Re: Bug: wic integration: Problem with wic-tools

[Henning Schild <henning.schild@siemens.com>]

Am Tue, 15 May 2018 12:47:41 +0200
schrieb "[ext] Andreas Reichel" <andreas.reichel.ext@siemens.com>:

> Hi,
> 
> my last attempt in building with current next results in:
> 
> ------------------
> ERROR: mc:xx-stretch:xx-product-1.0-r0 do_wic_image: Function failed:
> do_wic_image (log file is located
> at /build/tmp/work/xx-amd64/xx-product/temp/log.do_wic_image.8529)
> ERROR: Logfile of failure stored
> in: /build/tmp/work/xx-amd64/xx-product/temp/log.do_wic_image.8529
> Log data follows: | DEBUG: Executing shell function do_wic_image |
> INFO: Building wic-tools... | | Traceback (most recent call last):
> |   File "/em/isar/scripts/wic", line 326, in <module>
> |     sys.exit(main(sys.argv[1:]))
> |   File "/em/isar/scripts/wic", line 321, in main
> |     return hlp.invoke_subcommand(args, parser, hlp.wic_help_usage,
> subcommands) |   File "/em/isar/scripts/lib/wic/help.py", line 97, in
> invoke_subcommand |     subcommands.get(args[0],
> subcommand_error)[0](args[1:], usage) |   File
> "/em/isar/scripts/wic", line 201, in wic_create_subcommand |     if
> bitbake_main(BitBakeConfigParameters("bitbake wic-tools".split()), |
> NameError: name 'BitBakeConfigParameters' is not defined | WARNING:
> exit code 1 from a shell command. | ERROR: Function failed:
> do_wic_image (log file is located
> at /build/tmp/work/xx-amd64/xx-product/temp/log.do_wic_image.8529)
> ERROR: Task
> (multiconfig:xx-stretch:/em/isar-xx/recipes-core/images/xx-product.bb:do_wic_image)
> failed with exit code '1' NOTE: Tasks Summary: Attempted 86 tasks of
> which 0 didn't need to be rerun and 1 failed.
> 
> Summary: 1 task failed:
>   multiconfig:xx-stretch:/em/isar-xx/recipes-core/images/xx-product.bb:do_wic_image
> Summary: There were 24 WARNING messages shown.
> Summary: There was 1 ERROR message shown, returning a non-zero exit
> code.
> 
> ------------------
> 
> Seems there is an incompatibility between bitbake and wic versions in
> use?

This is wic trying to call bitbake, which is something that should
never happen if wic was called by bitbake. Even if it worked you would
fail running into the bitbake lock.

Read the wic code and you will see that you somehow ended up with an
invalid "native_sysroot" bringing you to where you run into the problem.

Henning

> I am investigating this, if nobody else already has an idea why this
> happens.
> 
> kr
> Andreas