Skip to contentSkip to conversation

drm: Fix bug in panel driver, update MIPI support macros

PATCH In review Brigham Campbell 3 participants 10 messages Jul 22, 2025 – Jul 29, 2025

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

v6 cover drm: Fix bug in panel driver, update MIPI support macros

This series removes the unintuitive mipi_dsi_generic_write_seq() macro and related mipi_dsi_generic_write_chatty() method from the drm subsystem. This is in accordance with a TODO item from Douglas Anderson in the drm subsystem documentation. Tejas Vipin (among others) has largely spearheaded this effort up until now, converting MIPI panel drivers one at a time.

The second patch of the series removes the last remaining references to mipi_dsi_generic_write_seq() in the jdi-lpm102a188a driver and updates the driver to use the undeprecated _multi variants of MIPI functions. It fixes a bug in the driver's unprepare function and cleans up duplicated code using the new mipi_dsi_dual* macros introduced in the first patch.

changes to v6:
- Fix various style and kerneldoc issues in patch 1/4
- Fix typo mpi_dsi_dual... -> mipi_dsi_dual...
- Fix incorrectly named "data" and "len" variables
- Make _seq argument of mipi_dsi_dual_dcs_write_seq_multi macro variadic
- Remove duplicate definition of mipi_dsi_dual_dcs_write_seq_multi macro from novatek display driver

changes to v5:
- Rework mipi_dsi_dual to explicitly not support passing macros into _func and add "dual" variants of the generic and dcs write macros.
- Make jdi-lpm102a188a use the new mipi_dsi_dual_generic_write_seq_multi macro.
- Make local struct variable in jdi-lpm102a188a conform to reverse christmas tree order.

changes to v4:
- Fix whitespace (I forgot to run checkpatch. Thanks for your patience as I familiarize myself with the kernel development process)
- Initialize mipi_dsi_multi_context struct

changes to v3:
- Define new mipi_dsi_dual macro in drm_mipi_dsi.h to reduce code duplication.
- Fix bug in lpm102a188a panel driver's unprepare function which causes it to return a nonsensical value.
- Make lpm102a188a panel driver's unprepare function send "display off" and "enter sleep mode" commands to both serial interfaces regardless of whether an error occurred when sending the last command.

changes to v2:
- Remove all usages of deprecated MIPI functions from jdi-lpm102a188a driver instead of just mipi_dsi_generic_write_seq().
- Update TODO item in drm documentation instead of removing it entirely.

Brigham Campbell (4):
drm: Create mipi_dsi_dual* macros
drm/panel: jdi-lpm102a188a: Fix bug and clean up driver
drm: Remove unused MIPI write seq and chatty functions
drm: docs: Update task from drm TODO list

 Documentation/gpu/todo.rst                    |  22 +-
 drivers/gpu/drm/drm_mipi_dsi.c                |  82 +-
 drivers/gpu/drm/panel/panel-jdi-lpm102a188a.c | 196 ++---
 drivers/gpu/drm/panel/panel-novatek-nt36523.c | 804 +++++++++---------
 include/drm/drm_mipi_dsi.h                    | 118 ++-
 5 files changed, 614 insertions(+), 608 deletions(-)

base-commit: ca2a6abdaee43808034cdb218428d2ed85fd3db8

Signature · 1 line
2.50.1
#1 lore raw

PATCH v6 1/4 drm: Create mipi_dsi_dual* macros

Create mipi_dsi_dual, mipi_dsi_dual_dcs_write_seq_multi, and mipi_dsi_dual_generic_write_seq_multi macros for panels which are driven by two parallel serial interfaces. This allows for the reduction of code duplication in drivers for these panels.

Remove mipi_dsi_dual_dcs_write_seq_multi definition from panel-novatek-nt36523.c to avoid the duplicate definition. Make novatek driver pass mipi_dsi_context struct as a pointer.

---
 drivers/gpu/drm/drm_mipi_dsi.c                |  48 ++
 drivers/gpu/drm/panel/panel-novatek-nt36523.c | 804 +++++++++---------
 include/drm/drm_mipi_dsi.h                    |  95 +++
 3 files changed, 541 insertions(+), 406 deletions(-)
drivers/gpu/drm/drm_mipi_dsi.c +48 -0
index a00d76443128..5b387543de28 100644
@@ -827,6 +827,30 @@ void mipi_dsi_generic_write_multi(struct mipi_dsi_multi_context *ctx,
 }
 EXPORT_SYMBOL(mipi_dsi_generic_write_multi);
 
+/**
+ * mipi_dsi_dual_generic_write_multi() - mipi_dsi_generic_write_multi() for
+ * two dsi channels, one after the other
+ * @ctx: Context for multiple DSI transactions
+ * @dsi1: First dsi channel to write buffer to
+ * @dsi2: Second dsi channel to write buffer to
+ * @payload: Buffer containing the payload
+ * @size: Size of payload buffer
+ *
+ * A wrapper around mipi_dsi_generic_write_multi() that allows the user to
+ * conveniently write to two dsi channels, one after the other.
+ */
+void mipi_dsi_dual_generic_write_multi(struct mipi_dsi_multi_context *ctx,
+				       struct mipi_dsi_device *dsi1,
+				       struct mipi_dsi_device *dsi2,
+				       const void *payload, size_t size)
+{
+	ctx->dsi = dsi1;
+	mipi_dsi_generic_write_multi(ctx, payload, size);
+	ctx->dsi = dsi2;
+	mipi_dsi_generic_write_multi(ctx, payload, size);
+}
+EXPORT_SYMBOL(mipi_dsi_dual_generic_write_multi);
+
 /**
  * mipi_dsi_generic_read() - receive data using a generic read packet
  * @dsi: DSI peripheral device
@@ -1006,6 +1030,30 @@ void mipi_dsi_dcs_write_buffer_multi(struct mipi_dsi_multi_context *ctx,
 }
 EXPORT_SYMBOL(mipi_dsi_dcs_write_buffer_multi);
 
+/**
+ * mipi_dsi_dual_dcs_write_buffer_multi - mipi_dsi_dcs_write_buffer_multi() for
+ * two dsi channels, one after the other
+ * @ctx: Context for multiple DSI transactions
+ * @dsi1: First dsi channel to write buffer to
+ * @dsi2: Second dsi channel to write buffer to
+ * @data: Buffer containing data to be transmitted
+ * @len: Size of transmission buffer
+ *
+ * A wrapper around mipi_dsi_dcs_write_buffer_multi() that allows the user to
+ * conveniently write to two dsi channels, one after the other.
+ */
+void mipi_dsi_dual_dcs_write_buffer_multi(struct mipi_dsi_multi_context *ctx,
+					  struct mipi_dsi_device *dsi1,
+					  struct mipi_dsi_device *dsi2,
+					  const void *data, size_t len)
+{
+	ctx->dsi = dsi1;
+	mipi_dsi_dcs_write_buffer_multi(ctx, data, len);
+	ctx->dsi = dsi2;
+	mipi_dsi_dcs_write_buffer_multi(ctx, data, len);
+}
+EXPORT_SYMBOL(mipi_dsi_dual_dcs_write_buffer_multi);
+
 /**
  * mipi_dsi_dcs_write() - send DCS write command
  * @dsi: DSI peripheral device
drivers/gpu/drm/panel/panel-novatek-nt36523.c +398 -406

824 lines not on this page · raw message

include/drm/drm_mipi_dsi.h +95 -0

113 lines not on this page · raw message

Signature · 1 line
2.50.1
Reviewed-by on v6 1/4
#2 lore raw

Re: [PATCH v6 1/4] drm: Create mipi_dsi_dual* macros

Hi,

Quoted 10 lines · Brigham Campbell

Create mipi_dsi_dual, mipi_dsi_dual_dcs_write_seq_multi, and mipi_dsi_dual_generic_write_seq_multi macros for panels which are driven by two parallel serial interfaces. This allows for the reduction of code duplication in drivers for these panels.

Remove mipi_dsi_dual_dcs_write_seq_multi definition from panel-novatek-nt36523.c to avoid the duplicate definition. Make novatek driver pass mipi_dsi_context struct as a pointer.

Quoted 7 lines
---
 drivers/gpu/drm/drm_mipi_dsi.c                |  48 ++
 drivers/gpu/drm/panel/panel-novatek-nt36523.c | 804 +++++++++---------
 include/drm/drm_mipi_dsi.h                    |  95 +++
 3 files changed, 541 insertions(+), 406 deletions(-)

Just because I was curious and wanted to confirm my claim that this would result in a nice space savings for novatek, I ran bloat-o-meter on the novatek module comparing before and after your patch:

add/remove: 0/0 grow/shrink: 0/4 up/down: 0/-8754 (-8754)

Function                                     old     new   delta
elish_csot_init_sequence.d                   758     379    -379
elish_boe_init_sequence.d                    846     423    -423
elish_csot_init_sequence                    9136    5380   -3756
elish_boe_init_sequence                    10192    5996   -4196

Total: Before=33880, After=25126, chg -25.84%

So the new way of defining mipi_dsi_dual_dcs_write_seq_multi() did indeed give a pretty sweet space savings! :-)

This patch looks good to me now. Thanks for putting up with all my requests. I know this is a lot bigger than you thought it would be when you posted your first patch, but I appreciate you sticking with it!

My plan would be to give this series another week on the list. If there is no additional feedback then I'll plan to land it.

-Doug

#3 lore raw

Re: [PATCH v6 1/4] drm: Create mipi_dsi_dual* macros

Quoted 11 lines · Doug Anderson

On Mon, Jul 21, 2025 at 6:53 PM Brigham Campbell <[email protected]> wrote:
add/remove: 0/0 grow/shrink: 0/4 up/down: 0/-8754 (-8754)

Function                                     old     new   delta
elish_csot_init_sequence.d                   758     379    -379
elish_boe_init_sequence.d                    846     423    -423
elish_csot_init_sequence                    9136    5380   -3756
elish_boe_init_sequence                    10192    5996   -4196

Total: Before=33880, After=25126, chg -25.84%

So the new way of defining mipi_dsi_dual_dcs_write_seq_multi() did indeed give a pretty sweet space savings! :-)

Interesting! With as much as I've heard about compiler optimizations, I'm kind of surprised that the compiler didn't do a better job before redefining mipi_dsi_dual_dcs_write_seq_multi(). It's nice to know what to expect when the rubber hits the proverbial road.

Quoted 9 lines

This patch looks good to me now. Thanks for putting up with all my requests. I know this is a lot bigger than you thought it would be when you posted your first patch, but I appreciate you sticking with it!

My plan would be to give this series another week on the list. If there is no additional feedback then I'll plan to land it.

Thanks, Doug. I appreciate your patient suggestions. Waiting another week for any additional comments or suggestions sounds good to me! Thanks for letting me know.

Brigham

#4 lore raw

PATCH v6 2/4 drm/panel: jdi-lpm102a188a: Fix bug and clean up driver

In reply to Brigham Campbell

Fix bug in unprepare() which causes the function's return value to be that of the last mipi "enter sleep mode" command.

Update driver to use the "multi" variant of MIPI functions in order to facilitate improved error handling and remove the panel's dependency on deprecated MIPI functions.

Use the new mipi_dsi_dual macro to reduce code duplication.

---
 drivers/gpu/drm/panel/panel-jdi-lpm102a188a.c | 196 ++++++------------
 1 file changed, 59 insertions(+), 137 deletions(-)
drivers/gpu/drm/panel/panel-jdi-lpm102a188a.c +59 -137
index 5f897e143758..83656bb4b0b2 100644
@@ -81,25 +81,25 @@ static int jdi_panel_disable(struct drm_panel *panel)
 static int jdi_panel_unprepare(struct drm_panel *panel)
 {
 	struct jdi_panel *jdi = to_panel_jdi(panel);
-	int ret;
 
-	ret = mipi_dsi_dcs_set_display_off(jdi->link1);
-	if (ret < 0)
-		dev_err(panel->dev, "failed to set display off: %d\n", ret);
+	/*
+	 * One context per panel since we'll continue trying to shut down the
+	 * other panel even if one isn't responding.
+	 */
+	struct mipi_dsi_multi_context dsi_ctx1 = { .dsi = jdi->link1 };
+	struct mipi_dsi_multi_context dsi_ctx2 = { .dsi = jdi->link2 };
 
-	ret = mipi_dsi_dcs_set_display_off(jdi->link2);
-	if (ret < 0)
-		dev_err(panel->dev, "failed to set display off: %d\n", ret);
+	mipi_dsi_dcs_set_display_off_multi(&dsi_ctx1);
+	mipi_dsi_dcs_set_display_off_multi(&dsi_ctx2);
 
 	/* Specified by JDI @ 50ms, subject to change */
 	msleep(50);
 
-	ret = mipi_dsi_dcs_enter_sleep_mode(jdi->link1);
-	if (ret < 0)
-		dev_err(panel->dev, "failed to enter sleep mode: %d\n", ret);
-	ret = mipi_dsi_dcs_enter_sleep_mode(jdi->link2);
-	if (ret < 0)
-		dev_err(panel->dev, "failed to enter sleep mode: %d\n", ret);
+	/* Doesn't hurt to try sleep mode even if display off fails */
+	dsi_ctx1.accum_err = 0;
+	dsi_ctx2.accum_err = 0;
+	mipi_dsi_dcs_enter_sleep_mode_multi(&dsi_ctx1);
+	mipi_dsi_dcs_enter_sleep_mode_multi(&dsi_ctx2);
 
 	/* Specified by JDI @ 150ms, subject to change */
 	msleep(150);
@@ -123,72 +123,46 @@ static int jdi_panel_unprepare(struct drm_panel *panel)
 	/* Specified by JDI @ 20ms, subject to change */
 	msleep(20);
 
-	return ret;
-}
-
-static int jdi_setup_symmetrical_split(struct mipi_dsi_device *left,
-				       struct mipi_dsi_device *right,
-				       const struct drm_display_mode *mode)
-{
-	int err;
-
-	err = mipi_dsi_dcs_set_column_address(left, 0, mode->hdisplay / 2 - 1);
-	if (err < 0) {
-		dev_err(&left->dev, "failed to set column address: %d\n", err);
-		return err;
-	}
-
-	err = mipi_dsi_dcs_set_column_address(right, 0, mode->hdisplay / 2 - 1);
-	if (err < 0) {
-		dev_err(&right->dev, "failed to set column address: %d\n", err);
-		return err;
-	}
-
-	err = mipi_dsi_dcs_set_page_address(left, 0, mode->vdisplay - 1);
-	if (err < 0) {
-		dev_err(&left->dev, "failed to set page address: %d\n", err);
-		return err;
-	}
-
-	err = mipi_dsi_dcs_set_page_address(right, 0, mode->vdisplay - 1);
-	if (err < 0) {
-		dev_err(&right->dev, "failed to set page address: %d\n", err);
-		return err;
-	}
-
 	return 0;
 }
 
-static int jdi_write_dcdc_registers(struct jdi_panel *jdi)
+static void jdi_setup_symmetrical_split(struct mipi_dsi_multi_context *dsi_ctx,
+					struct mipi_dsi_device *left,
+					struct mipi_dsi_device *right,
+					const struct drm_display_mode *mode)
+{
+	mipi_dsi_dual(mipi_dsi_dcs_set_column_address_multi,
+		      dsi_ctx, left, right,
+		      0, mode->hdisplay / 2 - 1);
+	mipi_dsi_dual(mipi_dsi_dcs_set_page_address_multi,
+		      dsi_ctx, left, right,
+		      0, mode->vdisplay - 1);
+}
+
+static void jdi_write_dcdc_registers(struct mipi_dsi_multi_context *dsi_ctx,
+				     struct jdi_panel *jdi)
 {
 	/* Clear the manufacturer command access protection */
-	mipi_dsi_generic_write_seq(jdi->link1, MCS_CMD_ACS_PROT,
-				   MCS_CMD_ACS_PROT_OFF);
-	mipi_dsi_generic_write_seq(jdi->link2, MCS_CMD_ACS_PROT,
-				   MCS_CMD_ACS_PROT_OFF);
+	mipi_dsi_dual_generic_write_seq_multi(dsi_ctx, jdi->link1, jdi->link2,
+					      MCS_CMD_ACS_PROT,
+					      MCS_CMD_ACS_PROT_OFF);
 	/*
-	 * Change the VGH/VGL divide rations to move the noise generated by the
+	 * Change the VGH/VGL divide ratios to move the noise generated by the
 	 * TCONN. This should hopefully avoid interaction with the backlight
 	 * controller.
 	 */
-	mipi_dsi_generic_write_seq(jdi->link1, MCS_PWR_CTRL_FUNC,
-				   MCS_PWR_CTRL_PARAM1_VGH_330_DIV |
-				   MCS_PWR_CTRL_PARAM1_DEFAULT,
-				   MCS_PWR_CTRL_PARAM2_VGL_410_DIV |
-				   MCS_PWR_CTRL_PARAM2_DEFAULT);
-
-	mipi_dsi_generic_write_seq(jdi->link2, MCS_PWR_CTRL_FUNC,
-				   MCS_PWR_CTRL_PARAM1_VGH_330_DIV |
-				   MCS_PWR_CTRL_PARAM1_DEFAULT,
-				   MCS_PWR_CTRL_PARAM2_VGL_410_DIV |
-				   MCS_PWR_CTRL_PARAM2_DEFAULT);
-
-	return 0;
+	mipi_dsi_dual_generic_write_seq_multi(dsi_ctx, jdi->link1, jdi->link2,
+					      MCS_PWR_CTRL_FUNC,
+					      MCS_PWR_CTRL_PARAM1_VGH_330_DIV |
+					      MCS_PWR_CTRL_PARAM1_DEFAULT,
+					      MCS_PWR_CTRL_PARAM2_VGL_410_DIV |
+					      MCS_PWR_CTRL_PARAM2_DEFAULT);
 }
 
 static int jdi_panel_prepare(struct drm_panel *panel)
 {
 	struct jdi_panel *jdi = to_panel_jdi(panel);
+	struct mipi_dsi_multi_context dsi_ctx = {};
 	int err;
 
 	/* Disable backlight to avoid showing random pixels
@@ -231,88 +205,36 @@ static int jdi_panel_prepare(struct drm_panel *panel)
 	 * put in place to communicate the configuration back to the DSI host
 	 * controller.
 	 */
-	err = jdi_setup_symmetrical_split(jdi->link1, jdi->link2,
-					  jdi->mode);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set up symmetrical split: %d\n",
-			err);
-		goto poweroff;
-	}
+	jdi_setup_symmetrical_split(&dsi_ctx, jdi->link1, jdi->link2,
+				    jdi->mode);
 
-	err = mipi_dsi_dcs_set_tear_scanline(jdi->link1,
-					     jdi->mode->vdisplay - 16);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set tear scanline: %d\n", err);
-		goto poweroff;
-	}
+	mipi_dsi_dual(mipi_dsi_dcs_set_tear_scanline_multi,
+		      &dsi_ctx, jdi->link1, jdi->link2,
+		      jdi->mode->vdisplay - 16);
 
-	err = mipi_dsi_dcs_set_tear_scanline(jdi->link2,
-					     jdi->mode->vdisplay - 16);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set tear scanline: %d\n", err);
-		goto poweroff;
-	}
+	mipi_dsi_dual(mipi_dsi_dcs_set_tear_on_multi,
+		      &dsi_ctx, jdi->link1, jdi->link2,
+		      MIPI_DSI_DCS_TEAR_MODE_VBLANK);
 
-	err = mipi_dsi_dcs_set_tear_on(jdi->link1,
-				       MIPI_DSI_DCS_TEAR_MODE_VBLANK);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set tear on: %d\n", err);
-		goto poweroff;
-	}
+	mipi_dsi_dual(mipi_dsi_dcs_set_pixel_format_multi,
+		      &dsi_ctx, jdi->link1, jdi->link2,
+		      MIPI_DCS_PIXEL_FMT_24BIT);
 
-	err = mipi_dsi_dcs_set_tear_on(jdi->link2,
-				       MIPI_DSI_DCS_TEAR_MODE_VBLANK);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set tear on: %d\n", err);
-		goto poweroff;
-	}
+	mipi_dsi_dual(mipi_dsi_dcs_exit_sleep_mode_multi,
+		      &dsi_ctx, jdi->link1, jdi->link2);
 
-	err = mipi_dsi_dcs_set_pixel_format(jdi->link1, MIPI_DCS_PIXEL_FMT_24BIT);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set pixel format: %d\n", err);
-		goto poweroff;
-	}
-
-	err = mipi_dsi_dcs_set_pixel_format(jdi->link2, MIPI_DCS_PIXEL_FMT_24BIT);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set pixel format: %d\n", err);
-		goto poweroff;
-	}
-
-	err = mipi_dsi_dcs_exit_sleep_mode(jdi->link1);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to exit sleep mode: %d\n", err);
-		goto poweroff;
-	}
-
-	err = mipi_dsi_dcs_exit_sleep_mode(jdi->link2);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to exit sleep mode: %d\n", err);
-		goto poweroff;
-	}
-
-	err = jdi_write_dcdc_registers(jdi);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to write dcdc registers: %d\n", err);
-		goto poweroff;
-	}
+	jdi_write_dcdc_registers(&dsi_ctx, jdi);
 	/*
-	 * We need to wait 150ms between mipi_dsi_dcs_exit_sleep_mode() and
-	 * mipi_dsi_dcs_set_display_on().
+	 * We need to wait 150ms between mipi_dsi_dcs_exit_sleep_mode_multi()
+	 * and mipi_dsi_dcs_set_display_on_multi().
 	 */
-	msleep(150);
+	mipi_dsi_msleep(&dsi_ctx, 150);
 
-	err = mipi_dsi_dcs_set_display_on(jdi->link1);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set display on: %d\n", err);
-		goto poweroff;
-	}
+	mipi_dsi_dual(mipi_dsi_dcs_set_display_on_multi,
+		      &dsi_ctx, jdi->link1, jdi->link2);
 
-	err = mipi_dsi_dcs_set_display_on(jdi->link2);
-	if (err < 0) {
-		dev_err(panel->dev, "failed to set display on: %d\n", err);
+	if (dsi_ctx.accum_err < 0)
 		goto poweroff;
-	}
 
 	jdi->link1->mode_flags &= ~MIPI_DSI_MODE_LPM;
 	jdi->link2->mode_flags &= ~MIPI_DSI_MODE_LPM;
Signature · 1 line
2.50.1
#5 lore raw

Re: [PATCH v6 2/4] drm/panel: jdi-lpm102a188a: Fix bug and clean up driver

Hi,

Quoted 10 lines · Brigham Campbell

Fix bug in unprepare() which causes the function's return value to be that of the last mipi "enter sleep mode" command.

Update driver to use the "multi" variant of MIPI functions in order to facilitate improved error handling and remove the panel's dependency on deprecated MIPI functions.

Use the new mipi_dsi_dual macro to reduce code duplication.

Quoted 7 lines
---
 drivers/gpu/drm/panel/panel-jdi-lpm102a188a.c | 196 ++++++------------
 1 file changed, 59 insertions(+), 137 deletions(-)

Just for fun, I ran bloat-o-meter to compare before and after.

add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-217 (-217)
Function old new delta
jdi_write_dcdc_registers.d 10 5 -5
jdi_panel_unprepare 312 260 -52
jdi_panel_prepare 1020 860 -160
Total: Before=4908, After=4691, chg -4.42%

...not as big of a savings, but that's not too surprising in this case since we don't have the whole pile of "seq" functions. Still the point was making the driver cleaner and the space savings is just an extra bonus!

-Doug

#6 lore raw

PATCH v6 3/4 drm: Remove unused MIPI write seq and chatty functions

In reply to Brigham Campbell

Remove the deprecated mipi_dsi_generic_write_seq() and mipi_dsi_generic_write_chatty() functions now that they are no longer used.

---
 drivers/gpu/drm/drm_mipi_dsi.c | 34 +++-------------------------------
 include/drm/drm_mipi_dsi.h     | 23 -----------------------
 2 files changed, 3 insertions(+), 54 deletions(-)
drivers/gpu/drm/drm_mipi_dsi.c +3 -31
index 5b387543de28..ca77f152404a 100644
@@ -772,41 +772,13 @@ ssize_t mipi_dsi_generic_write(struct mipi_dsi_device *dsi, const void *payload,
 EXPORT_SYMBOL(mipi_dsi_generic_write);
 
 /**
- * mipi_dsi_generic_write_chatty() - mipi_dsi_generic_write() w/ an error log
- * @dsi: DSI peripheral device
- * @payload: buffer containing the payload
- * @size: size of payload buffer
- *
- * Like mipi_dsi_generic_write() but includes a dev_err()
- * call for you and returns 0 upon success, not the number of bytes sent.
- *
- * Return: 0 on success or a negative error code on failure.
- */
-int mipi_dsi_generic_write_chatty(struct mipi_dsi_device *dsi,
-				  const void *payload, size_t size)
-{
-	struct device *dev = &dsi->dev;
-	ssize_t ret;
-
-	ret = mipi_dsi_generic_write(dsi, payload, size);
-	if (ret < 0) {
-		dev_err(dev, "sending generic data %*ph failed: %zd\n",
-			(int)size, payload, ret);
-		return ret;
-	}
-
-	return 0;
-}
-EXPORT_SYMBOL(mipi_dsi_generic_write_chatty);
-
-/**
- * mipi_dsi_generic_write_multi() - mipi_dsi_generic_write_chatty() w/ accum_err
+ * mipi_dsi_generic_write_multi() - mipi_dsi_generic_write() w/ accum_err
  * @ctx: Context for multiple DSI transactions
  * @payload: buffer containing the payload
  * @size: size of payload buffer
  *
- * Like mipi_dsi_generic_write_chatty() but deals with errors in a way that
- * makes it convenient to make several calls in a row.
+ * A wrapper around mipi_dsi_generic_write() that deals with errors in a way
+ * that makes it convenient to make several calls in a row.
  */
 void mipi_dsi_generic_write_multi(struct mipi_dsi_multi_context *ctx,
 				  const void *payload, size_t size)
include/drm/drm_mipi_dsi.h +0 -23
index f1dc822f69d6..ea523eb35b08 100644
@@ -285,8 +285,6 @@ void mipi_dsi_picture_parameter_set_multi(struct mipi_dsi_multi_context *ctx,
 
 ssize_t mipi_dsi_generic_write(struct mipi_dsi_device *dsi, const void *payload,
 			       size_t size);
-int mipi_dsi_generic_write_chatty(struct mipi_dsi_device *dsi,
-				  const void *payload, size_t size);
 void mipi_dsi_generic_write_multi(struct mipi_dsi_multi_context *ctx,
 				  const void *payload, size_t size);
 void mipi_dsi_dual_generic_write_multi(struct mipi_dsi_multi_context *ctx,
@@ -387,27 +385,6 @@ void mipi_dsi_dcs_set_tear_scanline_multi(struct mipi_dsi_multi_context *ctx,
 					  u16 scanline);
 void mipi_dsi_dcs_set_tear_off_multi(struct mipi_dsi_multi_context *ctx);
 
-/**
- * mipi_dsi_generic_write_seq - transmit data using a generic write packet
- *
- * This macro will print errors for you and will RETURN FROM THE CALLING
- * FUNCTION (yes this is non-intuitive) upon error.
- *
- * Because of the non-intuitive return behavior, THIS MACRO IS DEPRECATED.
- * Please replace calls of it with mipi_dsi_generic_write_seq_multi().
- *
- * @dsi: DSI peripheral device
- * @seq: buffer containing the payload
- */
-#define mipi_dsi_generic_write_seq(dsi, seq...)                                \
-	do {                                                                   \
-		static const u8 d[] = { seq };                                 \
-		int ret;                                                       \
-		ret = mipi_dsi_generic_write_chatty(dsi, d, ARRAY_SIZE(d));    \
-		if (ret < 0)                                                   \
-			return ret;                                            \
-	} while (0)
-
 /**
  * mipi_dsi_generic_write_seq_multi - transmit data using a generic write packet
  *
Signature · 1 line
2.50.1
Reviewed-by on v6 3/4
#7 lore raw

Re: [PATCH v6 3/4] drm: Remove unused MIPI write seq and chatty functions

Quoted 11 lines · Brigham Campbell

Remove the deprecated mipi_dsi_generic_write_seq() and mipi_dsi_generic_write_chatty() functions now that they are no longer used.

---
 drivers/gpu/drm/drm_mipi_dsi.c | 34 +++-------------------------------
 include/drm/drm_mipi_dsi.h     | 23 -----------------------
 2 files changed, 3 insertions(+), 54 deletions(-)
Signature · 2 lines
With best wishes
Dmitry
#8 lore raw

PATCH v6 4/4 drm: docs: Update task from drm TODO list

In reply to Brigham Campbell

Update TODO item from drm documentation to contain more applicable information regarding the removal of deprecated MIPI DSI functions and no longer reference functions which have already been removed from the kernel.

---
 Documentation/gpu/todo.rst | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)
Documentation/gpu/todo.rst +11 -11
index be8637da3fe9..92db80793bba 100644
@@ -497,19 +497,19 @@ Contact: Douglas Anderson <[email protected]>
 
 Level: Intermediate
 
-Transition away from using mipi_dsi_*_write_seq()
--------------------------------------------------
+Transition away from using deprecated MIPI DSI functions
+--------------------------------------------------------
 
-The macros mipi_dsi_generic_write_seq() and mipi_dsi_dcs_write_seq() are
-non-intuitive because, if there are errors, they return out of the *caller's*
-function. We should move all callers to use mipi_dsi_generic_write_seq_multi()
-and mipi_dsi_dcs_write_seq_multi() macros instead.
+There are many functions defined in ``drm_mipi_dsi.c`` which have been
+deprecated. Each deprecated function was deprecated in favor of its `multi`
+variant (e.g. `mipi_dsi_generic_write()` and `mipi_dsi_generic_write_multi()`).
+The `multi` variant of a function includes improved error handling and logic
+which makes it more convenient to make several calls in a row, as most MIPI
+drivers do.
 
-Once all callers are transitioned, the macros and the functions that they call,
-mipi_dsi_generic_write_chatty() and mipi_dsi_dcs_write_buffer_chatty(), can
-probably be removed. Alternatively, if people feel like the _multi() variants
-are overkill for some use cases, we could keep the mipi_dsi_*_write_seq()
-variants but change them not to return out of the caller.
+Drivers should be updated to use undeprecated functions. Once all usages of the
+deprecated MIPI DSI functions have been removed, their definitions may be
+removed from ``drm_mipi_dsi.c``.
 
 Contact: Douglas Anderson <[email protected]>
 
Signature · 1 line
2.50.1
#9 lore raw
In reply to Brigham Campbell

Hi,

Quoted 57 lines · Brigham Campbell

This series removes the unintuitive mipi_dsi_generic_write_seq() macro and related mipi_dsi_generic_write_chatty() method from the drm subsystem. This is in accordance with a TODO item from Douglas Anderson in the drm subsystem documentation. Tejas Vipin (among others) has largely spearheaded this effort up until now, converting MIPI panel drivers one at a time.

The second patch of the series removes the last remaining references to mipi_dsi_generic_write_seq() in the jdi-lpm102a188a driver and updates the driver to use the undeprecated _multi variants of MIPI functions. It fixes a bug in the driver's unprepare function and cleans up duplicated code using the new mipi_dsi_dual* macros introduced in the first patch.

changes to v6:
- Fix various style and kerneldoc issues in patch 1/4
- Fix typo mpi_dsi_dual... -> mipi_dsi_dual...
- Fix incorrectly named "data" and "len" variables
- Make _seq argument of mipi_dsi_dual_dcs_write_seq_multi macro variadic
- Remove duplicate definition of mipi_dsi_dual_dcs_write_seq_multi macro from novatek display driver

changes to v5:
- Rework mipi_dsi_dual to explicitly not support passing macros into _func and add "dual" variants of the generic and dcs write macros.
- Make jdi-lpm102a188a use the new mipi_dsi_dual_generic_write_seq_multi macro.
- Make local struct variable in jdi-lpm102a188a conform to reverse christmas tree order.

changes to v4:
- Fix whitespace (I forgot to run checkpatch. Thanks for your patience as I familiarize myself with the kernel development process)
- Initialize mipi_dsi_multi_context struct

changes to v3:
- Define new mipi_dsi_dual macro in drm_mipi_dsi.h to reduce code duplication.
- Fix bug in lpm102a188a panel driver's unprepare function which causes it to return a nonsensical value.
- Make lpm102a188a panel driver's unprepare function send "display off" and "enter sleep mode" commands to both serial interfaces regardless of whether an error occurred when sending the last command.

changes to v2:
- Remove all usages of deprecated MIPI functions from jdi-lpm102a188a driver instead of just mipi_dsi_generic_write_seq().
- Update TODO item in drm documentation instead of removing it entirely.

Brigham Campbell (4):
drm: Create mipi_dsi_dual* macros
drm/panel: jdi-lpm102a188a: Fix bug and clean up driver
drm: Remove unused MIPI write seq and chatty functions
drm: docs: Update task from drm TODO list

Quoted 6 lines
 Documentation/gpu/todo.rst                    |  22 +-
 drivers/gpu/drm/drm_mipi_dsi.c                |  82 +-
 drivers/gpu/drm/panel/panel-jdi-lpm102a188a.c | 196 ++---
 drivers/gpu/drm/panel/panel-novatek-nt36523.c | 804 +++++++++---------
 include/drm/drm_mipi_dsi.h                    | 118 ++-
 5 files changed, 614 insertions(+), 608 deletions(-)

Pushed all 4 to drm-misc-next. Technically one of them has a minor bugfix as mentioned in the commit message, but it's VERY minor (I didn't even think it needed a Fixes tag) so I didn't worry about trying to get it into drm-misc-fixes...

[1/4] drm: Create mipi_dsi_dual* macros

      commit: d94a2a00d2b8878678607c2969fee3b4e59126cb
[2/4] drm/panel: jdi-lpm102a188a: Fix bug and clean up driver
      commit: a6adf47d30cccaf3c1936ac9de94948c140e17dd
[3/4] drm: Remove unused MIPI write seq and chatty functions
      commit: 79b6bb18f849818140dd351f6e76a097efe99e9f
[4/4] drm: docs: Update task from drm TODO list
      commit: 85c23f28905cf20a86ceec3cfd7a0a5572c9eb13

-Doug