I think we should think about just fixing the problem rather than failing in a sanity check, given how easy it is to fix. It knows what needs doing, just do it, possibly with a warning or informational message about it.
Patch sent to oe-core mailing list
Alex, I think this was merged and fixed up, so we can probably resolve this, right?
Resolving as it has been merged and apparently no more performance issues or such
Reopening, because last part wasn't merged yet, I've sent it again: http://patchwork.openembedded.org/patch/103983/
Changing target milestone as bug was reopened pending integration of another change
Alex, will you send v2 with the change suggested by Chris?
The last part was merged, without my suggested change, see https://github.com/openembedded/openembedded-core/commit/95daa7c26d119ae11f1835ce0e895bbd4a3e9605.
Have a patch ready with suggested change, but haven't tested. Will resend to list this week
(In reply to comment #8) It has now been sent
(In reply to comment #9) > (In reply to comment #8) > > It has now been sent http://lists.openembedded.org/pipermail/openembedded-core/2015-October/111676.html
merged, closing
Verified on master efa08813d749c61303ebbadcf1a9f99affbdaed9