> > I agree, this would be a good change to make for long-term > maintainability. LGTM otherwise.
I have tried to implement such a check as well, and while this check is fine and works, the test of a patch with patchtest containing a variable from BB_RENAMED_VARIABLES that is not a pretest, i.e. the patch is already merged, the tinfoil setup and the subsequent parsing of the recipes while raise an issue for this test. e.g. ERROR: Variable SDK_LOCAL_CONF_BLACKLIST has been renamed to ESDK_LOCALCONF_REMOVE ... So on the one hand the issues is already cached by other parts of the buildsystem and on the other hand this will cause tests not related to renamed variables to fail. It was suggested on bugzilla by Trevor to merge this original test and open an issue in bugzilla to address the question of BB_RENAMED_VARIABLES separately. See [1] If you do agree, please tell me if a rebased version of the patch would be needed or not. - Cheers, Simone [1] https://bugzilla.yoctoproject.org/show_bug.cgi?id=15311#c5
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#192677): https://lists.openembedded.org/g/openembedded-core/message/192677 Mute This Topic: https://lists.openembedded.org/mt/103112424/21656 Group Owner: [email protected] Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
