Re: Linux/IA-64 byte order
Florian Lohoff (flo@rfc822.org)
Tue, 9 Mar 1999 17:49:12 +0100
On Tue, Mar 09, 1999 at 10:31:57AM +0100, Jes Sorensen wrote:
> Linus> On Tue, 9 Mar 1999, Alan Cox wrote:
> >> Do you even know if the option about endianness is software ? For
> >> all I know it might be wired down. I don't care what endian we use
> >> 8)
>
> Linus> I care, and it's going to be LE, simply because anything else
> Linus> is just asking for trouble with the x86 binary compatibility
> Linus> part.
>
> I might be missing a point, but why do _we_ care that much about x86
> compatibility? I see why M$ and others want it will all their closed
> source applications but for Linux it is just a matter of recompilation
> and that is what people really want to do. You say you care very
> little about kernel module binary compatibility, why do you suddenly
> care about compatibility of binary only user land applictions?
>
> I do not say we should use one byte order or the other, I just don't
> see x86 binaries as a big reason reason for choosing one over the
> other.
Thats the point - Currently only very few people still run
old pre 90 software for 8086 and it will be the same with Linux
and Merced. As soon as the Merced comes out no-one will use
old x86 binarys so i would prefer a little performance decrease
in x86 emulation than penalties with little endianess for years
in future. If we will have the chance of doing a cut to leave
stupid decisions behind us we should talk about REAL issues
not history of Computer Hardware.
Also with the Emulation issue - CPU get better performance and
a little x86 performance penalty causes the application to run
at same speed as long as they are not native code.
Flo
--
Florian Lohoff flo@rfc822.org +49-5241-470566
Good, Fast, Cheap: Pick any two (you can't have all three). (RFC 1925)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/