tegra: Add display driver and LCD support for Seaboard

This binding will apparently soon be in linux-next. Bring it in now since we need to do something, and may as well try to target what Linux will have.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v3:
- Add new commit for pwm binding and node

 arch/arm/dts/tegra20.dtsi                    |    7 +++++++
 doc/device-tree-bindings/pwm/tegra20-pwm.txt |   18 ++++++++++++++++++
 2 files changed, 25 insertions(+), 0 deletions(-)
 create mode 100644 doc/device-tree-bindings/pwm/tegra20-pwm.txt

arch/arm/dts/tegra20.dtsi +7 -0

index d936b1e..3221bc9 100644
@@ -211,4 +211,11 @@
 		compatible = "nvidia,tegra20-nand";
 		reg = <0x70008000 0x100>;
 	};
+
+	pwm: pwm@7000a000 {
+		compatible = "nvidia,tegra20-pwm";
+		reg = <0x7000a000 0x100>;
+		#pwm-cells = <2>;
+	};
+
 };

doc/device-tree-bindings/pwm/tegra20-pwm.txt new file +0 -0

new file mode 100644
index 0000000..bbbeedb

Quoted 42 lines · Simon Glass

This binding will apparently soon be in linux-next. Bring it in now since we need to do something, and may as well try to target what Linux will have.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v3:
- Add new commit for pwm binding and node

 arch/arm/dts/tegra20.dtsi                    |    7 +++++++
 doc/device-tree-bindings/pwm/tegra20-pwm.txt |   18 ++++++++++++++++++
 2 files changed, 25 insertions(+), 0 deletions(-)
 create mode 100644 doc/device-tree-bindings/pwm/tegra20-pwm.txt

arch/arm/dts/tegra20.dtsi +7 -0

index d936b1e..3221bc9 100644
@@ -211,4 +211,11 @@
              compatible = "nvidia,tegra20-nand";
              reg = <0x70008000 0x100>;
      };
+
+     pwm: pwm@7000a000 {
+             compatible = "nvidia,tegra20-pwm";
+             reg = <0x7000a000 0x100>;
+             #pwm-cells = <2>;
+     };
+
 };

doc/device-tree-bindings/pwm/tegra20-pwm.txt new file +6 -0

new file mode 100644
index 0000000..bbbeedb
@@ -0,0 +1,18 @@
+Tegra SoC PWFM controller
+
+Required properties:
+- compatible: should be one of:
+  - "nvidia,tegra20-pwm"
+  - "nvidia,tegra30-pwm"

doc/device-tree-bindings/pwm/tegra20-pwm.txt line 10

Quoted 4 lines

doc/device-tree-bindings/pwm/tegra20-pwm.txt new file +4 -0

+- reg: physical base address and length of the controller's registers
+- #pwm-cells: On Tegra the number of cells used to specify a PWM is 2. The
+  first cell specifies the per-chip index of the PWM to use and the second
+  cell is the duty cycle in nanoseconds.

This is a mistake in the Linux kernel documentation. The second cell specifies the period in nanoseconds, not the duty cycle. I'll fix this up in the kernel.

Thierry

Add LCD definitions and also a proposed binding for LCD displays.

The PWM is as per what will likely be committed to linux-next soon.

The displaymode binding comes from a proposal here:

http://lists.freedesktop.org/archives/dri-devel/2012-July/024875.html

The panel binding is new, and fills a need to specify the panel timings and other tegra-specific information. Should a binding appear that allows the pwm to handle this automatically, we can revisit this.

Add LCD definitions and also a proposed binding for LCD displays.

The PWM is as per what will likely be committed to linux-next soon.

The displaymode binding comes from a proposal here:

http://lists.freedesktop.org/archives/dri-devel/2012-July/024875.html

The panel binding is new, and fills a need to specify the panel timings and other tegra-specific information. Should a binding appear that allows the pwm to handle this automatically, we can revisit this.

arch/arm/dts/tegra20.dtsi +6 -0

 
+	host1x {
+		compatible = "nvidia,tegra20-host1x", "simple-bus";
+		reg = <0x50000000 0x00024000>;
+		interrupts = <0 65 0x04   /* mpcore syncpt */
+			      0 67 0x04>; /* mpcore general */
+

Quoted 21 lines · Simon Glass

Add LCD definitions and also a proposed binding for LCD displays.

The PWM is as per what will likely be committed to linux-next soon.

The displaymode binding comes from a proposal here:

http://lists.freedesktop.org/archives/dri-devel/2012-July/024875.html

The panel binding is new, and fills a need to specify the panel timings and other tegra-specific information. Should a binding appear that allows the pwm to handle this automatically, we can revisit this.

arch/arm/dts/tegra20.dtsi +6 -0

 
+     host1x {
+             compatible = "nvidia,tegra20-host1x", "simple-bus";
+             reg = <0x50000000 0x00024000>;
+             interrupts = <0 65 0x04   /* mpcore syncpt */
+                           0 67 0x04>; /* mpcore general */
+

arch/arm/dts/tegra20.dtsi line 236

Quoted 10 lines

arch/arm/dts/tegra20.dtsi +10 -0

+             #address-cells = <1>;
+             #size-cells = <1>;
+
+             ranges = <0x54000000 0x54000000 0x04000000>;
+
+             /* video-encoding/decoding */
+             mpe {
+                     reg = <0x54040000 0x00040000>;
+                     interrupts = <0 68 0x04>;
+             };

Shouldn't all of these nodes have status="disabled", since in general boards will want to opt-in to these modules. In fact, many of these nodes won't end up (ever?) being used in U-Boot; perhaps we should only add the nodes we care about.

arch/arm/dts/tegra20.dtsi line 276

Quoted 10 lines

arch/arm/dts/tegra20.dtsi +10 -0

+             /* display controllers */
+             dc@54200000 {
+                     compatible = "nvidia,tegra20-dc";
+                     reg = <0x54200000 0x00040000>;
+                     interrupts = <0 73 0x04>;
+
+                     rgb {
+                             status = "disabled";
+                     };
+             };

I think we definitely want status="disabled" in the dc nodes themselves, since we definitely want boards in U-Boot to control whether to enable the dc.

board/nvidia/dts/tegra20-seaboard.dts +6 -0

+			rgb {
+				status = "okay";
+				nvidia,panel = <&lcd_panel>;
+			};
+		};
+	};

board/nvidia/dts/tegra20-seaboard.dts line 168

Quoted 7 lines · Simon Glass

The Seaboard has a 1366x768 16bpp LCD. The backlight is controlled by one of the PWMs.

board/nvidia/dts/tegra20-seaboard.dts +2 -0

 
+     host1x {
+             dc@54200000 {

So based on my previous comment, I think you want status="okay" at this level too, perhaps even at the host1x level.

board/nvidia/dts/tegra20-seaboard.dts line 174

Quoted 7 lines

board/nvidia/dts/tegra20-seaboard.dts +6 -0

+                     rgb {
+                             status = "okay";
+                             nvidia,panel = <&lcd_panel>;
+                     };
+             };
+     };

This driver supports driving a single LCD and providing a U-Boot console on it.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v2:
- Update LCD driver to deal with new fdt bindings

Changes in v3:
- Move some fdt decode code from LCD driver to SOC display driver
- Rename fdt config structures
- Use new pwm binding from pre-linux-next
- Adjust LCD driver to use new SOC display driver structures

Changes in v4:
- Rebase on top of new tegra common directory structure
- Change LCD panel GPIOs to be entirely optional
- Only indicate that LCD init is complete when all stages complete

 drivers/video/Makefile |    1 +
 drivers/video/tegra.c  |  361
++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 362
insertions(+), 0 deletions(-)
 create mode 100644 drivers/video/tegra.c

drivers/video/Makefile +1 -0

index ebb6da8..78a7668 100644
@@ -49,6 +49,7 @@ COBJS-$(CONFIG_VIDEO_OMAP3) += omap3_dss.o
 COBJS-$(CONFIG_VIDEO_SED13806) += sed13806.o
 COBJS-$(CONFIG_VIDEO_SM501) += sm501.o
 COBJS-$(CONFIG_VIDEO_SMI_LYNXEM) += smiLynxEM.o videomodes.o
+COBJS-$(CONFIG_VIDEO_TEGRA) += tegra.o
 COBJS-$(CONFIG_VIDEO_VCXK) += bus_vcxk.o
 
 COBJS	:= $(sort $(COBJS-y))

drivers/video/tegra.c new file +141 -0

new file mode 100644
index 0000000..9df2bd3
@@ -0,0 +1,361 @@
+/*
+ * Copyright (c) 2011 The Chromium OS Authors.
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <fdtdec.h>
+#include <lcd.h>
+
+#include <asm/system.h>
+#include <asm/gpio.h>
+
+#include <asm/arch/clock.h>
+#include <asm/arch/funcmux.h>
+#include <asm/arch/pinmux.h>
+#include <asm/arch/pwm.h>
+#include <asm/arch/display.h>
+#include <asm/arch-tegra/timer.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+/* These are the stages we go throuh in enabling the LCD */
+enum stage_t {
+	STAGE_START,
+	STAGE_LVDS,
+	STAGE_BACKLIGHT_VDD,
+	STAGE_PWM,
+	STAGE_BACKLIGHT_EN,
+	STAGE_DONE,
+};
+
+static enum stage_t stage;	/* Current stage we are at */
+static unsigned long timer_next; /* Time we can move onto next stage */
+
+/* Our LCD config, set up in handle_stage() */
+static struct fdt_panel_config config;
+struct fdt_disp_config *disp_config;	/* Display controller config */
+
+enum {
+	/* Maximum LCD size we support */
+	LCD_MAX_WIDTH		= 1366,
+	LCD_MAX_HEIGHT		= 768,
+	LCD_MAX_LOG2_BPP	= 4,		/* 2^4 = 16 bpp */
+};
+
+int lcd_line_length;
+int lcd_color_fg;
+int lcd_color_bg;
+
+void *lcd_base;			/* Start of framebuffer memory	*/
+void *lcd_console_address;	/* Start of console buffer	*/
+
+short console_col;
+short console_row;
+
+vidinfo_t panel_info = {
+	/* Insert a value here so that we don't end up in the BSS */
+	.vl_col = -1,
+};
+
+char lcd_cursor_enabled;
+
+ushort lcd_cursor_width;
+ushort lcd_cursor_height;
+
+#ifndef CONFIG_OF_CONTROL
+#error "You must enable CONFIG_OF_CONTROL to get Tegra LCD support"
+#endif
+
+void lcd_cursor_size(ushort width, ushort height)
+{
+	lcd_cursor_width = width;
+	lcd_cursor_height = height;
+}
+
+void lcd_toggle_cursor(void)
+{
+	ushort x, y;
+	uchar *dest;
+	ushort row;
+
+	x = console_col * lcd_cursor_width;
+	y = console_row * lcd_cursor_height;
+	dest = (uchar *)(lcd_base + y * lcd_line_length + x * (1 << LCD_BPP) /
+			8);
+
+	for (row = 0; row < lcd_cursor_height; ++row, dest += lcd_line_length) {
+		ushort *d = (ushort *)dest;
+		ushort color;
+		int i;
+
+		for (i = 0; i < lcd_cursor_width; ++i) {
+			color = *d;
+			color ^= lcd_color_fg;
+			*d = color;
+			++d;
+		}
+	}
+}
+
+void lcd_cursor_on(void)
+{
+	lcd_cursor_enabled = 1;
+	lcd_toggle_cursor();
+}
+void lcd_cursor_off(void)
+{
+	lcd_cursor_enabled = 0;
+	lcd_toggle_cursor();
+}
+
+char lcd_is_cursor_enabled(void)
+{
+	return lcd_cursor_enabled;
+}
+
+static void update_panel_size(struct fdt_disp_config *config)
+{
+	panel_info.vl_col = config->width;
+	panel_info.vl_row = config->height;
+	panel_info.vl_bpix = config->log2_bpp;
+}
+
+/*
+ *  Main init function called by lcd driver.

drivers/video/tegra.c new file +42 -0

+	/* Make sure that we can acommodate the selected LCD */
+	assert(disp_config->width <= LCD_MAX_WIDTH);
+	assert(disp_config->height <= LCD_MAX_HEIGHT);
+	assert(disp_config->log2_bpp <= LCD_MAX_LOG2_BPP);
+	if (disp_config->width <= LCD_MAX_WIDTH
+			&& disp_config->height <= LCD_MAX_HEIGHT
+			&& disp_config->log2_bpp <= LCD_MAX_LOG2_BPP)
+		update_panel_size(disp_config);
+	size = lcd_get_size(&line_length);
+
+	debug("LCD frame buffer at %p\n", lcd_base);
+}
+
+ulong calc_fbsize(void)
+{
+	return (panel_info.vl_col * panel_info.vl_row *
+		NBITS(panel_info.vl_bpix)) / 8;
+}
+
+void lcd_setcolreg(ushort regno, ushort red, ushort green, ushort blue)
+{
+}
+
+void tegra_lcd_early_init(const void *blob)
+{
+	/*
+	 * Go with the maximum size for now. We will fix this up after
+	 * relocation. These values are only used for memory alocation.
+	 */
+	panel_info.vl_col = LCD_MAX_WIDTH;
+	panel_info.vl_row = LCD_MAX_HEIGHT;
+	panel_info.vl_bpix = LCD_MAX_LOG2_BPP;
+}
+
+/**
+ * Decode the panel information from the fdt.
+ *
+ * @param blob		fdt blob
+ * @param config	structure to store fdt config into
+ * @return 0 if ok, -ve on error
+ */
+static int fdt_decode_lcd(const void *blob, struct fdt_panel_config

*config) +{

drivers/video/tegra.c new file +39 -0

+	int display_node;
+
+	disp_config = tegra_display_get_config();
+	if (!disp_config) {
+		debug("%s: Display controller is not configured\n", __func__);
+		return -1;
+	}
+	display_node = disp_config->panel_node;
+	if (display_node < 0) {
+		debug("%s: No panel configuration available\n", __func__);
+		return -1;
+	}
+
+	config->pwm_channel = pwm_request(blob, display_node, "nvidia,pwm");
+	if (config->pwm_channel < 0) {
+		debug("%s: Unable to request PWM channel\n", __func__);
+		return -1;
+	}
+
+	config->cache_type = fdtdec_get_int(blob, display_node,
+					    "nvidia,cache-type",
+					    FDT_LCD_CACHE_WRITE_BACK_FLUSH);
+
+	/* These GPIOs are all optional */
+	fdtdec_decode_gpio(blob, display_node, "nvidia,backlight-enable-gpios",
+			    &config->backlight_en);
+	fdtdec_decode_gpio(blob, display_node, "nvidia,lvds-shutdown-gpios",
+			   &config->lvds_shutdown);
+	fdtdec_decode_gpio(blob, display_node, "nvidia,backlight-vdd-gpios",
+			   &config->backlight_vdd);
+	fdtdec_decode_gpio(blob, display_node, "nvidia,panel-vdd-gpios",
+			   &config->panel_vdd);
+
+	return fdtdec_get_int_array(blob, display_node, "nvidia,panel-timings",
+			config->panel_timings, FDT_LCD_TIMINGS);
+}
+
+/**
+ * Handle the next stage of device init

drivers/video/tegra.c new file +117 -0

+			printf("%s: Failed to probe display driver\n",
+			__func__);
+			return -1;
+		}
+
+		/* get panel details */
+		if (fdt_decode_lcd(blob, &config)) {
+			printf("No valid LCD information in device tree\n");
+			return -1;
+		}
+
+		/*
+		 * It is possible that the FDT has requested that the LCD be
+		 * disabled. We currently don't support this. It would require
+		 * changes to U-Boot LCD subsystem to have LCD support
+		 * compiled in but not used. An easier option might be to
+		 * still have a frame buffer, but leave the backlight off and
+		 * remove all mention of lcd in the stdout environment
+		 * variable.
+		 */
+
+		funcmux_select(PERIPH_ID_DISP1, FUNCMUX_DEFAULT);
+
+		fdtdec_setup_gpio(&config.panel_vdd);
+		fdtdec_setup_gpio(&config.lvds_shutdown);
+		fdtdec_setup_gpio(&config.backlight_vdd);
+		fdtdec_setup_gpio(&config.backlight_en);
+
+		/*
+		 * TODO: If fdt includes output flag we can omit this code
+		 * since fdtdec_setup_gpio will do it for us.
+		 */
+		if (fdt_gpio_isvalid(&config.panel_vdd))
+			gpio_direction_output(config.panel_vdd.gpio, 1);
+		if (fdt_gpio_isvalid(&config.lvds_shutdown))
+			gpio_direction_output(config.lvds_shutdown.gpio, 0);
+		if (fdt_gpio_isvalid(&config.backlight_vdd))
+			gpio_direction_output(config.backlight_vdd.gpio, 0);
+		if (fdt_gpio_isvalid(&config.backlight_en))
+			gpio_direction_output(config.backlight_en.gpio, 0);
+		break;
+	case STAGE_LVDS:
+		if (fdt_gpio_isvalid(&config.lvds_shutdown))
+			gpio_set_value(config.lvds_shutdown.gpio, 1);
+		break;
+	case STAGE_BACKLIGHT_VDD:
+		if (fdt_gpio_isvalid(&config.backlight_vdd))
+			gpio_set_value(config.backlight_vdd.gpio, 1);
+		break;
+	case STAGE_PWM:
+		/* Enable PWM at 15/16 high, 32768 Hz with divider 1 */
+		pinmux_set_func(PINGRP_GPU, PMUX_FUNC_PWM);
+		pinmux_tristate_disable(PINGRP_GPU);
+
+		pwm_enable(config.pwm_channel, 32768, 0xdf, 1);
+		break;
+	case STAGE_BACKLIGHT_EN:
+		if (fdt_gpio_isvalid(&config.backlight_en))
+			gpio_set_value(config.backlight_en.gpio, 1);
+		break;
+	case STAGE_DONE:
+		break;
+	}
+
+	/* set up timer for next stage */
+	timer_next = timer_get_us() + config.panel_timings[stage] * 1000;
+
+	/* move to next stage */
+	stage++;
+	return 0;
+}
+
+int tegra_lcd_check_next_stage(const void *blob, int wait)
+{
+	if (stage == STAGE_DONE)
+		return 0;
+
+	do {
+		/* wait if we need to */
+		debug("%s: stage %d\n", __func__, stage);
+		if (stage != STAGE_START) {
+			int delay = timer_next - timer_get_us();
+
+			if (delay > 0) {
+				if (wait)
+					udelay(delay);
+				else
+					return 0;
+			}
+		}
+
+		if (handle_stage(blob))
+			return -1;
+	} while (wait && stage != STAGE_DONE);
+	if (stage == STAGE_DONE)
+		debug("%s: LCD init complete\n", __func__);
+
+	return 0;
+}
+
+void lcd_enable(void)
+{
+	/*
+	 * Backlight and power init will be done separately in
+	 * tegra_lcd_check_next_stage(), which should be called in
+	 * board_late_init().
+	 *
+	 * U-Boot code supports only colour depth, selected at compile time.
+	 * The device tree setting should match this. Otherwise the display
+	 * will not look right, and U-Boot may crash.
+	 */
+	if (disp_config->log2_bpp != LCD_BPP) {
+		printf("%s: Error: LCD depth configured in FDT (%d = %dbpp)"
+			" must match setting of LCD_BPP (%d)\n", __func__,
+		       disp_config->log2_bpp, disp_config->bpp, LCD_BPP);
+	}
+}

Simon,

I tried to adapt this for paz00. I only get backlight and an all white screen.

On checking what's wrong I stumpled about something strange below...

Marc

Quoted 184 lines · On Thursday 27 September 2012 17:44:28 Simon Glass wrote:

This driver supports driving a single LCD and providing a U-Boot console on it.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v2:
- Update LCD driver to deal with new fdt bindings

Changes in v3:
- Move some fdt decode code from LCD driver to SOC display driver
- Rename fdt config structures
- Use new pwm binding from pre-linux-next
- Adjust LCD driver to use new SOC display driver structures

Changes in v4:
- Rebase on top of new tegra common directory structure
- Change LCD panel GPIOs to be entirely optional
- Only indicate that LCD init is complete when all stages complete

 drivers/video/Makefile |    1 +
 drivers/video/tegra.c  |  361
++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 362
insertions(+), 0 deletions(-)
 create mode 100644 drivers/video/tegra.c

drivers/video/Makefile +1 -0

index ebb6da8..78a7668 100644
@@ -49,6 +49,7 @@ COBJS-$(CONFIG_VIDEO_OMAP3) += omap3_dss.o
 COBJS-$(CONFIG_VIDEO_SED13806) += sed13806.o
 COBJS-$(CONFIG_VIDEO_SM501) += sm501.o
 COBJS-$(CONFIG_VIDEO_SMI_LYNXEM) += smiLynxEM.o videomodes.o
+COBJS-$(CONFIG_VIDEO_TEGRA) += tegra.o
 COBJS-$(CONFIG_VIDEO_VCXK) += bus_vcxk.o
 
 COBJS        := $(sort $(COBJS-y))

drivers/video/tegra.c new file +141 -0

new file mode 100644
index 0000000..9df2bd3
@@ -0,0 +1,361 @@
+/*
+ * Copyright (c) 2011 The Chromium OS Authors.
+ * See file CREDITS for list of people who contributed to this
+ * project.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston,
+ * MA 02111-1307 USA
+ */
+
+#include <common.h>
+#include <fdtdec.h>
+#include <lcd.h>
+
+#include <asm/system.h>
+#include <asm/gpio.h>
+
+#include <asm/arch/clock.h>
+#include <asm/arch/funcmux.h>
+#include <asm/arch/pinmux.h>
+#include <asm/arch/pwm.h>
+#include <asm/arch/display.h>
+#include <asm/arch-tegra/timer.h>
+
+DECLARE_GLOBAL_DATA_PTR;
+
+/* These are the stages we go throuh in enabling the LCD */
+enum stage_t {
+     STAGE_START,
+     STAGE_LVDS,
+     STAGE_BACKLIGHT_VDD,
+     STAGE_PWM,
+     STAGE_BACKLIGHT_EN,
+     STAGE_DONE,
+};
+
+static enum stage_t stage;   /* Current stage we are at */
+static unsigned long timer_next; /* Time we can move onto next stage */
+
+/* Our LCD config, set up in handle_stage() */
+static struct fdt_panel_config config;
+struct fdt_disp_config *disp_config; /* Display controller config */
+
+enum {
+     /* Maximum LCD size we support */
+     LCD_MAX_WIDTH           = 1366,
+     LCD_MAX_HEIGHT          = 768,
+     LCD_MAX_LOG2_BPP        = 4,            /* 2^4 = 16 bpp */
+};
+
+int lcd_line_length;
+int lcd_color_fg;
+int lcd_color_bg;
+
+void *lcd_base;                      /* Start of framebuffer memory  */
+void *lcd_console_address;   /* Start of console buffer      */
+
+short console_col;
+short console_row;
+
+vidinfo_t panel_info = {
+     /* Insert a value here so that we don't end up in the BSS */
+     .vl_col = -1,
+};
+
+char lcd_cursor_enabled;
+
+ushort lcd_cursor_width;
+ushort lcd_cursor_height;
+
+#ifndef CONFIG_OF_CONTROL
+#error "You must enable CONFIG_OF_CONTROL to get Tegra LCD support"
+#endif
+
+void lcd_cursor_size(ushort width, ushort height)
+{
+     lcd_cursor_width = width;
+     lcd_cursor_height = height;
+}
+
+void lcd_toggle_cursor(void)
+{
+     ushort x, y;
+     uchar *dest;
+     ushort row;
+
+     x = console_col * lcd_cursor_width;
+     y = console_row * lcd_cursor_height;
+     dest = (uchar *)(lcd_base + y * lcd_line_length + x * (1 << LCD_BPP) /
+                     8);
+
+     for (row = 0; row < lcd_cursor_height; ++row, dest += lcd_line_length) {
+             ushort *d = (ushort *)dest;
+             ushort color;
+             int i;
+
+             for (i = 0; i < lcd_cursor_width; ++i) {
+                     color = *d;
+                     color ^= lcd_color_fg;
+                     *d = color;
+                     ++d;
+             }
+     }
+}
+
+void lcd_cursor_on(void)
+{
+     lcd_cursor_enabled = 1;
+     lcd_toggle_cursor();
+}
+void lcd_cursor_off(void)
+{
+     lcd_cursor_enabled = 0;
+     lcd_toggle_cursor();
+}
+
+char lcd_is_cursor_enabled(void)
+{
+     return lcd_cursor_enabled;
+}
+
+static void update_panel_size(struct fdt_disp_config *config)
+{
+     panel_info.vl_col = config->width;
+     panel_info.vl_row = config->height;
+     panel_info.vl_bpix = config->log2_bpp;
+}
+
+/*
+ *  Main init function called by lcd driver.

drivers/video/tegra.c line 151

Quoted 10 lines

drivers/video/tegra.c new file +10 -0

+ *  Inits and then prints test pattern if required.
+ */
+
+void lcd_ctrl_init(void *lcdbase)
+{
+     int line_length, size;
+
+     assert(disp_config);
+
+     lcd_base = (void *)disp_config->frame_buffer;

this is always NULL, because it is defined in a circular manner (see arch/arm/cpu/armv7/tegra20/display.c:383) and ...

Quoted 86 lines

drivers/video/tegra.c new file +42 -0

+     /* Make sure that we can acommodate the selected LCD */
+     assert(disp_config->width <= LCD_MAX_WIDTH);
+     assert(disp_config->height <= LCD_MAX_HEIGHT);
+     assert(disp_config->log2_bpp <= LCD_MAX_LOG2_BPP);
+     if (disp_config->width <= LCD_MAX_WIDTH
+                     && disp_config->height <= LCD_MAX_HEIGHT
+                     && disp_config->log2_bpp <= LCD_MAX_LOG2_BPP)
+             update_panel_size(disp_config);
+     size = lcd_get_size(&line_length);
+
+     debug("LCD frame buffer at %p\n", lcd_base);
+}
+
+ulong calc_fbsize(void)
+{
+     return (panel_info.vl_col * panel_info.vl_row *
+             NBITS(panel_info.vl_bpix)) / 8;
+}
+
+void lcd_setcolreg(ushort regno, ushort red, ushort green, ushort blue)
+{
+}
+
+void tegra_lcd_early_init(const void *blob)
+{
+     /*
+      * Go with the maximum size for now. We will fix this up after
+      * relocation. These values are only used for memory alocation.
+      */
+     panel_info.vl_col = LCD_MAX_WIDTH;
+     panel_info.vl_row = LCD_MAX_HEIGHT;
+     panel_info.vl_bpix = LCD_MAX_LOG2_BPP;
+}
+
+/**
+ * Decode the panel information from the fdt.
+ *
+ * @param blob               fdt blob
+ * @param config     structure to store fdt config into
+ * @return 0 if ok, -ve on error
+ */
+static int fdt_decode_lcd(const void *blob, struct fdt_panel_config

*config) +{

drivers/video/tegra.c new file +43 -0

+     int display_node;
+
+     disp_config = tegra_display_get_config();
+     if (!disp_config) {
+             debug("%s: Display controller is not configured\n", __func__);
+             return -1;
+     }
+     display_node = disp_config->panel_node;
+     if (display_node < 0) {
+             debug("%s: No panel configuration available\n", __func__);
+             return -1;
+     }
+
+     config->pwm_channel = pwm_request(blob, display_node, "nvidia,pwm");
+     if (config->pwm_channel < 0) {
+             debug("%s: Unable to request PWM channel\n", __func__);
+             return -1;
+     }
+
+     config->cache_type = fdtdec_get_int(blob, display_node,
+                                         "nvidia,cache-type",
+                                         FDT_LCD_CACHE_WRITE_BACK_FLUSH);
+
+     /* These GPIOs are all optional */
+     fdtdec_decode_gpio(blob, display_node, "nvidia,backlight-enable-gpios",
+                         &config->backlight_en);
+     fdtdec_decode_gpio(blob, display_node, "nvidia,lvds-shutdown-gpios",
+                        &config->lvds_shutdown);
+     fdtdec_decode_gpio(blob, display_node, "nvidia,backlight-vdd-gpios",
+                        &config->backlight_vdd);
+     fdtdec_decode_gpio(blob, display_node, "nvidia,panel-vdd-gpios",
+                        &config->panel_vdd);
+
+     return fdtdec_get_int_array(blob, display_node, "nvidia,panel-timings",
+                     config->panel_timings, FDT_LCD_TIMINGS);
+}
+
+/**
+ * Handle the next stage of device init
+ */
+static int handle_stage(const void *blob)
+{
+     debug("%s: stage %d\n", __func__, stage);

drivers/video/tegra.c line 244

Quoted 6 lines

drivers/video/tegra.c new file +6 -0

+
+     /* do the things for this stage */
+     switch (stage) {
+     case STAGE_START:
+             /* Initialize the Tegra display controller */
+             if (tegra_display_probe(gd->fdt_blob, lcd_base)) {

.... here.

Quoted 110 lines

drivers/video/tegra.c new file +109 -0

+                     printf("%s: Failed to probe display driver\n",
+                     __func__);
+                     return -1;
+             }
+
+             /* get panel details */
+             if (fdt_decode_lcd(blob, &config)) {
+                     printf("No valid LCD information in device tree\n");
+                     return -1;
+             }
+
+             /*
+              * It is possible that the FDT has requested that the LCD be
+              * disabled. We currently don't support this. It would require
+              * changes to U-Boot LCD subsystem to have LCD support
+              * compiled in but not used. An easier option might be to
+              * still have a frame buffer, but leave the backlight off and
+              * remove all mention of lcd in the stdout environment
+              * variable.
+              */
+
+             funcmux_select(PERIPH_ID_DISP1, FUNCMUX_DEFAULT);
+
+             fdtdec_setup_gpio(&config.panel_vdd);
+             fdtdec_setup_gpio(&config.lvds_shutdown);
+             fdtdec_setup_gpio(&config.backlight_vdd);
+             fdtdec_setup_gpio(&config.backlight_en);
+
+             /*
+              * TODO: If fdt includes output flag we can omit this code
+              * since fdtdec_setup_gpio will do it for us.
+              */
+             if (fdt_gpio_isvalid(&config.panel_vdd))
+                     gpio_direction_output(config.panel_vdd.gpio, 1);
+             if (fdt_gpio_isvalid(&config.lvds_shutdown))
+                     gpio_direction_output(config.lvds_shutdown.gpio, 0);
+             if (fdt_gpio_isvalid(&config.backlight_vdd))
+                     gpio_direction_output(config.backlight_vdd.gpio, 0);
+             if (fdt_gpio_isvalid(&config.backlight_en))
+                     gpio_direction_output(config.backlight_en.gpio, 0);
+             break;
+     case STAGE_LVDS:
+             if (fdt_gpio_isvalid(&config.lvds_shutdown))
+                     gpio_set_value(config.lvds_shutdown.gpio, 1);
+             break;
+     case STAGE_BACKLIGHT_VDD:
+             if (fdt_gpio_isvalid(&config.backlight_vdd))
+                     gpio_set_value(config.backlight_vdd.gpio, 1);
+             break;
+     case STAGE_PWM:
+             /* Enable PWM at 15/16 high, 32768 Hz with divider 1 */
+             pinmux_set_func(PINGRP_GPU, PMUX_FUNC_PWM);
+             pinmux_tristate_disable(PINGRP_GPU);
+
+             pwm_enable(config.pwm_channel, 32768, 0xdf, 1);
+             break;
+     case STAGE_BACKLIGHT_EN:
+             if (fdt_gpio_isvalid(&config.backlight_en))
+                     gpio_set_value(config.backlight_en.gpio, 1);
+             break;
+     case STAGE_DONE:
+             break;
+     }
+
+     /* set up timer for next stage */
+     timer_next = timer_get_us() + config.panel_timings[stage] * 1000;
+
+     /* move to next stage */
+     stage++;
+     return 0;
+}
+
+int tegra_lcd_check_next_stage(const void *blob, int wait)
+{
+     if (stage == STAGE_DONE)
+             return 0;
+
+     do {
+             /* wait if we need to */
+             debug("%s: stage %d\n", __func__, stage);
+             if (stage != STAGE_START) {
+                     int delay = timer_next - timer_get_us();
+
+                     if (delay > 0) {
+                             if (wait)
+                                     udelay(delay);
+                             else
+                                     return 0;
+                     }
+             }
+
+             if (handle_stage(blob))
+                     return -1;
+     } while (wait && stage != STAGE_DONE);
+     if (stage == STAGE_DONE)
+             debug("%s: LCD init complete\n", __func__);
+
+     return 0;
+}
+
+void lcd_enable(void)
+{
+     /*
+      * Backlight and power init will be done separately in
+      * tegra_lcd_check_next_stage(), which should be called in
+      * board_late_init().
+      *
+      * U-Boot code supports only colour depth, selected at compile time.
+      * The device tree setting should match this. Otherwise the display

drivers/video/tegra.c line 361

Quoted 8 lines

drivers/video/tegra.c new file +8 -0

+      * will not look right, and U-Boot may crash.
+      */
+     if (disp_config->log2_bpp != LCD_BPP) {
+             printf("%s: Error: LCD depth configured in FDT (%d = %dbpp)"
+                     " must match setting of LCD_BPP (%d)\n", __func__,
+                    disp_config->log2_bpp, disp_config->bpp, LCD_BPP);
+     }
+}

Hi Marc,

Quoted 3 lines · Marc Dietrich

Simon,

I tried to adapt this for paz00. I only get backlight and an all white screen.

I'm not sure about that. One possibility is that your PWM is connected up differently. On Seaboard at least, I found that without the PWM I would get nothing on the display.

When the cursor position gets to the end of the LCD console we normally scroll by one line. This adds an option to increase that value.

Console scrolling is often slow, and if a large amount of output is being sent, increasing this option to 10 or so will speed things up considerably.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v3:
- Fix tiny bug in mult-line lcd scrolling

Changes in v4:
- Use CONFIG_CONSOLE_SCROLL_LINES instead of CONSOLE_SCROLL_LINES
- Put default CONFIG_CONSOLE_SCROLL_LINES at top of file

 README       |    6 ++++++
 common/lcd.c |   18 +++++++++++++++---
 2 files changed, 21 insertions(+), 3 deletions(-)

README +6 -0

index 1d0ce50..14b1f83 100644
@@ -1449,6 +1449,12 @@ The following options need to be configured:
 		here, since it is cheaper to change data cache settings on
 		a per-section basis.
 
+		CONFIG_CONSOLE_SCROLL_LINES
+
+		When the console need to be scrolled, this is the number of
+		lines to scroll by. It defaults to 1. Increasing this makes
+		the console jump but can help speed up operation when scrolling
+		is slow.
 
 - Splash Screen Support: CONFIG_SPLASH_SCREEN
 

common/lcd.c +5 -0

index 74fa3f9..460f0ca 100644
@@ -76,6 +76,11 @@
 #define CONFIG_LCD_ALIGNMENT PAGE_SIZE
 #endif
 
+/* By default we scroll by a single line */
+#ifndef CONFIG_CONSOLE_SCROLL_LINES
+#define CONFIG_CONSOLE_SCROLL_LINES 1
+#endif
+
 DECLARE_GLOBAL_DATA_PTR;
 
 ulong lcd_setmem (ulong addr);

common/lcd.c +5 -2

-	memset(CONSOLE_ROW_LAST, COLOR_MASK(lcd_color_bg), CONSOLE_ROW_SIZE);
+	memset(lcd_console_address + CONSOLE_SIZE - CONSOLE_ROW_SIZE * rows,
+		COLOR_MASK(lcd_color_bg),
+	       CONSOLE_ROW_SIZE * rows);
+
 	lcd_sync();
+	console_row -= rows;
 }
 
 /*----------------------------------------------------------------------*/

@@ -165,7 +178,6 @@ static inline void console_newline(void)
 	if (console_row >= CONSOLE_ROWS) {
 		/* Scroll everything up */
 		console_scrollup();
-		--console_row;
 	} else {
 		lcd_sync();
 	}

Am Donnerstag, den 27.09.2012, 17:44 -0700 schrieb Simon Glass:

Quoted 54 lines

When the cursor position gets to the end of the LCD console we normally scroll by one line. This adds an option to increase that value.

Console scrolling is often slow, and if a large amount of output is being sent, increasing this option to 10 or so will speed things up considerably.

Signed-off-by Simon Glass <[email protected]>

---
Changes in v3:
- Fix tiny bug in mult-line lcd scrolling

Changes in v4:
- Use CONFIG_CONSOLE_SCROLL_LINES instead of CONSOLE_SCROLL_LINES
- Put default CONFIG_CONSOLE_SCROLL_LINES at top of file

 README       |    6 ++++++
 common/lcd.c |   18 +++++++++++++++---
 2 files changed, 21 insertions(+), 3 deletions(-)

README +6 -0

index 1d0ce50..14b1f83 100644
@@ -1449,6 +1449,12 @@ The following options need to be configured:
              here, since it is cheaper to change data cache settings on
              a per-section basis.
 
+             CONFIG_CONSOLE_SCROLL_LINES
+
+             When the console need to be scrolled, this is the number of
+             lines to scroll by. It defaults to 1. Increasing this makes
+             the console jump but can help speed up operation when scrolling
+             is slow.
 
 - Splash Screen Support: CONFIG_SPLASH_SCREEN
 

common/lcd.c +5 -0

index 74fa3f9..460f0ca 100644
@@ -76,6 +76,11 @@
 #define CONFIG_LCD_ALIGNMENT PAGE_SIZE
 #endif
 
+/* By default we scroll by a single line */
+#ifndef CONFIG_CONSOLE_SCROLL_LINES
+#define CONFIG_CONSOLE_SCROLL_LINES 1
+#endif
+
 DECLARE_GLOBAL_DATA_PTR;
 
 ulong lcd_setmem (ulong addr);

common/lcd.c void lcd_set_flush_dcache(int flush) · line 141

Quoted 6 lines

common/lcd.c +2 -0

@@ -131,12 +136,20 @@ void lcd_set_flush_dcache(int flush)
 
 static void console_scrollup(void)
 {
+     const int rows = CONFIG_CONSOLE_SCROLL_LINES;
+
      /* Copy up rows ignoring the first one */

Comment is inaccurate as you possibly ignore more than the first line.

common/lcd.c void lcd_set_flush_dcache(int flush) · line 146

Quoted 6 lines

common/lcd.c +3 -1

-     memcpy(CONSOLE_ROW_FIRST, CONSOLE_ROW_SECOND, CONSOLE_SCROLL_SIZE);
+     memcpy(CONSOLE_ROW_FIRST,
+            lcd_console_address + CONSOLE_ROW_SIZE * rows,
+            CONSOLE_SIZE - CONSOLE_ROW_SIZE * rows);
 
      /* Clear the last one */

Same as above.

Quoted 10 lines

common/lcd.c +5 -1

-     memset(CONSOLE_ROW_LAST, COLOR_MASK(lcd_color_bg), CONSOLE_ROW_SIZE);
+     memset(lcd_console_address + CONSOLE_SIZE - CONSOLE_ROW_SIZE * rows,
+             COLOR_MASK(lcd_color_bg),
+            CONSOLE_ROW_SIZE * rows);
+
      lcd_sync();
+     console_row -= rows;
 }
 

common/lcd.c static inline void console_newline(void) · line 183

Quoted 9 lines

common/lcd.c +0 -1

 /*----------------------------------------------------------------------*/

@@ -165,7 +178,6 @@ static inline void console_newline(void)
      if (console_row >= CONSOLE_ROWS) {
              /* Scroll everything up */
              console_scrollup();
-             --console_row;
      } else {
              lcd_sync();
      }

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

I tested this on Seaboard, and found a couple of minor issues:

1) Nothing sets up the environment to enable LCD output; I had to execute:

setenv stdout serial,lcd

... in order to see anything on the display. I'd expect patch 16/16 to cause this to happen automatically?

It's probably best to fix this by revising patch 16 rather than later, since I presume this is trivial to fix?

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

I tested this on Seaboard, and found a couple of minor issues:

1) Nothing sets up the environment to enable LCD output; I had to execute:

setenv stdout serial,lcd

... in order to see anything on the display. I'd expect patch 16/16 to cause this to happen automatically?

It's probably best to fix this by revising patch 16 rather than later, since I presume this is trivial to fix?

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

I'll provide review comments on the patches themselves later.

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

Quoted 4 lines · Stephen Warren

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

...

Quoted 7 lines

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

A couple more points:

BTW, I tried the same tests on Springbank which is basically Seaboard with a slightly different PCB layout, and didn't see this issue. That leads me to suspect some timing issue interacting with the panel, which I believe is different albeit theoretically compatible resolution/scan-timing wise.

Second, I first applied Tom's Tegra30 patches and tested those, then applied your LCD series. The LCD series as-is breaks compilation of the Tegra30 support. I have no idea which order Tom might apply these in, but this is a heads-up there will be conflicts and one needs to be rebased on top of the other.

-----Original Message-----
From: sjg at google.com [mailto:sjg at google.com] On Behalf Of Simon Glass
Sent: Monday, October 08, 2012 1:06 PM
To: Stephen Warren
Cc: U-Boot Mailing List; Tom Warren
Subject: Re: [PATCH v4 0/16] tegra: Add display driver and LCD support for Seaboard

Hi Stephen,

Quoted 8 lines · Stephen Warren

Quoted 4 lines · Stephen Warren

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

...

Quoted 2 lines

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or

"reboot"

Quoted 2 lines

Quoted 2 lines

within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something

similar.

Quoted 11 lines

Quoted 3 lines

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

A couple more points:

BTW, I tried the same tests on Springbank which is basically Seaboard with a slightly different PCB layout, and didn't see this issue. That leads me to suspect some timing issue interacting with the panel, which I believe is different albeit theoretically compatible resolution/scan-timing wise.

OK. I wonder the panel does not get reset properly on Seaboard?

Quoted 1 line

Quoted 6 lines

Second, I first applied Tom's Tegra30 patches and tested those, then applied your LCD series. The LCD series as-is breaks compilation of the
Tegra30 support. I have no idea which order Tom might apply these in, but this is a heads-up there will be conflicts and one needs to be rebased on top of the other.

OK, let's see what Tom wants to do on that one.

Regards,
Simon

Hi Stephen,

Quoted 16 lines · Stephen Warren

Quoted 4 lines · Stephen Warren

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

...

Quoted 7 lines

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

A couple more points:

Quoted 5 lines

BTW, I tried the same tests on Springbank which is basically Seaboard with a slightly different PCB layout, and didn't see this issue. That leads me to suspect some timing issue interacting with the panel, which I believe is different albeit theoretically compatible resolution/scan-timing wise.

OK. I wonder the panel does not get reset properly on Seaboard?

Quoted 19 lines · Simon Glass

Hi Stephen,

Quoted 16 lines · Stephen Warren

Quoted 4 lines · Stephen Warren

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

...

Quoted 7 lines

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

A couple more points:

Quoted 7 lines

Quoted 5 lines

BTW, I tried the same tests on Springbank which is basically Seaboard with a slightly different PCB layout, and didn't see this issue. That leads me to suspect some timing issue interacting with the panel, which I believe is different albeit theoretically compatible resolution/scan-timing wise.

OK. I wonder the panel does not get reset properly on Seaboard?

I looked at the panel spec for the panels on the two boards. Both of them specify that when vdd_pnl is dropped to zero, you must wait at least 400mS before raising it again. Evidently the Seaboard panel actually cares about this, yet the Springbank panel seems to be more flexible than the specification. So, modifying drivers/video/tegra.c:handle_stage() case STAGE_START to:

    if (fdt_gpio_isvalid(&config.panel_vdd)) {
        udelay(400000);
        gpio_direction_output(config.panel_vdd.gpio, 1);
    }

... appears to solve the problem; I can run "reset" in U-Boot many many times without any issue after that change.

I guess this means that the nvidia,panel-timings DT property needs a pre-delay entry for STAGE_START, or something like that?

Also, I wondered if we shouldn't explicitly do:

    gpio_direction_output(config.panel_vdd.gpio, 0);
    udelay(400000);
    gpio_direction_output(config.panel_vdd.gpio, 1);

... although that doesn't make a difference; the particular GPIO used on Seaboard appears to default to 0. I'm not sure that all GPIOs are guaranteed to though.

Slightly related, I notice that handle_stage() always executes:

timer_next = timer_get_us() + config.panel_timings[stage] * 1000;

However, that function is called for stage==0..4 (or perhaps 0..5 given the set of case statements it supports, but not looking at the code at the call site), yet there are only entries 0..3 in that array; there's an overrun there.

Hi Stephen,

Quoted 52 lines · Stephen Warren

Quoted 19 lines · Simon Glass

Hi Stephen,

Quoted 16 lines · Stephen Warren

Quoted 4 lines · Stephen Warren

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

...

Quoted 7 lines

2) The display works fine for cold boot, or hitting the physical reset button, but when I execute "reset" at the U-Boot prompt, or "reboot" within Linux, the display is messed up; it looks like the LCD isn't able to sync to the timings sent by the display controller or something similar.

Since this is slightly a corner case, it may be OK to fix it later (but not much later; reboot is very useful to me.)

A couple more points:

Quoted 7 lines

Quoted 5 lines

BTW, I tried the same tests on Springbank which is basically Seaboard with a slightly different PCB layout, and didn't see this issue. That leads me to suspect some timing issue interacting with the panel, which I believe is different albeit theoretically compatible resolution/scan-timing wise.

OK. I wonder the panel does not get reset properly on Seaboard?

I looked at the panel spec for the panels on the two boards. Both of them specify that when vdd_pnl is dropped to zero, you must wait at least 400mS before raising it again. Evidently the Seaboard panel actually cares about this, yet the Springbank panel seems to be more flexible than the specification. So, modifying drivers/video/tegra.c:handle_stage() case STAGE_START to:

    if (fdt_gpio_isvalid(&config.panel_vdd)) {
        udelay(400000);
        gpio_direction_output(config.panel_vdd.gpio, 1);
    }

... appears to solve the problem; I can run "reset" in U-Boot many many times without any issue after that change.

I guess this means that the nvidia,panel-timings DT property needs a pre-delay entry for STAGE_START, or something like that?

Also, I wondered if we shouldn't explicitly do:

    gpio_direction_output(config.panel_vdd.gpio, 0);
    udelay(400000);
    gpio_direction_output(config.panel_vdd.gpio, 1);

Quoted 3 lines

... although that doesn't make a difference; the particular GPIO used on Seaboard appears to default to 0. I'm not sure that all GPIOs are guaranteed to though.

So reset is causing the line to go low?

OK, there's probably nothing
we can do here except wait for 400ms. Is there any way to detect this condition? Adding 400ms is quite a big jump in boot time.

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

Quoted 3 lines · Simon Glass

This series adds support for the Tegra2x's display peripheral. This supports the LCD display on Seaboard and we use this to enable console output in U-Boot on the LCD.

Aside from the comments I made just now, the whole series,

Acked-by Stephen Warren <[email protected]>

At this stage, I figure it's more important to get this feature checked in and enabled than anything else. For this version of the series, I only briefly scanned the patches, and mostly concentrated on the device tree files themselves. They're obviously close enough to the proposed kernel bindings that we should be able to migrate them easily enough to the finalized bindings when they're available.