04:20 <ndufresne> ok, stupid quesiton maybe, but how do we convince a 2015 hacked up uboot to inject recent kernel image/dtb ?

04:21 <ndufresne> * to injest

04:21 <ndufresne> I get some bad magic on the FDT side

04:25 <ndufresne> ok, progress

04:26 <ndufresne> seems like the fdt are warning because of the multi-dtb hack

04:27 <ndufresne> hmm, I think I have have wrongly picked my load address

04:53 <ndufresne> no luck today, it got stuck at "Starting kernel ... uboot time: 326591012 us" ...

05:13 <memeka> ndufresne: just curios, why all this hassle and not just get a 3399 board?

05:13 <memeka> the non-tvbox kind?

06:15 <chewitt> ndufresne: have a look at https://github.com/chewitt/LibreELEC.tv/tree/amlogic/projects/Amlogic/packages/aml_autoscript/scripts

12:14 <ndufresne> chewitt, that part is solved, my I think my kernel just does not boot, hard to say right now no traces ;-P

12:15 <ndufresne> memeka, because it's not an amlogic, would be hard to work on amlogic codec (in gstreamer) with a Rockchip 3399 ;-P

12:16 <ndufresne> I already have access to a Firefly and a libretech renegate, but codec aren't ready

14:04 <narmstrong> ndufresne: did you pass earlycon on the cmdline ?

14:07 <ndufresne> narmstrong, ah, earlyprintk isn't the same I guess ;-P

14:08 <ndufresne> what I was to do was to first ensure this build is fine on the potato, which I know works ;-P

14:08 <narmstrong> ndufresne: nop, earlycon uses the first defined uart earlycon driver

14:08 <ndufresne> (for the context, I works about 1h per day on this, so ....)

14:08 <ndufresne> narmstrong, ah, that might explain

14:08 <narmstrong> earlyprintk is legacy

15:27 <memeka> ndufresne: then existing boards?

16:07 <ndufresne> memeka, sometimes someone need to renew his knowledge, and playing with an unknown board on spare time works for that, because you get some blockers and that's where you learn

16:08 <ndufresne> (it's also a recycling project, that board, which btw, is just a q201 3g, was rotting in some drawer