1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
|
TESTS for CPPTRAJ:
Daniel R. Roe 2010
Updated 2016-02-02
This directory contains tests which check most of the functionality of cpptraj.
To run the tests, type 'make test' or 'make test.all'. In order to run the tests
in parallel, set the environment variable DO_PARALLEL to your MPI command, e.g.
make clean && DO_PARALLEL="mpiexec -n 2" make test.all
Note that in parallel, tests attempt to determine the number of threads being
used from the output from 'DO_PARALLEL nproc'. If your MPI run command produces
extraneous output to STDOUT (e.g. 'ibrun'), you should specify the number of
threads to be used by setting the N_THREADS environment variable, otherwise
tests will probably fail. If running tests for the hybrid MPI/OpenMP binary,
make sure OMP_NUM_THREADS is appropriately set.
The command 'make clean' can be used to clean up all test output.
For developers: tests can automatically be run with memory checking via valgrind
by typing 'test.vg'. This requires valgrind to be installed and in your PATH.
Tests are stored in directories with the naming convention 'Test_X', where 'X'
is a semi-descriptive name indicating what is tested. By default the STDOUT
output from each test is redirected to 'test.out'. If valgrind is
used STDERR is redirected to 'valgrind.out'. If not running tests within
AmberTools (i.e. in a standalone GitHub install), the test results are
summarized in Test_Results.dat and Test_Error.dat. Within AmberTools, test
failures are reported in $AMBERHOME/AmberTools/test/TEST_FAILURES.diff.
Any individual test can be run by itself, e.g. 'make test.general' (see the
Makefile for a full list of targets), or by changing to the test
directory and running './RunTest.sh'. There are several options which can be
passed to any 'RunTest.sh', available by typing './RunTest.sh --help'. These
options can also be passed to Makefile targets via the 'OPT' variable, e.g.
make test.all OPT=time
Note that all $TESTDIR/RunTest.sh tests require and depend on the existence of
MasterTest.sh, which sets up the run environment for each test in a consistent
way.
|
