[Webkit-unassigned] [Bug 18991] SquirrelFish: Major codegen issue in a.b=expr, a[b]=expr

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu May 15 06:31:15 PDT 2008


http://bugs.webkit.org/show_bug.cgi?id=18991





------- Comment #11 from cwzwarich at uwaterloo.ca  2008-05-15 06:31 PDT -------
This is really strange. When I ran it with network on, I got a regression, but
with network off I got a 0.1% *progression*:

TEST                   COMPARISON            FROM                 TO           
 DETAILS

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

** TOTAL **:           1.001x as fast    2360.8ms +/- 0.1%   2357.9ms +/- 0.1% 
   significant

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

  3d:                  *1.012x as slow*   305.9ms +/- 0.1%    309.6ms +/- 0.1% 
   significant
    cube:              *1.021x as slow*    92.4ms +/- 0.2%     94.3ms +/- 0.1% 
   significant
    morph:             *1.018x as slow*   109.8ms +/- 0.2%    111.7ms +/- 0.3% 
   significant
    raytrace:          -                  103.8ms +/- 0.2%    103.6ms +/- 0.1% 

  access:              1.013x as fast     357.3ms +/- 0.1%    352.8ms +/- 0.1% 
   significant
    binary-trees:      *1.009x as slow*    39.4ms +/- 0.4%     39.7ms +/- 0.3% 
   significant
    fannkuch:          1.034x as fast     140.5ms +/- 0.1%    135.9ms +/- 0.1% 
   significant
    nbody:             -                  140.8ms +/- 0.1%    140.6ms +/- 0.1% 
    nsieve:            -                   36.7ms +/- 0.4%     36.5ms +/- 0.4% 

  bitops:              1.010x as fast     265.9ms +/- 0.1%    263.3ms +/- 0.2% 
   significant
    3bit-bits-in-byte: -                   35.8ms +/- 0.7%     35.6ms +/- 0.4% 
    bits-in-byte:      -                   49.3ms +/- 0.3%     49.2ms +/- 0.2% 
    bitwise-and:       1.022x as fast      93.9ms +/- 0.1%     91.8ms +/- 0.4% 
   significant
    nsieve-bits:       1.002x as fast      86.9ms +/- 0.1%     86.7ms +/- 0.2% 
   significant

  controlflow:         *1.010x as slow*    36.0ms +/- 0.3%     36.3ms +/- 0.4% 
   significant
    recursive:         *1.010x as slow*    36.0ms +/- 0.3%     36.3ms +/- 0.4% 
   significant

  crypto:              ??                 160.2ms +/- 0.1%    160.4ms +/- 0.1% 
   not conclusive: might be *1.001x as slow*
    aes:               *1.010x as slow*    61.1ms +/- 0.2%     61.7ms +/- 0.2% 
   significant
    md5:               -                   48.4ms +/- 0.3%     48.4ms +/- 0.3% 
    sha1:              1.008x as fast      50.7ms +/- 0.2%     50.3ms +/- 0.3% 
   significant

  date:                1.002x as fast     193.5ms +/- 0.1%    193.1ms +/- 0.1% 
   significant
    format-tofte:      -                  120.0ms +/- 0.1%    120.0ms +/- 0.1% 
    format-xparb:      1.006x as fast      73.5ms +/- 0.2%     73.1ms +/- 0.1% 
   significant

  math:                1.006x as fast     265.1ms +/- 0.1%    263.5ms +/- 0.1% 
   significant
    cordic:            1.003x as fast      86.9ms +/- 0.2%     86.6ms +/- 0.2% 
   significant
    partial-sums:      -                  125.4ms +/- 0.1%    125.3ms +/- 0.1% 
    spectral-norm:     1.025x as fast      52.9ms +/- 0.2%     51.6ms +/- 0.3% 
   significant

  regexp:              -                  214.1ms +/- 0.2%    213.9ms +/- 0.2% 
    dna:               -                  214.1ms +/- 0.2%    213.9ms +/- 0.2% 

  string:              *1.004x as slow*   562.8ms +/- 0.1%    565.0ms +/- 0.1% 
   significant
    base64:            *1.007x as slow*    92.4ms +/- 0.2%     93.0ms +/- 0.1% 
   significant
    fasta:             *1.007x as slow*   117.1ms +/- 0.1%    118.0ms +/- 0.1% 
   significant
    tagcloud:          *1.005x as slow*   130.8ms +/- 0.2%    131.5ms +/- 0.2% 
   significant
    unpack-code:       -                  132.5ms +/- 0.2%    132.3ms +/- 0.1% 
    validate-input:    ??                  89.9ms +/- 0.4%     90.2ms +/- 0.4% 
   not conclusive: might be *1.004x as slow*


-- 
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