On Fri, 29 Oct 1999, Ian Field wrote:
> > >I using a netwinder with linux 2.2.12 and gcc 2.95 -
> > >How do I interface with the FPE to get floats? Is there a compiler switch?
> >
> > What do you mean by `interface to the FPE'? If you just write your code as
> > normal the FPE will automatically pick up on floating point operations.
>
> what kind of performance differences would someone see with say, mpg123 or
> xamp, if you just compiled it and caught the floating point instructions with
> exceptions, versus compiling with floating point emulation from the start?
Both won't run real-time anyway. You need to convert everything to fixed
point math if you need performance. I have a version of splay on
ftp.netwinder.org/users/n/nico that I converted to fixed point math if
interested.
Nicolas
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
