Re: [PAE] Initial patches for kvm(3) support

[Jean-Yves Migeon <jeanyves.migeon%free.fr@localhost>]

On Tue, 7 Sep 2010 18:34:22 +0100, David Laight <david%l8s.co.uk@localhost> 
wrote:
> On Mon, Sep 06, 2010 at 10:24:55PM +0200, Jean-Yves Migeon wrote:
>> 
>> Add to the fact that kvm code mixes pre-ANSI and
>> ANSIfied declarations (KNF), the patch is bigger than expected (I
>> deliberatly ANSIfied most of the code -- if not needed, I won't commit
>> this part). But I suppose __P() can go, nevertheless.
> 
> Somewhere I have two sed (I think) scripts that make a resonable
> job of removing __P() and converting K&R functions to ANSI.
> They are certainly less error-prone than trying to do it by hand!
> (They work when the arguments in the K&R defn are defined in the wrong
> order.)
> 
> Any KNF changes should probably be done first, and as separate commits
> with no other changes.

Of course. This patch creates churn in kvm(3), and I wanted opinions
first.

The ANSIfication is "needed" for coccinelle; I discovered that it is
having a hard time parsing pre-ANSI code (:

> Also I think you should be using uint32_t and uint64_t not long and
> long long.

Oops, that should have been the case -- thanks for pointing this out.

-- 
Jean-Yves Migeon
jeanyves.migeon%free.fr@localhost