LinuxBIOS with the CPU cache disabled 5-7 to when the OS was loaded on
a server board.  I did that by mistake once.  Totally smoked what the
OEM BIOS was doing.

Plus I find it very attractive to make the argument that OEM BIOS's
are slow simply because they are proprietary.

If someone can find a better explanation, I'd love to hear it.

Eric

"Preston L. Bannister" <[EMAIL PROTECTED]> writes:

> How about simply because the LinuxBIOS code:
> 
> 1.  Runs 32-bit instead of 16-bit code.
>     (I remember a 2x improvement rule-of-thumb).
> 
> 2.  Runs at full clock rate.
>     (Quite a difference between 6 and 600 Mhz - if different).
> 
> 3.  Enables the CPU cache.
>     (Could be another 10x or better factor there).
> 
> Not that I know exactly what your typical OEM BIOS does... :)
> 
> I'll bet the about pretty much covers the difference.
> 
> From: Eric W. Biederman
> > I have wondered for the longest while why linuxBIOS can so easily be
> > faster than the legacy BIOS's.  And I have come up with an interesting
> > hypothesis.  Other BIOSs are obfuscated to prevent reverse
> > engineering, and to hide their secrets.  My only guess is this tangled
> > up code slows them down.

Reply via email to