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

Bruno Cardoso Lopes bruno.cardoso at gmail.com
Fri Feb 11 07:02:04 PST 2011


Ok! Thanks!

On Fri, Feb 11, 2011 at 1:02 AM, Daniel Dunbar <daniel_dunbar at apple.com> wrote:
>
> On Feb 10, 2011, at 12:03 PM, Bruno Cardoso Lopes wrote:
>
>> Daniel,
>>
>> Can you provide me the .s of the mips llc run on this machine? That
>> would help me a lot! =D
>
> I don't have it handy, sorry. This test seems to be failing in a lot of different places though. Can you try with a different compiler or arch and see if you can repro? If you look at the logs on the google buildbots, it looks like the instruction is coming up with a completely different immediate value.
>
> I'm going to disable the test for now...
>
>  - Daniel
>
>> Thanks in advance!
>>
>> On Thu, Feb 10, 2011 at 6:00 PM, Daniel Dunbar <daniel_dunbar at apple.com> wrote:
>>> It is using gcc-4.0 on Leopard (IIRC).
>>>
>>> One typical reason for it to fail and not fail elsewhere is non-determinism in the code generator. Then there is the ever popular you've-found-a-new-way-to-get-gcc-to-miscompile-us set of problems.
>>>
>>>  - Daniel
>>>
>>> On Feb 10, 2011, at 11:43 AM, Bruno Cardoso Lopes wrote:
>>>
>>>> This seems odd. It works fine locally. Are there anymore buildbot
>>>> details that could help?
>>>>
>>>> On Thu, Feb 10, 2011 at 4:23 PM,  <daniel_dunbar at apple.com> wrote:
>>>>> The Buildbot has detected a new failure of clang-i386-darwin9 on smooshlab.
>>>>> Full details are available at:
>>>>>  http://smooshlab.apple.com:8010/builders/clang-i386-darwin9/builds/18767
>>>>>
>>>>> Buildbot URL: http://smooshlab.apple.com:8010/
>>>>>
>>>>> Buildslave for this Build: smoosh-15.apple.com
>>>>>
>>>>> Build Reason:
>>>>> Build Source Stamp: 125292
>>>>> Blamelist: bruno
>>>>>
>>>>> BUILD FAILED: failed test-llvm
>>>>>
>>>>> sincerely,
>>>>>  -The Buildbot
>>>>>
>>>>>
>>>>> ================================================================================
>>>>>
>>>>> CHANGES:
>>>>> Files:
>>>>>  lib/Target/Mips/MipsISelLowering.cpp
>>>>>  test/CodeGen/Mips/o32-cc.ll
>>>>> At: Thu 10 Feb 2011 10:13:31
>>>>> Changed By: bruno
>>>>> Comments: Fix a lot of o32 CC issues and add a bunch of tests. Patch by Akira Hatanaka with some small modifications by me.Properties:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> LOGS:
>>>>> Last 10 lines of 'stdio':
>>>>>        ********************
>>>>>        Failing Tests (1):
>>>>>            LLVM :: CodeGen/Mips/o32-cc.ll
>>>>>
>>>>>          Expected Passes    : 5214
>>>>>          Expected Failures  : 40
>>>>>          Unsupported Tests  : 539
>>>>>          Unexpected Failures: 1
>>>>>        make[1]: *** [check-local-lit] Error 1
>>>>>        make: *** [check] Error 2
>>>>>
>>>>> Last 10 lines of 'fail':
>>>>>        LLVM :: CodeGen/Mips/o32-cc.ll
>>>>>
>>>>> Last 10 lines of 'xfail':
>>>>>        LLVM :: MC/ARM/neont2-pairwise-encoding.s
>>>>>        LLVM :: MC/ARM/neont2-vst-encoding.s
>>>>>        LLVM :: MC/ARM/thumb2.s
>>>>>        LLVM :: MC/AsmParser/directive_lsym.s
>>>>>        LLVM :: MC/AsmParser/exprs-invalid.s
>>>>>        LLVM :: Transforms/GVN/rle-no-phi-translate.ll
>>>>>        LLVM :: Transforms/IndVarSimplify/loop_evaluate_6.ll
>>>>>        LLVM :: Transforms/MergeFunc/fold-weak.ll
>>>>>        LLVM :: Transforms/TailCallElim/nocapture.ll
>>>>>        LLVM :: Transforms/TailDup/2008-06-11-AvoidDupLoopHeader.ll
>>>>>
>>>>> Last 10 lines of 'unsupported':
>>>>>        LLVM :: LLVMC/C++/unknown_suffix.unk
>>>>>        LLVM :: LLVMC/C/emit-llvm-opt.c
>>>>>        LLVM :: LLVMC/C/emit-llvm.c
>>>>>        LLVM :: LLVMC/C/hello.c
>>>>>        LLVM :: LLVMC/C/include.c
>>>>>        LLVM :: LLVMC/C/opt-test.c
>>>>>        LLVM :: LLVMC/C/sink.c
>>>>>        LLVM :: LLVMC/C/wall.c
>>>>>        LLVM :: LLVMC/ObjC++/hello.mm
>>>>>        LLVM :: LLVMC/ObjC/hello.m
>>>>>
>>>>> Last 10 lines of 'o32-cc.ll':
>>>>>        <stdin>:941:2: note: scanning from here
>>>>>         ldc1 $f12, %lo($CPI23_0)($3)
>>>>>         ^
>>>>>        <stdin>:1007:2: note: possible intended match here
>>>>>         addiu $4, $zero, 82
>>>>>         ^
>>>>>        --
>>>>>
>>>>>        ********************
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> Bruno Cardoso Lopes
>>>> http://www.brunocardoso.cc
>>>
>>>
>>
>>
>>
>> --
>> Bruno Cardoso Lopes
>> http://www.brunocardoso.cc
>
>



-- 
Bruno Cardoso Lopes
http://www.brunocardoso.cc




More information about the llvm-testresults mailing list