Re: [PATCH] rootfs_ipk: respect ONLINE_PACKAGE_MANAGEMENT

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

On Wed, 2011-05-18 at 16:37 +0100, Phil Blundell wrote:
> On Tue, 2011-05-17 at 17:22 +0100, Richard Purdie wrote:
> > On Tue, 2011-05-17 at 15:50 +0100, Phil Blundell wrote:
> > > I guess I could teach remove_packaging_data_files() to not create the
> > > empty directory if O_P_M=="none".  Would you be happier with that?
> > 
> > Mostly. The question remains what happens if there are postinstalls that
> > are expecting to run on the device. I'm guessing that is why the mkdir
> > is there since at present we don't want to break the postinstalls.
> 
> The comment says something about ensuring that the lock directory
> exists, presumably for future invocations of opkg.  I guess this is to
> support O_P_M="add" kind of semantics.  
> 
> As far as I know there's no other reason that postinsts should
> expect/require the opkg directory to exist.  So I think it should
> probably be safe just to delete it if O_P_M="none".

Let me put this another way:

We have postinstalls that require to run on the target device. If
they're present, we need opkg there to run them, or at least some kind
of script to handle that and I think both cases have a requirement on
that directory (the rpm backend has a suitable script).

So the current patches which just don't create it don't really address
the problem. At the very least the image generation should error if
these types of postinstall are present which I guess is my real concern.

Cheers,

Richard