Re: [PATCH 1/6] OMAPDSS: DISPC: cleanup cpu_is_xxxx checks

[Tomi Valkeinen <tomi.valkeinen@ti.com>]

[-- Attachment #1: Type: text/plain, Size: 1373 bytes --]

On Wed, 2012-08-08 at 18:31 +0530, Mahapatra, Chandrabhanu wrote:

> Yes, this sounds good, but I was just following general order of
> structure and function declarations, structures initializations
> followed by functions.

Yep, it's a good rule. But there are exceptions when we'll get bloated
code by following the rule =).

> > During runtime we only require one of these, others can be discarded.
> > This can be accomplished with the combination of "__initdata" for these,
> > and "__init" for dispc_init_features().
> >
> 
> The same also applies for all structures in dss_features.c. Just a
> thought that __init and __initdata should have also been used there.

I agree. We should also see what things we can move from dss_features.c
into dss.c, dispc.c, etc. Some things in dss_features are quite global,
but some are really used only in one place in one file.

Although this also makes me wonder, is it better to have all the
hardware version information for all DSS modules in one place, as we
tried with dss_features, or is it better to have the HW information in
the respective DSS submodule, as you're doing with this patch.

Both have benefits. Even though with the latter method there's no one
place to look for DSS HW feature differences, I think it'll still give
us cleaner code. So let's continue on this track.

 Tomi


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 836 bytes --]