Tamarin:WeeklyUpdates: Difference between revisions

Jump to navigation Jump to search

Tamarin:WeeklyUpdates (view source)

Revision as of 21:25, 12 August 2008

2,611 bytes removed ,  12 August 2008
no edit summary
No edit summary
Line 10: Line 10:


'''Next meeting's Agenda Items:'''
'''Next meeting's Agenda Items:'''
*TBD


* (David) Explicit memory allocation.
= Aug 12th 2008 =
* (Mason) Progress on x86 Forth representation.
* (Moh)
** Update on Jungwoo's results with parallel JITing in TT.
*** Core-2 Duo is dual core, Silverthorne (Atom) is hyper-threaded: [[Image:Parallel_TT_JIT.PNG]]
** vprof is in TT. You can use it for value profiling. Extensions and more capabilities are underway.
** Improving InlineHashTable. Good speedup observed on Win32. https://bugzilla.mozilla.org/show_bug.cgi?id=449886
= July 22nd 2008 =


== Attendees ==
== Attendees ==


* intel: moh, carmen, shengnan, jungwoo
* intel: moh, carmen, shengnan, jungwoo
* mozilla: joel, dave, andreas
* mozilla: dave, andreas, brendan
* adobe: mason, dan, edwin s, rob, steve, vlad, jennifer, rick
* adobe: mason, edwin s, steven, vlad, scott, dan, thomas


== Updates ==
== Updates ==
*(Andreas) SSE2 calling conventions. SSE2-only JIT? Shall we specialize with SSE2-only?
*(David)  
** This may make slower processers even slower.  This would affect Flash.  This should be investigated though.  Would compile-time be enough? Need to look into this.  SSE2 or later machines would makes sense.
**Explicit memory allocation
** bug 440601
***Just made a quick patch on this.  This has also been discussed in previous meetings.  Plan is to commit the patch by next week.
*** gregor at UCI posted some python code to look at functions on mac -- might need volunteers to help with win and mac implementations
*(Mason)
*** would anyone be willing to buddy up with andreas for win and linux?
**Progress on x86 Forth representation.
**** edwin suggests andreas do a little profiling on some built-ins to see if performance is improved first.
*** Trying to get rid of the interpreter dispatch loop.  Has a basic version up now with 20% increase in speed in TT.  Hopes to have more test results by end of week.
**** performance-critical stuff seems to be in lib.c -- include them directly? seems to sound okay!
*(Moh)  
 
**Update on Jungwoo's results with parallel JITing in TT.
 
***Jungwoo has working now in TT. More work still needs to be done.
* (andreas/gregor) Register usage analysis for builtins
***Core-2 Duo is dual core, Silverthorne (Atom) is hyper-threaded.
** andreas will try to provide some numbers for the next call
**vprof is in TT. You can use it for value profiling. Extensions and more capabilities are underway.  
 
***Ed has also ported this to alternative platforms. Ed also worked on a timer to help see transitions for various components. Moh has also worked on some of the timers.
* (edwin) vprop for GCC
**Improving InlineHashTable. Good speedup observed on Win32. https://bugzilla.mozilla.org/show_bug.cgi?id=449886
** working on some patches for this
***Looking at 64bitHashTable as well.
*** will push patches, pending successful testing on windows
*(Steven)
 
**Taking fixes from Flash Player / TC and getting these integrated into the mainline.  May have to submit as multiple patches depending on how the check-in goes.
* (edwin) fragmento
*(Ed)
** assembler is long-lived but we want it to be short-lived
**Will be doing reviews of the reviews: VTune patch, etc.  Thanks to those who put those in while he was out.
** not re-using exit pages was using a lot of memory.
*(Mark)  
*** now, when recompiling a tree, optimizations make cache usage about 75% faster
**Waiting to hear on GlobalObjects issue with TC- any progress? Someone on Adobe side is looking into this (Tom Harwood), but might take a bit of time to review.
** andreas asks: will implemental growth patch still be needed?
*** mason did the testing on bug 443111 -- some tests fail on linux, but other platforms look fine.
*** edwin will look into these crashes.
*** andreas will wait to make his check-in until edwin has investigated a bit further.
** edwin also did a lot of code clean-up with this patch.
** what about getting rid of the exit block entirely? (rick)
*** this would be nice, but haven't thought about the details yet.
** in the end, edwin thinks we will want to use shared side exists.
*** andreas has not tried this yet.
 
* (moh) Speeding up the hot loop of MMGc PinStackObjects using SSE2 parallel compares
** https://bugzilla.mozilla.org/show_bug.cgi?id=446556
*** a large amount of stack scanning is happening. 
*** moh has gotten about 7% performance improvements so far
*** there is a similar loop in MarkItem, but adjusting that has not yielded as much improvement.
*** edwin thinks this is a valuable patch
 
*(Jungwoo) Update on parallel JIT'ing in TT. 
** Is trying moving all patches to interpreter side.
** 15-30% speedups could be possible!
*** edwin would like numbers with and without trees (these are without trees)
*** did intel look into MD5?  no, not yet,
 
*(Steven) Is tracking down performance issues for a Flex-based app for Tamarin Tracing.  New issues are coming up with TT vs. TC issues in doing this.
** working on improving superwords -- preliminary patch is posted for review
*** roughly 20% interpretor speed-up, but no tracing improvement
*** trying to identify some other areas to work on
 
*(Edwin) tri-server
** shared sandbox is online!  everyone is welcome to use this now.
** working on modifying build system to pull patches from different users' repositories for they get queued up properly.
 
* (dan) automated performance metrics
** we have now published a performance page.   
*** once a checkin is made, buildbot runs performance tests on various platforms and posts results.
*** weekly performance data is also summarized.
** how to track feedback to this system?  log in bugzilla.
 
*(dave) TraceMonkey:
** still working on additional patches  
***  will write up a wiki article soon
** basic ARM testing is underway
 
*(Mingqui) Submitted 2 patches a few weeks ago.  Steven at Adobe still needs to review these…
**https://bugzilla.mozilla.org/show_bug.cgi?id=438889
**https://bugzilla.mozilla.org/show_bug.cgi?id=438881


= Older meetings =
= Older meetings =
Sunterberg
66

edits

Retrieved from "https://wiki.mozilla.org/Special:MobileDiff/103852"

Navigation menu