00:33 <afaerber> xdarklight, just saying, whenever you're fixing some specific commit in a tree that's not rebasing, you should use Fixes, so then when someone is backporting commit X their scripts can find out that they need Y, too - haven't looked at the actual commit here

08:06 <narmstrong> xdarklight: did they also put a freq limitation on the vdec frequency ? They use 666Mhz on s905x

08:11 <xdarklight> narmstrong: at least that info is not in the .dts - I don't know whether that is coded in the driver somewhere

08:12 <xdarklight> oh wait

08:12 <xdarklight> vpu node, clk_level = <2>;

08:12 <xdarklight> that translates to 200.0M

08:13 <xdarklight> p212 = level 7, p281 = level 7, p241 = level 2

08:32 <xdarklight> narmstrong: is libretech going to release a board based on the S805X? I found this: https://libre.computer/products/boards/aml-s805x-ac/

09:09 <lvrp16> xdarklight: yes

09:36 <xdarklight> lvrp16: so price point will be even cheaper than S905W? meaning it'll compete with the allwinner devices?

09:36 <lvrp16> xdarklight: each has it's use case. s805x is very cut down.

09:37 <lvrp16> allwinner chips are for the most part fully featured

10:02 <narmstrong> lvrp16: what does the s805x datasheet about dissipation ? It could explain the lower frequencies used, the package could be way cheaper and could not dissipate like the one used in s905x

19:18 <lvrp16> narmstrong: idk about dissipation but the package is much smaller

19:22 <lvrp16> if my memory serves me correctly, s905x was 14mm x 14mm

19:22 <lvrp16> s805x is 11mm x 11mm

19:22 <lvrp16> same ball pitch so you have a lot fewer io pins