Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!ames!amdahl!uunet!vsi!friedl
From: friedl@vsi.COM (Stephen J. Friedl)
Newsgroups: comp.sys.att
Subject: Re: FPU on the UNIX pc (was Re: All right! and 1.2M floppies; HwNote05)
Summary: Careful with math chips
Message-ID: <898@vsi.COM>
Date: 23 Oct 88 16:16:17 GMT
References: <368@uncle.UUCP> <3462@rphroy.UUCP> <528@icus.islp.ny.us>
Organization: V-Systems, Inc. -- Santa Ana, CA
Lines: 23

In article <528@icus.islp.ny.us>, lenny@icus.islp.ny.us (Lenny Tropiano) writes:
> I'm not sure if this is actually true, but if someone was able to
> interface the MC68881 math acc. unit to the UNIX PC, the software problems
> might not be too hard.  Of course without system source, you wouldn't
> be able to recompile anything in the system to utilitize the FPU.  

One must be careful with plugging in math chips and stuff.  Many
times the operating system must provide support in the context
switch mechanism to save/restore the FP registers on-chip.  Without
this, a single process will work fine but multiple processes using
the FPU may fail because their registers get confused.

Also, some systems keep their floating point emulator in the kernel,
so putting in a math chip will indeed speed up things without a
recompile all around.

Disclaimer: I know nothing about the MC68881 or the specifics of the
UNIX-PC kernel, but this may be a consideration for the adventurous.

-- 
Steve Friedl    V-Systems, Inc.  +1 714 545 6442    3B2-kind-of-guy
friedl@vsi.com     {backbones}!vsi.com!friedl    attmail!vsi!friedl
---------Nancy Reagan on the Three Stooges: "Just say Moe"---------