On Tue, 2020-12-08 at 08:48 +0000, Paul Barker wrote:
> On Mon, 7 Dec 2020 at 12:46, Richard Purdie
> <richard.pur...@linuxfoundation.org> wrote:
> > On Mon, 2020-12-07 at 12:04 +0000, Paul Barker wrote:
> > > On Mon, 7 Dec 2020 at 10:29, Paul Barker <pbar...@konsulko.com>
> > > wrote:
> > > To follow up some more: The entries in PSEUDO_IGNORE_PATHS are
> > > treated
> > > as string prefixes within pseudo. So if
> > > "/home/pbarker/Projects/Yocto/meta-linux-mainline" is added to
> > > the
> > > ignore list it will exclude not just
> > > "/home/pbarker/Projects/Yocto/meta-linux-mainline/build" but also
> > > ""/home/pbarker/Projects/Yocto/meta-linux-mainline-build".
> > >
> > > I wonder if some more of those entries should have trailing
> > > slashes.
> >
> > In most (all?) cases it was very deliberate FWIW...
>
> That does make sense. Ignoring "${COREBASE}/meta" will also ignore
> most layers unpacked or cloned within the poky directory as their
> names start with "meta". However that does miss layers if you use a
> different directory structure which is what Peter's patch addresses
> (though I'm still not sure if there's an actual build failure with
> some layers which it is intended to fix or if it's just to make
> things
> consistent).
>
> The issue comes when you clone a layer as the top-level of your
> working tree and build within that. That's how I work with
> meta-sancloud & meta-linux-mainline. It's also what happens if you
> build using the kas config in meta-raspberrypi. So it's not uncommon.
>
> Investigating why the layer directories are being ignored I found
> this
> commit added the ignore of "${COREBASE}/meta":
>
> commit e0cb6dd689a362d8433caa14cc5a9fdd5eb44923
> Author: Richard Purdie <richard.pur...@linuxfoundation.org>
> Date: Wed Oct 7 23:08:45 2020 +0100
>
> bitbake.conf: Extend PSEUDO_IGNORE_PATHS to ${COREBASE}/meta
>
> Unfortunately, .pyc files can be generated in meta/lib/oe
> which corrupt the pseudo
> database so we need to extend the ignore list to cover this
> as well.
>
> Signed-off-by: Richard Purdie <
> richard.pur...@linuxfoundation.org>
>
> Could we instead ignore "${LAYERDIR}/lib" for each layer?
>
> An alternative would be to detect the case where TOPDIR or TMPDIR is
> beneath an entry in BBLAYERS and handle that as a special case.
I think we'll have to look at alternatives, yes. Thanks for reminding
me this was for the pyc files, I'd forgotten the exact reason.
The reason I added meta was because at least in core we have multiple
python lib directories (there are also things in scripts/).
Peter: Which files were you having a problem with in your layers?
We may need to go back to requiring layers to set this up correctly
themselves where needed?
Cheers,
Richard
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#145381):
https://lists.openembedded.org/g/openembedded-core/message/145381
Mute This Topic: https://lists.openembedded.org/mt/78640454/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
