On 26 July 2018 at 13:50, Andrzej Hajda a.hajda@samsung.com wrote:
On 26.07.2018 12:44, Krzysztof Kozlowski wrote:
On 26 July 2018 at 12:12, Andrzej Hajda a.hajda@samsung.com wrote:
HDMI-HPD was set active low, moreover by default pincontrol chip sets pull-down on the pin. As a result HDMI driver assumes TV is always connected regardless of actual state. The patch fixes it.
Signed-off-by: Andrzej Hajda a.hajda@samsung.com
arch/arm/boot/dts/exynos5250-arndale.dts | 4 +++- arch/arm/boot/dts/exynos5250-pinctrl.dtsi | 5 +++++ 2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/arch/arm/boot/dts/exynos5250-arndale.dts b/arch/arm/boot/dts/exynos5250-arndale.dts index 591023391f7d..26bae5157177 100644 --- a/arch/arm/boot/dts/exynos5250-arndale.dts +++ b/arch/arm/boot/dts/exynos5250-arndale.dts @@ -210,9 +210,11 @@ };
&hdmi {
pinctrl-names = "default";pinctrl-0 = <&hdmi_hpd>; status = "okay"; ddc = <&i2c_ddc>;
hpd-gpios = <&gpx3 7 GPIO_ACTIVE_LOW>;
hpd-gpios = <&gpx3 7 GPIO_ACTIVE_HIGH>;The gpx3-7 (HDMI-HPD) has external pull up to... so maybe it should be active low? How can it go high if it is already high? I am not sure if this matters because the driver ignores the flags and responds on both falling and rising edge.
HDMI specification says HPD is active high! Arndale schematics are quite misleading - pull-up resistors near level shifter are marked as not-connected :) so there are no pull-ups. And there is pull-down resistor on right side of the level shifter.
Ah, I was looking at schematics and looking and looking and still did not see "NC"... My mistake.
And finally with this patch it works as expected, without it it works sometimes.
Everything is good. I already sent last pull request so this will go after merge window, for v4.20 (or whatever number it will be).
Best regards, Krzysztof