Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.programming.threads > #1938
| From | aminer <aminer@toto.net> |
|---|---|
| Newsgroups | comp.programming.threads, comp.programming |
| Subject | Scalable Locks... |
| Date | 2013-10-21 13:02 -0700 |
| Organization | albasani.net |
| Message-ID | <l43mma$64h$1@news.albasani.net> (permalink) |
Cross-posted to 2 groups.
Hello all, Why not use the Windows critical section and that's all ? you have to understand that the Windows critical section is not starvation-free and it's not FIFO fair and it doesn't minimize efficiently the cache-coherence traffic cause it is using a backoff mechanism, so since the Windows critical section is using a backoff mechanism, that means that same thread will have the opportunity to enter many many times the same critical section, and that's what happen with Windows critical section, so as you have noticed this is fast cause the same same thread will reuse the same variable from the same local cache, but this is not FIFO fair and it's not starvation-free, so this why i have decided to contribute to the FreePascal and Lazarus community and i have implemented scalable Locks that are scalable and that minimizes efficiently the cache-coherence traffic and that are FIFO fair and starvation-free. And you can download them from my website: http://pages.videotron.com/aminer/ Thank you, Amine Moulay Ramdane.
Back to comp.programming.threads | Previous | Next | Find similar
Scalable Locks... aminer <aminer@toto.net> - 2013-10-21 13:02 -0700
csiph-web