Re: Hyperthreading on 2.4.18 and 2.4.20 (Redhat)
From: red floyd (no.spam_at_here.dude)
Date: Thu, 18 Mar 2004 18:09:25 GMT
Marc Schwartz wrote:
> On Thu, 18 Mar 2004 08:57:02 -0500, General Schvantzkoph wrote:
>>On Thu, 18 Mar 2004 16:00:24 +0800, Stephen SM WONG wrote:
>>>Interesting, but since you will not have (physical)
>>>processor affinity in hyperthreading situation, you might
>>>run into the case that only one (physical) processor is
>>>running 2 tasks on 2 hyperthreaded virtual processors, and
>>>hence you saw the slow down.
>>>My 2 cents.
>>>Stephen Wong @ Hong Kong.
>>It's certainly worth trying the same experiment with the 2.6.x kernel,
>>rumor has it that the 2.6 kernel has a better scheduler. However the
>>reason to believe that the individual thread performance will still suffer
>>significantly is that when the P4 is operating in hyperthreaded mode the
>>number of physical registers assigned to each thread reduced by 1/2. Also
>>the cache is shared between both threads so the cache is going to thrash.
> FWIW, recently on the Fedora lists, this had been discussed. There was
> one post from Alan Cox indicating that there were still issues with HT in
> the 2.6 kernel and that additional work on HT related scheduling was
> I do not have a multi-CPU system, but do have a new Dell 5150 laptop with
> a 3.2 Ghz P4 with HT running FC1. I have disabled HT in the system's BIOS
> due to known conflict issues with the 2.4 SMP kernel and am using the
> 2.4.x UP kernel.
> There was a recent article at 2CPU.com on 2.6 and HT:
> which you might find of interest.
> Marc Schwartz
I realize that this is a YMMV case... I'm running highly parallelized programs
-- the CPU is part of a Beowulf, actually, and I thought I'd play with HT on a
single node before enabling it on all cluster nodes.
What I have is a large number of LONG highly CPU-bound processes. Given that,
do you think the scheduler problems would be a real issue?