05:18 <bashintosh> Does anybody know how to disable HDMI and only leave eDP and VGA-out enabled on Firefly (RK3288)? I tried various things in the Firefly device tree (kernel) but no luck. I posted a thread on the Firefly forum -> http://bbs.t-firefly.com/forum.php?mod=viewthread&tid=574&extra=page%3D1

05:19 <amstan> bashintosh: heh, that's a tricky one

05:19 <amstan> bashintosh: how about on line 68

05:19 <amstan> you type status = "disabled"

05:21 <amstan> alright.. not tricky

05:21 <amstan> it was trickier to remove all signs that edp was ever there actually

05:22 <bashintosh> Was exactly the first thing I tried - the board boots but no video out on any port - somehow it seems that if you disable the HDMI node, VGA dies as well. This is even if you change the VGA source to 1 at line 30

05:23 <bashintosh> amstan: uh, that must have been a lot of clean-up

05:23 <amstan> oh, then that dts doesn't sound very healthy

05:24 <amstan> there's something that needs to be enabled for edp, that hdmi probably keeps it enabled

05:24 <amstan> so hdmi is probably supporting both, when edp should be responsible for it too

05:24 <amstan> bashintosh: see any errors?

05:24 <amstan> during bootup?

05:25 <bashintosh> I struggle to understand some sections - or better the relation between some nodes and the actual ports on the board - the fact is that the RK1000 seems to be quite involved - there is no TV-OUT on the Firefly so what does the RK1000 do? Is it related to VGA?

05:25 <amstan> oh, it's quite complicated, struggling is normal

05:25 <bashintosh> ah ah!

05:25 <bashintosh> true!

05:26 <amstan> bashintosh: this is edp only: https://chromium.googlesource.com/chromiumos/third_party/kernel/+/chromeos-3.14/arch/arm/boot/dts/rk3288-veyron-edp.dtsi

05:26 <bashintosh> amstan: yeah good tip, will dig into the boot logs as soon I get back to the hardware!

05:26 <amstan> for the chromebooks, but it's what needed to be done to separate edp without hdmi getting broken

05:26 <amstan> i don't have a similar fragment for hdmi though, because we always have that enabled

05:27 <amstan> bashintosh: have you also considered a non-kernel way to do this?

05:27 <amstan> with xrandr perhaps?

05:30 <amstan> another way you could trick it off is by misconfiguring the hpd pin

05:30 <bashintosh> That's interesting as option but too high level - involves X but I'd like to keep this change low level, as close as possible to the hardware. Maybe in the drivers...

05:30 <amstan> which is the way it knows there's something connected on hdmi

05:30 <amstan> misconfigure that and it won't know there's a cable connected

05:30 <bashintosh> amstan: that's a good one!

05:31 <amstan> bashintosh: what kernel is this anyway? i'm not familiar with any of those dt properties

05:31 <amstan> 3.10?

05:31 <amstan> i strongly suggest upgrading if so

05:31 <bashintosh> Ah sorry yeah 3.10 for Android 4.4

05:31 <bashintosh> to which ver?

05:31 <amstan> because the hdmi disabled would have just worked otherwise

05:32 <amstan> if you need android i guess you just need 3.10

05:32 <amstan> but ugh..

05:32 <bashintosh> amstan: I felt that would have been logic

05:33 <bashintosh> Yes Android, hopefully 5.x soon!

05:33 <amstan> :(

05:41 <bashintosh> I still have to try the preview version of Lollipop for the Firefly - wonder how much work is still needed to finish porting. Might be cool to help!

11:09 <naobsd> debug uart shares pins with sd slot?

11:09 <naobsd> wow

11:09 <naobsd> GPIO2_5/6

11:10 <naobsd> it means that we can get uart2 on all RK3368 devices which has sd slot

11:12 <ganbold> naobsd: so does rockchip replace their bootloader with u-boot or they have still their own bootloader?

11:13 <naobsd> I can see rk3368 support in RK's u-boot (in rk312x lollipop sdk), but I'm not sure which bootloader is used for production.

11:13 <naobsd> I guess it's u-boot, but no idea for now.

11:14 <naobsd> what I talked now is about uart2 pins

11:15 <naobsd> and I have no idea how to configure pins as uart on product which has sd slot with stock loader. pins should be configured as sd pins ;)

11:26 <naobsd> hmm, RK3368 box is $120?

11:40 <naobsd> mmind00: [PATCH 4/4] dt-bindings: document rk3368 R89 board from Rockchip <- s/R89/R88/

11:40 <mmind00> naobsd: nope ... the arm64 board is called R89

11:41 <naobsd> then, s/R88/R89/ in patches...

11:42 <mmind00> nope ... the rk3288 board is called R88, or did you see more r88 mentions in the other patches?

11:42 <mmind00> naobsd: oh you're right

11:42 <naobsd> "With these dts files, the R88 board is able to boot from an attached

11:42 <naobsd> usb device" very interesting

11:43 <mmind00> naobsd: totally confused now ... just looked at the board again ... really called r88

11:43 <naobsd> :)

11:44 <naobsd> mmind00: could you show me dmesg with linux-next ? :)

11:46 <mmind00> naobsd: https://bpaste.net/show/daf62f32a1bc

11:48 <naobsd> "LoadTrustBL" hmm

11:51 <naobsd> "SMP: Total of 8 processors activated." nice :)

11:53 <naobsd> mmind00: really interesting, thanks :)

12:01 <naobsd> "CPU: All CPU(s) started at EL1" hm

12:01 <naobsd> Exception Level

12:06 <naobsd> I don't know well about aarch64

12:06 <naobsd> TrustBL runs at EL3, U-Boot runs at EL2, Linux kernel runs at EL1?

12:07 <naobsd> I'll check RK's u-boot...

23:45 <bashintosh> Does anyone have the TRM for RK3288? Seems impossible to find..

23:51 <naobsd> bashintosh: http://bbs.t-firefly.com/forum.php?mod=viewthread&tid=275

23:58 <bashintosh> naobsd: thanks! I had that link - it's useful but missing lots of things, for example the USB controller and picoPHY architecture which I am looking into - I found the one for RK3188 and I assume that many sections are similar to RK3288 but it's hard to tell precisely. Here's a useful resource, but no RK3288 TRM :) http://rockchip.fr/