def test_incorrect_num_of_args():
src = "kernel void A(const int a) {}"
# too many inputs
with test.Raises(ValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[[1], [2], [3]],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
)
# too few inputs
with test.Raises(ValueError):
driver.DriveKernel(env.OclgrindOpenCLEnvironment(),
src, [],
gsize=(1, 1, 1),
lsize=(1, 1, 1))
# incorrect input width (3 ints instead of one)
with test.Raises(ValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[[1, 2, 3]],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
)
def test_empty_kernel():
src = " kernel void A() {} "
outputs = driver.DriveKernel(env.OclgrindOpenCLEnvironment(),
src, [],
gsize=(1, 1, 1),
lsize=(1, 1, 1))
assert len(outputs) == 0
def test_gsize_smaller_than_lsize():
src = "kernel void A() {}"
with test.Raises(ValueError):
driver.DriveKernel(env.OclgrindOpenCLEnvironment(),
src, [],
gsize=(4, 1, 1),
lsize=(8, 1, 1))
def test_zero_size_input():
src = "kernel void A(global int* a) {}"
with test.Raises(ValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[[]],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
)
def test_syntax_error():
src = "kernel void A(gl ob a l i nt* a) {}"
with testlib.DevNullRedirect():
with test.Raises(driver.OpenCLValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[[]],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
)
def test_timeout():
# non-terminating kernel
src = "kernel void A() { while (true) ; }"
with test.Raises(driver.Timeout):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
timeout=1,
)
def test_invalid_sizes():
src = "kernel void A() {}"
# invalid global size
with test.Raises(ValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[],
gsize=(0, -4, 1),
lsize=(1, 1, 1),
)
# invalid local size
with test.Raises(ValueError):
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[],
gsize=(1, 1, 1),
lsize=(-1, 1, 1),
)
def test_vector_input():
inputs = [[0, 1, 2, 3, 0, 1, 2, 3], [2, 4]]
inputs_orig = [[0, 1, 2, 3, 0, 1, 2, 3], [2, 4]]
outputs_gs = [[0, 2, 4, 6, 0, 4, 8, 12], [2, 4]]
src = """
kernel void A(global int* a, const int2 b) {
const int x_id = get_global_id(0);
const int y_id = get_global_id(1);
if (!y_id) {
a[x_id] *= b.x;
} else {
a[get_global_size(0) + x_id] *= b.y;
}
}
"""
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(4, 2, 1),
lsize=(1, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(inputs, inputs_orig)
testlib.Assert2DArraysAlmostEqual(outputs, outputs_gs)
# run kernel a second time with the previous outputs
outputs2 = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
outputs,
gsize=(4, 2, 1),
lsize=(1, 1, 1),
)
outputs2_gs = [[0, 4, 8, 12, 0, 16, 32, 48], [2, 4]]
testlib.Assert2DArraysAlmostEqual(outputs2, outputs2_gs)
def test_data_unchanged():
src = "kernel void A(global int* a, global int* b, const int c) {}"
inputs = data.MakeRand(src, 16)
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(16, 1, 1),
lsize=(1, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(outputs, inputs)
def test_iterative_increment():
src = "kernel void A(global int* a) { a[get_global_id(0)] += 1; }"
d_cl, d_host = [np.arange(16)], np.arange(16)
for _ in range(8):
d_host += 1 # perform computation on host
d_cl = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
d_cl,
gsize=(16, 1, 1),
lsize=(16, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(d_cl, [d_host])
def test_data_zerod():
# zero-ing a randomly initialized array
src = "kernel void A(global int* a) { a[get_global_id(0)] = 0; }"
inputs = data.MakeRand(src, 16)
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(16, 1, 1),
lsize=(4, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(outputs, [np.zeros(16)])
def test_comparison_against_pointer_warning():
src = """
kernel void A(global int* a) {
int id = get_global_id(0);
if (id < a) a += 1;
}
"""
driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
[[0]],
gsize=(1, 1, 1),
lsize=(1, 1, 1),
)
def test_gsize_smaller_than_data():
src = "kernel void A(global int* a) { a[get_global_id(0)] = 0; }"
inputs = [[5, 5, 5, 5, 5, 5, 5, 5]]
outputs_gs = [[0, 0, 0, 0, 5, 5, 5, 5]]
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(4, 1, 1),
lsize=(4, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(outputs, outputs_gs)
def test_vector_input_switch():
src = """
kernel void A(global int2* a) {
const int tid = get_global_id(0);
const int tmp = a[tid].x;
a[tid].x = a[tid].y;
a[tid].y = tmp;
}
"""
inputs = data.MakeArange(src, 4)
outputs_gs = [[1, 0, 3, 2, 5, 4, 7, 6]]
outputs = driver.DriveKernel(env.make_env(),
src,
inputs,
gsize=(4, 1, 1),
lsize=(4, 1, 1))
testlib.Assert2DArraysAlmostEqual(outputs, outputs_gs)
def test_profiling():
src = """
kernel void A(global int* a, constant int* b) {
const int id = get_global_id(0);
a[id] *= b[id];
}
"""
inputs = [np.arange(16), np.arange(16)]
outputs_gs = [np.arange(16)**2, np.arange(16)]
with testlib.DevNullRedirect():
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(16, 1, 1),
lsize=(16, 1, 1),
profiling=True,
)
testlib.Assert2DArraysAlmostEqual(outputs, outputs_gs)
def test_simple():
inputs = [[0, 1, 2, 3, 4, 5, 6, 7]]
inputs_orig = [[0, 1, 2, 3, 4, 5, 6, 7]]
outputs_gs = [[0, 2, 4, 6, 8, 10, 12, 14]]
src = """
kernel void A(global float* a) {
const int x_id = get_global_id(0);
a[x_id] *= 2.0;
}
"""
outputs = driver.DriveKernel(
env.OclgrindOpenCLEnvironment(),
src,
inputs,
gsize=(8, 1, 1),
lsize=(1, 1, 1),
)
testlib.Assert2DArraysAlmostEqual(inputs, inputs_orig)
testlib.Assert2DArraysAlmostEqual(outputs, outputs_gs)
def main(argv):
"""Main entry point."""
if len(argv) > 1:
app.Warning("Unknown arguments: '{}'.".format(" ".join(argv[1:])))
if FLAGS.ls_env:
env.PrintOpenClEnvironments()
# Read kernel source.
src = sys.stdin.read()
# Parse inputs from strings.
gsize = driver.NDRange.FromString(FLAGS.gsize)
lsize = driver.NDRange.FromString(FLAGS.lsize)
data_generator = data.Generator.FromString(FLAGS.generator)
env_ = env.make_env(devtype=FLAGS.devtype,
platform=FLAGS.platform,
device=FLAGS.device)
if FLAGS.compile_only:
inputs = []
else:
inputs = data.MakeData(
src=src,
size=FLAGS.size,
data_generator=data_generator,
scalar_val=FLAGS.scalar_val,
)
drive_args = {
"src": src,
"inputs": inputs,
"gsize": gsize,
"lsize": lsize,
"optimizations": not FLAGS.cl_opt,
"profiling": FLAGS.profiling,
"debug": FLAGS.debug,
"timeout": FLAGS.timeout,
}
if FLAGS.emit_c:
emit_c_args = {
"compile_only": FLAGS.compile_only,
"create_kernel": FLAGS.with_kernel,
}
print(cgen.emit_c(**drive_args, **emit_c_args))
else:
outputs = driver.DriveKernel(**drive_args, env=env_)
# Print result.
if FLAGS.binary:
d = pickle.dumps(outputs)
sys.stdout = io.TextIOWrapper(sys.stdout.detach(),
encoding="latin-1")
print(d.decode("latin-1"), end="", flush=True)
else:
np.set_printoptions(threshold=np.nan)
args_ = [
arg for arg in args.GetKernelArguments(src)
if not arg.address_space == "local"
]
assert len(args_) == len(outputs)
for arr, arg in zip(outputs, args_):
print(f"{arg.name}: {arr}")