19:42 <sjg1> Tartarus: re your CI idea, it would not be that hard to compare DT files against their kernel equivalent, I suspect

19:44 <Tartarus> sjg1: that only works if the DTs are already in sync. I was kinda thinking if we validated them vs the kernel.org bindings (and adding what's needed for u-boot, ones) it would fail an that's how we'd catch things

19:45 <vagrantc> probably could use some touch-up

19:51 <sjg1> Tartarus: Do the bindings actually validate everything these days? We don't have such validation in U-Boot yet. We should add it

19:52 <Tartarus> good question I don't know the answer to

20:03 <sjg1> Tartarus: I don't know either, but I don't think so

20:04 <sjg1> Tartarus: Re a source-code comparison, it might just create a lot of noise, or it might help, not sure. But we could have a whitelist by board and ratchet it down like we do with CONFIG

20:04 <Tartarus> sjg1: Yeah. My kind of hope is that if I poke people about this, that were unaware, they'll be more proactive in the future.

20:05 <Tartarus> I'm just going to assume some of the PowerPC history is why at least some parts of NXP were unaware

20:45 <override> hey folks, any way standarized ways of updating the device tree file through u-boot?

20:45 <override> kernel even... something more runtime, rather than updating/writing partitions/ disk images

21:53 <sjg1> override: Can you be more specific?

21:53 <sjg1> Do you mean the 'fdt' command?

22:01 <mrnuke> override: "load $fdt_addr_r some/devicetree.dtb" "fdt addr $fdt_addr_r" "fdt ls" "fdt help"