Jump to content

dolphs

Members
  • Posts

    272
  • Joined

  • Last visited

Everything posted by dolphs

  1. same here ( idle most time ) but must admit I always use "conservative" even on the H5 boards. BTW still did not have time to spin up a new build ( rc5 ) , come to that later ( weekend perhaps )
  2. cheers @lanefu but this is how I spin up my builds usually, therefore alas no go ( summary from first page below ) I will give it another shot Tuesday earliest I guess ( hoping also for rc5 )
  3. @sfx2000 @Kevin Sewell @guidol @5kft- Looks like NEO3 entered the arena with rk3328 chipset guess this will be adapted shortly by armbian in this section. I believe in theory this board could run on 1,4Ghz just like the NEO2 Black does, for now I skip this one
  4. cheers for that, since I am still learning this means a new checkout ( git clone --depth 1 https://github.com/armbian/build ) or just copy over your patch for time being " board-h6-improve-thermals.patch " , as these fail currently ( last time I built ): 1 out of 2 hunks FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6-cpu-opp.dtsi.rej 1 out of 1 hunk FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi.rej meaning eg " rm board-h6-improve-thermals.patch && wget https://raw.githubusercontent.com/5kft/build/86ad3c26c9f4835cd97e223223c72f0464173848/patch/kernel/sunxi-dev/board-h6-improve-thermals.patch " I did a fresh checkout , spun up a new compile job and verified the thermal patch does not throw any errors as was expected. Yet panfrost related patches to go, not sure about the "Unable to make Panfrost work on H6"-status since after May things are silent, perhaps @NicoD or @Werner have a good view of what is going on, on this one? Anyway in next attempt will exclude these to see what else is coming up. thanks @5kft [OK!] Processing file /home/armbian/patch/kernel/sunxi-dev/board-h6-improve-thermals.patch [NOK] Processing file /home/armbian/patch/kernel/sunxi-dev/board-h6-orangepi-lite2-fix-missing-all.patch 1 out of 1 hunk FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-lite2.dts.rej Processing file /home/armbian/patch/kernel/sunxi-dev/sun50i-h6-drm_panfrost-1-missing-remove-opp-table-in-case-of-failure.patch 2 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_devfreq.c.rej Processing file /home/armbian/patch/kernel/sunxi-dev/sun50i-h6-drm_panfrost-2-add-devfreq-regulator-support.patch 2 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_devfreq.c.rej 1 out of 1 hunk FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_device.h.rej So removed these three patches but alas still stuck ... make[5]: *** [scripts/Makefile.build:280: drivers/net/wireless/realtek/rtl8723cs/core/rtw_br_ext.o] Error 1 make[4]: *** [scripts/Makefile.build:497: drivers/net/wireless/realtek/rtl8723cs] Error 2 make[3]: *** [scripts/Makefile.build:497: drivers/net/wireless/realtek] Error 2 make[2]: *** [scripts/Makefile.build:497: drivers/net/wireless] Error 2 make[2]: *** Waiting for unfinished jobs.... make[1]: *** [scripts/Makefile.build:497: drivers/net] Error 2 make[1]: *** Waiting for unfinished jobs.... make: *** [Makefile:1758: drivers] Error 2 Not entirely sure if this is related to @martinayotte work done in 5.6 kernel cleanup/ fixes... Therefore added compilation.log, this is the one that FAILs. compilation.log
  5. quick summary how to repeat things: sed -i -e 's/orange-pi-5.7/orange-pi-5.8/g' sunxi64_common.inc ./compile.sh BRANCH=dev RELEASE=buster BUILD_MINIMAL=yes BUILD_DESKTOP=no KERNEL_ONLY=no KERNEL_CONFIGURE=yes EXTRAWIFI=no WIREGUARD=no BOARD=orangepioneplus # Compiling dev kernel [ 5.8.0-rc4 ] load .config-h6-default ( file attached in this topic ) # exit, start compiling kernel and building image Errors on following items, thus when patch directory is present : compilation.log : - lots of warnings thrown but ignoring these ( for the moment ) - yet these are of interest that need to be fixed: panfrost and wifi ( realtek ) stuff drivers/gpu/drm/panfrost/panfrost_devfreq.c:86:13: error: ‘struct panfrost_device’ has no member named ‘devfreq’; did you mean ‘pfdevfreq’? drivers/gpu/drm/panfrost/panfrost_devfreq.c:87:36: error: ‘struct panfrost_device’ has no member named ‘devfreq’; did you mean ‘pfdevfreq’? drivers/gpu/drm/panfrost/panfrost_devfreq.c:88:10: error: ‘struct panfrost_device’ has no member named ‘devfreq’; did you mean ‘pfdevfreq’? drivers/gpu/drm/panfrost/panfrost_devfreq.c:91:9: error: ‘ret’ undeclared (first use in this function); did you mean ‘net’? make[4]: *** [scripts/Makefile.build:280: drivers/gpu/drm/panfrost/panfrost_devfreq.o] Error 1 make[3]: *** [scripts/Makefile.build:497: drivers/gpu/drm/panfrost] Error 2 make[2]: *** [scripts/Makefile.build:497: drivers/gpu/drm] Error 2 make[1]: *** [scripts/Makefile.build:497: drivers/gpu] Error 2 drivers/net/wireless/realtek/rtl8723cs/include/rtw_security.h:241:8: error: redefinition of ‘struct sha256_state’ make[5]: *** [scripts/Makefile.build:280: drivers/net/wireless/realtek/rtl8723cs/core/rtw_br_ext.o] Error 1 make[4]: *** [scripts/Makefile.build:497: drivers/net/wireless/realtek/rtl8723cs] Error 2 make[3]: *** [scripts/Makefile.build:497: drivers/net/wireless/realtek] Error 2 make[2]: *** [scripts/Makefile.build:497: drivers/net/wireless] Error 2 make[1]: *** [scripts/Makefile.build:497: drivers/net] Error 2 make: *** [Makefile:1758: drivers] Error 2 patching.log : - those fail ( @5kft the first one is yours I assume ): Processing file /home/armbian/patch/kernel/sunxi-dev/board-h6-improve-thermals.patch 1 out of 2 hunks FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6-cpu-opp.dtsi.rej 1 out of 1 hunk FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6.dtsi.rej Processing file /home/armbian/patch/kernel/sunxi-dev/board-h6-orangepi-lite2-fix-missing-all.patch 1 out of 1 hunk FAILED -- saving rejects to file arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-lite2.dts.rej Processing file /home/armbian/patch/kernel/sunxi-dev/sun50i-h6-drm_panfrost-1-missing-remove-opp-table-in-case-of-failure.patch 2 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_devfreq.c.rej Processing file /home/armbian/patch/kernel/sunxi-dev/sun50i-h6-drm_panfrost-2-add-devfreq-regulator-support.patch 2 out of 3 hunks FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_devfreq.c.rej 1 out of 1 hunk FAILED -- saving rejects to file drivers/gpu/drm/panfrost/panfrost_device.h.rej Processing file /home/armbian/patch/misc/general-packaging-5.6.y.patch # either to be removed or updated for kernel 5.8 ( as well 5.7 ) # more plausible ones but not of interest atm - output.log : <snip snip> Displaying message: * [\e[32ml\e[0m][\e[32mc\e[0m] xxx-orangepiprime-addregulator.patch info Displaying message: * [\e[32ml\e[0m][\e[32mc\e[0m] xxx-orangepizero-plus2-default-regulator.patch info Displaying message: * [\e[32ml\e[0m][\e[32mc\e[0m] xxx-sun50i-h6-enable-higher-clock.patch info Displaying message: Compressing sources for the linux-source package Displaying message: Compiling dev kernel 5.8.0-rc4 info Displaying message: Compiler version aarch64-linux-gnu-gcc 8.3.0 info Displaying message: Using kernel config file config/kernel/linux-sunxi64-dev.config info Displaying message: Exporting new kernel config /home/armbian/output/config/linux-sunxi64-dev.config info Displaying message: ERROR in function compile_kernel compilation.sh:407 err Displaying message: Kernel was not built @host err - install.log : not present at this stage Please find compilation and patching logs attached. logs.zip
  6. Hi, Regarding this message "long" time ago I decided to give it another go So found myself some time to spin up a new build for rc4. ensuring " sed -i -e 's/orange-pi-5.7/orange-pi-5.8/g' sunxi64_common.inc " and renamed patch dir to patch.bak, building with these parameters: " ./compile.sh BRANCH=dev RELEASE=buster BUILD_MINIMAL=yes BUILD_DESKTOP=no KERNEL_ONLY=no KERNEL_CONFIGURE=yes EXTRAWIFI=no WIREGUARD=no BOARD=orangepioneplus " I attached " .config-h6-default "-file and should save us from pile of questions for the moment ( see also attached kernel58rc4_01.txt ) The build resulted in " Armbian_20.08.0-trunk_Orangepioneplus_buster_dev__minimal.img " that is true, but did not boot of course due to patches that need to be present However noticed following from the debug logs : - compilation.log shows no more errors, but quite some warnings I ignored - install.log shows a WARNING " apt does not have a stable CLI interface. Use with caution in scripts." that I ignored - patching.log/ - output.log : /home/armbian/patch/misc/general-packaging-5.6.y.patch: No such file or directory /home/armbian/lib/compilation.sh: line 741: /home/armbian/patch/misc/general-packaging-5.6.y.patch: No such file or directory [ warn ] * applying general-packaging-5.6.y.patch [ failed ] The patch dir got renamed to patch.bak so that is not the issue here, but this should be updated/ fixed for kernel 5.7 and kernel 5.8 ( jira story ? ) Next steps would be: - checking kernel settings more thoroughly, - take a look at the (redundant) patches in 5.8 ( and 5.7 ) - a bump for u-boot perhaps, as v2020.07 seems to be stable now. All Jira stories? Last but not least found out these are merged in ( mainlining effort ) 5.8 <*> SPI multiplexer support ( was in 5.7 already merged ) <*> Allwinner sun6i/sun8i/sun9i/sun50i Message Box [*] Allwinner H6 IOMMU Support rgd DVFS I suppose that is a newer driver in CPU Power Management ( Allwinner nvmem based SUN50I CPUFreq driver ) if I am not mistaken Not sure if @5kft 's contribution has been incorporated though? kernel58rc4_01.txt .config-h6-default
  7. oof lost u there I'm afraid. trying to be of help in some way, thanks so far to the ones reading this topic - however will need to do a "sed" still for dev ...
  8. hmm bummer still no luck with rc2, same status @megi this can be sorted ( also ) for kernel 5.8 please as @martinayotte requested for 5.7? TiA! ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun8i-mixer.ko] undefined! ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun6i_mipi_dsi.ko] undefined! ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun4i-tcon.ko] undefined! ERROR: modpost: "of_chosen" [drivers/phy/allwinner/phy-sun6i-mipi-dphy.ko] undefined!
  9. Ok that indeed explains that part, yihaa one hurdle, but more to overcome - cheers @martinayotte @megi - If not already done, could this be fixed as well for 5.8 release please? Will do another attempt once rc2 has been released
  10. Looks like these are the new names in kernel 5.8, as well IOMMU is present CONFIG_SUN50I_IOMMU=m drivers/gpu/drm/sun4i/sun8i-mixer.ko # replaced by CONFIG_DRM_SUN8I_MIXER=m drivers/gpu/drm/sun4i/sun6i_mipi_dsi.ko # replaced by CONFIG_DRM_SUN6I_DSI=m drivers/gpu/drm/sun4i/sun4i-tcon.ko # replaced by CONFIG_TOUCHSCREEN_SUN4I ? drivers/phy/allwinner/phy-sun6i-mipi-dphy.ko # replaced by CONFIG_PHY_SUN6I_MIPI_DPHY=m Got some more interesting options but they are attached in kernel58rc1_01.txt Yet will need to browse to all fresh meat available in this new kernel. kernel58rc1_01.txt
  11. Suppose we need to dig in to kernel config , thus " KERNEL_CONFIGURE=YES " and see what is going on there ... maybe later when I had some sleep ...
  12. @Werner - OK I did a " mv patch patch.bak " and gave "compile.sh" another spin From screen output <snip snip> OBJCOPY arch/arm64/boot/Image MODPOST Module.symvers [ error ] ERROR in function compile_kernel [ compilation.sh:383 ] [ error ] Kernel was not built [ @host ] [ o.k. ] Process terminated From logs I can see patching log more ore less clean ( duh ), except for these three which is apparently hard coded somewhere as "patch" directory is definitely not present : Processing file /home/armbian/patch/misc/general-packaging-5.6.y.patch Processing file /home/armbian/patch/misc/kali-wifi-injection-1.patch Processing file /home/armbian/patch/misc/kali-wifi-injection-2.patch so grepping for errors result in : :~/armbian/output/debug$ grep -i error *.log compilation.log:scripts/dtc/libfdt/fdt_strerror.c: In function ‘fdt_strerror’: compilation.log:scripts/dtc/libfdt/fdt_strerror.c:51:18: warning: comparison of integer expressions of different signedness: ‘int’ and ‘long unsigned int’ [-Wsign-compare] --- here is trouble --- compilation.log:ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun8i-mixer.ko] undefined! compilation.log:ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun6i_mipi_dsi.ko] undefined! compilation.log:ERROR: modpost: "of_chosen" [drivers/gpu/drm/sun4i/sun4i-tcon.ko] undefined! compilation.log:ERROR: modpost: "of_chosen" [drivers/phy/allwinner/phy-sun6i-mipi-dphy.ko] undefined! --- /here is trouble --- compilation.log:make[1]: *** [scripts/Makefile.modpost:111: Module.symvers] Error 1 compilation.log:make: *** [Makefile:1370: modules] Error 2 output.log:Displaying message: ERROR in function compile_kernel compilation.sh:383 err Also full compilation.log attached, but seems the four fellowes listed above seem to cause problems : compilation01-rc1.log
  13. OK new patch list I came up with that does not fly in 5.8-rc1:, - general-packaging-5.6.y.patch # processed though redundant/ to be revised for 5.8? - wifi-fix-staging-rtl8723cs-for-5.6.y.patch - 0004-sun4i-i2s-improvements.patch - 0008-arm64-dts-sun50i-a64-pine64-add-HDMI-audio-nodes.patch - 0010-general-h6-add-dma-i2c-ir-spi-uart.patch - 0101-arm64-dts-allwinner-a64-Enable-HDMI-output-on-A64-bo.patch - 0136-hdmi-audio-fixup.patch - 0534-arm64-dts-sun50i-h6-add-AC200-nodes.patch - add-BergMicro-SPI-flashes.patch - add-missing-sound-hdmi-H3+H5.patch - board-h6-orangepi-lite2-fix-missing-all.patch - board-h6-orangepioneplus-fix-missing-ethernet.patch - general-add-overlay-compilation-support.patch - xxx-add-bananapim2-zero-heartbeat-led.patch - xxx-orangepioneplus-enable-hdmi.patch - xxx-sun50i-h6-update-thermal-configuration.patch patching log is attached which resulted in ERRORS. The compilation log showed follow ERRORS rgd kernel: == kernel == scripts/Makefile.lib:8: 'always' is deprecated. Please use 'always-y' instead arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi:1265.21-1276.5: ERROR (duplicate_node_names): /soc/i2s@1c22800: Duplicate node name ERROR: Input tree has errors, aborting (use -f to force output) make[2]: *** [scripts/Makefile.lib:316: arch/arm64/boot/dts/allwinner/sun50i-a64-bananapi-m64.dtb] Error 2 make[2]: *** Waiting for unfinished jobs.... arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi:1265.21-1276.5: ERROR (duplicate_node_names): /soc/i2s@1c22800: Duplicate node name ERROR: Input tree has errors, aborting (use -f to force output) make[2]: *** [scripts/Makefile.lib:316: arch/arm64/boot/dts/allwinner/sun50i-a64-oceanic-5205-5inmfd.dtb] Error 2 arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi:1265.21-1276.5: ERROR (duplicate_node_names): /soc/i2s@1c22800: Duplicate node name ERROR: Input tree has errors, aborting (use -f to force output) make[2]: *** [scripts/Makefile.lib:316: arch/arm64/boot/dts/allwinner/sun50i-a64-nanopi-a64.dtb] Error 2 arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi:1265.21-1276.5: ERROR (duplicate_node_names): /soc/i2s@1c22800: Duplicate node name ERROR: Input tree has errors, aborting (use -f to force output) make[2]: *** [scripts/Makefile.lib:316: arch/arm64/boot/dts/allwinner/sun50i-a64-amarula-relic.dtb] Error 2 make[1]: *** [scripts/Makefile.build:497: arch/arm64/boot/dts/allwinner] Error 2 So we have a starting point ( square one? ), also noticed: Displaying message: ERROR in function compile_kernel compilation.sh:383 err build01-rc1
  14. Yes sorry for me being clumsy there, " ./compile.sh BRANCH=dev RELEASE=buster BUILD_MINIMAL=yes BUILD_DESKTOP=no KERNEL_ONLY=no KERNEL_CONFIGURE=no EXTRAWIFI=no WIREGUARD=no BOARD=orangepioneplus" # asuming wireguard is built in to kernel so flag can be set to NO was a no go obviously so will start looking at it , also when patch.bak is renamend back to patch, therefore compilation.log is bogus Pfew seen many patches went havoc from the logs - so would be likely flying with you any time soon again compilation.log
  15. hah I'd love to , however I am a total n00b when it comes to python this is what I basically concluded but I must have been missing some stuff somewhere along the road sudo apt update && sudo apt install patchutils python3-pip cd ~/armbian/config/sources/families/include sed -i -e 's/orange-pi-5.6/orange-pi-5.8/g' sunxi64_common.inc cd ~/armbian mv patch patch.bak wget https://github.com/EvilOlaf/refactorpatches/blob/master/refactor.py ~/armbian$ python3.8 refactor.py File "refactor.py", line 6 <!DOCTYPE html> ^ SyntaxError: invalid syntax Update : seen now checking out is not as simple as "wget ", yet better result: ykoot@armbian:~/armbian$ python3.8 refactor.py Patch comparison Options: 1 = print a list of patches sorted by their first target file Example: mypatch.patch -> mycode.c, mycode.h, myothercode.c, ... 2 = split patches targeting multiple files into seperated patches and then print a list sorted by the target file. Optionally show target files only that are altered by multiple patches. Example: mypatch.patch -> mycode.c my2ndpatch.patch -> mycode.c my_other_patch_that_fixes_something_else.patch -> mycode.c 0 = Exit
  16. Hi - Would it be possible to switch dev builds to 5.8 rc kernels? It looks like megous is already prepared for it ( atm rc1 ). Certainly I do realise the massive work filtering DT duplicates and applying fixes, etc ( the usual ) does come in here ... But rgd H6 this 5.8 kernel should hold important additions rgd: CPUFreq (DVFS), IOMMU, and MsgBox ( as well H3,H5 ) Therefore kindly requesting to skip 5.7 in dev and go straight for 5.8 which allows +/- two months of testing wich should be perfect for upcoming "Caple"-release in August.
  17. not entirely sure but imho as from kernel 5.7 SPI is supported. Suppose once megous is incorporated in the / your dev builds you can do another attempt
  18. is that issue current with kernel 4.4.180 and RTL8211E ( though for 5.6 ) ? Perhaps " ethtool -K eth0 tso off " disabling TCP Segmentation Offloading will solve this, however not sure how that impacts performance also if you consider this board will be primarily used as a wireguard vpn/ pihole server. Can one try to see if that helps/ is acceptable for speeds over 100Mbit?
  19. dolphs

    edited

    allritie then, proposed file change - hope did it proper way -
  20. dolphs

    edited

    was hoping 20.05.2 solved things but alas, we have to wait a little more: # sysctl -p ovpn.conf sysctl: setting key "net.ipv4.tcp_congestion_control": No such file or directory net.ipv4.tcp_congestion_control = bbr # uname -a Linux wgvpn01 5.4.43-sunxi64 #20.05.2 SMP Tue Jun 2 17:20:17 CEST 2020 aarch64 GNU/Linux
  21. dolphs

    edited

    meanwhile read here things have been reported earlier this week, sorry for the additional request
  22. dolphs

    edited

    Hi, seems upgrading to 30 May release 20.05.01 I lost "bbr" as congestion control setting. Can this be put back in please? - TiA! - root@wgvpn01:/etc/sysctl.d# uname -a Linux wgvpn01 5.4.43-rockchip64 #20.05.1 SMP PREEMPT Sat May 30 13:30:37 CEST 2020 aarch64 GNU/Linux root@wgvpn01:/etc/sysctl.d# sysctl net.ipv4.tcp_available_congestion_control net.ipv4.tcp_available_congestion_control = reno cubic root@wgvpn01:/etc/sysctl.d# sysctl -p wg.conf net.core.default_qdisc = fq net.core.rmem_max = 16777216 net.core.wmem_max = 16777216 net.ipv4.tcp_rmem = 4096 87380 16777216 net.ipv4.tcp_wmem = 4096 65535 16777216 sysctl: setting key "net.ipv4.tcp_congestion_control": No such file or directory net.ipv4.tcp_congestion_control = bbr net.ipv4.tcp_fastopen = 3 net.ipv4.tcp_no_metrics_save = 1 net.ipv4.tcp_slow_start_after_idle = 0 net.ipv4.tcp_mtu_probing = 1
  23. wow this looks very interesting, thanks so much @5kft, @Werner, @Igor for the ultra fast follow up. Now I wonder what the mainling effort planned in addition to this in kernel 5.8 ? Will flash my spare OpiOnePlus today or tomorrow, thanks again!
  24. Hi - Was wondering about the throttling issue encountered, also on H3/ H5 boards? Is there a fix meanwhile, also for the H6 boards or do we need to wait till kernel 5.8 has been completed ( DVFS aka Dynamic voltage and frequency scaling) ? Suppose @5kft knows more about this one? It is just a question and dont push my OPiOnePlus board to the limit, also because of conservative governor setting - works fine also on this H6 board - for wireguard ( just like my fav boards NEO2 Black ) purposes
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines