Talk:Read-copy-update

I just scanned this article several times looking for the spot where RCU is actually explained, and I keep missing it. The first part of the Overview section looks promising, but ultimately dances around the point it's trying to make.

{{quote|The {{mono|rcu_assign_pointer}} primitive is used to carry out this assignment, and ensures that the assignment is atomic in the sense that concurrent readers will either see a NULL pointer or a valid pointer to the new structure, but not some mash-up of the two values.}}

This is explaining very little; it simply introduces a magic primitive. Even more confusing is that [http://www.rdrop.com/users/paulmck/RCU/RCUdissertation.2004.07.14e1.pdf one of the references] asserts that no reader-side synchronization is needed, while the Overview section specifies the need for {{mono|rcu_read_lock}} and a corresponding unlock operation. After giving this all-too-high level overview, we plunge into the depths of Linux kernel code.

Another problem with the page is the use of peacock terms such as "excellent performance, scalability, and real-time response". QVVERTYVS (hm?) 13:37, 17 November 2015 (UTC)

Hmmmm... An alternative to the "peacock terms" would be something like this: "In the most aggressive implementations, which are actually used in practice in production, RCU readers can use the exact same sequence of machine instructions that would be used in a single-threaded environment." Would that be better, or would that still considered to be peacockery? PaulMcKenney (talk) 19:18, 27 June 2016 (UTC)

:[https://www.efficios.com/pub/rcu/urcu-main.pdf This paper] gives a much more precise overview than this wiki page. It's a bit dense, but it sets out the conditions and mechanism in/by which RCU works. QVVERTYVS (hm?) 13:55, 17 November 2015 (UTC)

I certainly have no objection to the description in that paper. That paper's [http://www.computer.org/cms/Computer.org/dl/trans/td/2012/02/extras/ttd2012020375s.pdf supplementary materials] might also be of interest.

Given that there is a quite a bit more RCU-related material available than ten years ago, and given that you have formally registered your discomfort with my editing this page, I will leave this to you. I do expect to check up on it from time to time, of course. PaulMcKenney (talk) 19:18, 27 June 2016 (UTC)

Hello fellow Wikipedians,

I have just added archive links to {{plural:2|one external link|2 external links}} on Read-copy-update. Please take a moment to review [https://en.wikipedia.org/w/index.php?diff=prev&oldid=707324546 my edit]. If necessary, add {{tlx|cbignore}} after the link to keep me from modifying it. Alternatively, you can add {{tlx|nobots|deny{{=}}InternetArchiveBot}} to keep me off the page altogether. I made the following changes:

• Added archive http://web.archive.org/web/20160105153047/http://www.eecg.toronto.edu/~tornado to http://www.eecg.toronto.edu/~tornado
• Added archive http://web.archive.org/web/20061010230626/http://www.eecg.toronto.edu:80/~tornado/ to http://www.eecg.toronto.edu/~tornado/

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{tl|Sourcecheck}}).

{{sourcecheck|checked=false}}

Cheers.—^{cyberbot II}_{Talk to my owner:Online} 05:07, 28 February 2016 (UTC)

Removed jargon and peacockry. Described abstractly what rcu is, where its used, and how it functions.

SecularJohannes (talk) 03:43, 14 March 2017 (UTC)