03:08 <bamse> mal: let's hope linus is in a good mood today, Cc'ed you on the pull request

11:20 <konradybcio> lumag: I did something and 8996 started booting again (without interconnect).. now i'm back to your dpu1 review comments, adding ad4 makes the device explode, resulting in rcu detecting stalls

11:21 <lumag> konradybcio, are you trying to get ad4 to work? Or just adding the IRQs?

11:21 <konradybcio> just adding the irqs

11:21 <lumag> The RCU stalls are usually a sympthom of something else

11:21 <lumag> Just ignore them for now.

11:21 <lumag> (ad4 irqs, not stalls)

11:21 <konradybcio> stalls are rather hard to ignore :P

11:26 <konradybcio> FYI i booted without any additional cmdline args

11:33 <konradybcio> well, speaking of "something else", I just rebooted and it doesn't work now

11:42 <konradybcio> pstore shows nothing, the log just randomly cut off.. may be due to some cpu instabilities as you suggested

11:42 <konradybcio> s/nothing/nothing interesting

13:27 <rawoul> I also saw the same RCU stall issues when I last tried booting my db820c

17:11 <mal> bamse: looks like that pull request went in already

18:17 <bamse> mal: yeah, looks like Linus picked it up right away, thanks for chasing me on getting that resolved

18:43 <steev> ayyy the good stuff has gone to the list

18:43 <steev> (for x13s)

19:09 <ajhalaney[m]> steev i know i need to find some time to test drive on my x13s (and maybe see what happens when i tickle pcie on a qdrive3/sa8540p)

19:14 <steev> ajhalaney[m]: if you're already running linux, those are already the patches in my tree, just that they now hit the list :)

19:15 <ajhalaney[m]> doh steev i didn't even check if there were any differences or not on the dts bits

19:15 <ajhalaney[m]> constant battle between trying to muck with x13s as a test machine or as my daily work driver lol

19:16 <steev> relatable

19:16 <steev> it would be nice to have a second one for the kali stuff, but meh

21:41 <lumag> kholk, Marijn[m] (and anybody else): do you remember by chance a device with non-default qcom,mdss-highest-bank-bit (IOW, memory type or configuration)?

21:59 <aka_[m]> And which is default?

22:02 <aka_[m]> If you mean .highest_bank_bit out of hw catalog then Bengal probably ship with 0x1

22:08 <konradybcio> it reflects the memory used

22:08 <konradybcio> lumag is wondering whether there are actually devices in the wild that use lpddr4x on 8350 instead of lpddr5 etc

22:14 <lumag> aka_[m], yes. There is a commonly found default (that e.g. we followed in the dpu_hw_catalog). Have you encountered a device, where the value used was different from the per-SoC default?

22:20 <aka_[m]> Ah so that's what you mean

22:20 <aka_[m]> Maybe we need to find device which ship not default conf of ram

22:21 <aka_[m]> Like soc being paired with lpddr4 and device with ddr3

22:21 <lumag> aka_[m], yep, that's the question. Konrad raised a question where we might need to have this as a DT property. I'm trying to understand whether it is really needed or not

22:21 <aka_[m]> For majority probably nope

22:21 <aka_[m]> Because qrds

22:22 <aka_[m]> konradybcio: so kexperias shipping non default memory confs?

22:33 <konradybcio> huawei 8350 (8325) phones may be using lpddr4 actually

22:33 <konradybcio> there are some google hits (from dubious sites though)

23:23 <aka_[m]> <konradybcio> "huawei 8350 (8325) phones may be..." <- Huauejs aren't unlockable so why bother

23:34 <konradybcio> I think they are, unless people posting WoA running on some were using some magic devices

23:35 <aka_[m]> I remember some ppl posting some lgs and it turned out they had soc swap on them