[PATCH weston 4/9] ivi-layout: abort without controller_module_init

[Pekka Paalanen]

On Mon, 06 Apr 2015 15:46:51 -0500
Derek Foreman <derekf at osg.samsung.com> wrote:

> On 02/04/15 07:11 AM, Pekka Paalanen wrote:
> > On Thu, 2 Apr 2015 01:04:12 +0000
> > "Tanibata, Nobuhiko (ADITJ/SWG)" <ntanibata at jp.adit-jv.com> wrote:
> > 
> >>
> >>
> >>> -----Original Message-----
> >>> From: wayland-devel
> >>> [mailto:wayland-devel-bounces at lists.freedesktop.org] On Behalf Of Pekka
> >>> Paalanen
> >>> Sent: Monday, March 30, 2015 6:21 PM
> >>> To: wayland-devel at lists.freedesktop.org
> >>> Cc: Pekka Paalanen; Nobuhiko Tanibata
> >>> Subject: [PATCH weston 4/9] ivi-layout: abort without
> >>> controller_module_init
> >>>
> >>> From: Pekka Paalanen <pekka.paalanen at collabora.co.uk>
> >>>
> >>> When loading a controller module, if we do not find a controller_module_init
> >>> symbol, return failure to the caller instead of ignoring the failure.
> >>>
> >>> Signed-off-by: Pekka Paalanen <pekka.paalanen at collabora.co.uk>
> >> [ntanibata] 
> >> Tested-by: Nobuhiko Tanibata <NOBUHIKO_TANIBATA at xddp.denso.co.jp>
> >>
> > 
> > I've pushed two more patches from this series:
> > ivi-shell: add cmdline option for controller module
> > ivi-layout: abort without controller_module_init
> >    4ac06ff..97246c0  master -> master
> > 
> > Patches 5-9 are remaining, now in force-pushed:
> > http://cgit.collabora.com/git/user/pq/weston.git/log/?h=ivi-test-5
> > 
> > The force-push is a simple rebase to account for added tags in commit
> > messages. No code changes.
> 
> Just finished looking at 5-9...  Just a couple of questions:
> 
> in 8, What does 0xffc01200 mean?

It's a random value I pulled from my hat. It only has to be a ivi-id
segment that is not used by anything else. I'm not aware of any
registry for those, and we can change this value here at any time if
needed, I believe.

So, it's just whatever works, and different enough.

I'll add a comment to that.

> in 9 you have copies of pretty much the same assert macros from 8, do
> you think they're useful enough to put somewhere shared instead of
> duplicating them?

The difference is that one eventually uses wl_resource_post_error() and
the other weston_compositor_exit_with_code(). I suppose we could share
just the macros by assuming the fail-function to be called has a
particular name... but in this case duplicating those few lines avoid
assuming any particular function names.

I would find it a little awkward to see a iassert(foo), then have to go
look into header what it does, and find multiple different
implementations of the function it calls. In this case I think
duplicating is better than sharing.

> 
> Reviewed-by: Derek Foreman <derekf at osg.samsung.com>

Thanks!
pq