On Wed, 2023-06-21 at 10:58 +0200, Martin Jansa wrote: > There is maybe unexpected change in behavior caused by this. > > Before it was triggering error only for .patch files in oe-core, > while now it seems to apply for all recipes from oe-core layer > (independently from where the .patch file was added), so e.g. meta- > raspberrypi is now causing bluez5 do_patch to fail: > > bluez5-5.66: Missing Upstream-Status in patch > meta-raspberrypi/recipes-connectivity/bluez5/bluez5/0001-bcm43xx-Add-bcm43xx-3wire-variant.patch > Please add according to > https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines#Patch_Header_Recommendations:_Upstream-Status > . [patch-status] > bluez5-5.66: Missing Upstream-Status in patch > meta-raspberrypi/recipes-connectivity/bluez5/bluez5/0002-bcm43xx-The-UART-speed-must-be-reset-after-the-firmw.patch > Please add according to > https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines#Patch_Header_Recommendations:_Upstream-Status > . [patch-status] > bluez5-5.66: Missing Upstream-Status in patch > meta-raspberrypi/recipes-connectivity/bluez5/bluez5/0003-Increase-firmware-load-timeout-to-30s.patch > Please add according to > https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines#Patch_Header_Recommendations:_Upstream-Status > . [patch-status] > bluez5-5.66: Missing Upstream-Status in patch > meta-raspberrypi/recipes-connectivity/bluez5/bluez5/0004-Move-the-hciattach-firmware-into-lib-firmware.patch > Please add according to > https://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines#Patch_Header_Recommendations:_Upstream-Status > . [patch-status]
I did mention this in the docs update patch to the migration guide I sent yesterday. It wasn't entirely expected but is a side effect from the way the layer override works as it is based on the recipe itself, not where the patches may reside. We noticed it on the autobuilder for meta-agl for example and it also broke from of the selftests. Overall, I don't think this should be a huge issue once we update the hopefully small number of patches to core recipes in layers and the layer mechanism is an overall win in the end? You could work around this with things like: ERROR_QA:remove:pn-bluez = "patch-status" to, not that I'd recommend it. Cheers, Richard
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#183178): https://lists.openembedded.org/g/openembedded-core/message/183178 Mute This Topic: https://lists.openembedded.org/mt/99623855/21656 Group Owner: openembedded-core+ow...@lists.openembedded.org Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
