discussion forum


message

Name: Stephan
Date: June 17, 2003 at 02:38:54
Subject: Re: Hyperthreading and 'lock' prefix
In Reply To: Re: Hyperthreading and 'lock' prefix by
james on June 16, 2003 at 12:29:35
Text: | However, architecturally
| the LOCK is required, to ensure compatibility -- no ifs or buts.

But.... ;-), yes, you are absolutely right.

| Performance isn't that good an excuse for writing implementation-
| dependent code. Also, bundling more than one version of the code
| and selecting the desired variant at run-time requires that you
| come up with an absolutely reliable processor/feature detection.
| Which you won't be able to, unless you can foresee the future.

For this special application, the software will be bundled with the
hardware, so I either able to select a fitting processor or to
adjust the software for a special system.

| Are you absolutely certain that you can't improve the algorithm,
| to avoid this particular performance issue?

I am sure that the algorithm could be rewritten to avoid this
problem, but that would cost a lot of effort. (There are other
options, e.g. I might stick the conflicting threads to one
processor and let other independent threads run on the other,
but that confines the scheduler.)


post a followup message
(Be nice... or be blocked. Be technical... or be erased.)

Name: optional link title:
eMail: optional link URL:
Subject: optional image URL:
  Insert line breaks by hand when only about one inch remains at the right side. Otherwise your message will be unreadable.
Text:
 

  Note: The above eMail form fields may look unaligned if you are using a browser other than Netscape Navigator version 3.0.


currently posted followup messages
(You may have to press the RELOAD button of your browser.)


main page