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
<hacker420[m]> <barni2000[m]> "> <@hacker420:kde.org> ```..." <- yeah
f_ has quit [Remote host closed the connection]
f_ has joined #msm8937-mainline
f_ has quit [Remote host closed the connection]
f_ has joined #msm8937-mainline
<hacker420[m]> <barni2000[m]> "panel names for cedric should..." <- > <@barni2000:matrix.org> panel names for cedric should be... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/YnVfGWSjWYbdYtaVFXXzKfKI>)
<hacker420[m]> think it should react right away
<barni2000[m]> for me it is disabling backlight and enables it back
<barni2000[m]> * it back on boot
<hacker420[m]> maybe I need to toss this out/
<hacker420[m]> I have this for display
<hacker420[m]> maybe the supply or gpio is an issue?
<barni2000[m]> nah it is fine
<barni2000[m]> disable wled node
<hacker420[m]> with status disabled or comment it out
<barni2000[m]> both fine
<hacker420[m]> well disabled still does nothing
<barni2000[m]> it means the same
<barni2000[m]> maybe it is freezing like land
<barni2000[m]> 3.299331] [pmOS-rd]: Illegal instruction
<barni2000[m]> [ 3.300135] Internal error: Oops - Undefined instruction: 0000000002000000 [#1] PREEMPT SMP
<barni2000[m]> [ 3.300148] Unable to handle kernel paging request at virtual address 60ffff00002298b8
<barni2000[m]> finally i have usable ramoops from land
<barni2000[m]> i guess something can be with reserved memory ranges
<barni2000[m]> can you use debug-shell without panel?
<hacker420[m]> I'd need functional USB first
<hacker420[m]> btw did you send the montana changes to your fork?
<barni2000[m]> no
<barni2000[m]> * not yet
<barni2000[m]> <hacker420[m]> "I'd need functional USB first" <- No it has touch keyboard now
<hacker420[m]> and how exactly will I access it without panel
<barni2000[m]> Simpledrm
<hacker420[m]> right
<hacker420[m]> so disable panel and try simpledrm?
<hacker420[m]> or have both modules enabled
<hacker420[m]> will it auto select panel or simpledrm?
<barni2000[m]> Dissble panel first, you should check for freeze
<barni2000[m]> It just validation it is booting fine
<barni2000[m]> Dont forget to add rmi_i2c and rmi_core to initfs
<hacker420[m]> <barni2000[m]> "Dissble panel first, you..." <- so disable it in the dts, disable the module or both?
<barni2000[m]> Dts
<hacker420[m]> don't I also need to put simpledrm in the init modules file
<barni2000[m]> No
<barni2000[m]> It is not compiled as a module
<hacker420[m]> what
<hacker420[m]> to have simpledrm as built in
<hacker420[m]> I'd need to set drm to built in
<hacker420[m]> if I set drm to built in then I can switch between module and built in
<hacker420[m]> yeah I think it's stuck somewhere
<barni2000[m]> Do you see some logs on your screen?
<barni2000[m]> What config do you use?
<hacker420[m]> <barni2000[m]> "What config do you use?" <- uhh I'll send you
<barni2000[m]> ❯ cat .output/.config | grep DRM_SIMPLE
<barni2000[m]> # CONFIG_DRM_SIMPLE_BRIDGE is not set
<barni2000[m]> CONFIG_DRM_SIMPLEDRM=y
<hacker420[m]> let me try yours
<barni2000[m]> many stuff is missing from your config
<barni2000[m]> usb
<barni2000[m]> * usb for exampler
<barni2000[m]> * usb for example
<barni2000[m]> a53pll
<barni2000[m]> # CONFIG_DRM_MSM_DSI_28NM_PHY is not set
<hacker420[m]> let's see
<barni2000[m]> * # CONFIG\_DRM\_MSM\_DSI\_28NM\_PHY is not set
<hacker420[m]> maybe my config was fucked after all
<barni2000[m]> * # CONFIG_DRM_MSM_DSI_28NM_PHY is not set
<barni2000[m]> without 28nm dsi complied panel will never work
<hacker420[m]> kek
<hacker420[m]> let's test simplefb first
<barni2000[m]> ok
<hacker420[m]> ok
<hacker420[m]> so now it reboots
<hacker420[m]> hmmm
<hacker420[m]> after a bit
<barni2000[m]> is the frequencies still disabled?
<hacker420[m]> wait they aren't
<hacker420[m]> guess when I pulled the changes got wiped
<barni2000[m]> i will need to create separated opp for msm8940
<hacker420[m]> hm
<hacker420[m]> do I need to put something in cmdline to output log to screen?
<hacker420[m]> right now got a brief blue flash and now the screen is blank with backlight on
<hacker420[m]> under simpledrm
<hacker420[m]> might just go to panel right away
<barni2000[m]> telnet?
<hacker420[m]> eh right forgot the hook
<hacker420[m]> fuck it, go big or go home
<hacker420[m]> let me test the panel
<hacker420[m]> but shouldn't I need something on cmdline to make it output to screen
<hacker420[m]> like console=tty0 or something?
<hacker420[m]> aj
<barni2000[m]> no
<hacker420[m]> > in addition if you want the kernel logs to print to the screen make sure that console=tty0 is set in the kernel command line.
<barni2000[m]> i see logs without it
<hacker420[m]> on what phone
<barni2000[m]> enable debug-shell and pmos.debug-shell
<hacker420[m]> because cedric is fucked in that aspect
<hacker420[m]> it forces console=null
<barni2000[m]> hacker420[m]: montana, land, santoni, riva
<hacker420[m]> hm
<hacker420[m]> montana works without it
<hacker420[m]> weird
<barni2000[m]> let me check the params
<hacker420[m]> I had to do that on op6 too to get logs on screen
<barni2000[m]> PMOS_NOSPLASH
<hacker420[m]> let me add it and see what happens
<barni2000[m]> and remove PMOS_NOOUTPUT_REDIRECT
<hacker420[m]> hacker420[m]: ok so just backlight again
<hacker420[m]> but a usb device popped up
<barni2000[m]> telnet
<hacker420[m]> will need to see with debug-shell enabled
<barni2000[m]> telnet 172.16.42.1
<hacker420[m]> ok weird
<hacker420[m]> now I get no device
<hacker420[m]> nope
<hacker420[m]> nothing
<hacker420[m]> let's remove it and try agaim
<hacker420[m]> let's remove it and try again
<hacker420[m]> ok what
<hacker420[m]> now I get no device
<hacker420[m]> is this usb port so flaky or what
<hacker420[m]> nothing on screen either
f_ is now known as g_
g_ is now known as f_
<barni2000[m]> try ti enable simpledrm first you can find the corret parameters in the lk2nd log
<barni2000[m]> land needs couple of reboot to work
<barni2000[m]> but panel also works
<barni2000[m]> [ 76.588247] panel-xiaomi-land-nt35521s-ebbg 1a94000.dsi.0: Failed to set display off: -110
<barni2000[m]> it is not a serious error just for proof it is working
<barni2000[m]> i have uploaded more panel configs
<barni2000[m]> montana, land
<barni2000[m]> ugglite and ugg is left
<barni2000[m]> next step is sending lk2nd supports
<hacker420[m]> <barni2000[m]> "try ti enable simpledrm first..." <- ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/hXBFkILFJihRAfuZUsGwPFUM>)
<hacker420[m]> [160] MDP continuous splash detected: pipe RGB_0, base: 0x90001000, stride: 3240, src: 1080x1920 (0,0), img: 1080x1920, out: 1080x1920 (0,0), format: 0x2243f (bpp: 3)
<hacker420[m]> [170] Display refresh: cmd mode: 0, auto refresh: 0 (sel: 0x1f20)
<hacker420[m]> I get nothing on screen
<hacker420[m]> is the panel initializing before it or what
<barni2000[m]> other thing don't enable two together because it will cause error
<hacker420[m]> wdym
<hacker420[m]> both panel and simpledrm?
<hacker420[m]> in kernel config?
<hacker420[m]> or dts
<barni2000[m]> no in dt
<hacker420[m]> panel is commented out
<barni2000[m]> dt will tell what should be loaded you don't need to disable it in the kernel
<hacker420[m]> I still just get that brief blue flash and just the backlight
<barni2000[m]> what about telnet?
<hacker420[m]> I get zero usb devices
<barni2000[m]> did you installed this postmarketos-mkinitfs-hook-debug-shell-0.7.0-r0
<barni2000[m]> hacker420[m]: i think it is crashing
<hacker420[m]> barni2000[m]: even with that I got no devices
<hacker420[m]> barni2000[m]: maybe but then why did I get a usb device that one time
<hacker420[m]> but with no IP
<barni2000[m]> where you disabled the freqs, you should only disable them in cpu opp
<barni2000[m]> and leave them in pll opp
<hacker420[m]> yes
<hacker420[m]> that is what I did
<hacker420[m]> didn't touch pll
<barni2000[m]> ok it seems fine
<hacker420[m]> the problem is I get no output
<hacker420[m]> just backlight
<barni2000[m]> no usb?
<hacker420[m]> nope
<hacker420[m]> I have no idea if it's due to some change or flakiness
<barni2000[m]> try to disable emmc and sdcard
<barni2000[m]> kept only the simpledrm and the touch
<barni2000[m]> * the simpledrm,, * usb and the
<hacker420[m]> get debug-shell too?
<barni2000[m]> i have enabled debug-shell
<hacker420[m]> nope nothing
<barni2000[m]> idk you should trying more or give up my devices has no issue
<barni2000[m]> btw did you put panel drivers in your initfs?
<hacker420[m]> ok
<hacker420[m]> I have a device now
<hacker420[m]> barni2000[m]: yes
<hacker420[m]> but can't telnet in
<hacker420[m]> can't bring it up either
<hacker420[m]> despite nm showing it as connected
<hacker420[m]> can't ping it either
<hacker420[m]> but like why am I not getting output
<barni2000[m]> did you added pmos.debug-shell param?
<hacker420[m]> no
<hacker420[m]> what is that param even
<barni2000[m]> it is enabling the debug shell
<hacker420[m]> huh?
<hacker420[m]> shouldn't it auto enable if you use debug-shell?
<barni2000[m]> not anymore
<hacker420[m]> ah
<hacker420[m]> ok I have debug shell
<hacker420[m]> https://bpa.st/CAYQ
<hacker420[m]> what now
<hacker420[m]> it's like simpledrm doesn't even show up
<hacker420[m]> I can try reenabling storage
<hacker420[m]> ok another approach
<hacker420[m]> reenabled panel and storage
<hacker420[m]> disabled simple framebuffer in dts
<barni2000[m]> you have crashing in your logs
<hacker420[m]> because storage is disabled
<barni2000[m]> at debug shell it does not matter
<hacker420[m]> hm
<barni2000[m]> you have same crashing like i have at land
<hacker420[m]> and how did you fix it?
<hacker420[m]> if you did
<barni2000[m]> couple of reebots will fix it
<barni2000[m]> s/reebots/reboots/
<barni2000[m]> idk what cause it
<hacker420[m]> would explain the weird booting\
<barni2000[m]> but panel works btw
<barni2000[m]> it cause freeze next reboot is not working on land
<barni2000[m]> try to remove ramoops region
<barni2000[m]> and don't use the cmdline param
<barni2000[m]> it seems land is fine now, lets try couple more reboot
<hacker420[m]> it doesn't help the damn usb is flaky
<hacker420[m]> and that's a problem with my phone
<hacker420[m]> not the kernel here
<barni2000[m]> i would say change the charging port but moto using idiotic iphone style assemblement
<hacker420[m]> barni2000[m]: which param
<barni2000[m]> you need to remove the display first to fix anything
<barni2000[m]> hacker420[m]: lk2nd.pass-ramoops ...
<hacker420[m]> no?
<hacker420[m]> barni2000[m]: I don't have that
<hacker420[m]> my usb is on a separate board
<barni2000[m]> i don't speak about the pcb
<barni2000[m]> only lg puts it on the same board
<hacker420[m]> tbh
<hacker420[m]> these boards are so cheap that I don't see the point of replacing just the plug
<barni2000[m]> but for change battery or anything you need to remove display first
<hacker420[m]> looks at cedric with removable back
<barni2000[m]> at xiaomis you need to remove the back cover
<barni2000[m]> oh cedric is repairable here is montana /watch?v=A2-rT93YZo8
<barni2000[m]> potter is same sh**
<barni2000[m]> g5 plus
<hacker420[m]> ok I reenabled panel and storage
<hacker420[m]> now I get no usb device
<hacker420[m]> what the hell
<barni2000[m]> because it is panicking
<barni2000[m]> s/panicking/panicing/
<barni2000[m]> * because it has panic
<hacker420[m]> except I already tried multiple reboots
<hacker420[m]> the question is what causes the panic
<hacker420[m]> panel or storage
<barni2000[m]> i think none of them
<barni2000[m]> but land also has booting issues
<barni2000[m]> try fw_devlink=permissive
<barni2000[m]> maybe it is same initfs stuff what i have with some devices on msm8953
<barni2000[m]> not helps than maybe go in recovery
<barni2000[m]> and try after
<barni2000[m]> ok, no i have done about 10x reboot but it is freezing :(
<hacker420[m]> <barni2000[m]> "try fw_devlink=permissive" <- ok this gave me a device much quicker
<hacker420[m]> and I got telnet
<hacker420[m]> https://bpa.st/VLHA
<hacker420[m]> what do I try now?
<hacker420[m]> continue boot?
<hacker420[m]> don't see any panics
<barni2000[m]> it is not booting anymore i have same issue like you
<barni2000[m]> what lsmod shows to you?
<hacker420[m]> in debug shell
<barni2000[m]> no msm no panel
<hacker420[m]> I could try to manually modprobe it
<barni2000[m]> try maybe the modules are not there
<hacker420[m]> modprobe: ERROR: ctx=0xffffaede2880 path=/lib/modules/6.10.0-msm8917/kernel/drivers/soc/qcom/mdt_loader.ko error=No such file or directory
<hacker420[m]> modprobe: ERROR: ctx=0xffffaede2880 path=/lib/modules/6.10.0-msm8917/kernel/drivers/soc/qcom/mdt_loader.ko error=No such file or directory
<hacker420[m]> modprobe: ERROR: could not insert 'msm': Unknown symbol in module, or unknown parameter (see dmesg)
<barni2000[m]> it was packed wrong module
<barni2000[m]> did you installed the latest built kernel in the chroot?
<hacker420[m]> yes
<hacker420[m]> I can retry
<barni2000[m]> ok
<hacker420[m]> I just run build with envkernel
<hacker420[m]> then pmb install
<hacker420[m]> and then flash rootfs
<barni2000[m]> / # ls /mnt/pmbootstrap/packages/aarch64/
<barni2000[m]> s///`#/, s/#//, s////`/
<barni2000[m]> chroot -r
<barni2000[m]> * pmbootstrap chroot -r
<hacker420[m]> wait hold on
<barni2000[m]> you only need just the boot image from the chroot
<hacker420[m]> could be that it has just that in initramfs?
<barni2000[m]> it seems lot of stuff is missing
<hacker420[m]> probably because it's initramfs
<hacker420[m]> not the proper rootfs
<barni2000[m]> i would show my initfs but i cannot atm
<barni2000[m]> i am going to switch to montana
<hacker420[m]> weird
<hacker420[m]> if I try to continue boot it dumps me back to debug shell
<hacker420[m]> Wait
<hacker420[m]> I think I see the issue
<hacker420[m]> I guess the modules-initfs thing is deciding what ti put in initramfs?
<barni2000[m]> hacker420[m]: it was copied to /usr/share/mkinitfs/modules/ as 00-motorola-montana.modules
<barni2000[m]> there is an 00-default.modles there also
<barni2000[m]> s/modles/modules/
<hacker420[m]> think my apkbuild for the device package is wrong
<hacker420[m]> hang on
<barni2000[m]> i am only modifying these files in chroot for testing
<barni2000[m]> it is much quicker way to test stuff
<barni2000[m]> and storage is not needed for debug-shell
<barni2000[m]> evrything is in the memory
<barni2000[m]> you can rebuild initfs with mkinitfs
<barni2000[m]> i don't understand why montana is stable
<hacker420[m]> so you had no console=tty0 on montana in cmdline set?
<barni2000[m]> no
<barni2000[m]> just PMOS_NO_SHELL
<hacker420[m]> what
<hacker420[m]> you mean no_splash?
<barni2000[m]> * just PMOS_NO_SPLASH
<barni2000[m]> yes i was just miss typed that
<hacker420[m]> wait wait wait hang on
<hacker420[m]> after I manually modprobed it
<hacker420[m]> backlight went out
<hacker420[m]> does that mean I need gpu fw?
<hacker420[m]> but any idea why I can't contine boot?
<hacker420[m]> is it that parameter?
<hacker420[m]> barni2000 you have gpu fw installed?
<barni2000[m]> no it is not needed for msmdrmfb
<hacker420[m]> hm
<hacker420[m]> so then what npw
<hacker420[m]> *now
<hacker420[m]> for some reason it doesn't init when in initfs automatically
<hacker420[m]> could be that alignment exception
<hacker420[m]> also
<hacker420[m]> I don't see storage
<barni2000[m]> btw storage is disabled on montana atm
<hacker420[m]> because?
<hacker420[m]> causes crashes?
<barni2000[m]> no i have just enabled the minimal stuff what is enough for test panel
<hacker420[m]> so right now the thing to figure out'
<hacker420[m]> what is that alignment exception about
<hacker420[m]> and why doesn't panel auto probe
<hacker420[m]> wait
<barni2000[m]> hacker420[m]: check your compatibles
<hacker420[m]> storage is disabled as in dts or config
<hacker420[m]> barni2000[m]: in dts?
<barni2000[m]> dts
<barni2000[m]> hacker420[m]: dts
<barni2000[m]> seems fine
<hacker420[m]> initfs
<hacker420[m]> ~ # ls /lib/modules/6.10.0-msm8917/kernel/drivers/gpu/drm/panel/msm8937-generated/
<hacker420[m]> panel-motorola-cedric-497-tianma.ko
<hacker420[m]> Hold on
<hacker420[m]> panel is working
<hacker420[m]> backlight is dead
<hacker420[m]> doesn't change the fact it won't auto probe
<barni2000[m]> it is very dimmed?
<barni2000[m]> * is it very dimmed?
<hacker420[m]> yes
<hacker420[m]> do I try to enable this?
<hacker420[m]> or is it dim on montana too
<barni2000[m]> no montana works fine without wled
<barni2000[m]> you can try to enable wled but 29500 is not valid value use 29600
<barni2000[m]> and check the other values also
<barni2000[m]> enabled-strings and num-strings means the same
<barni2000[m]> only use num-strings
<barni2000[m]> [00 01 10] in downstream means 3 [00 01] means 2
<hacker420[m]> that's in downstream
<barni2000[m]> it is just an override
<barni2000[m]> this is from a decompiled dt
<barni2000[m]> * [00 01 02] in downstream means 3 [00 01] means 2
<barni2000[m]> qcom,led-strings-list = [00 01 02]; this means qcom,num-strings=3
<hacker420[m]> so this
<barni2000[m]> pfet, cabc, num-strings=2, 980 mA, 25000, 29500 mv (ovp) -> 29600 mv
<barni2000[m]> so current limit is 25000
<barni2000[m]> cabc pfet
<hacker420[m]> hm?
<barni2000[m]> qcom,external-pfet
<barni2000[m]> seems fine
<hacker420[m]> ok this time it auto probed
<hacker420[m]> but it's still dark
<hacker420[m]> but
<barni2000[m]> hacker420[m]: > <@hacker420:kde.org> ```... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/BIlXHnpszomQBvssGSQaMoya>)
<barni2000[m]> pmi8950 only have sc-irq i am thinking about report this but it is working anyway
<hacker420[m]> so what now
<barni2000[m]> idk it should be fixed in the panel driver i think
<barni2000[m]> maybe you can try --dumb-dcs param
<hacker420[m]> in the panel driver generator?
<barni2000[m]> yes
<hacker420[m]> can you try to regenerate it for me?
<barni2000[m]> is it still not working?
<hacker420[m]> nope
<barni2000[m]> what abou in docker?
<barni2000[m]> s/abou/about/
<hacker420[m]> running docker for one program feels overkill
* barni2000[m] posted a file: panel-motorola-cedric-nt35596-tianma.c (6KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/JDbcpKOTcYMcfdBnbYQEcmxa >
<hacker420[m]> turns out the generator does work
<hacker420[m]> I have no idea what it crashes on there
<barni2000[m]> land works again
<barni2000[m]> i guess something was in wrong state maybe i should investigate for resets
<hacker420[m]> do I try with wled on first?
jojo_autoboy[m] has joined #msm8937-mainline
<jojo_autoboy[m]> hmm speaking of land what's the current state of things
<barni2000[m]> booting, panel works
<barni2000[m]> unstable a bit
<barni2000[m]> both clusters is working but 8937 cannot reach higher than 1.2GHz i think because missing saw2 and cpr driver
<barni2000[m]> * think because of missing saw2
<barni2000[m]> i was forgot to add mdss reset it helps some times to avoid some interrupt storm caused by bootloaders
<jojo_autoboy[m]> esh
<barni2000[m]> but it not helps on land :(
<hacker420[m]> barni2000 yeah the --dumb-dcs thing worked
<barni2000[m]> cool
<barni2000[m]> land has this issue
<barni2000[m]> hacker420[m]: if you enable wifi, storage and install firmware-qcom-adreno-a530 than you can play supertuxkart now :D
<hacker420[m]> looks like I still need this fw_devlink=permissive
<hacker420[m]> and I can't seem to boot to rootfs
<hacker420[m]> it does give me a device
<hacker420[m]> but I can't ssh or telnet in
<hacker420[m]> removed the debug shell param
<hacker420[m]> nothing on screen either
<barni2000[m]> maybe the same issue what i have on land
<hacker420[m]> maybe
<barni2000[m]> you can try limit CPUs more
<hacker420[m]> is there some cmdline param for it
<barni2000[m]> i mean in opp
<hacker420[m]> ah
<hacker420[m]> so I leave figuring out rootfs boot for later
<barni2000[m]> ok