Hyperthreading Kernel Configuration - 5.1

[Marc G. Fournier]

On Thu, 2 Oct 2003, Damian Gerow wrote:

> Thus spake John R. Shannon (john at johnrshannon.com) [02/10/03 13:05]:
> > On a new computer, dmesg shows:
> >
> > CPU: Intel(R) Pentium(R) 4 CPU 2.80GHz (2793.01-MHz 686-class CPU)
> >   Origin = "GenuineIntel"  Id = 0xf29  Stepping = 9
> >   Features=0xbfebfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE>
> >   Hyperthreading: 2 logical CPUs
> > acpi_cpu0: <CPU> port 0x530-0x537 on acpi0
> > acpi_cpu1: <CPU> port 0x530-0x537 on acpi0
> >
> > Should options  SMP  and  APIC_IO  be enabled in kernel?
>
> In short: yes.  And then you need to look at the machdep.* sysctl nobs,
> there's one you need to enable in there (I've forgotten which one).
>
> But that leads me to a secondary question: is enabling HTT really worth the
> time?  I know that people have said that using HTT can actually make your
> system slower -- is this an implementation issue, or did Intel really
> release something that degrades performance?

My understanding is that the speed improvements (or degradation) depend on
the use of the machine ... for instance, I've heard that a high I/O server
will be slower with HTT enabled, and, from my experience with one such, it
is so ... I'm not sure what circumstances would show improvements though
...