On Wednesday, October 14, 2020 at 12:00:30 AM UTC+5:30 Jan Kiszka wrote: > On 13.10.20 20:22, Jan Kiszka wrote: > > On 07.10.20 12:49, vijaikumar....@gmail.com wrote: > >> > >> > >> On Tuesday, October 6, 2020 at 2:43:04 PM UTC+5:30 i...@radix50.net > wrote: > >> > >> On Mon, Oct 05, 2020 at 09:55:40PM -0700, vijaikumar....@gmail.com > >> wrote: > >> > > http://ci.isar-build.org:8080/job/isar_vkk_devel/85/ > >> > > >> > CI build succeeded. > >> > >> Thanks Vijai Kumar. I don't have objections to this one. I assume > >> Henning's > >> comments have been addressed in 07/08. We'll test on hikey and let > >> you know. > >> > >> > >> Yes. It is addressed. Sure let me know how it goes on Hikey. > >> > > > > My integration experiment revealed some issue, but I do not understand > > yet, where: If you build [1], which is the combination of your series > > with [2], I get > > > > | DEBUG: Executing shell function do_wic_image > > | INFO: Creating image(s)... > > | > > | ERROR: _exec_cmd: tune2fs -L fsbl1 > /tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/tmp.wic.9q0k2f75/tf-a-stm32mp157c-ev1.stm32.1 > returned '1' instead of 0 > > | output: /usr/sbin/tune2fs: Bad magic number in super-block while > trying to open > /tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/tmp.wic.9q0k2f75/tf-a-stm32mp157c-ev1.stm32.1 > > > | tune2fs 1.44.5 (15-Dec-2018) > > | > > | stat: missing operand > > | Try 'stat --help' for more information. > > | chown: invalid user: '/meta' > > | ls: cannot access > '/work/build/tmp/deploy/buildchroot-target/debian-buster-armhf/tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/*.direct': > No such file or directory > > | mv: missing destination file operand after > '/work/build/tmp/deploy/images/stm32mp15x/isar-image-base-debian-buster-stm32mp15x.wic.img' > > > | Try 'mv --help' for more information. > > | WARNING: exit code 1 from a shell command. > > | ERROR: Execution of > '/work/build/tmp/work/debian-buster-armhf/isar-image-base-stm32mp15x-wic-img/1.0-r0/temp/run.do_wic_image.17923' > failed with exit code 1: > > | INFO: Creating image(s)... > > | > > | ERROR: _exec_cmd: tune2fs -L fsbl1 > /tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/tmp.wic.9q0k2f75/tf-a-stm32mp157c-ev1.stm32.1 > returned '1' instead of 0 > > | output: /usr/sbin/tune2fs: Bad magic number in super-block while > trying to open > /tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/tmp.wic.9q0k2f75/tf-a-stm32mp157c-ev1.stm32.1 > > > | tune2fs 1.44.5 (15-Dec-2018) > > | > > | stat: missing operand > > | Try 'stat --help' for more information. > > | chown: invalid user: '/meta' > > | ls: cannot access > '/work/build/tmp/deploy/buildchroot-target/debian-buster-armhf/tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/*.direct': > No such file or directory > > | mv: missing destination file operand after > '/work/build/tmp/deploy/images/stm32mp15x/isar-image-base-debian-buster-stm32mp15x.wic.img' > > > | Try 'mv --help' for more information. > > | WARNING: exit code 1 from a shell command. > > | > > ERROR: Task > (mc:stm32mp15x-buster:/repo/meta-isar/recipes-core/images/isar-image-base.bb:do_wic_image) > failed with exit code '1' > > > > Any ideas where to dig? > > It must be related to the raw partitions we are defining for the > bootloader artifacts. But I just copied from > > > https://github.com/STMicroelectronics/meta-st-stm32mp/blob/dunfell/wic/sdcard-stm32mp157c-dk2-optee-1GB.wks.in > > It looked indeed strange to me that those raw partitions where declared > with "--fstype=ext4", but removing that does not solve the issue. > Hi Jan, I had a quick run through, looks assigning a label to the image is the issue. if you drop the --label and --fstype build is fine. On further investigation the file /tmp/tmp.z4IScQtkyS/isar-image-base-debian-buster-stm32mp15x.wic/tmp.wic.9q0k2f75/tf-a-stm32mp157c-ev1.stm32.1 is really just the data file and hence tune2fs rightly complains and returns a non-zero exit code. I am currently building using meta-st-stm32mp, to see if this actually works there. If so how. Thanks, Vijai Kumar K > Is your wic revision in line with or even ahead of dunfell? > > Jan > > > > > Jan > > > > [1] https://github.com/siemens/isar/commits/jan/wic-testing > > [2] https://groups.google.com/forum/#!topic/isar-users/ijj6mdBfb2w > > > > -- > Siemens AG, T RDA IOT > Corporate Competence Center Embedded Linux >