MLton / Code (SVN - read-only) Commit Log

Commit	Date
[r5839] by fluet Don't hardcode the expected target sizes. This has some overlap with the LookupConstants functionality; unfortunately, we need to know the right sizes for mplimb, objptr, header, and seqIndex in order to parseAndElaborate the Basis Library (via the MLB path variables) to discover the other constants.	2007-08-09 01:33:59	Tree
[r5838] by fluet Type alias for SML/NJ compile	2007-08-09 01:04:29	Tree
[r5837] by fluet Update benchmark counts for benchmarks taking < 10s; don't want startup/shutdown costs to dominate	2007-08-08 23:45:59	Tree
[r5836] by fluet New defaults for inlining. This doesn't uniformly improve performance, but it eliminates some particularly bad x86_64 benchmark performance problems (at the expense of introducing a performance problem with output1). A more systematic inlining heuristic should be on the TODO list, but this is passable for the Aug. 2007 release. Benchmarks: ============================================================ SHADOW (Dual-processor AMD Opteron 2.0GHz, 8GB Memory, Fedora Core 7) Linux shadow 2.6.22.1-41.fc7 #1 SMP Fri Jul 27 18:21:43 EDT 2007 x86_64 x86_64 x86_64 GNU/Linux gcc (GCC) 4.1.2 20070502 (Red Hat 4.1.2-12) MLton0 -- ~/devel/mlton/mlton-20051202-1.i386-linux/build/bin/mlton MLton1 -- ~/devel/mlton/mlton.svn.trunk/build/bin/mlton MLton2 -- ~/devel/mlton/mlton.svn.trunk.inline/build/bin/mlton run time ratio benchmark MLton0 MLton1 MLton2 barnes-hut 1.00 0.92 0.92 boyer 1.00 1.24 1.13 checksum 1.00 0.61 0.42 count-graphs 1.00 2.24 1.39 DLXSimulator 1.00 1.59 1.51 fft 1.00 1.01 1.00 fib 1.00 1.08 1.19 flat-array 1.00 0.99 0.99 hamlet 1.00 1.72 1.40 imp-for 1.00 0.91 0.92 knuth-bendix 1.00 1.43 1.33 lexgen 1.00 1.48 1.23 life 1.00 1.64 1.16 logic 1.00 1.34 1.52 mandelbrot 1.00 0.86 0.90 matrix-multiply 1.00 0.82 0.80 md5 1.00 1.93 0.10 merge 1.00 1.64 1.72 mlyacc 1.00 1.52 1.42 model-elimination 1.00 1.41 1.30 mpuz 1.00 1.67 1.34 nucleic 1.00 1.11 0.94 output1 1.00 1.07 1.90 peek 1.00 1.74 1.75 psdes-random 1.00 1.16 1.13 ratio-regions 1.00 1.22 0.77 ray 1.00 1.11 1.11 raytrace 1.00 1.10 1.30 simple 1.00 1.40 1.11 smith-normal-form 1.00 0.82 0.82 tailfib 1.00 1.22 1.23 tak 1.00 1.22 1.25 tensor 1.00 1.40 1.31 tsp 1.00 1.06 1.05 tyan 1.00 1.54 1.44 vector-concat 1.00 0.82 0.97 vector-rev 1.00 1.17 1.14 vliw 1.00 2.04 1.85 wc-input1 1.00 1.23 1.23 wc-scanStream 1.00 1.01 1.00 zebra 1.00 1.54 1.55 zern 1.00 1.21 1.39 size benchmark MLton0 MLton1 MLton2 barnes-hut 119,430 168,798 167,742 boyer 144,869 215,393 213,649 checksum 65,597 95,825 93,681 count-graphs 76,789 120,593 119,937 DLXSimulator 144,257 195,996 196,204 fft 75,108 117,087 117,487 fib 58,029 95,873 93,569 flat-array 58,077 95,857 93,073 hamlet 1,246,688 1,554,825 1,504,601 imp-for 57,901 95,745 93,393 knuth-bendix 124,269 171,116 171,980 lexgen 216,614 283,867 285,419 life 76,117 119,665 118,081 logic 115,605 167,345 177,809 mandelbrot 58,013 95,713 93,233 matrix-multiply 59,540 96,369 95,249 md5 93,317 128,172 127,036 merge 59,597 97,985 94,897 mlyacc 516,486 644,507 662,011 model-elimination 646,219 860,898 850,418 mpuz 60,845 99,569 99,857 nucleic 207,760 271,680 269,104 output1 96,519 135,120 136,816 peek 91,969 127,244 132,604 psdes-random 58,669 96,721 96,465 ratio-regions 83,997 128,961 120,977 ray 198,158 256,920 245,128 raytrace 278,187 359,698 372,946 simple 236,873 329,497 343,593 smith-normal-form 192,389 266,972 272,076 tailfib 57,709 95,473 93,137 tak 58,109 95,905 93,585 tensor 113,380 155,875 162,515 tsp 97,789 139,083 139,627 tyan 149,593 205,276 212,556 vector-concat 59,301 97,441 94,881 vector-rev 58,545 96,753 94,641 vliw 406,705 522,650 519,402 wc-input1 117,601 164,282 159,338 wc-scanStream 124,769 173,498 170,106 zebra 156,293 212,476 212,700 zern 104,246 132,902 132,438 compile time benchmark MLton0 MLton1 MLton2 barnes-hut 3.99 9.70 9.28 boyer 4.19 10.31 9.80 checksum 2.85 7.51 7.11 count-graphs 3.24 8.12 8.01 DLXSimulator 4.47 10.72 10.40 fft 3.06 8.10 7.75 fib 2.85 7.51 7.15 flat-array 2.88 7.33 7.14 hamlet 42.27 48.64 46.43 imp-for 2.91 7.34 7.26 knuth-bendix 3.70 9.06 9.01 lexgen 5.19 12.10 11.70 life 3.11 8.20 7.82 logic 3.73 9.01 9.24 mandelbrot 2.87 7.16 7.08 matrix-multiply 2.89 7.70 7.33 md5 3.21 8.48 8.01 merge 2.85 7.35 7.27 mlyacc 11.72 25.75 25.73 model-elimination 11.77 25.04 24.28 mpuz 2.93 7.47 7.33 nucleic 5.34 11.14 10.88 output1 3.23 8.22 8.17 peek 3.16 8.39 8.01 psdes-random 2.89 7.56 7.23 ratio-regions 3.46 8.58 8.56 ray 4.64 11.48 11.08 raytrace 6.38 14.74 14.35 simple 5.39 12.97 12.65 smith-normal-form 4.64 11.33 10.91 tailfib 2.87 7.41 7.27 tak 2.82 7.55 7.12 tensor 4.16 10.01 9.78 tsp 3.37 8.60 8.48 tyan 4.35 10.35 10.45 vector-concat 2.90 7.52 7.16 vector-rev 2.89 7.13 7.07 vliw 8.65 19.47 19.15 wc-input1 3.57 9.10 8.74 wc-scanStream 3.92 9.29 8.98 zebra 4.35 10.65 10.45 zern 3.20 8.49 8.09 run time benchmark MLton0 MLton1 MLton2 barnes-hut 16.94 15.56 15.62 boyer 16.23 20.20 18.37 checksum 43.15 26.34 18.27 count-graphs 21.14 47.42 29.43 DLXSimulator 16.93 26.96 25.56 fft 14.30 14.37 14.36 fib 34.67 37.47 41.13 flat-array 8.49 8.38 8.36 hamlet 15.86 27.24 22.22 imp-for 28.84 26.21 26.65 knuth-bendix 17.72 25.42 23.62 lexgen 19.64 29.15 24.24 life 8.40 13.81 9.77 logic 18.15 24.37 27.68 mandelbrot 23.98 20.69 21.48 matrix-multiply 3.36 2.76 2.69 md5 32.35 62.47 3.38 merge 14.15 23.20 24.40 mlyacc 17.63 26.83 24.97 model-elimination 28.08 39.51 36.58 mpuz 21.87 36.41 29.32 nucleic 16.66 18.48 15.71 output1 7.22 7.71 13.70 peek 19.96 34.74 34.89 psdes-random 16.29 18.90 18.34 ratio-regions 24.32 29.79 18.63 ray 15.23 16.93 16.96 raytrace 16.18 17.87 21.07 simple 20.26 28.33 22.48 smith-normal-form 10.24 8.37 8.39 tailfib 19.36 23.63 23.72 tak 13.00 15.91 16.22 tensor 17.30 24.26 22.70 tsp 20.38 21.67 21.31 tyan 18.24 28.11 26.34 vector-concat 30.09 24.75 29.19 vector-rev 20.34 23.85 23.23 vliw 17.92 36.57 33.06 wc-input1 28.28 34.82 34.88 wc-scanStream 14.68 14.77 14.72 zebra 26.30 40.57 40.67 zern 17.81 21.50 24.81 mlton.svn.trunk MLton MLTONVERSION (built Wed Aug 08 14:37:03 2007 on shadow) MLton starting Compile SML starting pre codegen starting pre codegen finished in 184.89 + 86.68 (32% GC) amd64 code gen starting amd64 code gen finished in 329.74 + 6.21 (2% GC) Compile SML finished in 514.72 + 92.89 (15% GC) Compile and Assemble starting Compile and Assemble finished in 20.98 + 0.76 (3% GC) Link starting Link finished in 0.89 + 0.00 (0% GC) MLton finished in 537.15 + 94.66 (15% GC) mlton.svn.trunk.inline MLton MLTONVERSION (built Wed Aug 08 15:07:30 2007 on shadow) MLton starting Compile SML starting pre codegen starting pre codegen finished in 193.23 + 93.84 (33% GC) amd64 code gen starting amd64 code gen finished in 282.40 + 14.68 (5% GC) Compile SML finished in 475.70 + 108.52 (19% GC) Compile and Assemble starting Compile and Assemble finished in 19.57 + 0.03 (0% GC) Link starting Link finished in 0.90 + 0.00 (0% GC) MLton finished in 497.01 + 109.50 (18% GC) ============================================================ FENRIR (Dual-processor Dual-core Intel Xeon 2.0GHz, 2GB Memory, MacOS 10.4) Darwin fenrir.uchicago.edu 8.10.1 Darwin Kernel Version 8.10.1: Wed May 23 16:33:00 PDT 2007; root:xnu-792.22.5~1/RELEASE_I386 i386 i386 i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build 5367) MLton0 -- ~/devel/mlton/mlton-20061119-1.x86-darwin/build/bin/mlton MLton1 -- ~/devel/mlton/mlton.svn.trunk/build/bin/mlton MLton2 -- ~/devel/mlton/mlton.svn.trunk.inline/build/bin/mlton run time ratio benchmark MLton0 MLton1 MLton2 barnes-hut 1.00 1.05 1.02 boyer 1.00 0.89 0.79 checksum 1.00 1.09 1.06 count-graphs 1.00 1.72 1.02 DLXSimulator 1.00 1.03 1.04 fft 1.00 1.01 1.00 fib 1.00 0.99 0.88 flat-array 1.00 1.00 0.99 hamlet 1.00 1.12 0.94 imp-for 1.00 1.00 1.00 knuth-bendix 1.00 1.06 1.03 lexgen 1.00 1.01 0.92 life 1.00 0.96 0.96 logic 1.00 1.08 1.06 mandelbrot 1.00 1.00 1.15 matrix-multiply 1.00 0.99 0.65 md5 1.00 1.05 0.08 merge 1.00 0.94 0.97 mlyacc 1.00 1.03 1.07 model-elimination 1.00 1.08 1.08 mpuz 1.00 1.00 1.01 nucleic 1.00 1.16 0.95 output1 1.00 0.98 0.69 peek 1.00 0.94 0.94 psdes-random 1.00 0.94 0.91 ratio-regions 1.00 1.02 0.68 ray 1.00 0.99 0.95 raytrace 1.00 1.01 0.97 simple 1.00 0.58 0.60 smith-normal-form 1.00 1.00 1.00 tailfib 1.00 1.00 1.00 tak 1.00 0.91 0.92 tensor 1.00 1.00 1.00 tsp 1.00 1.03 1.03 tyan 1.00 1.05 1.06 vector-concat 1.00 1.00 1.12 vector-rev 1.00 1.00 1.01 vliw 1.00 1.46 1.24 wc-input1 1.00 1.12 1.00 wc-scanStream 1.00 1.04 1.04 zebra 1.00 1.01 1.06 zern 1.00 1.00 1.04 size benchmark MLton0 MLton1 MLton2 barnes-hut 114,688 163,840 167,936 boyer 155,648 204,800 204,800 checksum 61,440 106,496 106,496 count-graphs 77,824 126,976 126,976 DLXSimulator 147,456 196,608 196,608 fft 77,824 122,880 126,976 fib 57,344 106,496 102,400 flat-array 57,344 106,496 102,400 hamlet 1,327,104 1,359,872 1,335,296 imp-for 57,344 106,496 102,400 knuth-bendix 126,976 176,128 176,128 lexgen 221,184 270,336 274,432 life 77,824 126,976 126,976 logic 118,784 167,936 172,032 mandelbrot 61,440 106,496 102,400 matrix-multiply 61,440 106,496 106,496 md5 90,112 139,264 139,264 merge 61,440 106,496 106,496 mlyacc 540,672 589,824 602,112 model-elimination 684,032 720,896 729,088 mpuz 61,440 110,592 110,592 nucleic 229,376 282,624 278,528 output1 94,208 143,360 143,360 peek 90,112 139,264 143,360 psdes-random 61,440 106,496 106,496 ratio-regions 86,016 135,168 131,072 ray 208,896 237,568 237,568 raytrace 290,816 319,488 331,776 simple 241,664 286,720 307,200 smith-normal-form 208,896 258,048 262,144 tailfib 57,344 106,496 102,400 tak 57,344 106,496 102,400 tensor 110,592 163,840 167,936 tsp 94,208 143,360 143,360 tyan 151,552 200,704 208,896 vector-concat 61,440 106,496 106,496 vector-rev 57,344 106,496 106,496 vliw 430,080 450,560 466,944 wc-input1 118,784 167,936 167,936 wc-scanStream 126,976 172,032 172,032 zebra 159,744 208,896 208,896 zern 106,496 135,168 135,168 compile time benchmark MLton0 MLton1 MLton2 barnes-hut 3.35 5.22 5.18 boyer 3.42 5.25 5.33 checksum 2.44 4.16 4.22 count-graphs 2.70 4.44 4.57 DLXSimulator 3.61 5.53 5.58 fft 2.61 4.34 4.48 fib 2.44 4.14 4.21 flat-array 2.43 4.18 4.22 hamlet 32.15 21.47 21.24 imp-for 2.44 4.16 4.23 knuth-bendix 3.07 4.95 5.03 lexgen 4.21 6.29 6.43 life 2.64 4.38 4.47 logic 3.10 4.88 5.18 mandelbrot 2.47 4.20 4.26 matrix-multiply 2.47 4.20 4.28 md5 2.72 4.58 4.61 merge 2.47 4.17 4.27 mlyacc 8.66 13.66 13.86 model-elimination 9.07 11.88 12.59 mpuz 2.51 4.24 4.31 nucleic 4.39 6.23 6.30 output1 2.71 4.58 4.63 peek 2.69 4.49 4.62 psdes-random 2.48 4.22 4.29 ratio-regions 2.88 4.78 4.92 ray 3.84 6.01 6.12 raytrace 5.01 7.33 7.83 simple 4.36 6.74 6.96 smith-normal-form 3.72 5.80 5.91 tailfib 2.47 4.20 4.26 tak 2.47 4.20 4.27 tensor 3.39 5.29 5.49 tsp 2.86 4.71 4.84 tyan 3.65 5.58 5.77 vector-concat 2.49 4.23 4.31 vector-rev 2.49 4.22 4.28 vliw 6.53 9.53 10.04 wc-input1 3.02 4.92 4.96 wc-scanStream 3.10 5.03 5.10 zebra 3.55 5.84 5.90 zern 2.74 4.55 4.66 run time benchmark MLton0 MLton1 MLton2 barnes-hut 11.68 12.21 11.93 boyer 11.35 10.10 8.92 checksum 28.52 31.15 30.33 count-graphs 12.04 20.73 12.32 DLXSimulator 10.67 10.99 11.05 fft 12.43 12.52 12.38 fib 21.54 21.36 18.98 flat-array 4.62 4.62 4.58 hamlet 10.28 11.46 9.65 imp-for 13.37 13.38 13.38 knuth-bendix 11.27 11.96 11.63 lexgen 10.90 11.03 10.07 life 5.81 5.58 5.56 logic 10.78 11.63 11.44 mandelbrot 16.54 16.54 18.99 matrix-multiply 1.90 1.89 1.24 md5 23.63 24.86 1.87 merge 8.79 8.30 8.52 mlyacc 12.45 12.85 13.28 model-elimination 21.31 23.08 22.94 mpuz 12.38 12.39 12.54 nucleic 11.54 13.41 10.92 output1 6.82 6.66 4.73 peek 21.64 20.28 20.40 psdes-random 14.39 13.54 13.10 ratio-regions 10.66 10.87 7.29 ray 14.18 14.05 13.51 raytrace 11.66 11.77 11.30 simple 19.68 11.41 11.90 smith-normal-form 14.08 14.13 14.13 tailfib 13.62 13.62 13.62 tak 7.42 6.76 6.85 tensor 20.93 20.93 20.93 tsp 22.42 23.13 23.10 tyan 11.49 12.03 12.16 vector-concat 16.54 16.50 18.54 vector-rev 9.68 9.67 9.76 vliw 9.83 14.32 12.17 wc-input1 14.36 16.09 14.29 wc-scanStream 8.98 9.33 9.31 zebra 15.15 15.34 16.07 zern 14.04 14.06 14.58 mlton.svn.trunk MLton MLTONVERSION (built Wed Aug 08 14:25:28 2007 on fenrir.uchicago.edu) MLton starting Compile SML starting pre codegen starting pre codegen finished in 67.83 + 57.61 (46% GC) x86 code gen starting x86 code gen finished in 116.47 + 9.56 (8% GC) Compile SML finished in 184.35 + 67.17 (27% GC) Compile and Assemble starting Compile and Assemble finished in 13.85 + 0.05 (0% GC) Link starting Link finished in 0.42 + 0.00 (1% GC) MLton finished in 198.71 + 67.69 (25% GC) mlton.svn.trunk.inline MLton MLTONVERSION (built Wed Aug 08 14:43:39 2007 on fenrir.uchicago.edu) MLton starting Compile SML starting pre codegen starting pre codegen finished in 69.91 + 67.70 (49% GC) x86 code gen starting x86 code gen finished in 117.76 + 5.19 (4% GC) Compile SML finished in 187.71 + 72.89 (28% GC) Compile and Assemble starting Compile and Assemble finished in 13.67 + 0.15 (1% GC) Link starting Link finished in 0.42 + 0.00 (0% GC) MLton finished in 201.89 + 73.77 (27% GC)	2007-08-08 23:40:56	Tree
[r5835] by vesak Removed join-generics, which is no longer used. Reorganized the exports.	2007-08-08 17:25:24	Tree
[r5834] by vesak Renamed several signatures, structures, and functors in the generic library for brevity and, more importantly, for improved clarity.	2007-08-08 08:35:36	Tree
[r5833] by vesak Added mapState to Reader and Writer.	2007-08-08 08:31:28	Tree
[r5832] by vesak Added a WithDebug functor for checking the uniqueness of labels and constructors.	2007-08-07 19:09:19	Tree
[r5831] by vesak Replaced Tie.unit and Tie.option by the more general id combinator.	2007-08-07 18:19:06	Tree
[r5830] by vesak Examples.	2007-08-07 08:10:39	Tree

[r5839] by fluet

Don't hardcode the expected target sizes.

This has some overlap with the LookupConstants functionality;
unfortunately, we need to know the right sizes for mplimb, objptr,
header, and seqIndex in order to parseAndElaborate the Basis Library
(via the MLB path variables) to discover the other constants.

2007-08-09 01:33:59

Tree

[r5838] by fluet

Type alias for SML/NJ compile

2007-08-09 01:04:29

Tree

[r5837] by fluet

Update benchmark counts for benchmarks taking < 10s; don't want startup/shutdown costs to dominate

2007-08-08 23:45:59

Tree

[r5836] by fluet

New defaults for inlining. This doesn't uniformly improve
performance, but it eliminates some particularly bad x86_64 benchmark
performance problems (at the expense of introducing a performance
problem with output1).

A more systematic inlining heuristic should be on the TODO list, but
this is passable for the Aug. 2007 release.

Benchmarks:
============================================================

SHADOW (Dual-processor AMD Opteron 2.0GHz, 8GB Memory, Fedora Core 7)

Linux shadow 2.6.22.1-41.fc7 #1 SMP Fri Jul 27 18:21:43 EDT 2007 x86_64 x86_64 x86_64 GNU/Linux

gcc (GCC) 4.1.2 20070502 (Red Hat 4.1.2-12)

MLton0 -- ~/devel/mlton/mlton-20051202-1.i386-linux/build/bin/mlton
MLton1 -- ~/devel/mlton/mlton.svn.trunk/build/bin/mlton
MLton2 -- ~/devel/mlton/mlton.svn.trunk.inline/build/bin/mlton
run time ratio
benchmark MLton0 MLton1 MLton2
barnes-hut 1.00 0.92 0.92
boyer 1.00 1.24 1.13
checksum 1.00 0.61 0.42
count-graphs 1.00 2.24 1.39
DLXSimulator 1.00 1.59 1.51
fft 1.00 1.01 1.00
fib 1.00 1.08 1.19
flat-array 1.00 0.99 0.99
hamlet 1.00 1.72 1.40
imp-for 1.00 0.91 0.92
knuth-bendix 1.00 1.43 1.33
lexgen 1.00 1.48 1.23
life 1.00 1.64 1.16
logic 1.00 1.34 1.52
mandelbrot 1.00 0.86 0.90
matrix-multiply 1.00 0.82 0.80
md5 1.00 1.93 0.10
merge 1.00 1.64 1.72
mlyacc 1.00 1.52 1.42
model-elimination 1.00 1.41 1.30
mpuz 1.00 1.67 1.34
nucleic 1.00 1.11 0.94
output1 1.00 1.07 1.90
peek 1.00 1.74 1.75
psdes-random 1.00 1.16 1.13
ratio-regions 1.00 1.22 0.77
ray 1.00 1.11 1.11
raytrace 1.00 1.10 1.30
simple 1.00 1.40 1.11
smith-normal-form 1.00 0.82 0.82
tailfib 1.00 1.22 1.23
tak 1.00 1.22 1.25
tensor 1.00 1.40 1.31
tsp 1.00 1.06 1.05
tyan 1.00 1.54 1.44
vector-concat 1.00 0.82 0.97
vector-rev 1.00 1.17 1.14
vliw 1.00 2.04 1.85
wc-input1 1.00 1.23 1.23
wc-scanStream 1.00 1.01 1.00
zebra 1.00 1.54 1.55
zern 1.00 1.21 1.39
size
benchmark MLton0 MLton1 MLton2
barnes-hut 119,430 168,798 167,742
boyer 144,869 215,393 213,649
checksum 65,597 95,825 93,681
count-graphs 76,789 120,593 119,937
DLXSimulator 144,257 195,996 196,204
fft 75,108 117,087 117,487
fib 58,029 95,873 93,569
flat-array 58,077 95,857 93,073
hamlet 1,246,688 1,554,825 1,504,601
imp-for 57,901 95,745 93,393
knuth-bendix 124,269 171,116 171,980
lexgen 216,614 283,867 285,419
life 76,117 119,665 118,081
logic 115,605 167,345 177,809
mandelbrot 58,013 95,713 93,233
matrix-multiply 59,540 96,369 95,249
md5 93,317 128,172 127,036
merge 59,597 97,985 94,897
mlyacc 516,486 644,507 662,011
model-elimination 646,219 860,898 850,418
mpuz 60,845 99,569 99,857
nucleic 207,760 271,680 269,104
output1 96,519 135,120 136,816
peek 91,969 127,244 132,604
psdes-random 58,669 96,721 96,465
ratio-regions 83,997 128,961 120,977
ray 198,158 256,920 245,128
raytrace 278,187 359,698 372,946
simple 236,873 329,497 343,593
smith-normal-form 192,389 266,972 272,076
tailfib 57,709 95,473 93,137
tak 58,109 95,905 93,585
tensor 113,380 155,875 162,515
tsp 97,789 139,083 139,627
tyan 149,593 205,276 212,556
vector-concat 59,301 97,441 94,881
vector-rev 58,545 96,753 94,641
vliw 406,705 522,650 519,402
wc-input1 117,601 164,282 159,338
wc-scanStream 124,769 173,498 170,106
zebra 156,293 212,476 212,700
zern 104,246 132,902 132,438
compile time
benchmark MLton0 MLton1 MLton2
barnes-hut 3.99 9.70 9.28
boyer 4.19 10.31 9.80
checksum 2.85 7.51 7.11
count-graphs 3.24 8.12 8.01
DLXSimulator 4.47 10.72 10.40
fft 3.06 8.10 7.75
fib 2.85 7.51 7.15
flat-array 2.88 7.33 7.14
hamlet 42.27 48.64 46.43
imp-for 2.91 7.34 7.26
knuth-bendix 3.70 9.06 9.01
lexgen 5.19 12.10 11.70
life 3.11 8.20 7.82
logic 3.73 9.01 9.24
mandelbrot 2.87 7.16 7.08
matrix-multiply 2.89 7.70 7.33
md5 3.21 8.48 8.01
merge 2.85 7.35 7.27
mlyacc 11.72 25.75 25.73
model-elimination 11.77 25.04 24.28
mpuz 2.93 7.47 7.33
nucleic 5.34 11.14 10.88
output1 3.23 8.22 8.17
peek 3.16 8.39 8.01
psdes-random 2.89 7.56 7.23
ratio-regions 3.46 8.58 8.56
ray 4.64 11.48 11.08
raytrace 6.38 14.74 14.35
simple 5.39 12.97 12.65
smith-normal-form 4.64 11.33 10.91
tailfib 2.87 7.41 7.27
tak 2.82 7.55 7.12
tensor 4.16 10.01 9.78
tsp 3.37 8.60 8.48
tyan 4.35 10.35 10.45
vector-concat 2.90 7.52 7.16
vector-rev 2.89 7.13 7.07
vliw 8.65 19.47 19.15
wc-input1 3.57 9.10 8.74
wc-scanStream 3.92 9.29 8.98
zebra 4.35 10.65 10.45
zern 3.20 8.49 8.09
run time
benchmark MLton0 MLton1 MLton2
barnes-hut 16.94 15.56 15.62
boyer 16.23 20.20 18.37
checksum 43.15 26.34 18.27
count-graphs 21.14 47.42 29.43
DLXSimulator 16.93 26.96 25.56
fft 14.30 14.37 14.36
fib 34.67 37.47 41.13
flat-array 8.49 8.38 8.36
hamlet 15.86 27.24 22.22
imp-for 28.84 26.21 26.65
knuth-bendix 17.72 25.42 23.62
lexgen 19.64 29.15 24.24
life 8.40 13.81 9.77
logic 18.15 24.37 27.68
mandelbrot 23.98 20.69 21.48
matrix-multiply 3.36 2.76 2.69
md5 32.35 62.47 3.38
merge 14.15 23.20 24.40
mlyacc 17.63 26.83 24.97
model-elimination 28.08 39.51 36.58
mpuz 21.87 36.41 29.32
nucleic 16.66 18.48 15.71
output1 7.22 7.71 13.70
peek 19.96 34.74 34.89
psdes-random 16.29 18.90 18.34
ratio-regions 24.32 29.79 18.63
ray 15.23 16.93 16.96
raytrace 16.18 17.87 21.07
simple 20.26 28.33 22.48
smith-normal-form 10.24 8.37 8.39
tailfib 19.36 23.63 23.72
tak 13.00 15.91 16.22
tensor 17.30 24.26 22.70
tsp 20.38 21.67 21.31
tyan 18.24 28.11 26.34
vector-concat 30.09 24.75 29.19
vector-rev 20.34 23.85 23.23
vliw 17.92 36.57 33.06
wc-input1 28.28 34.82 34.88
wc-scanStream 14.68 14.77 14.72
zebra 26.30 40.57 40.67
zern 17.81 21.50 24.81

mlton.svn.trunk
MLton MLTONVERSION (built Wed Aug 08 14:37:03 2007 on shadow)
MLton starting
Compile SML starting
pre codegen starting
pre codegen finished in 184.89 + 86.68 (32% GC)
amd64 code gen starting
amd64 code gen finished in 329.74 + 6.21 (2% GC)
Compile SML finished in 514.72 + 92.89 (15% GC)
Compile and Assemble starting
Compile and Assemble finished in 20.98 + 0.76 (3% GC)
Link starting
Link finished in 0.89 + 0.00 (0% GC)
MLton finished in 537.15 + 94.66 (15% GC)
mlton.svn.trunk.inline
MLton MLTONVERSION (built Wed Aug 08 15:07:30 2007 on shadow)
MLton starting
Compile SML starting
pre codegen starting
pre codegen finished in 193.23 + 93.84 (33% GC)
amd64 code gen starting
amd64 code gen finished in 282.40 + 14.68 (5% GC)
Compile SML finished in 475.70 + 108.52 (19% GC)
Compile and Assemble starting
Compile and Assemble finished in 19.57 + 0.03 (0% GC)
Link starting
Link finished in 0.90 + 0.00 (0% GC)
MLton finished in 497.01 + 109.50 (18% GC)

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

FENRIR (Dual-processor Dual-core Intel Xeon 2.0GHz, 2GB Memory, MacOS 10.4)

Darwin fenrir.uchicago.edu 8.10.1 Darwin Kernel Version 8.10.1: Wed May 23 16:33:00 PDT 2007; root:xnu-792.22.5~1/RELEASE_I386 i386 i386

i686-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build 5367)

MLton0 -- ~/devel/mlton/mlton-20061119-1.x86-darwin/build/bin/mlton
MLton1 -- ~/devel/mlton/mlton.svn.trunk/build/bin/mlton
MLton2 -- ~/devel/mlton/mlton.svn.trunk.inline/build/bin/mlton
run time ratio
benchmark MLton0 MLton1 MLton2
barnes-hut 1.00 1.05 1.02
boyer 1.00 0.89 0.79
checksum 1.00 1.09 1.06
count-graphs 1.00 1.72 1.02
DLXSimulator 1.00 1.03 1.04
fft 1.00 1.01 1.00
fib 1.00 0.99 0.88
flat-array 1.00 1.00 0.99
hamlet 1.00 1.12 0.94
imp-for 1.00 1.00 1.00
knuth-bendix 1.00 1.06 1.03
lexgen 1.00 1.01 0.92
life 1.00 0.96 0.96
logic 1.00 1.08 1.06
mandelbrot 1.00 1.00 1.15
matrix-multiply 1.00 0.99 0.65
md5 1.00 1.05 0.08
merge 1.00 0.94 0.97
mlyacc 1.00 1.03 1.07
model-elimination 1.00 1.08 1.08
mpuz 1.00 1.00 1.01
nucleic 1.00 1.16 0.95
output1 1.00 0.98 0.69
peek 1.00 0.94 0.94
psdes-random 1.00 0.94 0.91
ratio-regions 1.00 1.02 0.68
ray 1.00 0.99 0.95
raytrace 1.00 1.01 0.97
simple 1.00 0.58 0.60
smith-normal-form 1.00 1.00 1.00
tailfib 1.00 1.00 1.00
tak 1.00 0.91 0.92
tensor 1.00 1.00 1.00
tsp 1.00 1.03 1.03
tyan 1.00 1.05 1.06
vector-concat 1.00 1.00 1.12
vector-rev 1.00 1.00 1.01
vliw 1.00 1.46 1.24
wc-input1 1.00 1.12 1.00
wc-scanStream 1.00 1.04 1.04
zebra 1.00 1.01 1.06
zern 1.00 1.00 1.04
size
benchmark MLton0 MLton1 MLton2
barnes-hut 114,688 163,840 167,936
boyer 155,648 204,800 204,800
checksum 61,440 106,496 106,496
count-graphs 77,824 126,976 126,976
DLXSimulator 147,456 196,608 196,608
fft 77,824 122,880 126,976
fib 57,344 106,496 102,400
flat-array 57,344 106,496 102,400
hamlet 1,327,104 1,359,872 1,335,296
imp-for 57,344 106,496 102,400
knuth-bendix 126,976 176,128 176,128
lexgen 221,184 270,336 274,432
life 77,824 126,976 126,976
logic 118,784 167,936 172,032
mandelbrot 61,440 106,496 102,400
matrix-multiply 61,440 106,496 106,496
md5 90,112 139,264 139,264
merge 61,440 106,496 106,496
mlyacc 540,672 589,824 602,112
model-elimination 684,032 720,896 729,088
mpuz 61,440 110,592 110,592
nucleic 229,376 282,624 278,528
output1 94,208 143,360 143,360
peek 90,112 139,264 143,360
psdes-random 61,440 106,496 106,496
ratio-regions 86,016 135,168 131,072
ray 208,896 237,568 237,568
raytrace 290,816 319,488 331,776
simple 241,664 286,720 307,200
smith-normal-form 208,896 258,048 262,144
tailfib 57,344 106,496 102,400
tak 57,344 106,496 102,400
tensor 110,592 163,840 167,936
tsp 94,208 143,360 143,360
tyan 151,552 200,704 208,896
vector-concat 61,440 106,496 106,496
vector-rev 57,344 106,496 106,496
vliw 430,080 450,560 466,944
wc-input1 118,784 167,936 167,936
wc-scanStream 126,976 172,032 172,032
zebra 159,744 208,896 208,896
zern 106,496 135,168 135,168
compile time
benchmark MLton0 MLton1 MLton2
barnes-hut 3.35 5.22 5.18
boyer 3.42 5.25 5.33
checksum 2.44 4.16 4.22
count-graphs 2.70 4.44 4.57
DLXSimulator 3.61 5.53 5.58
fft 2.61 4.34 4.48
fib 2.44 4.14 4.21
flat-array 2.43 4.18 4.22
hamlet 32.15 21.47 21.24
imp-for 2.44 4.16 4.23
knuth-bendix 3.07 4.95 5.03
lexgen 4.21 6.29 6.43
life 2.64 4.38 4.47
logic 3.10 4.88 5.18
mandelbrot 2.47 4.20 4.26
matrix-multiply 2.47 4.20 4.28
md5 2.72 4.58 4.61
merge 2.47 4.17 4.27
mlyacc 8.66 13.66 13.86
model-elimination 9.07 11.88 12.59
mpuz 2.51 4.24 4.31
nucleic 4.39 6.23 6.30
output1 2.71 4.58 4.63
peek 2.69 4.49 4.62
psdes-random 2.48 4.22 4.29
ratio-regions 2.88 4.78 4.92
ray 3.84 6.01 6.12
raytrace 5.01 7.33 7.83
simple 4.36 6.74 6.96
smith-normal-form 3.72 5.80 5.91
tailfib 2.47 4.20 4.26
tak 2.47 4.20 4.27
tensor 3.39 5.29 5.49
tsp 2.86 4.71 4.84
tyan 3.65 5.58 5.77
vector-concat 2.49 4.23 4.31
vector-rev 2.49 4.22 4.28
vliw 6.53 9.53 10.04
wc-input1 3.02 4.92 4.96
wc-scanStream 3.10 5.03 5.10
zebra 3.55 5.84 5.90
zern 2.74 4.55 4.66
run time
benchmark MLton0 MLton1 MLton2
barnes-hut 11.68 12.21 11.93
boyer 11.35 10.10 8.92
checksum 28.52 31.15 30.33
count-graphs 12.04 20.73 12.32
DLXSimulator 10.67 10.99 11.05
fft 12.43 12.52 12.38
fib 21.54 21.36 18.98
flat-array 4.62 4.62 4.58
hamlet 10.28 11.46 9.65
imp-for 13.37 13.38 13.38
knuth-bendix 11.27 11.96 11.63
lexgen 10.90 11.03 10.07
life 5.81 5.58 5.56
logic 10.78 11.63 11.44
mandelbrot 16.54 16.54 18.99
matrix-multiply 1.90 1.89 1.24
md5 23.63 24.86 1.87
merge 8.79 8.30 8.52
mlyacc 12.45 12.85 13.28
model-elimination 21.31 23.08 22.94
mpuz 12.38 12.39 12.54
nucleic 11.54 13.41 10.92
output1 6.82 6.66 4.73
peek 21.64 20.28 20.40
psdes-random 14.39 13.54 13.10
ratio-regions 10.66 10.87 7.29
ray 14.18 14.05 13.51
raytrace 11.66 11.77 11.30
simple 19.68 11.41 11.90
smith-normal-form 14.08 14.13 14.13
tailfib 13.62 13.62 13.62
tak 7.42 6.76 6.85
tensor 20.93 20.93 20.93
tsp 22.42 23.13 23.10
tyan 11.49 12.03 12.16
vector-concat 16.54 16.50 18.54
vector-rev 9.68 9.67 9.76
vliw 9.83 14.32 12.17
wc-input1 14.36 16.09 14.29
wc-scanStream 8.98 9.33 9.31
zebra 15.15 15.34 16.07
zern 14.04 14.06 14.58

mlton.svn.trunk
MLton MLTONVERSION (built Wed Aug 08 14:25:28 2007 on fenrir.uchicago.edu)
MLton starting
Compile SML starting
pre codegen starting
pre codegen finished in 67.83 + 57.61 (46% GC)
x86 code gen starting
x86 code gen finished in 116.47 + 9.56 (8% GC)
Compile SML finished in 184.35 + 67.17 (27% GC)
Compile and Assemble starting
Compile and Assemble finished in 13.85 + 0.05 (0% GC)
Link starting
Link finished in 0.42 + 0.00 (1% GC)
MLton finished in 198.71 + 67.69 (25% GC)
mlton.svn.trunk.inline
MLton MLTONVERSION (built Wed Aug 08 14:43:39 2007 on fenrir.uchicago.edu)
MLton starting
Compile SML starting
pre codegen starting
pre codegen finished in 69.91 + 67.70 (49% GC)
x86 code gen starting
x86 code gen finished in 117.76 + 5.19 (4% GC)
Compile SML finished in 187.71 + 72.89 (28% GC)
Compile and Assemble starting
Compile and Assemble finished in 13.67 + 0.15 (1% GC)
Link starting
Link finished in 0.42 + 0.00 (0% GC)
MLton finished in 201.89 + 73.77 (27% GC)

2007-08-08 23:40:56

Tree

[r5835] by vesak

Removed join-generics, which is no longer used.
Reorganized the exports.

2007-08-08 17:25:24

Tree

[r5834] by vesak

Renamed several signatures, structures, and functors in the generic
library for brevity and, more importantly, for improved clarity.

2007-08-08 08:35:36

Tree

[r5833] by vesak

Added mapState to Reader and Writer.

2007-08-08 08:31:28

Tree

[r5832] by vesak

Added a WithDebug functor for checking the uniqueness of labels and
constructors.

2007-08-07 19:09:19

Tree

[r5831] by vesak

Replaced Tie.unit and Tie.option by the more general id combinator.

2007-08-07 18:19:06

Tree

[r5830] by vesak

Examples.

2007-08-07 08:10:39

Tree

MLton Code (SVN - read-only)

A whole-program optimizing compiler for Standard ML

Code (SVN - read-only) Commit Log