I use the blow command to create a /home partition which can grow to fill the available space, on the Yocto arm64 platform.
part /home --source rootfs --rootfs-dir={IMAGE-ROOTFS}/home --fstype=ext4 --label home --part-name home --part-type xxx-xxx-xxx --size=1024 --grow
But building pops the unrecognized argument --grow.
Is there any problem with the "part" command in Yocto wic?
Related
I would like to run a Yocto image in QEMU but the way how it's described in the documentation doesn't work.
To verify I'm not doing anything wrong i followed the steps in the quick build guide:
install required packages
clone poky
checkout correct version
source build environment
set machine to qemux86 in local.conf
add sstate-mirrors and allow parallel build
start bitbake core-image-sato
do something else for the next few hours
when I try now to run that image in qemu like describe in the documentation:
runqemu qemux86
I just get the following output, nothing happens:
runqemu - INFO - Running MACHINE=qemux86 bitbake -e...
runqemu - INFO - Continuing with the following parameters:
KERNEL: [/mnt/wwn-0x50014ee0576fe9ef- part1/test_python3_in_yocto/build/tmp/deploy/images/qemux86/bzImage--4.14.76+git0+3435617380_2c5caa7e84-r0-qemux86-20190305114605.bin]
MACHINE: [qemux86]
FSTYPE: [ext4]
ROOTFS: [/mnt/wwn-0x50014ee0576fe9ef-part1/test_python3_in_yocto/build/tmp/deploy/images/qemux86/core-image-base-qemux86-20190305151244.rootfs.ext4]
CONFFILE: [/mnt/wwn-0x50014ee0576fe9ef-part1/test_python3_in_yocto/build/tmp/deploy/images/qemux86/core-image-base-qemux86-20190305151244.qemuboot.conf]
runqemu - INFO - Setting up tap interface under sudo
runqemu - INFO - Network configuration: 192.168.7.2::192.168.7.1:255.255.255.0
runqemu - INFO - Running /mnt/wwn-0x50014ee0576fe9ef-part1/test_python3_in_yocto/build/tmp/work/x86_64-linux/qemu-helper-native/1.0-r1/recipe-sysroot-native/usr/bin/qemu-system-i386 -device virtio-net-pci,netdev=net0,mac=52:54:00:12:34:02 -netdev tap,id=net0,ifname=tap0,script=no,downscript=no -drive file=/mnt/wwn-0x50014ee0576fe9ef-part1/test_python3_in_yocto/build/tmp/deploy/images/qemux86/core-image-base-qemux86-20190305151244.rootfs.ext4,if=virtio,format=raw -vga vmware -show-cursor -usb -device usb-tablet -device virtio-rng-pci -cpu pentium2 -m 256 -serial mon:vc -serial null -kernel /mnt/wwn-0x50014ee0576fe9ef-part1/test_python3_in_yocto/build/tmp/deploy/images/qemux86/bzImage--4.14.76+git0+3435617380_2c5caa7e84-r0-qemux86-20190305114605.bin -append 'root=/dev/vda rw highres=off mem=256M ip=192.168.7.2::192.168.7.1:255.255.255.0 vga=0 uvesafb.mode_option=640x480-32 oprofile.timer=1 uvesafb.task_timeout=-1 '
When I try to run qemu without graphics I get a kernel panic:
runqemu nographic qemux86
...
[ 6.171521] EXT4-fs (vda): mounted filesystem with ordered data mode. Opts: (null)
[ 6.172937] VFS: Mounted root (ext4 filesystem) on device 253:0.
[ 6.175806] devtmpfs: error mounting -2
[ 6.237143] Freeing unused kernel memory: 852K
[ 6.238001] Write protecting the kernel text: 8752k
[ 6.238722] Write protecting the kernel read-only data: 2376k
[ 6.244382] Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation/admin.
[ 6.245455] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.14.76-yocto-standard #1
[ 6.245913] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org 04/01/4
[ 6.246730] Call Trace:
[ 6.247788] dump_stack+0x58/0x72
[ 6.248071] ? rest_init+0x90/0xc0
[ 6.248320] panic+0x94/0x1c6
[ 6.248529] ? rest_init+0xc0/0xc0
[ 6.248807] kernel_init+0xda/0xf0
[ 6.249046] ret_from_fork+0x2e/0x38
[ 6.249834] Kernel Offset: 0xd800000 from 0xc1000000 (relocation range: 0xc0000000-0xd07dbfff)
[ 6.250595] ---[ end Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentat.
Is there something missing in the documentation?
I also tried different images...
core-image-sato
core-image-base
core-image-minimal
and finally I tried it with version 2.5.2 (sumo) instead of 2.6.1 (thud)... but no change...
When I googled for that issue I didn't really find anything helpful except increasing the memory, which didn't change anything, so I hope anyone here someone knows whats wrong...
Probably missing an appendage to the second parameter.
Browse to the folder in the error message, and check the directory qemux86 exists, chances are it doesn't.
I had to
runqemu qemux86-64
and not
runqemu qemux86
If you haven't changed MACHINE variable in your local.conf file in build directory, check out the file and make sure your image name was identical.
I am building Yocto for AGL image (for more details: automotivelinux.org).
The below error occurred during the build progress (do_rootfs).
In packagegroup-agl-demo-platform.bb, declared packagegroup-agl-image-ivi as a runtime dependent package.
RDEPENDS_${PN} += "\
packagegroup-agl-image-ivi \
"
I can build successfully the packagegroup-agl-image-ivi separately. But when building the whole agl-demo-platform image, happened as follows:
ERROR: agl-demo-platform-1.0-r0 do_rootfs: Unable to install packages. Command '/LTSI4.9/LTSI4.4/build/tmp/work/m3ulcb-agl-linux/agl-demo-platform/1.0-r0/opkg.conf -t /LTSI4.9/build/tmp/work/m3ulcb-agl-linux/agl-demo-platform/1.0-r0/temp/ipktemp/ -o /LTSI4.9/build/tmp/work/m3ulcb-agl-linux/agl-demo-platform/1.0-r0/rootfs --force_postinstall --prefer-arch-to-version install
run-postinsts
screen
kernel-modules
packagegroup-agl-devel
packagegroup-core-eclipse-debug
mc packagegroup-core-tools-profile
kernel-module-vsp2 kernel-module-pvrsrvkm
kernel-module-vspm-if
opkg packagegroup-core-tools-debug
psplash kernel-module-vspm
packagegroup-core-ssh-openssh
packagegroup-agl-demo-platform
omx-user-module kernel-devicetree'
returned 1:
Solver encountered 1 problem(s):
Problem 1/1:
- package packagegroup-agl-demo-platform-1.0-r0.all requires packagegroup-agl-image-ivi, but none of the providers can be installed
Solution 1:
- do not ask to install a package providing packagegroup-agl-demo-platform
ERROR: agl-demo-platform-1.0-r0 do_rootfs: Function failed: do_rootfs
ERROR: Logfile of failure stored in: /LTSI4.9/build/tmp/work/m3ulcb-agl-linux/agl-demo-platform/1.0-r0/temp/log.do_rootfs.14498
ERROR: Task (/LTSI4.9/meta-agl-demo/recipes-platform/images/agl-demo-platform.bb:do_rootfs) failed with exit code '1'
Can anyone help me out in this case ?
I tried 02 ways as follows. They did work.
First method, I cleaned all relative packages and rebuild the whole image.
$ bitbake -c cleanall -c cleansstate <recipes>
recipes consisted of all dependent & runtime dependent packages. But it was a little bit confused to inexperienced users to determine which ones.
Second method, I wiped out the build/tmp/, cache/, sstate-cache/ folders, and re-build all Yocto packages.
There were nothing happening any more. It was really a bad idea if be in critical period of time, but if have free time, be helpful.
I am very new to yocto world, I am following below instructions to compile yocto.
http://doc.qt.io/QtForDeviceCreation/qtee-custom-embedded-linux-image.html
When I compiled yocto(boot2qt) from qt repository I got sdcard image. When I flashed sdcard to memory card I observed that there is no files in /boot/ directory kernel and dtb files are present in another partition mmc1p1. My u-boot configuration is as below
setenv bootargs 'console=ttymxc1,115200 root=/dev/mmcblk0p2 rootwait fixrtc rw'
setenv bootcmd 'mmc dev 1; ext2load mmc 1:2 10800000 /boot/zImage;ext2load mmc 1:2 0x12000000 /boot/imx6q-sabrelitesd.dtb; bootz 0x10800000 - 0x12000000'
saveenv
And its not booting because there is no files in /boot/ directory. Then I changed my uboot configuration as below
setenv bootcmd 'mmc dev 1; ext2load mmc 1:2 10800000 /zImage;ext2load mmc 1:2 0x12000000 /imx6q-sabrelitesd.dtb; bootz 0x10800000 - 0x12000000'
and I am getting file not found error, so its clear that I have mentioned
root=/dev/mmcblk0p2 and I am trying to load kernel from mmc0p1
How to fix this issue? Should I have to do something on yocto configuration and make kernel and dtb files to be available in /boot/ directory or I have to change u-boot configuration to load kernel from different partition and rootfs from different partition?
The problem here is you are still trying to load kernel and dtb from partition 2, you have to update bootcmd. And as per my knowledge with yocto partition1 will be in FAT file system so you have to use fatload.
setenv bootcmd 'mmc dev 1; fatload mmc 1:1 10800000 /zImage; fatload mmc 1:1 0x12000000 /imx6q-sabrelitesd.dtb; bootz 0x10800000 - 0x12000000'
I believe you are using .SDCARD image from yocto. So then follow the below command to flash your image to sdcard. First delete all your sdcard partitions and umount the sdcard. then use lsblk command to know the /dev/sdb|c|d or mmcblk0 etc
sudo dd if=fsl-image-gui-<MACHINE_NAME>.sdcard of=/dev/sd[b|c|d|e] bs=1MB
[ 2% 247/11700] target SharedLib: lib...libart_intermediates/LINKED/libart.so)
FAILED: /bin/bash -c "prebuilts/clang/host/linux-x86/clang-2690385/bin/clang++ -nostdlib -Wl,-soname,libart.so -Wl,--gc-sections -shared -Lout/target/product/generic/obj/lib out/target/product/generic/obj/lib/crtbegin_so.o out/target/product/generic/obj/SHARED_LIBRARIES/libart_intermediates/interpreter/mterp/out/mterp_arm.o
prebuilts/gcc/linux-x86/arm/arm-linux-androideabi-4.9/arm-linux-androideabi/bin/ld: error: out/target/product/generic/obj/SHARED_LIBRARIES/libart_intermediates/type_lookup_table.o: file is empty
prebuilts/gcc/linux-x86/arm/arm-linux-androideabi-4.9/arm-linux-androideabi/bin/ld: error: out/target/product/generic/obj/SHARED_LIBRARIES/libart_intermediates/verifier/instruction_flags.o: file is empty
art/runtime/base/scoped_flock.cc:37: error: undefined reference to 'art::OS::OpenFileWithFlags(char const*, int)'
art/runtime/check_jni.cc:3240: error: undefined reference to 'art::operator<<(std::__1::basic_ostream >&, art::Primitive::Type const&)'
art/runtime/check_jni.cc:3056: error: undefined reference to 'art::operator<<(std::__1::basic_ostream >&,
art/runtime/class_linker.cc:2166: error: undefined reference to 'art::ObjectLock::ObjectLock(art::Thread*, art::Handle)'
art/runtime/class_linker.cc:2169: error: undefined reference to 'art::ObjectLock::WaitIgnoringInterrupts()'
clang++: error: linker command failed with exit code 1 (use -v to see invocation)
[ 2% 247/11700] Building with Jack: o...t_intermediates/with-local/classes.dex
ninja: build stopped: subcommand failed.
build/core/ninja.mk:148: recipe for target 'ninja_wrapper' failed
make: *** [ninja_wrapper] Error 1
make failed to build some targets (06:55 (mm:ss))
Getting this error when i run make -j16
Help me what to do !
Seems like some of your object files have been corrupted. I suggest you either clean the dir out/target/product/generic/obj/SHARED_LIBRARIES/libart_intermediates or/and clean ccache ccache -C
First advice is not to build with make -j16 as it takes too much ram.
For example compiling on 8 cores needs 64 GB of ram (tested on my server)
Try just compiling on a single core or at most on two. Another solution would be to call the command m instead of make to let it automatically find optimal settings on your hardware.
Another thing could be that you are missing some packages. Are you sure these packages are installed?
git-core gnupg flex bison build-essential zip curl zlib1g-dev gcc-multilib g++-multilib libc6-dev-i386 lib32ncurses5-dev x11proto-core-dev libx11-dev lib32z1-dev libgl1-mesa-dev libxml2-utils xsltproc unzip fontconfig
They could have different names for your distro (for example if you use Arch).
Also to note:
Compiling should take between 2.5 to 4 hours depending on your cpu performance. - It needs 100GB for downloading the android sources and 100-150GB for compilation.
You also need at least 12 GB or more to compile on a single core.
There is some method to compile it on 8GB of RAM by using swap space but this is extremely slow so not recommended.
My personal hardware and times:
Intel xeon e5-2680v3 8 core cpu
40 GB ram
Compile time on make -j1: 2 hours
I have a Raspberry Pi with the default, store-bought operating systems on it. I want to wipe the SD card clean so that I can put in a new operating system, but I want to preserve the original OS in a backup disc image. I planned to store it in a .bin file. The SD card has two partitions.
I used the following command to figure out which drive is the SD card.
sudo dmesg | tail
--output--
[ 2954.642182] sd 3:0:0:0: [sdb] Attached SCSI removable disk _
[ 2955.149750] EXT4-fs (sdb2): mounted filesystem with ordered data mode. Opts: (null)
I believe this tells me that it is under dev/sdb2, but I also tried dev/sdb, ~/dev/sdb and ~/dev/sdb2. I used the following command to create the image:
dd if="dev/sdb2" of="~/Desktop/Pi Backup/Pi.bin"
But when I try to do this it returns the error message
dd: opening `dev/sdb2': No such file or directory
I'm running Linux Mint, Cinnamon.
Any help is appreciated.
Instead of doing:
sudo dd if="/dev/mmcblk0p1" of="Pi_1.bin"
sudo dd if="/dev/mmcblk0p2" of="Pi_2.bin"
try:
sudo dd if="/dev/mmcblk0" of="Pi.bin"
p1 and p2 are the partitions in that device and you want to make an image of the entire device.
All devices are under /dev
dev is looking for dev under the current directory and ~/dev is looking for dev under your home directory.
/dev/sdb2 if the second partition, I would expect use have /dev/sdb1 (the first partition) too.
sudo dmesg | tail -30
will give you the last 30 lines, then you should be able to see the sdb1 too.
I'm on Mint 14 and I did mount and the SDcard shows as 2 partitions like below
/dev/mmcblk0p1 on /media/nig/3312-932F type vfat (rw,nosuid,nodev,uid=1000,gid=1000,shortname=mixed,dmask=0077,utf8=1,showexec,flush,uhelper=udisks2)
/dev/mmcblk0p2 on /media/nig/b7b5ddff-ddb4-48dd-84d2-dd47bf00564a type ext4 (rw,nosuid,nodev,uhelper=udisks2)
so I then did
sudo dd if="/dev/mmcblk0p1" of="Pi_1.bin"
sudo dd if="/dev/mmcblk0p2" of="Pi_2.bin"
seemed to work, not sure about restoring, not tried that yet