Open tranzmatt opened 5 years ago
Hi, we will be switching to warrior. Watch for this PR to get merged https://github.com/balena-os/balena-allwinner/pull/84
Will this fix the error above? Because I'm trying this from the warrior_update branch and it seems that the overlay sub-dir isn't being properly pulled in.
It would. The PR would be merged only when all the supported boards compile ok.
I've tried the latest warrior_update branch and am still getting the same error. Trying to merge in the PR into master doesn't seem to work as it get complaints about the layer compatibility, even after I try to sync submodules.
On Tue, Oct 8, 2019 at 11:00 AM Florin Sarbu notifications@github.com wrote:
It would. The PR would be merged only when all the supported boards compile ok.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/balena-os/balena-allwinner/issues/86?email_source=notifications&email_token=AAGTZCSQ5LZM52BYDLFTYW3QNSOCHA5CNFSM4I6J5OBKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEAUPM4Y#issuecomment-539555443, or mute the thread https://github.com/notifications/unsubscribe-auth/AAGTZCSJPBWPD4OADFDBLFDQNSOCHANCNFSM4I6J5OBA .
BTW, I tried to build the same orange-pi-lite under the old master branch using a 16.04 Docker build environment, but it doesn't work either. :(
Commit 2cf0d9452b51b1105675943e98cb05f0cf19fadc seemed to work, but I had to edit a bbappend file to remove a patch conflict:
diff --git a/layers/meta-balena-allwinner/recipes-kernel/linux/linux-mainline%.bbappend b/layers/meta-balena-allwinner/recipes-kernel/linux/linux-mainline%.bbappend index 2e47bb5..d4e57ae 100644 --- a/layers/meta-balena-allwinner/recipes-kernel/linux/linux-mainline%.bbappend +++ b/layers/meta-balena-allwinner/recipes-kernel/linux/linux-mainline%.bbappend @@ -2,9 +2,6 @@ inherit kernel-resin
FILESEXTRAPATHS_prepend := "${THISDIR}/linux-mainline:"
On Wed, Oct 9, 2019 at 9:25 AM Matthew Clark linux.matthew@gmail.com wrote:
I've tried the latest warrior_update branch and am still getting the same error. Trying to merge in the PR into master doesn't seem to work as it get complaints about the layer compatibility, even after I try to sync submodules.
On Tue, Oct 8, 2019 at 11:00 AM Florin Sarbu notifications@github.com wrote:
It would. The PR would be merged only when all the supported boards compile ok.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/balena-os/balena-allwinner/issues/86?email_source=notifications&email_token=AAGTZCSQ5LZM52BYDLFTYW3QNSOCHA5CNFSM4I6J5OBKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEAUPM4Y#issuecomment-539555443, or mute the thread https://github.com/notifications/unsubscribe-auth/AAGTZCSJPBWPD4OADFDBLFDQNSOCHANCNFSM4I6J5OBA .
Stay tuned as we review the PR. It will get merged soon enough
I'm having a build error for the orange-pi-lite when trying to use the warrior branch
| DTC arch/arm/boot/dts/sun8i-h3-orangepi-lite.dtb | NOTE: make -j 12 HOSTCC=gcc -isystem/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/usr/include -O2 -pipe -L/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/usr/lib -L/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/lib -Wl,--enable-new-dtags -Wl,-rpath-link,/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/usr/lib -Wl,-rpath-link,/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/lib -Wl,-rpath,/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/usr/lib -Wl,-rpath,/home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/recipe-sysroot-native/lib -Wl,-O1 -Wl,--allow-shlib-undefined -Wl,--dynamic-linker=/home/builder/Code/Yocto/Shared/tmp/sysroots-uninative/x86_64-linux/lib/ld-linux-x86-64.so.2 HOSTCPP=gcc -E overlay/sun8i-h3-analog-codec.dtbo | make[2]: *** No rule to make target 'overlay/sun8i-h3-analog-codec.dtbo'. Stop. | make[1]: *** [Makefile:146: sub-make] Error 2 | make: *** [Makefile:24: __sub-make] Error 2 | ERROR: oe_runmake failed | WARNING: /home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/temp/run.do_compile.14276:1 exit 1 from 'exit 1' | ERROR: Function failed: do_compile (log file is located at /home/builder/Code/Yocto/Shared/tmp/work/orange_pi_lite-poky-linux-gnueabi/linux-mainline/4.19.63-r0/temp/log.do_compile.14276) ERROR: Task (/home/builder/Code/Yocto/Balena.io/balena-allwinner.warrior/build/../layers/meta-sunxi/recipes-kernel/linux/linux-mainline_4.19.63.bb:do_compile) failed with exit code '1' NOTE: Tasks Summary: Attempted 345 tasks of which 341 didn't need to be rerun and 1 failed.
I've been able to build the same device using Armbian, but I can't see a substantial difference between its patched kernel source and the balena version. Has anyone run into this?