Re: CVS commit: src/sys/arch/amd64/amd64

[Maxime Villard <max%m00nbsd.net@localhost>]

Le 22/02/2018 à 17:31, Christos Zoulas a écrit :
> In article <7f4de63c-e782-14e6-5554-9b9d23471e6d%m00nbsd.net@localhost>,
> Maxime Villard  <max%m00nbsd.net@localhost> wrote:
> > Le 22/02/2018 à 15:54, Christos Zoulas a écrit :
> > > In article <20180222140848.70E95F94A%cvs.NetBSD.org@localhost>,
> > > Martin Husemann <source-changes-d%NetBSD.org@localhost> wrote:
> > > > -=-=-=-=-=-
> > > >
> > > > Module Name:	src
> > > > Committed By:	martin
> > > > Date:		Thu Feb 22 14:08:48 UTC 2018
> > > >
> > > > Modified Files:
> > > > 	src/sys/arch/amd64/amd64: locore.S
> > > >
> > > > Log Message:
> > > > Protect the SVS part of SYSCALL_ENTRY by #ifdef SVS to make non-SVS
> > > > kernels compile again.
> > >
> > > The combination of "#ifdef" and ".if" makes the code more horrific.
> > > Can we use one and not the other? Preferrably "#ifdef" since we already
> > > use it extensively?
> >
> > In this case the ifdef just had to be put around the declaration.
> >
> > You can't replace .if by #ifdef, there are two SYSCALL_ENTRY declarations,
> > and we give a different argument depending on whether we want the SVS code
> > to be in the macro or not.
>
> The question is do we want to keep using both cpp and assembly macros.

Why wouldn't we? I don't see the problem.

> The use of assembly macros is recent, the cpp one has always been there.
> I.e. until recently we were not using .macro or .if, now we are.

... And? There is only one place where we use .if instead of #if, because there
is a good reason for doing so. It doesn't occur to me we need to replace all
the other #ifs by .ifs as a result.

Maxime