05:25 <Jookia> editing the device tree to remove unused tcons and mixers seems to have actually fixed my lcd issue

05:46 <Jookia> it's looking like if i strip out anything mixer1 related the issue gets fixed

06:15 <Jookia> just for notes it looks like this also solves another person's completely different but fatal LCD issue

06:16 <Jookia> while my LCD was only blue tinted, theirs was fully green and unusable

08:30 <tokyovigilante> weird, although how sure are you your device has mixer1? the H616 et al only seem to have mixer0 (which I assume means they can only render the same content to each output, ie display mirroring only, no support for two seperate FBs)

08:30 <Jookia> not sure. the datasheet i suppose

09:25 <apritzel> Jookia: is that possibly some difference between the T507 and H616? The H616 user manual shows one TCON connected to the RT-mixer, whereas the T5 series manual shows two "DE"s, in "Figure 4-1. DE Block Diagram"

09:25 <Jookia> hmm. this is the D1/T113

09:26 <apritzel> ah, right, let me check there

09:34 <apritzel> so indeed D1/T113-s3 have two mixers, H616/T507 have only one

09:36 <Jookia> my u-boot code works fine because we only use mixer0

10:14 <apritzel> but there is nothing actually using mixer1, right? So it is just described in the DT, and the Linux driver initialises it, I guess, but it's not referenced otherwise in the DT?

10:15 <Jookia> tcons and hdmi have endpoints for mixer1

10:35 <apritzel> but HDMI is not enabled in your DT, right?

10:35 <Jookia> no

21:04 <apritzel> macromorgan, tokyovigilante, loki666: can you test pure mainline on the Anbernic devices? -rc3 is missing the AXP717 fix series, of which I believe patch 1/5 is critical

21:04 <apritzel> "mfd: axp20x: AXP717: Fix missing IRQ status registers range"

21:05 <apritzel> I don't remember exactly what the symptom was, whether it was just the power key(?) not working, or the whole AXP717 driver bailing out and the board not coming up?

21:20 <tokyovigilante> I’m away til Monday, but can do. I have had previous iterations running with just the mainlined portions, but all my recent work has been with -rc2 with those patches on top as my working tree. I have not seen any circumstance previously where the driver flatly failed, but loki666 has had quite a bit of trouble, not sure whether that was that missing patch set or just a config issue.

21:24 <apritzel> tokyovigilante: thanks! I try to remember whether the driver failed probing at all, or whether it was just the IRQ parts not working. I thinks it's the former, because initialising the IRQs didn't work, and IIRC this returns an error code