* Jim Wilson -- Sunday 26 May 2002 16:20: > Melchior FRANZ <[EMAIL PROTECTED]> said: > > > But this was indeed with -ffast-math (2.95.2, Linux). I recompiled both SimGear > > and FlightGear without that flag and the c310 works now, while the a4 still > reports > > a SOLUTION FAILURE, albeit with different values (always the same, of course): > > This is strange.
I'd say it is alarming. The report with -ffast-math and without shows quite different results. And this behavious is not covered by the description of this compiler flag, which is supposed to mainly drop things like validity checks, such as negative values in sqrt(). m. _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
