[MPlayer-dev-eng] [PATCH] More accurate af_volume

[D Richard Felker III]

On Sun, Feb 20, 2005 at 09:26:49PM +0100, Alexander Strasser wrote:
> D Richard Felker III wrote ( On Sun, Feb 20, 2005 at 02:17:11PM -0500 ):
> > On Sun, Feb 20, 2005 at 07:36:25PM +0100, Romain Dolbeau wrote:
> > > Reimar Döffinger wrote:
> > > 
> > > >Is there an architecture where this is not the case?
> > > 
> > > all ILP64 OS/arch. Alpha with almost any OSes is one.
> > > 
> > > At least Solaris 9 on Sparc V9 and Linux on Itanium seems
> > > to default to I32LP64...
> > 
> > This is all non-gcc. On gcc, int is always 32bit, for good reason.
> > If you make int any bigger, there's at least one size (either 16bit or
> > 32bit) that can't exist at all within the standard types.. this is an
> > unfortunate limitation of the C language.
> 
> That of course makes sense, but i think we should still not assume it is
> always 32bit when writing code and using the standard type int.

I didn't read the code well, but I fail to see how it assumed int is
32bit. As far as I could tell all it needed was at least 32 bits,
which is a given.

Rich