kernel-tuner 0.1.2__tar.gz → 0.1.3__tar.gz

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 1.1
 Name: kernel_tuner
-Version: 0.1.2
+Version: 0.1.3
 Summary: A simple CUDA/OpenCL kernel tuner in Python
 Home-page: http://benvanwerkhoven.github.io/kernel_tuner/
 Author: Ben van Werkhoven
@@ -44,41 +44,34 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
         Installation
         ------------
         
-        To install from PyPi:  
+        The easiest way to install the Kernel Tuner is using pip:
         
-        - ``pip install kernel_tuner``  
+        .. code-block:: bash
         
-        To install from the source (and get the tutorial and examples):  
+            pip install kernel_tuner
         
-        - clone the repository  
-         
-          - With a GitHub account: ``git clone git@github.com:benvanwerkhoven/kernel_tuner.git``  
-          - Without GitHub account: ``git clone https://github.com/benvanwerkhoven/kernel_tuner.git``  
-           
-        - change into the top-level directory  
+        But you can also install from the git repository. This way you also get the
+        examples and the tutorials:
         
-          - ``cd kernel_tuner``  
-           
-        - install using  
-         
-          - ``pip install .``  
+        .. code-block:: bash
         
-        Dependencies
-        ------------
-        Python 2.7 or Python 3.5
-        
-        PyCuda and/or PyOpenCL (https://mathema.tician.de/software/)
-        
-        - To tune CUDA kernels
+            git clone https://github.com/benvanwerkhoven/kernel_tuner.git
+            cd kernel_tuner
+            pip install -r requirements.txt
+            pip install .
+            
+        To tune CUDA kernels:
         
-          - Make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed 
-          - You can install PyCuda using ``pip install pycuda`` 
+          - First, make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed
+          - You can install PyCuda using ``pip install pycuda``
         
-        - To tune OpenCL kernels
+        To tune OpenCL kernels:
         
-          - Make sure you have an OpenCL compiler for your intended OpenCL platform
+          - First, make sure you have an OpenCL compiler for your intended OpenCL platform
           - You can install PyOpenCL using ``pip install pyopencl``
         
+        If you need more information about how to install the Kernel Tuner and all dependencies see the `installation guide <http://benvanwerkhoven.github.io/kernel_tuner/install.html>`__
+        
         Example usage
         -------------
         
@@ -104,7 +97,7 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
             args = [c, a, b, n]
         
             tune_params = dict()
-            tune_params["block_size_x"] = [128+64*i for i in range(15)]
+            tune_params["block_size_x"] = [32, 64, 128, 256, 512]
         
             tune_kernel("vector_add", kernel_string, size, args, tune_params)
         
@@ -136,7 +129,7 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
         Tuning host and kernel code
         ---------------------------
         
-        It is also possible to tune for combinations of tunable parameters in
+        It is possible to tune for combinations of tunable parameters in
         both host and kernel code. This allows for a number of powerfull things,
         such as tuning the number of streams for a kernel that uses CUDA Streams
         or OpenCL Command Queues to overlap transfers between host and device
@@ -160,33 +153,10 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
             answer = [a+b, None, None]  # the order matches the arguments (in args) to the kernel
             tune_kernel("vector_add", kernel_string, size, args, tune_params, answer=answer)
         
-        Contribution guide
-        ------------------
-        
-        The kernel tuner follows the Google Python style guide, with Sphinxdoc
-        docstrings for module public functions. If you want to contribute to the
-        project please fork it, create a branch including your addition, and
-        create a pull request.
-        
-        The tests use relative imports and can be run directly after making
-        changes to the code. To run all tests use ``nosetests`` in the main
-        directory. To run the examples after code changes, you need to run
-        ``pip install --upgrade .`` in the main directory. Documentation is
-        generated by typing ``make html`` in the doc directory, the contents of
-        ``doc/build/html/`` should then be copied to ``sphinxdoc`` directory of
-        the ``gh-pages`` branch.
-        
-        Before creating a pull request please ensure the following:
-        
-        * You have written unit tests to test your additions and all unit tests pass
-        * The examples still work and produce the same (or better) results
-        * The code is compatible with both Python 2.7 and Python 3.5
-        * An entry about the change or addition is created in CHANGELOG.md
-        
-        Contributing authors so far:
+        Contributing
+        ------------
         
-        * Ben van Werkhoven
-        * Berend Weel
+        Please see the `Contributions Guide <http://benvanwerkhoven.github.io/kernel_tuner/contributing.html>`__.
         
         Citation
         --------

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/README.rst

@@ -36,41 +36,34 @@ The full documentation is available
 Installation
 ------------
 
-To install from PyPi:  
+The easiest way to install the Kernel Tuner is using pip:
 
-- ``pip install kernel_tuner``  
+.. code-block:: bash
 
-To install from the source (and get the tutorial and examples):  
+    pip install kernel_tuner
 
-- clone the repository  
- 
-  - With a GitHub account: ``git clone git@github.com:benvanwerkhoven/kernel_tuner.git``  
-  - Without GitHub account: ``git clone https://github.com/benvanwerkhoven/kernel_tuner.git``  
-   
-- change into the top-level directory  
+But you can also install from the git repository. This way you also get the
+examples and the tutorials:
 
-  - ``cd kernel_tuner``  
-   
-- install using  
- 
-  - ``pip install .``  
+.. code-block:: bash
 
-Dependencies
-------------
-Python 2.7 or Python 3.5
-
-PyCuda and/or PyOpenCL (https://mathema.tician.de/software/)
-
-- To tune CUDA kernels
+    git clone https://github.com/benvanwerkhoven/kernel_tuner.git
+    cd kernel_tuner
+    pip install -r requirements.txt
+    pip install .
+    
+To tune CUDA kernels:
 
-  - Make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed 
-  - You can install PyCuda using ``pip install pycuda`` 
+  - First, make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed
+  - You can install PyCuda using ``pip install pycuda``
 
-- To tune OpenCL kernels
+To tune OpenCL kernels:
 
-  - Make sure you have an OpenCL compiler for your intended OpenCL platform
+  - First, make sure you have an OpenCL compiler for your intended OpenCL platform
   - You can install PyOpenCL using ``pip install pyopencl``
 
+If you need more information about how to install the Kernel Tuner and all dependencies see the `installation guide <http://benvanwerkhoven.github.io/kernel_tuner/install.html>`__
+
 Example usage
 -------------
 
@@ -96,7 +89,7 @@ The following shows a simple example for tuning a CUDA kernel:
     args = [c, a, b, n]
 
     tune_params = dict()
-    tune_params["block_size_x"] = [128+64*i for i in range(15)]
+    tune_params["block_size_x"] = [32, 64, 128, 256, 512]
 
     tune_kernel("vector_add", kernel_string, size, args, tune_params)
 
@@ -128,7 +121,7 @@ kernels and the scripts to tune them.
 Tuning host and kernel code
 ---------------------------
 
-It is also possible to tune for combinations of tunable parameters in
+It is possible to tune for combinations of tunable parameters in
 both host and kernel code. This allows for a number of powerfull things,
 such as tuning the number of streams for a kernel that uses CUDA Streams
 or OpenCL Command Queues to overlap transfers between host and device
@@ -152,33 +145,10 @@ expected output of the kernel. Input arguments are replaced with None.
     answer = [a+b, None, None]  # the order matches the arguments (in args) to the kernel
     tune_kernel("vector_add", kernel_string, size, args, tune_params, answer=answer)
 
-Contribution guide
-------------------
-
-The kernel tuner follows the Google Python style guide, with Sphinxdoc
-docstrings for module public functions. If you want to contribute to the
-project please fork it, create a branch including your addition, and
-create a pull request.
-
-The tests use relative imports and can be run directly after making
-changes to the code. To run all tests use ``nosetests`` in the main
-directory. To run the examples after code changes, you need to run
-``pip install --upgrade .`` in the main directory. Documentation is
-generated by typing ``make html`` in the doc directory, the contents of
-``doc/build/html/`` should then be copied to ``sphinxdoc`` directory of
-the ``gh-pages`` branch.
-
-Before creating a pull request please ensure the following:
-
-* You have written unit tests to test your additions and all unit tests pass
-* The examples still work and produce the same (or better) results
-* The code is compatible with both Python 2.7 and Python 3.5
-* An entry about the change or addition is created in CHANGELOG.md
-
-Contributing authors so far:
+Contributing
+------------
 
-* Ben van Werkhoven
-* Berend Weel
+Please see the `Contributions Guide <http://benvanwerkhoven.github.io/kernel_tuner/contributing.html>`__.
 
 Citation
 --------

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/kernel_tuner/c.py

@@ -61,14 +61,22 @@ class CFunctions(object):
                 self.arg_mapping[str(ctype_args[-1])] = arg.shape
             elif numpy.isscalar(arg):
                 if hasattr(arg, 'dtype'):
-                    if str(arg.dtype).startswith('int'):
-                        ctype_args.append(int(arg))
-                    elif str(arg.dtype).startswith('float'):
-                        ctype_args.append(float(arg))
+                    np_to_c_type_map = { numpy.int32: C.c_int32,
+                            numpy.int64: C.c_int64,
+                            numpy.float32: C.c_float,
+                            numpy.float64: C.c_double }
+                    if type(arg) in np_to_c_type_map:
+                        ctype_args.append(np_to_c_type_map[type(arg)](arg))
                     else:
                         raise TypeError("Argument is scalar with a dtype, but does not start with int or float")
                 else:
-                    ctype_args.append(arg)
+                    if isinstance(arg, int):
+                        ctype_args.append(C.c_int(arg))
+                    elif isinstance(arg, float):
+                        ctype_args.append(C.c_float(arg))
+                    else:
+                        raise TypeError("Argument is scalar without dtype, and is not instance of int or float")
+                self.arg_mapping[str(ctype_args[-1])] = ()
             else:
                 raise TypeError("Argument is not a numpy.ndarray and is not a scalar %s" % type(arg))
 
@@ -196,6 +204,9 @@ class CFunctions(object):
         :returns: A robust average of values returned by the C function.
         :rtype: float
         """
+        logging.debug("run_kernel")
+        logging.debug("arguments=" + str([str(arg) for arg in c_args]))
+
         time = func(*c_args)
 
         #I would like to replace the following with actually capturing

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/kernel_tuner/interface.py

@@ -338,8 +338,11 @@ def run_kernel(kernel_name, original_kernel, problem_size, arguments,
     #copy data in GPU memory back to the host
     results = []
     for i, arg in enumerate(arguments):
-        results.append(numpy.zeros_like(arg))
-        dev.memcpy_dtoh(results[-1], gpu_args[i])
+        if numpy.isscalar(arg):
+            results.append(arg)
+        else:
+            results.append(numpy.zeros_like(arg))
+            dev.memcpy_dtoh(results[-1], gpu_args[i])
     return results
 
 

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/kernel_tuner/util.py

@@ -5,18 +5,27 @@ import os
 import errno
 
 def get_instance_string(params):
+    """ combine the parameters to a string mostly used for debug output
+        use of OrderedDict is advised
+    """
     return "_".join([str(i) for i in params.values()])
 
 def get_config_string(params):
+    """ return a compact string representation of a dictionary """
     return "".join([k + "=" + str(v) + ", " for k,v in params.items()])
 
 def get_kernel_string(original_kernel):
+    """ retrieves kernel string from a file if the string passed looks like filename
+        if the string does look like a filename, but the file does not exist, it
+        is assumed that the string is not a filename after all.
+    """
     kernel_string = original_kernel
     if looks_like_a_filename(original_kernel):
         kernel_string = read_file(original_kernel) or original_kernel
     return kernel_string
 
 def delete_temp_file(filename):
+    """ delete a temporary file, don't complain if is no longer exists """
     try:
         os.remove(filename)
     except OSError as e:
@@ -24,6 +33,7 @@ def delete_temp_file(filename):
             raise e
 
 def get_temp_filename():
+    """ return a string in the form of temp_X, where X is a large integer """
     random_large_int = numpy.random.randint(low=100, high=100000000000)
     return 'temp_' + str(random_large_int)
 
@@ -33,7 +43,7 @@ def looks_like_a_filename(original_kernel):
     if isinstance(original_kernel, str):
         result = True
         #test if not too long
-        if len(original_kernel) > 100:
+        if len(original_kernel) > 250:
             result = False
         #test if not contains special characters
         for c in "();{}\\":
@@ -48,13 +58,15 @@ def looks_like_a_filename(original_kernel):
     return result
 
 def read_file(filename):
+    """ return the contents of the file named filename or None if file not found """
     if os.path.isfile(filename):
         with open(filename, 'r') as f:
             return f.read()
 
 def write_file(filename, string):
-    #ugly fix, hopefully we can find a better one
+    """ dump the contents of string to a file called filename """
     import sys
+    #ugly fix, hopefully we can find a better one
     if sys.version_info[0] >= 3:
         with open(filename, 'w', encoding="utf-8") as f:
             f.write(string)
@@ -82,8 +94,7 @@ def get_problem_size(problem_size, params):
     if isinstance(problem_size, (str,int,numpy.integer)):
         problem_size = (problem_size, )
     current_problem_size = [1, 1, 1]
-    for i in range(len(problem_size)):
-        s = problem_size[i]
+    for i, s in enumerate(problem_size):
         if isinstance(s, str):
             current_problem_size[i] = int(eval(replace_param_occurrences(s,params)))
         elif isinstance(s, int) or isinstance(s, numpy.integer):
@@ -177,6 +188,7 @@ def replace_param_occurrences(string, params):
     return string
 
 def check_restrictions(restrictions, element, keys, verbose):
+    """ check whether a specific instance meets the search space restrictions """
     params = dict(zip(keys, element))
     for restrict in restrictions:
         if not eval(replace_param_occurrences(restrict, params)):
@@ -187,25 +199,26 @@ def check_restrictions(restrictions, element, keys, verbose):
     return True
 
 def check_argument_list(args):
+    """ raise an exception if a kernel argument is of unsupported type """
     for (i, arg) in enumerate(args):
         if not isinstance(arg, (numpy.ndarray, numpy.generic)):
             raise TypeError("Argument at position " + str(i) + " of type: " + str(type(arg)) + " should be of type numpy.ndarray or numpy scalar")
 
 def setup_block_and_grid(dev, problem_size, grid_div_z, grid_div_y, grid_div_x, params, instance_string, verbose):
-        """compute problem size, thread block and grid dimensions for this kernel"""
-        threads = get_thread_block_dimensions(params)
-        if numpy.prod(threads) > dev.max_threads:
-            if verbose:
-                print("skipping config", instance_string, "reason: too many threads per block")
-            return None, None
-        current_problem_size = get_problem_size(problem_size, params)
-        grid = get_grid_dimensions(current_problem_size, params, grid_div_z, grid_div_y, grid_div_x)
-        return threads, grid
+    """compute problem size, thread block and grid dimensions for this kernel"""
+    threads = get_thread_block_dimensions(params)
+    if numpy.prod(threads) > dev.max_threads:
+        if verbose:
+            print("skipping config", instance_string, "reason: too many threads per block")
+        return None, None
+    current_problem_size = get_problem_size(problem_size, params)
+    grid = get_grid_dimensions(current_problem_size, params, grid_div_z, grid_div_y, grid_div_x)
+    return threads, grid
 
 def setup_kernel_strings(kernel_name, original_kernel, params, grid, instance_string):
-        """create configuration specific kernel string"""
-        kernel_string = prepare_kernel_string(original_kernel, params, grid)
-        name = kernel_name + "_" + instance_string
-        kernel_string = kernel_string.replace(kernel_name, name)
-        return name, kernel_string
+    """create configuration specific kernel string"""
+    kernel_string = prepare_kernel_string(original_kernel, params, grid)
+    name = kernel_name + "_" + instance_string
+    kernel_string = kernel_string.replace(kernel_name, name)
+    return name, kernel_string
 

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/kernel_tuner.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 1.1
 Name: kernel-tuner
-Version: 0.1.2
+Version: 0.1.3
 Summary: A simple CUDA/OpenCL kernel tuner in Python
 Home-page: http://benvanwerkhoven.github.io/kernel_tuner/
 Author: Ben van Werkhoven
@@ -44,41 +44,34 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
         Installation
         ------------
         
-        To install from PyPi:  
+        The easiest way to install the Kernel Tuner is using pip:
         
-        - ``pip install kernel_tuner``  
+        .. code-block:: bash
         
-        To install from the source (and get the tutorial and examples):  
+            pip install kernel_tuner
         
-        - clone the repository  
-         
-          - With a GitHub account: ``git clone git@github.com:benvanwerkhoven/kernel_tuner.git``  
-          - Without GitHub account: ``git clone https://github.com/benvanwerkhoven/kernel_tuner.git``  
-           
-        - change into the top-level directory  
+        But you can also install from the git repository. This way you also get the
+        examples and the tutorials:
         
-          - ``cd kernel_tuner``  
-           
-        - install using  
-         
-          - ``pip install .``  
+        .. code-block:: bash
         
-        Dependencies
-        ------------
-        Python 2.7 or Python 3.5
-        
-        PyCuda and/or PyOpenCL (https://mathema.tician.de/software/)
-        
-        - To tune CUDA kernels
+            git clone https://github.com/benvanwerkhoven/kernel_tuner.git
+            cd kernel_tuner
+            pip install -r requirements.txt
+            pip install .
+            
+        To tune CUDA kernels:
         
-          - Make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed 
-          - You can install PyCuda using ``pip install pycuda`` 
+          - First, make sure you have the `CUDA Toolkit <https://developer.nvidia.com/cuda-toolkit>`_ installed
+          - You can install PyCuda using ``pip install pycuda``
         
-        - To tune OpenCL kernels
+        To tune OpenCL kernels:
         
-          - Make sure you have an OpenCL compiler for your intended OpenCL platform
+          - First, make sure you have an OpenCL compiler for your intended OpenCL platform
           - You can install PyOpenCL using ``pip install pyopencl``
         
+        If you need more information about how to install the Kernel Tuner and all dependencies see the `installation guide <http://benvanwerkhoven.github.io/kernel_tuner/install.html>`__
+        
         Example usage
         -------------
         
@@ -104,7 +97,7 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
             args = [c, a, b, n]
         
             tune_params = dict()
-            tune_params["block_size_x"] = [128+64*i for i in range(15)]
+            tune_params["block_size_x"] = [32, 64, 128, 256, 512]
         
             tune_kernel("vector_add", kernel_string, size, args, tune_params)
         
@@ -136,7 +129,7 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
         Tuning host and kernel code
         ---------------------------
         
-        It is also possible to tune for combinations of tunable parameters in
+        It is possible to tune for combinations of tunable parameters in
         both host and kernel code. This allows for a number of powerfull things,
         such as tuning the number of streams for a kernel that uses CUDA Streams
         or OpenCL Command Queues to overlap transfers between host and device
@@ -160,33 +153,10 @@ Description: A simple CUDA/OpenCL kernel tuner in Python
             answer = [a+b, None, None]  # the order matches the arguments (in args) to the kernel
             tune_kernel("vector_add", kernel_string, size, args, tune_params, answer=answer)
         
-        Contribution guide
-        ------------------
-        
-        The kernel tuner follows the Google Python style guide, with Sphinxdoc
-        docstrings for module public functions. If you want to contribute to the
-        project please fork it, create a branch including your addition, and
-        create a pull request.
-        
-        The tests use relative imports and can be run directly after making
-        changes to the code. To run all tests use ``nosetests`` in the main
-        directory. To run the examples after code changes, you need to run
-        ``pip install --upgrade .`` in the main directory. Documentation is
-        generated by typing ``make html`` in the doc directory, the contents of
-        ``doc/build/html/`` should then be copied to ``sphinxdoc`` directory of
-        the ``gh-pages`` branch.
-        
-        Before creating a pull request please ensure the following:
-        
-        * You have written unit tests to test your additions and all unit tests pass
-        * The examples still work and produce the same (or better) results
-        * The code is compatible with both Python 2.7 and Python 3.5
-        * An entry about the change or addition is created in CHANGELOG.md
-        
-        Contributing authors so far:
+        Contributing
+        ------------
         
-        * Ben van Werkhoven
-        * Berend Weel
+        Please see the `Contributions Guide <http://benvanwerkhoven.github.io/kernel_tuner/contributing.html>`__.
         
         Citation
         --------

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/kernel_tuner.egg-info/SOURCES.txt

@@ -14,6 +14,7 @@ kernel_tuner/util.py
 kernel_tuner.egg-info/PKG-INFO
 kernel_tuner.egg-info/SOURCES.txt
 kernel_tuner.egg-info/dependency_links.txt
+kernel_tuner.egg-info/requires.txt
 kernel_tuner.egg-info/top_level.txt
 kernel_tuner/runners/__init__.py
 kernel_tuner/runners/random_sample.py

kernel_tuner-0.1.3/kernel_tuner.egg-info/requires.txt

@@ -0,0 +1 @@
+numpy>=1.7.1

kernel_tuner-0.1.3/requirements.txt

@@ -0,0 +1 @@
+numpy>=1.7.1

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/setup.py

@@ -8,9 +8,13 @@ def readme():
     with open('README.rst') as f:
         return f.read()
 
+def required():
+    with open('requirements.txt') as f:
+        return f.read().splitlines()
+
 setup(
     name = "kernel_tuner",
-    version = "0.1.2",
+    version = "0.1.3",
     author = "Ben van Werkhoven",
     author_email = "b.vanwerkhoven@esciencecenter.nl",
     description = ("A simple CUDA/OpenCL kernel tuner in Python"),
@@ -19,6 +23,7 @@ setup(
     url = "http://benvanwerkhoven.github.io/kernel_tuner/",
     packages=['kernel_tuner', 'kernel_tuner.runners'],
     long_description=readme(),
+    install_requires=required(),
     classifiers=[
         'Environment :: Console',
         'Intended Audience :: Developers',

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/test/test_c_functions.py

@@ -48,14 +48,14 @@ def test_ready_argument_list2():
     output_arg1 = numpy.ctypeslib.as_array(output[0], shape=arg1.shape)
 
     assert output_arg1.dtype == 'float32'
-    assert isinstance(output[1], int)
-    assert isinstance(output[2], float)
-    assert isinstance(output[3], int)
+    assert isinstance(output[1], C.c_int)
+    assert isinstance(output[2], C.c_float)
+    assert isinstance(output[3], C.c_int)
 
     assert all(output_arg1 == arg1)
-    assert output[1] == arg2
-    assert output[2] == arg3
-    assert output[3] == arg4
+    assert output[1].value == arg2
+    assert output[2].value == arg3
+    assert output[3].value == arg4
 
 
 def test_ready_argument_list3():

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/test/test_interface.py

@@ -96,11 +96,11 @@ def test_run_kernel(dev_interface):
     args = [n]
     tune_params = dict()
     tune_params["block_size_x"] = 128
-    run_kernel("fake_kernel", kernel_string, problem_size, args, tune_params)
+    answer = run_kernel("fake_kernel", kernel_string, problem_size, args, tune_params)
 
     assert dev.compile.call_count == 1
     dev.run_kernel.assert_called_once_with('compile', 'ready_argument_list', (128, 1, 1), (10, 1, 1))
-    dev.memcpy_dtoh.assert_called_once_with(numpy.zeros(1), 'r')
+    assert answer[0] == n
 
 @patch('kernel_tuner.core.CudaFunctions')
 def test_check_kernel_correctness(dev_interface):

kernel_tuner-0.1.2/requirements.txt

@@ -1,7 +0,0 @@
-numpy>=1.7.1
-mock>=2.0.0
-nose>=1.3.7
-pytest>=3.0.3
-pycuda>=2016.1.1
-pyopencl>=2015.2.4
-

{kernel_tuner-0.1.2 → kernel_tuner-0.1.3}/setup.cfg

@@ -2,7 +2,7 @@
 description-file = README.rst
 
 [egg_info]
-tag_build = 
-tag_date = 0
 tag_svn_revision = 0
+tag_date = 0
+tag_build =