Date: Thu, 6 Jul 2017 18:40:13 -0300 From: "Dr. Rolf Jansen" <rj@obsigna.com> To: freebsd-arm@FreeBSD.org Subject: Re: Enabling ADC on a Beaglebone Black running FreeBSD 12.0-CURRENT (BEAGLEBONE) Message-ID: <6DE3312D-BA5B-4192-88C4-1F0C800CBFF6@obsigna.com> In-Reply-To: <79EE9798-BCF0-4585-93F5-4E604142561F@obsigna.com> References: <0C4DCBB9-2642-4B0F-B15B-4139D5D8B249@obsigna.com> <271AFD8F-BD2C-445C-AB95-D7D07593E487@obsigna.com> <5D2FEB0D-64F3-488C-8458-85E7DF10EFB7@obsigna.com> <20170330202858.GA22253@bluezbox.com> <79EE9798-BCF0-4585-93F5-4E604142561F@obsigna.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> Am 30.03.2017 um 18:11 schrieb Dr. Rolf Jansen <rj@obsigna.com>:
>=20
> Am 30.03.2017 um 17:28 schrieb Oleksandr Tymoshenko =
<gonzo@bluezbox.com>:
>> Dr. Rolf Jansen (rj@obsigna.com) wrote:
>>> Today I updated once again my Beaglebone Black by
>>> merging-in the latest FreeBSD 12.0-CURRENT (BEAGLEBONE)
>>> snapshot, and once again I enabled the ADCs on the
>>> Beaglebone in the device tree blob am335x-boneblack.dtb,
>>> and the ADC is still working fine.
>>>=20
>>> Even if it is not that a big hassle to modify the device
>>> tree blob, I am curious on why the ADC has been disabled
>>> in the blob in the first place, end even more, given the
>>> fact that the device ti_adc driver is built-in to kernel
>>> and once enabled, the ADC is functional.
>>>=20
>>> Didn't it work at some time in the past? Now it is.
>>>=20
>>> What is missing to activate the ADC in the device tree
>>> blob by default?
>>=20
>> Few months ago FreeBSD switched to using upstream DTB files
>> instead of custom-made ones. For some reason ADC is disabled
>> in upstream. If you're running recent FreeBSD you can use
>> dtb overlays to enable ADC without hassle of maintaining
>> custom dts file. You can do following:
>>=20
>> 1. Create am335x-beaglebone-tscadc.dts with following
>> content:
>>=20
>> /dts-v1/;
>> /plugin/;
>>=20
>> / {
>> compatible =3D "ti,am335x-bone-green", "ti,am335x-bone-black", =
"ti,am335x-bone", "ti,am33xx";
>>=20
>> fragment@0 {
>> target =3D <&tscadc>;
>> __overlay__ {
>> status =3D "okay";
>> adc {
>> ti,adc-channels =3D <0 1 2 3 4 5 6>;
>> };
>> };
>> };
>> };
>>=20
>> 2. Compile overlay:
>> $ dtc -I dts -O dtb -o am335x-beaglebone-tscadc.dtbo =
am335x-beaglebone-tscadc.dts
>>=20
>> 3. Copy it to /boot/dtb/ directory on your BBB
>>=20
>> 4. Enable overlay in /boot/loader.conf by adding following line:
>>=20
>> fdt_overlays=3D"am335x-beaglebone-tscadc.dtbo"
>=20
>=20
> Oleksandr, thank you very much for drawing my attention to the dtb =
overlay facility.
>=20
> I created and enabled the adc-dtb-overlay file as you explained above, =
and it works.
>=20
> Best regards
>=20
> Rolf
With the last two 12.0-CURRENT BBB snapshots (r320360 and r320599), the =
dtb oberlay facility is broken. I can get the above overlay working only =
if I compile it with the dtc tool from the previous June, 19th snapshot =
(r320095), and in addition it is necessary to re-add the phandle to the =
tscadc@44e0d000 definitions of the base dtb file. Said phandle was =
removed together with almost all phandles of other device definitions =
between r320095 and r320360.
Since I need to edit the base dtb anyway, for the time being, I enable =
the ADC and its channels right in there. This one can be compiled with =
the new dtc, and in addition the ADC phandle is not necessary.
By the way, the decompiler of the new dtc commes with a formatting =
glitch. The sections are no more devided by an empty line, instead an =
empty line is introduced below the section headers -- this looks spoiled =
somehow. Anyway, here are my working changes:=20
--- am335x-boneblack.dts.orig 2017-07-06 18:32:10.321019000 -0300
+++ am335x-boneblack.dts 2017-07-06 16:00:22.255994000 -0300
@@ -1711,7 +1711,7 @@
interrupt-parent =3D <0x1>;
interrupts =3D <0x10>;
ti,hwmods =3D "adc_tsc";
- status =3D "disabled";
+ status =3D "okay";
dmas =3D <0x29 0x35 0x0 0x29 0x39 0x0>;
dma-names =3D "fifo0", "fifo1";
tsc {
@@ -1722,8 +1722,19 @@
=20
#io-channel-cells =3D <0x1>;
compatible =3D "ti,am3359-adc";
+ ti,adc-channels =3D <0 1 2 3 4 5 6 7>;
};
};
+ pruss@4A300000 {
+
+ compatible =3D "ti,pruss-v2";
+ ti,deassert-hard-reset =3D "pruss", "pruss";
+ reg =3D <0x4a300000 0x80000>;
+ ti,pintc-offset =3D <0x20000>;
+ interrupt-parent =3D <0x1>;
+ interrupts =3D <0x14 0x15 0x16 0x17 0x18 0x19 =
0x1A 0x1B>;
+ status =3D "okay";
+ };
gpmc@50000000 {
=20
compatible =3D "ti,am3352-gpmc";
@@ -1969,6 +1980,7 @@
ehrpwm2_tbclk =3D =
"/ocp/l4_wkup@44c00000/scm@210000/scm_conf@0/clocks/ehrpwm2_tbclk@44e10664=
";
ehrpwm0_tbclk =3D =
"/ocp/l4_wkup@44c00000/scm@210000/scm_conf@0/clocks/ehrpwm0_tbclk@44e10664=
";
tscadc =3D "/ocp/tscadc@44e0d000";
+ pruss =3D "/ocp/pruss@4A300000";
clkout2_pin =3D =
"/ocp/l4_wkup@44c00000/scm@210000/pinmux@800/pinmux_clkout2_pin";
aes0_fck =3D =
"/ocp/l4_wkup@44c00000/scm@210000/scm_conf@0/clocks/aes0_fck";
gpio3_dbclk =3D =
"/ocp/l4_wkup@44c00000/prcm@200000/clocks/gpio3_dbclk@b4";
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?6DE3312D-BA5B-4192-88C4-1F0C800CBFF6>