Bernd Paysan wrote: > > On Sunday 28 March 2004 22:03, Dennis Ruffer wrote: > > If not a ffcall problem, why does it come and go simply by installing > > ffcall? DaR > > We try to work around the broken long long. Usually "broken long long" for > us means long long=long (on 64 bit architectures), so we use long for that. > However, in your case, sizeof(long long)=0, so there's something much more > serious going on. What we can try to do is to avoid long long for ffcalls > if it is so severely broken. > > What you can do is to find out *why* long long is so seriously broken. After > all, GCC for PowerPC (32 bit version) did have a working long long with 64 > bit support for a long time. Why is it broken now?
It is broken in gcc-2.95 for PPC (i.e., for quite some time), and the best way to deal with that seemed to be to turn use of long long off for PPC (doubles are not so common that the slowdown is noticable for most people). - anton --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
