I made it a little further by adding --no-cache-dir to the pip3 install command.  That got rid fo the warning about not being able to access the .cache/pip.  However I still have the error:

| ERROR: torch-1.10.0-cp36-cp36m-linux_aarch64.whl is not a supported wheel on this platform.

Hi,
I'm attempting to create a recipe for pytorch for the TX2i using a wheel file obtained from here:

https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-10-now-available/72048

Here is what I have so far for the recipe:

```

DESCRIPTION = "NVIDIA's Python Torch"

HOMEPAGE = "https://nvidia.com"

LICENSE = "BSD-3-Clause"

LIC_FILES_CHKSUM = "file://../LICENSE;md5=91a5dfdaccf53b27488cb3a639e986d5"

inherit setuptools3

SRC_URI = "\

        file://torch-1.10.0-cp36-cp36m-linux_aarch64.whl \

        file://LICENSE \

"

COMPATIBLE_MACHINE = "jetson-tx2-devkit-tx2i"

PACKAGE_ARCH = "${MACHINE_ARCH}"

S = "${WORKDIR}/${PN}-${PV}"

do_configure() {

        :

}

do_compile() {

        :

}

do_install() {

        pip3 install ${WORKDIR}/torch-1.10.0-cp36-cp36m-linux_aarch64.whl

}

DEPENDS = "python3-pip-native"

```

When I build I get the following ambiguous error during do_install:

| WARNING: The directory '/home/ddbabich/.cache/pip' or its parent directory is not owned or is not writable by the current user. The cache has been disabled. Check the permissions and owner of that directory. If executing pip with sudo, you should use sudo's -H flag.

| ERROR: torch-1.10.0-cp36-cp36m-linux_aarch64.whl is not a supported wheel on this platform.

Hello,

I'm on dunfell release and building mender image with following wks file:
| + echo '### Contents of wks file ###'
| ### Contents of wks file ###
| + cat /home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/mender-sdimg.wks
| part --source rawcopy
--sourceparams="file=/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/uboot.env"
--ondisk "mmcblk0" --align 8192 --no-table
| part --source rootfs --rootfs-dir
/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/bootfs.image_sdimg
--ondisk "mmcblk0" --fstype=vfat --label boot --align 8192
--fixed-size 16 --active
| part --source rawcopy
--sourceparams="file=/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/deploy-core-image-minimal-image-complete/core-image-minimal-raspberrypi4.ext4"
--ondisk "mmcblk0" --align 8192 --fixed-size 172032k
| part --source rawcopy
--sourceparams="file=/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/deploy-core-image-minimal-image-complete/core-image-minimal-raspberrypi4.ext4"
--ondisk "mmcblk0" --align 8192 --fixed-size 172032k
| part swap --ondisk "mmcblk0" --fstype=swap --label swap --align 8192
--size 512
| part --source rawcopy
--sourceparams="file=/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/deploy-core-image-minimal-image-complete/core-image-minimal-raspberrypi4.dataimg"
--ondisk "mmcblk0" --align 8192 --fixed-size 128
|
| bootloader --ptable msdos
| + echo '### End of contents of wks file ###'
| ### End of contents of wks file ###

Without a swap partition image is built fine without any issue. When
swap partition is added to wks file I'm getting following error:

| INFO: Creating image(s)...
|
| Traceback (most recent call last):
| File "/home/marek/projects/kas-test/poky/scripts/wic", line 542, in <module>
| sys.exit(main(sys.argv[1:]))
| File "/home/marek/projects/kas-test/poky/scripts/wic", line 537, in main
| return hlp.invoke_subcommand(args, parser, hlp.wic_help_usage,
subcommands)
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/help.py",
line 83, in invoke_subcommand
| subcmd[0](args, usage)
| File "/home/marek/projects/kas-test/poky/scripts/wic", line 219,
in wic_create_subcommand
| engine.wic_create(wks_file, rootfs_dir, bootimg_dir, kernel_dir,
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/engine.py",
line 190, in wic_create
| plugin.do_create()
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/plugins/imager/direct.py",
line 96, in do_create
| self.create()
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/plugins/imager/direct.py",
line 180, in create
| self._image.prepare(self)
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/plugins/imager/direct.py",
line 354, in prepare
| part.prepare(imager, imager.workdir, imager.oe_builddir,
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/partition.py",
line 182, in prepare
| plugin.do_prepare_partition(self, srcparams_dict, creator,
| File "/home/marek/projects/kas-test/poky/scripts/lib/wic/plugins/source/rootfs.py",
line 112, in do_prepare_partition
| copyhardlinktree(orig_dir, new_rootfs)
| File "/home/marek/projects/kas-test/poky/meta/lib/oe/path.py",
line 121, in copyhardlinktree
| subprocess.check_output(cmd, shell=True, stderr=subprocess.STDOUT)
| File "/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/recipe-sysroot-native/usr/lib/python3.8/subprocess.py",
line 411, in check_output
| return run(*popenargs, stdout=PIPE, timeout=timeout, check=True,
| File "/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/recipe-sysroot-native/usr/lib/python3.8/subprocess.py",
line 512, in run
| raise CalledProcessError(retcode, process.args,
| subprocess.CalledProcessError: Command 'cd
/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/bootfs.image_sdimg;
find . -type d -print | tar --xattrs --xattrs-include='*' -cf - -S -C
/home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/bootfs.image_sdimg
-p --no-recursion --files-from - | tar --xattrs --xattrs-include='*'
-xhf - -C /home/marek/projects/kas-test/build/tmp/work/raspberrypi4-poky-linux-gnueabi/core-image-minimal/1.0-r0/deploy-core-image-minimal-image-complete/core-image-minimal-raspberrypi4-20211115212420-sdimg/tmp.wic.1oe9hwjw/rootfs2'
returned non-zero exit status 2.

I tried to debug it but didn't find anything obvious. I'm sure this
was working fine in previous releases. Any ideas?

Thanks and BR,

marek

--
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

Hi,

I was able to build my package in Zeus, I am now upgrading from Zeus
to Honister, the same package I could build in Zeus now failed in
Honister at do_package where I installed my libraries and symbolic
libraries links to my own RW locations:

ERROR: QA Issue: Files/directories were installed but not shipped

I purposely separate my application libraries and symbolic libraries
links from traditional system RO library location to my application RW
locations, I know how to handle it, it works perfectly built in Zeus,
I don't need the Yocto QA to police it, I tried to add an empty
do_package_qa() in my receipt, it does not work either. Are there any
ways to disable package QA?

Appreciate your advice either how to fix it or disable package QA.

Okay, I seem to have achieved this by adding:

ALLOW_EMPTY_${PN}="1"

To the adh-user-add recipe that everything is depending on (including the image recipe).

I'm a bit confused as to why that was such a hassle to achieve something I would have thought was a simple adding of a few users. Definitely would've appreciated a warning that the build system was attempting to add an empty package or something like that. I wouldn't think I'm on the fringe's of Yocto development here, but perhaps my use case is resulting in an 'empty package' being created via my image recipe that most other use-cases aren't triggering?

I would be very glad to hear any input despite having this resolved.

Hi,

I’m attempting to create a few service users for our software products. Initially I had inherited useradd in each recipe that required the user, then had a corresponding user added from within the image recipe. I know this is not ideal, however I have been unable to solve some issues I’m having so I’m hoping someone can help.

I'm on Dunfell branch.

I now have a recipe that creates two users, that all other recipes DEPENDS on. Here is the contents of that recipe (adh-user-add.bb):

FILESEXTRAPATHS_prepend := "${THISDIR}:"

SUMMARY = "Add adh and adhX user for other recipes"

LICENSE="CLOSED"

PV = "${ADH_ADH_SVN_REVISION}"

require require/revisions.inc

PACKAGE_ARCH = "${MACHINE_ARCH}"

inherit useradd

USERADD_PACKAGES = "${PN}"

USERADD_PARAM_${PN} = "--system --uid 420 -g adh adh; --system --uid 421 -g adhX adhX "

GROUPADD_PARAM_${PN} = "-g 420 adh; -g 421 adhX "

Each recipe that needs these users simply DEPENDS on it from within its own recipe. I understand that is the preferred way of doing this?

However, inside my image (development-image.bb) once its built, my /etc/passwd file does not contain these users at all. I have tried the following things:

1. IMAGE_INSTALL_append += “ adh-user-add”
2. DEPENDS += “ adh-user-add”
3. RDEPENDS += “ adh-user-add”

Each time I get a passwd file inside of adh-user-add/recipe-sysroots and development-image/recipe-sysroots that contains both users, however the passwd file on the device is missing those users entirely.

My image recipe looks like this:

SUMMARY = "Development image"

inherit core-image extrausers

require require/revisions.inc

PV = "${CUBE_IMAGE_VERSION}"

PR = "${CUBE_IMAGE_REVISION}"

IMAGE_FSTYPES_append += "wic wic.xz"

WKS_FILE = "sgc30cube.wks"

IMAGE_INSTALL = "packagegroup-core-boot ${CORE_IMAGE_EXTRA_INSTALL}  "

IMAGE_INSTALL_append = " \

                                                 *** packages removed ***

                                                 "

IMAGE_FEATURES_append = " ssh-server-dropbear"

MACHINE_FEATURES_remove = "kernel26 nc screen ext2 alsa gpu touchscreen qemu-usermode"

DEPENDS += "adh-user-add"

IMAGE_FEATURES_append = " debug-tweaks"

I feel like I'm going crazy trying to figure this out. I wish I could figure out how the final passwd file is generated. I've looked in some logs that I've seen suggested online, but have been unable to ascertain if there are any errors occurring.

Any help would be greatly appreciated,

Thanks,

Taylor W

Hi Joseph,
thank you for replying.
Shall I use the oe-pkgdata-util find-path this way?
oe-pkgdata-util  find-path /usr/bin/hexdump
Because I'm encountering an error
ERROR: Unable to find any package producing path /usr/bin/hexdump
So I think I am doing something wrong

(re-adding list)

Am 19.11.2021 um 10:11 schrieb lucapirozzi via lists.yoctoproject.org:

Hi Joseph,

thank you for replying.

Shall I use oe-pkgdata-util find-patch this way?

oe-pkgdata-util find-path /usr/bin/hexdump

Because it gives me an error
ERROR: Unable to find any package producing path /usr/bin/hexdump

You have to use the path you found hexdump at. Check in your running
target by using "which", or inspect the image contents.

https://github.com/agherzan/meta-raspberrypi/pull/951

Hi Joseph,

thank you for replying.

Shall I use oe-pkgdata-util find-patch this way?

oe-pkgdata-util  find-path /usr/bin/hexdump

Because it gives me an error
ERROR: Unable to find any package producing path /usr/bin/hexdump

So I think I am doing something wrong. Maybe the folder I am launching the script from?

Hi everyone,

I'm trying to remove some application, some of them packages and other being busybox parts.

I'm trying to remove hexdump from my image and I configured every busybox defconfig to have CONFIG_HEXDUMP and its REVERSE CONFIG to =n.

But I'm finding hexdump in my image filesystem nevertheless.

Any help?

Thank you and Regards.