t3006-resource-advanced.t produces errors to stderr when run with verbose and debug set

[SteVwonder]

When running the testsuite with debug=t and verbose=t the t3006-resource-advanced.t test produces the following error output, which causes (erroneous?) test failures:

herbein1@80012f26c745:/usr/src/t$ verbose=t debug=t make check TESTS=t3006-resource-advanced.t
make  t3006-resource-advanced.t
make[1]: Entering directory '/usr/src/t'
make[1]: Nothing to be done for 't3006-resource-advanced.t'.
make[1]: Leaving directory '/usr/src/t'
make  check-TESTS
make[1]: Entering directory '/usr/src/t'
make[2]: Entering directory '/usr/src/t'
ERROR: t3006-resource-advanced.t 3 # OUT-OF-ORDER (expecting 1)
ERROR: t3006-resource-advanced.t 4 # OUT-OF-ORDER (expecting 2)
ERROR: t3006-resource-advanced.t - too few tests run (expected 4, got 2)
============================================================================
Testsuite summary for flux-sched 0.10.0-38-g8f6f4e7c
============================================================================
# TOTAL: 3
# PASS:  0
# SKIP:  0
# XFAIL: 0
# FAIL:  0
# XPASS: 0
# ERROR: 3
============================================================================
See t/test-suite.log
============================================================================
make[2]: *** [Makefile:692: test-suite.log] Error 1
make[2]: Leaving directory '/usr/src/t'
make[1]: *** [Makefile:800: check-TESTS] Error 2
make[1]: Leaving directory '/usr/src/t'
make: *** [Makefile:867: check-am] Error 2

❯ cat ../t3006-resource-advanced.log                                                                                                                                                                                           21:02:46 ()
ERROR: by_subplan: planner_multi_avail_during returned -1.
Numerical result out of range.
by_subplan: planner_multi_avail_during returned -1.
Numerical result out of range.
<snip>
ERROR: by_subplan: planner_multi_avail_during returned -1.
Numerical result out of range.
by_subplan: planner_multi_avail_during returned -1.
Numerical result out of range.
<snip>
sharness: loading extensions from /usr/src/t/sharness.d/flux-sharness.sh
sharness: loading extensions from /usr/src/t/sharness.d/sched-sharness.sh
expecting success:
    sed "s~@TEST_SRCDIR@~${SHARNESS_TEST_SRCDIR}~g" ${cmds001} > cmds001 &&
    ${query} -L ${grugs} -S CA -P high -t 001.R.out < cmds001 &&
    test_cmp 001.R.out ${exp_dir}/001.R.out

INFO: Loading a matcher: CA
resource-query> resource-query> resource-query> ok 1 - match allocate with advanced critieria including BB (pol=hi)

expecting success:
    sed "s~@TEST_SRCDIR@~${SHARNESS_TEST_SRCDIR}~g" ${cmds002} > cmds002 &&
    ${query} -L ${grugs} -S CA -P low -t 002.R.out < cmds002 &&
    test_cmp 002.R.out ${exp_dir}/002.R.out

INFO: Loading a matcher: CA
resource-query> resource-query> resource-query> ok 2 - match allocate with advanced critieria including BB (pol=low)

skipping test: match for a large system with disaggregated resources (pol=high)
    sed "s~@TEST_SRCDIR@~${SHARNESS_TEST_SRCDIR}~g" ${cmds003} > cmds003 &&
    ${query} -L ${disag} -S CA -P high -t 003.R.out -r 400000< cmds003 &&
    test_cmp 003.R.out ${exp_dir}/003.R.out

ok 3 # skip match for a large system with disaggregated resources (pol=high) (missing LONGTEST)
ERROR: t3006-resource-advanced.t 3 # OUT-OF-ORDER (expecting 1)

skipping test: match for a large system with disaggregated resources (pol=low)
    sed "s~@TEST_SRCDIR@~${SHARNESS_TEST_SRCDIR}~g" ${cmds004} > cmds004 &&
    ${query} -L ${disag} -S CA -P low -t 004.R.out -r 400000 < cmds004 &&
    test_cmp 004.R.out ${exp_dir}/004.R.out

ok 4 # skip match for a large system with disaggregated resources (pol=low) (missing LONGTEST)
ERROR: t3006-resource-advanced.t 4 # OUT-OF-ORDER (expecting 2)

# passed all 4 test(s)
1..4
ERROR: t3006-resource-advanced.t - too few tests run (expected 4, got 2)

The 001.R.out and 002.R.out files generated by these tests are populated with what looks like correct content. So I'm guessing this is a valid control flow path that is errantly being marked as a failure, leading to extraneous error printing.

[dongahn]

@SteVwonder Probably not an issue any longer?