[llvm-testresults] Grawp nightly tester results
daniel_dunbar at apple.com
daniel_dunbar at apple.com
Fri Dec 18 03:33:47 PST 2009
http://smooshlab.apple.com/perf.cgi/db_nt_internal/nightlytest/3859/
Name: grawp.apple.com
Nickname: Grawp:4
Run: 3859, Start Time: 2009-12-18 00:18:31, End Time: 2009-12-18 03:33:22
Comparing To: 3832, Start Time: 2009-12-17 00:34:41, End Time: 2009-12-17 03:39:51
--- Changes Summary ---
New Test Passes: 0
New Test Failures: 0
Added Tests: 9
Removed Tests: 0
Significant Changes: 15
--- Tests Summary ---
Total Tests: 3879
Total Test Failures: 136
Total Test Failures By Type:
CBE: 52
JIT: 26
JIT codegen: 58
--- Changes Detail ---
New Test Passes:
New Test Failures:
Added Tests:
SingleSource/Benchmarks/Misc/fp-convert [JIT, CBE, JIT codegen, LLC, LLC-BETA, LLC compile, Bitcode, GCCAS, LLC_BETA compile]
Removed Tests:
Significant Changes in Test Results:
LLC-BETA:
MultiSource/Applications/lambda-0_1_3/lambda: 6.97% (6.4600 => 6.9100)
Bitcode:
MultiSource/Benchmarks/tramp3d-v4/tramp3d-v4: 5.36% (1888624.0000 => 1989792.0000)
JIT:
MultiSource/Benchmarks/OptimizerEval/optimizer-eval: 6.90% (0.2900 => 0.3100)
MultiSource/Benchmarks/Prolangs-C/TimberWolfMC/timberwolfmc: 9.26% (0.5400 => 0.5900)
MultiSource/Benchmarks/Prolangs-C/agrep/agrep: 8.70% (0.2300 => 0.2500)
MultiSource/Benchmarks/mafft/pairlocalalign: -6.23% (36.1300 => 33.8800)
SingleSource/Benchmarks/Misc/himenobmtxpa: 10.68% (2.8100 => 3.1100)
SingleSource/Benchmarks/Shootout/lists: 19.11% (7.1700 => 8.5400)
SingleSource/Benchmarks/Shootout/methcall: -23.88% (8.6700 => 6.6000)
JIT codegen:
MultiSource/Applications/kimwitu++/kc: 6.70% (5.3635 => 5.7226)
MultiSource/Benchmarks/MallocBench/gs/gs: 5.03% (0.9164 => 0.9625)
MultiSource/Benchmarks/Prolangs-C/TimberWolfMC/timberwolfmc: 7.85% (0.4918 => 0.5304)
MultiSource/Benchmarks/mediabench/mpeg2/mpeg2dec/mpeg2decode: 5.03% (0.4412 => 0.4634)
CBE:
Externals/SPEC/CINT2006/458_sjeng/458_sjeng: 5.22% (6.1300 => 6.4500)
MultiSource/Benchmarks/Ptrdist/ft/ft: 5.94% (1.0100 => 1.0700)
More information about the llvm-testresults
mailing list