01:07 <iscle> I have some good news! The USB DMA driver appears to be partially working :) However, it crashes after finishing the transfer because of some null pointer I have yet to find/fix. However, I like it, I'm on the right track!

01:08 <iscle> Time to get some sleep now :)

03:38 <tokyovigilante> apritzel: I found that on the H616/H700, audio playback was 50% of correct, and needed the PLL_AUDIO_1X and PLL_AUDIO_4X dividers to be doubled to compensate

07:05 <machinehum> https://lore.kernel.org/lkml/568ABCBD.3070909@gmail.com/T/#m115673e150607f500bc2e4272f51c56c2d1b85fe

07:06 <machinehum> Is that "useful" to anyone? I've found a bunch of ttyxyz nodes in my /dev/ (far more than my desktop machine, far more than 64) and it's polluting the directory

07:08 <machinehum> Also, my latest Sunxi creation if anyone is interested: https://www.reddit.com/r/hacking/comments/1fpu2du/flipper_blackhat/

07:13 <gamiee> machinehum : this is amazing I love it

07:13 <gamiee> Great job

07:20 <machinehum> ahah thanks :)

07:21 <machinehum> Seems like people were interested in it, so I'll probably be making more

08:36 <dsimic> machinehum: hmm, I went quickly through the ML thread that you linked above

08:37 <dsimic> basically, my train of thought could be expressed with these messages already in the thread:

08:37 <dsimic> https://lore.kernel.org/lkml/20160104225544.158dc847@lxorguk.ukuu.org.uk/

08:37 <dsimic> https://lore.kernel.org/lkml/20160105152547.GC7251@kroah.com/

08:37 <dsimic> https://lore.kernel.org/lkml/20160105204745.760727c1@lxorguk.ukuu.org.uk/

10:38 <apritzel> tokyovigilante: ah, I see, thanks. I don't see that same bug on the H616, but it looks a bit fishy: we force several PLL_AUDIO bits to fixed values, but am not sure we communicate the implications correctly to the kernel

10:54 <apritzel> so if anyone is playing with audio on the D1/T113s, please have a look at this patch: https://lore.kernel.org/linux-sunxi/20241001105016.1068558-1-andre.przywara@arm.com/T/#u

10:58 <apritzel> this is a clear fix, but it would double the PLL frequency, so I am wondering if we need to adjust somewhere else, or whether it never really worked before