This important email came from David Miller.
It bounced since david apparently changed email addresses
(David, you are subscribed as:
davem@caip.rutgers.edu
rather than:
davem@cobaltmicro.com
You may unsub/resub yourself to fix it, it is open)
Anyway, here it is, so the list can see it:
----- Forwarded message from owner-linux@cthulhu -----
From: "David S. Miller" <davem@dm.cobaltmicro.com>
To: ralf@uni-koblenz.de
CC: adevries@engsoc.carleton.ca, linux@cthulhu.engr.sgi.com
In-reply-to: <19980507192145.65233@uni-koblenz.de> (ralf@uni-koblenz.de)
Subject: Re: errors...
References: <Pine.LNX.3.95.980507112045.20653B-100000@lager.engsoc.carleton.ca>
<19980507175205.27567@uni-koblenz.de>
<199805071609.JAA02646@dm.cobaltmicro.com> <19980507192145.65233@uni-koblenz.de>
Date: Thu, 7 May 1998 19:21:45 +0200
From: ralf@uni-koblenz.de
I've just taken a look at the sources and found that the various
architectures seem to handle this issue differently by default.
Sparc disables the exceptions, Intel enables all IEEE754
exceptions, MIPS only division by zero and overflow and Alpha
handles things a bit different anyway ...
Seems there still is no common dominator on what should be default
and I don't think either 0x00000600 or 0x0 is clearly right or
wrong.
Log into an IRIX machine, compile a program that prints out the FPU
csr value you see handed to you in main(), 'nuff said, it's zero, and
this is what I did. All the weird FPU exceptions go away and everyone
is happy.
Later,
David S. Miller
davem@dm.cobaltmicro.com
----- End of forwarded message from owner-linux@cthulhu -----
--
Peace, Ariel
|