts: renesas: r8a77995: draak: Enable LVDS1 encoder

 .../arm64/boot/dts/renesas/r8a77990-ebisu.dts |  2 +
 .../arm64/boot/dts/renesas/r8a77995-draak.dts |  7 ++
 drivers/gpu/drm/Makefile                      |  2 +-
 drivers/gpu/drm/rcar-du/Kconfig               |  1 +
 drivers/gpu/drm/rcar-du/rcar_du_crtc.c        | 34 +++++++
 drivers/gpu/drm/rcar-du/rcar_du_drv.h         |  3 +
 drivers/gpu/drm/rcar-du/rcar_du_encoder.c     |  4 +-
 drivers/gpu/drm/rcar-du/rcar_du_encoder.h     |  3 +-
 drivers/gpu/drm/rcar-du/rcar_du_kms.c         | 52 +---------
 drivers/gpu/drm/rcar-du/rcar_lvds.c           | 94 ++++++++++++++++---
 drivers/gpu/drm/rcar-du/rcar_lvds.h           | 27 ++++++
 11 files changed, 164 insertions(+), 65 deletions(-)
 create mode 100644 drivers/gpu/drm/rcar-du/rcar_lvds.h

-- 
Regards,

Laurent Pinchart

[PATCH v2 1/6] drm: rcar-du: Simplify encoder registration

[Laurent Pinchart]

Before the driver fully moved to drm_bridge and drm_panel, it was
necessary to parse DT and locate encoder and connector nodes. The
connector node is now unused and can be removed as a parameter to
rcar_du_encoder_init(). As a consequence rcar_du_encoders_init_one() can
be greatly simplified, removing most of the DT parsing.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_encoder.c |  3 +-
 drivers/gpu/drm/rcar-du/rcar_du_encoder.h |  3 +-
 drivers/gpu/drm/rcar-du/rcar_du_kms.c     | 52 ++---------------------
 3 files changed, 6 insertions(+), 52 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
index f16209499117..ba2b38f76b54 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
@@ -30,8 +30,7 @@ static const struct drm_encoder_funcs encoder_funcs = {
 
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_output output,
-			 struct device_node *enc_node,
-			 struct device_node *con_node)
+			 struct device_node *enc_node)
 {
 	struct rcar_du_encoder *renc;
 	struct drm_encoder *encoder;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
index 552f2a02e5b5..df9be4524301 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h
@@ -26,7 +26,6 @@ struct rcar_du_encoder {
 
 int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 			 enum rcar_du_output output,
-			 struct device_node *enc_node,
-			 struct device_node *con_node);
+			 struct device_node *enc_node);
 
 #endif /* __RCAR_DU_ENCODER_H__ */
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index 015c421451dc..1bd75187d039 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -330,17 +330,10 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu,
 				     enum rcar_du_output output,
 				     struct of_endpoint *ep)
 {
-	struct device_node *connector = NULL;
-	struct device_node *encoder = NULL;
-	struct device_node *ep_node = NULL;
-	struct device_node *entity_ep_node;
 	struct device_node *entity;
 	int ret;
 
-	/*
-	 * Locate the connected entity and infer its type from the number of
-	 * endpoints.
-	 */
+	/* Locate the connected entity and initialize the encoder. */
 	entity = of_graph_get_remote_port_parent(ep->local_node);
 	if (!entity) {
 		dev_dbg(rcdu->dev, "unconnected endpoint %pOF, skipping\n",
@@ -356,50 +349,13 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu,
 		return -ENODEV;
 	}
 
-	entity_ep_node = of_graph_get_remote_endpoint(ep->local_node);
-
-	for_each_endpoint_of_node(entity, ep_node) {
-		if (ep_node == entity_ep_node)
-			continue;
-
-		/*
-		 * We've found one endpoint other than the input, this must
-		 * be an encoder. Locate the connector.
-		 */
-		encoder = entity;
-		connector = of_graph_get_remote_port_parent(ep_node);
-		of_node_put(ep_node);
-
-		if (!connector) {
-			dev_warn(rcdu->dev,
-				 "no connector for encoder %pOF, skipping\n",
-				 encoder);
-			of_node_put(entity_ep_node);
-			of_node_put(encoder);
-			return -ENODEV;
-		}
-
-		break;
-	}
-
-	of_node_put(entity_ep_node);
-
-	if (!encoder) {
-		dev_warn(rcdu->dev,
-			 "no encoder found for endpoint %pOF, skipping\n",
-			 ep->local_node);
-		of_node_put(entity);
-		return -ENODEV;
-	}
-
-	ret = rcar_du_encoder_init(rcdu, output, encoder, connector);
+	ret = rcar_du_encoder_init(rcdu, output, entity);
 	if (ret && ret != -EPROBE_DEFER)
 		dev_warn(rcdu->dev,
 			 "failed to initialize encoder %pOF on output %u (%d), skipping\n",
-			 encoder, output, ret);
+			 entity, output, ret);
 
-	of_node_put(encoder);
-	of_node_put(connector);
+	of_node_put(entity);
 
 	return ret;
 }
-- 
Regards,

Laurent Pinchart

[PATCH v2 2/6] drm: rcar-du: lvds: Don't fail probe if output is not connected on D3/E3

[Laurent Pinchart]

On the D3 and E3 SoCs the LVDS encoder has an extended internal PLL and
supplies a clock to the DU. That clock is used not only for the LVDS
outputs but also for the DPAD output. The LVDS encoder thus needs to be
available to the DU even when its output is disabled. Don't fail probe
in that case on D3 and E3.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
Changes since v1:

- Typo fix in the commit message
---
 drivers/gpu/drm/rcar-du/rcar_lvds.c | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/rcar-du/rcar_lvds.c b/drivers/gpu/drm/rcar-du/rcar_lvds.c
index caea2c590b7a..b21906b81f10 100644
--- a/drivers/gpu/drm/rcar-du/rcar_lvds.c
+++ b/drivers/gpu/drm/rcar-du/rcar_lvds.c
@@ -544,7 +544,10 @@ static int rcar_lvds_attach(struct drm_bridge *bridge)
 		return drm_bridge_attach(bridge->encoder, lvds->next_bridge,
 					 bridge);
 
-	/* Otherwise we have a panel, create a connector. */
+	/* Otherwise if we have a panel, create a connector. */
+	if (!lvds->panel)
+		return 0;
+
 	ret = drm_connector_init(bridge->dev, connector, &rcar_lvds_conn_funcs,
 				 DRM_MODE_CONNECTOR_LVDS);
 	if (ret < 0)
@@ -592,7 +595,8 @@ static int rcar_lvds_parse_dt(struct rcar_lvds *lvds)
 	local_output = of_graph_get_endpoint_by_regs(lvds->dev->of_node, 1, 0);
 	if (!local_output) {
 		dev_dbg(lvds->dev, "unconnected port@1\n");
-		return -ENODEV;
+		ret = -ENODEV;
+		goto done;
 	}
 
 	/*
@@ -642,6 +646,15 @@ static int rcar_lvds_parse_dt(struct rcar_lvds *lvds)
 	of_node_put(remote_input);
 	of_node_put(remote);
 
+	/*
+	 * On D3/E3 the LVDS encoder provides a clock to the DU, which can be
+	 * used for the DPAD output even when the LVDS output is not connected.
+	 * Don't fail probe in that case as the DU will need the bridge to
+	 * control the clock.
+	 */
+	if (lvds->info->quirks & RCAR_LVDS_QUIRK_EXT_PLL)
+		return ret == -ENODEV ? 0 : ret;
+
 	return ret;
 }
 
-- 
Regards,

Laurent Pinchart

[PATCH v2 3/6] drm: rcar-du: lvds: Add API to enable/disable clock output

[Laurent Pinchart]

On the D3 and E3 platforms, the LVDS internal PLL supplies the pixel
clock to the DU. This works automatically for LVDS outputs as the LVDS
encoder is enabled through the bridge API, enabling the internal PLL and
clock output. However, when using the DU DPAD output with the LVDS
outputs turned off, the LVDS PLL needs to be controlled manually. Add an
API to do so, to be called by the DU driver.

The drivers/gpu/drm/rcar-du/ directory has to be treated as obj-y
unconditionally, as the LVDS driver could be built-in while the DU
driver is compiled as a module.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
Changes since v1:

- Provide stubs when the LVDS driver is not available
- Export the LVDS driver API symbols
- Fix dependency with the DU and LVDS drivers to prevent DU=y && LVDS=m
- Always treat drivers/gpu/drm/rcar-du/ as obj-y
---
 drivers/gpu/drm/Makefile            |  2 +-
 drivers/gpu/drm/rcar-du/Kconfig     |  1 +
 drivers/gpu/drm/rcar-du/rcar_lvds.c | 77 +++++++++++++++++++++++++----
 drivers/gpu/drm/rcar-du/rcar_lvds.h | 27 ++++++++++
 4 files changed, 96 insertions(+), 11 deletions(-)
 create mode 100644 drivers/gpu/drm/rcar-du/rcar_lvds.h

diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index ce8d1d384319..59e426704755 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -81,7 +81,7 @@ obj-$(CONFIG_DRM_UDL) += udl/
 obj-$(CONFIG_DRM_AST) += ast/
 obj-$(CONFIG_DRM_ARMADA) += armada/
 obj-$(CONFIG_DRM_ATMEL_HLCDC)	+= atmel-hlcdc/
-obj-$(CONFIG_DRM_RCAR_DU) += rcar-du/
+obj-y			+= rcar-du/
 obj-$(CONFIG_DRM_SHMOBILE) +=shmobile/
 obj-y			+= omapdrm/
 obj-$(CONFIG_DRM_SUN4I) += sun4i/
diff --git a/drivers/gpu/drm/rcar-du/Kconfig b/drivers/gpu/drm/rcar-du/Kconfig
index 225141656e19..7c36e2777a15 100644
--- a/drivers/gpu/drm/rcar-du/Kconfig
+++ b/drivers/gpu/drm/rcar-du/Kconfig
@@ -4,6 +4,7 @@ config DRM_RCAR_DU
 	depends on DRM && OF
 	depends on ARM || ARM64
 	depends on ARCH_RENESAS || COMPILE_TEST
+	imply DRM_RCAR_LVDS
 	select DRM_KMS_HELPER
 	select DRM_KMS_CMA_HELPER
 	select DRM_GEM_CMA_HELPER
diff --git a/drivers/gpu/drm/rcar-du/rcar_lvds.c b/drivers/gpu/drm/rcar-du/rcar_lvds.c
index b21906b81f10..7b9d8cccead4 100644
--- a/drivers/gpu/drm/rcar-du/rcar_lvds.c
+++ b/drivers/gpu/drm/rcar-du/rcar_lvds.c
@@ -22,6 +22,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_panel.h>
 
+#include "rcar_lvds.h"
 #include "rcar_lvds_regs.h"
 
 struct rcar_lvds;
@@ -182,8 +183,9 @@ struct pll_info {
 
 static void rcar_lvds_d3_e3_pll_calc(struct rcar_lvds *lvds, struct clk *clk,
 				     unsigned long target, struct pll_info *pll,
-				     u32 clksel)
+				     u32 clksel, bool dot_clock_only)
 {
+	unsigned int div7 = dot_clock_only ? 1 : 7;
 	unsigned long output;
 	unsigned long fin;
 	unsigned int m_min;
@@ -217,9 +219,9 @@ static void rcar_lvds_d3_e3_pll_calc(struct rcar_lvds *lvds, struct clk *clk,
 	 *                     `------------> | |
 	 *                                    |/
 	 *
-	 * The /7 divider is optional when the LVDS PLL is used to generate a
-	 * dot clock for the DU RGB output, without using the LVDS encoder. We
-	 * don't support this configuration yet.
+	 * The /7 divider is optional, it is enabled when the LVDS PLL is used
+	 * to drive the LVDS encoder, and disabled when  used to generate a dot
+	 * clock for the DU RGB output, without using the LVDS encoder.
 	 *
 	 * The PLL allowed input frequency range is 12 MHz to 192 MHz.
 	 */
@@ -279,7 +281,7 @@ static void rcar_lvds_d3_e3_pll_calc(struct rcar_lvds *lvds, struct clk *clk,
 				 * the PLL, followed by a an optional fixed /7
 				 * divider.
 				 */
-				fout = fvco / (1 << e) / 7;
+				fout = fvco / (1 << e) / div7;
 				div = DIV_ROUND_CLOSEST(fout, target);
 				diff = abs(fout / div - target);
 
@@ -300,7 +302,7 @@ static void rcar_lvds_d3_e3_pll_calc(struct rcar_lvds *lvds, struct clk *clk,
 
 done:
 	output = fin * pll->pll_n / pll->pll_m / (1 << pll->pll_e)
-	       / 7 / pll->div;
+	       / div7 / pll->div;
 	error = (long)(output - target) * 10000 / (long)target;
 
 	dev_dbg(lvds->dev,
@@ -310,17 +312,18 @@ static void rcar_lvds_d3_e3_pll_calc(struct rcar_lvds *lvds, struct clk *clk,
 		pll->pll_m, pll->pll_n, pll->pll_e, pll->div);
 }
 
-static void rcar_lvds_pll_setup_d3_e3(struct rcar_lvds *lvds, unsigned int freq)
+static void __rcar_lvds_pll_setup_d3_e3(struct rcar_lvds *lvds,
+					unsigned int freq, bool dot_clock_only)
 {
 	struct pll_info pll = { .diff = (unsigned long)-1 };
 	u32 lvdpllcr;
 
 	rcar_lvds_d3_e3_pll_calc(lvds, lvds->clocks.dotclkin[0], freq, &pll,
-				 LVDPLLCR_CKSEL_DU_DOTCLKIN(0));
+				 LVDPLLCR_CKSEL_DU_DOTCLKIN(0), dot_clock_only);
 	rcar_lvds_d3_e3_pll_calc(lvds, lvds->clocks.dotclkin[1], freq, &pll,
-				 LVDPLLCR_CKSEL_DU_DOTCLKIN(1));
+				 LVDPLLCR_CKSEL_DU_DOTCLKIN(1), dot_clock_only);
 	rcar_lvds_d3_e3_pll_calc(lvds, lvds->clocks.extal, freq, &pll,
-				 LVDPLLCR_CKSEL_EXTAL);
+				 LVDPLLCR_CKSEL_EXTAL, dot_clock_only);
 
 	lvdpllcr = LVDPLLCR_PLLON | pll.clksel | LVDPLLCR_CLKOUT
 		 | LVDPLLCR_PLLN(pll.pll_n - 1) | LVDPLLCR_PLLM(pll.pll_m - 1);
@@ -329,6 +332,9 @@ static void rcar_lvds_pll_setup_d3_e3(struct rcar_lvds *lvds, unsigned int freq)
 		lvdpllcr |= LVDPLLCR_STP_CLKOUTE | LVDPLLCR_OUTCLKSEL
 			 |  LVDPLLCR_PLLE(pll.pll_e - 1);
 
+	if (dot_clock_only)
+		lvdpllcr |= LVDPLLCR_OCKSEL;
+
 	rcar_lvds_write(lvds, LVDPLLCR, lvdpllcr);
 
 	if (pll.div > 1)
@@ -342,6 +348,57 @@ static void rcar_lvds_pll_setup_d3_e3(struct rcar_lvds *lvds, unsigned int freq)
 		rcar_lvds_write(lvds, LVDDIV, 0);
 }
 
+static void rcar_lvds_pll_setup_d3_e3(struct rcar_lvds *lvds, unsigned int freq)
+{
+	__rcar_lvds_pll_setup_d3_e3(lvds, freq, false);
+}
+
+/* -----------------------------------------------------------------------------
+ * Clock - D3/E3 only
+ */
+
+int rcar_lvds_clk_enable(struct drm_bridge *bridge, unsigned long freq)
+{
+	struct rcar_lvds *lvds = bridge_to_rcar_lvds(bridge);
+	int ret;
+
+	if (WARN_ON(!(lvds->info->quirks & RCAR_LVDS_QUIRK_EXT_PLL)))
+		return -ENODEV;
+
+	dev_dbg(lvds->dev, "enabling LVDS PLL, freq=%luHz\n", freq);
+
+	WARN_ON(lvds->enabled);
+
+	ret = clk_prepare_enable(lvds->clocks.mod);
+	if (ret < 0)
+		return ret;
+
+	__rcar_lvds_pll_setup_d3_e3(lvds, freq, true);
+
+	lvds->enabled = true;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rcar_lvds_clk_enable);
+
+void rcar_lvds_clk_disable(struct drm_bridge *bridge)
+{
+	struct rcar_lvds *lvds = bridge_to_rcar_lvds(bridge);
+
+	if (WARN_ON(!(lvds->info->quirks & RCAR_LVDS_QUIRK_EXT_PLL)))
+		return;
+
+	dev_dbg(lvds->dev, "disabling LVDS PLL\n");
+
+	WARN_ON(!lvds->enabled);
+
+	rcar_lvds_write(lvds, LVDPLLCR, 0);
+
+	clk_disable_unprepare(lvds->clocks.mod);
+
+	lvds->enabled = false;
+}
+EXPORT_SYMBOL_GPL(rcar_lvds_clk_disable);
+
 /* -----------------------------------------------------------------------------
  * Bridge
  */
diff --git a/drivers/gpu/drm/rcar-du/rcar_lvds.h b/drivers/gpu/drm/rcar-du/rcar_lvds.h
new file mode 100644
index 000000000000..a709cae1bc32
--- /dev/null
+++ b/drivers/gpu/drm/rcar-du/rcar_lvds.h
@@ -0,0 +1,27 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * rcar_lvds.h  --  R-Car LVDS Encoder
+ *
+ * Copyright (C) 2013-2018 Renesas Electronics Corporation
+ *
+ * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com)
+ */
+
+#ifndef __RCAR_LVDS_H__
+#define __RCAR_LVDS_H__
+
+struct drm_bridge;
+
+#if IS_ENABLED(CONFIG_DRM_RCAR_LVDS)
+int rcar_lvds_clk_enable(struct drm_bridge *bridge, unsigned long freq);
+void rcar_lvds_clk_disable(struct drm_bridge *bridge);
+#else
+static inline int rcar_lvds_clk_enable(struct drm_bridge *bridge,
+				       unsigned long freq)
+{
+	return -ENOSYS;
+}
+static inline void rcar_lvds_clk_disable(struct drm_bridge *bridge) { }
+#endif /* CONFIG_DRM_RCAR_LVDS */
+
+#endif /* __RCAR_LVDS_H__ */
-- 
Regards,

Laurent Pinchart

[PATCH v2 4/6] drm: rcar-du: Turn LVDS clock output on/off for DPAD0 output on D3/E3

[Laurent Pinchart]

On the D3 and E3 SoCs the LVDS PLL clock output provides the dot clock
to the DU channels, even when the LVDS outputs are not in use. Enable
and disable the LVDS clock output when enabling or disabling a CRTC
connected to the DPAD0 output.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
 drivers/gpu/drm/rcar-du/rcar_du_crtc.c    | 34 +++++++++++++++++++++++
 drivers/gpu/drm/rcar-du/rcar_du_drv.h     |  3 ++
 drivers/gpu/drm/rcar-du/rcar_du_encoder.c |  1 +
 3 files changed, 38 insertions(+)

diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index 771b460c7216..3018e4500aa1 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -26,6 +26,7 @@
 #include "rcar_du_plane.h"
 #include "rcar_du_regs.h"
 #include "rcar_du_vsp.h"
+#include "rcar_lvds.h"
 
 static u32 rcar_du_crtc_read(struct rcar_du_crtc *rcrtc, u32 reg)
 {
@@ -657,8 +658,27 @@ static void rcar_du_crtc_atomic_enable(struct drm_crtc *crtc,
 				       struct drm_crtc_state *old_state)
 {
 	struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
+	struct rcar_du_crtc_state *rstate = to_rcar_crtc_state(crtc->state);
+	struct rcar_du_device *rcdu = rcrtc->group->dev;
 
 	rcar_du_crtc_get(rcrtc);
+
+	/*
+	 * On D3/E3 the dot clock is provided by the LVDS encoder attached to
+	 * the DU channel. We need to enable its clock output explicitly if
+	 * the LVDS output is disabled.
+	 */
+	if (rcdu->info->lvds_clk_mask & BIT(rcrtc->index) &&
+	    rstate->outputs == BIT(RCAR_DU_OUTPUT_DPAD0)) {
+		struct rcar_du_encoder *encoder =
+			rcdu->encoders[RCAR_DU_OUTPUT_LVDS0 + rcrtc->index];
+		const struct drm_display_mode *mode =
+			&crtc->state->adjusted_mode;
+
+		rcar_lvds_clk_enable(encoder->base.bridge,
+				     mode->clock * 1000);
+	}
+
 	rcar_du_crtc_start(rcrtc);
 }
 
@@ -666,10 +686,24 @@ static void rcar_du_crtc_atomic_disable(struct drm_crtc *crtc,
 					struct drm_crtc_state *old_state)
 {
 	struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
+	struct rcar_du_crtc_state *rstate = to_rcar_crtc_state(old_state);
+	struct rcar_du_device *rcdu = rcrtc->group->dev;
 
 	rcar_du_crtc_stop(rcrtc);
 	rcar_du_crtc_put(rcrtc);
 
+	if (rcdu->info->lvds_clk_mask & BIT(rcrtc->index) &&
+	    rstate->outputs == BIT(RCAR_DU_OUTPUT_DPAD0)) {
+		struct rcar_du_encoder *encoder =
+			rcdu->encoders[RCAR_DU_OUTPUT_LVDS0 + rcrtc->index];
+
+		/*
+		 * Disable the LVDS clock output, see
+		 * rcar_du_crtc_atomic_enable().
+		 */
+		rcar_lvds_clk_disable(encoder->base.bridge);
+	}
+
 	spin_lock_irq(&crtc->dev->event_lock);
 	if (crtc->state->event) {
 		drm_crtc_send_vblank_event(crtc, crtc->state->event);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index 6c187d0bf7c2..1327cd0df90a 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -22,6 +22,7 @@ struct device;
 struct drm_device;
 struct drm_property;
 struct rcar_du_device;
+struct rcar_du_encoder;
 
 #define RCAR_DU_FEATURE_CRTC_IRQ_CLOCK	BIT(0)	/* Per-CRTC IRQ and clock */
 #define RCAR_DU_FEATURE_VSP1_SOURCE	BIT(1)	/* Has inputs from VSP1 */
@@ -81,6 +82,8 @@ struct rcar_du_device {
 	struct rcar_du_crtc crtcs[RCAR_DU_MAX_CRTCS];
 	unsigned int num_crtcs;
 
+	struct rcar_du_encoder *encoders[RCAR_DU_OUTPUT_MAX];
+
 	struct rcar_du_group groups[RCAR_DU_MAX_GROUPS];
 	struct rcar_du_vsp vsps[RCAR_DU_MAX_VSPS];
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
index ba2b38f76b54..13c3c7382514 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c
@@ -41,6 +41,7 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu,
 	if (renc == NULL)
 		return -ENOMEM;
 
+	rcdu->encoders[output] = renc;
 	renc->output = output;
 	encoder = rcar_encoder_to_drm_encoder(renc);
 
-- 
Regards,

Laurent Pinchart

[PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Laurent Pinchart]

The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
output when the LVDS0 encoder is used. Enable it despite its output not
being connected.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
Changes since v1:

- Add a comment in the DT to explain why the LVDS1 encoder needs to be
  enabled.
---
 arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts b/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts
index 62bdddcbbae7..800a537bbbe0 100644
--- a/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts
+++ b/arch/arm64/boot/dts/renesas/r8a77990-ebisu.dts
@@ -443,6 +443,13 @@
 };
 
 &lvds1 {
+	/*
+	 * Even though the LVDS1 output is not connected, the encoder must be
+	 * enabled to supply a pixel clock to the DU for the DPAD output when
+	 * LVDS0 is in use.
+	 */
+	status = "okay";
+
 	clocks = <&cpg CPG_MOD 727>,
 		 <&x13_clk>,
 		 <&extal_clk>;
-- 
Regards,

Laurent Pinchart

[PATCH v2 6/6] arm64: dts: renesas: r8a77995: draak: Enable LVDS1 encoder

[Laurent Pinchart]

The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
output when the LVDS0 encoder is used. Enable it despite its output not
being connected.

Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
---
Changes since v1:

- Add a comment in the DT to explain why the LVDS1 encoder needs to be
  enabled.
---
 arch/arm64/boot/dts/renesas/r8a77995-draak.dts | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/arch/arm64/boot/dts/renesas/r8a77995-draak.dts b/arch/arm64/boot/dts/renesas/r8a77995-draak.dts
index 89df9bc844c0..3c9ec67e342b 100644
--- a/arch/arm64/boot/dts/renesas/r8a77995-draak.dts
+++ b/arch/arm64/boot/dts/renesas/r8a77995-draak.dts
@@ -356,6 +356,13 @@
 };
 
 &lvds1 {
+	/*
+	 * Even though the LVDS1 output is not connected, the encoder must be
+	 * enabled to supply a pixel clock to the DU for the DPAD output when
+	 * LVDS0 is in use.
+	 */
+	status = "okay";
+
 	clocks = <&cpg CPG_MOD 727>,
 		 <&x12_clk>,
 		 <&extal_clk>;
-- 
Regards,

Laurent Pinchart

Re: [PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Simon Horman]

On Wed, Jan 23, 2019 at 12:54:04AM +0200, Laurent Pinchart wrote:
> The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> output when the LVDS0 encoder is used. Enable it despite its output not
> being connected.
> 
> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> ---
> Changes since v1:
> 
> - Add a comment in the DT to explain why the LVDS1 encoder needs to be
>   enabled.

Thanks,

This looks fine to me but I will wait to see if there are other reviews
before applying.

Reviewed-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v2 6/6] arm64: dts: renesas: r8a77995: draak: Enable LVDS1 encoder

[Simon Horman]

On Wed, Jan 23, 2019 at 12:54:05AM +0200, Laurent Pinchart wrote:
> The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> output when the LVDS0 encoder is used. Enable it despite its output not
> being connected.
> 
> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> ---
> Changes since v1:
> 
> - Add a comment in the DT to explain why the LVDS1 encoder needs to be
>   enabled.

Thanks,

This looks fine to me but I will wait to see if there are other reviews
before applying.

Reviewed-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Laurent Pinchart]

Hi Simon,

On Wed, Jan 23, 2019 at 09:56:57AM +0100, Simon Horman wrote:
> On Wed, Jan 23, 2019 at 12:54:04AM +0200, Laurent Pinchart wrote:
> > The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> > output when the LVDS0 encoder is used. Enable it despite its output not
> > being connected.
> > 
> > Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > ---
> > Changes since v1:
> > 
> > - Add a comment in the DT to explain why the LVDS1 encoder needs to be
> >   enabled.
> 
> Thanks,
> 
> This looks fine to me but I will wait to see if there are other reviews
> before applying.
> 
> Reviewed-by: Simon Horman <horms+renesas@verge.net.au>

Please note that this will likely cause probe failures if applied
without the driver patches from this series. Would it be OK to merge the
DT changes through the DRM tree ?

-- 
Regards,

Laurent Pinchart

Re: [PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Simon Horman]

Hi Laurent,

On Wed, Jan 23, 2019 at 11:55:52AM +0200, Laurent Pinchart wrote:
> Hi Simon,
> 
> On Wed, Jan 23, 2019 at 09:56:57AM +0100, Simon Horman wrote:
> > On Wed, Jan 23, 2019 at 12:54:04AM +0200, Laurent Pinchart wrote:
> > > The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> > > output when the LVDS0 encoder is used. Enable it despite its output not
> > > being connected.
> > > 
> > > Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > > ---
> > > Changes since v1:
> > > 
> > > - Add a comment in the DT to explain why the LVDS1 encoder needs to be
> > >   enabled.
> > 
> > Thanks,
> > 
> > This looks fine to me but I will wait to see if there are other reviews
> > before applying.
> > 
> > Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
> 
> Please note that this will likely cause probe failures if applied
> without the driver patches from this series. Would it be OK to merge the
> DT changes through the DRM tree ?

I'm not sure that the probe failures are a problem unless they move
something that was in a working state to a non-working state. Nonetheless
they are at the very least not very nice.

I generally shy away from having DT patches merged in other trees
to avoid the chance of merge conflicts (that need to be resolved by Linus).
What is your feeling on the risk of such a conflict?

Re: [PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Laurent Pinchart]

Hi Simon,

On Wed, Jan 23, 2019 at 11:03:04AM +0100, Simon Horman wrote:
> On Wed, Jan 23, 2019 at 11:55:52AM +0200, Laurent Pinchart wrote:
> > On Wed, Jan 23, 2019 at 09:56:57AM +0100, Simon Horman wrote:
> >> On Wed, Jan 23, 2019 at 12:54:04AM +0200, Laurent Pinchart wrote:
> >>> The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> >>> output when the LVDS0 encoder is used. Enable it despite its output not
> >>> being connected.
> >>> 
> >>> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> >>> ---
> >>> Changes since v1:
> >>> 
> >>> - Add a comment in the DT to explain why the LVDS1 encoder needs to be
> >>>   enabled.
> >> 
> >> Thanks,
> >> 
> >> This looks fine to me but I will wait to see if there are other reviews
> >> before applying.
> >> 
> >> Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
> > 
> > Please note that this will likely cause probe failures if applied
> > without the driver patches from this series. Would it be OK to merge the
> > DT changes through the DRM tree ?
> 
> I'm not sure that the probe failures are a problem unless they move
> something that was in a working state to a non-working state. Nonetheless
> they are at the very least not very nice.
> 
> I generally shy away from having DT patches merged in other trees
> to avoid the chance of merge conflicts (that need to be resolved by Linus).
> What is your feeling on the risk of such a conflict?

I would have said low, but given that it's too late to get the DT
changes in v5.1 anyway, it's no big deal. The driver patches have been
merged for v5.1, so could you queue the DT changes for v5.2 ?

-- 
Regards,

Laurent Pinchart

Re: [PATCH v2 0/6] R-Car DU DPAD support for D3 and E3

[Kuninori Morimoto]

Hi Laurent

> For your convenience the patches are available from
> 
>         git://linuxtv.org/pinchartl/media.git drm/du/d3e3

It seems we can't find this branch/tag on your branch
Is it removed ? or not yet pushed ?

	https://git.linuxtv.org/pinchartl/media.git/refs/heads

Best regards
---
Kuninori Morimoto

Re: [PATCH v2 0/6] R-Car DU DPAD support for D3 and E3

[Laurent Pinchart]

Hi Morimoto-san,

On Mon, Feb 25, 2019 at 04:44:55PM +0900, Kuninori Morimoto wrote:
> 
> Hi Laurent
> 
> > For your convenience the patches are available from
> > 
> >         git://linuxtv.org/pinchartl/media.git drm/du/d3e3
> 
> It seems we can't find this branch/tag on your branch
> Is it removed ? or not yet pushed ?
> 
> 	https://git.linuxtv.org/pinchartl/media.git/refs/heads

The branch has been merged to drm-next (for the DU driver part) and to
Simon's next branch (for the DT part), so I've removed it from my tree.
You can find a merge of those two branches in my repository in the
drm/du/base branch.

-- 
Regards,

Laurent Pinchart

Re: [PATCH v2 0/6] R-Car DU DPAD support for D3 and E3

[Kuninori Morimoto]

Hi Laurent

> > > For your convenience the patches are available from
> > > 
> > >         git://linuxtv.org/pinchartl/media.git drm/du/d3e3
> > 
> > It seems we can't find this branch/tag on your branch
> > Is it removed ? or not yet pushed ?
> > 
> > 	https://git.linuxtv.org/pinchartl/media.git/refs/heads
> 
> The branch has been merged to drm-next (for the DU driver part) and to
> Simon's next branch (for the DT part), so I've removed it from my tree.
> You can find a merge of those two branches in my repository in the
> drm/du/base branch.

OK, Thanks.
Nice to know


Best regards
---
Kuninori Morimoto

Re: [PATCH v2 5/6] arm64: dts: renesas: r8a77990: ebisu: Enable LVDS1 encoder

[Simon Horman]

On Fri, Feb 15, 2019 at 01:44:43PM +0200, Laurent Pinchart wrote:
> Hi Simon,
> 
> On Wed, Jan 23, 2019 at 11:03:04AM +0100, Simon Horman wrote:
> > On Wed, Jan 23, 2019 at 11:55:52AM +0200, Laurent Pinchart wrote:
> > > On Wed, Jan 23, 2019 at 09:56:57AM +0100, Simon Horman wrote:
> > >> On Wed, Jan 23, 2019 at 12:54:04AM +0200, Laurent Pinchart wrote:
> > >>> The LVDS1 encoder must supply a pixel clock to the DU for the DPAD
> > >>> output when the LVDS0 encoder is used. Enable it despite its output not
> > >>> being connected.
> > >>> 
> > >>> Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > >>> ---
> > >>> Changes since v1:
> > >>> 
> > >>> - Add a comment in the DT to explain why the LVDS1 encoder needs to be
> > >>>   enabled.
> > >> 
> > >> Thanks,
> > >> 
> > >> This looks fine to me but I will wait to see if there are other reviews
> > >> before applying.
> > >> 
> > >> Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
> > > 
> > > Please note that this will likely cause probe failures if applied
> > > without the driver patches from this series. Would it be OK to merge the
> > > DT changes through the DRM tree ?
> > 
> > I'm not sure that the probe failures are a problem unless they move
> > something that was in a working state to a non-working state. Nonetheless
> > they are at the very least not very nice.
> > 
> > I generally shy away from having DT patches merged in other trees
> > to avoid the chance of merge conflicts (that need to be resolved by Linus).
> > What is your feeling on the risk of such a conflict?
> 
> I would have said low, but given that it's too late to get the DT
> changes in v5.1 anyway, it's no big deal. The driver patches have been
> merged for v5.1, so could you queue the DT changes for v5.2 ?

Thanks, I've applied patches 5 and 6 of this series to my
renesas tree for inclusion in v5.2.