[mdlug] [Fwd: Re: [opensuse] 64 bit vrs 32bit advantages speedetc.]

[Ingles, Raymond]

> From: Adam Tauno Williams

> Ah, the rush to conspiracy.  Perhaps (a) it is a bug, or (b) a
> disagreement over the spec, or (c) they've encountered SSE/SSE2/SSE3
> implementations that don't work as reported so they err on the side of
> paranoia

 ...except Intel's never claimed (c), nor has anyone found any examples
of (c). As to (a), it appears the first iteration *was* a bug - programs
would *crash* on AMD chips instead of just running slower. The patches
produced by the third party show that (b) doesn't apply - altering the
compiler to remove just the "GenuineIntel" check does not impact anything
else, and "GenuineIntel" is *not* part of the SSE spec.

> And if it is (a) or (b) it may take a v-e-r-y long time to resolve;
> understandably so.

 Read the web page. The fixes are so simple they are in Perl scripts.

 Sincerely,

 Ray Ingles                                            (313) 227-2317

 Novice electrician:
   Why should I waste time and money getting the right color wire?
   The electricity doesn't care.
 Experienced electrician:
   Electricity is color-blind.  Electricians are not.
                         -- unknown

 Novice programmer:
   Why should I waste time commenting code? The computer doesn't care.
 Experienced programmer:
   The computer is comment-blind.  Programmers are not.
                -- Dr. Robert J. Meier 1980
The contents of this e-mail are intended for the named addressee only. It contains information that may be confidential. Unless you are the named addressee or an authorized designee, you may not copy or use it, or disclose it to anyone else. If you received it in error please notify us immediately and then destroy it.