Re: Hyperthreading on 2.4.18 and 2.4.20 (Redhat)

From: red floyd (no.spam_at_here.dude)
Date: 03/18/04


Date: Thu, 18 Mar 2004 03:37:53 GMT

General Schvantzkoph wrote:

> On Wed, 17 Mar 2004 22:56:25 +0000, red floyd wrote:
>
>
>>I have an RH8 box, with 2 P4 Xeons. I recently enabled hyperthreading
>>in the BIOS, but am still only seeing 2 CPUs.
>>
>>I have tried Redhat kernels 2.4.18-14 and 2.4.20-28.8. I looked at the
>>source in arch/i386/kernel/{setup,smpboot}.c, and it looks like there's
>>an error in the hyperthreading detection logic. There's some major
>>changes to this code in the vanilla 2.4.25 kernel.
>>
>>Anyone got HT working on an SMP box with a stock RH kernel?
>>Alternatively, what's the diff between a RH kernel and a vanilla kernel?
>>If I put a vanilla kernel on this box, will it break any RH specific stuff?
>>
>>Thanks in advance
>
>
> A standard kernel will work fine, go ahead and build a 2.4.25 and give it
> a try. BTW hyperthreading doesn't work very well, my tests showed that
> hyperthreading increades throughput by only 10% while reducing individual
> thread performance by 25%.
>

Is that in comparison to the UP kernel? I'm already running an SMP kernel,
since I have a dual CPU box.



Relevant Pages

  • Re: hdparm -t yields incorrect timings when Intel hyperthreading is enabled
    ... the system doesn't appear to be THAT much more unresponsive when hyperthreading is enabled over when disabled. ... Let's say that hdparm thinks its transferring bytes for '2' seconds, but in reality the system is blocking somewhere and the full '2' seconds is not actually being used. ... There are quite a few differences in the kernel log ... because Intel did something funny with microcode updates for your processor ...
    (Debian-User)
  • Re: hdparm -t yields incorrect timings when Intel hyperthreading is enabled
    ... What it reeks of is a linux kernel issue. ... I'm not going to replace firmware just to hope that it will fix any issue, much less one with hyperthreading. ... With all this stuff flying around, I thought that maybe I was confused and hdparm might actually show the correct disk bandwidth on the 3.2.0-4 kernel. ... About git bissect: ...
    (Debian-User)
  • Re: hyper threading.
    ... So what that says it that the kernel network code doesn't get any ... >> multiple physical processors. ... >> an obvious performance gain, and hyperthreading does, too, although ... > an obvious performance gain, and hyperthreading does, too, although it's ...
    (freebsd-questions)
  • Re: hdparm -t yields incorrect timings when Intel hyperthreading is enabled
    ... with hyperthreading disabled and one enabled. ... affected by enabling hyperthreading, 3.12 kernel. ... About git bissect: ... git bissect stuff, compiling, etc) on a safe, known release/stable kernel. ...
    (Debian-User)
  • Re: Hyperthreading on 2.4.18 and 2.4.20 (Redhat)
    ... but am still only seeing 2 CPUs. ... > an error in the hyperthreading detection logic. ... > Anyone got HT working on an SMP box with a stock RH kernel? ... > If I put a vanilla kernel on this box, will it break any RH specific stuff? ...
    (comp.os.linux)