On Sat, Aug 19, 2000 at 12:17:46AM +0200, Ingo Molnar wrote:
> the 2.4 kernel's scalability of the most common workloads is limited by
> hardware (ie. the kernel lets those workloads scale 'down to the metal').
> The 2.4 kernel scales quite decently on 8 CPUs. I would be surprised if we
> had any serious problem at 32 or 64 CPUs. (Linux right now is
> architecturally limited to 32 CPUs on 32-bit systems and 64 CPUs on 64-bit
> systems - because we have some per-CPU data structures put into word-size
> bitfields.) But fortunately there is still scalability work to be done - a
> perfect kernel would be extremely boring! :-)
On Thursday Kanoj, Cort Dougan, DaveM and me had dinner in Mountain View
and we decieded to tackle this bitmask problem for 2.4. Linux 2.6 or 3.0
will be to long in the future for SGI, so we think we have to already do
this for 2.4.
Ralf
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Aug 23 2000 - 21:00:06 EST