Closed bathmatt closed 7 years ago
Never saw that before. Can I see the full function call? Can you push this to a branch that I can see (perhaps inside SNL SRN)?
It is in trilinos' git develop that I'm seeing this.
FUNCTION(DREKAR_ADD_EXODIFF_TEST IN_TEST_NAME IN_EXECUTABLE
IN_INPUT_XML_FILE IN_OUTPUT_EXO IN_GOLD_STANDARD_EXO IN_EXODIFF_FILE)
IF (${PROJECT_NAME}_VERBOSE_CONFIGURE)
MESSAGE("\nDREKAR_ADD_EXODIFF_TEST: ${IN_TEST_NAME}\n")
ENDIF()
MESSAGE("\nDREKAR_ADD_EXODIFF_TEST: ${IN_TEST_NAME}\n")
TRIBITS_ADD_ADVANCED_TEST(
${IN_TEST_NAME}
OVERALL_WORKING_DIRECTORY "TEST_NAME"
SKIP_CLEAN_OVERALL_WORKING_DIRECTORY TRUE
TEST_0 EXEC ${IN_EXECUTABLE} NOEXEPREFIX NOEXESUFFIX
ARGS --i=${IN_INPUT_XML_FILE}
PASS_REGULAR_EXPRESSION "Drekar run completed."
TEST_1 EXEC ${SEACAS_BINARY_DIR}/applications/epu/epu NOEXEPREFIX
NOEXESUFFIX
ARGS "-auto" "${IN_OUTPUT_EXO}.${MPI_EXEC_MAX_NUMPROCS}."
NUM_MPI_PROCS 1
TEST_2 EXEC ${SEACAS_BINARY_DIR}/applications/exodiff/exodiff
NOEXEPREFIX NOEXESUFFIX
ARGS -file ${IN_EXODIFF_FILE} ${IN_OUTPUT_EXO}
${IN_GOLD_STANDARD_EXO}
NUM_MPI_PROCS 1
)
ENDFUNCTION()
On Tue, Feb 28, 2017 at 7:33 PM, Roscoe A. Bartlett < notifications@github.com> wrote:
Never saw that before. Can I see the full function call? Can you push this to a branch that I can see (perhaps inside SNL SRN)?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/TriBITSPub/TriBITS/issues/180#issuecomment-283226987, or mute the thread https://github.com/notifications/unsubscribe-auth/AOPDIJixXskQROtyx539cuf-RtEgUcduks5rhNj_gaJpZM4MPAmP .
@bathmatt,
You need to change:
SKIP_CLEAN_OVERALL_WORKING_DIRECTORY TRUE
to
SKIP_CLEAN_OVERALL_WORKING_DIRECTORY
as documented at:
The way the parser used by this works is very strange but what it is doing is it is doing is removing the option SKIP_CLEAN_OVERALL_WORKING_DIRECTORY
and then lumping TRUE
with the argument OVERALL_WORKING_DIRECTORY TEST_NAME TRUE
.
We need to add better checking for these arguments. We got that parser from someone else back in 2008 before I even started working with CMake. Now that CMake provides a native parser function:
we might look into using that to see if it does something more sane (and might even go faster).
@bathmatt, is this fixed on your end? Can I close this Issue?
BTW, I created #181 to replace the input function parser which might help catch and report problems like this a little better.
Thanks...
On Thu, Mar 2, 2017 at 12:30 PM, Roscoe A. Bartlett < notifications@github.com> wrote:
@bathmatt https://github.com/bathmatt, is this fixed on your end? Can I close this Issue?
BTW, I created #181 https://github.com/TriBITSPub/TriBITS/issues/181 to replace the input function parser which might help catch and report problems like this a little better.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/TriBITSPub/TriBITS/issues/180#issuecomment-283754656, or mute the thread https://github.com/notifications/unsubscribe-auth/AOPDII53hjV3tUkjevRsOGute9AJ6MO9ks5rhxjlgaJpZM4MPAmP .
Thanks...
Closing as complete after confirmation from @bathmatt.
If I add a test as follows
I get a cmake file that says Drekar_plasma_oscillation_rtc_2D.cmake:SET(OVERALL_WORKING_DIRECTORY "TEST_NAMETRUE")
Why did it add the TRUE at the end of TEST_NAME?