OpenCL

These notes will introduce OpenCL

ITCS 6/8010 CUDA Programming, UNC-Charlotte, B. Wilkinson, April 7, 2011, [Link]

OpenCL
(Open Computing Language)
A standard based upon C for portable parallel applications Task parallel and data parallel applications Focuses on multi platform support (multiple CPUs, GPUs, )

Development initiated by Apple.

Developed by Khromos group who also managed OpenGL OpenCL 1.0 2008. Released with Max OS 10.6 (Snow Leopard) OpenCL 1.1 June 2010 Similarities with CUDA Implementation available for NVIDIA GPUs
Wikipedia OpenCL [Link] 2

OpenCL Programming Model

Uses data parallel programming model, similar to CUDA Host program launches kernel routines as in CUDA, but allows for just-in-time compilation during host execution. OpenCL work items corresponds to CUDA threads OpenCL work groups corresponds to CUDA thread blocks Work items in same work group can be synchronized with a barrier as in CUDA.
3

Sample OpenCL code to add two vectors

To illustrate OpenCL commands, will used OpenCl code to add two vectors, A and B which are transferred to the device (GPU) and the result, C, returned to host (CPU), similar to CUDA vector addition

Structure of OpenCL main program

Get information about platform and devices available on system

Select devices to use

Create an OpenCL command queue Create memory buffers on device

Transfer data from host to device memory buffers

Create kernel program object Build (compile) kernel in-line (or load precompiled binary) Create OpenCL kernel object Set kernel arguments

Execute kernel
Read kernel memory and copy to host memory. 5

Platform
"The host plus a collection of devices managed by the OpenCL framework that allow an application to share resources and execute kernels on devices in the platform." Platforms represented by a cl_platform object, initialized with clGetPlatformID()

[Link]

Simple code for identifying platform

//Platform
cl_platform_id platform;

clGetPlatformIDs (1, &platform, NULL);

Returns number of OpenCL platforms available. If NULL, ignored.

Number of platform entries

List of OpenCL platforms found. (Platform IDs) In our case just one platform, identified by &platform

Context
The environment within which the kernels execute and the domain in which synchronization and memory management is defined. The context includes a set of devices, the memory accessible to those devices, the corresponding memory properties and one or more command-queues used to schedule execution of a kernel(s) or operations on memory objects.

The OpenCL Specification version 1.1 [Link]

Code for context

//Context cl_context_properties props[3]; props[0] = (cl_context_properties) CL_CONTEXT_PLATFORM; props[1] = (cl_context_properties) platform; props[2] = (cl_context_properties) 0; cl_context GPUContext = clCreateContextFromType(props,CL_DEVICE_TYPE_GPU,NULL,NULL,NULL); //Context info size_t ParmDataBytes; clGetContextInfo(GPUContext,CL_CONTEXT_DEVICES,0,NULL,&ParmDataBytes); cl_device_id* GPUDevices = (cl_device_id*)malloc(ParmDataBytes); clGetContextInfo(GPUContext,CL_CONTEXT_DEVICES,ParmDataBytes,GPUDevices, NULL); 9

Command Queue
An object that holds commands that will be executed on a specific device.

The command-queue is created on a specific device in a context. Commands to a command-queue are queued in-order but may be executed in-order or out-of-order. ...

The OpenCL Specification version 1.1 [Link]

10

Simple code for creating a command queue

// Create command-queue cl_command_queue GPUCommandQueue = clCreateCommandQueue(GPUContext,GPUDevices[0],0,NULL);

11

Allocating memory on device

Use clCreatBuffer:
cl_mem clCreateBuffer(cl_context context, cl_mem_flags flags,
No of bytes in buffer memory object
Returns memory object OpenCL context, from clCreateContextFromType()

size_t size, void *host_ptr, cl_int *errcode_ret)

Bit field to specify type of allocation/usage (CL_MEM_READ_WRITE ,) Ptr to buffer data (May be previously allocated.)

Returns error code if an error

12

Sample code for allocating memory on device for source data

// source data on host, two vectors int *A, A = new B = new for(int A[i] B[i] } *B; int[N]; int[N]; i = 0; i < N; i++) { = rand()%1000; = rand()%1000;

// Allocate GPU memory for source vectors cl_mem GPUVector1 = clCreateBuffer(GPUContext,CL_MEM_READ_ONLY | CL_MEM_COPY_HOST_PTR,sizeof(int)*N, A, NULL); cl_mem GPUVector2 = clCreateBuffer(GPUContext,CL_MEM_READ_ONLY | CL_MEM_COPY_HOST_PTR,sizeof(int)*N, B, NULL); 13

Sample code for allocating memory on device for results on GPU

// Allocate GPU memory for output vector

cl_mem GPUOutputVector = clCreateBuffer(GPUContext,CL_MEM_WRITE_ONLY,sizeof(int)*N, NULL,NULL);

14

Kernel Program

Simple programs might be in the same file as the host code as our CUDA examples. In that case need to formed into strings in a character array. If in a separate file, can read that file into host program as a character string

15

OpenCL qualifier to indicate kernel code

Kernel program

If in same program as host, kernel needs to be strings (I think it can be a single string)

const char* OpenCLSource[] = { "__kernel void vectorAdd (const __global int* a,", " const __global int* b,", OpenCL qualifier to " __global int* c)", indicate kernel memory "{", (Memory objects " unsigned int gid = get_global_id(0);", allocated from global memory pool) " c[gid] = a[gid] + b[gid];", "}" };
Returns global work-item ID in given dimension (0 here)

int main(int argc, char **argv){ Double underscores optional in OpenCL qualifiers } 16

Kernel in a separate file

// Load the kernel source code into the array source_str FILE *fp; char *source_str; size_t source_size; fp = fopen("vector_add_kernel.cl", "r"); if (!fp) { fprintf(stderr, "Failed to load kernel.\n"); exit(1); } source_str = (char*)malloc(MAX_SOURCE_SIZE); source_size = fread( source_str, 1, MAX_SOURCE_SIZE, fp); fclose( fp );
[Link] 17

Create kernel program object

const char* OpenCLSource[] = { }; This example uses a single file for both host and kernel code. Can use clCreateprogramWithSource() with a separate kernel file read into host program

int main(int argc, char **argv) // Create OpenCL program object

Used to return error code if error

cl_program OpenCLProgram = clCreateProgramWithSource(GPUContext,7,OpenCLSource,NULL,NULL);

Number of strings in kernel program array Used if strings not null-terminated to given length of strings 18

Build kernel program

// Build the program (OpenCL JIT compilation) clBuildProgram(OpenCLProgram,0,NULL,NULL,NULL,NULL);
Arguments for notification routine Function ptr to notification routine called with build complete. Then clBuildProgram will return immediately, otherwise only when build complete 19

Program object from clCreateProgramwithSource

Number of devices

Build options

List of devices, if more than one

Creating Kernel Objects

// Create a handle to the compiled OpenCL function
cl_kernel OpenCLVectorAdd = clCreateKernel(OpenCLProgram, "vectorAdd", NULL);

Built prgram from clBuildProgram

Function name with __kernel qualifier

Return error code

20

Set Kernel Arguments

// Set kernel arguments
clSetKernelArg(OpenCLVectorAdd,0,sizeof(cl_mem), (void*)&GPUVector1); clSetKernelArg(OpenCLVectorAdd,1,sizeof(cl_mem), (void*)&GPUVector2);

clSetKernelArg(OpenCLVectorAdd,2,sizeof(cl_mem), (void*)&GPUOutputVector);
Pointer to data for argument, from clCreateBuffer() Kernel object from clCreateKernel() Which argument Size of argument

21

Enqueue a command to execute kernel on device

// Launch the kernel size_t WorkSize[1] = {N}; // Total number of work items size_t localWorkSize[1]={256}; //No of work items in work group

Kernel object from clCreatKernel()

// Launch the kernel

Dimensions of work items

clEnqueueNDRangeKernel(GPUCommandQueue,OpenCLVectorAdd,1,NULL, WorkSize, localWorkSize, 0, NULL, NULL);

Offset used with work item Event wait list Event

Array describing no of global work items

Array describing no of work items that make up a work group

Number of events to complete before this commands

22

Function to copy from buffer object to host memory

The following function enqueue commands to read from a buffer object to host memory:
cl_int clEnqueueReadBuffer (cl_command_queue command_queue, cl_mem buffer, cl_bool blocking_read, size_t offset, size_t cb, void *ptr, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event)

The OpenCL Specification version 1.1 [Link]

23

Function to copy from host memory to buffer object

The following function enqueue commands to write to a buffer object from host memory:
cl_int clEnqueueWriteBuffer (cl_command_queue command_queue, cl_mem buffer, cl_bool blocking_write, size_t offset, size_t cb, const void *ptr, cl_uint num_events_in_wait_list, const cl_event *event_wait_list, cl_event *event)

The OpenCL Specification version 1.1 [Link]

24

Copy data back from kernel

// Copy the output back to CPU memory
int *C; C = new int[N];
Command queue from clCreateCommandQueue Device buffer from clCreateBuffer

clEnqueueReadBuffer(GPUCommandQueue,GPUOutputVector, CL_TRUE, 0, N*sizeof(int), C, 0, NULL, NULL);

Number of Event wait events to list complete before this commands

Read is blocking

Byte Size of data to offset in read in bytes buffer

Pointer to buffer in host to write data

Event

25

Results from GPU

cout << "C[ << 0 << "]: " << A[0] <<"+"<< B[0] <<"=" << C[0]
<< "\n";

cout << "C[ << N-1 << "]: << A[N-1] << "+ << B[N-1] << "=" << C[N-1] << "\n";

C++ here
26

Clean-up
// Cleanup free(GPUDevices); clReleaseKernel(OpenCLVectorAdd); clReleaseProgram(OpenCLProgram); clReleaseCommandQueue(GPUCommandQueue); clReleaseContext(GPUContext); clReleaseMemObject(GPUVector1); clReleaseMemObject(GPUVector2); clReleaseMemObject(GPUOutputVector);

27

Compiling
Need OpenCL header: #include <CL/cl.h> (For mac: #include <OpenCL/opencl.h> ) and link to the OpenCL library. Compile OpenCL host program main.c using gcc, two phases:

gcc -c -I /path-to-include-dir-with-cl.h/ main.c -o main.o gcc -L /path-to-lib-folder-with-OpenCL-libfile/ -l OpenCL main.o -o host

Ref: [Link] 28

Make File
(Program called scalarmulocl)
CC = g++ LD = g++ -lm CFLAGS = -Wall -shared CDEBUG = LIBOCL = -L/nfs-home/mmishra2/NVIDIA_GPU_Computing_SDK/OpenCL/common/lib INCOCL = -I/nfs-home/mmishra2/NVIDIA_GPU_Computing_SDK/OpenCL/common/inc SRCS = [Link] OBJS = scalarmulocl.o EXE = scalarmulocl.a all: $(EXE) $(OBJS): $(SRCS) $(CC) $(CFLAGS) $(INCOCL) -I/usr/include -c $(SRCS) $(EXE): $(OBJS) $(LD) -L/usr/local/lib $(OBJS) $(LIBOCL) -o $(EXE) -l OpenCL clea: rm -f $(OBJS) *~ clear References: [Link] Submitted by: Manisha Mishra 29

Compiling and Executing the program

To compile: make
To Run: ./scalarmulocl.a Snapshot:

Submitted by: Manisha Mishra

30

Questions

More Information

Chapter 11 of Programming Massively Parallel Processors by D. B. Kirk and W-M W. Hwu, Morgan Kaufmann, 2010

32

clGetPlatformIDs
Obtain the list of platforms available.
cl_int clGetPlatformIDs(cl_uint num_entries, cl_platform_id *platforms, cl_uint *num_platforms)
Parameters

num_entries The number of cl_platform_id entries that can be added to platforms. If platforms is not NULL, the num_entries must be greater than zero.
platforms Returns a list of OpenCL platforms found. The cl_platform_id values returned in platforms can be used to identify a specific OpenCL platform. If platforms argument is NULL, this argument is ignored. The number of OpenCL platforms returned is the mininum of the value specified by num_entries or the number of OpenCL platforms available. num_platforms Returns the number of OpenCL platforms available. If num_platforms is NULL, this argument is ignored. [Link] 33

Includes

#include <stdio.h> #include <stdlib.h> #include <CL/cl.h> #include <iostream> using namespace std; //OpenCL header for C //C++ input/output

34