On Fri, 6 Mar 2026, Richard Purdie via lists.openembedded.org wrote: > On Sun, 2026-03-01 at 19:38 -0500, Trevor Woerner via lists.openembedded.org > wrote: > > By convention, wic will look for a "wic" directory in the top-level > > of all BBLAYERS in order to find *wks files. oe-core was the only > > exception; it stored its *wks files in scripts/lib/wic/canned-wks. > > Take the *wks, *wks.in, *inc, and *cfg files that were in the canned- > > wks location and place them in a "wic" top-level directory in the oe- > > core layer, thereby following the convention of every other layer. > > This commit message is not strictly correct. The current code looks in > "wic" or "scripts/lib/wic/canned-wks" in every layer, not just oe-core. > > You're dropping the scripts/lib/wic/canned-wks search location for all > layers. > > I agree it is unlikely other layers are using that but the commit > message does need to be correct. That in turn will lead to better > migration notes for the release. > > The bigger issue is that if this isn't documented correctly, what else > are the patches doing which I haven't spotted? :/.
So far in a quick ad hoc search of layers I've got cloned locally I've found meta-sifive, meta-qcom, and meta-phytec that are using scripts/lib/wic/canned-wks. That suggests dropping support for that location does need to be clearly called out in the migration guide. Scott
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#232607): https://lists.openembedded.org/g/openembedded-core/message/232607 Mute This Topic: https://lists.openembedded.org/mt/118086346/21656 Group Owner: [email protected] Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
