Re: standard runtime for (possible/hopeful) 64bit kernels

[Paul Koning <Paul_Koning%dell.com@localhost>]

>>>>> "Toru" == Toru Nishimura <locore64%alkyltechnology.com@localhost> writes:

 >> That sounds right.  N64 is the harder job because you have to fix
 >> all the code that assumes sizeof(void *) == sizeof(int),

 Toru> I'd say correct NetBSD apps do not have such the evil
 Toru> assumption since the day when NetBSD/alpha was roled out to
 Toru> public ...

That may be, but not all apps are correct in this respect...

 >> Given that lots of embedded apps make little use of the FPU (even
 >> if there is one), I like O64 because it's much less work.

 Toru> Um I hesitate to say in this way, it's a matter for how to
 Toru> reduce porting burden of incorrectly written SW from 32bit
 Toru> world, isn't it?

No.  O64 and N32 are identical as far as the pointer vs. integer size
issue I mentioned -- in both cases, sizeof(int) == sizeof (long) ==
sizeof (void *).  Only N64 (and O64 with -mlong64) have the 64 bit
long and pointer.

By "less work" I meant a lot fewer files in the NetBSD code base to
change.  O64 touches only a fraction of the assembly language files,
while N32 or N64 touches most of them.

      paul