03:49 <chewitt> @lvrp16 ^

16:29 <aeroraptor> chewitt: hah, in #librecomputing lvrp16 had suggested I ask that question in here....

18:55 <xdarklight> aeroraptor: out of curiosity: what are you trying to achieve? the public S905X datasheet is "vague" (I'm being kind with my wording here) what this pad supports and how to configure it. our pinctrl driver is missing at least the pinmux function configuration

19:01 <aeroraptor> I want to use it to characterize the stability of the SoC and its crystal (by feeding it into a frequency counter, then performing analysis on the results) so that I can properly quantify changes to the system after I replace the crystal with a much better TCXO

19:02 <aeroraptor> there are a few different ways to measure a system's stability, and my plan is to use a few of them before and after the swap so that I can see how consistent the changes are

19:03 <aeroraptor> and being able to pull the 24MHz directly through the SoC is one of the best indicators

19:07 <xdarklight> aeroraptor: if I understand the vendor driver code (which is not given ;-)) I think all that's needed is setting the correct bits in the register you mentioned. if it's only an experiment then you can try this hack: https://pastebin.com/raw/DjhHGSrN (not tested at all)

19:07 <xdarklight> aeroraptor: my understanding is that this should output XTAL (24MHz) on GPIOCLK_0

19:12 <aeroraptor> neat, thank you for pulling this together - it is indeed just for an experiment.

19:13 <aeroraptor> I haven't done any kernel development on this board yet (only had it a few days) so it's probably going to take me a little while to actually test this

19:16 <xdarklight> sure, keep me updated as I'm curious what you'll find (whether this works or not)

19:21 <aeroraptor> yeah absolutely, I'm pulling down the libretech-linux repo to the board now... gonna take a little while

19:22 <aeroraptor> one of these days I might get around to getting my distcc environment running again so that this sort of thing can be a bit less painful

19:37 <xdarklight> when you're saying "libretech-linux repo": which one is that? for clarification: my patch is based on mainline (Linus Torvalds' Linux tree)

19:55 <aeroraptor> https://github.com/libre-computer-project/libretech-linux

19:56 <aeroraptor> meson-eeclk.c is still present at drivers/clk/meson/

19:57 <xdarklight> if you're using the master branch then it'll be fine, it seems to be following mainline

19:58 <aeroraptor> right on

21:41 <aeroraptor> well I think I'm going to have to table this for a while, I'm getting nowhere fast on building this kernel

21:42 <aeroraptor> and even then I don't know how to point the system at it, I have no idea how the le potato picks its kernel to boot from