03:14 <bamse> konradybcio: low to high would match what we have everywhere else...

06:25 <aka_[m]> <bamse> "konradybcio: low to high would..." <- "its complicated"

06:25 <aka_[m]> some dts indeed have few opp tables where they do high to low

07:28 <krzk> konradybcio: I remember something that Linux drivers preferred one sorting over another

07:29 <krzk> konradybcio: logically, if we sort unit addresses from low to high, then opps should be sorted as well low to high.

07:29 <krzk> konradybcio: and bindings in opp/ follow this approach, I think

07:34 <aka_[m]> yaml have low-high example in it

09:53 <phh> while we're adding quirks in DTS, should I add one for https://github.com/JamiKettunen/linux-mainline-oneplus5/commit/088eaa9153803e2b028e092f88539036442da4a3 ?

10:19 <phh> hum actually, looking at vendor kernel, it looks like they are using QMI_WLFW_FW_MEM_READY_IND_V01 rather than QMI_WLFW_MSA_READY_IND_V01 to move to the bdf loading state, I could maybe do that rather than a quirk, but I don't know how to estimate the breakage I'll induce on other devices...

10:55 <phh> hum yeah no, I don't get that even either, even with req.mem_ready_enable_valid/req.mem_ready_enable = 1, and after resizing wlfw_ind_register_req_msg_v01 to match the vendor one. quirk it is.