JVMLangSummit - User contributions [en]

What the JVM needs

2010-07-28T21:51:32Z

Hiroshi: /* types */

== Notes from Rich Hickey's workshop ==

=== user mode limitations ===
* please optimize this
* fences
* typestate: now immutable - especially arrays (typically encapsulated in non-writing header, a la java.lang.String)
* [[Why Tailcalls]]: as a guarantee, not just an optimization
* dynamic vs. inline
* safepoints

=== numbers ===
* fixnums (say, 28-bit small ints encoded as (x >> 28) == -1) http://blogs.sun.com/jrose/entry/fixnums_in_the_vm
* need a uniform entry/exit types (Object => boxing)
* Hacker's delight overflow detection is OK (only for add/sub); need a static method that throws overflow-exception?
* Unsigned <

=== storage management ===
* floating garbage foo(o, o=null); is tailcall a solution?
* concurrent collection of permgen

=== limits ===
* 31-bit arrays
* 16-bit instructions
* static data structures (array literals)

=== types ===
* structs (efficient pointerless element selection)
* value types (below)

The C#'s struct might be a source of ideas (though that's not necessarily how we want it)

=== value types ===
no mutability, acmp-identity, id-hashcode, monitorenter
* immutable arrays == array-like value types
* issue: intermediate early array construction states interact with memory model
* perhaps need a typestate distinction: under construction vs. immutable

Continuations in Servers

2010-07-27T17:33:19Z

Hiroshi:

'''Speaker'''

Hiroshi Yamauchi, Google

'''Abstract'''

We are looking at a problem in servers about scalability vs.
productivity. There are two types of architectures: synchronous (aka
blocking, 'thread-per-request') and asynchronous (aka non-blocking,
'select-server' style, event-driven) servers. The key distinction lies
in how threads are mapped to requests and whether to block or not in
network I/O (eg RPC or database access). The former is better in terms
of productivity because the code gets simpler. The latter is better in
scalability because a thread does not have to be tied to each request.
We are investigating a continuation-based solution that gives us the
best of both worlds: simpler code and better scalability. A synthetic
rpc server has been 'converted' to use continuations and showed good
performance. An experiment with a production server has been ongoing.
We've been using a modified version of the old MLVM callcc patches.

[[Media:ContinuationInServers.pdf‎]]

File:ContinuationInServers.pdf

2010-07-27T17:31:08Z

Hiroshi:

Continuations in Servers

2010-07-27T17:18:23Z

Hiroshi:

Continuations in Servers

2010-07-27T16:03:16Z

Hiroshi:

Hiroshi Yamauchi, Google

We are looking at a problem in servers about scalability vs.
productivity. There are two types of architectures: synchronous (aka
blocking, 'thread-per-request') and asynchronous (aka non-blocking,
'select-server' style, event-driven) servers. The key distinction lies
in how threads are mapped to requests and whether to block or not in
network I/O (eg RPC or database access). The former is better in terms
of productivity because the code gets simpler. The latter is better in
scalability because a thread does not have to be tied to each request.
We are investigating a continuation-based solution that gives us the
best of both worlds: simpler code and better scalability. A synthetic
rpc server has been 'converted' to use continuations and showed good
performance. An experiment with a production server has been ongoing.
We've been using a modified version of the old MLVM callcc patches.

Continuations in Servers

2010-07-27T16:02:44Z

Hiroshi: New page: We are looking at a problem in servers about scalability vs. productivity. There are two types of architectures: synchronous (aka blocking, 'thread-per-request') and asynchronous (aka non-...

We are looking at a problem in servers about scalability vs.
productivity. There are two types of architectures: synchronous (aka
blocking, 'thread-per-request') and asynchronous (aka non-blocking,
'select-server' style, event-driven) servers. The key distinction lies
in how threads are mapped to requests and whether to block or not in
network I/O (eg RPC or database access). The former is better in terms
of productivity because the code gets simpler. The latter is better in
scalability because a thread does not have to be tied to each request.
We are investigating a continuation-based solution that gives us the
best of both worlds: simpler code and better scalability. A synthetic
rpc server has been 'converted' to use continuations and showed good
performance. An experiment with a production server has been ongoing.
We've been using a modified version of the old MLVM callcc patches.

2012 Main Page

2010-07-27T16:01:06Z

Hiroshi: /* Agenda for the 2010 JVM Language Summit */

== [[#SUMMIT ATTENDEES: READ THIS FIRST|SUMMIT ATTENDEES: CLICK THIS FIRST]] ==

==Agenda for the 2010 [http://jvmlangsummit.com/ JVM Language Summit]==

{| border="1"
!
!Monday 7/26
!Tuesday 7/27
!Wednesday 7/28
|-
| 8:30|| align="center" colspan="3" style="color:green; background-color:#ffffcc;" | Breakfast (30 mins) 
|-
| 9:00|| Brian Goetz, Welcome || rowspan="2" | Doug Lea (Keynote) [[Engineering Fine-Grained Parallelism in Java]] || Jochen Theodorou [[Improvements in Groovy performance]]
|-
| 9:30|| Charles Nutter [[Trending Toward the Middle: The Best of Static and Dynamic]] || Rémi Forax [[PHP.reboot: a post JSR292 dynamic language]]
|-
| 10:00|| Eric Caspole [[Improvements in OpenJDK useful for JVM languages]] || Joshua Bloch [[Performance Anxiety]] || Per Bothner [[Kawa and gnu.bytecode update]]
|-
| 10:30|| align="center" colspan="3" style="color:green; background-color:#ffffcc;" | Break (30 mins) 
|-
| 11:00|| John Rose [[Gathering the threads: JVM Futures]] || rowspan="2" style="color:lightblue; background-color:green;" | W1A: John Rose W1B: Robert McKinney || rowspan="2" style="color:lightblue; background-color:green;" | W1A: Rich Hickey W1B: ''Available''
|-
| 11:30|| Christian Wimmer [[Register Allocation on SSA form for Java Just-in-Time Compilation]]
|-
| 12:00|| align="center" rowspan="2" style="color:green; background-color:#ffffcc;" | Lunch (60 mins) || align="center" colspan="2" style="color:green; background-color:#ffffcc;" | Lunch (30 mins) 
|-
| 12:30|| Mark Reinhold JDK7 || Adam Messinger Oracle
|-
| 1:00|| Lukas Stadler [[Coroutines for the Java Platform]] || Neal Gafter [[LINQ: Language Features for concurrency (among other things)]] || David Pollak [[Faking closures on the JVM isn't as simple as it looks]]
|-
| 1:30|| Prashant Deva [[Chronon - Time Travelling Debugger]] || John Field [[The Thorn Programming Language: Robust Distributed Scripting on the JVM]] || Christine Flood [[Project Fortress compiler implementation status]]
|-
| 2:00|| Fredrik Öhrström [[Efficient compilation of Lambdas using MethodHandles and JRockit]] || align="center" colspan="2" style="color:green; background-color:#ffffcc;" | Break (30 mins)
|-
| 2:30|| rowspan="2" style="color:lightblue; background-color:green;" | W1A: Fredrik Öhrström Compile your favourite Lambda expression! W1B: Ben Titzer Virgil on the JVM || Dan Heidinga [[MethodHandles: an IBM implementation]] || Andy Clement [[Mixed language project compilation in Eclipse: Java and Groovy]]
|-
| 3:00|| Cliff Click, Performance || Oleg Pliss [[Symmetric multilanguage VM architecture]]
|-
| 3:30|| align="center" style="color:green; background-color:#ffffcc;" | Break (30 mins) || Kresten Krab Thorup [[Erjang - A JVM-based Erlang VM]] || Michael Van De Vanter [[The Maxine Inspector: a Specialized Tool for VM Development]]
|-
| 4:00|| rowspan="2" style="color:lightblue; background-color:green;" | W2A: Charles Nutter W2B: Hiroshi Yamauchi [[Continuations in Servers]] || rowspan="2" style="color:lightblue; background-color:green;" | W2A: Kresten Krab Thorup W2B: John Field || rowspan="2" style="color:lightblue; background-color:green;" | W2A: Cliff Click W2B: Per Bothner Speedy Scripting
|-
| 4:30
|-
| 5:00|| Break || Lightning talks || Final break
|-
| 6:00|| || Dinner at Faultline ||
|}

== Resources ==

* Main pages for the JVM Language Summit in [[2008_Main_Page | 2008]] and [[2009_Main_Page | 2009]].

== SUMMIT ATTENDEES: READ THIS FIRST ==

If you have a http://java.net/ (or OpenJDK) user name, please use that to [http://wiki.jvmlangsummit.com/index.php?title=Special:UserLogin&type=signup register on this wiki].

If you logged in for the first time as a jvmlang participant, please [http://wiki.jvmlangsummit.com/index.php?title=Special:UserLogin&type=signup register as yourself on this wiki].

The initial jvmlang participant account does not have full write privileges; please use it only for self-registering.

* [http://wiki.jvmlangsummit.com/index.php?title=Special:UserLogin&type=signup First register your real ID] then log out and log back in using the tiny login link at the upper right.
* Consult the [http://meta.wikimedia.org/wiki/Help:Contents User's Guide] for information on using the wiki software.
* [http://www.mediawiki.org/wiki/Manual:Configuration_settings Configuration settings list]
* [http://www.mediawiki.org/wiki/Manual:FAQ MediaWiki FAQ]
* [http://lists.wikimedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list]