From e36182cd8c5dd5c6d0aa1f77cf5cdca87e8bb1f0 Mon Sep 17 00:00:00 2001
From: Joseph Redmon <pjreddie@gmail.com>
Date: Fri, 21 Nov 2014 23:35:19 +0000
Subject: [PATCH] cleaned up data parsing a lot. probably nothing broken?
---
src/axpy.c | 27 ++++++++++++++++++++-------
1 files changed, 20 insertions(+), 7 deletions(-)
diff --git a/src/axpy.c b/src/axpy.c
index c4ec1eb..21293b3 100644
--- a/src/axpy.c
+++ b/src/axpy.c
@@ -1,24 +1,24 @@
#include "mini_blas.h"
-inline void axpy_cpu(int N, float ALPHA, float *X, int INCX, float *Y, int INCY)
+void axpy_cpu(int N, float ALPHA, float *X, int INCX, float *Y, int INCY)
{
int i;
for(i = 0; i < N; ++i) Y[i*INCY] += ALPHA*X[i*INCX];
}
-inline void scal_cpu(int N, float ALPHA, float *X, int INCX)
+void scal_cpu(int N, float ALPHA, float *X, int INCX)
{
int i;
for(i = 0; i < N; ++i) X[i*INCX] *= ALPHA;
}
-inline void copy_cpu(int N, float *X, int INCX, float *Y, int INCY)
+void copy_cpu(int N, float *X, int INCX, float *Y, int INCY)
{
int i;
for(i = 0; i < N; ++i) Y[i*INCY] = X[i*INCX];
}
-inline float dot_cpu(int N, float *X, int INCX, float *Y, int INCY)
+float dot_cpu(int N, float *X, int INCX, float *Y, int INCY)
{
int i;
float dot = 0;
@@ -65,6 +65,11 @@
void axpy_ongpu(int N, float ALPHA, cl_mem X, int INCX, cl_mem Y, int INCY)
{
+ axpy_ongpu_offset(N,ALPHA,X,0,INCX,Y,0,INCY);
+}
+
+void axpy_ongpu_offset(int N, float ALPHA, cl_mem X, int OFFX, int INCX, cl_mem Y, int OFFY, int INCY)
+{
cl_setup();
cl_kernel kernel = get_axpy_kernel();
cl_command_queue queue = cl.queue;
@@ -73,19 +78,25 @@
cl.error = clSetKernelArg(kernel, i++, sizeof(N), (void*) &N);
cl.error = clSetKernelArg(kernel, i++, sizeof(ALPHA), (void*) &ALPHA);
cl.error = clSetKernelArg(kernel, i++, sizeof(X), (void*) &X);
+ cl.error = clSetKernelArg(kernel, i++, sizeof(OFFX), (void*) &OFFX);
cl.error = clSetKernelArg(kernel, i++, sizeof(INCX), (void*) &INCX);
cl.error = clSetKernelArg(kernel, i++, sizeof(Y), (void*) &Y);
+ cl.error = clSetKernelArg(kernel, i++, sizeof(OFFY), (void*) &OFFY);
cl.error = clSetKernelArg(kernel, i++, sizeof(INCY), (void*) &INCY);
check_error(cl);
const size_t global_size[] = {N};
- clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
+ cl.error = clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
check_error(cl);
}
void copy_ongpu(int N, cl_mem X, int INCX, cl_mem Y, int INCY)
{
+ copy_ongpu_offset(N,X,0,INCX,Y,0,INCY);
+}
+void copy_ongpu_offset(int N, cl_mem X, int OFFX, int INCX, cl_mem Y, int OFFY, int INCY)
+{
cl_setup();
cl_kernel kernel = get_copy_kernel();
cl_command_queue queue = cl.queue;
@@ -93,14 +104,16 @@
cl_uint i = 0;
cl.error = clSetKernelArg(kernel, i++, sizeof(N), (void*) &N);
cl.error = clSetKernelArg(kernel, i++, sizeof(X), (void*) &X);
+ cl.error = clSetKernelArg(kernel, i++, sizeof(OFFX), (void*) &OFFX);
cl.error = clSetKernelArg(kernel, i++, sizeof(INCX), (void*) &INCX);
cl.error = clSetKernelArg(kernel, i++, sizeof(Y), (void*) &Y);
+ cl.error = clSetKernelArg(kernel, i++, sizeof(OFFY), (void*) &OFFY);
cl.error = clSetKernelArg(kernel, i++, sizeof(INCY), (void*) &INCY);
check_error(cl);
const size_t global_size[] = {N};
- clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
+ cl.error = clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
check_error(cl);
}
void scal_ongpu(int N, float ALPHA, cl_mem X, int INCX)
@@ -118,7 +131,7 @@
const size_t global_size[] = {N};
- clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
+ cl.error = clEnqueueNDRangeKernel(queue, kernel, 1, 0, global_size, 0, 0, 0, 0);
check_error(cl);
}
#endif
--
Gitblit v1.10.0