Port-amd64 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Changing x87 precision to full 63bit as default



On Tue, Nov 12, 2013 at 06:00:01PM +0900, tsugutomo.enami%jp.sony.com@localhost 
wrote:
> > There are a number of options here. With default options, GCC at least
> > will not store assignments to memory (see -ffloat-store and
> > -fexcess-precision=standard) without some force. The canonical way is to
> > assign to a volatile variable, see libm/src/math_private.h
> > STRICT_ASSIGN. It would be nice to use fenv.h for this, but support for
> > that is even less wide spread...
> 
> Thanks, but for the gauche's case, forcing the float value to memory
> didn't fix the problem.  It looks like we need to set 387 mode to 53 bit
> precision explicitly by using fpsetprec() to avoid unwanted rounding due
> to 63 bit precision.

What exactly have you tried?

> (Here is detailed description of the problem by the author:
> http://blog.practical-scheme.net/gauche/20130128-ratnum-bignum )

Long rant without meet ;)

Joerg


Home | Main Index | Thread Index | Old Index