Port-i386 archive

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

Re: MATH_EMULATE




To: der Mouse <mouse%Rodents.Montreal.QC.CA@localhost>

Subject: Re: MATH_EMULATE

From: Vit Herman <v.herman%sh.cvut.cz@localhost>

Date: Tue, 15 Jan 2008 19:49:51 +0100 (CET)


On Tue, 15 Jan 2008, der Mouse wrote:


The major way I would foresee its rotting would be via -msoft-float.
If my understanding that this would be the only way we'd run on FPUless
x86, then this would be the only user of i386 -msoft-float....


I thought that kernel is always built with -msoft-float?


I mean, how much could such compile options get incompatibile with
possible source changes -


asm()s using FPU instructions, is the first way that comes to mind.


But that was an issue always - e
ven back when the emulation worked, right?  The only way to solve such thing would be to have complete emulation of  FPU (we could countinue with MMX, SSE...).

Vit Herman



Follow-Ups:

Re: MATH_EMULATE
From: Perry E. Metzger


References:

MATH_EMULATE
From: Andrew Doran

Re: MATH_EMULATE
From: Perry E. Metzger

Re: MATH_EMULATE
From: Jared D. McNeill

Re: MATH_EMULATE
From: Perry E. Metzger

Re: MATH_EMULATE
From: Jared D. McNeill

Re: MATH_EMULATE
From: Perry E. Metzger

Re: MATH_EMULATE
From: Pierre Dubuc

Re: MATH_EMULATE
From: Vit Herman

Re: MATH_EMULATE
From: Matthias Drochner

Re: MATH_EMULATE
From: Perry E. Metzger

Re: MATH_EMULATE
From: der Mouse

Re: MATH_EMULATE
From: Perry E. Metzger

Re: MATH_EMULATE
From: Vit Herman

Re: MATH_EMULATE
From: der Mouse




Prev by Date: Re: MATH_EMULATE

Next by Date: Re: MATH_EMULATE

Previous by Thread: Re: MATH_EMULATE

Next by Thread: Re: MATH_EMULATE

Indexes:

reverse Date

reverse Thread

Old Index



Home | Main Index | Thread Index | Old Index