02:05 <TahomaSoft> Anyone have any experience with the RK3568? I'm working on supporting the LinkStar H68K1492v1 and am stuck getting the third of three PCIe buses to work. I could use some advice on working with the DTS.

02:10 <CounterPillow> mainline or vendor kernel?

02:13 <TahomaSoft> Mainline Openwrt kernel, which currently is 6.6.92

02:21 <TahomaSoft> The objective is to get the work into mainline Openwrt and patched into the mainline kernel

02:25 <CounterPillow> referencing the odroid m1 dts for clues may be useful, but my general advice is to make sure that the pcie bus you're looking at in a schematic corresponds to he one you're fiddling with in a dts, and also that the combphys are on

02:26 <CounterPillow> if it's one of the PCIe2.1/SATA/USB or PCIe2.1/SATA controllers, then it needs the combphys to be status = "okay"; as well

02:28 <TahomaSoft> It is currently the one of the PCIe 3 controllers that is giving me the problem. The PCIe 2 bus comes up. I'll take a look at the odroid m1, and sounds good on schematic and combphys ideas too.

02:29 <CounterPillow> for the PCIe3 ones, double-check that you've got the PCIe3 PHY enabled in the kernel config. CONFIG_PHY_ROCKCHIP_SNPS_PCIE3

02:31 <TahomaSoft> Affirmative on the SNPS_PCIE3=y

02:33 <TahomaSoft> Thank you

07:12 <mmind00> TahomaSoft: depending on what you read first (here or my mail from just now), make sure the num-lanes are set to <1> ... because pcie3x2 and pcie3x1 share the 2 available lanes

08:26 <qschulz> linkmauve: patches welcome to U-Boot's docs :)

09:28 <TahomaSoft> Thank you @mmind00!

16:26 <CounterPillow> mmind00: enjoy the massive changelog in v2 of the PWM series :D

16:27 <mmind00> CounterPillow: but 200 lines less than v1 :-)

16:31 <mmind00> CounterPillow: really like the bitfield addition :-)

16:32 <mmind00> although that macro-voodoo is cause for headaches

16:38 <CounterPillow> I have a slight suspicion that since this is fixed size, we could make at least the non-const one a static inline __pure function with all the typechecking and normal syntax that entails and compilers will emit mostly the same code, but I don't want to rock the boat too much

17:38 <diederik> Dex2x: nice :)

22:30 <macromorgan> detlevc: Will you be upstreaming the patches from here? https://gitlab.collabora.com/hardware-enablement/rockchip-3588/linux/-/commit/fc9878ba4612e4b4c75dd4096b7a2fdf382145c2

22:30 <macromorgan> I got DSI working on my panel finally for the rk3588s, and those patches help from what I understand