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
<hacker420[m]> hm
<hacker420[m]> I assume I need gpu firmware?
<hacker420[m]> lanik123:
<barni2000[m]> msm module needs it?
<hacker420[m]> idk
<hacker420[m]> that's why I am asking
<barni2000[m]> i could test on an msm8953 device
<barni2000[m]> but as far as i know it should work without it
<barni2000[m]> only panels depends on msm module which is not loading autmatically
<hacker420[m]> I assume I should automatically get a boot log?
<hacker420[m]> on display
<hacker420[m]> because right now I have a blank display with backlight on
<barni2000[m]> maybe it crashing or the compatible is not correct
<barni2000[m]> do you have usbnet?
<hacker420[m]> barni2000[m]: nope
<barni2000[m]> usb related stuff should be very similar to santoni
<barni2000[m]> maybe it freezing check pstore or try to add debug-shell to initramfs
<hacker420[m]> gonna give it a few minutes first
<hacker420[m]> nope
<hacker420[m]> gonna toss this here for sanity checking
<hacker420[m]> Danct12: btw what is this `data too large` error when pulling ramoops
<barni2000[m]> msmgpio should be tlmm
<hacker420[m]> not in lanik's fork
<hacker420[m]> it tells me that it's not a valid label if I use tlmm
<hacker420[m]> not sure if it's some formatting error or what
<barni2000[m]> i think it uses older msm8937.dtsi
<barni2000[m]> so msmgpio is fine in that case
<Danct12[m]> <hacker420[m]> "Danct12: btw what is this `..." <- no data
<lanik123[m]> <hacker420[m]> "lanik123:..." <- It depends what you will test. For display they are always required
<lanik123[m]> <barni2000[m]> "msmgpio should be tlmm" <- Ye, it's formatting typo. I will do dts refactoring soon
<lanik123[m]> <hacker420[m]> "msm8937-motorola-cedric.dts" <- lol, for what I make msm8937-pm8937.dtsi ?
<hacker420[m]> lanik123[m]: ahh\
<hacker420[m]> will retest in a bit
<dr_shui[m]> <lanik123[m]> "https://github.com/Lanik123/..."; <- did you need lk2nd to boot this?
<lanik123[m]> dr_shui[m]: No, I didn’t remove qcom headers from SOCs dts
<hacker420[m]> well still blank screen with backlight so fart
<hacker420[m]> s/fart/far/
<hacker420[m]> maybe I am missing some kernel config option
<dr_shui[m]> <lanik123[m]> "No, I didn’t remove qcom headers..." <- what should i do to pin usb at host mode? I haven't seen &usb node on `msm8917-xiaomi-rolex.dts` file
<barni2000[m]> usb is not working on rolex
<barni2000[m]> rolex have only 1 pmic
<dr_shui[m]> oh
<dr_shui[m]> sad
<barni2000[m]> charger driver should be modified i think
<dr_shui[m]> charger driver not important to me, but usb
<dr_shui[m]> is very important to connecting other devides
<dr_shui[m]> s/devides/devices/
<barni2000[m]> dr_shui[m]: yes but you need to define usb_vbos
<M0xCAFEBABE[m]> barni2000[m]: yeah, rolex's charger driver provides a extcon interface directly or indirectly in downstream, required for USB to work
<barni2000[m]> s/usb_vbos/usb\_vbus/
<M0xCAFEBABE[m]> * in downstream k4.x, required
<M0xCAFEBABE[m]> M0xCAFEBABE[m]: not just rolex, but also any other devices which doesn't ship PMI chip (which doesn't have QPNP smbcharger)
<barni2000[m]> yes ugg, ugglite and riva has same issue
<barni2000[m]> thats why i stuck with the porting
<barni2000[m]> i have tried to define linux-extcon-gpio but it is not enough
<M0xCAFEBABE[m]> what GPIO do you use for it...
<M0xCAFEBABE[m]> the USB detection mechanism isn't GPIO i think
<M0xCAFEBABE[m]> but this
<M0xCAFEBABE[m]> the charger driver reads register from the charger chip, and sets usb status depending on it
<barni2000[m]> there is and usb_id pinctrl in downstream
<barni2000[m]> s/and/an/, s/usb_id/usb\_id/
<barni2000[m]> mainline bq2560x is broken
<barni2000[m]> or not implement everything
<M0xCAFEBABE[m]> barni2000[m]: at least I've never messed with it when I was bringing up downstream k4.x... the LineageOS/android_kernel_xiaomi_msm8937 is made by me tho
<barni2000[m]> M0xCAFEBABE[m]: i know
<M0xCAFEBABE[m]> <barni2000[m]> "there is and usb_id pinctrl in..." <- I've only found it on ugg/ugglite device specific dts, and it's unused at all
f_ is now known as f_estive
<barni2000[m]> yes i don't find for riva at the past i have experimenting with ugglite the most
<M0xCAFEBABE[m]> barni2000[m]: > <@barni2000:matrix.org> ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/JlrSOyxaZcSMIxrVLLFbyVCG>)
<barni2000[m]> this is for ugglite
<M0xCAFEBABE[m]> it really isn't based on GPIO lollllllll...
<M0xCAFEBABE[m]> barni2000[m]: > <@barni2000:matrix.org> ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/WYXzcPRuQeecXeOnWLKUZcfL>)
<M0xCAFEBABE[m]> didn't even find any reference to it
<barni2000[m]> but why they modified the pinctrl than, nvm original xiaomi sources is a mess
<M0xCAFEBABE[m]> <M0xCAFEBABE[m]> "https://github.com/LineageOS/..."; <- and this is the real logic
<M0xCAFEBABE[m]> barni2000[m]: these ODMs are always been retarded
<barni2000[m]> motorolla sources are relativly clean
<barni2000[m]> * motorola sources are relatively clean
<barni2000[m]> M0xCAFEBABE[m]: can it be used in mainline charging driver?
<M0xCAFEBABE[m]> barni2000[m]: idk, only thing I'm sure is it works in this way
<M0xCAFEBABE[m]> if it's missing, implement it
<barni2000[m]> i will check
<barni2000[m]> after i have finished testing kernel updates for msm8953
<M0xCAFEBABE[m]> barni2000[m]: btw, which msm8953 device do you have?
<barni2000[m]> 3 mido, 2 tissot, 2 ysl, 1 vince, 1 sakura(with daisy panel), motorola potter, asus zs550kl, huawei nova (can-l11)
<M0xCAFEBABE[m]> barni2000[m]: wew, so many msm8953 devices... btw, I have a vince too
<barni2000[m]> last two are a garbage because bootloader cannot be unlocked :(
<barni2000[m]> s/a//
<barni2000[m]> i have some mainboard also but all has some defects
<barni2000[m]> btw xiaomis are community supported
<jojo_autoboy[m]> there's a chance the huawei could be unlocked if they left secure boot off on it
<barni2000[m]> jojo_autoboy[m]: i cannot go in fastboot because it reboots instantly
<jojo_autoboy[m]> there's always edl
<barni2000[m]> if you have suitable programer
<barni2000[m]> for xiaomis easy to get but for huawei idk
<jojo_autoboy[m]> if it's got secure boot off you could use any programmer with it
<jojo_autoboy[m]> iirc
<barni2000[m]> first i need to tear down my device
<barni2000[m]> i wanted to check my asus also but i don't find information about the testpoints
<dr_shui[m]> Danct12: have any progress on santoni’s wlan?
<barni2000[m]> <M0xCAFEBABE[m]> "and this is the real logic" <- what should be the extcon?
<M0xCAFEBABE[m]> barni2000[m]: I was talking about how the hardware works, not driver framework stuff
<M0xCAFEBABE[m]> <barni2000[m]> "/* OTG reporting */..." <- > <@barni2000:matrix.org> ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/mnRIhztgUsRwGIyOkgLgjadU>)
<barni2000[m]> ok, i am going to experimenting with ugglite
<hacker420[m]> <dr_shui[m]> "Danct12: have any progress on..." <- well it works for me on 8937 cedric
<hacker420[m]> so I assume santoni should work too?
<Sid_key[m]> <hacker420[m]> "so I assume santoni should..." <- Should
<Sid_key[m]> <jojo_autoboy[m]> "if it's got secure boot off..." <- Most msm89x7 has sb on
<hacker420[m]> <lanik123[m]> "It depends what you will test..." <- ah
<hacker420[m]> hm
<hacker420[m]> so will it display anything without it or not
<hacker420[m]> or will it just cut out when it gets to DE
<hacker420[m]> lanik123:
<barni2000[m]> <Sid_key[m]> "Most msm89x7 has sb on" <- above msm8916 all soc has sb on afaik
<hacker420[m]> barni2000[m]: unless you have a device with it off
<barni2000[m]> hacker420[m]: its very rare
<hacker420[m]> yes
<hacker420[m]> but not impossible to get
<lanik123[m]> <hacker420[m]> "or will it just cut out when..." <- lol what, gpu fw is always used while drm is running and without them display cannot work(if it not simpledrm)
<hacker420[m]> ah
<hacker420[m]> btw what should I look for?
<hacker420[m]> as in filenames
<lanik123[m]> hacker420[m]: You can load adreno fw via defconchig
<lanik123[m]> s/defconchig/defconfig/
<hacker420[m]> yes but where do I get said fw from
<hacker420[m]> directly from device somewhere?
<hacker420[m]> or from the vendor repos
<barni2000[m]> /lib/firmware/qcom/a530_pm4.fw
<barni2000[m]> /lib/firmware/postmarketos/a505_zap.b02
<barni2000[m]> /lib/firmware/postmarketos/a505_zap.mdt
<lanik123[m]> s/plain/tree/
<barni2000[m]> install linux-firmware-qcom
<barni2000[m]> that will give you a530 files
<barni2000[m]> and you need to copy your zap files from the device because they are signed
<hacker420[m]> barni2000[m]: > <@barni2000:matrix.org> ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/cfcqlmtlpXkPjttrgPlHUFrd>)
<hacker420[m]> but the a505 zap files aren't there
<lanik123[m]> lanik123[m]: I also have them in my kernel https://github.com/Lanik123/linux/tree/msm89x7-6.6-playground/firmware/qcom
<hacker420[m]> unless they are somewhere else
<hacker420[m]> so well
<hacker420[m]> where do I get the zap files
<lanik123[m]> hacker420[m]: lol, zap it's not implemented yet for a505
<hacker420[m]> that looks fine?
<hacker420[m]> lanik123[m]: so just get these files?
<lanik123[m]> lanik123[m]: Try without them
<hacker420[m]> without what
<hacker420[m]> the zap files or fw
<lanik123[m]> without zap fw
<lanik123[m]> hacker420[m]: +
<hacker420[m]> anything else I need to install into the rootfs?
<hacker420[m]> or can I just run the same rootfs
<barni2000[m]> a505 has not got zap?
<lanik123[m]> hacker420[m]: Gpu fw is enough
<hacker420[m]> so no rootfs? modification needed
<lanik123[m]> barni2000[m]: It's got but zap in mainline is realy broken
<hacker420[m]> * so no rootfs modification needed?
<barni2000[m]> lanik123[m]: msm8953 is using zap
<lanik123[m]> barni2000[m]: ))))))))))))))))
<lanik123[m]> lanik123[m]: Give me a sec
<lanik123[m]> Where u see secure context for zap?))))
<barni2000[m]> gpu not works with out zap files
<barni2000[m]> i have tested befor
<barni2000[m]> s/befor/before/
<lanik123[m]> barni2000[m]: Becouse u tested on phone with new tz
<hacker420[m]> so doing no rootfs modification
<hacker420[m]> let's hope it got built in
<barni2000[m]> lanik123[m]: thats why one of my phone worked without it and other is not
<lanik123[m]> lanik123[m]: On older phones zap is often not supported at all.
<barni2000[m]> i have asked the guys in msm8953-mainline about it when i cleaned up the device packages and they don't know the awnser so it is related to the tz
<hacker420[m]> blank screen without reinstalling the rootfsd
<hacker420[m]> let me try reinstalling it with the new kernel package
<lanik123[m]> barni2000[m]: Also it can related to bl. You can ask ArtemSid(or CAFEBABE), he will tell you for sure
<hacker420[m]> hm do I still need to install firmware-qcom into the rootfs then?
<barni2000[m]> you can chroot in it
<barni2000[m]> pmbootstrap chroot -r
<lanik123[m]> hacker420[m]: Without logs I can't say what the reason is
<hacker420[m]> let me flash rootfs just to see
<hacker420[m]> without installing firmware-qcom
<hacker420[m]> as it should be built into the kernel
<hacker420[m]> if everything went to plan
<hacker420[m]> let me wait a bit
<hacker420[m]> so far blank screen
<barni2000[m]> is your panel driver in the initramfs?
<hacker420[m]> if I am booting the boot.img and selected the panel in kconfig then it should be
<barni2000[m]> mkinitfs makes the boot.img so if you not defined the modules it will not load automatically
<barni2000[m]> or everything compiled builtin?
<hacker420[m]> barni2000[m]: yes
<barni2000[m]> what about the msm module?
<barni2000[m]> sry, i am slow you were quick
<barni2000[m]> what about mdp5?
<dr_shui[m]> <hacker420[m]> "so I assume santoni should..." <- Where can I get 8940’s firmware ?
<barni2000[m]> modem partition
<barni2000[m]> or from NON-HLOS.bin
<dr_shui[m]> Use msm-firmware-loader?
<barni2000[m]> that will also work
<barni2000[m]> unfortunately i cannot enable the usb if i comment out the extcon i have the interface for usbnet but it is not working
<hacker420[m]> happy new year
<hacker420[m]> 2024 will be the year of 8937 mainline
<jojo_autoboy[m]> let's beat the 8916 in support