[llvm-testresults] Grawp-NO-LTO-PIC nightly tester results

daniel_dunbar at apple.com daniel_dunbar at apple.com
Tue Jan 12 09:53:01 PST 2010


http://smooshlab.apple.com/perf.cgi/db_nt_internal/nightlytest/4557/
Name:  grawp.apple.com
Nickname: Grawp-NO-LTO-PIC:4

Run: 4557, Start Time: 2010-01-12 06:43:24, End Time: 2010-01-12 09:52:42
Comparing To: 4523, Start Time: 2010-01-11 06:40:52, End Time: 2010-01-11 09:43:57

--- Changes Summary ---
New Test Passes: 0
New Test Failures: 0
Added Tests: 0
Removed Tests: 0
Significant Changes: 16

--- Tests Summary ---
Total Tests: 3888
Total Test Failures: 931

Total Test Failures By Type:
  CBE: 63
  JIT: 432
  JIT codegen: 432
  LLC: 1
  LLC compile: 1
  LLC-BETA: 1
  LLC_BETA compile: 1

--- Changes Detail ---
New Test Passes:

New Test Failures:

Added Tests:

Removed Tests:

Significant Changes in Test Results:
LLC:
 Externals/SPEC/CINT2006/445_gobmk/445_gobmk: 12.50% (0.2400 => 0.2700)
 Externals/SPEC/CINT2006/471_omnetpp/471_omnetpp: -5.88% (2.2100 => 2.0800)
 MultiSource/Benchmarks/BitBench/drop3/drop3: 16.67% (0.4800 => 0.5600)
 MultiSource/Benchmarks/Olden/perimeter/perimeter: -17.07% (0.4100 => 0.3400)
Bitcode:
 MultiSource/Benchmarks/Olden/perimeter/perimeter: -6.93% (6928.0000 => 6448.0000)
LLC-BETA:
 MultiSource/Applications/siod/siod: -14.70% (5.1700 => 4.4100)
 MultiSource/Benchmarks/BitBench/drop3/drop3: 9.80% (0.5100 => 0.5600)
 MultiSource/Benchmarks/Bullet/bullet: -6.44% (8.8500 => 8.2800)
 MultiSource/Benchmarks/Olden/perimeter/perimeter: -14.63% (0.4100 => 0.3500)
 MultiSource/Benchmarks/Trimaran/enc-pc1/enc-pc1: -20.83% (1.6800 => 1.3300)
 SingleSource/Benchmarks/Shootout-C++/matrix: 31.72% (3.0900 => 4.0700)
 SingleSource/Benchmarks/Shootout/matrix: 28.77% (3.5800 => 4.6100)
 SingleSource/Benchmarks/Shootout/strcat: 7.14% (0.2800 => 0.3000)
CBE:
 MultiSource/Applications/kimwitu++/kc: 6.67% (0.3000 => 0.3200)
 MultiSource/Applications/siod/siod: 12.10% (4.3800 => 4.9100)
 MultiSource/Benchmarks/Ptrdist/ft/ft: -10.43% (1.1500 => 1.0300)



More information about the llvm-testresults mailing list