Jump to content

Talk:Time Stamp Counter

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Anti-reverse Engineering

[edit]

Would be worth documenting RTDSC's use in anti-reverse engineering techniques. Faissaloo (talk) 08:56, 26 April 2019 (UTC)[reply]

Merger from RDTSC

[edit]

This article should be merged with RTDSC

Indeed it should. One wonders why it took two years for someone to start working on it. --Aurochs (Talk | Block)
Okay, I've dumped most of the text from RDTSC into this article, and most references to the instruction with references to the counter itself. I think most of the intro text should be moved into a section, since at the moment it forms the bulk of the article. I'm still debating whether or not we really need the code examples, since anybody interested in using it could easily figure it out on their own. --Aurochs (Talk | Block)

I'm very disappointed, that you have removed all the code! Now I have to search for it! —Preceding unsigned comment added by 194.237.142.7 (talk) 08:21, 14 October 2008 (UTC)[reply]

Older copies of the RDTSC article can still be found in its history here. rCX (talk) 19:34, 14 October 2008 (UTC)[reply]

Yeah, I completely disagree that anybody interested in it could figure it out on their own; it's handy for people who don't know x86 assembler, and just want to drop it into some high level language. I think we should bring the code examples back.

Then put it on Wikibooks. It doesn't belong here. --Aurochs (Talk | Block)

Don't agree: Sections of code that show the use of something are (a) verifiable and (b) often essential to understanding the point, purpose and use of it. If you apply this "How To" principle then 90% of the wiki-held documentation on mathematics would have to also disappear. Similarly, descriptions of famous algorithms like the various sort methods (e.g. Prof. Wirth's implementation) would disappear. Listen to the computer enthusiasts and software professionals when they say "keep the code examples"! Show some common sense - not the fine print of the rule book. And please remove the rather ugly banner about this on the actual page. — Preceding unsigned comment added by 86.164.178.110 (talk) 05:42, 20 November 2011 (UTC)[reply]

You are correct in noting that it's difficult to describe a procedure for doing something (an algorithm) without providing an example of said procedure. However, the time stamp counter is not a procedure in itself; it is a component of a procedure. Throwing a bunch of code in that does nothing more than read out the time stamp counter does not help with understanding the point, purpose, or use of the time stamp counter, any more than code that simply loads the instruction pointer into the accumulator would help with understanding the point, purpose, or use of the instruction pointer. Now, simple pseudocode could be written that uses a generic time stamp counter to determine the number of clock cycles elapsed during a loop, for instance, and this would be acceptable in the article (at least from my point of view). --Aurochs (Talk | Block) 17:49, 28 August 2012 (UTC)[reply]

Nehalem has always-running-TSC on

[edit]

http://lwn.net/Articles/345007/ says that the intel Nehalem has an always running TSC. maybe this should be added. 194.36.2.101 (talk) 15:49, 3 August 2009 (UTC)[reply]

Tacts not ticks

[edit]

The article states that the number of ticks (100-nanosecond units) is counted, while in fact, rdtsc gives the number of processor cycles - tacts since the last reset. — Preceding unsigned comment added by 194.204.37.249 (talk) 17:34, 29 December 2011 (UTC)[reply]