[cgl_discussion] Hyperthreading project home?

Mika Kukkonen mika at osdl.org
Mon Nov 25 09:43:46 PST 2002


On ma, 2002-11-25 at 08:41, Stephen Hemminger wrote:
(...)
> As far as I can tell, HT has been done as part of the mainline kernel
> development and that is probably best, since nothing about it is CGL (or
> DCL) specific.

HT was debated somewhat when it was added to the feature list, and I
opposed the plan to add it. Reasons why I let it go in the end were:
  a) In theory, it is not specific to x86, but other architectures 
     should be able to use same technology.
  b) With some strech of imagination, it can be considered relevant to
     CGL, where the heat and size do have some importance (in theory
     HT addresses those issues, i.e. 2-way SMP w/ HT vs. 4-way SMP).
  c) As it was only getting into mainline, adding it to our list did
     not cause too much extra burden for us.
  d) There was more than one member company wanting this feature in.

I do admit that this is the utmost borderline case between CGL features
and features that should not be in CGL, i.e. HT is about
as far as possible I am willing to stretch the definition of CGL
feature.

--MiKu




More information about the cgl_discussion mailing list