<p>Am 25.08.2017 01:33 schrieb "Ralf Quint" <<a href="mailto:freedos.la@gmail.com">freedos.la@gmail.com</a>>:<br>
><br>
> On 8/24/2017 2:45 PM, Ched wrote:<br>
> > And under "modern" Windows, the extended type, which is fully<br>
> > supported by the FPU, is *degraded* to double. So, I can't do high<br>
> > precision computation with fpc even if the native hardware permits<br>
> > such computations. I'm glued to XP as the full capabilities of the FPU<br>
> > were on the hands of programmers, years ago.<br>
> It's not a problem of "modern" Windows, but a problem of any 64bit x86<br>
> OS that in that "long" mode the FPU does NOT support the extended data<br>
> type (80 bits, 10 bytes). In 32bit mode, the FPU is using the "old" x87<br>
> FP unit on the chip, in 64bit mode, it is using the SSE FP unit, which<br>
> doesn't have those 80 bit registers, it's 64 bit only (and several times<br>
> faster).<br>
> So that you can use that with your Windows XP version is likely due to<br>
> the fact that this is a 32bit version, not XP Professional 64 (or<br>
> Windows Server 2003 64 for that matter).<br>
><br>
> There are apparently some ways to enable the FP calculation to use the<br>
> x87 FPU and therefor the possibility of those 80 bit registers on Inter<br>
> chips (so far), but for one, this runs significantly slower that the SSE<br>
> FPU, and then this is apparently not supported by (all) AMD CPUs, so you<br>
> are limiting yourself also on which systems your code can run...</p>
<p>Plase note that all non-Windows 64-bit systems have no problems with enabling the FPU and even Windows needs to handle it due to 32-bit software requiring it.<br>
The only x86_64 OS we support that doesn't have Extended enabled (by default) *is* Win64.</p>
<p>Regards,<br>
Sven</p>