This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: Possible bug in sysdeps/unix/sysv/linux/powerpc/mmap64.c
On Sun, May 20, 2001 at 09:21:24PM -0700, Geoff Keating wrote:
> > Date: Sun, 20 May 2001 17:28:00 -0700
> > From: "H . J . Lu" <hjl@lucon.org>
> > Cc: Geoff Keating <geoffk@redhat.com>, drepper@cygnus.com,
> > howarth@bromo.med.uc.edu, libc-alpha@sourceware.cygnus.com
>
> > On Sun, May 20, 2001 at 05:20:45PM -0700, Richard Henderson wrote:
> > > On Sun, May 20, 2001 at 09:53:01AM -0700, H . J . Lu wrote:
> > > > I believe every port should implelement the mmap2 system call.
> > >
> > > And the 64-bit ports would do this why?
> > >
> >
> > Of course, mmap2 is for 32-bit ports. Ok, let me rephrase
> >
> > 1. All 32-bit ports should have the mmap2 system call and use
> > sysdeps/unix/sysv/linux/mmap64.c.
>
> New kernel ports, 32-bit and 64-bit, should use a 64-bit off_t in all
> kernel interfaces. There is no reason to do otherwise. So the
> compatibility code you see does not need to work for more platforms
> than the ones it already works on.
Let me rephrase again:
1. All ports with 32bit off_t should have the mmap2 system call and use
sysdeps/unix/sysv/linux/mmap64.c.
1. All ports with 64bit off_t should have a dummy
sysdeps/unix/sysv/linux/$arch/mmap64.c.
It is true today. What my patch does it to remove the duplication. Why
do you like duplication so much?
H.J.