tag:blogger.com,1999:blog-2935119175410671378.post7817229353837751657..comments2023-03-22T03:23:08.549-06:00Comments on High Performance Linux: NatSys Lock-free Queue vs LMAX Disruptor vs boost::lockfree::queueAlexander Krizhanovskyhttp://www.blogger.com/profile/00939006050444455233noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-2935119175410671378.post-5885704477277636462018-04-26T10:05:08.012-06:002018-04-26T10:05:08.012-06:00This comment has been removed by a blog administrator.Anonymoushttps://www.blogger.com/profile/05704159231627046186noreply@blogger.comtag:blogger.com,1999:blog-2935119175410671378.post-34108793986244336342014-04-03T06:40:06.772-06:002014-04-03T06:40:06.772-06:00Hi Igor,
that was not my idea to reference varon-...Hi Igor,<br /><br />that was not my idea to reference varon-t's C implementation as LMAX Disrutor. Firstly, I was asked many times how our approach (algorithm, not particular C or Java implementation) differs from Disruptor design. And secondly akadien referred the C implementation as a plug for C at Hacker News (https://news.ycombinator.com/item?id=7042771).Alexander Krizhanovskyhttps://www.blogger.com/profile/00939006050444455233noreply@blogger.comtag:blogger.com,1999:blog-2935119175410671378.post-45755624053526178312014-04-03T04:29:28.993-06:002014-04-03T04:29:28.993-06:00Alexander, code review is great as usual, but the ...Alexander, code review is great as usual, but the article has following flaw: you are seemingly attributing varon-t's C implementation as LMAX Java original Distruptor ("LMAX Disruptor .. Its source code available"), which isn't correct.Anonymoushttps://www.blogger.com/profile/06873928316234508392noreply@blogger.comtag:blogger.com,1999:blog-2935119175410671378.post-4937176940890647952014-03-29T07:53:19.897-06:002014-03-29T07:53:19.897-06:00This comment has been removed by the author.Alexander Krizhanovskyhttps://www.blogger.com/profile/00939006050444455233noreply@blogger.comtag:blogger.com,1999:blog-2935119175410671378.post-90308863384756043672014-01-22T04:41:55.069-07:002014-01-22T04:41:55.069-07:00Hi Ahmad,
probably I've never faced "cac...Hi Ahmad,<br /><br />probably I've never faced "cache block" in literature, and "cache line" could be misleading, but in most cases it's clear from context whether an author writes about physical placeholder in a cache or a payload. So I just prefer to hold common practise.Alexander Krizhanovskyhttps://www.blogger.com/profile/00939006050444455233noreply@blogger.comtag:blogger.com,1999:blog-2935119175410671378.post-68909391336873292002014-01-21T13:54:34.380-07:002014-01-21T13:54:34.380-07:00Interesting article!
Just have a side comment. I ...Interesting article!<br /><br />Just have a side comment. I noticed you use the words "cache line". However, I think this should be "cache block". This is a general misconception. Cache line and cache block are different. The cache line is the physical placeholder inside a cache, however, a cache block is the payload inside that physical placeholder. People tend to confuse them together since they are both of the same size. Hence, movement is supposed to be for cache blocks and not for cache lines :)Ahmad Samihhttps://www.blogger.com/profile/07349145523408167849noreply@blogger.com