[llvm-testresults] Geryon-X86-64-PIC i386 nightly tester results

Dan Gohman gohman at apple.com
Fri Jul 4 08:32:09 PDT 2008


No, it did not appear to be real.

Also, in last night's Geryon-X86-64-PIC run, both hmmer and bzip2
bounced back.

But 464.h264ref took a 9% dive. But not in LLC-BETA.

On LLC-BETA, 400.perlbench took a 9% dive, and
300.twolf bot 9% better. Neither of those showed up in LLC.

Dan

On Fri, July 4, 2008 12:58 am, Evan Cheng wrote:
> Ping. Is 456.hmmer regression real?
>
> Evan
>
> On Jul 3, 2008, at 6:31 AM, Apache wrote:
>
>> http://llvm.org/nightlytest/test.php?machine=283&night=6569
>> Name: clattner2.apple.com
>> Nickname: Geryon-X86-64-PIC
>> Buildstatus: OK
>>
>> New Test Passes:
>> None
>>
>> New Test Failures:
>> None
>>
>> Added Tests:
>> None
>>
>> Removed Tests:
>> Benchmarks/BitBench/huffman/huffman
>>
>>
>> Significant changes in test results:
>> CBE:
>> singlesource/Benchmarks/BenchmarkGame/recursive: -23.37% (1.84 =>
>> 2.27)
>> singlesource/Benchmarks/Dhrystone/fldry: -14.29% (1.40 => 1.60)
>> singlesource/Benchmarks/Misc/flops: 16.45% (11.31 => 9.45)
>> singlesource/Benchmarks/Shootout/ary3: -6.20% (3.55 => 3.77)
>> singlesource/Benchmarks/Shootout/lists: 6.02% (5.65 => 5.31)
>> multisource/Benchmarks/BitBench/five11/five11: 7.84% (6.25 => 5.76)
>> external/SPEC/CINT2006/456.hmmer/456.hmmer: -9.39% (5.75 => 6.29)
>> external/SPEC/CINT2006/458.sjeng/458.sjeng: 5.95% (6.55 => 6.16)
>> LLC:
>> singlesource/Benchmarks/BenchmarkGame/recursive: -22.40% (1.83 =>
>> 2.24)
>> singlesource/Benchmarks/Shootout-C++/fibo: -5.42% (6.46 => 6.81)
>> singlesource/Benchmarks/Shootout/lists: 9.63% (6.44 => 5.82)
>> singlesource/Benchmarks/Shootout/random: 6.24% (4.17 => 3.91)
>> multisource/Applications/SIBsim4/SIBsim4: -10.55% (3.98 => 4.40)
>> multisource/Benchmarks/BitBench/five11/five11: 7.22% (6.23 => 5.78)
>> multisource/Benchmarks/SciMark2-C/scimark2: -5.76% (25.88 => 27.37)
>> external/SPEC/CINT2000/181.mcf/181.mcf: 6.96% (5.89 => 5.48)
>> external/SPEC/CINT2000/256.bzip2/256.bzip2: -7.67% (10.17 => 10.95)
>> external/SPEC/CINT2006/456.hmmer/456.hmmer: -5.04% (4.76 => 5.00)
>> LLC-BETA:
>> singlesource/Benchmarks/BenchmarkGame/recursive: -15.93% (1.82 =>
>> 2.11)
>> singlesource/Benchmarks/Shootout/ary3: -5.95% (3.53 => 3.74)
>> multisource/Applications/JM/lencod/lencod: 9.74% (8.11 => 7.32)
>> multisource/Benchmarks/sim/sim: -5.36% (4.85 => 5.11)
>> external/SPEC/CINT2000/300.twolf/300.twolf: -8.36% (3.59 => 3.89)
>> GCCAS:
>> multisource/Applications/SPASS/SPASS: 10.58% (6.8541 => 6.1290)
>> multisource/Benchmarks/tramp3d-v4/tramp3d-v4: 5.05% (8.6165 => 8.1815)
>> external/SPEC/CINT2006/464.h264ref/464.h264ref: -6.19% (4.4489 =>
>> 4.7241)
>> LLC-BETA compile:
>> external/SPEC/CFP2006/447.dealII/447.dealII: -9.11% (5.3046 => 5.7881)
>> LLC compile:
>> external/SPEC/CINT2006/400.perlbench/400.perlbench: 5.55% (20.1315
>> => 19.0135)
>>
>> _______________________________________________
>> llvm-testresults mailing list
>> llvm-testresults at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-testresults
>
>





More information about the llvm-testresults mailing list