> I can't think of a reason not too use one; however, if the P-4 runs at
>1.5gHz and has additional enhancements such as fewer clocks per
>instruction, more instructions, registers ( it has a bunch more floating
>point registers than pentium-pro-3, athlon), faster front-side bus, bigger
>cache, etc. etc. and perhaps it's the direction things will migrate, then
>it may make sense to go in that direction instead of the 1.2 gHz athlon.
I agree with you for these enhancement : faster front-side bus, bigger cache
But disagree with additional instructions, more registers, more fp registers, because
these particularity must be used by the compiler !
I don't think that gcc or other compiler take care of these cpu changement.
More, if you think using more registers, more fpu registers, the kernel should take
care of this to save these ones during task switching !
I had this problem when I wanted to use SIMD instructions and registers for the PIII,
the kernel was not adapted !
So, I don't think using P4 will enhance anything more than PIII or Athlon at same
speed .
>
>I have a couple of Athlon 1.1gHz machines which we've just installed RH
>6.2 on with unusual difficulty. The generic installed RH6.2 kernel
>crashes and dumps the registers. We got it running with a custom kernel
>from my 600mHz athlon system and then building a custom one for the
>machine.
>
When using a distrib kernel, take care of MTRR using for kernel less than 2.4
If the kernel is compiled with mtrr support, you will have a crash when booting on AMD
Athlon.
( mtrr, which is Intel specific, is called FastPages on amd cpu and driven directly by
the cpu)
So just recompile your kernel without mtrr support for amd cpu.
Chris.
--
Kumsta christophe
Real-time image processing
RTAI/RTlinux
maf-roda
[EMAIL PROTECTED]
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
--
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
