On Sun, Dec 09, 2012 at 11:27:14AM -0600, David Young wrote:
> On Sat, Dec 08, 2012 at 12:36:31PM +0000, KIYOHARA Takashi wrote:
> > Module Name: src
> > Committed By: kiyohara
> > Date: Sat Dec 8 12:36:31 UTC 2012
> >
> > Modified Files:
> > src/sys/arch/i386/i386: db_interface.c trap.c
> > src/sys/arch/x86/x86: bus_dma.c cpu.c platform.c
> >
> > Log Message:
> > #ifdef - #endif-ed. NMCA, NISA, NNPX, NIOAPIC, LAPIC, MPBIOS and
> > MULTIPROCESSOR.
>
> Please, add no new #ifdefs. Let config(1) and the linker configure the
> kernel.
Indeed.
I doubt that any modern systems are so short of memory that removing
the code fragments from switch statements (because the event can't
happen) matters.
One good reason for removing code is that is isn't likely to be used
on anything anyone might try to run in on, has probably bitrotted
(because it hasn't been run for years), might be a security risk,
and complicates the other code paths.
Which is, more or less, why i386 support was removed.
There is probably other 'support' for systems that no one has any more,
and really aren't of enough historical interest to keep supporting.
David
--
David Laight: da...@l8s.co.uk
