[PATCH] ARM: dts: am57xx-beagle-x15: fix HDMI LS_OE GPIO

BeagleBoard-X15
1

I believe we have A2s being shipped as well. we dont want to break A2s either -> I propose we refactor the dts to introduce revb1 X15.

2

There are at-least 200 A2's that have been shipped to end users.

http://uk.rs-online.com/web/p/processor-microcontroller-development-kits/8874764/

Regards,

3

Thanks. Tomi: if you are ok - I already do have a refactoring code from older kernel. I will rebase to latest and repost.

Tony: is there a branch youd' like me to base the changes on?

4

Sounds good to me.

Tomi

5

* Nishanth Menon <nm@ti.com> [160901 07:15]:

> > >
> > > LS_OE GPIO was changed from gpio6_28 to gpio2_30 for the Rev B1 board,
> > > and the current .dts file has the old GPIO.
> > >
> > > Rev B1 is the earliest board supported, so we can just change the .dts.
> > >
> > > This fixes "omapdss error: HDMI I2C Master Error" which causes failure
> > > to read EDID from the monitor.

...

> > I believe we have A2s being shipped as well. we dont want to break A2s
> > either -> I propose we refactor the dts to introduce revb1 X15.
>
> There are at-least 200 A2's that have been shipped to end users.
>
> http://uk.rs-online.com/web/p/processor-microcontroller-development-kits/8874764/

Thanks. Tomi: if you are ok - I already do have a refactoring code from
older kernel. I will rebase to latest and repost.

Tony: is there a branch youd' like me to base the changes on?

Probably best to do it against current omapf-for-v4.9/dt as it has all
the skeleton.dtsi removal now applied.

Regards,

Tony

6

Ok will do, will try to post this later today or early tomorrow.

7

* Nishanth Menon <nm@ti.com> [160901 07:15]:

LS_OE GPIO was changed from gpio6_28 to gpio2_30 for the Rev B1 board,
and the current .dts file has the old GPIO.

Rev B1 is the earliest board supported, so we can just change the
.dts.

This fixes "omapdss error: HDMI I2C Master Error" which causes failure
to read EDID from the monitor.

...

I believe we have A2s being shipped as well. we dont want to break A2s
either -> I propose we refactor the dts to introduce revb1 X15.

There are at-least 200 A2's that have been shipped to end users.

http://uk.rs-online.com/web/p/processor-microcontroller-development-kits/8874764/

Thanks. Tomi: if you are ok - I already do have a refactoring code from
older kernel. I will rebase to latest and repost.

Tony: is there a branch youd' like me to base the changes on?

Probably best to do it against current omapf-for-v4.9/dt as it has all
the skeleton.dtsi removal now applied.

Ok will do, will try to post this later today or early tomorrow.

I'm getting this error again with the latest X15 image on my A2
version board. I don't see this patch in the mainline kernel. Did it
never get applied?

Thanks
Bill

8

With am57xx-beagle-x15.dts? That seems be the A2 version.

9

I'm getting this error again with the latest X15 image on my A2
version board. I don't see this patch in the mainline kernel. Did it
never get applied?

With am57xx-beagle-x15.dts? That seems be the A2 version.

Yes.
I see the following in the boot message

loading /boot/dtbs/4.9.35-ti-r44/am57xx-beagle-x15.dtb ...

I can get around this by inserting/uninserting an sdcard once the
kernel starts booting.

This is with the Debian 9.0 2017-07-02 4GB SD LXQT image.

Thanks
Bill

10

Weird, ldo1 is set to always on:

https://github.com/RobertCNelson/linux-stable-rcn-ee/blob/4.9.35-ti-r44/arch/arm/boot/dts/am57xx-beagle-x15-common.dtsi#L276-L283

Something else is going on..

Regards,