[LLVMdev] Lots of regtest failures on PPC64/Linux

[David Blaikie]

missing-abstract-variable is a recent one I introduced - looking into it.

On Wed, Jun 4, 2014 at 2:39 AM, İsmail Dönmez <ismail at donmez.ws> wrote:
> Hi Hal,
>
> These tests failures go away when I disable static libs aka
> -DBUILD_SHARED_LIBS=OFF , with that only 2 regtest failures are left:
>
>
> [ 1314s] FAILED: cd /home/abuild/rpmbuild/BUILD/llvm/stage2/test &&
> /usr/bin/python /home/abuild/rpmbuild/BUILD/llvm/utils/lit/lit.py -sv
> --param
> llvm_site_config=/home/abuild/rpmbuild/BUILD/llvm/stage2/test/lit.site.cfg
> --param
> llvm_unit_site_config=/home/abuild/rpmbuild/BUILD/llvm/stage2/test/Unit/lit.site.cfg
> /home/abuild/rpmbuild/BUILD/llvm/stage2/test
> [ 1314s] lit.py: run.py:223: note: failed to initialize multiprocessing
> [ 1314s] -- Testing: 10750 tests, 16 threads --
> [ 1314s] Testing: 0 .. 10.. 20.. 30.. 40.. 50
> [ 1314s] FAIL: LLVM :: DebugInfo/missing-abstract-variable.ll (5826 of
> 10750)
> [ 1314s] ******************** TEST 'LLVM ::
> DebugInfo/missing-abstract-variable.ll' FAILED ********************
> [ 1314s] Script:
> [ 1314s] --
> [ 1314s] /home/abuild/rpmbuild/BUILD/llvm/stage2/./bin/llc -O0 -filetype=obj
> <
> /home/abuild/rpmbuild/BUILD/llvm/test/DebugInfo/missing-abstract-variable.ll
> | /home/abuild/rpmbuild/BUILD/llvm/stage2/./bin/llvm-dwarfdump
> -debug-dump=info - | /home/abuild/rpmbuild/BUILD/llvm/stage2/./bin/FileCheck
> /home/abuild/rpmbuild/BUILD/llvm/test/DebugInfo/missing-abstract-variable.ll
> [ 1314s] --
> [ 1314s] Exit Code: 1
> [ 1314s]
> [ 1314s] Command Output (stderr):
> [ 1314s] --
> [ 1314s]
> /home/abuild/rpmbuild/BUILD/llvm/test/DebugInfo/missing-abstract-variable.ll:74:10:
> error: expected string not found in input
> [ 1314s] ; CHECK: DW_TAG_formal_parameter
> [ 1314s]          ^
> [ 1314s] <stdin>:85:67: note: scanning from here
> [ 1314s]  DW_AT_abstract_origin [DW_FORM_ref4] (cu + 0x002a => {0x0000002a})
> [ 1314s]                                                                   ^
> [ 1314s] <stdin>:91:13: note: possible intended match here
> [ 1314s] 0x000000b7: DW_TAG_lexical_block [8] *
> [ 1314s]             ^
> [ 1314s]
> [ 1314s] --
>
>
> and
>
> [ 1314s] ******************** TEST 'LLVM-Unit ::
> ADT/ADTTests/HashingTest.HashCombineBasicTest' FAILED ********************
> [ 1314s] Note: Google Test filter = HashingTest.HashCombineBasicTest
> [ 1314s] [==========] Running 1 test from 1 test case.
> [ 1314s] [----------] Global test environment set-up.
> [ 1314s] [----------] 1 test from HashingTest
> [ 1314s] [ RUN      ] HashingTest.HashCombineBasicTest
> [ 1314s] ../unittests/ADT/HashingTest.cpp:413: Failure
> [ 1314s] Value of: hash_combine(bigarr[0], bigarr[1], l3)
> [ 1314s]   Actual: 5088237075164901046
> [ 1314s] Expected: hash_combine_range(bigarr, bigarr + 10)
> [ 1314s] Which is: 16538947639308476466
> [ 1314s] [  FAILED  ] HashingTest.HashCombineBasicTest (0 ms)
> [ 1314s] [----------] 1 test from HashingTest (0 ms total)
> [ 1314s]
> [ 1314s] [----------] Global test environment tear-down
> [ 1314s] [==========] 1 test from 1 test case ran. (0 ms total)
> [ 1314s] [  PASSED  ] 0 tests.
> [ 1314s] [  FAILED  ] 1 test, listed below:
> [ 1314s] [  FAILED  ] HashingTest.HashCombineBasicTest
> [ 1314s]
> [ 1314s]  1 FAILED TEST
>
>
>
>
>
> On Mon, May 12, 2014 at 4:02 PM, Hal Finkel <hfinkel at anl.gov> wrote:
>>
>> ----- Original Message -----
>> > From: "İsmail Dönmez" <ismail at donmez.ws>
>> > To: "Hal Finkel" <hfinkel at anl.gov>
>> > Cc: "LLVM Developers Mailing List" <llvmdev at cs.uiuc.edu>
>> > Sent: Monday, May 12, 2014 7:18:48 AM
>> > Subject: Re: Lots of regtest failures on PPC64/Linux
>> >
>> >
>> > Hi Hal,
>> >
>> >
>> >
>> >
>> >
>> > On Tue, Apr 1, 2014 at 8:34 PM, İsmail Dönmez < ismail at donmez.ws >
>> > wrote:
>> >
>> >
>> >
>> > Hi,
>> >
>> >
>> >
>> >
>> > On Tue, Apr 1, 2014 at 3:26 PM, Hal Finkel < hfinkel at anl.gov > wrote:
>> >
>> >
>> > İsmail,
>> >
>> > I still don't see these errors locally. Can you try with an
>> > autoconf-based build and see if they're somehow related to
>> > (triggered by) the way that cmake builds LLVM?
>> >
>> >
>> >
>> >
>> > On the same machine autoconf build is fine where as cmake has
>> > failures, reproduced simply with
>> >
>> >
>> > cmake -DLLVM_ENABLE_ASSERTIONS=ON -DCMAKE_BUILD_TYPE=Release ..
>> > make -j10
>> > make check-all -j10
>> >
>> >
>> > Did you have a chance to test with cmake? Current clang trunk still
>> > shows 47 test failures on our PPC64 systems.
>>
>> Unfortunately no, this is on my TODO list, but I've not gotten to it yet.
>> Hopefully next week.
>>
>>  -Hal
>>
>> >
>> >
>> > Thanks.
>> >
>>
>> --
>> Hal Finkel
>> Assistant Computational Scientist
>> Leadership Computing Facility
>> Argonne National Laboratory
>
>
>
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
>