[llvm-testresults] buildbot failure in llvm on llvm-gcc-i686-darwin10-selfhost

[Bob Wilson]

On Apr 7, 2010, at 11:07 AM, Jakob Stoklund Olesen wrote:

> 
> On Apr 7, 2010, at 10:26 AM, buildbot at google1.osuosl.org wrote:
> 
>> The Buildbot has detected a new failure of llvm-gcc-i686-darwin10-selfhost on llvm.
>> Full details are available at:
>> http://google1.osuosl.org:8011/builders/llvm-gcc-i686-darwin10-selfhost/builds/3596
>> 
>> Buildbot URL: http://google1.osuosl.org:8011/
>> 
>> Buildslave for this Build: dunbar-darwin10
>> 
>> Build Reason: 
>> Build Source Stamp: 100632
>> Blamelist: stoklund
>> 
>> BUILD FAILED: failed compile.llvm-gcc.stage1
> 
> I claim a cosmic ray.
> 
> The next build seems to be doing fine: http://google1.osuosl.org:8011/builders/llvm-gcc-i686-darwin10-selfhost/builds/3597

I have seen this (or something similar) before, so I had a look at this in case it is a race condition like we had in libgomp.  I don't see any evidence for that.  For some reason, the _AC_COMPILER_EXEEXT test failed to link a trivial program, so configure decided that it couldn't link anything.  I'll watch for this again, but for now I agree about the cosmic ray theory.