00:00 <apritzel> and the AXP717 is a bigger PMIC, which might be related to the AXP2202.

00:01 <apritzel> We have seen that in the past: AXP1530 is apparently just another name for the AXP313, IIRC

00:25 <apritzel> macromorgan: there are also several hits in a github codesearch for axp2202

01:19 <macromorgan> okay, I'll check for that... I'm working on bringup of an H700 device and that was one of the main bits that stuck out to me

01:19 <macromorgan> The AXP2202, the display engine, and the audio codec look like they'll be needed but are missing from mainline. I'm not sure how similar it is to other existing H61x series stuff, but the address space is all the same.

01:34 <macromorgan> good call on the searching for the axp2022 in github. It's not perfect but I've found a few 4.9 kernels and even a 5.4 BSP kernel with a driver for that PMIC.

01:34 <macromorgan> only downside is, without a datasheet I'm not sure I'd be willing to take on the burden of upstreaming it...

01:35 <macromorgan> https://github.com/sputnik2019/tina-linux-5.4/blob/main/drivers/mfd/axp2101-i2c.c#L73

10:59 <apritzel> macromorgan: this is just the MFD driver part, which is fairly easy, though still quite far from mainline

11:00 <apritzel> macromorgan: having a BSP *regulator* driver should give you all the bits you need for mainlining, the manual doesn't add too much to it, since in general the AXPs are quite similar

11:01 <apritzel> have a look at that: https://lore.kernel.org/linux-sunxi/20230524000012.15028-1-andre.przywara@arm.com/

11:01 <apritzel> that's the AXP313 patch series. You could copy&paste that and replace the details with what you learned from the BSP source

11:02 <apritzel> at the end it's about what rails there are, where their control bits live and what the ranges and encodings for each of them are

11:03 <apritzel> if you are lucky the BSP driver actually uses similar data structures as mainline, then you might be able to copy those bits in