Re: [OE-core] [PATCH] wic: do not ignore ROOTFS_SIZE if the rootfs is modified

[Trevor Woerner <twoerner@gmail.com>]

Hi Mathieu,

Thanks for looking at my patch!

On Wed 2025-04-09 @ 04:29:57 PM, Mathieu Dubois-Briand wrote:
> On Tue Apr 8, 2025 at 10:57 PM CEST, Trevor Woerner via lists.openembedded.org wrote:
> I do have some build issue on the autobuilder:
> 
> 2025-04-09 11:14:21,897 - oe-selftest - INFO - wic.Wic.test_exclude_path_with_extra_space (subunit.RemotedTestCase)
> 2025-04-09 11:14:21,898 - oe-selftest - INFO -  ... FAIL
> ...
> [Wed Apr  9 11:01:08 2025] (1309377/1309395): 192.168.7.4:53040 UA:elfutils/0.192,Linux/x86_64,/ XFF: GET /buildid/da3aea1c8160034e1b0961a7f2b5ce5c5a95553b/debuginfo 200 237848 0+9ms
> ERROR: When reparsing /tmp/selftest-fetchw3a6djws/test.bb:do_checkuri, the basehash value changed from 84b9e23a83a764e7a5d8a96ee523199e2b1c604f4280a45030e248de0ff52962 to 4123da244f465feffecb2b4fd2b132b64851981ec2fb70bf9445df8b0ea6a69e. The metadata is not deterministic and this needs to be fixed.
> ERROR: The following commands may help:
> ERROR: $ bitbake test -cdo_checkuri -Snone
> ERROR: Then:
> ERROR: $ bitbake test -cdo_checkuri -Sprintdiff
> 
> ERROR: When reparsing /tmp/selftest-fetchw3a6djws/test.bb:do_fetch, the basehash value changed from 6d55d6743729b615749dc016857d7e5f9c884a8b92f0d57f68e743f4910333d3 to fa7d6ea0563b60362418d2a6c5e41a6684183d01ec578b65abd0c64a9cfba7cf. The metadata is not deterministic and this needs to be fixed.
> ERROR: The following commands may help:
> ERROR: $ bitbake test -cdo_fetch -Snone
> ERROR: Then:
> ERROR: $ bitbake test -cdo_fetch -Sprintdiff
> 
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/35/builds/1329

Strange. This *sounds* like what happens when I update my layers while a build
is running. I wouldn't expect the AB to do something like that in the middle
of a build?

> I tried to reproduce this locally, and reproduced a fail with
> "oe-selftest -r wic.Wic.test_exclude_path_with_extra_space", but with a
> different error:
> 
> | INFO: Creating image(s)...
> |
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/imager/direct.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/bootimg-biosplusefi.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/empty.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/bootimg-efi.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/bootimg-pcbios.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/isoimage-isohybrid.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/bootimg-partition.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/rootfs.py
> | DEBUG: loading plugin module /home/mdubois-briand/swat/oe-sefltest/poky-contrib/scripts/lib/wic/plugins/source/rawcopy.py
> | DEBUG: _exec_cmd: install -d /home/mdubois-briand/swat/oe-sefltest/poky-contrib/build-st/tmp/work/qemux86_64-poky-linux/core-image-minimal/1.0/tmp-wic/boot.1
> | DEBUG: ['install', '-d', '/home/mdubois-briand/swat/oe-sefltest/poky-contrib/build-st/tmp/work/qemux86_64-poky-linux/core-image-minimal/1.0/tmp-wic/boot.1']
> | DEBUG: _exec_cmd: output for install -d /home/mdubois-briand/swat/oe-sefltest/poky-contrib/build-st/tmp/work/qemux86_64-poky-linux/core-image-minimal/1.0/tmp-wic/boot.1 (rc = 0):
> | ERROR: No boot files defined, IMAGE_BOOT_FILES unset for entry #1
> 
> So maybe I'm missing something here.
> 
> The exact branch I used for this build should be available in
> https://web.git.yoctoproject.org/poky-ci-archive/tag/?h=autobuilder.yoctoproject.org/valkyrie/a-full-1372,
> but the mirror seems out of sync. I hope it will be there a bit later
> today.
> 
> Can you have a look at this, please?

This is my first oe-selftest, so I'm not entirely surprised. It succeeded for
me, I thought I had it figured out, but I've probably done something wrong in
my testing.

Thanks, I'll take another look at this with your branch.

Best regards,
	Trevor