So setting the processor type basically just sets the data
alignment boundaries? Rumours have been floating around for a long time
that the CPU selection subtly influences code in other parts of the
kernel. As I'm not proficient enough to go through the code and look for
such instances I have to take what you kernel guys tell me on faith. :)
>Coming up in the kernel, I think the whole chip definition section should
>be overhauled into some type of sassy multi-selection deal.
If the only effect would be cosmetic then I can see why this
wasn't ever really done. Although considering there are Cyrix patches
floating around, it sounds like a good idea to me.
>Bzzzt. ;-)
>I know that stepping 1 (I think?) had a bug where a bit of the MSR
>information
>was swapped (I think it was discussed here too).. but other than that,
>the K5's
>I've messed with have run anything Pentium great. It's not really 486
>technology at all.. it's a nifty cybil of a chip.. RISC running CISC..
>Considering that it can equal a higher mhz Pentium at a lower actual mhz
>rating
>is impressive alone..
>
>Sure.. the FPU is slower than the Pentium, but the Pentium is a perverted
>chip
>anyway.. floating points faster than integers??
>
>Well, I'm going off on an offtopic tangent, so I'll stop now. :)
I didn't think it was too far off topic. Its nice to get answers
to questions that have been bugging me for a long time. :) Especially
with the varied selection of CPU's these days, its nice to get some
definitive answers on how best to utilize them under Linux.
-Scott
-- Scott Lampert fortunato@heavymetal.org http://www.ioa.com/~fortunato "Sing the Hare Hare, Dance the Hoochie Koo"