12:35 <warpme> apritzel: i have Q regarding booting tx1 from emmc: i got state where tx1 emmc has written mainline bootloader + my distro rootfs. spl from emmc goes ok but uboot seems to using wrong device (so can't find partitions etc). Uart says: https://gist.github.com/warpme/13db7046a0df4ee96ea1d07051390b03 The same uboot binary boots ok mainline kernel + rootfs from usb key. Do you have idea where issue might be?

13:09 <warpme> qll - got it working. (so now watching 4k move on this tiny half credit-card box) :)

13:19 <karlp> so what was it?

13:19 <karlp> I remember lots of qurikiness converting most examples for sdcards into emmc, and underrstanding the way emmc hard partitions are different to file system partitions within an emmc partition

13:50 <warpme> karlp: i was bitten again by h616 uboot emmc issue :-(

17:06 <Jookia> does anyone have a H6 or H616 they could test some HDMI support patches on? :)

17:18 <dok> Jookia: i got a xunlong orangepi lite2 (sun50i... h6)

17:19 <Jookia> dok: nice! when i post the patch series i'll link you and give instructions for adding a device node

17:20 <Jookia> unfortunately i don't have a board to actually debug the HDMI patch, so i'm just making it HOPEFULLY work

17:21 <dok> just mind that i never ran anything on this devices, except my barebox port

17:22 <dok> so please include instructions on how to: build & execute

17:23 <Jookia> oh that'll be a bit trickier. i'll try

17:31 <dok> if i have any question i'll ask you

18:16 <Jookia> i just realized for the H6 and H616 i need different HDMI clocks

18:59 <Jookia> actually i'll just ignore HDMI for now

19:30 <apritzel> Jookia: if you have any patches, I'd be more than happy to test them

19:30 <apritzel> or even fix them

19:31 <Jookia> apritzel: oh, i deleted them already. oops. i'll check my git history

19:31 <Jookia> i'll re-write the code and append it to the patch series im doing

19:45 <Jookia> oh, my patchset is basically done. guess i'll take a quick break then add the HDMI code then send it off

19:51 <Jookia> apritzel: what's the general feeling about posting WIP patches that you can't test? i can't actually test the H6 support for my patchset but it's basically the same as the D1 so i'm adding it 'for free'. but i can't test HDMI at all

19:54 <Jookia> i feel like it's still probably worth it just because of the amount of time they take :\

19:56 <apritzel> Jookia: yes, definitely, and RFC is the probably the right thing to use

19:56 <apritzel> I would be very happy to test H6 and H616 HDMI patches, and have quite some boards around here

19:57 <Jookia> do you have any H6 with LCD boards to test? :)

19:58 <apritzel> ah, no. I mean the PineH64 has LCD connectors, but I don't have any actual displays

19:58 <Jookia> that's fine

19:59 <Jookia> if it doesn't work i'm sure someone will complain

19:59 <Jookia> testing on *both* H6 and H616 would really help as each has its own clocking config

20:01 <apritzel> plenty of both here ;-)

20:01 <Jookia> what are your thoughts on using DM in SPL to solve the SPI NAND problem where we'd have to add yet another chip-specific piece of spl code for spi nand?

20:02 <Jookia> chips like the T113 seem to be able to handle DM in SPL

20:02 <apritzel> I am quite hesitant to do this, as this just creates a bunch of new problems

20:03 <Jookia> oh ok

20:03 <apritzel> for a start: there are still the old SoCs, which don't have enough SRAM, so we would need to support both non-DM and DM SPL, which increases the test matrix

20:03 <apritzel> secondly: a lot of hardcoded SPL info is not in the DT, like DRAM parameters, PLL config, and so on

20:04 <apritzel> and also: I consider the SPL a continuation of the BROM, just with DRAM enabled

20:04 <Jookia> i thought there was a push to move to SPL_DM?

20:04 <apritzel> and the BROM is per SoC, so board agnostic, and doesn't need any DT

20:05 <apritzel> push from whom? smaeul_ wanted it, and maybe Simon Glass, but there is no real push otherwise, I'd say

20:05 <Jookia> ah ok

20:05 <apritzel> and my feeling is we just swap one set of problem for another one

20:06 <Jookia> i think some code needs refactoring for a good SPL SPI NAND situation but i have no idea where to start

20:25 <apritzel> Jookia: btw: if you have WIP patches that you want to have tested, just push them to some repo (your github, for instance), and drop a link here

20:25 <apritzel> that's probably less work than preparing a full series, especially if you are unsure about things

20:37 <Jookia> apritzel: i need to get in to the habit of preparing a series anyway