[Webkit-unassigned] [Bug 15996] New: Apply further sanity to match() function (for 10%!? speedup in dna-regexp?)

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Nov 15 02:01:04 PST 2007


http://bugs.webkit.org/show_bug.cgi?id=15996

           Summary: Apply further sanity to match() function (for 10%!?
                    speedup in dna-regexp?)
           Product: WebKit
           Version: 525+ (Nightly build)
          Platform: Macintosh
        OS/Version: Mac OS X 10.4
            Status: NEW
          Severity: Normal
          Priority: P2
         Component: JavaScriptCore
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: eric at webkit.org


Apply further sanity to match() function (for 10%!? speedup in dna-regexp?)

This was not intended as a speedup patch.  However SunSpider claims it's a 10%
speedup to regexp-dna.  It also claims it's a regression to some other random
(non-regexp  using!) tests.  I blame bad codegen.  Or the toothfairy.  I can't
decide which.

TEST                   COMPARISON            FROM                 TO           
 DETAILS

=============================================================================

** TOTAL **:           0.3% faster     4377.8ms +/- 0.1%   4365.4ms +/- 0.2%   
 significant

=============================================================================

  3d:                  1.9% *slower*    497.6ms +/- 0.5%    507.0ms +/- 2.4%   
 significant
    cube:              4.6% *slower*    166.8ms +/- 0.6%    174.4ms +/- 6.9%   
 significant
    morph:             ??               160.4ms +/- 0.4%    160.8ms +/- 0.3%   
 not conclusive: might be 0.2% *slower*
    raytrace:          0.8% *slower*    170.4ms +/- 0.7%    171.8ms +/- 0.3%   
 significant

  access:              0.8% faster      717.4ms +/- 0.5%    711.4ms +/- 0.2%   
 significant
    binary-trees:      -                108.8ms +/- 1.7%    108.6ms +/- 0.6% 
    fannkuch:          1.4% faster      347.6ms +/- 0.2%    342.8ms +/- 0.2%   
 significant
    nbody:             -                182.0ms +/- 0.5%    181.8ms +/- 0.3% 
    nsieve:            1.0% faster       79.0ms +/- 0.0%     78.2ms +/- 0.7%   
 significant

  bitops:              ??               621.4ms +/- 0.4%    622.8ms +/- 1.2%   
 not conclusive: might be 0.2% *slower*
    3bit-bits-in-byte: -                116.4ms +/- 1.6%    115.6ms +/- 1.0% 
    bits-in-byte:      ??               153.4ms +/- 0.9%    155.6ms +/- 4.6%   
 not conclusive: might be 1.4% *slower*
    bitwise-and:       -                214.2ms +/- 0.3%    214.2ms +/- 0.3% 
    nsieve-bits:       -                137.4ms +/- 0.5%    137.4ms +/- 0.5% 

  controlflow:         -                155.8ms +/- 0.7%    155.4ms +/- 0.7% 
    recursive:         -                155.8ms +/- 0.7%    155.4ms +/- 0.7% 

  crypto:              -                345.0ms +/- 0.5%    343.6ms +/- 0.8% 
    aes:               -                 97.8ms +/- 0.6%     97.6ms +/- 0.7% 
    md5:               -                124.2ms +/- 0.4%    123.6ms +/- 0.9% 
    sha1:              -                123.0ms +/- 0.7%    122.4ms +/- 2.1% 

  date:                ??               368.0ms +/- 0.3%    369.2ms +/- 0.7%   
 not conclusive: might be 0.3% *slower*
    format-tofte:      1.5% *slower*    160.6ms +/- 0.4%    163.0ms +/- 1.3%   
 significant
    format-xparb:      -                207.4ms +/- 0.7%    206.2ms +/- 0.3% 

  math:                3.3% *slower*    510.8ms +/- 0.4%    527.6ms +/- 0.5%   
 significant
    cordic:            7.0% *slower*    213.8ms +/- 0.5%    228.8ms +/- 0.6%   
 significant
    partial-sums:      -                174.2ms +/- 0.3%    173.8ms +/- 0.3% 
    spectral-norm:     1.8% *slower*    122.8ms +/- 0.8%    125.0ms +/- 0.7%   
 significant

  regexp:              10.0% faster     296.0ms +/- 0.3%    266.4ms +/- 0.4%   
 significant
    dna:               10.0% faster     296.0ms +/- 0.3%    266.4ms +/- 0.4%   
 significant

  string:              -                865.8ms +/- 0.5%    862.0ms +/- 0.2% 
    base64:            1.3% faster      122.6ms +/- 0.9%    121.0ms +/- 0.0%   
 significant
    fasta:             0.4% *slower*    205.2ms +/- 0.3%    206.0ms +/- 0.4%   
 significant
    tagcloud:          -                197.0ms +/- 1.2%    196.2ms +/- 0.8% 
    unpack-code:       -                202.4ms +/- 0.5%    202.4ms +/- 0.3% 
    validate-input:    1.6% faster      138.6ms +/- 1.2%    136.4ms +/- 1.2%   
 significant


-- 
Configure bugmail: http://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.



More information about the webkit-unassigned mailing list