ChanServ 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]> how's the status on latest branch?
<M0xCAFEBABE[m]> xiaomi devices specifically
<M0xCAFEBABE[m]> does display and USB work fine?
<barni2000[m]> you can use my dev branch cpufreq, gpu, usb is working but you will need lk2nd
<barni2000[m]> ig you want to test with android but here is the wip pmos packages https://gitlab.postmarketos.org/postmarketOS/pmaports/-/merge_requests/5693
<M0xCAFEBABE[m]> <barni2000[m]> "you can use my dev branch..." <- what's done in lk2nd...
<M0xCAFEBABE[m]> > <@barni2000:matrix.org> you can use my dev branch cpufreq, gpu, usb is working but you will need lk2nd
<M0xCAFEBABE[m]> * what exactly is done in lk2nd...
<barni2000[m]> cpufreq init
<barni2000[m]> 8937 and 8940 uses generic a53pll
<barni2000[m]> 8917 does not needs it
<barni2000[m]> you can use maxcpus=4
<M0xCAFEBABE[m]> what about USB?
<barni2000[m]> it is working
<M0xCAFEBABE[m]> I mean... what USB stuff is done in lk2nd, that's required for mainline
<barni2000[m]> M0xCAFEBABE[m]: nothing
<barni2000[m]> it is needed for cpufreq and panel selection
<M0xCAFEBABE[m]> > usb is working but you will need lk2nd
<M0xCAFEBABE[m]> ok I misunderstood it then, great, thanks
<M0xCAFEBABE[m]> * > usb is working but you will need lk2nd
<M0xCAFEBABE[m]> ok I misunderstood it then, great, thanks
<barni2000[m]> sorry maybe i was not exact enough
<M0xCAFEBABE[m]> btw are those really necessary? fails to apply as-is on 6.12.0-rc3
<M0xCAFEBABE[m]> 8e2777a4a47cf Revert "cpuidle: psci: Move enabling OSI mode after power domains creation"
<M0xCAFEBABE[m]> 55331ab0f7ab3 Revert "drm/msm/dsi: split bridge pre_enable/post_disable"
<barni2000[m]> this 55331ab0f7ab3 Revert "drm/msm/dsi: split bridge pre_enable/post_disable" yes
<barni2000[m]> other idk
<M0xCAFEBABE[m]> uhm
<barni2000[m]> wait the issue is this repo based on msm8916-mainline/linux
<barni2000[m]> so if they do some hacks it should be reverted
<barni2000[m]> idk which commit you have been started
<M0xCAFEBABE[m]> started picking from 8e2777a4a47cf Revert "cpuidle: psci: Move enabling OSI mode after power domains creation"
<M0xCAFEBABE[m]> well they did this
<M0xCAFEBABE[m]> 3198657051f64 drm/msm/dsi: split bridge pre_enable/post_disable
<M0xCAFEBABE[m]> 0b988e13c6c29 HACK: drm/msm: Run mdp5 encoder enable()/disable() after bridge hooks
<barni2000[m]> none of them is neede
<barni2000[m]> s/neede/needed/
<barni2000[m]> only improve power management commit
<barni2000[m]> i am working on msm8917 upstreaminig now days so there will be a lot of changes there btw
<barni2000[m]> yes, it has been usptreamed already 6.13 will contains it
<barni2000[m]> s/it/they/, s/has/have/
<M0xCAFEBABE[m]> nice
<M0xCAFEBABE[m]> are our GPUs supported in upstream Mesa yet?
<barni2000[m]> git version yes
<M0xCAFEBABE[m]> both 308 and 505?
<barni2000[m]> yes
<M0xCAFEBABE[m]> nice
<M0xCAFEBABE[m]> now I should use lk2nd from msm8916-mainline? can I keep using msm8953-mainline one for now? (I did a lot of changes in my fork which would take some time to port them...)
<barni2000[m]> unfortunately msm8953 does not contains necessary changes everything is implemented in msm8916
<M0xCAFEBABE[m]> * am I supposed to use lk2nd from msm8916-mainline now? can I keep using msm8953-mainline one for now? (I did a lot of changes in my fork which would take some time to port them...)
<M0xCAFEBABE[m]> oh okay
<barni2000[m]> btw msm8953-mainline/lk2nd should be archived because everything was migrated to msm8916-mainline/lk2nd
<M0xCAFEBABE[m]> <barni2000[m]> "yes" <- does it need the same hacks as msm8953?
<M0xCAFEBABE[m]> s/hacks/workarounds/
<barni2000[m]> generally a506 does not needs hacks, but same. imo you should start without hacks
<barni2000[m]> a306a (308) will crashing more
<M0xCAFEBABE[m]> <barni2000[m]> "it is working" <- usb gadget mode works fine too?
<barni2000[m]> you mean like usbnet or otg?
<M0xCAFEBABE[m]> yes
<M0xCAFEBABE[m]> actually I'd use like ADB or MTP, etc
<M0xCAFEBABE[m]> just depends on if usb gadget is not broken in drivers...
<barni2000[m]> adb, mtp should work
<barni2000[m]> otg is not working
<M0xCAFEBABE[m]> oh
<Andr[m]> <barni2000[m]> "ig you want to test with android..." <- I tested these packages with an additional dts file for motorola,cedric (the one that was provided here some time ago) and lk2nd 19.0. But it didn't work. There was a crash after about 1 second and a boot loop. I tried lk2nd's ramoops, but no log was written to the memory.
<Andr[m]> Has anyone else tried to get cedric running with lk2nd and the provided kernel?
<M0xCAFEBABE[m]> <barni2000[m]> "otg is not working" <- do u have built lk2nd.img? i don't want to clone and compile rn
<M0xCAFEBABE[m]> oh, thanks
<barni2000[m]> <Andr[m]> "I tested these packages with..." <- > <@aandre:tu-chemnitz.de> I tested these packages with an additional dts file for motorola,cedric (the one that was provided here some time ago) and lk2nd 19.0. But it... (full message at <https://matrix.org/oftc/media/v1/media/download/AVv7zElbIQpkDmNzOduZCwcXF53_NKYSr3uK9RfZarWc_mQJiCbgTBPNt9G-F-ELU_dzzKemkUbDo_i6F0Kz_atCeTIPDEFQAG1hdHJpeC5vcmcvWUhuVkRodUNHdVF6dkdjTVZBWmVoZ09i>)
<barni2000[m]> every other msm8937 device is working so issue will be in your dt
<M0xCAFEBABE[m]> <barni2000[m]> "https://github.com/msm8916-..."; <- does this lk2nd still support appended dtb?
<hacker420[m]> barni2000[m]: yeah the dts works fine here
<M0xCAFEBABE[m]> > <@barni2000:matrix.org> https://github.com/msm8916-mainline/lk2nd/releases
<M0xCAFEBABE[m]> * does this lk2nd still support booting Image.gz-dtb?
<Andr[m]> barni2000[m]: just to be sure. You tested with qcom and msm89x7 selected during `pmbootstrap init`, right?
<Andr[m]> hacker420[m]: could you push it to the msm89x7-linux repo?
<barni2000[m]> Andr[m]: is it contains cedric panel drivers?
<hacker420[m]> Andr[m]: the reason I didn't is that my internet sucks ass
<barni2000[m]> generic package does not contain panel drivers
<hacker420[m]> so downloading the full git history is painful'
<barni2000[m]> use --depth=1 first
<barni2000[m]> and after you can make a pull
<hacker420[m]> well yes I did
<hacker420[m]> my repo for now is a mess
<hacker420[m]> with the random charger patches for battery and charging
<Andr[m]> barni2000[m]: USB networking would be sufficient
<hacker420[m]> barni2000: are the smbcharger commits in yet?
<barni2000[m]> no
<barni2000[m]> i don't want to add more commits atm
<barni2000[m]> i am busy with upstreaming msm8917
<hacker420[m]> ah alright
<hacker420[m]> well here's the dts
<hacker420[m]> you'll need to remove the charger and battery stuff André
<Andr[m]> thanks, I'll compare it to my version, when I have some time
<barni2000[m]> check your rootfs /boot
<Andr[m]> Are qcom,msm-id and board-id needed, when using lk2nd?
<M0xCAFEBABE[m]> weird, the same boot image with both msm89x7 and msm8953 dts included, boots fine on msm8953, on msm89x7 lk2nd says dtb not found, boots fine using stock fastboot
<barni2000[m]> <Andr[m]> "Are qcom,msm-id and board-id..." <- maybe
<barni2000[m]> xiaomi-sakura can boot with daisy's board-id so it is does not matter
<barni2000[m]> dtb was set in lk2nd dt?
<barni2000[m]> * was dtb set in lk2nd dt?
<M0xCAFEBABE[m]> usb doesn't work on santoni..?
<M0xCAFEBABE[m]> [ 25.110909][ T143] platform 78db000.usb: deferred probe pending: (reason unknown)
<barni2000[m]> M0xCAFEBABE[m]: > <@me_cafebabe:matrix.org> usb doesn't work on santoni..?
<barni2000[m]> it works for me
<barni2000[m]> > [ 25.110909][ T143] platform 78db000.usb: deferred probe pending: (reason unknown)
<M0xCAFEBABE[m]> M0xCAFEBABE[m]: > <@me_cafebabe:matrix.org> usb doesn't work on santoni..?
<M0xCAFEBABE[m]> > [ 25.110909][ T143] platform 78db000.usb: deferred probe pending: (reason unknown)
<M0xCAFEBABE[m]> meanwhile it works on msm8953
<M0xCAFEBABE[m]> same smbchg
<barni2000[m]> it depends from extcon
<barni2000[m]> msm8953 using dwc3 so that is totally different
<barni2000[m]> you will need msm8941-misc
<M0xCAFEBABE[m]> oh yeah nvm I didn't enable that thing
<M0xCAFEBABE[m]> I thought I already did it because pmi8950.dtsi is included by msm8953-xiaomi-common.dtsi too
<barni2000[m]> two repo is different a bit
<M0xCAFEBABE[m]> pmi8950_smbcharger isn't even enabled... on land and santoni
<barni2000[m]> yes because in this repo charging driver is missing
<barni2000[m]> * this repo fg and charging driver
<M0xCAFEBABE[m]> ah ok
<M0xCAFEBABE[m]> so does USB work on msm8917 devices (custom charger&fg) now?
<barni2000[m]> yes usb is working everywhere
<M0xCAFEBABE[m]> * devices (which has custom charger&fg, * custom charger&fg chips) now?
<barni2000[m]> usb is critical for me
<M0xCAFEBABE[m]> setting 78db000.usb as usb controller work
<M0xCAFEBABE[m]> or it has different address/name on mainline?
<M0xCAFEBABE[m]> * usb controller doesn't work
<M0xCAFEBABE[m]> meanwhile 7000000.usb worked for msm8953
<M0xCAFEBABE[m]> I don't have UART RX soldered on the device and I can't shell into it and find out the correct usb controller...
<M0xCAFEBABE[m]> msm8953 has phy@79000 but 7000000.usb (usb@7000000) is the correct one, uhh...
<M0xCAFEBABE[m]> could u do ls /sys/class/udc for me on your working device? thanks
<M0xCAFEBABE[m]> * could u show ls /sys/class/udc on your working device? thanks
<barni2000[m]> a moment
<barni2000[m]> `
<M0xCAFEBABE[m]> oh, I didn't expect it's in such name lol, thanks