[Qemu-devel] x86 FPU Emulation

[Qemu-devel] x86 FPU Emulation

[The Qube]

Using any DOS/DOS-like setup under qemu x86 (no kqemu/qvm86) and running 
calc387 will produce non-sane results.  For example, "calc387 2^4" produces 
17.7635.  "calc387 4^2" produces 17.7635 as well.  I've verified this is not 
an issue with calc387 by  running the above two calculations under dosbox 
0.63 with the dynamic core and geting "16" back for both..  I've test this 
on an Athlon XP system.  runcom doesn't work with calc387, so qemu-i386 
isn't testable.

calc387's homepage (near the bottom): http://www.geocities.com/craig_hessel/

Re: [Qemu-devel] x86 FPU Emulation

[Christian MICHON]

I see the same problems on bartpe (XP winpe compatible).
By the way, on XP (guest or host), you need to type
calc387 2^^4.

But I've tested other software, notably 3D images raytracers,
using FP operations. No such problems. Images look good.
Yet I guess I should do a pixel per pixel comparison to
check any truncation.

Another possibility would be to compile the asm of  calc387
on linux to check the values.

Christian

On 7/17/05, The Qube <qube99@hotmail.com> wrote:
> Using any DOS/DOS-like setup under qemu x86 (no kqemu/qvm86) and running
> calc387 will produce non-sane results.  For example, "calc387 2^4" produces
> 17.7635.  "calc387 4^2" produces 17.7635 as well.  I've verified this is not
> an issue with calc387 by  running the above two calculations under dosbox
> 0.63 with the dynamic core and geting "16" back for both..  I've test this
> on an Athlon XP system.  runcom doesn't work with calc387, so qemu-i386
> isn't testable.
> 
> calc387's homepage (near the bottom): http://www.geocities.com/craig_hessel/
> 
> 
> 
> 
> _______________________________________________
> Qemu-devel mailing list
> Qemu-devel@nongnu.org
> http://lists.nongnu.org/mailman/listinfo/qemu-devel
> 


-- 
Christian