03:42 <Guest16> I will try applying your role switch patch

10:06 <Guest16> CounterPillow is there any way I can force the device to just stay in host mode and ignore all other usbc stuff? i tried adding fusb302 nodes to the dt but that didnt seem to help.

10:06 <Guest16> I saw  some work Kwiboo did https://libera.catirclogs.org/linux-rockchip/2024-05-03#. was there any success to that?

10:06 <Guest16> Or can i add a temporary patch in the naneng combophy to force the device to be in host mode?

12:09 <CounterPillow> dr_mode = "host"; in the DT should do it I think

17:52 <Guest16> CounterPillow: The tree should already have that. I've tried adding the FUSB302 entried to the device tree but they dont seem to help

20:24 <Ermine> Hello, so I was fucking with MPP kernel driver this week. It pokes in power management and clk and opp stuff (and in also that rockchip_dmc thing...). Are there power management bits that need to be upstreamed, in case if future upstream rkvenc/rkvdec drivers would need them?

20:24 <Ermine> At least rockchip tree has more smc calls defined than mainline

23:38 <CounterPillow> rkvdec for several SoCs and codecs either already exists upstream (rk3399, rk3328) or has been sent to the list (rk3588/rk3576) by detlevc, so I think we're good on that front. Usually it's best to implement some of the more intricate power/clock stuff when you have a way to test it upstream, i.e. a user of it; vendor code has a tendency to do more things than necessary or do them in a different way than what upstream would want for

23:38 <CounterPillow> such a solution