Tamarin:WeeklyUpdates: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
m (Fix link to the tracing paper.)
No edit summary
Line 12: Line 12:




= 18th December 2007 =


== Attendees ==
= 8th January 2008 =


* Marsha
* Brett
* Ed
* Steven
* Brendan
* Moh
* Benjamin
* Andreas Gal
* Dan Schaffer
* Jeff


== tamarin-tracing ==
== ATTENDEES ==


* Adobe will push later today a branch of the code in tamarin-central (T-C), which introduces a trace-based just-in-time compiler.
* Research (http://www.ics.uci.edu/~franz/Site/pubs-pdf/ICS-TR-07-10.pdf) and experience with Tamarin indicate that tracing provides significant size and performance advantages over traditional static JIT compilers, especially when executing untyped JavaScript.


* The new code will reside in a repository at:
 
  http://hg.mozilla.org/tamarin-tracing


* Tamarin-tracing (T-T) shares a common root with T-C and so the complete history of changes is recorded. Fixes from both branches will be merged to the other over time.
* Tom
* T-T is currently optimized to run effectively on mobile devices. As such, its performance characteristics favor memory efficiency over speed.
* Steven
* Not API compatible with T-C
* Ben
* Calls to builtins can be optimized by rewriting builtins in interpreted (trace-able) byte code. This also leads to smaller object files
* Jason
* The interpreter is and indirect threading Forth engine.
* Brent
* Erik
* Edwin
* Jeff
* Rick
* Marsha
* Moh
* Mark
* Brendan


== Status ==


'''Mark H'''
== ANNOUNCEMENTS ==
* Q. How different are the APIs? A. Pretty different. They are not compatible, but the builts have already been ported, and you'll only need to port you own host objects.


* Shouldn't be too bad then.


* ChessApp working a bit better. JScript can call Tamarin and vice-versa. This allows workaround of missing event handling across engines since JScript event handlers can call Tamarin functions. Close to having Tamarin move chess pieces around the board.
* Tamarin summit February 1st at Adobe – SF


'''Benjamin'''


* From earlier email:
== STATUS ==
 
 
'''Moh'''
 
* Escalated release of vtune
* Runtime dll added to standard release of vtune
* Need contact person at mozilla, expert in licensing, as a contact. Suggest Brendan.
* Name of dll and header
* Release team of VTune in Russia. Via vtune update.
* Analyzer/bin/Jitpi.dll


Just a note: I'm working on a project to add valgrind memory annotations to
MMgc. This will allow valgrind to note improper use of GC memory.


The first step is quite simple: it will use the valgrind client requests
'''Marsha'''
described at
http://valgrind.org/docs/manual/mc-manual.html#mc-manual.clientreqs to
declare freed memory as unaccessable.


The second step will be to check memory overruns (writes of multiple GC
* additional modification of source code
allocations from a single call).
* release debugger build of tamarin
* variable overhead sometimes as high ~50%, reduced to ~20%, but wondering if someone can help reduce the overhead
* Tom and Edwin have volunteered
* Open bug and post patch for Tom and Edwin to see


The third step will be a warning system for correct incremental marking. If
'''Mark'''
you write a pointer-like value to GC memory without using a write barrier,
it will warn. This has to be a conservative warning, of course.


I hope to have step I done by the first week of January.
* chess app should be up and running
* brent owns it for now


- --BDS
'''Edwin'''


'''Jason'''
* optimistic binding pending for TT
* about 30% are running as fast for TC
* Rick: is comparision fair (e.g. sse on or off in TC)
* 5% to 25% in jitted code
* make better by calling helper less of ten
* Brendan: how can we help; maybe with bugs
* Bugs are being moved into bugzilla so take a look


* Stage 1 (ActionMonkey) just about complete. Waiting for reviews. "maybegc" api done. 4 or 5 patches ready to land.
'''Steven'''


* Stage 2, incremental GC, is next
* linux build of TT in process; could use help
* hopefully have something working in a day or two


'''Moh'''
'''Benjamin'''


* Profiled JSBench (http://gforge.ssllab.org/gf/project/jsbench/scmsvn/) from the UCI team. Shows one helper method (doubleToAtom) is responsible for ~46% of the time in one of the tests (Moldyn).
* question about strings
* Will study further to look for optimization opporunities
* bug about Utf8 and Utf16
* mozilla might want to share string class with tamarin
* BE: might want to share interface rather than implementation
* BE: Utf8 is generally a win
* ES: we might consider an implementation that allows multiple encodings


= Older meetings =
= Older meetings =


* [[Tamarin:WeeklyUpdates/2007-12-18|December 18, 2007]]
* [[Tamarin:WeeklyUpdates/2007-12-01|December 11, 2007]]
* [[Tamarin:WeeklyUpdates/2007-12-01|December 11, 2007]]
* [[Tamarin:WeeklyUpdates/2007-12-04|December 4, 2007]]
* [[Tamarin:WeeklyUpdates/2007-12-04|December 4, 2007]]

Revision as of 18:37, 9 January 2008

These updates concern Tamarin and related projects only.

Meeting Details

  • 2:00pm Pacific Time (21:00 UTC) on Tuesdays
    • (5PM Eastern US, 11PM Oslo, 6AM (Wed) Seoul, 7AM (Wed) Melbourne)
  • Meeting ID: 8262746 (TAMARIN)
    • California: 408-536-9900
    • Toll-Free(US & Canada): 877-220-5439
    • International: +1-408-536-9900
  • Duration: 60 minutes
  • join irc.mozilla.org #tamarin for attendence taking and questions


8th January 2008

ATTENDEES

  • Tom
  • Steven
  • Ben
  • Jason
  • Brent
  • Erik
  • Edwin
  • Jeff
  • Rick
  • Marsha
  • Moh
  • Mark
  • Brendan


ANNOUNCEMENTS

  • Tamarin summit February 1st at Adobe – SF


STATUS

Moh

  • Escalated release of vtune
  • Runtime dll added to standard release of vtune
  • Need contact person at mozilla, expert in licensing, as a contact. Suggest Brendan.
  • Name of dll and header
  • Release team of VTune in Russia. Via vtune update.
  • Analyzer/bin/Jitpi.dll


Marsha

  • additional modification of source code
  • release debugger build of tamarin
  • variable overhead sometimes as high ~50%, reduced to ~20%, but wondering if someone can help reduce the overhead
  • Tom and Edwin have volunteered
  • Open bug and post patch for Tom and Edwin to see

Mark

  • chess app should be up and running
  • brent owns it for now

Edwin

  • optimistic binding pending for TT
  • about 30% are running as fast for TC
  • Rick: is comparision fair (e.g. sse on or off in TC)
  • 5% to 25% in jitted code
  • make better by calling helper less of ten
  • Brendan: how can we help; maybe with bugs
  • Bugs are being moved into bugzilla so take a look

Steven

  • linux build of TT in process; could use help
  • hopefully have something working in a day or two

Benjamin

  • question about strings
  • bug about Utf8 and Utf16
  • mozilla might want to share string class with tamarin
  • BE: might want to share interface rather than implementation
  • BE: Utf8 is generally a win
  • ES: we might consider an implementation that allows multiple encodings

Older meetings