Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.java.programmer > #15311 > unrolled thread
| Started by | "Aaron W. Hsu" <arcfide@sacrideo.us> |
|---|---|
| First post | 2012-06-15 17:33 -0500 |
| Last post | 2012-06-19 07:46 -0700 |
| Articles | 4 on this page of 84 — 19 participants |
Back to article view | Back to comp.lang.java.programmer
Recommendations for Lightweight Threading? "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-15 17:33 -0500
Re: Recommendations for Lightweight Threading? markspace <-@.> - 2012-06-15 15:55 -0700
Re: Recommendations for Lightweight Threading? "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-15 18:12 -0500
Re: Recommendations for Lightweight Threading? markspace <-@.> - 2012-06-15 16:31 -0700
Re: Recommendations for Lightweight Threading? "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-15 20:00 -0500
Re: Recommendations for Lightweight Threading? Robert Klemme <shortcutter@googlemail.com> - 2012-06-16 14:39 +0200
Re: Recommendations for Lightweight Threading? Daniel Pitts <newsgroup.nospam@virtualinfinity.net> - 2012-06-16 12:13 -0700
Re: Recommendations for Lightweight Threading? Roedy Green <see_website@mindprod.com.invalid> - 2012-06-16 00:57 -0700
Re: Recommendations for Lightweight Threading? Lew <lewbloch@gmail.com> - 2012-06-15 15:57 -0700
Re: Recommendations for Lightweight Threading? "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-15 18:12 -0500
Re: Recommendations for Lightweight Threading? Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-15 20:19 -0400
Re: Recommendations for Lightweight Threading? "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-15 19:59 -0500
Re: Recommendations for Lightweight Threading? Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-15 21:37 -0400
Controlling the Garbage Collector "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-16 11:51 -0500
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-16 14:24 -0400
Re: Controlling the Garbage Collector markspace <-@.> - 2012-06-16 12:24 -0700
Re: Controlling the Garbage Collector Daniel Pitts <newsgroup.nospam@virtualinfinity.net> - 2012-06-16 13:14 -0700
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-16 16:35 -0400
Re: Controlling the Garbage Collector "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-16 19:43 -0500
Re: Controlling the Garbage Collector Wanja Gayk <brixomatic@yahoo.com> - 2012-06-23 13:36 +0200
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-23 15:39 +0200
Re: Controlling the Garbage Collector markspace <-@.> - 2012-06-16 14:34 -0700
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-18 04:32 -0700
Re: Controlling the Garbage Collector Roedy Green <see_website@mindprod.com.invalid> - 2012-06-24 14:31 -0700
Re: Controlling the Garbage Collector Arne Vajhøj <arne@vajhoej.dk> - 2012-06-20 21:19 -0400
Re: Controlling the Garbage Collector "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-21 13:24 -0500
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-21 11:37 -0700
Re: Controlling the Garbage Collector Fred Greer <fggreer@nospam.invalid> - 2012-06-21 21:20 +0000
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-21 15:24 -0400
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-21 23:46 +0200
Re: Controlling the Garbage Collector Jukka Lahtinen <jtfjdehf@hotmail.com.invalid> - 2012-06-25 15:28 +0300
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-25 09:05 -0400
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-25 17:01 +0000
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-25 13:45 -0400
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-25 13:49 -0400
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-25 20:41 +0200
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-26 12:25 +0000
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-26 06:46 -0700
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-26 16:26 +0000
Re: Controlling the Garbage Collector Eric Sosman <esosman@ieee-dot-org.invalid> - 2012-06-26 13:07 -0400
Re: Controlling the Garbage Collector Robert Klemme <shortcutter@googlemail.com> - 2012-06-26 22:28 +0200
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-26 23:49 +0000
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-26 18:20 -0700
Re: Controlling the Garbage Collector Highway to Hell <HtH49439112@gmail.com> - 2012-06-26 21:52 -0400
Re: Controlling the Garbage Collector Gene Wirchenko <genew@ocis.net> - 2012-06-26 20:01 -0700
Re: Controlling the Garbage Collector Highway to Hell <HtH49439112@gmail.com> - 2012-06-26 23:23 -0400
Re: Controlling the Garbage Collector Gene Wirchenko <genew@ocis.net> - 2012-06-27 09:05 -0700
Re: Controlling the Garbage Collector Martin Gregorie <martin@address-in-sig.invalid> - 2012-06-27 20:15 +0000
Re: Controlling the Garbage Collector Gene Wirchenko <genew@ocis.net> - 2012-06-27 13:52 -0700
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-27 13:41 -0700
Re: Controlling the Garbage Collector Gene Wirchenko <genew@ocis.net> - 2012-06-27 19:02 -0700
Re: Controlling the Garbage Collector Martin Gregorie <martin@address-in-sig.invalid> - 2012-06-28 21:45 +0000
Re: [OT] Driver's license restrictions (Was: Controlling the Garbage Collector) Lew <lewbloch@gmail.com> - 2012-06-28 15:16 -0700
Re: [OT] Driver's license restrictions (Was: Controlling the Garbage Collector) Martin Gregorie <martin@address-in-sig.invalid> - 2012-06-29 02:10 +0000
Re: [OT] Driver's license restrictions (Was: Controlling the Garbage Collector) Gene Wirchenko <genew@ocis.net> - 2012-06-28 19:57 -0700
Re: [OT] Driver's license restrictions glen herrmannsfeldt <gah@ugcs.caltech.edu> - 2012-06-29 04:06 +0000
Re: [OT] Driver's license restrictions Tim Slattery <Slattery_T@bls.gov> - 2012-06-29 08:35 -0400
Re: [OT] Driver's license restrictions (Was: Controlling the Garbage Collector) Tim Slattery <Slattery_T@bls.gov> - 2012-06-29 08:33 -0400
Re: Controlling the Garbage Collector Tim Slattery <Slattery_T@bls.gov> - 2012-06-29 08:30 -0400
Re: Controlling the Garbage Collector Martin Gregorie <martin@address-in-sig.invalid> - 2012-06-29 23:04 +0000
Re: Controlling the Garbage Collector Highway to Hell <HtH49439112@gmail.com> - 2012-06-27 16:53 -0400
Re: Controlling the Garbage Collector Leif Roar Moldskred <leifm@dimnakorr.com> - 2012-06-27 11:32 -0500
Re: Controlling the Garbage Collector Highway to Hell <HtH49439112@gmail.com> - 2012-06-27 16:54 -0400
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-27 11:45 -0700
Re: Controlling the Garbage Collector Highway to Hell <HtH49439112@gmail.com> - 2012-06-27 16:55 -0400
Re: Controlling the Garbage Collector glen herrmannsfeldt <gah@ugcs.caltech.edu> - 2012-06-27 02:06 +0000
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-27 16:34 +0000
Re: Controlling the Garbage Collector Leif Roar Moldskred <leifm@dimnakorr.com> - 2012-06-27 11:45 -0500
Re: Controlling the Garbage Collector Wanja Gayk <brixomatic@yahoo.com> - 2012-07-02 11:21 +0200
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-26 13:52 -0700
Re: Controlling the Garbage Collector Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2012-06-26 23:40 +0000
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-26 16:48 -0700
Re: Controlling the Garbage Collector markspace <-@.> - 2012-06-21 15:15 -0700
Re: Controlling the Garbage Collector Lew <lewbloch@gmail.com> - 2012-06-21 15:33 -0700
Re: Controlling the Garbage Collector "Aaron W. Hsu" <arcfide@sacrideo.us> - 2012-06-21 21:24 -0500
Re: Controlling the Garbage Collector markspace <-@.> - 2012-06-21 15:23 -0700
Re: Recommendations for Lightweight Threading? Wanja Gayk <brixomatic@yahoo.com> - 2012-06-17 15:49 +0200
Re: Recommendations for Lightweight Threading? Roedy Green <see_website@mindprod.com.invalid> - 2012-06-16 01:00 -0700
Re: Recommendations for Lightweight Threading? Roedy Green <see_website@mindprod.com.invalid> - 2012-06-16 01:04 -0700
Re: Recommendations for Lightweight Threading? Wanja Gayk <brixomatic@yahoo.com> - 2012-06-16 04:03 +0200
Re: Recommendations for Lightweight Threading? Kevin McMurtrie <mcmurtrie@pixelmemory.us> - 2012-06-15 22:27 -0700
Re: Recommendations for Lightweight Threading? Robert Klemme <shortcutter@googlemail.com> - 2012-06-16 14:39 +0200
Re: Recommendations for Lightweight Threading? Kevin McMurtrie <mcmurtrie@pixelmemory.us> - 2012-06-18 19:37 -0700
Re: Recommendations for Lightweight Threading? Robert Klemme <shortcutter@googlemail.com> - 2012-06-19 07:46 -0700
Page 5 of 5 — ← Prev page 1 2 3 4 [5]
| From | Kevin McMurtrie <mcmurtrie@pixelmemory.us> |
|---|---|
| Date | 2012-06-15 22:27 -0700 |
| Message-ID | <4fdc1941$0$86808$742ec2ed@news.sonic.net> |
| In reply to | #15311 |
In article <YvmdncQlVsKmJUbSnZ2dnUVZ_t4AAAAA@giganews.com>, "Aaron W. Hsu" <arcfide@sacrideo.us> wrote: > I am considering moving one of my projects from C to Java, but I am > hoping to find a high-performance threading implementation, or something > along the lines of libqthread, which offers Fill-Empty bit blocking and > good cooperative lightweight threading as a library. > > Is there a current "best" solution when doing many threaded programs in > Java? By many threads I mean many more than the cores or machines on the > network. Something that scales up efficiently to distributed computing > would be nice as well. It's possible to roll your own concurrency classes using compare-and-set or safe races. Custom code with reduced functionality can beat the general JVM classes by a lot. Minimizing thread interaction in high-performance code helps even more. Writing to shared memory results in a hardware stall to sync CPU caches. The bigger the system, the worse those cache syncs are. -- I will not see posts from Google because I must filter them as spam
[toc] | [prev] | [next] | [standalone]
| From | Robert Klemme <shortcutter@googlemail.com> |
|---|---|
| Date | 2012-06-16 14:39 +0200 |
| Message-ID | <a43d3iF1ajU1@mid.individual.net> |
| In reply to | #15323 |
On 16.06.2012 07:27, Kevin McMurtrie wrote: > It's possible to roll your own concurrency classes using compare-and-set > or safe races. Custom code with reduced functionality can beat the > general JVM classes by a lot. That sounds interesting: please share concrete examples where you were able to beat std lib "by a lot". > Minimizing thread interaction in high-performance code helps even more. > Writing to shared memory results in a hardware stall to sync CPU caches. > The bigger the system, the worse those cache syncs are. Well, that's one of the basic principles Doug Lea calls "thread confinement". Not sharing is always more efficient than sharing when it comes to multiple threads. Kind regards robert -- remember.guy do |as, often| as.you_can - without end http://blog.rubybestpractices.com/
[toc] | [prev] | [next] | [standalone]
| From | Kevin McMurtrie <mcmurtrie@pixelmemory.us> |
|---|---|
| Date | 2012-06-18 19:37 -0700 |
| Message-ID | <4fdfe5cd$0$34406$742ec2ed@news.sonic.net> |
| In reply to | #15329 |
In article <a43d3iF1ajU1@mid.individual.net>, Robert Klemme <shortcutter@googlemail.com> wrote: > On 16.06.2012 07:27, Kevin McMurtrie wrote: > > > It's possible to roll your own concurrency classes using compare-and-set > > or safe races. Custom code with reduced functionality can beat the > > general JVM classes by a lot. > > That sounds interesting: please share concrete examples where you were > able to beat std lib "by a lot". Many of the utility classes in the concurrency package use locking because it's required for implementing general List, Map, and the blocking APIs. If you needed only a small subset of features, you could eliminate the locks. Eliminating the locks would reduce overhead and, in many cases, improve the JIT compiler optimizations. The classic example is data that has an incredibly high read to write ratio. That makes it possible to use copy-on-write rather than locking. You can also eliminate locks for blocking methods when there is one thread that may block. The blocking thread calls LockSupport.park() and the unblocking thread calls LockSupport.unpark(). This comes in handy for one worker thread performing many very small tasks. The Java classes don't have this drastic limitation of a single thread so they need locking. (This can work for multiple threads but the complexity and overhead may get out of control.) > > > Minimizing thread interaction in high-performance code helps even more. > > Writing to shared memory results in a hardware stall to sync CPU caches. > > The bigger the system, the worse those cache syncs are. > > Well, that's one of the basic principles Doug Lea calls "thread > confinement". Not sharing is always more efficient than sharing when it > comes to multiple threads. > > Kind regards > > robert -- I will not see posts from Google because I must filter them as spam
[toc] | [prev] | [next] | [standalone]
| From | Robert Klemme <shortcutter@googlemail.com> |
|---|---|
| Date | 2012-06-19 07:46 -0700 |
| Message-ID | <31ca1101-5c32-49e8-ba9f-57eac03abd43@googlegroups.com> |
| In reply to | #15406 |
On Tuesday, June 19, 2012 4:37:00 AM UTC+2, Kevin McMurtrie wrote: > In article <a43d3iF1ajU1@mid.individual.net>, > Robert Klemme <shortcutter@googlemail.com> wrote: > > > On 16.06.2012 07:27, Kevin McMurtrie wrote: > > > > > It's possible to roll your own concurrency classes using compare-and-set > > > or safe races. Custom code with reduced functionality can beat the > > > general JVM classes by a lot. > > > > That sounds interesting: please share concrete examples where you were > > able to beat std lib "by a lot". > > Many of the utility classes in the concurrency package use locking > because it's required for implementing general List, Map, and the > blocking APIs. If you needed only a small subset of features, you could > eliminate the locks. Eliminating the locks would reduce overhead and, > in many cases, improve the JIT compiler optimizations. > > The classic example is data that has an incredibly high read to write > ratio. That makes it possible to use copy-on-write rather than locking. In which cases did these classes not work or not work well for you? java.util.concurrent.CopyOnWriteArrayList java.util.concurrent.CopyOnWriteArraySet etc. http://docs.oracle.com/javase/6/docs/api/java/util/concurrent/package-summary.html java.util.concurrent.atomic.AtomicReferenceArray etc. http://docs.oracle.com/javase/6/docs/api/java/util/concurrent/atomic/package-summary.html > You can also eliminate locks for blocking methods when there is one > thread that may block. The blocking thread calls LockSupport.park() and > the unblocking thread calls LockSupport.unpark(). This comes in handy > for one worker thread performing many very small tasks. The Java > classes don't have this drastic limitation of a single thread so they > need locking. (This can work for multiple threads but the complexity > and overhead may get out of control.) I see. Thank you! Kind regards robert
[toc] | [prev] | [standalone]
Page 5 of 5 — ← Prev page 1 2 3 4 [5]
Back to top | Article view | comp.lang.java.programmer
csiph-web