Re: [PATCH 1/6] linux-yocto-dev: conditionally add coreutils to depends

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Tue, 2024-10-15 at 16:56 -0400, Bruce Ashfield wrote:
> On Tue, Oct 15, 2024 at 10:15 AM Richard Purdie
> <richard.purdie@linuxfoundation.org> wrote:
> > 
> > On Mon, 2024-10-14 at 09:45 -0400, bruce.ashfield@gmail.com wrote:
> > > From: Bruce Ashfield <bruce.ashfield@gmail.com>
> > > 
> > > 6.11+ uses truncate as part of the build. While truncate has been
> > > added to HOSTTOOLS in oe-core master, it still is possible to
> > > build a kernel recipe newer than 6.11+ on a release that doesn't
> > > have the HOSTTOOLS tweak.
> > > 
> > > We can test for truncate and condtionally add the dependency.
> > > 
> > > Signed-off-by: Bruce Ashfield <bruce.ashfield@gmail.com>
> > > ---
> > >  meta/recipes-kernel/linux/linux-yocto-dev.bb | 4 ++++
> > >  1 file changed, 4 insertions(+)
> > > 
> > > diff --git a/meta/recipes-kernel/linux/linux-yocto-dev.bb b/meta/recipes-kernel/linux/linux-yocto-dev.bb
> > > index 292897ce43..752d108bd8 100644
> > > --- a/meta/recipes-kernel/linux/linux-yocto-dev.bb
> > > +++ b/meta/recipes-kernel/linux/linux-yocto-dev.bb
> > > @@ -44,6 +44,10 @@ PACKAGECONFIG[dt-validation] = ",,python3-dtschema-native"
> > >  # we need the wrappers if validation isn't in the packageconfig
> > >  DEPENDS += "${@bb.utils.contains('PACKAGECONFIG', 'dt-validation', '', 'python3-dtschema-wrapper-native', d)}"
> > > 
> > > +# 6.11+ uses truncate in the build, so we need to either have it in host tools, or
> > > +# build it ourselves
> > > +DEPENDS += "${@bb.utils.contains('HOSTTOOLS', 'truncate', '', 'coreutils-native', d)}"
> > > +
> > >  COMPATIBLE_MACHINE = "^(qemuarmv5|qemuarm|qemuarm64|qemux86|qemuppc|qemumips|qemumips64|qemux86-64|qemuriscv32|qemuriscv64|qemuloongarch64)$"
> > > 
> > >  KERNEL_DEVICETREE:qemuarmv5 = "arm/versatile-pb.dtb"
> > 
> > This is for master so truncate should always be in HOSTTOOLS? I don't
> > understand why we need this there?
> 
> It came out of a discussion to do a reference for people that are building
> newer kernels on older releases (of which there seem to be quite a few).
> 
> So either we'd have to bacckport the hosttools addition (which is likely
> possible), or I could throw together that reference of how you'd build it
> conditionally.

I think the addition was backported for scarthgap so we're good there.
The main question would be around kirkstone where this commit could be
more appropriate if someone were building 6.11+.

I'm find with having this patch around as a proof of concept but I
don't think it should go into master...

Cheers,

Richard