[OE-core] [PATCH] rootfs_ipk.bbclass: Ensure bad recommendations persist in the status file

[Richard Purdie]

On Tue, 2014-08-26 at 02:51 +0200, Andreas Oberritter wrote:
> Hello Richard,
> 
> sorry for digging out such an old message, but please see below.
> 
> On 12.11.2011 09:36, Richard Purdie wrote:
> > Currently bad recommendations are added to the status file with status
> > "ok". After a single opkg command, whilst it will ignore the recommendation,
> > the status changes to "installed" even if the recommended package was not
> > installed. Whilst this is likely a glitch in opkg's logic, the correct
> > way to persist the information in the status file is to set the status
> > to "hold" as deinstall packages with that status remain. With this change
> > the bad recommendations persist accross multiple opkg runs and the system
> > behaves as expected.
> > 
> > [YOCTO #1758]
> > 
> > Signed-off-by: Richard Purdie <richard.purdie at linuxfoundation.org>
> > ---
> > diff --git a/meta/classes/rootfs_ipk.bbclass b/meta/classes/rootfs_ipk.bbclass
> > index 3b4c392..4a5a2dd 100644
> > --- a/meta/classes/rootfs_ipk.bbclass
> > +++ b/meta/classes/rootfs_ipk.bbclass
> > @@ -44,7 +44,7 @@ fakeroot rootfs_ipk_do_rootfs () {
> >  		pkginfo="`opkg-cl ${IPKG_ARGS} info $i`"
> >  		if [ ! -z "$pkginfo" ]; then
> >  			echo "$pkginfo" | grep -e '^Package:' -e '^Architecture:' -e '^Version:' >> $STATUS
> > -			echo "Status: deinstall ok not-installed" >> $STATUS
> > +			echo "Status: deinstall hold not-installed" >> $STATUS
> 
> I just implemented bad recommendations for deb (in dora) and by looking
> at dpkg's source code I realized that "hold" is only a valid value for
> the first field. Maybe you meant to write "Status: hold ok
> not-installed" instead? Valid values for the second field are ok and
> reinstreq.
> 
> The code above doesn't exist in master anymore, but the status line
> hasn't changed in the newer version implemented in python.

I'm not sure what opkg accepts or what is correct. Paul, cc'd may have a
better idea though. What we have seems to work, that doesn't necessarily
mean its correct though...

Cheers,

Richard