CODEWITHSUNDEEP
pythonpyopencl
Nicolas Heimann
Nicolas Heimann

Reputation: 2581

pyopencl global_work_offset kernel argument

I want to use the global_work_offset argument from OpenCL API function clEnqueueNDRangeKernel. I can't figure out how to do that within pyopencl API. Here is a demo code where I want to add an offset of 2 to kernel invocation so get_global_id(0) starts at 2 instead of 0:

import pyopencl as cl 
import pyopencl.array 
import numpy as np

platform = cl.get_platforms()[0]
devices = platform.get_devices()[1] #gpu
context = cl.Context(devices=[devices])
queue =  cl.CommandQueue(context)

kernel = cl.Program(context, """
    __kernel void derp(global char* a) {
        a[get_global_id(0)] = 1;
    }""").build()

buffarr = cl.array.zeros(queue, 4, dtype=np.uint8)
kernel.derp(queue, (2,), None, buffarr.data)

np_data = buffarr.get()

# within this demo the buffer contains currently [1,1,0,0]
assert np.array_equal(np_data, [0,0,1,1])

How to change the code so that assertion won't fail? I Don't want to add an extra argument here to the kernel code.

Upvotes: 3

Views: 478

Answers (1)

Gomiero
Gomiero

Reputation: 2300

As per the documentation, you may pass the global_offset as a named parameter.

The kernel's call becomes:

kernel.derp(queue, (4, 1), None, buffarr.data, global_offset=[2, 0])

The program with the change:

import pyopencl as cl
import pyopencl.array
import numpy as np


platform = cl.get_platforms()[2]
print(platform)
devices = platform.get_devices()[0] #gpu
context = cl.Context(devices=[devices])
queue =  cl.CommandQueue(context)

kernel = cl.Program(context, """
    __kernel void derp(global char* a) {
        a[get_global_id(0)] = 1;
    }""").build()


buffarr = cl.array.zeros(queue, 4, dtype=np.uint8)

# (4, 1) ==> shape of the buffer
kernel.derp(queue, (4, 1), None, buffarr.data, global_offset=[2, 0])

np_data = buffarr.get()
print(np_data)
# within this demo the buffer contains currently [1,1,0,0]
assert np.array_equal(np_data, [0,0,1,1])
print("Ok")

After the execution:

On device 0

<pyopencl.Platform 'Intel(R) OpenCL' at 0x60bdc0>
[0 0 1 1]
Ok

On device 1

<pyopencl.Platform 'Experimental OpenCL 2.0 CPU Only Platform' at 0xb60a20>
[0 0 1 1]
Ok

On device 2

<pyopencl.Platform 'NVIDIA CUDA' at 0xff0440>
[0 0 1 1]
Ok

tested with python 2.7.11 [MSC v.1500 64 bit (AMD64)] - pyopencl (2015, 1)

Upvotes: 3

Related Questions