Danct12 changed the topic of #msm8937-mainline to: Boot Linux on your MSM8917/37/40 and QM215 mobile! | GitHub: https://github.com/msm89x7-mainline | Logs: https://oftc.irclog.whitequark.org/msm8937-mainline | Bridged to #msm8937-mainline:kde.org on Matrix
<M0xCAFEBABE[m]> <barni2000[m]> "Gpu works and issue with..." <- hey, do you have rebased our kernel on latest linux version?
<barni2000[m]> Only on rc5
<barni2000[m]> * Only on rc6
<M0xCAFEBABE[m]> barni2000[m]: 6.9.0-rc6?
<barni2000[m]> Yes
<M0xCAFEBABE[m]> on top of msm8953-mainline? (I heard that you're doing this)
<barni2000[m]> It is on top of msm8916-mainline
<M0xCAFEBABE[m]> oh
<M0xCAFEBABE[m]> could you share your repo? I wanna try it, thanks
<M0xCAFEBABE[m]> s/your/the/, s/repo/branch/
<barni2000[m]> I have also rebase lanik's branch you will find that also there
<barni2000[m]> s/rebase/rebased/
<danascape[m]> I also rebased lanik patches
<M0xCAFEBABE[m]> danascape[m]: onto which linux version? did you boot it?
<danascape[m]> M0xCAFEBABE[m]: Sadly not, I had some work after that, so couldnt debug
<danascape[m]> on land and xoop
<M0xCAFEBABE[m]> I see
<danascape[m]> I am working on yocto and uboot these days mainly
<danascape[m]> i was able to telnet on 4.19 kernel on land, so probably Ill push that on postmarketos this weekend
<barni2000[m]> Oh 4.19 so old, so sad downstream still there
<danascape[m]> I am with you in this
<jojo_autoboy[m]> doesn't the cpu still barely work in mainline or did i miss something
<danascape[m]> but we use 4.19 as a daily driver and the device was launched with 3.18
<barni2000[m]> Tomorrow if i will have a little time i am going to try merge the to branch together and fix santoni
<barni2000[m]> and ugglite
<danascape[m]> we have 2 downstreams ported 4.9 and 4.19 and they are very stable, thanks to 0xCAFEBABE
<M0xCAFEBABE[m]> danascape[m]: yeah, in downstream it's the same kernel version as Snapdragon 865, the SoCs older than 865 are still using even older kernel versions XD
<M0xCAFEBABE[m]> those older SoCs now requires this and that hacks/backports to boot up newer android versions, fortunately we still could have one more newer android version decently
<jojo_autoboy[m]> i wonder if 4.19 is new enough to get freedreno working on it
<M0xCAFEBABE[m]> * android version working decently
<danascape[m]> actually
<M0xCAFEBABE[m]> > <@danascape:matrix.org> but we use 4.19 as a daily driver and the device was launched with 3.18
<M0xCAFEBABE[m]> * yeah, in downstream it's the same kernel version as Snapdragon 865, the SoCs older than 865 are still struggling with even older kernel versions XD
<jojo_autoboy[m]> hm yeah seems like the a5xx drivers are there
<danascape[m]> Nice nice
<danascape[m]> I know someone who got that on realme gt2 pro but thats irrelevant here
<jojo_autoboy[m]> newer devices use drm iirc
<danascape[m]> What are the requirements for drm
<danascape[m]> Tho
<jojo_autoboy[m]> * use drm in downstream iirc
<M0xCAFEBABE[m]> danascape[m]: he's talking about MDSS vs msm drm ig
<danascape[m]> Yess
<M0xCAFEBABE[m]> yeah, I've seen a sm6xxx soc still using mdss (camera_v2 as well)
<jojo_autoboy[m]> i guess
<jojo_autoboy[m]> like basically get gpu on pmos with the 4.19 kernel
<danascape[m]> I guess, lemme push land, then Ill work on that
<danascape[m]> debug shell hooks fails to load init for some reason and no pstore info
<M0xCAFEBABE[m]> s/yeah,//, s/camera_v2/camera\_v2/
<danascape[m]> <M0xCAFEBABE[m]> "yeah, I've seen a sm6xxx soc..." <- My billie uses that
<barni2000[m]> <jojo_autoboy[m]> "like basically get gpu on pmos..." <- Why not with 6.x kernel?
<danascape[m]> Are their any docs related to gpu
<jojo_autoboy[m]> <jojo_autoboy[m]> "doesn't the cpu still barely..." <- .
<barni2000[m]> Yes msm8937 msm8940 has some issues with a53-pll and a53-hfpll
<barni2000[m]> But downstream is a wasting effort
<jojo_autoboy[m]> ive never seen a mainline kernel get to 100% power wise
<danascape[m]> Btw, how about we start by pushing gcc and basic framebuffer on mainline?
<barni2000[m]> gcc is there
<danascape[m]> 8837? No?
<danascape[m]> *8937
<barni2000[m]> compatible = "qcom,gcc-msm8937", "syscon";
<danascape[m]> Ohhh
<barni2000[m]> The issue with mdss clocks are dsi clocks are missing from the clock-controller node
<barni2000[m]> But hard to search from phone atm
<barni2000[m]> btw framebuffer also works with simpledrm
<danascape[m]> mmm 👀
<barni2000[m]> Just land has different address
<danascape[m]> ye
<danascape[m]> ill check it after my classes
<barni2000[m]> Btw most of the stuff should be able to enable easily because msm8937 is very close to msm8916 and msm8953 and those are in pretty good shape
<barni2000[m]> In downstream 53 and 37 using the same branch
<danascape[m]> <barni2000[m]> "Btw most of the stuff should..." <- So umm, basically gcc-8917 is common for 8937?
<danascape[m]> s/8917/8916/
<danascape[m]> so ig bringup of danct12 is more justified than us rather having our own clk drivers?
<barni2000[m]> <danascape[m]> "So umm, basically gcc-8917 is..." <- They are similar but msm8937 has different clocks
<barni2000[m]> 8917 and 8937 has much more common
<danascape[m]> Ah yes
<danascape[m]> My bad
<jojo_autoboy[m]> isn't the 8917 just a lower bin of the 37
<barni2000[m]> There are more difference afaik
<barni2000[m]> msm8917 is 4 core and has different gpu
<barni2000[m]> s/is/has/, s/core/cores/
ungeskriptet has joined #msm8937-mainline