Menu
▾
▴
[Jikesrvm-regression] [AIX] regression SUCCEEDED
|
From: Jalapeno (M. H. <jal...@wa...> - 2002-02-21 16:24:19
|
cvs checkout performed on Wed Feb 20 23:06:49 EST 2002
Ran 167
Passed 167
Failed 0
Performance Summary
Results for SPECjvm98
----------------------------------
FastAdaptiveSemispace Result: 24.50
FastSemispace Result: 23.82
FastMarkSweep Result: 21.49
Results for SPECjbb2000
----------------------------------
FastMarkSweep Valid run, Score is 3274
FastSemispace Valid run, Score is 3548
FastAdaptiveSemispace Valid run, Score is 3344
SPECjvm98 Performance Details
-----------------------------
RESULT: FastSemispace SPECjvm98
compress Time: 35.376 Ratio: 33.21
jess Time: 15.683 Ratio: 24.23
db Time: 60.696 Ratio: 8.32
javac Time: 26.279 Ratio: 16.17
mpegaudio Time: 22.064 Ratio: 49.85
mtrt Time: 8.658 Ratio: 53.13
jack Time: 29.993 Ratio: 15.17
Bottom Line: Result: 23.82
RESULT: FastMarkSweep SPECjvm98
compress Time: 35.751 Ratio: 32.87
jess Time: 18.388 Ratio: 20.67
db Time: 76.315 Ratio: 6.62
javac Time: 29.974 Ratio: 14.18
mpegaudio Time: 22.109 Ratio: 49.75
mtrt Time: 9.551 Ratio: 48.16
jack Time: 32.838 Ratio: 13.86
Bottom Line: Result: 21.49
RESULT: FastAdaptiveSemispace SPECjvm98
compress Time: 33.741 Ratio: 34.82
jess Time: 13.984 Ratio: 27.17
db Time: 54.937 Ratio: 9.19
javac Time: 26.755 Ratio: 15.88
mpegaudio Time: 22.144 Ratio: 49.67
mtrt Time: 9.253 Ratio: 49.71
jack Time: 29.297 Ratio: 15.53
Bottom Line: Result: 24.50
Opt Level 0 compilation breakdown for SPECjvm98 size 100
Compilation Subsystem Report
Comp #Meths Time bcb/ms mcb/bcb MCKB BCKB
Base 122 94 460.18 11.71 494.7 42.2
Opt 2082 17779 10.09 4.46 781.7 175.1
Explanation of Metrics
#Meths: Total number of methods compiled by the compiler
Time: Total compilation time in milliseconds
bcb/ms: Number of bytecode bytes complied per millisecond
mcb/bcb: Ratio of machine code bytes to bytecode bytes
MCKB: Total number of machine code bytes generated in kilobytes
BCKB: Total number of bytecode bytes compiled in kilobytes
Optimizing Compiler SubSystem
Phase Time
(ms) (0fTotal)
Convert Bytecodes to HIR
Generate HIR 1499 9.01%
Branch Optimizations 225 1.35%
TOTAL 1724 10.36%
Yield Point Insertion 74 0.44%
Field Analysis 72 0.43%
Convert HIR to LIR
Expand Runtime Services 599 3.60%
HIR Operator Expansion 345 2.07%
Branch Optimizations 202 1.21%
TOTAL 1148 6.90%
Convert LIR to MIR
SplitBasicBlock 53 0.32%
Instruction Selection
Reduce Operators 82 0.49%
ConvertALUOps 24 0.14%
Normalize Constants 309 1.86%
Live Handlers 27 0.16%
DepGraph & BURS 3382 20.33%
Complex Operators 431 2.59%
TOTAL 4257 25.59%
TOTAL 4310 25.91%
NullCheckCombining 233 1.40%
Register Mapping
Expand Calling Convention 217 1.30%
Live Analysis 2128 12.79%
Register Allocation
Register Allocation Prep 371 2.23%
Live Interval Analysis 323 1.94%
Linear Scan 4961 29.82%
TOTAL 5656 34.00%
Insert Prologue/Epilogue 411 2.47%
TOTAL 8413 50.58%
Generate Machine Code
Final MIR Expansion 191 1.15%
Assembler Driver 465 2.80%
TOTAL 657 3.95%
TOTAL COMPILATION TIME 16634
Opt Level 1 compilation breakdown for SPECjvm98 size 100
Compilation Subsystem Report
Comp #Meths Time bcb/ms mcb/bcb MCKB BCKB
Base 125 95 456.13 11.72 495.8 42.3
Opt 1541 46423 3.68 5.87 979.8 166.9
Explanation of Metrics
#Meths: Total number of methods compiled by the compiler
Time: Total compilation time in milliseconds
bcb/ms: Number of bytecode bytes complied per millisecond
mcb/bcb: Ratio of machine code bytes to bytecode bytes
MCKB: Total number of machine code bytes generated in kilobytes
BCKB: Total number of bytecode bytes compiled in kilobytes
Optimizing Compiler SubSystem
Phase Time
(ms) (0fTotal)
Convert Bytecodes to HIR
Generate HIR 3623 8.04%
Branch Optimizations 461 1.02%
TOTAL 4084 9.06%
Tail Recursion Elimination 55 0.12%
Yield Point Insertion 80 0.18%
Simple Opts 872 1.93%
Escape Transformations 772 1.71%
Branch Optimizations 204 0.45%
Static Splitting 53 0.12%
Simple Opts 421 0.93%
Local CopyProp 1886 4.18%
Local ConstantProp 226 0.50%
Local CSE 1215 2.69%
Field Analysis 66 0.15%
Convert HIR to LIR
Expand Runtime Services 3228 7.16%
Branch Optimizations 230 0.51%
Local Cast Optimizations 72 0.16%
HIR Operator Expansion 425 0.94%
Branch Optimizations 345 0.77%
TOTAL 4303 9.54%
Local CopyProp 6676 14.81%
Local ConstantProp 422 0.94%
Local CSE 1074 2.38%
Simple Opts 946 2.10%
Code Reordering 2042 4.53%
Branch Optimizations 331 0.73%
Convert LIR to MIR
SplitBasicBlock 61 0.14%
Instruction Selection
Reduce Operators 107 0.24%
ConvertALUOps 18 0.04%
Normalize Constants 462 1.02%
Live Handlers 20 0.04%
DepGraph & BURS 4550 10.09%
Complex Operators 670 1.49%
TOTAL 5829 12.93%
TOTAL 5890 13.06%
NullCheckCombining 307 0.68%
Register Mapping
Expand Calling Convention 886 1.97%
Live Analysis 3323 7.37%
Register Allocation
Register Allocation Prep 436 0.97%
Live Interval Analysis 519 1.15%
Linear Scan 6491 14.40%
TOTAL 7447 16.52%
Insert Prologue/Epilogue 421 0.93%
TOTAL 12078 26.79%
Branch Optimizations 251 0.56%
Generate Machine Code
Final MIR Expansion 200 0.44%
Assembler Driver 622 1.38%
TOTAL 822 1.82%
TOTAL COMPILATION TIME 45089
Opt Level 2 compilation breakdown for SPECjvm98 size 100
Compilation Subsystem Report
Comp #Meths Time bcb/ms mcb/bcb MCKB BCKB
Base 145 99 453.89 11.57 507.9 43.9
Opt 1535 166757 1.01 6.06 993.0 163.9
Explanation of Metrics
#Meths: Total number of methods compiled by the compiler
Time: Total compilation time in milliseconds
bcb/ms: Number of bytecode bytes complied per millisecond
mcb/bcb: Ratio of machine code bytes to bytecode bytes
MCKB: Total number of machine code bytes generated in kilobytes
BCKB: Total number of bytecode bytes compiled in kilobytes
Optimizing Compiler SubSystem
Phase Time
(ms) (0fTotal)
Convert Bytecodes to HIR
Generate HIR 3589 2.11%
Branch Optimizations 468 0.27%
TOTAL 4057 2.38%
Tail Recursion Elimination 53 0.03%
Yield Point Insertion 76 0.04%
Simple Opts 877 0.51%
Escape Transformations 777 0.46%
Branch Optimizations 204 0.12%
Static Splitting 53 0.03%
CFGTransformations 2752 1.61%
Simple Opts 456 0.27%
Loop Unrolling 3133 1.84%
SSA
HIR SSA transformations
Local CopyProp 2093 1.23%
Local ConstantProp 257 0.15%
Dominators + LpStrTree 1584 0.93%
Dominance Frontier 555 0.33%
Array SSA Load Eliminati 47788 28.04%
RedundantBranchEliminati
Ensure SSA 2336 1.37%
Global Value Number 2773 1.63%
RBE Transform 87 0.05%
TOTAL 5197 3.05%
SSA Tune Up
SSA Tune UpPreparati 25 0.01%
Enter SSA 2197 1.29%
Simple Opts 505 0.30%
SSA Expression Foldi 369 0.22%
TOTAL 3097 1.82%
SSA Tune Up
SSA Tune UpPreparati 23 0.01%
Enter SSA 2080 1.22%
Simple Opts 439 0.26%
SSA Expression Foldi 337 0.20%
TOTAL 2881 1.69%
Global Code Placement
GCP Preparation 28 0.02%
Enter SSA 7806 4.58%
Global CSE 34 0.02%
LICM 11003 6.46%
GCP Finalization 28 0.02%
TOTAL 18902 11.09%
Leave SSA 3607 2.12%
TOTAL 85967 50.44%
Coalesce Moves 1975 1.16%
Post SSA cleanup
Local CopyProp 580 0.34%
Local ConstantProp 212 0.12%
Simple Opts 457 0.27%
Escape Transformations 716 0.42%
Branch Optimizations 207 0.12%
TOTAL 2174 1.28%
TOTAL 90117 52.87%
Local CopyProp 576 0.34%
Local ConstantProp 145 0.09%
Local CSE 1083 0.64%
Field Analysis 65 0.04%
Convert HIR to LIR
Expand Runtime Services 3293 1.93%
Branch Optimizations 224 0.13%
Local Cast Optimizations 73 0.04%
HIR Operator Expansion 445 0.26%
Branch Optimizations 357 0.21%
TOTAL 4393 2.58%
SSA
LIR SSA transformations
Dominators + LpStrTree 2524 1.48%
Dominance Frontier 1082 0.63%
Global Code Placement
GCP Preparation 28 0.02%
Enter SSA 2998 1.76%
Global CSE 1534 0.90%
LICM 11965 7.02%
GCP Finalization 29 0.02%
TOTAL 16557 9.71%
Leave SSA 5949 3.49%
TOTAL 26113 15.32%
Coalesce Moves 4372 2.57%
Post SSA cleanup
Local CopyProp 2537 1.49%
Local ConstantProp 279 0.16%
Simple Opts 959 0.56%
Branch Optimizations 302 0.18%
TOTAL 4078 2.39%
TOTAL 34564 20.28%
Local CopyProp 737 0.43%
Local ConstantProp 158 0.09%
Local CSE 839 0.49%
Simple Opts 768 0.45%
Code Reordering 2037 1.20%
Branch Optimizations 294 0.17%
Convert LIR to MIR
SplitBasicBlock 60 0.04%
Instruction Selection
Reduce Operators 106 0.06%
ConvertALUOps 18 0.01%
Normalize Constants 463 0.27%
Live Handlers 2592 1.52%
DepGraph & BURS 4756 2.79%
Complex Operators 673 0.39%
TOTAL 8608 5.05%
TOTAL 8669 5.09%
NullCheckCombining 313 0.18%
Register Mapping
Expand Calling Convention 901 0.53%
Live Analysis 3364 1.97%
Register Allocation
Register Allocation Prep 432 0.25%
Live Interval Analysis 531 0.31%
Linear Scan 6495 3.81%
TOTAL 7459 4.38%
Insert Prologue/Epilogue 426 0.25%
TOTAL 12151 7.13%
Branch Optimizations 242 0.14%
Generate Machine Code
Final MIR Expansion 204 0.12%
Assembler Driver 632 0.37%
TOTAL 837 0.49%
TOTAL COMPILATION TIME 170439
|