OMAPDSS: HDMI PHY burnout fix for 3.2 stable

Started by Tomi Valkeinen on Mar 1, 2012 · 20 messages from 2 authors · Last reply Mar 10, 2012

Series

7 patches, 6 files

+105 −28

Subsystems (4)

Show 1 more

Maintainers (8)

Show 4 more
STALE5361d PENDING · 5 of 7 reviewed

Status: Pending, 5 of 7 reviewed.

open the first message on its own page

  1. [PATCH 1/7] OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios
    Related patches
  2. [PATCH 2/7] OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD
    Related patches
  3. [PATCH 3/7] OMAPDSS: remove wrong HDMI HPD muxing
    Related patches
  4. [PATCH 4/7] OMAP: 4430SDP/Panda: setup HDMI GPIO muxes
    Related patches
  5. [PATCH 5/7] OMAP: 4430SDP/Panda: add HDMI HPD gpio
    Related patches
  6. [PATCH 6/7] OMAPDSS: HDMI: PHY burnout fix
    Related patches
  7. [PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix
    Related patches
  8. [PATCH 1/7] OMAP: DSS2: HDMI: use default dividers
    Related patches
  9. [PATCH 2/7] OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios
    Related patches
  10. [PATCH 3/7] OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD
    Related patches
  11. [PATCH 4/7] OMAPDSS: remove wrong HDMI HPD muxing
    Related patches
  12. [PATCH 5/7] OMAP: 4430SDP/Panda: setup HDMI GPIO muxes
    Related patches
  13. [PATCH 6/7] OMAP: 4430SDP/Panda: add HDMI HPD gpio
    Related patches
  14. [PATCH 7/7] OMAPDSS: HDMI: PHY burnout fix
    Related patches
Message details and related content
The main patch in this set is the "PHY burnout fix", which implements a fix for
a hardware bug in the OMAP4 HDMI PHY which may cause physical damage to the
board if the HDMI PHY is already enabled when the HDMI cable is plugged in. The
possible damage breaks HDMI output hardware, otherwise the board is unaffected.

The preceding patches are small cleanups/fixes for HDMI GPIOs so that the fix
can be implemented. And the fix introduced a small bug with hot plug detect,
which is fixed by the patch from Rob. Rob's patch is not yet in upstream, but
has been accepted by fbdev maintainer.

 Tomi

Rob Clark (1):
  OMAPDSS: HDMI: hot plug detect fix

Tomi Valkeinen (6):
  OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios
  OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD
  OMAPDSS: remove wrong HDMI HPD muxing
  OMAP: 4430SDP/Panda: setup HDMI GPIO muxes
  OMAP: 4430SDP/Panda: add HDMI HPD gpio
  OMAPDSS: HDMI: PHY burnout fix

 arch/arm/mach-omap2/board-4430sdp.c       |   22 +++++---
 arch/arm/mach-omap2/board-omap4panda.c    |   22 +++++---
 drivers/video/omap2/dss/hdmi.c            |    3 +
 drivers/video/omap2/dss/ti_hdmi.h         |    4 ++
 drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c |   77 ++++++++++++++++++++++++-----
 include/video/omapdss.h                   |    5 ++
 6 files changed, 105 insertions(+), 28 deletions(-)

-- 
1.7.4.1

[PATCH 1/7] OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios

Message details and related content
Instead of freeing the GPIOs individually, use gpio_free_array().

Backported from 575753e3bea3b67eef8e454fb87f719e3f7da599

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +--
 arch/arm/mach-omap2/board-omap4panda.c |    3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 5156468..8434b0c 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -628,8 +628,7 @@ static int sdp4430_panel_enable_hdmi(struct omap_dss_device *dssdev)
 
 static void sdp4430_panel_disable_hdmi(struct omap_dss_device *dssdev)
 {
-	gpio_free(HDMI_GPIO_LS_OE);
-	gpio_free(HDMI_GPIO_HPD);
+	gpio_free_array(sdp4430_hdmi_gpios, ARRAY_SIZE(sdp4430_hdmi_gpios));
 }
 
 static struct nokia_dsi_panel_data dsi1_panel = {
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index a8c2c42..1ac7a22 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -512,8 +512,7 @@ static int omap4_panda_panel_enable_hdmi(struct omap_dss_device *dssdev)
 
 static void omap4_panda_panel_disable_hdmi(struct omap_dss_device *dssdev)
 {
-	gpio_free(HDMI_GPIO_LS_OE);
-	gpio_free(HDMI_GPIO_HPD);
+	gpio_free_array(panda_hdmi_gpios, ARRAY_SIZE(panda_hdmi_gpios));
 }
 
 static struct omap_dss_device  omap4_panda_hdmi_device = {
-- 
1.7.4.1

[PATCH 2/7] OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD

Message details and related content
The GPIO 60 on 4430sdp and Panda is not HPD GPIO, as currently marked in
the board files, but CT_CP_HPD, which is used to enable/disable HPD
functionality.

This patch renames the GPIO.

Backported from 3932a32fcf5393f8be70ac99dc718ad7ad0a415b

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    4 ++--
 arch/arm/mach-omap2/board-omap4panda.c |    4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 8434b0c..4333115 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -52,7 +52,7 @@
 #define ETH_KS8851_QUART		138
 #define OMAP4_SFH7741_SENSOR_OUTPUT_GPIO	184
 #define OMAP4_SFH7741_ENABLE_GPIO		188
-#define HDMI_GPIO_HPD 60 /* Hot plug pin for HDMI */
+#define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
 #define DISPLAY_SEL_GPIO	59	/* LCD2/PicoDLP switch */
 #define DLP_POWER_ON_GPIO	40
@@ -610,7 +610,7 @@ static void sdp4430_hdmi_mux_init(void)
 }
 
 static struct gpio sdp4430_hdmi_gpios[] = {
-	{ HDMI_GPIO_HPD,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_hpd"   },
+	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_ls_oe" },
 };
 
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 1ac7a22..d8e20f1 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -51,7 +51,7 @@
 #define GPIO_HUB_NRESET		62
 #define GPIO_WIFI_PMENA		43
 #define GPIO_WIFI_IRQ		53
-#define HDMI_GPIO_HPD 60 /* Hot plug pin for HDMI */
+#define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
 
 /* wl127x BT, FM, GPS connectivity chip */
@@ -494,7 +494,7 @@ static void omap4_panda_hdmi_mux_init(void)
 }
 
 static struct gpio panda_hdmi_gpios[] = {
-	{ HDMI_GPIO_HPD,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_hpd"   },
+	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ls_oe" },
 };
 
-- 
1.7.4.1

[PATCH 3/7] OMAPDSS: remove wrong HDMI HPD muxing

Message details and related content
"hdmi_hpd" pin is muxed to INPUT and PULLUP, but the pin is not
currently used, and in the future when it is used, the pin is used as a
GPIO and is board specific, not an OMAP4 wide thing.

So remove the muxing for now.

Backported from 7bb122d155f742fe2d79849090c825be7b4a247e

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    4 ----
 arch/arm/mach-omap2/board-omap4panda.c |    4 ----
 2 files changed, 0 insertions(+), 8 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 4333115..385c1fd 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -597,12 +597,8 @@ static void __init omap_sfh7741prox_init(void)
 
 static void sdp4430_hdmi_mux_init(void)
 {
-	/* PAD0_HDMI_HPD_PAD1_HDMI_CEC */
-	omap_mux_init_signal("hdmi_hpd",
-			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_cec",
 			OMAP_PIN_INPUT_PULLUP);
-	/* PAD0_HDMI_DDC_SCL_PAD1_HDMI_DDC_SDA */
 	omap_mux_init_signal("hdmi_ddc_scl",
 			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_ddc_sda",
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index d8e20f1..e0d0abc 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -481,12 +481,8 @@ int __init omap4_panda_dvi_init(void)
 
 static void omap4_panda_hdmi_mux_init(void)
 {
-	/* PAD0_HDMI_HPD_PAD1_HDMI_CEC */
-	omap_mux_init_signal("hdmi_hpd",
-			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_cec",
 			OMAP_PIN_INPUT_PULLUP);
-	/* PAD0_HDMI_DDC_SCL_PAD1_HDMI_DDC_SDA */
 	omap_mux_init_signal("hdmi_ddc_scl",
 			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_ddc_sda",
-- 
1.7.4.1

[PATCH 4/7] OMAP: 4430SDP/Panda: setup HDMI GPIO muxes

Message details and related content
The HDMI GPIO pins LS_OE and CT_CP_HPD are not currently configured.
This patch configures them as output pins.

Backported from 78a1ad8f12db70b8b0a4548b90704de08ee216ce

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +++
 arch/arm/mach-omap2/board-omap4panda.c |    3 +++
 2 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 385c1fd..7bcafb9 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -824,6 +824,9 @@ static void omap_4430sdp_display_init(void)
 	sdp4430_hdmi_mux_init();
 	sdp4430_picodlp_init();
 	omap_display_init(&sdp4430_dss_data);
+
+	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
 }
 
 #ifdef CONFIG_OMAP_MUX
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index e0d0abc..27fc162 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -541,6 +541,9 @@ void omap4_panda_display_init(void)
 
 	omap4_panda_hdmi_mux_init();
 	omap_display_init(&omap4_panda_dss_data);
+
+	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
 }
 
 static void __init omap4_panda_init(void)
-- 
1.7.4.1

[PATCH 5/7] OMAP: 4430SDP/Panda: add HDMI HPD gpio

Message details and related content
Both Panda and 4430SDP use GPIO 63 as HDMI hot-plug-detect. Configure
this GPIO in the board files.

Backported from aa74274b464d4aa24703963ac89a0ee942d5d267

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +++
 arch/arm/mach-omap2/board-omap4panda.c |    3 +++
 2 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 7bcafb9..5c2aaa0 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -54,6 +54,7 @@
 #define OMAP4_SFH7741_ENABLE_GPIO		188
 #define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
+#define HDMI_GPIO_HPD  63 /* Hotplug detect */
 #define DISPLAY_SEL_GPIO	59	/* LCD2/PicoDLP switch */
 #define DLP_POWER_ON_GPIO	40
 
@@ -608,6 +609,7 @@ static void sdp4430_hdmi_mux_init(void)
 static struct gpio sdp4430_hdmi_gpios[] = {
 	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_ls_oe" },
+	{ HDMI_GPIO_HPD, GPIOF_DIR_IN, "hdmi_gpio_hpd" },
 };
 
 static int sdp4430_panel_enable_hdmi(struct omap_dss_device *dssdev)
@@ -827,6 +829,7 @@ static void omap_4430sdp_display_init(void)
 
 	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
 	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_HPD, OMAP_PIN_INPUT_PULLDOWN);
 }
 
 #ifdef CONFIG_OMAP_MUX
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 27fc162..e913c80 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -53,6 +53,7 @@
 #define GPIO_WIFI_IRQ		53
 #define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
+#define HDMI_GPIO_HPD  63 /* Hotplug detect */
 
 /* wl127x BT, FM, GPS connectivity chip */
 static int wl1271_gpios[] = {46, -1, -1};
@@ -492,6 +493,7 @@ static void omap4_panda_hdmi_mux_init(void)
 static struct gpio panda_hdmi_gpios[] = {
 	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ls_oe" },
+	{ HDMI_GPIO_HPD, GPIOF_DIR_IN, "hdmi_gpio_hpd" },
 };
 
 static int omap4_panda_panel_enable_hdmi(struct omap_dss_device *dssdev)
@@ -544,6 +546,7 @@ void omap4_panda_display_init(void)
 
 	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
 	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_HPD, OMAP_PIN_INPUT_PULLDOWN);
 }
 
 static void __init omap4_panda_init(void)
-- 
1.7.4.1

[PATCH 6/7] OMAPDSS: HDMI: PHY burnout fix

Message details and related content
A hardware bug in the OMAP4 HDMI PHY may cause physical damage to the
board if the HDMI PHY is already enabled when the HDMI cable is plugged
in. The possible damage breaks HDMI output, otherwise the board is
unaffected.

This patch solves the problem by adding hot-plug-detection into the HDMI
IP driver. This is not a real HPD support in the sense that nobody else
than the IP driver gets to know about the HPD events, but is only meant
to fix the HW bug.

The strategy is simple: If the display device is turned off by the user,
the PHY power is set to OFF. When the display device is turned on by the
user, the PHY power is set either to LDOON or TXON, depending on whether
the HDMI cable is connected.

The reason to avoid PHY OFF when the display device is on, but the cable
is disconnected, is that when the PHY is turned OFF, the HDMI IP is not
"ticking" and thus the DISPC does not receive pixel clock from the HDMI
IP. This would, for example, prevent any VSYNCs from happening, and
would thus affect the users of omapdss. By using LDOON when the cable is
disconnected we'll avoid the HW bug, but keep the HDMI working as usual
from the user's point of view.

Backported from c49d005b6cc8491fad5b24f82805be2d6bcbd3dd

Signed-off-by: Tomi Valkeinen <redacted>
---
 arch/arm/mach-omap2/board-4430sdp.c       |    5 ++
 arch/arm/mach-omap2/board-omap4panda.c    |    5 ++
 drivers/video/omap2/dss/hdmi.c            |    3 +
 drivers/video/omap2/dss/ti_hdmi.h         |    4 ++
 drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c |   68 +++++++++++++++++++++++++++--
 include/video/omapdss.h                   |    5 ++
 6 files changed, 86 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 5c2aaa0..02cd29a 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -742,6 +742,10 @@ static void sdp4430_lcd_init(void)
 		pr_err("%s: Could not get lcd2_reset_gpio\n", __func__);
 }
 
+static struct omap_dss_hdmi_data sdp4430_hdmi_data = {
+	.hpd_gpio = HDMI_GPIO_HPD,
+};
+
 static struct omap_dss_device sdp4430_hdmi_device = {
 	.name = "hdmi",
 	.driver_name = "hdmi_panel",
@@ -749,6 +753,7 @@ static struct omap_dss_device sdp4430_hdmi_device = {
 	.platform_enable = sdp4430_panel_enable_hdmi,
 	.platform_disable = sdp4430_panel_disable_hdmi,
 	.channel = OMAP_DSS_CHANNEL_DIGIT,
+	.data = &sdp4430_hdmi_data,
 };
 
 static struct picodlp_panel_data sdp4430_picodlp_pdata = {
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index e913c80..51b1c93 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -513,6 +513,10 @@ static void omap4_panda_panel_disable_hdmi(struct omap_dss_device *dssdev)
 	gpio_free_array(panda_hdmi_gpios, ARRAY_SIZE(panda_hdmi_gpios));
 }
 
+static struct omap_dss_hdmi_data omap4_panda_hdmi_data = {
+	.hpd_gpio = HDMI_GPIO_HPD,
+};
+
 static struct omap_dss_device  omap4_panda_hdmi_device = {
 	.name = "hdmi",
 	.driver_name = "hdmi_panel",
@@ -520,6 +524,7 @@ static struct omap_dss_device  omap4_panda_hdmi_device = {
 	.platform_enable = omap4_panda_panel_enable_hdmi,
 	.platform_disable = omap4_panda_panel_disable_hdmi,
 	.channel = OMAP_DSS_CHANNEL_DIGIT,
+	.data = &omap4_panda_hdmi_data,
 };
 
 static struct omap_dss_device *omap4_panda_dss_devices[] = {
diff --git a/drivers/video/omap2/dss/hdmi.c b/drivers/video/omap2/dss/hdmi.c
index c56378c..7099c31 100644
--- a/drivers/video/omap2/dss/hdmi.c
+++ b/drivers/video/omap2/dss/hdmi.c
@@ -490,6 +490,7 @@ bool omapdss_hdmi_detect(void)
 
 int omapdss_hdmi_display_enable(struct omap_dss_device *dssdev)
 {
+	struct omap_dss_hdmi_data *priv = dssdev->data;
 	int r = 0;
 
 	DSSDBG("ENTER hdmi_display_enable\n");
@@ -502,6 +503,8 @@ int omapdss_hdmi_display_enable(struct omap_dss_device *dssdev)
 		goto err0;
 	}
 
+	hdmi.ip_data.hpd_gpio = priv->hpd_gpio;
+
 	r = omap_dss_start_device(dssdev);
 	if (r) {
 		DSSERR("failed to start device\n");
diff --git a/drivers/video/omap2/dss/ti_hdmi.h b/drivers/video/omap2/dss/ti_hdmi.h
index 2c3443d..ec337b5d 100644
--- a/drivers/video/omap2/dss/ti_hdmi.h
+++ b/drivers/video/omap2/dss/ti_hdmi.h
@@ -121,6 +121,10 @@ struct hdmi_ip_data {
 	const struct ti_hdmi_ip_ops *ops;
 	struct hdmi_config cfg;
 	struct hdmi_pll_info pll_data;
+
+	/* ti_hdmi_4xxx_ip private data. These should be in a separate struct */
+	int hpd_gpio;
+	bool phy_tx_enabled;
 };
 int ti_hdmi_4xxx_phy_enable(struct hdmi_ip_data *ip_data);
 void ti_hdmi_4xxx_phy_disable(struct hdmi_ip_data *ip_data);
diff --git a/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c b/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
index e1a6ce5..3683404 100644
--- a/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
+++ b/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
@@ -28,6 +28,7 @@
 #include <linux/delay.h>
 #include <linux/string.h>
 #include <linux/seq_file.h>
+#include <linux/gpio.h>
 
 #include "ti_hdmi_4xxx_ip.h"
 #include "dss.h"
@@ -223,6 +224,49 @@ void ti_hdmi_4xxx_pll_disable(struct hdmi_ip_data *ip_data)
 	hdmi_set_pll_pwr(ip_data, HDMI_PLLPWRCMD_ALLOFF);
 }
 
+static int hdmi_check_hpd_state(struct hdmi_ip_data *ip_data)
+{
+	unsigned long flags;
+	bool hpd;
+	int r;
+	/* this should be in ti_hdmi_4xxx_ip private data */
+	static DEFINE_SPINLOCK(phy_tx_lock);
+
+	spin_lock_irqsave(&phy_tx_lock, flags);
+
+	hpd = gpio_get_value(ip_data->hpd_gpio);
+
+	if (hpd = ip_data->phy_tx_enabled) {
+		spin_unlock_irqrestore(&phy_tx_lock, flags);
+		return 0;
+	}
+
+	if (hpd)
+		r = hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_TXON);
+	else
+		r = hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_LDOON);
+
+	if (r) {
+		DSSERR("Failed to %s PHY TX power\n",
+				hpd ? "enable" : "disable");
+		goto err;
+	}
+
+	ip_data->phy_tx_enabled = hpd;
+err:
+	spin_unlock_irqrestore(&phy_tx_lock, flags);
+	return r;
+}
+
+static irqreturn_t hpd_irq_handler(int irq, void *data)
+{
+	struct hdmi_ip_data *ip_data = data;
+
+	hdmi_check_hpd_state(ip_data);
+
+	return IRQ_HANDLED;
+}
+
 int ti_hdmi_4xxx_phy_enable(struct hdmi_ip_data *ip_data)
 {
 	u16 r = 0;
@@ -232,10 +276,6 @@ int ti_hdmi_4xxx_phy_enable(struct hdmi_ip_data *ip_data)
 	if (r)
 		return r;
 
-	r = hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_TXON);
-	if (r)
-		return r;
-
 	/*
 	 * Read address 0 in order to get the SCP reset done completed
 	 * Dummy access performed to make sure reset is done
@@ -257,12 +297,32 @@ int ti_hdmi_4xxx_phy_enable(struct hdmi_ip_data *ip_data)
 	/* Write to phy address 3 to change the polarity control */
 	REG_FLD_MOD(phy_base, HDMI_TXPHY_PAD_CFG_CTRL, 0x1, 27, 27);
 
+	r = request_threaded_irq(gpio_to_irq(ip_data->hpd_gpio),
+			NULL, hpd_irq_handler,
+			IRQF_DISABLED | IRQF_TRIGGER_RISING |
+			IRQF_TRIGGER_FALLING, "hpd", ip_data);
+	if (r) {
+		DSSERR("HPD IRQ request failed\n");
+		hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_OFF);
+		return r;
+	}
+
+	r = hdmi_check_hpd_state(ip_data);
+	if (r) {
+		free_irq(gpio_to_irq(ip_data->hpd_gpio), ip_data);
+		hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_OFF);
+		return r;
+	}
+
 	return 0;
 }
 
 void ti_hdmi_4xxx_phy_disable(struct hdmi_ip_data *ip_data)
 {
+	free_irq(gpio_to_irq(ip_data->hpd_gpio), ip_data);
+
 	hdmi_set_phy_pwr(ip_data, HDMI_PHYPWRCMD_OFF);
+	ip_data->phy_tx_enabled = false;
 }
 
 static int hdmi_core_ddc_init(struct hdmi_ip_data *ip_data)
diff --git a/include/video/omapdss.h b/include/video/omapdss.h
index 378c7ed..6582c45 100644
--- a/include/video/omapdss.h
+++ b/include/video/omapdss.h
@@ -575,6 +575,11 @@ struct omap_dss_device {
 	int (*get_backlight)(struct omap_dss_device *dssdev);
 };
 
+struct omap_dss_hdmi_data
+{
+	int hpd_gpio;
+};
+
 struct omap_dss_driver {
 	struct device_driver driver;
 
-- 
1.7.4.1

[PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix

Message details and related content
From: Rob Clark <redacted>

The "OMAPDSS: HDMI: PHY burnout fix" commit switched the HDMI driver
over to using a GPIO for plug detect.  Unfortunately the ->detect()
method was not also updated, causing HDMI to no longer work for the
omapdrm driver (because it would actually check if a connection was
detected before attempting to enable display).

Signed-off-by: Rob Clark <redacted>
Signed-off-by: Tomi Valkeinen <redacted>
---
 drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c |    9 +--------
 1 files changed, 1 insertions(+), 8 deletions(-)
diff --git a/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c b/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
index 3683404..aad48a1 100644
--- a/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
+++ b/drivers/video/omap2/dss/ti_hdmi_4xxx_ip.c
@@ -479,14 +479,7 @@ int ti_hdmi_4xxx_read_edid(struct hdmi_ip_data *ip_data,
 
 bool ti_hdmi_4xxx_detect(struct hdmi_ip_data *ip_data)
 {
-	int r;
-
-	void __iomem *base = hdmi_core_sys_base(ip_data);
-
-	/* HPD */
-	r = REG_GET(base, HDMI_CORE_SYS_SYS_STAT, 1, 1);
-
-	return r = 1;
+	return gpio_get_value(ip_data->hpd_gpio);
 }
 
 static void hdmi_core_init(struct hdmi_core_video_config *video_cfg,
-- 
1.7.4.1

[PATCH 0/7] OMAPDSS: HDMI PHY burnout fix for 3.0 stable

Message details and related content
This series is for 3.0 stable.

The main patch in this set is the "PHY burnout fix", which implements a fix for
a hardware bug in the OMAP4 HDMI PHY which may cause physical damage to the
board if the HDMI PHY is already enabled when the HDMI cable is plugged in. The
possible damage breaks HDMI output hardware, otherwise the board is unaffected.

The "use default dividers" patch fixes a problem with setting up the HDMI PLL
dividers on Pandaboard. Without the patch the HDMI output does not work at all
on Panda.

The preceding patches are small cleanups/fixes for HDMI GPIOs so that the fix
can be implemented.

 Tomi

Tomi Valkeinen (7):
  OMAP: DSS2: HDMI: use default dividers
  OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios
  OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD
  OMAPDSS: remove wrong HDMI HPD muxing
  OMAP: 4430SDP/Panda: setup HDMI GPIO muxes
  OMAP: 4430SDP/Panda: add HDMI HPD gpio
  OMAPDSS: HDMI: PHY burnout fix

 arch/arm/mach-omap2/board-4430sdp.c    |   31 +++++------
 arch/arm/mach-omap2/board-omap4panda.c |   22 +++++---
 drivers/video/omap2/dss/hdmi.c         |   86 +++++++++++++++++++++++++++++--
 include/video/omapdss.h                |    5 ++
 4 files changed, 113 insertions(+), 31 deletions(-)

-- 
1.7.4.1

[PATCH 1/7] OMAP: DSS2: HDMI: use default dividers

Message details and related content
Use default regn and regm2 dividers in the hdmi driver if the board file
does not define them.

Pandaboard's board file was missing the dividers, causing HDMI output
not to work, so this patch fixes the problem.

Backported from 8d88767a4377171752c22ac39bcb2b505eb751da

Cc: Mythri P K <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tomi Valkeinen <redacted>
---
 arch/arm/mach-omap2/board-4430sdp.c |    9 ---------
 drivers/video/omap2/dss/hdmi.c      |   15 +++++++++++++--
 2 files changed, 13 insertions(+), 11 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 63de2d3..f515fa2 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -617,15 +617,6 @@ static struct omap_dss_device sdp4430_hdmi_device = {
 	.name = "hdmi",
 	.driver_name = "hdmi_panel",
 	.type = OMAP_DISPLAY_TYPE_HDMI,
-	.clocks	= {
-		.dispc	= {
-			.dispc_fclk_src	= OMAP_DSS_CLK_SRC_FCK,
-		},
-		.hdmi	= {
-			.regn	= 15,
-			.regm2	= 1,
-		},
-	},
 	.platform_enable = sdp4430_panel_enable_hdmi,
 	.platform_disable = sdp4430_panel_disable_hdmi,
 	.channel = OMAP_DSS_CHANNEL_DIGIT,
diff --git a/drivers/video/omap2/dss/hdmi.c b/drivers/video/omap2/dss/hdmi.c
index b0555f4..f3369cf 100644
--- a/drivers/video/omap2/dss/hdmi.c
+++ b/drivers/video/omap2/dss/hdmi.c
@@ -40,6 +40,9 @@
 #include "hdmi.h"
 #include "dss_features.h"
 
+#define HDMI_DEFAULT_REGN 15
+#define HDMI_DEFAULT_REGM2 1
+
 static struct {
 	struct mutex lock;
 	struct omap_display_platform_data *pdata;
@@ -1069,7 +1072,11 @@ static void hdmi_compute_pll(struct omap_dss_device *dssdev, int phy,
 	 * Input clock is predivided by N + 1
 	 * out put of which is reference clk
 	 */
-	pi->regn = dssdev->clocks.hdmi.regn;
+	if (dssdev->clocks.hdmi.regn = 0)
+		pi->regn = HDMI_DEFAULT_REGN;
+	else
+		pi->regn = dssdev->clocks.hdmi.regn;
+
 	refclk = clkin / (pi->regn + 1);
 
 	/*
@@ -1077,7 +1084,11 @@ static void hdmi_compute_pll(struct omap_dss_device *dssdev, int phy,
 	 * Multiplying by 100 to avoid fractional part removal
 	 */
 	pi->regm = (phy * 100 / (refclk)) / 100;
-	pi->regm2 = dssdev->clocks.hdmi.regm2;
+
+	if (dssdev->clocks.hdmi.regm2 = 0)
+		pi->regm2 = HDMI_DEFAULT_REGM2;
+	else
+		pi->regm2 = dssdev->clocks.hdmi.regm2;
 
 	/*
 	 * fractional multiplier is remainder of the difference between
-- 
1.7.4.1

[PATCH 2/7] OMAP: 4430SDP/Panda: use gpio_free_array to free HDMI gpios

Message details and related content
Instead of freeing the GPIOs individually, use gpio_free_array().

Backported from 575753e3bea3b67eef8e454fb87f719e3f7da599

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +--
 arch/arm/mach-omap2/board-omap4panda.c |    3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index f515fa2..7feefb8 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -609,8 +609,7 @@ static int sdp4430_panel_enable_hdmi(struct omap_dss_device *dssdev)
 
 static void sdp4430_panel_disable_hdmi(struct omap_dss_device *dssdev)
 {
-	gpio_free(HDMI_GPIO_LS_OE);
-	gpio_free(HDMI_GPIO_HPD);
+	gpio_free_array(sdp4430_hdmi_gpios, ARRAY_SIZE(sdp4430_hdmi_gpios));
 }
 
 static struct omap_dss_device sdp4430_hdmi_device = {
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 0cfe200..2e9a6f36 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -645,8 +645,7 @@ static int omap4_panda_panel_enable_hdmi(struct omap_dss_device *dssdev)
 
 static void omap4_panda_panel_disable_hdmi(struct omap_dss_device *dssdev)
 {
-	gpio_free(HDMI_GPIO_LS_OE);
-	gpio_free(HDMI_GPIO_HPD);
+	gpio_free_array(panda_hdmi_gpios, ARRAY_SIZE(panda_hdmi_gpios));
 }
 
 static struct omap_dss_device  omap4_panda_hdmi_device = {
-- 
1.7.4.1

[PATCH 3/7] OMAP: 4430SDP/Panda: rename HPD GPIO to CT_CP_HPD

Message details and related content
The GPIO 60 on 4430sdp and Panda is not HPD GPIO, as currently marked in
the board files, but CT_CP_HPD, which is used to enable/disable HPD
functionality.

This patch renames the GPIO.

Backported from 3932a32fcf5393f8be70ac99dc718ad7ad0a415b

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    4 ++--
 arch/arm/mach-omap2/board-omap4panda.c |    4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 7feefb8..d0a49c3 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -49,7 +49,7 @@
 #define ETH_KS8851_QUART		138
 #define OMAP4_SFH7741_SENSOR_OUTPUT_GPIO	184
 #define OMAP4_SFH7741_ENABLE_GPIO		188
-#define HDMI_GPIO_HPD 60 /* Hot plug pin for HDMI */
+#define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
 
 static const int sdp4430_keymap[] = {
@@ -591,7 +591,7 @@ static void sdp4430_hdmi_mux_init(void)
 }
 
 static struct gpio sdp4430_hdmi_gpios[] = {
-	{ HDMI_GPIO_HPD,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_hpd"   },
+	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_ls_oe" },
 };
 
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 2e9a6f36..bc98749 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -52,7 +52,7 @@
 #define GPIO_HUB_NRESET		62
 #define GPIO_WIFI_PMENA		43
 #define GPIO_WIFI_IRQ		53
-#define HDMI_GPIO_HPD 60 /* Hot plug pin for HDMI */
+#define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
 
 /* wl127x BT, FM, GPS connectivity chip */
@@ -627,7 +627,7 @@ static void omap4_panda_hdmi_mux_init(void)
 }
 
 static struct gpio panda_hdmi_gpios[] = {
-	{ HDMI_GPIO_HPD,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_hpd"   },
+	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ls_oe" },
 };
 
-- 
1.7.4.1

[PATCH 4/7] OMAPDSS: remove wrong HDMI HPD muxing

Message details and related content
"hdmi_hpd" pin is muxed to INPUT and PULLUP, but the pin is not
currently used, and in the future when it is used, the pin is used as a
GPIO and is board specific, not an OMAP4 wide thing.

So remove the muxing for now.

Backported from 7bb122d155f742fe2d79849090c825be7b4a247e

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    4 ----
 arch/arm/mach-omap2/board-omap4panda.c |    4 ----
 2 files changed, 0 insertions(+), 8 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index d0a49c3..9bc418e 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -578,12 +578,8 @@ static void __init omap_sfh7741prox_init(void)
 
 static void sdp4430_hdmi_mux_init(void)
 {
-	/* PAD0_HDMI_HPD_PAD1_HDMI_CEC */
-	omap_mux_init_signal("hdmi_hpd",
-			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_cec",
 			OMAP_PIN_INPUT_PULLUP);
-	/* PAD0_HDMI_DDC_SCL_PAD1_HDMI_DDC_SDA */
 	omap_mux_init_signal("hdmi_ddc_scl",
 			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_ddc_sda",
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index bc98749..86a2d30 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -614,12 +614,8 @@ int __init omap4_panda_dvi_init(void)
 
 static void omap4_panda_hdmi_mux_init(void)
 {
-	/* PAD0_HDMI_HPD_PAD1_HDMI_CEC */
-	omap_mux_init_signal("hdmi_hpd",
-			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_cec",
 			OMAP_PIN_INPUT_PULLUP);
-	/* PAD0_HDMI_DDC_SCL_PAD1_HDMI_DDC_SDA */
 	omap_mux_init_signal("hdmi_ddc_scl",
 			OMAP_PIN_INPUT_PULLUP);
 	omap_mux_init_signal("hdmi_ddc_sda",
-- 
1.7.4.1

[PATCH 5/7] OMAP: 4430SDP/Panda: setup HDMI GPIO muxes

Message details and related content
The HDMI GPIO pins LS_OE and CT_CP_HPD are not currently configured.
This patch configures them as output pins.

Backported from 78a1ad8f12db70b8b0a4548b90704de08ee216ce

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +++
 arch/arm/mach-omap2/board-omap4panda.c |    3 +++
 2 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 9bc418e..669e329 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -631,6 +631,9 @@ void omap_4430sdp_display_init(void)
 {
 	sdp4430_hdmi_mux_init();
 	omap_display_init(&sdp4430_dss_data);
+
+	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
 }
 
 #ifdef CONFIG_OMAP_MUX
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 86a2d30..5d519fb 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -674,6 +674,9 @@ void omap4_panda_display_init(void)
 
 	omap4_panda_hdmi_mux_init();
 	omap_display_init(&omap4_panda_dss_data);
+
+	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
 }
 
 static void __init omap4_panda_init(void)
-- 
1.7.4.1

[PATCH 6/7] OMAP: 4430SDP/Panda: add HDMI HPD gpio

Message details and related content
Both Panda and 4430SDP use GPIO 63 as HDMI hot-plug-detect. Configure
this GPIO in the board files.

Backported from aa74274b464d4aa24703963ac89a0ee942d5d267

Signed-off-by: Tomi Valkeinen <redacted>
Acked-by: Tony Lindgren <tony@atomide.com>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    3 +++
 arch/arm/mach-omap2/board-omap4panda.c |    3 +++
 2 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 669e329..0ee578a 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -51,6 +51,7 @@
 #define OMAP4_SFH7741_ENABLE_GPIO		188
 #define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
+#define HDMI_GPIO_HPD  63 /* Hotplug detect */
 
 static const int sdp4430_keymap[] = {
 	KEY(0, 0, KEY_E),
@@ -589,6 +590,7 @@ static void sdp4430_hdmi_mux_init(void)
 static struct gpio sdp4430_hdmi_gpios[] = {
 	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH,	"hdmi_gpio_ls_oe" },
+	{ HDMI_GPIO_HPD, GPIOF_DIR_IN, "hdmi_gpio_hpd" },
 };
 
 static int sdp4430_panel_enable_hdmi(struct omap_dss_device *dssdev)
@@ -634,6 +636,7 @@ void omap_4430sdp_display_init(void)
 
 	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
 	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_HPD, OMAP_PIN_INPUT_PULLDOWN);
 }
 
 #ifdef CONFIG_OMAP_MUX
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 5d519fb..2b3f7e0 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -54,6 +54,7 @@
 #define GPIO_WIFI_IRQ		53
 #define HDMI_GPIO_CT_CP_HPD 60 /* HPD mode enable/disable */
 #define HDMI_GPIO_LS_OE 41 /* Level shifter for HDMI */
+#define HDMI_GPIO_HPD  63 /* Hotplug detect */
 
 /* wl127x BT, FM, GPS connectivity chip */
 static int wl1271_gpios[] = {46, -1, -1};
@@ -625,6 +626,7 @@ static void omap4_panda_hdmi_mux_init(void)
 static struct gpio panda_hdmi_gpios[] = {
 	{ HDMI_GPIO_CT_CP_HPD, GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ct_cp_hpd" },
 	{ HDMI_GPIO_LS_OE,	GPIOF_OUT_INIT_HIGH, "hdmi_gpio_ls_oe" },
+	{ HDMI_GPIO_HPD, GPIOF_DIR_IN, "hdmi_gpio_hpd" },
 };
 
 static int omap4_panda_panel_enable_hdmi(struct omap_dss_device *dssdev)
@@ -677,6 +679,7 @@ void omap4_panda_display_init(void)
 
 	omap_mux_init_gpio(HDMI_GPIO_LS_OE, OMAP_PIN_OUTPUT);
 	omap_mux_init_gpio(HDMI_GPIO_CT_CP_HPD, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_GPIO_HPD, OMAP_PIN_INPUT_PULLDOWN);
 }
 
 static void __init omap4_panda_init(void)
-- 
1.7.4.1

[PATCH 7/7] OMAPDSS: HDMI: PHY burnout fix

Message details and related content
A hardware bug in the OMAP4 HDMI PHY may cause physical damage to the
board if the HDMI PHY is already enabled when the HDMI cable is plugged
in. The possible damage breaks HDMI output, otherwise the board is
unaffected.

This patch solves the problem by adding hot-plug-detection into the HDMI
IP driver. This is not a real HPD support in the sense that nobody else
than the IP driver gets to know about the HPD events, but is only meant
to fix the HW bug.

The strategy is simple: If the display device is turned off by the user,
the PHY power is set to OFF. When the display device is turned on by the
user, the PHY power is set either to LDOON or TXON, depending on whether
the HDMI cable is connected.

The reason to avoid PHY OFF when the display device is on, but the cable
is disconnected, is that when the PHY is turned OFF, the HDMI IP is not
"ticking" and thus the DISPC does not receive pixel clock from the HDMI
IP. This would, for example, prevent any VSYNCs from happening, and
would thus affect the users of omapdss. By using LDOON when the cable is
disconnected we'll avoid the HW bug, but keep the HDMI working as usual
from the user's point of view.

Backported from c49d005b6cc8491fad5b24f82805be2d6bcbd3dd

Signed-off-by: Tomi Valkeinen <redacted>
---
 arch/arm/mach-omap2/board-4430sdp.c    |    5 ++
 arch/arm/mach-omap2/board-omap4panda.c |    5 ++
 drivers/video/omap2/dss/hdmi.c         |   71 ++++++++++++++++++++++++++++++--
 include/video/omapdss.h                |    5 ++
 4 files changed, 82 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/board-4430sdp.c b/arch/arm/mach-omap2/board-4430sdp.c
index 0ee578a..14a5971 100644
--- a/arch/arm/mach-omap2/board-4430sdp.c
+++ b/arch/arm/mach-omap2/board-4430sdp.c
@@ -610,6 +610,10 @@ static void sdp4430_panel_disable_hdmi(struct omap_dss_device *dssdev)
 	gpio_free_array(sdp4430_hdmi_gpios, ARRAY_SIZE(sdp4430_hdmi_gpios));
 }
 
+static struct omap_dss_hdmi_data sdp4430_hdmi_data = {
+	.hpd_gpio = HDMI_GPIO_HPD,
+};
+
 static struct omap_dss_device sdp4430_hdmi_device = {
 	.name = "hdmi",
 	.driver_name = "hdmi_panel",
@@ -617,6 +621,7 @@ static struct omap_dss_device sdp4430_hdmi_device = {
 	.platform_enable = sdp4430_panel_enable_hdmi,
 	.platform_disable = sdp4430_panel_disable_hdmi,
 	.channel = OMAP_DSS_CHANNEL_DIGIT,
+	.data = &sdp4430_hdmi_data,
 };
 
 static struct omap_dss_device *sdp4430_dss_devices[] = {
diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c
index 2b3f7e0..107dfc3 100644
--- a/arch/arm/mach-omap2/board-omap4panda.c
+++ b/arch/arm/mach-omap2/board-omap4panda.c
@@ -646,6 +646,10 @@ static void omap4_panda_panel_disable_hdmi(struct omap_dss_device *dssdev)
 	gpio_free_array(panda_hdmi_gpios, ARRAY_SIZE(panda_hdmi_gpios));
 }
 
+static struct omap_dss_hdmi_data omap4_panda_hdmi_data = {
+	.hpd_gpio = HDMI_GPIO_HPD,
+};
+
 static struct omap_dss_device  omap4_panda_hdmi_device = {
 	.name = "hdmi",
 	.driver_name = "hdmi_panel",
@@ -653,6 +657,7 @@ static struct omap_dss_device  omap4_panda_hdmi_device = {
 	.platform_enable = omap4_panda_panel_enable_hdmi,
 	.platform_disable = omap4_panda_panel_disable_hdmi,
 	.channel = OMAP_DSS_CHANNEL_DIGIT,
+	.data = &omap4_panda_hdmi_data,
 };
 
 static struct omap_dss_device *omap4_panda_dss_devices[] = {
diff --git a/drivers/video/omap2/dss/hdmi.c b/drivers/video/omap2/dss/hdmi.c
index f3369cf..fadd6a0 100644
--- a/drivers/video/omap2/dss/hdmi.c
+++ b/drivers/video/omap2/dss/hdmi.c
@@ -29,6 +29,7 @@
 #include <linux/mutex.h>
 #include <linux/delay.h>
 #include <linux/string.h>
+#include <linux/gpio.h>
 #include <video/omapdss.h>
 #if defined(CONFIG_SND_OMAP_SOC_OMAP4_HDMI) || \
 	defined(CONFIG_SND_OMAP_SOC_OMAP4_HDMI_MODULE)
@@ -54,6 +55,9 @@ static struct {
 	u8 edid_set;
 	bool custom_set;
 	struct hdmi_config cfg;
+
+	int hpd_gpio;
+	bool phy_tx_enabled;
 } hdmi;
 
 /*
@@ -278,6 +282,47 @@ static int hdmi_pll_reset(void)
 	return 0;
 }
 
+static int hdmi_check_hpd_state(void)
+{
+	unsigned long flags;
+	bool hpd;
+	int r;
+	/* this should be in ti_hdmi_4xxx_ip private data */
+	static DEFINE_SPINLOCK(phy_tx_lock);
+
+	spin_lock_irqsave(&phy_tx_lock, flags);
+
+	hpd = gpio_get_value(hdmi.hpd_gpio);
+
+	if (hpd = hdmi.phy_tx_enabled) {
+		spin_unlock_irqrestore(&phy_tx_lock, flags);
+		return 0;
+	}
+
+	if (hpd)
+		r = hdmi_set_phy_pwr(HDMI_PHYPWRCMD_TXON);
+	else
+		r = hdmi_set_phy_pwr(HDMI_PHYPWRCMD_LDOON);
+
+	if (r) {
+		DSSERR("Failed to %s PHY TX power\n",
+				hpd ? "enable" : "disable");
+		goto err;
+	}
+
+	hdmi.phy_tx_enabled = hpd;
+err:
+	spin_unlock_irqrestore(&phy_tx_lock, flags);
+	return r;
+}
+
+static irqreturn_t hpd_irq_handler(int irq, void *data)
+{
+	hdmi_check_hpd_state();
+
+	return IRQ_HANDLED;
+}
+
 static int hdmi_phy_init(void)
 {
 	u16 r = 0;
@@ -286,10 +331,6 @@ static int hdmi_phy_init(void)
 	if (r)
 		return r;
 
-	r = hdmi_set_phy_pwr(HDMI_PHYPWRCMD_TXON);
-	if (r)
-		return r;
-
 	/*
 	 * Read address 0 in order to get the SCP reset done completed
 	 * Dummy access performed to make sure reset is done
@@ -311,6 +352,23 @@ static int hdmi_phy_init(void)
 	/* Write to phy address 3 to change the polarity control */
 	REG_FLD_MOD(HDMI_TXPHY_PAD_CFG_CTRL, 0x1, 27, 27);
 
+	r = request_threaded_irq(gpio_to_irq(hdmi.hpd_gpio),
+			NULL, hpd_irq_handler,
+			IRQF_DISABLED | IRQF_TRIGGER_RISING |
+			IRQF_TRIGGER_FALLING, "hpd", NULL);
+	if (r) {
+		DSSERR("HPD IRQ request failed\n");
+		hdmi_set_phy_pwr(HDMI_PHYPWRCMD_OFF);
+		return r;
+	}
+
+	r = hdmi_check_hpd_state();
+	if (r) {
+		free_irq(gpio_to_irq(hdmi.hpd_gpio), NULL);
+		hdmi_set_phy_pwr(HDMI_PHYPWRCMD_OFF);
+		return r;
+	}
+
 	return 0;
 }
 
@@ -361,7 +419,9 @@ static int hdmi_pll_program(struct hdmi_pll_info *fmt)
 
 static void hdmi_phy_off(void)
 {
+	free_irq(gpio_to_irq(hdmi.hpd_gpio), NULL);
 	hdmi_set_phy_pwr(HDMI_PHYPWRCMD_OFF);
+	hdmi.phy_tx_enabled = false;
 }
 
 static int hdmi_core_ddc_edid(u8 *pedid, int ext)
@@ -1236,12 +1296,15 @@ void omapdss_hdmi_display_set_timing(struct omap_dss_device *dssdev)
 
 int omapdss_hdmi_display_enable(struct omap_dss_device *dssdev)
 {
+	struct omap_dss_hdmi_data *priv = dssdev->data;
 	int r = 0;
 
 	DSSDBG("ENTER hdmi_display_enable\n");
 
 	mutex_lock(&hdmi.lock);
 
+	hdmi.hpd_gpio = priv->hpd_gpio;
+
 	r = omap_dss_start_device(dssdev);
 	if (r) {
 		DSSERR("failed to start device\n");
diff --git a/include/video/omapdss.h b/include/video/omapdss.h
index 892b97f..c0d8014 100644
--- a/include/video/omapdss.h
+++ b/include/video/omapdss.h
@@ -514,6 +514,11 @@ struct omap_dss_device {
 	int (*get_backlight)(struct omap_dss_device *dssdev);
 };
 
+struct omap_dss_hdmi_data
+{
+	int hpd_gpio;
+};
+
 struct omap_dss_driver {
 	struct device_driver driver;
 
-- 
1.7.4.1

Re: [PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix

Message details and related content
On Thu, Mar 01, 2012 at 02:26:35PM +0200, Tomi Valkeinen wrote:
From: Rob Clark <redacted>

The "OMAPDSS: HDMI: PHY burnout fix" commit switched the HDMI driver
over to using a GPIO for plug detect.  Unfortunately the ->detect()
method was not also updated, causing HDMI to no longer work for the
omapdrm driver (because it would actually check if a connection was
detected before attempting to enable display).

Signed-off-by: Rob Clark <redacted>
Signed-off-by: Tomi Valkeinen <redacted>
You forgot to tell me what the git commit id is for this patch (it's
ca888a7958b3d808e4efd08ceff88913f4212c69, right?)

And why isn't this needed for the 3.0 kernel as well?

thanks,

greg k-h

Re: [PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix

Message details and related content
On Wed, 2012-03-07 at 12:01 -0800, Greg KH wrote:
On Thu, Mar 01, 2012 at 02:26:35PM +0200, Tomi Valkeinen wrote:
quoted
From: Rob Clark <redacted>

The "OMAPDSS: HDMI: PHY burnout fix" commit switched the HDMI driver
over to using a GPIO for plug detect.  Unfortunately the ->detect()
method was not also updated, causing HDMI to no longer work for the
omapdrm driver (because it would actually check if a connection was
detected before attempting to enable display).

Signed-off-by: Rob Clark <redacted>
Signed-off-by: Tomi Valkeinen <redacted>
You forgot to tell me what the git commit id is for this patch (it's
ca888a7958b3d808e4efd08ceff88913f4212c69, right?)
Yes, that's the one. It wasn't in Linus's tree yet, only in fbdev tree,
so I wasn't sure what the commit id is.
And why isn't this needed for the 3.0 kernel as well?
The detect() function is not present in 3.0, so there was nothing to
break.

 Tomi

Re: [PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix

Message details and related content
On Thu, Mar 08, 2012 at 09:35:13AM +0200, Tomi Valkeinen wrote:
On Wed, 2012-03-07 at 12:01 -0800, Greg KH wrote:
quoted
On Thu, Mar 01, 2012 at 02:26:35PM +0200, Tomi Valkeinen wrote:
quoted
From: Rob Clark <redacted>

The "OMAPDSS: HDMI: PHY burnout fix" commit switched the HDMI driver
over to using a GPIO for plug detect.  Unfortunately the ->detect()
method was not also updated, causing HDMI to no longer work for the
omapdrm driver (because it would actually check if a connection was
detected before attempting to enable display).

Signed-off-by: Rob Clark <redacted>
Signed-off-by: Tomi Valkeinen <redacted>
You forgot to tell me what the git commit id is for this patch (it's
ca888a7958b3d808e4efd08ceff88913f4212c69, right?)
Yes, that's the one. It wasn't in Linus's tree yet, only in fbdev tree,
so I wasn't sure what the commit id is.
Then you should not have sent it to me, as if I were to take it then, I
could not have :(
quoted
And why isn't this needed for the 3.0 kernel as well?
The detect() function is not present in 3.0, so there was nothing to
break.
Ok, so everything I've queued up is all that is needed, right?

greg k-h

Re: [PATCH 7/7] OMAPDSS: HDMI: hot plug detect fix

Message details and related content
On Thu, 2012-03-08 at 07:29 -0800, Greg KH wrote:
On Thu, Mar 08, 2012 at 09:35:13AM +0200, Tomi Valkeinen wrote:
quoted
On Wed, 2012-03-07 at 12:01 -0800, Greg KH wrote:
quoted
On Thu, Mar 01, 2012 at 02:26:35PM +0200, Tomi Valkeinen wrote:
quoted
From: Rob Clark <redacted>

The "OMAPDSS: HDMI: PHY burnout fix" commit switched the HDMI driver
over to using a GPIO for plug detect.  Unfortunately the ->detect()
method was not also updated, causing HDMI to no longer work for the
omapdrm driver (because it would actually check if a connection was
detected before attempting to enable display).

Signed-off-by: Rob Clark <redacted>
Signed-off-by: Tomi Valkeinen <redacted>
You forgot to tell me what the git commit id is for this patch (it's
ca888a7958b3d808e4efd08ceff88913f4212c69, right?)
Yes, that's the one. It wasn't in Linus's tree yet, only in fbdev tree,
so I wasn't sure what the commit id is.
Then you should not have sent it to me, as if I were to take it then, I
could not have :(
Oh, ok. I thought the "patch-must-be-in-mainline"-rule was not a totally
strict one, so I decided to include it in this case as the patch was a
rather trivial one and already in the fbdev tree (I mentioned it in the
intro mail).

I guess I got lucky and the patch got into mainline before you took the
patches.
quoted
quoted
And why isn't this needed for the 3.0 kernel as well?
The detect() function is not present in 3.0, so there was nothing to
break.
Ok, so everything I've queued up is all that is needed, right?
Yes, looks correct. Thanks!

 Tomi

Keyboard shortcuts
hback out one level
jnext message in thread / next series-diff position
kprevious message in thread / previous series-diff position
ldrill in
Escclose help / fold thread tree
?toggle this help