[CMake] Strange 2.8.4 memcheck result on CDash

[David Cole]

Can you send me (offlist) copies of the DynamicAnalysis.xml files from
each case? Do you still have them?

If not, no worried. I'll try to reproduce this on our CMake dashboard
by submitting a 2nd valgrind dashboard: same script, just drive it
with 2.8.4. It may take me a few days to get around to this, but I'll
keep you posted.


Thanks,
David


On Wed, Feb 23, 2011 at 4:31 PM, Allen D Byrne <byrn at hdfgroup.org> wrote:
> Further results:
>
> 2.8.2 works with MemoryCheck
>
> 2.8.4 does not display even with MemoryCheck.
>
> Here is out CTestConfig.cmake file:
>
> #############################################
>
> SET (CTEST_PROJECT_NAME "HDF5.1.8")
>
> SET (CTEST_NIGHTLY_START_TIME "20:00:00 CST")
>
> SET (CTEST_DROP_METHOD "http")
>
> SET (CTEST_DROP_SITE "nei.hdfgroup.uiuc.edu")
>
> SET (CTEST_DROP_LOCATION "/cdash/submit.php?project=HDF5.1.8")
>
> SET (CTEST_DROP_SITE_CDASH TRUE)
>
> SET (UPDATE_TYPE svn)
>
> SET (VALGRIND_COMMAND "/usr/bin/valgrind")
>
> SET (VALGRIND_COMMAND_OPTIONS "-v --tool=memcheck --leak-check=full
> --track-fds=yes --num-callers=50 --show-reachable=yes --track-origins=yes
> --malloc-fill=0xff --free-fill=0xfe")
>
> SET (MEMORYCHECK_COMMAND "/usr/bin/valgrind")
>
> SET (MEMORYCHECKCOMMAND "/usr/bin/valgrind")
>
> SET (CTEST_MEMORYCHECK_COMMAND_OPTIONS "-v --tool=memcheck --leak-check=full
> --track-fds=yes --num-callers=50 --show-reachable=yes --track-origins=yes
> --malloc-fill=0xff --free-fill=0xfe")
>
> SET (CTEST_MEMORYCHECKCOMMAND_OPTIONS "-v --tool=memcheck --leak-check=full
> --track-fds=yes --num-callers=50 --show-reachable=yes --track-origins=yes
> --malloc-fill=0xff --free-fill=0xfe")
>
> SET (CTEST_TESTING_TIMEOUT 3600)
>
> SET (DART_TESTING_TIMEOUT 3600)
>
> #############################################
>
> Allen
>
>> It looks to me from:
>
>>
>
>> git gui blame -- Source/cmCTest.cxx
>
>>
>
>> that this is the behavior that has always been -- ever since the -D
>
>> MemCheck -D MemoryCheck stuff has been added to ctest several years
>
>> ago.
>
>>
>
>> Are you sure that nothing else changed (besides CMake from 2.8.2 to
>
>> 2.8.4 in this scenario....)?
>
>>
>
>> Are we sure that executing "ctest -D NightlyMemCheck" with ctest 2.8.2
>
>> was doing what you thought it was doing?
>
>>
>
>> Let me know what you find out after your results are in...
>
>>
>
>>
>
>> Thanks,
>
>> David
>
>>