On Thu, 2020-04-16 at 20:39 +0200, Quentin Schulz wrote:
> Yocto bug: https://bugzilla.yoctoproject.org/show_bug.cgi?id=10130
> 
> Move logic checking that all packages licenses are only a subset of
> recipe licenses from base.bbclass to the package.bbclass so that it's
> evaluated only once, during packaging.
> 
> As explained in the linked bugzilla entry, if a package license is not
> part of the recipe license, the warning message gets shown an
> unreasonable amount of time because it's evaluated every time a recipe
> is parsed.
> 
> Signed-off-by: Quentin Schulz <[email protected]>
> ---
> 
> - We may not want this patch if we find it important that this warning is
> displayed at parsing time and not only after packaging has been done
> (which means dryruns aren't going to detect this issue anymore).
> - Do we want separate patches for each class?
> - The original bug report isn't very explicit, so hopefully I'm actually
> fixing it :) adding Paul, the reporter, in Cc so he can confirm.
> 
>  meta/classes/base.bbclass    | 13 -------------
>  meta/classes/package.bbclass | 13 ++++++++++++-
>  2 files changed, 12 insertions(+), 14 deletions(-)
> 
> diff --git a/meta/classes/base.bbclass b/meta/classes/base.bbclass
> index 45f9435fd8..7aa2e144eb 100644
> --- a/meta/classes/base.bbclass
> +++ b/meta/classes/base.bbclass
> @@ -584,19 +584,6 @@ python () {
>                          bb.debug(1, "Skipping recipe %s because of 
> incompatible license(s): %s" % (pn, ' '.join(incompatible_lic)))
>                          raise bb.parse.SkipRecipe("it has incompatible 
> license(s): %s" % ' '.join(incompatible_lic))
>  
> -        # Try to verify per-package (LICENSE_<pkg>) values. LICENSE should 
> be a
> -        # superset of all per-package licenses. We do not do advanced 
> (pattern)
> -        # matching of license expressions - just check that all license 
> strings
> -        # in LICENSE_<pkg> are found in LICENSE.
> -        license_set = oe.license.list_licenses(license)
> -        for pkg in d.getVar('PACKAGES').split():
> -            pkg_license = d.getVar('LICENSE_' + pkg)
> -            if pkg_license:
> -                unlisted = oe.license.list_licenses(pkg_license) - 
> license_set
> -                if unlisted:
> -                    bb.warn("LICENSE_%s includes licenses (%s) that are not "
> -                            "listed in LICENSE" % (pkg, ' '.join(unlisted)))
> -
>      needsrcrev = False
>      srcuri = d.getVar('SRC_URI')
>      for uri in srcuri.split():
> diff --git a/meta/classes/package.bbclass b/meta/classes/package.bbclass
> index d4c6a90e84..1e87108e43 100644
> --- a/meta/classes/package.bbclass
> +++ b/meta/classes/package.bbclass
> @@ -1322,9 +1322,20 @@ python populate_packages () {
>      os.umask(oldumask)
>      os.chdir(workdir)
>  
> -    # Handle LICENSE_EXCLUSION
> +    license_set = oe.license.list_licenses(d.getVar('LICENSE'))
>      package_list = []
>      for pkg in packages:
> +        # Try to verify per-package (LICENSE_<pkg>) values. LICENSE should 
> be a
> +        # superset of all per-package licenses. We do not do advanced 
> (pattern)
> +        # matching of license expressions - just check that all license 
> strings
> +        # in LICENSE_<pkg> are found in LICENSE.
> +        pkg_license = d.getVar('LICENSE_' + pkg)
> +        if pkg_license:
> +            unlisted = oe.license.list_licenses(pkg_license) - license_set
> +            if unlisted:
> +                bb.warn("LICENSE_%s includes licenses (%s) that are not "
> +                        "listed in LICENSE" % (pkg, ' '.join(unlisted)))
> +        # Handle LICENSE_EXCLUSION
>          licenses = d.getVar('LICENSE_EXCLUSION-' + pkg)
>          if licenses:
>              msg = "Excluding %s from packaging as it has incompatible 
> license(s): %s" % (pkg, licenses)

I think that this check really belongs in insane.bbclass as its a
validation checking issue. I'd love to move this away from parse time
and to runtime only! I do think its something for the insane class
framework where we can configure it to warn or error for example.

Cheers,

Richard



-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#137259): 
https://lists.openembedded.org/g/openembedded-core/message/137259
Mute This Topic: https://lists.openembedded.org/mt/73061286/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub  
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to