[MLton-devel] mutator card-marking costs

Stephen Weeks MLton@mlton.org
Fri, 16 Aug 2002 08:34:02 -0700


Here are the results of comparing MLton 20020728 with my current
working version, where the the generational GC is hardwired off.  So,
the only difference between compiling -generational true and
-generational false is that the mutator does card marking (and the GC
allocates the cardmap).  There is only one benchmark, vector-concat,
where the card marking really hurts -- and that makes no sense, sadly,
because in vector-concat it is integer arrays being concatenated.
I'll look into that one. 

Overall, the mutator cost of card marking is squat both in code size
and run time.

MLton0 -- /usr/local/bin/mlton 
MLton1 -- mlton -generational false
MLton2 -- mlton -generational true

run time ratio
benchmark         MLton1 MLton2
barnes-hut           1.0    1.0
boyer                1.1    0.9
checksum             1.0    1.0
count-graphs         1.0    1.1
DLXSimulator         1.0    1.1
fft                  1.0    1.0
fib                  1.0    1.0
hamlet               1.1    1.1
imp-for              1.0    1.0
knuth-bendix         1.0    1.0
lexgen               1.0    0.9
life                 1.0    1.1
logic                0.9    1.0
mandelbrot           1.0    1.0
matrix-multiply      1.0    1.0
md5                  1.0    1.0
merge                0.9    0.9
mlyacc               1.0    1.1
mpuz                 1.0    1.0
nucleic              1.2    1.2
peek                 1.0    1.0
psdes-random         1.0    1.0
ratio-regions        1.0    1.0
ray                  1.0    1.0
raytrace             1.0    1.0
simple               1.0    1.0
smith-normal-form    1.0    1.0
tailfib              1.0    1.0
tak                  1.0    1.0
tensor               1.0    1.0
tsp                  1.0    1.0
tyan                 0.9    1.0
vector-concat        1.0    1.4
vector-rev           1.0    1.0
vliw                 1.0    1.0
wc-input1            1.0    0.9
wc-scanStream        1.0    1.1
zebra                1.0    1.0
zern                 1.0    1.0

size
benchmark            MLton0    MLton1    MLton2
barnes-hut           61,736    66,576    66,864
boyer               124,315   129,187   131,395
checksum             30,947    35,819    35,819
count-graphs         50,499    55,371    55,531
DLXSimulator         93,550    98,390    98,550
fft                  40,007    44,879    44,879
fib                  30,979    35,819    35,819
hamlet            1,090,668 1,095,508 1,097,140
imp-for              31,011    35,851    35,851
knuth-bendix         70,620    75,460    75,492
lexgen              153,225   158,097   158,737
life                 47,011    51,883    51,883
logic                87,571    92,443    92,763
mandelbrot           31,043    35,915    35,915
matrix-multiply      31,491    36,331    36,331
md5                  39,676    44,516    44,548
merge                32,195    37,067    37,067
mlyacc              466,313   471,185   472,977
mpuz                 34,979    39,851    39,851
nucleic              69,667    74,539    74,539
peek                 38,620    43,460    43,492
psdes-random         32,003    36,875    36,875
ratio-regions        49,283    54,155    54,347
ray                  88,732    93,604    94,052
raytrace            237,689   242,561   243,009
simple              185,607   190,447   190,511
smith-normal-form   139,740   144,580   144,868
tailfib              30,691    35,563    35,563
tak                  31,107    35,979    35,979
tensor               60,675    65,515    65,579
tsp                  45,084    49,924    50,660
tyan                 89,036    93,876    95,028
vector-concat        31,459    36,331    36,331
vector-rev           31,331    36,171    36,171
vliw                295,941   300,813   302,797
wc-input1            53,913    58,753    58,913
wc-scanStream        54,585    59,425    59,617
zebra               116,564   121,404   122,172
zern                 37,510    42,350    42,350


-------------------------------------------------------
This sf.net email is sponsored by: OSDN - Tired of that same old
cell phone?  Get a new here for FREE!
https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390
_______________________________________________
MLton-devel mailing list
MLton-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mlton-devel