[llvm-testresults] buildbot failure in smooshlab on llvm-gcc-i386-darwin9

[Chris Lattner]

On Jan 9, 2011, at 10:49 AM, Evan Cheng wrote:
> Linux x86_64 tester is also showing non-determinstic behavior. Various buildbots have been red for multiple days.

I've been following all the testers that originally blamed me.  All of them seem to have come back except for these two:
http://smooshlab.apple.com:8010/builders/llvm-gcc-powerpc-darwin9
http://google1.osuosl.org:8011/builders/llvm-gcc-i386-linux-selfhost

> Chris, PPC tester is blaming you if I am reading it right.
> http://smooshlab.apple.com:8010/builders/llvm-gcc-powerpc-darwin9/builds/4890
> 
> Can we temporarily disable these memcpy / memset optimizations and see if bots turn green?

Yes, I will investigate and try to find the right thing to revert.  Sorry for the breakage.  I'm a bit concerned that there is a deeper problem that I just perturbed though.

-Chris

> 
> Evan
> 
> http://google1.osuosl.org:8011/builders/llvm-gcc-x86_64-linux-selfhost
> 
> On Jan 8, 2011, at 5:50 PM, Jakob Stoklund Olesen wrote:
> 
>> 
>> On Jan 8, 2011, at 5:26 PM, Chris Lattner wrote:
>> 
>>> 
>>> On Jan 8, 2011, at 5:15 PM, Jakob Stoklund Olesen wrote:
>>> 
>>>> 
>>>> On Jan 8, 2011, at 3:41 PM, daniel_dunbar at apple.com wrote:
>>>> 
>>>>> The Buildbot has detected a new failure of llvm-gcc-i386-darwin9 on smooshlab.
>>>>> Full details are available at:
>>>>> http://smooshlab.apple.com:8010/builders/llvm-gcc-i386-darwin9/builds/12499
>>>>> 
>>>>> Buildbot URL: http://smooshlab.apple.com:8010/
>>>>> 
>>>>> Buildslave for this Build: smoosh-03.apple.com
>>>>> 
>>>>> Build Reason: 
>>>>> Build Source Stamp: 123102
>>>>> Blamelist: coppro,stoklund
>>>>> 
>>>>> BUILD FAILED: failed compile.llvm-gcc.stage1
>>>> 
>>>> This is a bootstrap comparison failure.
>>>> 
>>>> I don't think my changes caused that, and other build bots are blaming Chris' earlier patches.
>>> 
>>> This doesn't blame me, and the other llvm-gcc failures are now green.
>> 
>> I think we may have code generation dependent on address space layout somewhere. This build has a bootstrap comparison failure in stage 2:
>> 
>> http://google1.osuosl.org:8011/builders/llvm-gcc-i386-linux-selfhost/builds/555
>> 
>> The next build fails in stage 1, still comparison failure:
>> 
>> http://google1.osuosl.org:8011/builders/llvm-gcc-i386-linux-selfhost/builds/556
>> 
>> Last good build was: http://google1.osuosl.org:8011/builders/llvm-gcc-i386-linux-selfhost/builds/553
>> Build #554 failed with an assertion in the optimizer.
>> 
>> 
>> _______________________________________________
>> llvm-testresults mailing list
>> llvm-testresults at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-testresults
>