[PATCH] arm: dts: bcm283x: Allow UARTs to work before relocation

[Simon Glass]

Hi Tom,

On Thu, 2 Jan 2020 at 20:59, Tom Rini <trini at konsulko.com> wrote:
>
> On Thu, Jan 02, 2020 at 07:28:50PM -0700, Simon Glass wrote:
> > Hi Tom,
> >
> > On Thu, 2 Jan 2020 at 19:27, Simon Glass <sjg at chromium.org> wrote:
> > >
> > > Hi Tom,
> > >
> > > On Thu, 2 Jan 2020 at 16:11, Tom Rini <trini at konsulko.com> wrote:
> > > >
> > > > On Tue, Dec 31, 2019 at 02:28:19PM -0700, Simon Glass wrote:
> > > > > Hi Matthias,
> > > > >
> > > > > On Tue, 10 Dec 2019 at 02:43, Matthias Brugger <mbrugger at suse.com> wrote:
> > > > > >
> > > > > > Hi Simon,
> > > > > >
> > > > > > On 02/12/2019 16:45, Tom Rini wrote:
> > > > > > > On Sun, Dec 01, 2019 at 07:33:56PM -0700, Simon Glass wrote:
> > > > > > >
> > > > > > >> At present the pinctrl nodes are not enabled in pre-relocation U-Boot so
> > > > > > >> the UARTs do not correctly select the pinconfig to enable the UART pins.
> > > > > > >> Fix this so that the U-Boot banner is printed.
> > > > > > >>
> > > > > > >> Signed-off-by: Simon Glass <sjg at chromium.org>
> > > > > > >> Fixes: 9821636b64 (bcm2835_pinctrl: Probe pre-reloc)
> > > > > > >> ---
> > > > > > >>
> > > > > > >>  arch/arm/dts/bcm283x-u-boot.dtsi | 8 ++++++++
> > > > > > >>  1 file changed, 8 insertions(+)
> > > > > > >>
> > > > > > >> diff --git a/arch/arm/dts/bcm283x-u-boot.dtsi b/arch/arm/dts/bcm283x-u-boot.dtsi
> > > > > > >> index 36548dad62..68d03627f4 100644
> > > > > > >> --- a/arch/arm/dts/bcm283x-u-boot.dtsi
> > > > > > >> +++ b/arch/arm/dts/bcm283x-u-boot.dtsi
> > > > > > >> @@ -19,3 +19,11 @@
> > > > > > >>  &gpio {
> > > > > > >>      u-boot,dm-pre-reloc;
> > > > > > >>  };
> > > > > > >> +
> > > > > > >> +&uart0_gpio14 {
> > > > > > >> +    u-boot,dm-pre-reloc;
> > > > > > >> +};
> > > > > > >> +
> > > > > > >> +&uart1_gpio14 {
> > > > > > >> +    u-boot,dm-pre-reloc;
> > > > > > >> +};
> > > > > > >
> > > > > > > I think this is superseded by the RPi PR that I had been testing and
> > > > > > > just now pushed.  Can you confirm that master is fine on your Pis as
> > > > > > > well?  I gather you hit this failure doing pytest on the board, which is
> > > > > > > also how I found it.  Thanks!
> > > > > > >
> > > > > >
> > > > > > Can you confirm if this is working with master branch or if we still need your
> > > > > > patch. In which situation would we need your patch then?
> > > > >
> > > > > Unfortunately with the 'next' branch I still need this patch. WIthout
> > > > > it I don't get the pre-relocation serial output and tests fail.
> > > >
> > > > Which Pi exactly?  My 3B works with rpi_3_32b, rpi_3 and rpi_arm64
> > > > defconfigs and test.py.
> > >
> > > It is a rpi_3b I believe, and I am using the rpi_3_32b.
> > >
> > > When I cherry-pick the patch the banner appears but before that it
> > > doesn't. Shall I check it again?
> >
> > Actually, can you please confirm the tree you are using? I was testing
> > with -next. I wonder if I was missing a patch?
>
> I just re-confirmed by checking out current 'next' and re-running
> test.py on all of the above combinations and they pass.

I'm reviving this thread as I still have problems. I have these in config.txt:

enable_uart =1
gpu_freq=250

Have you tried with those?

Separately, on latest mainline I don't see any serial output at all
from U-Boot with this config although linux does boot. I'm taking a
look as the culprit is the ofdata commit.

Regards,
Simon