Re: build failure in amd64-current in fss.c

[Eduardo Horvath <eeh%NetBSD.org@localhost>]

On Wed, 30 Nov 2011, Manuel Bouyer wrote:

> On Wed, Nov 30, 2011 at 09:59:48AM +0100, Martin Husemann wrote:
> > On Tue, Nov 29, 2011 at 10:27:37PM -0800, Hisashi T Fujinaka wrote:
> > > /usr/src/sys/dev/fss.c: In function 'fss_ioctl':
> > > /usr/src/sys/dev/fss.c:367:2: error: duplicate case value
> > > /usr/src/sys/dev/fss.c:341:2: error: previously used here
> > > 
> > > Seems like case statements can't tell the difference between FSSIOCGET50
> > > and FSSIOCSET. Is that a compiler bug?
> > 
> > I guess the size of the ioctl structure did not change for 64bit ports - 
> > duh!
> 
> Ops, yes, time_t was already 64bits for LP64 ports.
> 
> > Manuel, is compat needed in this case at all?
> 
> No, if the structures have the same size, there's no compat issues.
> 
> > If not, we could ifndef _LP64
> > that case...
> 
> I guess there's no much other choise ... I fixed it this way.

I think it's probably more complicated than that.  sparc64 used to have a 
32-bit time_t before everything switched to 64 bits.

Eduardo