[PATCH 0/7] drm/bridge: tc358768: Long command support

[PATCH 0/7] drm/bridge: tc358768: Long command support

[Tomi Valkeinen]

This series makes some small improvements to the tc358768 driver, and
then adds support for long commands, commands that have more than 8
bytes of payload.

This has been tested on Toradex Verdin AM62 board, with a ST7703 based
DSI panel which requires initial configuration using commands that have
8+ bytes.

I tested the following combinations, by observing the panel visually and
the DSI lanes with an oscilloscope:

- HS command transfer, continuous clock
- HS command transfer, non-continuous clock
- LP command transfer, continuous clock
- LP command transfer, non-continuous clock

All except LP + continuous clock work fine. While observing the lanes
with normal oscilloscope is a very high level and vague view of what
exactly is going on the lanes, it still looks good to me: initial
commands are sent in LP, and then video data is being sent in HS, and
clock lane is continuous. However, the panel stays black, so something
is not quite right.

As it doesn't make sense to send commands in LP during initial
configuration, especially with longer commands, I will leave that
particular combination unresolved. I'd be interested to hear if it works
on some other panel.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
Parth Pancholi (1):
      drm/bridge: tc358768: Set pre_enable_prev_first for reverse order

Tomi Valkeinen (6):
      drm/bridge: tc358768: Fix typo in TC358768_DSI_CONTROL_DIS_MODE
      drm/bridge: tc358768: Separate indirect register writes
      drm/bridge: tc358768: Support non-continuous clock
      drm/bridge: tc358768: Add LP mode command support
      drm/bridge: tc358768: Separate video format config
      drm/bridge: tc358768: Add support for long command tx via video buffer

 drivers/gpu/drm/bridge/tc358768.c | 202 +++++++++++++++++++++++++++++---------
 1 file changed, 157 insertions(+), 45 deletions(-)
---
base-commit: e5f0a698b34ed76002dc5cff3804a61c80233a7a
change-id: 20251021-tc358768-93eb23689b7a

Best regards,
-- 
Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

[PATCH 1/7] drm/bridge: tc358768: Fix typo in TC358768_DSI_CONTROL_DIS_MODE

[Tomi Valkeinen]

It's "DSI_MODE", not "DIS_MODE".

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index fbdc44e16229..c95d164bd3a0 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -115,7 +115,7 @@
 #define TC358768_DSI_HACT		0x062C
 
 /* TC358768_DSI_CONTROL (0x040C) register */
-#define TC358768_DSI_CONTROL_DIS_MODE	BIT(15)
+#define TC358768_DSI_CONTROL_DSI_MODE	BIT(15)
 #define TC358768_DSI_CONTROL_TXMD	BIT(7)
 #define TC358768_DSI_CONTROL_HSCKMD	BIT(5)
 #define TC358768_DSI_CONTROL_EOTDIS	BIT(0)
@@ -1082,7 +1082,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	tc358768_write(priv, TC358768_DSI_CONFW, val);
 
 	val = TC358768_DSI_CONFW_MODE_CLR | TC358768_DSI_CONFW_ADDR_DSI_CONTROL;
-	val |= TC358768_DSI_CONTROL_DIS_MODE; /* DSI mode */
+	val |= TC358768_DSI_CONTROL_DSI_MODE;
 	tc358768_write(priv, TC358768_DSI_CONFW, val);
 
 	ret = tc358768_clear_error(priv);

-- 
2.43.0

Re: [PATCH 1/7] drm/bridge: tc358768: Fix typo in TC358768_DSI_CONTROL_DIS_MODE

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:22:57PM +0300, Tomi Valkeinen wrote:
> It's "DSI_MODE", not "DIS_MODE".
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 2/7] drm/bridge: tc358768: Set pre_enable_prev_first for reverse order

[Tomi Valkeinen]

From: Parth Pancholi <parth.pancholi@toradex.com>

Enable the pre_enable_prev_first flag on the tc358768 bridge to reverse
the pre-enable order, calling bridge pre_enable before panel prepare.
This ensures the bridge is ready before sending panel init commands in
the case of panels sending init commands in panel prepare function.

Signed-off-by: Parth Pancholi <parth.pancholi@toradex.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index c95d164bd3a0..dab9cdf5cb98 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -448,6 +448,8 @@ static int tc358768_dsi_host_attach(struct mipi_dsi_host *host,
 						    DRM_MODE_CONNECTOR_DSI);
 		if (IS_ERR(bridge))
 			return PTR_ERR(bridge);
+
+		bridge->pre_enable_prev_first = true;
 	}
 
 	priv->output.dev = dev;

-- 
2.43.0

Re: [PATCH 2/7] drm/bridge: tc358768: Set pre_enable_prev_first for reverse order

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:22:58PM +0300, Tomi Valkeinen wrote:
> From: Parth Pancholi <parth.pancholi@toradex.com>
> 
> Enable the pre_enable_prev_first flag on the tc358768 bridge to reverse
> the pre-enable order, calling bridge pre_enable before panel prepare.
> This ensures the bridge is ready before sending panel init commands in
> the case of panels sending init commands in panel prepare function.
> 
> Signed-off-by: Parth Pancholi <parth.pancholi@toradex.com>
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 3/7] drm/bridge: tc358768: Separate indirect register writes

[Tomi Valkeinen]

Some registers can only be written indirectly, using DSI_CONFW register.
We don't have many uses for those registers (in fact, only DSI_CONTROL
is currently written), but the code to do those writes inline is a bit
confusing.

Add a new function, tc358768_confw_update_bits() which can be used to
write the bits indirectly. Only DSI_CONTROL is currently supported.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 52 +++++++++++++++++++++++++++++----------
 1 file changed, 39 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index dab9cdf5cb98..755ed6483b2e 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -123,7 +123,7 @@
 /* TC358768_DSI_CONFW (0x0500) register */
 #define TC358768_DSI_CONFW_MODE_SET	(5 << 29)
 #define TC358768_DSI_CONFW_MODE_CLR	(6 << 29)
-#define TC358768_DSI_CONFW_ADDR_DSI_CONTROL	(0x3 << 24)
+#define TC358768_DSI_CONFW_ADDR(x)	((x) << 24)
 
 /* TC358768_DSICMD_TX (0x0600) register */
 #define TC358768_DSI_CMDTX_DC_START	BIT(0)
@@ -232,6 +232,36 @@ static void tc358768_update_bits(struct tc358768_priv *priv, u32 reg, u32 mask,
 		tc358768_write(priv, reg, tmp);
 }
 
+static void tc358768_confw_update_bits(struct tc358768_priv *priv, u16 reg,
+				       u16 mask, u16 val)
+{
+	u8 confw_addr;
+	u32 confw_val;
+
+	switch (reg) {
+	case TC358768_DSI_CONTROL:
+		confw_addr = 0x3;
+		break;
+	default:
+		priv->error = -EINVAL;
+		return;
+	}
+
+	if (mask != val) {
+		confw_val = TC358768_DSI_CONFW_MODE_CLR |
+			TC358768_DSI_CONFW_ADDR(confw_addr) |
+			mask;
+		tc358768_write(priv, TC358768_DSI_CONFW, confw_val);
+	}
+
+	if (val & mask) {
+		confw_val = TC358768_DSI_CONFW_MODE_SET |
+			TC358768_DSI_CONFW_ADDR(confw_addr) |
+			(val & mask);
+		tc358768_write(priv, TC358768_DSI_CONFW, confw_val);
+	}
+}
+
 static void tc358768_dsicmd_tx(struct tc358768_priv *priv)
 {
 	u32 val;
@@ -693,7 +723,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	struct tc358768_priv *priv = bridge_to_tc358768(bridge);
 	struct mipi_dsi_device *dsi_dev = priv->output.dev;
 	unsigned long mode_flags = dsi_dev->mode_flags;
-	u32 val, val2, lptxcnt, hact, data_type;
+	u32 val, mask, val2, lptxcnt, hact, data_type;
 	s32 raw_val;
 	struct drm_crtc_state *crtc_state;
 	struct drm_connector_state *conn_state;
@@ -1065,13 +1095,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	tc358768_write(priv, TC358768_DSI_START, 0x1);
 
 	/* Configure DSI_Control register */
-	val = TC358768_DSI_CONFW_MODE_CLR | TC358768_DSI_CONFW_ADDR_DSI_CONTROL;
-	val |= TC358768_DSI_CONTROL_TXMD | TC358768_DSI_CONTROL_HSCKMD |
-	       0x3 << 1 | TC358768_DSI_CONTROL_EOTDIS;
-	tc358768_write(priv, TC358768_DSI_CONFW, val);
-
-	val = TC358768_DSI_CONFW_MODE_SET | TC358768_DSI_CONFW_ADDR_DSI_CONTROL;
-	val |= (dsi_dev->lanes - 1) << 1;
+	val = (dsi_dev->lanes - 1) << 1;
 
 	val |= TC358768_DSI_CONTROL_TXMD;
 
@@ -1081,11 +1105,13 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	if (dsi_dev->mode_flags & MIPI_DSI_MODE_NO_EOT_PACKET)
 		val |= TC358768_DSI_CONTROL_EOTDIS;
 
-	tc358768_write(priv, TC358768_DSI_CONFW, val);
+	mask = TC358768_DSI_CONTROL_TXMD | TC358768_DSI_CONTROL_HSCKMD |
+	       0x3 << 1 | TC358768_DSI_CONTROL_EOTDIS;
+
+	tc358768_confw_update_bits(priv, TC358768_DSI_CONTROL, mask, val);
 
-	val = TC358768_DSI_CONFW_MODE_CLR | TC358768_DSI_CONFW_ADDR_DSI_CONTROL;
-	val |= TC358768_DSI_CONTROL_DSI_MODE;
-	tc358768_write(priv, TC358768_DSI_CONFW, val);
+	tc358768_confw_update_bits(priv, TC358768_DSI_CONTROL,
+				   TC358768_DSI_CONTROL_DSI_MODE, 0);
 
 	ret = tc358768_clear_error(priv);
 	if (ret)

-- 
2.43.0

Re: [PATCH 3/7] drm/bridge: tc358768: Separate indirect register writes

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:22:59PM +0300, Tomi Valkeinen wrote:
> Some registers can only be written indirectly, using DSI_CONFW register.
> We don't have many uses for those registers (in fact, only DSI_CONTROL
> is currently written), but the code to do those writes inline is a bit
> confusing.
> 
> Add a new function, tc358768_confw_update_bits() which can be used to
> write the bits indirectly. Only DSI_CONTROL is currently supported.
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 4/7] drm/bridge: tc358768: Support non-continuous clock

[Tomi Valkeinen]

The driver prints a warning if MIPI_DSI_CLOCK_NON_CONTINUOUS is set, and
falls back to continuous clock mode. This was added in commit
fbc5a90e82c1 ("drm/bridge: tc358768: Disable non-continuous clock mode").

However, there have been multiple changes to the driver since then, and
at least in my setup, non-continuous clock mode works: I can see an
image on the panel, and I can see the clock lanes being non-continuous
with an oscilloscope.

So, let's enable MIPI_DSI_CLOCK_NON_CONTINUOUS support.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Cc: Dmitry Osipenko <digetx@gmail.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index 755ed6483b2e..a276fbc75dde 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -722,7 +722,6 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 {
 	struct tc358768_priv *priv = bridge_to_tc358768(bridge);
 	struct mipi_dsi_device *dsi_dev = priv->output.dev;
-	unsigned long mode_flags = dsi_dev->mode_flags;
 	u32 val, mask, val2, lptxcnt, hact, data_type;
 	s32 raw_val;
 	struct drm_crtc_state *crtc_state;
@@ -744,11 +743,6 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	u32 dsi_vsdly;
 	const u32 internal_dly = 40;
 
-	if (mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) {
-		dev_warn_once(dev, "Non-continuous mode unimplemented, falling back to continuous\n");
-		mode_flags &= ~MIPI_DSI_CLOCK_NON_CONTINUOUS;
-	}
-
 	tc358768_hw_enable(priv);
 
 	ret = tc358768_sw_reset(priv);
@@ -1032,7 +1026,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	tc358768_write(priv, TC358768_HSTXVREGEN, val);
 
 	tc358768_write(priv, TC358768_TXOPTIONCNTRL,
-		       (mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) ? 0 : BIT(0));
+		       (dsi_dev->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) ? 0 : BIT(0));
 
 	/* TXTAGOCNT[26:16] RXTASURECNT[10:0] */
 	val = tc358768_ps_to_ns((lptxcnt + 1) * hsbyteclk_ps * 4);
@@ -1099,7 +1093,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 
 	val |= TC358768_DSI_CONTROL_TXMD;
 
-	if (!(mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS))
+	if (!(dsi_dev->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS))
 		val |= TC358768_DSI_CONTROL_HSCKMD;
 
 	if (dsi_dev->mode_flags & MIPI_DSI_MODE_NO_EOT_PACKET)

-- 
2.43.0

Re: [PATCH 4/7] drm/bridge: tc358768: Support non-continuous clock

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:23:00PM +0300, Tomi Valkeinen wrote:
> The driver prints a warning if MIPI_DSI_CLOCK_NON_CONTINUOUS is set, and
> falls back to continuous clock mode. This was added in commit
> fbc5a90e82c1 ("drm/bridge: tc358768: Disable non-continuous clock mode").
> 
> However, there have been multiple changes to the driver since then, and
> at least in my setup, non-continuous clock mode works: I can see an
> image on the panel, and I can see the clock lanes being non-continuous
> with an oscilloscope.
> 
> So, let's enable MIPI_DSI_CLOCK_NON_CONTINUOUS support.
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
> Cc: Dmitry Osipenko <digetx@gmail.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 5/7] drm/bridge: tc358768: Add LP mode command support

[Tomi Valkeinen]

Currently the driver ignores MIPI_DSI_MODE_LPM and always uses HS mode.
Add code to enable HS mode in pre_enable() only if MIPI_DSI_MODE_LPM is
not set, and always enable HS mode in enable() for video transmission.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index a276fbc75dde..a7a14c125ac4 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -1091,7 +1091,8 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	/* Configure DSI_Control register */
 	val = (dsi_dev->lanes - 1) << 1;
 
-	val |= TC358768_DSI_CONTROL_TXMD;
+	if (!(dsi_dev->mode_flags & MIPI_DSI_MODE_LPM))
+		val |= TC358768_DSI_CONTROL_TXMD;
 
 	if (!(dsi_dev->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS))
 		val |= TC358768_DSI_CONTROL_HSCKMD;
@@ -1123,6 +1124,11 @@ static void tc358768_bridge_atomic_enable(struct drm_bridge *bridge,
 		return;
 	}
 
+	/* Enable HS mode for video TX */
+	tc358768_confw_update_bits(priv, TC358768_DSI_CONTROL,
+				   TC358768_DSI_CONTROL_TXMD,
+				   TC358768_DSI_CONTROL_TXMD);
+
 	/* clear FrmStop and RstPtr */
 	tc358768_update_bits(priv, TC358768_PP_MISC, 0x3 << 14, 0);
 

-- 
2.43.0

Re: [PATCH 5/7] drm/bridge: tc358768: Add LP mode command support

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:23:01PM +0300, Tomi Valkeinen wrote:
> Currently the driver ignores MIPI_DSI_MODE_LPM and always uses HS mode.
> Add code to enable HS mode in pre_enable() only if MIPI_DSI_MODE_LPM is
> not set, and always enable HS mode in enable() for video transmission.
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 6/7] drm/bridge: tc358768: Separate video format config

[Tomi Valkeinen]

Sending long commands using the video buffer (to be implemented in
following patches) requires setting TC358768_DATAFMT and
TC358768_DSITX_DT registers for command transfer. The same registers
also need to be configured properly for video transfer.

The long commands will be sent between the bridge's pre_enable() and
enable(), and currently we configure the registers for video transfer in
pre_enable(). Thus, they would be overwritten by the long command
transfer code.

To prevent that from happening, set those registers for video transfer
in enable(), not in pre_enable().

Based on code from Parth Pancholi <parth.pancholi@toradex.com>

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 51 ++++++++++++++++++++++++++++-----------
 1 file changed, 37 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index a7a14c125ac4..e1ed4003b3c5 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -722,7 +722,7 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 {
 	struct tc358768_priv *priv = bridge_to_tc358768(bridge);
 	struct mipi_dsi_device *dsi_dev = priv->output.dev;
-	u32 val, mask, val2, lptxcnt, hact, data_type;
+	u32 val, mask, val2, lptxcnt, hact;
 	s32 raw_val;
 	struct drm_crtc_state *crtc_state;
 	struct drm_connector_state *conn_state;
@@ -768,30 +768,20 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	dsiclk = priv->dsiclk;
 	hsbyteclk = dsiclk / 4;
 
-	/* Data Format Control Register */
-	val = BIT(2) | BIT(1) | BIT(0); /* rdswap_en | dsitx_en | txdt_en */
 	switch (dsi_dev->format) {
 	case MIPI_DSI_FMT_RGB888:
-		val |= (0x3 << 4);
 		hact = vm.hactive * 3;
-		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_24;
 		break;
 	case MIPI_DSI_FMT_RGB666:
-		val |= (0x4 << 4);
 		hact = vm.hactive * 3;
-		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_18;
 		break;
 
 	case MIPI_DSI_FMT_RGB666_PACKED:
-		val |= (0x4 << 4) | BIT(3);
 		hact = vm.hactive * 18 / 8;
-		data_type = MIPI_DSI_PIXEL_STREAM_3BYTE_18;
 		break;
 
 	case MIPI_DSI_FMT_RGB565:
-		val |= (0x5 << 4);
 		hact = vm.hactive * 2;
-		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_16;
 		break;
 	default:
 		dev_err(dev, "Invalid data format (%u)\n",
@@ -947,9 +937,6 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 	/* VSDly[9:0] */
 	tc358768_write(priv, TC358768_VSDLY, dsi_vsdly - internal_dly);
 
-	tc358768_write(priv, TC358768_DATAFMT, val);
-	tc358768_write(priv, TC358768_DSITX_DT, data_type);
-
 	/* Enable D-PHY (HiZ->LP11) */
 	tc358768_write(priv, TC358768_CLW_CNTRL, 0x0000);
 	/* Enable lanes */
@@ -1113,6 +1100,39 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 		dev_err(dev, "Bridge pre_enable failed: %d\n", ret);
 }
 
+static void tc358768_config_video_format(struct tc358768_priv *priv)
+{
+	struct mipi_dsi_device *dsi_dev = priv->output.dev;
+	u32 val, data_type;
+
+	/* Data Format Control Register */
+	val = BIT(2) | BIT(1) | BIT(0); /* rdswap_en | dsitx_en | txdt_en */
+	switch (dsi_dev->format) {
+	case MIPI_DSI_FMT_RGB888:
+		val |= (0x3 << 4);
+		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_24;
+		break;
+	case MIPI_DSI_FMT_RGB666:
+		val |= (0x4 << 4);
+		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_18;
+		break;
+	case MIPI_DSI_FMT_RGB666_PACKED:
+		val |= (0x4 << 4) | BIT(3);
+		data_type = MIPI_DSI_PIXEL_STREAM_3BYTE_18;
+		break;
+	case MIPI_DSI_FMT_RGB565:
+		val |= (0x5 << 4);
+		data_type = MIPI_DSI_PACKED_PIXEL_STREAM_16;
+		break;
+	default:
+		dev_err(priv->dev, "Invalid data format (%u)\n", dsi_dev->format);
+		return;
+	}
+
+	tc358768_write(priv, TC358768_DATAFMT, val);
+	tc358768_write(priv, TC358768_DSITX_DT, data_type);
+}
+
 static void tc358768_bridge_atomic_enable(struct drm_bridge *bridge,
 					  struct drm_atomic_state *state)
 {
@@ -1124,6 +1144,9 @@ static void tc358768_bridge_atomic_enable(struct drm_bridge *bridge,
 		return;
 	}
 
+	/* Configure video format registers */
+	tc358768_config_video_format(priv);
+
 	/* Enable HS mode for video TX */
 	tc358768_confw_update_bits(priv, TC358768_DSI_CONTROL,
 				   TC358768_DSI_CONTROL_TXMD,

-- 
2.43.0

Re: [PATCH 6/7] drm/bridge: tc358768: Separate video format config

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:23:02PM +0300, Tomi Valkeinen wrote:
> Sending long commands using the video buffer (to be implemented in
> following patches) requires setting TC358768_DATAFMT and
> TC358768_DSITX_DT registers for command transfer. The same registers
> also need to be configured properly for video transfer.
> 
> The long commands will be sent between the bridge's pre_enable() and
> enable(), and currently we configure the registers for video transfer in
> pre_enable(). Thus, they would be overwritten by the long command
> transfer code.
> 
> To prevent that from happening, set those registers for video transfer
> in enable(), not in pre_enable().
> 
> Based on code from Parth Pancholi <parth.pancholi@toradex.com>
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

[PATCH 7/7] drm/bridge: tc358768: Add support for long command tx via video buffer

[Tomi Valkeinen]

TC358768 has two ways to send DSI commands: 1) buffer the payload data
into registers (DSICMD_WDx), which supports up to 8 bytes of payload, 2)
buffer the payload data into the video buffer, which supports up to 1024
bytes of payload.

The driver currently supports method 1).

Add support for transmitting long DSI commands (more than 8 bytes, up to
1024 bytes) using the video buffer. This mode can only be used before
the actual video transmission is enabled, i.e. the initial configuration.

Original version from Parth Pancholi <parth.pancholi@toradex.com>

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
---
 drivers/gpu/drm/bridge/tc358768.c | 79 ++++++++++++++++++++++++++++++++++-----
 1 file changed, 70 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/bridge/tc358768.c b/drivers/gpu/drm/bridge/tc358768.c
index e1ed4003b3c5..e0b5a4b5abbe 100644
--- a/drivers/gpu/drm/bridge/tc358768.c
+++ b/drivers/gpu/drm/bridge/tc358768.c
@@ -45,6 +45,9 @@
 
 /* Debug (16-bit addressable) */
 #define TC358768_VBUFCTRL		0x00E0
+#define TC358768_VBUFCTRL_VBUF_EN	BIT(15)
+#define TC358768_VBUFCTRL_TX_EN		BIT(14)
+#define TC358768_VBUFCTRL_MASK		BIT(13)
 #define TC358768_DBG_WIDTH		0x00E2
 #define TC358768_DBG_VBLANK		0x00E4
 #define TC358768_DBG_DATA		0x00E8
@@ -537,9 +540,21 @@ static ssize_t tc358768_dsi_host_transfer(struct mipi_dsi_host *host,
 		return -ENOTSUPP;
 	}
 
+	if (msg->tx_len > 1024) {
+		dev_warn(priv->dev, "Maximum 1024 byte MIPI tx is supported\n");
+		return -EINVAL;
+	}
+
 	if (msg->tx_len > 8) {
-		dev_warn(priv->dev, "Maximum 8 byte MIPI tx is supported\n");
-		return -ENOTSUPP;
+		u32 confctl;
+
+		tc358768_read(priv, TC358768_CONFCTL, &confctl);
+
+		if (confctl & BIT(6)) {
+			dev_warn(priv->dev,
+				 "Video is currently active. Unable to transmit long command\n");
+			return -EBUSY;
+		}
 	}
 
 	ret = mipi_dsi_create_packet(&packet, msg);
@@ -552,23 +567,66 @@ static ssize_t tc358768_dsi_host_transfer(struct mipi_dsi_host *host,
 		tc358768_write(priv, TC358768_DSICMD_WC, 0);
 		tc358768_write(priv, TC358768_DSICMD_WD0,
 			       (packet.header[2] << 8) | packet.header[1]);
-	} else {
-		int i;
-
+		tc358768_dsicmd_tx(priv);
+	} else if (packet.payload_length <= 8) {
 		tc358768_write(priv, TC358768_DSICMD_TYPE,
 			       (0x40 << 8) | (packet.header[0] & 0x3f));
 		tc358768_write(priv, TC358768_DSICMD_WC, packet.payload_length);
-		for (i = 0; i < packet.payload_length; i += 2) {
+
+		for (int i = 0; i < packet.payload_length; i += 2) {
 			u16 val = packet.payload[i];
 
 			if (i + 1 < packet.payload_length)
 				val |= packet.payload[i + 1] << 8;
-
 			tc358768_write(priv, TC358768_DSICMD_WD0 + i, val);
 		}
-	}
 
-	tc358768_dsicmd_tx(priv);
+		tc358768_dsicmd_tx(priv);
+	} else {
+		unsigned long tx_sleep_us;
+		size_t len;
+
+		/* For packets over 8 bytes we need to use the video buffer */
+		tc358768_write(priv, TC358768_DATAFMT, BIT(0));	/* txdt_en */
+		tc358768_write(priv, TC358768_DSITX_DT, packet.header[0] & 0x3f);
+		tc358768_write(priv, TC358768_CMDBYTE, packet.payload_length);
+		tc358768_write(priv, TC358768_VBUFCTRL, TC358768_VBUFCTRL_VBUF_EN);
+
+		/*
+		 * Write the payload in 2-byte chunks, and pad with zeroes to
+		 * align to 4 bytes.
+		 */
+		len = ALIGN(packet.payload_length, 4);
+
+		for (int i = 0; i < len; i += 2) {
+			u16 val = 0;
+
+			if (i < packet.payload_length)
+				val |= packet.payload[i];
+			if (i + 1 < packet.payload_length)
+				val |= packet.payload[i + 1] << 8;
+
+			tc358768_write(priv, TC358768_DBG_DATA, val);
+		}
+
+		/* Start transmission */
+		tc358768_write(priv, TC358768_VBUFCTRL,
+			       TC358768_VBUFCTRL_VBUF_EN |
+			       TC358768_VBUFCTRL_TX_EN |
+			       TC358768_VBUFCTRL_MASK);
+
+		/*
+		 * The TC358768 spec says to wait until the transmission has
+		 * been finished, estimating the sleep time based on the payload
+		 * and clock rates. We use a simple safe estimate of 2us per
+		 * byte (LP mode transmission).
+		 */
+		tx_sleep_us = packet.payload_length * 2;
+		usleep_range(tx_sleep_us, tx_sleep_us * 2);
+
+		tc358768_write(priv, TC358768_VBUFCTRL, TC358768_VBUFCTRL_MASK);
+		tc358768_write(priv, TC358768_VBUFCTRL, 0); /* Stop transmission */
+	}
 
 	ret = tc358768_clear_error(priv);
 	if (ret)
@@ -752,6 +810,9 @@ static void tc358768_bridge_atomic_pre_enable(struct drm_bridge *bridge,
 		return;
 	}
 
+	/* Release RstPtr so that the video buffer can be used for DSI commands */
+	tc358768_update_bits(priv, TC358768_PP_MISC, BIT(14), 0);
+
 	connector = drm_atomic_get_new_connector_for_encoder(state, bridge->encoder);
 	conn_state = drm_atomic_get_new_connector_state(state, connector);
 	crtc_state = drm_atomic_get_new_crtc_state(state, conn_state->crtc);

-- 
2.43.0

Re: [PATCH 7/7] drm/bridge: tc358768: Add support for long command tx via video buffer

[Francesco Dolcini]

On Tue, Oct 21, 2025 at 04:23:03PM +0300, Tomi Valkeinen wrote:
> TC358768 has two ways to send DSI commands: 1) buffer the payload data
> into registers (DSICMD_WDx), which supports up to 8 bytes of payload, 2)
> buffer the payload data into the video buffer, which supports up to 1024
> bytes of payload.
> 
> The driver currently supports method 1).
> 
> Add support for transmitting long DSI commands (more than 8 bytes, up to
> 1024 bytes) using the video buffer. This mode can only be used before
> the actual video transmission is enabled, i.e. the initial configuration.
> 
> Original version from Parth Pancholi <parth.pancholi@toradex.com>
> 
> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>

Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com>

Re: [PATCH 0/7] drm/bridge: tc358768: Long command support

[João Paulo Gonçalves]

Hi Tomi,

On Tue, Oct 21, 2025 at 04:22:56PM +0300, Tomi Valkeinen wrote:
> This series makes some small improvements to the tc358768 driver, and
> then adds support for long commands, commands that have more than 8
> bytes of payload.
> 
> This has been tested on Toradex Verdin AM62 board, with a ST7703 based
> DSI panel which requires initial configuration using commands that have
> 8+ bytes.
> 

Tested-by: João Paulo Gonçalves <joao.goncalves@toradex.com> # Toradex Verdin AM62

Best Regards,
João Paulo Gonçalves