On Mon, 2021-09-20 at 09:32 -0700, Khem Raj wrote: > > On 9/20/21 5:51 AM, Richard Purdie wrote: > > On Mon, 2021-09-20 at 13:46 +0100, Richard Purdie via lists.openembedded.org > > wrote: > > > The 'meta' is old and not very useful. It empties PACKAGES and creates > > > recursive do_build dependencies. > > > > > > We shouldn't need such recursive build dependencies any more so simplify > > > the code. > > > > > > This does cause behaviour changes as some dependencies are no longer > > > built. > > > It did show up issues with qemu-helper-native handling for example but > > > those > > > issues look like real races and the underlying dependency chains were not > > > correct. > > > > > > Signed-off-by: Richard Purdie <richard.pur...@linuxfoundation.org> > > > --- > > > meta/classes/populate_sdk_base.bbclass | 4 +++- > > > 1 file changed, 3 insertions(+), 1 deletion(-) > > > > > > diff --git a/meta/classes/populate_sdk_base.bbclass > > > b/meta/classes/populate_sdk_base.bbclass > > > index ccfe2232895..49e166e697d 100644 > > > --- a/meta/classes/populate_sdk_base.bbclass > > > +++ b/meta/classes/populate_sdk_base.bbclass > > > @@ -1,4 +1,6 @@ > > > -inherit meta image-postinst-intercepts image-artifact-names > > > +PACKAGES = "" > > > + > > > +inherit image-postinst-intercepts image-artifact-names > > > > > > # Wildcards specifying complementary packages to install for every > > > package that has been explicitly > > > # installed into the rootfs > > > > This change is probably high impact in that it changes behaviour the most > > of any > > of the changes. It does make image builds use a lot less sstate and in > > theory, > > if there are issues with failures from this, there are probably existing > > races. > > > > It does also have CI impact as it changes the things which are built as not > > all > > build tasks now run. Open to opinions on whether we should do this now or > > not. > > > > I definitely like this change and want it in upcoming LTS in 2022, not > comfortable with rushing it into this release. It will cause fair bit of > work for rest of layers in ecosystem.
The question is whether this is a real world issue for anyone with CI out there. I've been looking at the YP autobuilder and I'm not sure it is. Yes, it does change things but not really in a way I'm too worried about. There is some breakage in this series but it is mainly from the layer.conf change. So the question is whether anyone does see real issues in their CI from this? There is an easy change if it does cause an issue which I've now documented in the commit log as I replied to Joshua. Cheers, Richard
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#156197): https://lists.openembedded.org/g/openembedded-core/message/156197 Mute This Topic: https://lists.openembedded.org/mt/85739675/21656 Group Owner: openembedded-core+ow...@lists.openembedded.org Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
