00:08 <brads> xdarklight: It would to nice to see the DWC2 fix applied to stable. Also ethernet devicetree fix for for performance issues would be good also (although can mostly mod the devicetree).

00:49 <chewitt> xdarklight: thanks for pushing that fix through

00:50 <chewitt> xdarklight: how's video for m8* coming along?

00:51 <chewitt> LE plans on dropping 3.10 kernel devices from its line-up after our pending 9.0 release

00:51 <chewitt> the older 8726MX devices aren't anything to lose sleep over

00:52 <chewitt> but we have quite a few people with S812 boxes that it would be nice to resurrect support for

02:52 <brads> I think I understand the strange ethernet LED issue I have been seeing on the Odroid C2. I have CONFIG_IP_PNP configured along with realtek phy and stmmac drivers compiled in rather than modules for early boot nfs root filesystem support.

02:57 <brads> Normally the registers for RTL8211F led (LED control register (LCR)) are set in mainline uboot driver and carried on to linux, but if I happen to reset the phy later (trying to load the module version) it reverts to default realtek LCR registers

03:01 <brads> This is where mainline uboot sets it - http://git.denx.de/?p=u-boot.git;a=blob;f=drivers/net/phy/realtek.c;h=b3e6578df9acbf26878dd00ba9a95cebc3c7ce86;hb=HEAD#l145

03:09 <brads> 0x617f seems to be what uboot sets and 0x8A46 is rtl8211x default in linux (which is what I seem to see if it gets reset after being reset early boot) - http://files.pine64.org/doc/datasheet/rock64/RTL8211F-CG-Realtek.pdf

03:09 <brads> Section 8.4.27. (p45)

03:58 <brads> I guess kernel does not touch led's its in uboot and the os

04:05 <brads> would be nice to have ethernet led config in board dt for rtl8211F maybe?

04:09 <brads> requiring changes to the realtek phy driver?

06:06 <Net|> https://netpipe.ca/aha/programs.tar.gz thats my stash of program source ment to be viewed as an archive

08:20 <xdarklight> brads: good job detective ;)

08:21 <xdarklight> brads: yes, the realtek PHY driver requires changes for that - but those changes should be straight forward

08:37 <warpme_> guys: what is status mainline 4.20 kernel regarding HDMI audio on s905? Should I apply any patches like for 4.19 (i.e. https://github.com/jeromebrunet/linux/commits/v4.19/aml/gx-audio-wip)?

08:49 <xdarklight> warpme_: any kind of audio for S802, S805, S812, S905, S905X and S912 is not in mainline yet (and also not queued for v4.21)

08:49 <xdarklight> (S905D as well...)

08:50 <warpme_> xdarklight: ok got it. but for 4.19 patches from Neil are working ok for me. Unfortunatelly I can’t find such patches for 4.20…

08:51 <xdarklight> warpme_: I'm not sure how much has changed between 4.19 and 4.20. in other words: the 4.19 patches *may* still apply onto 4.20

08:52 <warpme_> yes, I applied them - but kernel reports no any sound card…

08:57 <warpme_> btw: what is reason not mainlining sound support for amlogic?

08:58 <xdarklight> warpme_: bad luck then :/ I guess Jerome or Neil will update the audio patches for the yocto build

08:59 <xdarklight> warpme_: last time I asked the reason why audio support is not mainlined yet is because the code still contains hacks

08:59 <xdarklight> and there's no time to fix these