Skip to contentSkip to conversation

drm/panel: auo novatek 1080p video mode panel

PATCH In review Rob Clark 3 participants 10 messages Jul 21, 2015 – Aug 17, 2015

Subsystems (5)
lore
Share & tools
lore

Rob Clark

author
#0 lore raw

PATCH 1/1 drm/panel: auo novatek 1080p video mode panel

This is one of several different panels that are used on the Sony Xperia Z3 phone. It can operate in either command or video mode, although so far only video mode is implemented (since that is the mode that the downstream kernel version I happened to be looking at was using, and that is where I got the programming sequences for the panel).

---
Couple Notes:
1) programming sequences and basically everything I know about the panel came from downstream android kernel. I've started a wiki page to document how to translate from downstream kernel-msm way of doing things to upstream panel framework, which may be useful for others wanting to port downstream panel drivers for snapdragon devices:

https://github.com/freedreno/freedreno/wiki/DSI-Panel-Driver-Porting

2) The Sony phones at least (not sure if this is common) use one of several different panels, with runtime probing. Depending on the device they seem to either use a gpio (simple) or send some DSI commands to read back the panel-id. My rough thinking here about how to handle this is to implement a "panel-meta" driver (or maybe one each for the different probing methods), which would take a list of phandles to the actual candidate panels, and fwd the panel fxn calls to the chosen panel after probing.

 .../bindings/panel/auo,novatek-1080p.txt           |  25 ++
 drivers/gpu/drm/panel/Kconfig                      |   9 +
 drivers/gpu/drm/panel/Makefile                     |   1 +
 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c    | 470 +++++++++++++++++++++
 4 files changed, 505 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt
 create mode 100644 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +25 -0
new file mode 100644
index 0000000..8a53093
@@ -0,0 +1,25 @@
+AU Optronics Corporation 1080x1920 DSI panel
+
+This panel supports both video and command mode (although currently only video
+mode is implemented in the driver.
+
+Required properties:
+- compatible: should be "auo,novatek-1080p-vid"
+
+Optional properties:
+- power-supply: phandle of the regulator that provides the supply voltage
+- reset-gpio: phandle of gpio for reset line
+- backlight: phandle of the backlight device attached to the panel
+
+Example:
+
+	dsi@54300000 {
+		panel: panel@0 {
+			compatible = "auo,novatek-1080p-vid";
+			reg = <0>;
+
+			power-supply = <...>;
+			reset-gpio = <...>;
+			backlight = <...>;
+		};
+	};
drivers/gpu/drm/panel/Kconfig +9 -0
index 6d64c7b..89f0e8c 100644
@@ -43,4 +43,13 @@ config DRM_PANEL_SHARP_LQ101R1SX01
 	  To compile this driver as a module, choose M here: the module
 	  will be called panel-sharp-lq101r1sx01.
 
+config DRM_PANEL_AUO_NOVATEK_1080P
+	tristate "AUO Novatek 1080p video mode panel"
+	depends on OF
+	depends on DRM_MIPI_DSI
+	depends on BACKLIGHT_CLASS_DEVICE
+	help
+	  Say Y here if you want to enable support for AUO 1080p DSI panel
+	  as found in some Sony XPERIA Z3 devices
+
 endmenu
drivers/gpu/drm/panel/Makefile +1 -0
index 4b2a043..cbcfedf 100644
@@ -2,3 +2,4 @@ obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
 obj-$(CONFIG_DRM_PANEL_LD9040) += panel-ld9040.o
 obj-$(CONFIG_DRM_PANEL_S6E8AA0) += panel-s6e8aa0.o
 obj-$(CONFIG_DRM_PANEL_SHARP_LQ101R1SX01) += panel-sharp-lq101r1sx01.o
+obj-$(CONFIG_DRM_PANEL_AUO_NOVATEK_1080P) += panel-auo-novatek-1080p.o
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c new file +470 -0

470 lines not on this page · raw message

Signature · 1 line
2.4.3
#1 lore raw
Quoted 41 lines · Rob Clark

This is one of several different panels that are used on the Sony Xperia Z3 phone. It can operate in either command or video mode, although so far only video mode is implemented (since that is the mode that the downstream kernel version I happened to be looking at was using, and that is where I got the programming sequences for the panel).

---
Couple Notes:
1) programming sequences and basically everything I know about the panel came from downstream android kernel. I've started a wiki page to document how to translate from downstream kernel-msm way of doing things to upstream panel framework, which may be useful for others wanting to port downstream panel drivers for snapdragon devices:

https://github.com/freedreno/freedreno/wiki/DSI-Panel-Driver-Porting

2) The Sony phones at least (not sure if this is common) use one of several different panels, with runtime probing. Depending on the device they seem to either use a gpio (simple) or send some DSI commands to read back the panel-id. My rough thinking here about how to handle this is to implement a "panel-meta" driver (or maybe one each for the different probing methods), which would take a list of phandles to the actual candidate panels, and fwd the panel fxn calls to the chosen panel after probing.

 .../bindings/panel/auo,novatek-1080p.txt           |  25 ++
 drivers/gpu/drm/panel/Kconfig                      |   9 +
 drivers/gpu/drm/panel/Makefile                     |   1 +
 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c    | 470 +++++++++++++++++++++
 4 files changed, 505 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt
 create mode 100644 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +0 -0
new file mode 100644
index 0000000..8a53093

Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt line 7

Quoted 7 lines
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +7 -0
@@ -0,0 +1,25 @@
+AU Optronics Corporation 1080x1920 DSI panel
+
+This panel supports both video and command mode (although currently only video
+mode is implemented in the driver.
+
+Required properties:
+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Also, what's the -vid suffix for?

Quoted 24 lines
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +17 -0
+Optional properties:
+- power-supply: phandle of the regulator that provides the supply voltage
+- reset-gpio: phandle of gpio for reset line
+- backlight: phandle of the backlight device attached to the panel
+
+Example:
+
+	dsi@54300000 {
+		panel: panel@0 {
+			compatible = "auo,novatek-1080p-vid";
+			reg = <0>;
+
+			power-supply = <...>;
+			reset-gpio = <...>;
+			backlight = <...>;
+		};
+	};
drivers/gpu/drm/panel/Kconfig +0 -0
index 6d64c7b..89f0e8c 100644
@@ -43,4 +43,13 @@ config DRM_PANEL_SHARP_LQ101R1SX01
 	  To compile this driver as a module, choose M here: the module
 	  will be called panel-sharp-lq101r1sx01.

drivers/gpu/drm/panel/Kconfig config DRM_PANEL_SHARP_LQ101R1SX01 · line 54

Quoted 10 lines
drivers/gpu/drm/panel/Kconfig +9 -0
 
+config DRM_PANEL_AUO_NOVATEK_1080P
+	tristate "AUO Novatek 1080p video mode panel"
+	depends on OF
+	depends on DRM_MIPI_DSI
+	depends on BACKLIGHT_CLASS_DEVICE
+	help
+	  Say Y here if you want to enable support for AUO 1080p DSI panel
+	  as found in some Sony XPERIA Z3 devices
+

Can we sort this alphabetically, please?

drivers/gpu/drm/panel/Makefile obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o · line 5

Quoted 10 lines
drivers/gpu/drm/panel/Kconfig +0 -0
 endmenu
drivers/gpu/drm/panel/Makefile +1 -0
index 4b2a043..cbcfedf 100644
@@ -2,3 +2,4 @@ obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
 obj-$(CONFIG_DRM_PANEL_LD9040) += panel-ld9040.o
 obj-$(CONFIG_DRM_PANEL_S6E8AA0) += panel-s6e8aa0.o
 obj-$(CONFIG_DRM_PANEL_SHARP_LQ101R1SX01) += panel-sharp-lq101r1sx01.o
+obj-$(CONFIG_DRM_PANEL_AUO_NOVATEK_1080P) += panel-auo-novatek-1080p.o

This too.

Actually, nevermind. I have local patches to add vendor prefixes more consistently so that we can actually sort properly. I can fix that up in your patch when I apply.

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 57

Quoted 9 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +8 -0
+static int auo_panel_init(struct auo_panel *auo)
+{
+	struct mipi_dsi_device *dsi = auo->dsi;
+	int ret;
+
+	dsi->mode_flags |= MIPI_DSI_MODE_LPM;
+
+	ret = mipi_dsi_generic_write(dsi, (u8[]){ 0xb0, 0x04 }, 2);

I find this notation hard to read. Have you considered moving this into some sort of table that you can loop through? Or perhaps add some helpers, say, mipi_dsi_generic_writeb() and mipi_dsi_dcs_writeb() to help make this more readable?

Quoted 81 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +81 -0
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0xe0 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xb5, (u8[]){ 0x86 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xb6, (u8[]){ 0x77 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xb8, (u8[]){ 0xad }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x20 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_enter_sleep_mode(dsi);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x24 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xc6, (u8[]){ 0x00 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xc5, (u8[]){ 0x32 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0x92, (u8[]){ 0x92 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x10 }, 1);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_set_tear_on(dsi, MIPI_DSI_DCS_TEAR_MODE_VBLANK);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_TEAR_SCANLINE,
+			(u8[]){ 0x03, 0x00 }, 2);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0x3b, (u8[]){ 0x03, 0x30, 0x06 }, 3);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xbb, (u8[]){ 0x10 }, 1);
+	if (ret < 0)
+		return ret;
+	msleep(1);
+
+	ret = mipi_dsi_dcs_exit_sleep_mode(dsi);
+	if (ret < 0)
+		return ret;
+	msleep(30);

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 148

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+
+	return 0;
+}
+
+static int auo_panel_on(struct auo_panel *auo)
+{
+	struct mipi_dsi_device *dsi = auo->dsi;
+	int ret;
+
+	dsi->mode_flags |= MIPI_DSI_MODE_LPM;

This is weird.

Quoted 5 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+	ret = mipi_dsi_dcs_set_display_on(dsi);
+	if (ret < 0)
+		return ret;
+
+	msleep(40);

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 164

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+
+	return 0;
+}
+
+static int auo_panel_off(struct auo_panel *auo)
+{
+	struct mipi_dsi_device *dsi = auo->dsi;
+	int ret;
+
+	dsi->mode_flags &= ~MIPI_DSI_MODE_LPM;

And this even more. Doesn't the panel work when you simply send everything in low-power mode?

Quoted 38 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +38 -0
+static int auo_panel_prepare(struct drm_panel *panel)
+{
+	struct auo_panel *auo = to_auo_panel(panel);
+	int ret;
+
+	if (auo->prepared)
+		return 0;
+
+	DRM_DEBUG("prepare\n");
+
+	if (auo->reset_gpio) {
+		gpiod_set_value(auo->reset_gpio, 0);
+		msleep(5);
+	}
+
+	ret = regulator_enable(auo->supply);
+	if (ret < 0)
+		return ret;
+
+	msleep(20);
+
+	if (auo->reset_gpio) {
+		gpiod_set_value(auo->reset_gpio, 1);
+		msleep(10);
+	}
+
+	msleep(150);
+
+	ret = auo_panel_init(auo);
+	if (ret) {
+		dev_err(panel->dev, "failed to init panel: %d\n", ret);
+		goto poweroff;
+	}
+
+	ret = auo_panel_on(auo);
+	if (ret) {
+		dev_err(panel->dev, "failed to set panel on: %d\n", ret);
+		goto poweroff;

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 274

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+	}
+
+	auo->prepared = true;
+
+	return 0;
+
+poweroff:
+	regulator_disable(auo->supply);
+	if (auo->reset_gpio)
+		gpiod_set_value(auo->reset_gpio, 0);

You should be able to do without the check here, because gpiod_set_value() will simply nop if the GPIO is NULL.

I assume you may not want to do it above because of the additional delays that are only relevant if you have a reset GPIO.

Quoted 19 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +19 -0
+	return ret;
+}
+
+static int auo_panel_enable(struct drm_panel *panel)
+{
+	struct auo_panel *auo = to_auo_panel(panel);
+
+	if (auo->enabled)
+		return 0;
+
+	DRM_DEBUG("enable\n");
+
+	if (auo->backlight) {
+		auo->backlight->props.power = FB_BLANK_UNBLANK;
+		backlight_update_status(auo->backlight);
+	}
+
+	auo->enabled = true;
+

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 352

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+	return 0;
+}
+
+static int auo_panel_add(struct auo_panel *auo)
+{
+	struct device *dev= &auo->dsi->dev;
+	struct device_node *np;
+	int ret;
+
+	auo->mode = &default_mode;

This seems to be unused.

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 364

Quoted 12 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +12 -0
+
+	auo->supply = devm_regulator_get(dev, "power");
+	if (IS_ERR(auo->supply))
+		return PTR_ERR(auo->supply);
+
+	auo->reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_LOW);
+	if (IS_ERR(auo->reset_gpio)) {
+		dev_err(dev, "cannot get reset-gpios %ld\n",
+			PTR_ERR(auo->reset_gpio));
+		auo->reset_gpio = NULL;
+	} else {
+		gpiod_direction_output(auo->reset_gpio, 0);

Isn't that what GPIOD_OUT_LOW already does?

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

author
#2 lore raw
Quoted 45 lines · Thierry Reding
Quoted 44 lines · Rob Clark

This is one of several different panels that are used on the Sony Xperia Z3 phone. It can operate in either command or video mode, although so far only video mode is implemented (since that is the mode that the downstream kernel version I happened to be looking at was using, and that is where I got the programming sequences for the panel).

---
Couple Notes:
1) programming sequences and basically everything I know about the panel came from downstream android kernel. I've started a wiki page to document how to translate from downstream kernel-msm way of doing things to upstream panel framework, which may be useful for others wanting to port downstream panel drivers for snapdragon devices:

https://github.com/freedreno/freedreno/wiki/DSI-Panel-Driver-Porting

2) The Sony phones at least (not sure if this is common) use one of several different panels, with runtime probing. Depending on the device they seem to either use a gpio (simple) or send some DSI commands to read back the panel-id. My rough thinking here about how to handle this is to implement a "panel-meta" driver (or maybe one each for the different probing methods), which would take a list of phandles to the actual candidate panels, and fwd the panel fxn calls to the chosen panel after probing.

 .../bindings/panel/auo,novatek-1080p.txt           |  25 ++
 drivers/gpu/drm/panel/Kconfig                      |   9 +
 drivers/gpu/drm/panel/Makefile                     |   1 +
 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c    | 470 +++++++++++++++++++++
 4 files changed, 505 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt
 create mode 100644 drivers/gpu/drm/panel/panel-auo-novatek-1080p.c
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +3 -0
new file mode 100644
index 0000000..8a53093
@@ -0,0 +1,25 @@
+AU Optronics Corporation 1080x1920 DSI panel
+
+This panel supports both video and command mode (although currently only video
Quoted 8 lines

Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt line 7

Quoted 4 lines
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +4 -0
+mode is implemented in the driver.
+
+Required properties:
+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

Quoted 54 lines
Quoted 24 lines
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +17 -0
+Optional properties:
+- power-supply: phandle of the regulator that provides the supply voltage
+- reset-gpio: phandle of gpio for reset line
+- backlight: phandle of the backlight device attached to the panel
+
+Example:
+
+     dsi@54300000 {
+             panel: panel@0 {
+                     compatible = "auo,novatek-1080p-vid";
+                     reg = <0>;
+
+                     power-supply = <...>;
+                     reset-gpio = <...>;
+                     backlight = <...>;
+             };
+     };
drivers/gpu/drm/panel/Kconfig +0 -0
index 6d64c7b..89f0e8c 100644
@@ -43,4 +43,13 @@ config DRM_PANEL_SHARP_LQ101R1SX01
        To compile this driver as a module, choose M here: the module
        will be called panel-sharp-lq101r1sx01.

drivers/gpu/drm/panel/Kconfig config DRM_PANEL_SHARP_LQ101R1SX01 · line 54

Quoted 10 lines
drivers/gpu/drm/panel/Kconfig +9 -0
 
+config DRM_PANEL_AUO_NOVATEK_1080P
+     tristate "AUO Novatek 1080p video mode panel"
+     depends on OF
+     depends on DRM_MIPI_DSI
+     depends on BACKLIGHT_CLASS_DEVICE
+     help
+       Say Y here if you want to enable support for AUO 1080p DSI panel
+       as found in some Sony XPERIA Z3 devices
+

Can we sort this alphabetically, please?

drivers/gpu/drm/panel/Makefile obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o · line 5

Quoted 10 lines
drivers/gpu/drm/panel/Kconfig +0 -0
 endmenu
drivers/gpu/drm/panel/Makefile +1 -0
index 4b2a043..cbcfedf 100644
@@ -2,3 +2,4 @@ obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
 obj-$(CONFIG_DRM_PANEL_LD9040) += panel-ld9040.o
 obj-$(CONFIG_DRM_PANEL_S6E8AA0) += panel-s6e8aa0.o
 obj-$(CONFIG_DRM_PANEL_SHARP_LQ101R1SX01) += panel-sharp-lq101r1sx01.o
+obj-$(CONFIG_DRM_PANEL_AUO_NOVATEK_1080P) += panel-auo-novatek-1080p.o

This too.

Actually, nevermind. I have local patches to add vendor prefixes more consistently so that we can actually sort properly. I can fix that up in your patch when I apply.

Quoted 15 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 57

Quoted 9 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +8 -0
+static int auo_panel_init(struct auo_panel *auo)
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
+     dsi->mode_flags |= MIPI_DSI_MODE_LPM;
+
+     ret = mipi_dsi_generic_write(dsi, (u8[]){ 0xb0, 0x04 }, 2);

I find this notation hard to read. Have you considered moving this into some sort of table that you can loop through? Or perhaps add some helpers, say, mipi_dsi_generic_writeb() and mipi_dsi_dcs_writeb() to help make this more readable?

Yeah, helper macro thing might be a reasonable idea. The table option makes it hard to use the helpers for things that are not non-standard, or when you need delays, etc..

Quoted 104 lines
Quoted 81 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +81 -0
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0xe0 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xb5, (u8[]){ 0x86 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xb6, (u8[]){ 0x77 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xb8, (u8[]){ 0xad }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x20 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_enter_sleep_mode(dsi);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x24 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xfb, (u8[]){ 0x01 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xc6, (u8[]){ 0x00 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xc5, (u8[]){ 0x32 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0x92, (u8[]){ 0x92 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xff, (u8[]){ 0x10 }, 1);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_set_tear_on(dsi, MIPI_DSI_DCS_TEAR_MODE_VBLANK);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, MIPI_DCS_SET_TEAR_SCANLINE,
+                     (u8[]){ 0x03, 0x00 }, 2);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0x3b, (u8[]){ 0x03, 0x30, 0x06 }, 3);
+     if (ret < 0)
+             return ret;
+
+     ret = mipi_dsi_dcs_write(dsi, 0xbb, (u8[]){ 0x10 }, 1);
+     if (ret < 0)
+             return ret;
+     msleep(1);
+
+     ret = mipi_dsi_dcs_exit_sleep_mode(dsi);
+     if (ret < 0)
+             return ret;
+     msleep(30);

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 148

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+
+     return 0;
+}
+
+static int auo_panel_on(struct auo_panel *auo)
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
+     dsi->mode_flags |= MIPI_DSI_MODE_LPM;

This is weird.

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +10 -0
+     ret = mipi_dsi_dcs_set_display_on(dsi);
+     if (ret < 0)
+             return ret;
+
+     msleep(40);
+
+     return 0;
+}
+
+static int auo_panel_off(struct auo_panel *auo)
Quoted 8 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 164

Quoted 5 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
+     dsi->mode_flags &= ~MIPI_DSI_MODE_LPM;

And this even more. Doesn't the panel work when you simply send everything in low-power mode?

I wouldn't expect low power mode to have enough bandwidth for the video signal.. but otoh it seems like I need to use lpm for power-on/off sequence. Maybe we should wrap that up in a helper to enable/disable lpm? But that seemed a bit overkill.

Quoted 43 lines
Quoted 43 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +43 -0
+static int auo_panel_prepare(struct drm_panel *panel)
+{
+     struct auo_panel *auo = to_auo_panel(panel);
+     int ret;
+
+     if (auo->prepared)
+             return 0;
+
+     DRM_DEBUG("prepare\n");
+
+     if (auo->reset_gpio) {
+             gpiod_set_value(auo->reset_gpio, 0);
+             msleep(5);
+     }
+
+     ret = regulator_enable(auo->supply);
+     if (ret < 0)
+             return ret;
+
+     msleep(20);
+
+     if (auo->reset_gpio) {
+             gpiod_set_value(auo->reset_gpio, 1);
+             msleep(10);
+     }
+
+     msleep(150);
+
+     ret = auo_panel_init(auo);
+     if (ret) {
+             dev_err(panel->dev, "failed to init panel: %d\n", ret);
+             goto poweroff;
+     }
+
+     ret = auo_panel_on(auo);
+     if (ret) {
+             dev_err(panel->dev, "failed to set panel on: %d\n", ret);
+             goto poweroff;
+     }
+
+     auo->prepared = true;
+
+     return 0;
Quoted 8 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 274

Quoted 5 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+
+poweroff:
+     regulator_disable(auo->supply);
+     if (auo->reset_gpio)
+             gpiod_set_value(auo->reset_gpio, 0);

You should be able to do without the check here, because gpiod_set_value() will simply nop if the GPIO is NULL.

ok

Quoted 26 lines

I assume you may not want to do it above because of the additional delays that are only relevant if you have a reset GPIO.

Quoted 23 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +23 -0
+     return ret;
+}
+
+static int auo_panel_enable(struct drm_panel *panel)
+{
+     struct auo_panel *auo = to_auo_panel(panel);
+
+     if (auo->enabled)
+             return 0;
+
+     DRM_DEBUG("enable\n");
+
+     if (auo->backlight) {
+             auo->backlight->props.power = FB_BLANK_UNBLANK;
+             backlight_update_status(auo->backlight);
+     }
+
+     auo->enabled = true;
+
+     return 0;
+}
+
+static int auo_panel_add(struct auo_panel *auo)
Quoted 8 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 352

Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +6 -0
+{
+     struct device *dev= &auo->dsi->dev;
+     struct device_node *np;
+     int ret;
+
+     auo->mode = &default_mode;

This seems to be unused.

yeah, I think just cargo cult'd from sharp panel..

Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +6 -0
+
+     auo->supply = devm_regulator_get(dev, "power");
+     if (IS_ERR(auo->supply))
+             return PTR_ERR(auo->supply);
+
+     auo->reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_LOW);
Quoted 9 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 364

Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +6 -0
+     if (IS_ERR(auo->reset_gpio)) {
+             dev_err(dev, "cannot get reset-gpios %ld\n",
+                     PTR_ERR(auo->reset_gpio));
+             auo->reset_gpio = NULL;
+     } else {
+             gpiod_direction_output(auo->reset_gpio, 0);

Isn't that what GPIOD_OUT_LOW already does?

hmm, maybe?

BR,
-R

#3 lore raw
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[..]

Quoted 11 lines
Quoted 5 lines
Quoted 1 line

+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

We're working on it.

Quoted 5 lines
Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

An alternative would be to make it a bool property, to indicate video mode - following how the framework is implemented.

[..]

Quoted 16 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+static int auo_panel_init(struct auo_panel *auo)
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
Quoted 9 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 57

Quoted 3 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +3 -0
+     dsi->mode_flags |= MIPI_DSI_MODE_LPM;
+
+     ret = mipi_dsi_generic_write(dsi, (u8[]){ 0xb0, 0x04 }, 2);

I find this notation hard to read. Have you considered moving this into some sort of table that you can loop through? Or perhaps add some helpers, say, mipi_dsi_generic_writeb() and mipi_dsi_dcs_writeb() to help make this more readable?

Quoted 4 lines

Yeah, helper macro thing might be a reasonable idea. The table option makes it hard to use the helpers for things that are not non-standard, or when you need delays, etc..

I agree with you here, we don't want lists of data that the driver has to interpret into writes (of various types) and delays.

Quoted 4 lines
Quoted 4 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 60

Quoted 4 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +3 -0
+     if (ret < 0)
+             return ret;
+

Regards,
Bjorn

author
#4 lore raw
Quoted 11 lines · Bjorn Andersson
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[..]

Quoted 6 lines
Quoted 5 lines
Quoted 1 line

+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Quoted 7 lines
Quoted 5 lines

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

We're working on it.

Cool, thx

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

An alternative would be to make it a bool property, to indicate video mode - following how the framework is implemented.

I was debating both approaches.. I think I ended up going with the compat name so that we could, if wanted, have the two split out into different drivers, depending on how much was in common. Not even really sure if that is worth worrying about or not.

BR,
-R

Quoted 32 lines

[..]

Quoted 16 lines
Quoted 6 lines
Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+static int auo_panel_init(struct auo_panel *auo)
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
Quoted 9 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 57

Quoted 3 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +3 -0
+     dsi->mode_flags |= MIPI_DSI_MODE_LPM;
+
+     ret = mipi_dsi_generic_write(dsi, (u8[]){ 0xb0, 0x04 }, 2);

I find this notation hard to read. Have you considered moving this into some sort of table that you can loop through? Or perhaps add some helpers, say, mipi_dsi_generic_writeb() and mipi_dsi_dcs_writeb() to help make this more readable?

Quoted 4 lines

Yeah, helper macro thing might be a reasonable idea. The table option makes it hard to use the helpers for things that are not non-standard, or when you need delays, etc..

I agree with you here, we don't want lists of data that the driver has to interpret into writes (of various types) and delays.

Quoted 4 lines
Quoted 4 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 60

Quoted 4 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +3 -0
+     if (ret < 0)
+             return ret;
+

Regards,
Bjorn

#5 lore raw
In reply to Bjorn Andersson
Quoted 19 lines · Bjorn Andersson
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[..]

Quoted 11 lines
Quoted 5 lines
Quoted 1 line

+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

We're working on it.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

An alternative would be to make it a bool property, to indicate video mode - following how the framework is implemented.

Please, let's not do either. This doesn't belong in the compatible string. The compatible string specifies the panel, and the panel supports both video and command modes. That's implied by the compatible string.

Which mode to use is a configuration or policy decision and therefore doesn't belong in device tree. It should be up to the display driver to determine what the preferred mode of operation is.

Thierry

author
#6 lore raw
Quoted 29 lines · Thierry Reding
Quoted 19 lines · Bjorn Andersson
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[..]

Quoted 11 lines
Quoted 5 lines
Quoted 1 line

+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

We're working on it.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

An alternative would be to make it a bool property, to indicate video mode - following how the framework is implemented.

Quoted 8 lines

Please, let's not do either. This doesn't belong in the compatible string. The compatible string specifies the panel, and the panel supports both video and command modes. That's implied by the compatible string.

Which mode to use is a configuration or policy decision and therefore doesn't belong in device tree. It should be up to the display driver to determine what the preferred mode of operation is.

I would call it a "system integrator decision".. and at least
currently we don't have anywhere better than DT for that. Maybe it's one of those "if all you have is a hammer, everything looks like a nail" things..

BR,
-R

Quoted 1 line

Thierry

#7 lore raw
Quoted 36 lines · Rob Clark
Quoted 34 lines · Thierry Reding
Quoted 19 lines · Bjorn Andersson
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[..]

Quoted 11 lines
Quoted 5 lines
Quoted 1 line

+- compatible: should be "auo,novatek-1080p-vid"

This looks a little generic for a compatible string. Can't we get at the specific panel model number that's been used? What if AUO ever produced some other Novatek panel with a 1080p resolution?

Maybe Sony or someone else can chime in? That somewhat generic name was all I could get from downstream android kernel. I'm sure there is a better possible name, although I have no means to find that out myself.

We're working on it.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Also, what's the -vid suffix for?

the same panel seems to also work in cmd mode.. so idea was to have -vid and -cmd compat strings to choose which mode to operate in.

An alternative would be to make it a bool property, to indicate video mode - following how the framework is implemented.

Please, let's not do either. This doesn't belong in the compatible string. The compatible string specifies the panel, and the panel supports both video and command modes. That's implied by the compatible string.

Quoted 8 lines
Quoted 3 lines

Which mode to use is a configuration or policy decision and therefore doesn't belong in device tree. It should be up to the display driver to determine what the preferred mode of operation is.

I would call it a "system integrator decision".. and at least
currently we don't have anywhere better than DT for that. Maybe it's one of those "if all you have is a hammer, everything looks like a nail" things..

If it were a system integrator decision then I'd agree it should be parameterizable in DT. But to my knowledge you can run any command mode capable display in video mode just fine.

Thierry

#8 lore raw
In reply to Rob Clark
Quoted 2 lines · Rob Clark
Quoted 1 line · Thierry Reding

On Tue, Jul 21, 2015 at 03:36:02PM -0400, Rob Clark wrote:

[...]

Quoted 28 lines
Quoted 19 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 148

Quoted 6 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c new file +6 -0
+static int auo_panel_on(struct auo_panel *auo)
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
+     dsi->mode_flags |= MIPI_DSI_MODE_LPM;

This is weird.

Quoted 10 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c new file +10 -0
+     ret = mipi_dsi_dcs_set_display_on(dsi);
+     if (ret < 0)
+             return ret;
+
+     msleep(40);
+
+     return 0;
+}
+
+static int auo_panel_off(struct auo_panel *auo)
Quoted 8 lines

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 164

Quoted 5 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c new file +5 -0
+{
+     struct mipi_dsi_device *dsi = auo->dsi;
+     int ret;
+
+     dsi->mode_flags &= ~MIPI_DSI_MODE_LPM;

And this even more. Doesn't the panel work when you simply send everything in low-power mode?

Quoted 4 lines

I wouldn't expect low power mode to have enough bandwidth for the video signal.. but otoh it seems like I need to use lpm for power-on/off sequence. Maybe we should wrap that up in a helper to enable/disable lpm? But that seemed a bit overkill.

I think there's a misunderstanding, which arguable might stem from a lack of documentation. The intention for MIPI_DSI_MODE_LPM was to be used in conjunction with "host-driven" command mode.

Perhaps I should elaborate on the vocabulary here: Tegra supports two types of command mode: "host-driven" and "DC-driven". Host driven command mode is used to perform panel setup (using DCS and vendor- specific commands). "DC-driven" command mode is used to update the framebuffer using write_memory_start and write_memory_continue DCS commands directly generated by the DSI controller.

In the latter case you'd obviously want to run in high-speed mode to achieve the throughput necessary to drive you panel at the requested resolution and framerate. In the former your device should be able to receive command in both high speed and low power modes. However some hardware is known not to work with high speed command transmission. MIPI_DSI_MODE_LPM is targetted at these cases, so that display drivers know not to attempt high-speed transmission of initial command packets.

Note how MIPI_DSI_MODE_LPM translates to MIPI_DSI_MSG_USE_LPM when transferring messages (see mipi_dsi_device_transfer()). Looking at the comment for the MIPI_DSI_MODE_LPM definition I realize that it isn't very precise, but I have trouble coming up with anything better. Perhaps:

	/* transmit command messages (non-video data) in low power mode */
	#define MIPI_DSI_MODE_LPM		BIT(11)

Any ideas?

On a semi-related note, some of the other flags are rather badly documented. I do see that both Exynos and MSM implement most of these (specifically the MIPI_DSI_MODE_VIDEO_H* ones), perhaps the comments for all of those should be revisited. Ideally they'd be annotated with a reference to the spec, like we do for MIPI_DSI_CLOCK_NON_CONTINUOUS.

Thierry

#9 lore raw
In reply to Rob Clark

On Tue 21 Jul 12:36 PDT 2015, Rob Clark wrote:

[..]

Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt line 7

Quoted 9 lines

+++ b/Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt

Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +7 -0
@@ -0,0 +1,25 @@
+AU Optronics Corporation 1080x1920 DSI panel
+
+This panel supports both video and command mode (although currently only video
+mode is implemented in the driver.
+
+Required properties:
+- compatible: should be "auo,novatek-1080p-vid"

The panel name is AUO H515DAN02.0

Quoted 6 lines
Documentation/devicetree/bindings/panel/auo,novatek-1080p.txt new file +6 -0
+
+Optional properties:
+- power-supply: phandle of the regulator that provides the supply voltage
+- reset-gpio: phandle of gpio for reset line
+- backlight: phandle of the backlight device attached to the panel
+

[..]

Quoted 1 line
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +0 -0

[..]

Quoted 3 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +3 -0
+
+static int auo_panel_init(struct auo_panel *auo)
+{

[..]

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 130

Quoted 5 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +5 -0
+	ret = mipi_dsi_dcs_write(dsi, 0x3b, (u8[]){ 0x03, 0x30, 0x06 }, 3);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_dcs_write(dsi, 0xbb, (u8[]){ 0x10 }, 1);

This should be 0x3 for video mode and 0x10 for command mode.

drivers/gpu/drm/panel/panel-auo-novatek-1080p.c line 142

Quoted 12 lines
drivers/gpu/drm/panel/panel-auo-novatek-1080p.c +12 -0
+	if (ret < 0)
+		return ret;
+	msleep(1);
+
+	ret = mipi_dsi_dcs_exit_sleep_mode(dsi);
+	if (ret < 0)
+		return ret;
+	msleep(30);
+
+	return 0;
+}
+

Regards,
Bjorn