hughperkins
commented
9 years ago Note that it is sufficient to run the original sger example, but just check clFinish afterwards. Both rowmajor and columnmajor cause out-of-bounds access, on these tiny 5x5 matrices :-(
/* ************************************************************************
* Copyright 2013 Advanced Micro Devices, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
* ************************************************************************/
#include <sys/types.h>
#include <stdio.h>
#include <string.h>
#include <math.h>
/* Include CLBLAS header. It automatically includes needed OpenCL header,
* so we can drop out explicit inclusion of cl.h header.
*/
#include <clBLAS.h>
/* This example uses predefined matrices and their characteristics for
* simplicity purpose.
*/
static const clblasOrder order = clblasRowMajor;
static const size_t M = 5;
static const size_t N = 5;
static const cl_float alpha = 10;
static cl_float A[] = {
11, 12, 13, 14, 15,
21, 22, 23, 24, 25,
31, 32, 33, 34, 35,
41, 42, 43, 44, 45,
81, 22, 33, 14, 75
};
static const size_t lda = 5; /* i.e. lda = N */
static const cl_float X[] = {
11,
21,
31,
41,
51,
91,
};
static const int incx = 1;
static const cl_float Y[] = {
45,
23,
39,
45,
50,
10,
};
static const int incy = 1;
static void
printResult(void)
{
size_t i, j;
printf("\nResult:\n");
for (i = 0; i < M; i++) {
for(j = 0; j < N; j++)
printf("\t%f", A[ i*N + j ]);
printf("\n");
}
}
int
main(void)
{
cl_int err;
cl_platform_id platform = 0;
cl_device_id device = 0;
cl_context_properties props[3] = { CL_CONTEXT_PLATFORM, 0, 0 };
cl_context ctx = 0;
cl_command_queue queue = 0;
cl_mem bufA, bufX, bufY;
cl_event event = NULL;
int ret = 0;
/* Setup OpenCL environment. */
err = clGetPlatformIDs(1, &platform, NULL);
if (err != CL_SUCCESS) {
printf( "clGetPlatformIDs() failed with %d\n", err );
return 1;
}
err = clGetDeviceIDs(platform, CL_DEVICE_TYPE_GPU, 1, &device, NULL);
if (err != CL_SUCCESS) {
printf( "clGetDeviceIDs() failed with %d\n", err );
return 1;
}
props[1] = (cl_context_properties)platform;
ctx = clCreateContext(props, 1, &device, NULL, NULL, &err);
if (err != CL_SUCCESS) {
printf( "clCreateContext() failed with %d\n", err );
return 1;
}
queue = clCreateCommandQueue(ctx, device, 0, &err);
if (err != CL_SUCCESS) {
printf( "clCreateCommandQueue() failed with %d\n", err );
clReleaseContext(ctx);
return 1;
}
/* Setup clblas. */
err = clblasSetup();
if (err != CL_SUCCESS) {
printf("clblasSetup() failed with %d\n", err);
clReleaseCommandQueue(queue);
clReleaseContext(ctx);
return 1;
}
/* Prepare OpenCL memory objects and place matrices inside them. */
bufA = clCreateBuffer(ctx, CL_MEM_READ_WRITE, M * lda * sizeof(cl_float),
NULL, &err);
bufX = clCreateBuffer(ctx, CL_MEM_READ_ONLY, ( 1 + ( M - 1 )*abs( incx ) ) * sizeof(cl_float),
NULL, &err);
bufY = clCreateBuffer(ctx, CL_MEM_READ_ONLY, ( 1 + ( N - 1 )*abs( incy ) ) * sizeof(cl_float),
NULL, &err);
err = clEnqueueWriteBuffer(queue, bufA, CL_TRUE, 0,
M * lda * sizeof(cl_float), A, 0, NULL, NULL);
err = clEnqueueWriteBuffer(queue, bufX, CL_TRUE, 0,
( 1 + ( M - 1 )*abs( incx ) ) * sizeof(cl_float), X, 0, NULL, NULL);
err = clEnqueueWriteBuffer(queue, bufY, CL_TRUE, 0,
( 1 + ( N - 1 )*abs( incy ) ) * sizeof(cl_float), Y, 0, NULL, NULL);
/* Call clblas function. */
err = clFinish(queue);
printf("err %i\n", err);
err = clblasSger(order, M, N, alpha, bufX, 0, incx, bufY, 0, incy,
bufA, 0, lda, 1, &queue, 0, NULL, &event);
if (err != CL_SUCCESS) {
printf("clblasSger() failed with %d\n", err);
ret = 1;
}
else {
/* Wait for calculations to be finished. */
err = clWaitForEvents(1, &event);
/* Fetch results of calculations from GPU memory. */
err = clEnqueueReadBuffer(queue, bufA, CL_TRUE, 0, (M * lda * sizeof(cl_float)),
A, 0, NULL, NULL);
/* At this point you will get the result of SGER placed in A array. */
printResult();
}
err = clFinish(queue);
printf("err %i\n", err);
/* Release OpenCL memory objects. */
clReleaseMemObject(bufY);
clReleaseMemObject(bufX);
clReleaseMemObject(bufA);
/* Finalize work with clblas. */
clblasTeardown();
/* Release OpenCL working objects. */
clReleaseCommandQueue(queue);
clReleaseContext(ctx);
return ret;
}
pavanky
TimmyLiu
following code falis, with error -36 'invalid queue'. This usually menas an out of bounds array access occurred.