[llvm-testresults] pollyperf-O3-polly-detect__clang_DEV__x86_64 test results

llvm-testresults at cs.uiuc.edu llvm-testresults at cs.uiuc.edu
Mon Jan 27 19:49:42 PST 2014


http://llvm.org/perf/db_default/v4/nts/21415?compare_to=21407&baseline=19899
Nickname: pollyperf-O3-polly-detect__clang_DEV__x86_64:35

Comparing:
     Run: 21415, Order: 200257, Start Time: 2014-01-27 23:10:31, End Time: 2014-01-28 03:54:44
      To: 21407, Order: 200231, Start Time: 2014-01-27 18:14:18, End Time: 2014-01-27 22:58:29
Baseline: 19899, Order: 198654, Start Time: 2014-01-07 01:22:04, End Time: 2014-01-07 06:18:02

===============
Tests Summary
===============

Performance Regressions: 18 (21 on baseline)
Performance Improvements: 0 (66 on baseline)
Unchanged Tests: 1092 (1023 on baseline)
Total Tests: 1110

===========================
Run-Over-Run Changes Detail
===========================

Performance Regressions - Compile Time
--------------------------------------
  SingleSource/Benchmarks/BenchmarkGame/nsieve-bits: 80.00% (0.0200 => 0.0360, std.dev.: 0.0036)
  SingleSource/UnitTests/2003-05-07-VarArgs: 50.00% (0.0240 => 0.0360, std.dev.: 0.0027)
  SingleSource/Benchmarks/McGill/queens: 40.00% (0.0400 => 0.0560, std.dev.: 0.0049)
  MultiSource/Benchmarks/Olden/bisort/bisort: 35.71% (0.0560 => 0.0760, std.dev.: 0.0054)
  SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu: 22.22% (0.0720 => 0.0880, std.dev.: 0.0020)
  MultiSource/Benchmarks/Olden/mst/mst: 19.35% (0.1240 => 0.1480, std.dev.: 0.0074)
  SingleSource/Benchmarks/Polybench/medley/floyd-warshall/floyd-warshall: 18.75% (0.0640 => 0.0760, std.dev.: 0.0045)
  SingleSource/Benchmarks/CoyoteBench/lpbench: 14.81% (0.1080 => 0.1240, std.dev.: 0.0038)
  SingleSource/Benchmarks/Shootout-C++/wc: 6.45% (0.2480 => 0.2640, std.dev.: 0.0041)
  SingleSource/Benchmarks/Shootout-C++/ary: 5.80% (0.2760 => 0.2920, std.dev.: 0.0036)
  MultiSource/Benchmarks/MallocBench/cfrac/cfrac: 5.67% (0.9880 => 1.0440, std.dev.: 0.0207)
  SingleSource/Benchmarks/Shootout-C++/strcat: 5.26% (0.2280 => 0.2400, std.dev.: 0.0042)
  SingleSource/Benchmarks/Adobe-C++/stepanov_abstraction: 2.61% (0.6120 => 0.6280, std.dev.: 0.0050)
  MultiSource/Benchmarks/TSVC/StatementReordering-flt/StatementReordering-flt: 2.27% (1.0561 => 1.0801, std.dev.: 0.0072)
  MultiSource/Benchmarks/TSVC/Symbolics-dbl/Symbolics-dbl: 1.82% (1.0961 => 1.1161, std.dev.: 0.0063)
  MultiSource/Benchmarks/TSVC/InductionVariable-dbl/InductionVariable-dbl: 1.41% (1.1361 => 1.1521, std.dev.: 0.0047)

Performance Regressions - Execution Time
----------------------------------------
  SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu: 300.00% (0.0040 => 0.0160, std.dev.: 0.0037)
  MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael: 60.00% (0.0200 => 0.0320, std.dev.: 0.0036)

================================
Run-Over-Baseline Changes Detail
================================

Performance Regressions - Compile Time
--------------------------------------
  SingleSource/UnitTests/2006-12-07-Compare64BitConstant: 300.00% (0.0040 => 0.0160, std.dev.: 0.0030)
  SingleSource/Benchmarks/McGill/chomp: 16.67% (0.1920 => 0.2240, std.dev.: 0.0070)
  MultiSource/Benchmarks/McCat/04-bisect/bisect: 13.16% (0.1520 => 0.1720, std.dev.: 0.0056)
  SingleSource/Benchmarks/Shootout-C++/ary: 4.29% (0.2800 => 0.2920, std.dev.: 0.0036)
  MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael: 4.12% (0.6800 => 0.7080, std.dev.: 0.0088)

Performance Regressions - Execution Time
----------------------------------------
  SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu: 300.00% (0.0040 => 0.0160, std.dev.: 0.0037)
  SingleSource/Benchmarks/Shootout/sieve: 157.80% (6.6644 => 17.1811, std.dev.: 0.0031)
  MultiSource/Benchmarks/Ptrdist/yacr2/yacr2: 11.59% (0.8281 => 0.9241, std.dev.: 0.0047)
  MultiSource/Benchmarks/Olden/perimeter/perimeter: 9.33% (0.3000 => 0.3280, std.dev.: 0.0080)
  MultiSource/Benchmarks/MiBench/telecomm-CRC32/telecomm-CRC32: 6.78% (0.2360 => 0.2520, std.dev.: 0.0039)
  MultiSource/Benchmarks/ASCI_Purple/SMG2000/smg2000: 5.93% (3.9842 => 4.2203, std.dev.: 0.0232)
  SingleSource/Benchmarks/Dhrystone/fldry: 4.00% (0.6000 => 0.6240, std.dev.: 0.0043)
  MultiSource/Benchmarks/Trimaran/netbench-crc/netbench-crc: 3.89% (0.7200 => 0.7480, std.dev.: 0.0103)
  MultiSource/Benchmarks/Ptrdist/anagram/anagram: 2.90% (1.2401 => 1.2761, std.dev.: 0.0126)
  MultiSource/Benchmarks/TSVC/InductionVariable-dbl/InductionVariable-dbl: 2.55% (5.0203 => 5.1483, std.dev.: 0.0043)
  SingleSource/Benchmarks/McGill/chomp: 2.16% (0.9241 => 0.9441, std.dev.: 0.0051)
  MultiSource/Applications/JM/lencod/lencod: 1.92% (8.1245 => 8.2805, std.dev.: 0.0197)
  MultiSource/Applications/aha/aha: 1.39% (2.8802 => 2.9202, std.dev.: 0.0033)
  SingleSource/Benchmarks/Shootout-C++/hash2: 1.09% (2.9362 => 2.9682, std.dev.: 0.0045)
  MultiSource/Applications/hexxagon/hexxagon: 1.04% (11.5567 => 11.6767, std.dev.: 0.0146)
  MultiSource/Applications/lambda-0_1_3/lambda: 1.02% (5.0963 => 5.1483, std.dev.: 0.0134)

Performance Improvements - Compile Time
---------------------------------------
  SingleSource/UnitTests/2009-12-07-StructReturn: -80.00% (0.0200 => 0.0040, std.dev.: 0.0051)
  MultiSource/Benchmarks/FreeBench/pcompress2/pcompress2: -12.20% (0.3280 => 0.2880, std.dev.: 0.0134)
  SingleSource/Benchmarks/CoyoteBench/lpbench: -8.82% (0.1360 => 0.1240, std.dev.: 0.0038)
  MultiSource/Benchmarks/TSVC/ControlLoops-flt/ControlLoops-flt: -8.31% (1.2521 => 1.1481, std.dev.: 0.0146)
  MultiSource/Benchmarks/TSVC/CrossingThresholds-dbl/CrossingThresholds-dbl: -8.22% (1.2161 => 1.1161, std.dev.: 0.0133)
  MultiSource/Benchmarks/TSVC/InductionVariable-flt/InductionVariable-flt: -8.12% (1.2321 => 1.1321, std.dev.: 0.0122)
  MultiSource/Benchmarks/TSVC/Recurrences-dbl/Recurrences-dbl: -8.04% (1.1441 => 1.0521, std.dev.: 0.0148)
  MultiSource/Benchmarks/TSVC/LinearDependence-dbl/LinearDependence-dbl: -8.00% (1.3001 => 1.1961, std.dev.: 0.0113)
  MultiSource/Benchmarks/TSVC/Reductions-dbl/Reductions-dbl: -7.79% (1.2841 => 1.1841, std.dev.: 0.0139)
  MultiSource/Benchmarks/TSVC/Searching-flt/Searching-flt: -7.75% (1.1361 => 1.0481, std.dev.: 0.0127)
  MultiSource/Benchmarks/TSVC/Equivalencing-flt/Equivalencing-flt: -7.67% (1.2001 => 1.1081, std.dev.: 0.0095)
  MultiSource/Benchmarks/TSVC/Reductions-flt/Reductions-flt: -7.45% (1.2881 => 1.1921, std.dev.: 0.0098)
  MultiSource/Benchmarks/TSVC/Searching-dbl/Searching-dbl: -7.39% (1.1361 => 1.0521, std.dev.: 0.0113)
  MultiSource/Benchmarks/TSVC/Recurrences-flt/Recurrences-flt: -7.29% (1.1521 => 1.0681, std.dev.: 0.0105)
  MultiSource/Benchmarks/TSVC/LoopRestructuring-flt/LoopRestructuring-flt: -7.10% (1.2401 => 1.1521, std.dev.: 0.0104)
  MultiSource/Benchmarks/TSVC/NodeSplitting-dbl/NodeSplitting-dbl: -7.09% (1.1841 => 1.1001, std.dev.: 0.0147)
  MultiSource/Benchmarks/TSVC/IndirectAddressing-flt/IndirectAddressing-flt: -7.07% (1.1881 => 1.1041, std.dev.: 0.0150)
  MultiSource/Benchmarks/MiBench/telecomm-gsm/telecomm-gsm: -7.07% (1.5280 => 1.4200, std.dev.: 0.0336)
  MultiSource/Benchmarks/TSVC/StatementReordering-flt/StatementReordering-flt: -6.90% (1.1601 => 1.0801, std.dev.: 0.0072)
  MultiSource/Benchmarks/TSVC/LoopRerolling-flt/LoopRerolling-flt: -6.83% (1.1721 => 1.0921, std.dev.: 0.0093)
  MultiSource/Benchmarks/TSVC/LoopRestructuring-dbl/LoopRestructuring-dbl: -6.82% (1.2321 => 1.1481, std.dev.: 0.0109)
  MultiSource/Benchmarks/TSVC/ControlLoops-dbl/ControlLoops-dbl: -6.71% (1.2521 => 1.1681, std.dev.: 0.0091)
  MultiSource/Benchmarks/TSVC/CrossingThresholds-flt/CrossingThresholds-flt: -6.67% (1.2001 => 1.1201, std.dev.: 0.0129)
  MultiSource/Benchmarks/TSVC/ControlFlow-dbl/ControlFlow-dbl: -6.65% (1.3841 => 1.2921, std.dev.: 0.0091)
  MultiSource/Benchmarks/TSVC/NodeSplitting-flt/NodeSplitting-flt: -6.46% (1.1761 => 1.1001, std.dev.: 0.0100)
  MultiSource/Benchmarks/TSVC/Symbolics-flt/Symbolics-flt: -6.46% (1.1761 => 1.1001, std.dev.: 0.0098)
  MultiSource/Benchmarks/TSVC/Packing-dbl/Packing-dbl: -6.36% (1.1321 => 1.0601, std.dev.: 0.0091)
  MultiSource/Benchmarks/TSVC/StatementReordering-dbl/StatementReordering-dbl: -6.32% (1.1401 => 1.0681, std.dev.: 0.0076)
  MultiSource/Benchmarks/TSVC/InductionVariable-dbl/InductionVariable-dbl: -6.19% (1.2281 => 1.1521, std.dev.: 0.0047)
  MultiSource/Benchmarks/TSVC/IndirectAddressing-dbl/IndirectAddressing-dbl: -6.14% (1.1721 => 1.1001, std.dev.: 0.0084)
  MultiSource/Benchmarks/TSVC/Equivalencing-dbl/Equivalencing-dbl: -6.08% (1.1841 => 1.1121, std.dev.: 0.0099)
  MultiSource/Benchmarks/TSVC/LinearDependence-flt/LinearDependence-flt: -5.88% (1.2921 => 1.2161, std.dev.: 0.0113)
  MultiSource/Benchmarks/TSVC/GlobalDataFlow-dbl/GlobalDataFlow-dbl: -5.84% (1.2321 => 1.1601, std.dev.: 0.0108)
  MultiSource/Benchmarks/TSVC/Expansion-dbl/Expansion-dbl: -5.77% (1.2481 => 1.1761, std.dev.: 0.0106)
  MultiSource/Benchmarks/TSVC/Expansion-flt/Expansion-flt: -5.77% (1.2481 => 1.1761, std.dev.: 0.0095)
  MultiSource/Benchmarks/TSVC/Packing-flt/Packing-flt: -5.59% (1.1441 => 1.0801, std.dev.: 0.0084)
  MultiSource/Benchmarks/TSVC/LoopRerolling-dbl/LoopRerolling-dbl: -5.54% (1.1561 => 1.0921, std.dev.: 0.0087)
  MultiSource/Benchmarks/TSVC/GlobalDataFlow-flt/GlobalDataFlow-flt: -5.52% (1.2321 => 1.1641, std.dev.: 0.0088)
  MultiSource/Benchmarks/TSVC/ControlFlow-flt/ControlFlow-flt: -5.26% (1.3681 => 1.2961, std.dev.: 0.0061)
  MultiSource/Benchmarks/TSVC/Symbolics-dbl/Symbolics-dbl: -5.10% (1.1761 => 1.1161, std.dev.: 0.0063)
  MultiSource/Benchmarks/nbench/nbench: -5.04% (1.3521 => 1.2840, std.dev.: 0.0147)
  SingleSource/Benchmarks/Shootout-C++/strcat: -4.76% (0.2520 => 0.2400, std.dev.: 0.0042)
  MultiSource/Applications/ClamAV/clamscan: -4.12% (19.2243 => 18.4323, std.dev.: 0.0596)
  MultiSource/Benchmarks/MallocBench/espresso/espresso: -3.46% (6.8280 => 6.5920, std.dev.: 0.0456)
  MultiSource/Benchmarks/MiBench/consumer-lame/consumer-lame: -3.15% (5.3360 => 5.1680, std.dev.: 0.0305)
  SingleSource/Benchmarks/Misc-C++/stepanov_container: -3.14% (1.0201 => 0.9881, std.dev.: 0.0103)
  MultiSource/Applications/d/make_dparser: -2.92% (4.1081 => 3.9881, std.dev.: 0.0346)
  MultiSource/Benchmarks/MallocBench/gs/gs: -2.90% (5.3840 => 5.2280, std.dev.: 0.0355)
  MultiSource/Applications/SPASS/SPASS: -2.55% (18.7966 => 18.3164, std.dev.: 0.0623)
  MultiSource/Benchmarks/mafft/pairlocalalign: -2.35% (11.3843 => 11.1163, std.dev.: 0.0277)
  MultiSource/Benchmarks/7zip/7zip-benchmark: -2.33% (49.1529 => 48.0088, std.dev.: 0.0751)
  MultiSource/Applications/treecc/treecc: -2.11% (2.6480 => 2.5920, std.dev.: 0.0213)
  MultiSource/Benchmarks/MiBench/consumer-typeset/consumer-typeset: -1.99% (16.7083 => 16.3763, std.dev.: 0.0375)
  MultiSource/Benchmarks/Bullet/bullet: -1.83% (39.0726 => 38.3566, std.dev.: 0.0649)
  MultiSource/Benchmarks/PAQ8p/paq8p: -1.61% (2.7282 => 2.6842, std.dev.: 0.0135)
  MultiSource/Applications/siod/siod: -1.60% (2.7562 => 2.7121, std.dev.: 0.0123)
  MultiSource/Applications/JM/lencod/lencod: -1.43% (19.5406 => 19.2605, std.dev.: 0.0417)
  MultiSource/Applications/JM/ldecod/ldecod: -1.40% (8.5802 => 8.4602, std.dev.: 0.0340)
  SingleSource/Benchmarks/Adobe-C++/simple_types_loop_invariant: -1.27% (2.2081 => 2.1801, std.dev.: 0.0100)
  SingleSource/Benchmarks/Adobe-C++/loop_unroll: -1.21% (4.2803 => 4.2283, std.dev.: 0.0141)

Performance Improvements - Execution Time
-----------------------------------------
  MultiSource/Benchmarks/FreeBench/analyzer/analyzer: -10.71% (0.1120 => 0.1000, std.dev.: 0.0031)
  SingleSource/UnitTests/Vectorizer/gcc-loops: -6.37% (5.2763 => 4.9403, std.dev.: 0.0839)
  MultiSource/Applications/siod/siod: -5.36% (2.9842 => 2.8242, std.dev.: 0.0218)
  SingleSource/Benchmarks/Shootout-C++/matrix: -4.61% (3.4722 => 3.3122, std.dev.: 0.0525)
  MultiSource/Benchmarks/TSVC/LinearDependence-flt/LinearDependence-flt: -2.13% (4.5123 => 4.4163, std.dev.: 0.0033)
  SingleSource/Benchmarks/Adobe-C++/functionobjects: -1.15% (3.4802 => 3.4402, std.dev.: 0.0035)

Report Time: 2.22s
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-testresults/attachments/20140127/2762009a/attachment.html>


More information about the llvm-testresults mailing list