Tamarin:WeeklyUpdates: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
Line 31: Line 31:




= 6th May 2008 =
= 20th May 2008 =


== Attendees ==
== Attendees ==


* Adobe: Ed, Scott, Rick, Steven, Tom
(Note, unfortunately I neglected to take attendance; these are the people I know were there. If you were present and not listed, please add yourself)
* Mozilla: Dave, Jason, Jim, Brian Crowder
 
* Intel: Shengnan Cong, Marsha Eng, Jungwoo Ha, Moh Haghighat, Mingqiu Sun
* Adobe: Rick, Steven
* Mozilla: Ben, Jim
* Intel: Shengnan, Moh


== Updates ==
== Updates ==


** Moh (Intel)-  
** Moh (Intel)-  
*** Major performance gains from optimized doubleToInt32 (double d) using integer code. https://bugzilla.mozilla.org/show_bug.cgi?id=412978
*** An easy performance gain for string comparisons was offered in bug 433799. Steven to push change in.
**** This is used in SpiderMonkey and Tamarin.  Intel’s numeric team has helped out with this issue. Great performance gains thus far- Ed has also gotten these consistent results
*** runtests.py produces inaccurate results per bug 434118. A fix was offered. Adobe to review the fix and push this week.
***** Mozilla to see how this works with Firefox
*** Intel wants to investigate doing a simple post-codegen pass for easy & fast optimizations (eg code alignment, peephole), definitely worth investigating — Rick to follow up with Intel offline, going to open a bug on it.
**** Now it needs to be tested!
*** Intel wants to add LIR_file, LIR_line instructions for VTune — Rick thought was already there but apparently not in the right form (not being transmitted from IL to LIR). Problem is harder than it sounds; Rick to think about it and work with Intel offline. Shengnan has initial patch she’s going to attach to the existing bug.
*** CMOV is no longer considered harmful!
 
**** A discussion on the performance considerations of using CMOV. Intel recommends using CMOV except in certain highly predictable cases.
** Jim (Mozilla)-
*** Working on bug 409021, merging MMgc from TC to TT; mostly done but still some bugs to track down (also, a few patches are still awaiting review by Tommy). Rick and Steven advised about greedy and pedantic modes, which are likely to help track down his remaining bugs.


** Steven (Adobe)-
** Steven (Adobe)-
*** Working on a patch for an interpreter speed-up. Currently is only working on Mac. Patch should be viewable later today.  
*** Restoring some functionality that was in TC but got lost on the way to TT (introspection and lazy type resolution). These features are probably not very important for Mozilla but are very important for Flash.
**** Progress is being made!


** Rick (Adobe)-
** Rick (Adobe)-
*** Is continuing to work on trees. Expects to land patch soon and go over updates in the next few weeks.
*** Rick gave a synopsis of trace tree work. It’s not working yet but hopefully will be by next week. Note, merging trees isn’t part of this (eg the Game Of Life problem) but is phase 2 of this problem (ie this work is a prerequisite for having the information needed to merge trees)
 
*** At Moh’s request, Rick also talked some about how register allocation is currently done. Moh is of the opinion there's room for improvement.
** Jim (Mozilla)-
*** Update on regression of performance.  In order for Tamarin to get into Mozilla, then patches can’t land if they slow things down.  Take a benchmark approach and get TT to be on par w/ SpiderMonkey- then it can be switched over.
**** Mozilla will contribute speed optimizations to Adobe.  Adobe needs to be comfortable making optimized code “clean” going forward, while making sure the numbers hit.
***** Adobe is building out its own internal performance measurements now. Going forward, make sure Mozilla and Adobe have agreement on benchmarks.
 
** Tom (Adobe)-
*** Has posted a wiki page on Hosting APIs that he wants Mozilla to review.  He will also send around to the list.
 
** Mark Hammond-
*** How to best evangelize Tamarin through ScreamingMonkey? How to get folks using it?
**** The IE DOM works as hopes (minus some expected issues).  Can we evangelize and package?  Yes, but who can assist? John R and other evangelists can blog on this. Brendan to follow-up with Mark on this.


= Older meetings =
= Older meetings =

Revision as of 22:18, 20 May 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)
  • Location: Tel: 866 705 2554 (us), 913 227 1201 (int’l)
    • Passcode: 9140087
  • Duration: 60 minutes
  • join irc.mozilla.org #tamarin for attendence taking and questions
  • Last meeting's Agenda Items:
    • (Moh/Intel)-
    • (Mark Hammond)-
      • How to best evangelize Tamarin through ScreamingMonkey?


20th May 2008

Attendees

(Note, unfortunately I neglected to take attendance; these are the people I know were there. If you were present and not listed, please add yourself)

  • Adobe: Rick, Steven
  • Mozilla: Ben, Jim
  • Intel: Shengnan, Moh

Updates

    • Moh (Intel)-
      • An easy performance gain for string comparisons was offered in bug 433799. Steven to push change in.
      • runtests.py produces inaccurate results per bug 434118. A fix was offered. Adobe to review the fix and push this week.
      • Intel wants to investigate doing a simple post-codegen pass for easy & fast optimizations (eg code alignment, peephole), definitely worth investigating — Rick to follow up with Intel offline, going to open a bug on it.
      • Intel wants to add LIR_file, LIR_line instructions for VTune — Rick thought was already there but apparently not in the right form (not being transmitted from IL to LIR). Problem is harder than it sounds; Rick to think about it and work with Intel offline. Shengnan has initial patch she’s going to attach to the existing bug.
    • Jim (Mozilla)-
      • Working on bug 409021, merging MMgc from TC to TT; mostly done but still some bugs to track down (also, a few patches are still awaiting review by Tommy). Rick and Steven advised about greedy and pedantic modes, which are likely to help track down his remaining bugs.
    • Steven (Adobe)-
      • Restoring some functionality that was in TC but got lost on the way to TT (introspection and lazy type resolution). These features are probably not very important for Mozilla but are very important for Flash.
    • Rick (Adobe)-
      • Rick gave a synopsis of trace tree work. It’s not working yet but hopefully will be by next week. Note, merging trees isn’t part of this (eg the Game Of Life problem) but is phase 2 of this problem (ie this work is a prerequisite for having the information needed to merge trees)
      • At Moh’s request, Rick also talked some about how register allocation is currently done. Moh is of the opinion there's room for improvement.

Older meetings