ChanServ changed the topic of #linux-msm to:
Daanct12 has joined #linux-msm
Daanct12 has quit [Quit: Leaving]
jhovold has joined #linux-msm
<jhovold> quit
jhovold has quit [Quit: WeeChat 3.6]
<Mis012[m]> rude
<aka_[m]> Mis012[m]: but quite on point, sometimes don't we all just wanna stop.
<Mis012[m]> not really? :P
<aka_[m]> can we change power domain performance state from userspace?
Caterpillar has quit [Remote host closed the connection]
pevik_ has joined #linux-msm
pevik_ has quit [Ping timeout: 480 seconds]
Caterpillar has joined #linux-msm
pevik_ has joined #linux-msm
<aka_[m]> Marijn: have you tried to boot recent rc on 8976?
<aka_[m]> I managed to get gnome-mobile to boot once but now it flips errors about vblank timeouts
<aka_[m]> maybe i should import some hacks from 8953 >_>
<aka_[m]> added backlight and rebooted and it work for now, however gnome-mobile is quite fuckin unstable on MDP backend without GPU
<aka_[m]> keeping cci mux at max isnt probably most power aware thing but i guess nothing better for now
<Marijn[m]> aka_: no it's (almost) Christmas, I haven't had time for that
<Marijn[m]> Maybe you've improperly rebased our mdp5 commits?
<aka_[m]> i havent used any of yours mdp5 commits
<Marijn[m]> Have you tested with modetest or kmscube to rule out gnome nonsense?
<aka_[m]> its pure next
<Marijn[m]> We haven't landed mdss/mdp5 in -next yet so then you're on framebuffer?
<aka_[m]> uh nope
<aka_[m]> added entire mdss node
<aka_[m]> and iommu
<aka_[m]> tomorrow going to try kick in with gpu without hacks and then rebase other commits for wcnss modems and so on
<aka_[m]> same IPA ver so maybe it will work with 8953 stuff
<Marijn[m]> aka_[m]: That's what I mean, mdss/mdp5 in DT
<Marijn[m]> We had some patches in the driver but as you might remember they landed two years ago I think 🤔
<aka_[m]> yea, thats why im unaware of any driver changes
<Marijn[m]> I can later check if I have anything locally, but I don't think so
<Marijn[m]> I reverted the commit that forced minimum 30fps refresh on cmdmode panels, just to make sure the hardware interrupt is properly working
<aka_[m]> no premium hardware here xD
<aka_[m]> 75$ device
<aka_[m]> brand new
<Marijn[m]> Is cmdmode premium? 😂
<aka_[m]> isnt it
<aka_[m]> its quite popular with oled panels
<aka_[m]> which come at premium
<aka_[m]> yea, there are cheap oleds but why bother
<Marijn[m]> All Sony panels actually, including lcd
<Marijn[m]> But the implementation is quite meh, even downstream
<aka_[m]> sony still premium brand.
<aka_[m]> if its not qrd based then its premium
<Marijn[m]> Afaik 10 ii/iii/iv is almost qrd
<JIaxyga[m]> Marijn: So all the same there are some small successes with dsc?
<Marijn[m]> Have you tried my patchset?
<JIaxyga[m]> Marijn[m]: I haven't tried anything since last time. I will be busy for the next week, and then I will be completely free for all sorts of experiments
flto has joined #linux-msm
<JIaxyga[m]> I am too busy due to study/work this month
<Marijn[m]> I do need you on a recent Linux with my patchset and a proper panel driver to be able to do any conclusive comparisons
<Marijn[m]> But then I am also way too busy and have way too many other patches in swing... Perhaps finish those first and see what Neil comes up with for sm8550 DSC
<Marijn[m]> None of pmos folks have said anything on DSC pre-sm8150 either :(
<aka_[m]> I have question.
<aka_[m]> If i want to implement apcs_mux_like thing but with specific divider set i should use clk_regmap_mux_div for that?
<aka_[m]> there is cci apcs mux on 8953 which is supplied via dedicated PLL which also supply C0/C1
<aka_[m]> i need it to always set DIV to 2.5 on MUX
<aka_[m]> or it appears i could mash apcs-msm8916 for that.
<aka_[m]> Sadly not sure if its proper thing for driving cpu
Daanct12 has joined #linux-msm
<aka_[m]> i need to check because i could also reuse it for 8953 even for cores too
<aka_[m]> just supply it with apps-ipq driver
Danct12 has quit [Ping timeout: 480 seconds]
<aka_[m]> any qcom employee can tell anything about CORE_CBCR branch clock which sit 0x8 after apcs mux?
<aka_[m]> by default it appears to be disabled
<aka_[m]> we don't do anything to it with apcs-msm8916 however apcs-ipq8016 marks this one as critical due to "reboots"
<aka_[m]> i meant ipq6018? or so
<aka_[m]> ok i read downstream code and it doesnt make much sense because i get no basics.... (full message at <>)
<aka_[m]> well on 8953 we have branch enabled via lk2nd
<aka_[m]> hmm, this stuff seems to be case for 8939, 8916 while having that clock not enable it in a7 ds code.
<aka_[m]> however it is enabled on 8939
<aka_[m]> shawnguo: i rememember you said you are no longer working on 8939 but do you remember anything about that branch clock?
<aka_[m]> I see some 8939 maintained repo but they set just c1(a53 cluster)
<aka_[m]> vknecht: you had 8939 right
flto has quit [Quit: Leaving]
pevik_ has quit [Ping timeout: 480 seconds]
flto has joined #linux-msm