Re: CVS commit: src/sys/arch

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

Le 15/06/2017 à 14:15, Joerg Sonnenberger a écrit :
> On Thu, Jun 15, 2017 at 07:29:22AM +0200, Maxime Villard wrote:
> > Le 14/06/2017 à 15:34, Joerg Sonnenberger a écrit :
> > > On Wed, Jun 14, 2017 at 12:27:24PM +0000, Maxime Villard wrote:
> > > > Module Name:	src
> > > > Committed By:	maxv
> > > > Date:		Wed Jun 14 12:27:24 UTC 2017
> > > >
> > > > Modified Files:
> > > > 	src/sys/arch/amd64/include: param.h
> > > > 	src/sys/arch/i386/include: param.h
> > > > 	src/sys/arch/x86/x86: x86_machdep.c
> > > >
> > > > Log Message:
> > > > Define MAXPHYSMEM globally.
> > >
> > > Given that we have an option called PHYSMEM_MAX_SIZE, I find the name a
> > > bit unfortunate.
> >
> > Indeed, but they don't mean the same thing: MAXPHYSMEM is the limit of
> > the implementation, and PHYSMEM_MAX_SIZE is user-set. I took MAXPHYSMEM
> > because there was already MAXIOMEM nearby.
>
> Yes, they have slightly different meaning. But that doesn't make the
> name less confusion. If there is no MI reason, maybe PHYSMEM_PMAP_LIMIT?
> That would reflect *why* the limit exists as well.

In fact, it would be nice to have a MD name for this, and an associated
man page to make things clear on each port. Otherwise I'm not sure the
"pmap" in the name is appropriate; actually I don't know where this 16TB
comes from, but as far as I can tell, it is not related to pmap.