Jump to content

thc013

Members
  • Posts

    52
  • Joined

  • Last visited

Everything posted by thc013

  1. well about hacked together they all use the same bsp to get their precompiled libraries and they all use the same dowmstream kernel . there is just some diffrences from wich date they based their kernel from the dowmstream one , and some drivers for khadas edge2 the mcu or small difrent lcd's screens in their shop or other wifi or net (opi5) and 1 use the "not official" panfork mod and the other the default mali blob and pd issue 5B i dont know use a xiamo charger wich got free with a phone and happy does 20V and the sd reader is crap and the nanopi kernel is the most modded one with their openwrt things etc and for the OS it is just how you configure it and if you like the khadas ubunutu more as stock raxda os then use khadas ubuntu on a 5B how is the av1 and avs streaming is going on other distros as armbian or wich board has it supported ?
  2. @maka they sure did not 🤣 yours is just a bad decompiled one you stll don't listen and for your information that dts in mainline is from last year. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/arch/arm64/boot/dts/rockchip?h=v6.2-rc7&id=2e0537b16b2557974f81db008a51f41c838dcb81 and it says Signed-off-by: Piotr Oniszczuk <piotr.oniszczuk@gmail.com> Signed-off-by: Andy Yan <andyshrk@163.com> and there is no maka in it
  3. @maka 🤣 if the rk35xx line is the whole arm world , and it is just old news did this a year ago a tvbox can just run linux same as a sbc it only how much work you wanna put into it . but other as nonsens what did you bring in ?
  4. @maka if you think every box is the same your complete wrong !!! only the t95 ,tr43 and x96 have box-demo as base dts so and it is not fully supported there is still no npu in mainline and video decoding is nah it works thats it @mmie4jbcu the x88 is based on the rk3566-evb5-lp4x-v10 so a board with a io-domain from a rk809 will do such as indeed a m2 from firefly but for making a proper dts compare it downstream with a rk3566-evb5-lp4x-v10. and the factory dtb has 2 dts files use the second @hotnikq some people as @tmm1 give you a very good hint wich is the base dts and a decompiled dts is not a proper dts it is a start , and what you now got is nothing and if you looked around you see there is also a android 12 image for your box with a little updated dts and is easy to compare to the armbian legacy build just decompile the dtb of rk3566-evb3-ddr3-v10 and compare it against the decompiled of the android 12 image . ow and the h96 factory dtb has 2 dts files one with a tcs4525 config and the other has a sti8070 so check your pcb wich it has . wonder what the dmesg says about the included dtb ow btw it is thc013 aka shaggy013 🤗 h96-v56-sti8070.dtb h96-v56-tcs4525.dtb
  5. lol what a mess do just a proper reverse of the dts then just using a decompiled one .
  6. your confusing dtb and dts and you realy have no clue when you say compare a 4.19 dtb with a mainline dtb your still wanna go from step 1 to end result wich is not gonna work the problem is that people dont read and are lazy !
  7. h96 v56 has a rk809 the h96 v58 (rk3588) has a pmic x88 has no pmic from other boards i cannot see it so quick quartz64-a has rk817 firefly has rk809 etc etc at least warpme did a proper dtb downstream > bad decompiled dts > dts proper downstream ( adapt one from device tree folder or make dts wich at the end match compiled with original dtb ) conversion and then changed it for mainline in other words he did ALL the steps then just 1 not working hdmi is just 1 name change in a dts from below 4.19.206 would you also like a explanation about ? "a decompiled dtb is not a dts for use in kernel and it is sure not gonna work with a 4.19 on a 5.10 kernel" or your gonna call the Ghostbusters Hotnikq is not gonna find it in my outdated guide to build from bsp that is a guide of building from bsp and has nothing to do with reverse engineer a dtb . Tmm1 is not gonna make it either as long he thinks that some editing in a decompiled dts is gonna make it work . and posting about a end result is not gonna help either as long you did not all the steps from the dtb of your board . And why "devs" have no interest is they dont have that board for testing and helping remotly is just not gonna work , just buy a sbc from one of the vendors like pine64 raxda firefly etc they just as cheap as a tvbox and 100% more support
  8. more nonsens maka ? it is a dts for a x96_x6 wich dont have a pmic because it is a cheaptv box and indeed too dangereous for you !
  9. lol that is all not gonna work same as with that h96 v56 a decompiled dtb is not a dts for use in kernel and it is sure not gonna work with a 4.19 on a 5.10 kernel just check if your board has a pmic and if it is a rk817 or rk809 otherwise things get very hot for 4.19 below 219 there are some name changes in the dts to work on +219 version and for 5.10 you have to adjust some settings in storage devices
  10. It works even the graphics damn how i as noob gonna without asking compare dts ,defconfig against a proper downstream one from a sensei for a quartz64 , mainstream is no issue lot of sensei for that .
  11. yeah that is just mainline kernel would be weird if it need testing after so long time where is the dowmstream kernel ?
  12. i got the model you should have by now or is on the way the firefly dts is weird core and then base and then model pcie was a little fixed months ago and lately someone noticed data corruption so they fixed that in the last 5.16, 5.17 last what i read about is that the patches are not included yet but you can find them , i see you included these usb3 and pcie patch. the 4.19 downstream rockchip also adjusted some pcie settings i wouldnt know if that dts works or not , proberly some litte changes it is a dts from the 5.13-14 and yeah comparing a downstream dts with a mainline is not 1 to 1 but you know that , but for mainline it is easy you have a good working dts for the quartz64 to begin a port for a m2 only a shematic a your almost done
  13. i didnt it depends how smart or stupid os-assistant program/script is not uboot . and in my opinion you could solve your answer with just a google search got 4 partitions and editing a extlinux.conf is easy .
  14. Huh i am not a owner of a m2 but i am a rk3566 user pci ? pcie not working ? https://gitlab.com/pine64-org/quartz-bsp/linux-next/-/commit/0e2d1d8ff6cfe4bca99de4906d6eb14a2416b2c6 well my pcie dvbc tv card works from what i understand it is not 100% stable with high loads , more the fine tunning and would be fixed if another thing got posted first in the kernel ow and hdmi in 5.10 does work , only trouble i have is with rga and the npu
  15. a 3A power supply is enough you can check L2101 inductor the opi 4 is little bad with noisy power and active cooling is needed for working under heavy load on a rk3399 i compile gentoo on a opi4 wit all usb populated and a mini pcie wifi (yeah rust,gcc,clang spidermpnkey is no fun ) idle 30c /full load 60 c 5.10 is the most stable
  16. not usb3 dp alt-dp power rk3399-orangepi-4.dtb rk3399-orangepi-4.dts
  17. yeah the admin of that forum and you got 423 channels and tvheadend is fun there is a wiki a manual and a lot of info how it works the wizard is just to get it working , and then you have to tune it to your desires map the services to channels has the option check mux changes to be enabled or not etc,etc and epg is not that fast and works when you have it setup . just check what the mux is to start to scan to get a mux list , is the provider id right ,is there a encryption involved , etc etc
  18. it is not a bootsplash problem at boot the kernel searches for device and activates them and if a device takes time to get seen by the kernel the kernel get a amount of time to wait so when a device for example my usb3 got called before the bootsplash there is no bootsplash because kernel is waiting for the usb before he wants to put something on screen. you can see that effect if you boot through a console then you see the kernel spit out messages sometimes fast sometimes slower and it happens even more when you reboot but that is more less nowdays because the drivers are getting better
  19. ow iam not a little it is just a very old law they call it the law of ohm if i got 3 amp at 4.8v or i get 3amp at 5v makes a difference off 0.2v what should the R do ? "voltage drop since I'm going thru a voltage meter" realy ? https://www.dummies.com/programming/electronics/how-to-measure-voltage-on-an-electronic-circuit/ and to inform you 4.8v indicates that you power-supply can't supply 5.0v under load and it is not that panfrost or the codecs are a mess they are just come clear every day even that i know can get a little grip on it . and if you read the posts of EndlessEden you see he is missing the connection between kernel and ffmpeg. and he can remove from dts rockchip,defaultmode = <16>; /* CEA 1920x1080@60Hz */ and this hdmi { /delete-node/ hdmi-i2c-xfer; }; it is useless it will become a lot more clear for him if he compiles ffmpeg with the libreelec patches on his armbian build from source instead of the armbian normal ffmpeg otherwise you get ffmpeg calling the wrong codec paths and also integrate some of the libreelec patches to the armbian kernel otherwise ffmpeg cant let the kernel know to switch to other vpu state kodi doesnt matter much as long as the drmprime patches are included and these didnt change lateley and if he does that he can come to same kernel level as librele or armbian normal 5.10.34 or are we 35 already above 5.10 to 5.11 and 5.12 take a look at minimyth2 sources not realy much different with patches it has a other layout how he comes to the same result . or even my sensei has a libre for alwinner on 5.12 and me i just on 5.12 gentoo system with armbian-librelelec-minimyth2 mixed together and instead of 40 mousepointers when i move the mouse or green screen or v4l2 errors in gbm i now have only no picture on hdr ,sound works and kodi says it is using the correct codec didnt check with v4l utils yet can watch tv again and play most of my series and movies. and in 5.10 everything works. and why overclock even the codec core get underclocked to be smoother
  20. such a useless discussion was the 3.06 at 4.87 or 4.82 and where is the 5v how is you chain to the power grid . and by the looks of it it is powered through a usb hub with powered devices throuh selfmade pcb,s what is the losses of that . and great start of the chain of power you think ? u=ixr i=u/r what is the cos what is the foam layer of the beer from the board ? https://www.dc.systems/dcwiki/power-quality-a-typical-ac-problem https://www.youtube.com/watch?v=pKXPqApOYfk https://www.youtube.com/watch?v=wopmEyZKnYo how is your power supply build you see you got multiple choices to get a 5v on all the opi runs and the opi is delivering power to the ssd,oled,tft, pcie wifi , fan and still hapily compiling gentoo with distcc no pump overclocking isnt just stable power. where was your chiplet on the wafer ? how is the power chain on the board for example the power of the ram on the opi is shared with the rest of the 1.8v rail while other boards have a seperate regulator for that . and does the cpu run from the 5v source directly so you can undervolt it for example . and how is your cooling ? https://github.com/warpme/minimyth2 and even mythtv is still alive and it is another way to run your board as mediabox . how is mainline going almost everything is out of staging or will be soon ?
  21. https://github.com/armbian/build/blob/master/patch/kernel/archive/rockchip64-5.11/rk3399-add-sclk-i2sout-src-clock.patch so i dont know wich distro he is using but he just could copy the patches from armbian ony thig he had to do is choose a dtb or adjust it nothing more
  22. ah you downgraded your kernel with the cp command lucky you still got the modules installed of that kernel. so now your back at 5.9.14 and linux-headers isnt the problem. boot from sd do a blkid write it down or copy it , mount the rootdev look wich kernelmodules are installed in /lib/modules/<kernelname> or copy your modules from sd to rootdev edit armbianenv.txt change the uuid to wich rootdev it should boot.
  23. well you got a mac adress so hardware is seen did you do a apt update /apt upgrade seems more a systemd fault
  24. yeah that is rubbish armbianio and if wiringpi or wiringnp doesnt work or you edit the script of it or just calculate the gpio pin and use the normal linux commands to use a gpio pin like this https://wiki.radxa.com/Rockpi4/hardware/gpio https://www.programmersought.com/article/92031588023/ and ofcourse use the pinout of your board shematic
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines