You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
237 lines
6.6 KiB
Plaintext
237 lines
6.6 KiB
Plaintext
/*
|
|
* spGPU - Sparse matrices on GPU library.
|
|
*
|
|
* Copyright (C) 2010 - 2014
|
|
* Davide Barbieri - University of Rome Tor Vergata
|
|
*
|
|
* This program is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU General Public License
|
|
* version 3 as published by the Free Software Foundation.
|
|
*
|
|
* This program is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
* GNU General Public License for more details.
|
|
*/
|
|
|
|
#include "cudadebug.h"
|
|
#include "cudalang.h"
|
|
#include <stdio.h>
|
|
extern "C"
|
|
{
|
|
#include "core.h"
|
|
#include "hell.h"
|
|
int getGPUSharedMemPerBlock();
|
|
}
|
|
|
|
#include "debug.h"
|
|
|
|
#define VALUE_TYPE double
|
|
#define TYPE_SYMBOL D
|
|
#define TEX_FETCH_TYPE int2
|
|
#include "hell_spmv_base.cuh"
|
|
|
|
|
|
|
|
#if defined(NEW_MM)
|
|
|
|
#define MMBSZ 8
|
|
|
|
#undef GEN_SPGPU_HELL_NAME
|
|
#define GEN_SPGPU_HELL_NAME(x) CONCAT(CONCAT(spgpu,x),hellspmm)
|
|
#undef GEN_SPGPU_HELL_NAME_VANILLA
|
|
#define GEN_SPGPU_HELL_NAME_VANILLA(x) CONCAT(CONCAT(spgpu,x),hellspmm_vanilla)
|
|
|
|
|
|
__global__ void
|
|
CONCAT(GEN_SPGPU_HELL_NAME(TYPE_SYMBOL), _krn)
|
|
(int count, VALUE_TYPE *z, int zPitch, const VALUE_TYPE *y, int yPitch,
|
|
VALUE_TYPE alpha, const VALUE_TYPE* cM, const int* rP,
|
|
int hackSize, const int* hackOffsets, const int* rS, int rows,
|
|
const VALUE_TYPE *x, int xPitch,
|
|
VALUE_TYPE beta, int baseIndex)
|
|
{
|
|
VALUE_TYPE *pz,*px,*py;
|
|
VALUE_TYPE zProd = CONCAT(zero_,VALUE_TYPE)();
|
|
VALUE_TYPE yVal;
|
|
__shared__ VALUE_TYPE temp[MMBSZ][THREAD_BLOCK];
|
|
|
|
int i = threadIdx.x + blockIdx.x * (THREAD_BLOCK);
|
|
|
|
if (i < rows) {
|
|
int j;
|
|
int hackId = i / hackSize;
|
|
int hackLaneId = i % hackSize;
|
|
|
|
int hackOffset;
|
|
unsigned int laneId = threadIdx.x % 32;
|
|
if (laneId == 0)
|
|
hackOffset = hackOffsets[hackId];
|
|
//__syncthreads();
|
|
hackOffset = __shfl_sync(0xFFFFFFFF,hackOffset, 0) + hackLaneId;
|
|
|
|
rP += hackOffset;
|
|
cM += hackOffset;
|
|
|
|
int rowSize = rS[i];
|
|
for (int k=0; k<count; k++) {
|
|
temp[k][threadIdx.x] = CONCAT(zero_,VALUE_TYPE)();
|
|
}
|
|
|
|
for (int j = 0; j < rowSize; j++) {
|
|
int pointer;
|
|
VALUE_TYPE value;
|
|
VALUE_TYPE fetch;
|
|
|
|
pointer = rP[0] - baseIndex;
|
|
rP += hackSize;
|
|
|
|
value = cM[0];
|
|
cM += hackSize;
|
|
|
|
px = (VALUE_TYPE *) x;
|
|
for (int k=0; k<count; k++) {
|
|
fetch = px[pointer];
|
|
temp[k][threadIdx.x] =
|
|
CONCAT(VALUE_TYPE, _fma)(value, fetch, temp[k][threadIdx.x]);
|
|
px = px + xPitch;
|
|
}
|
|
}
|
|
// Since z and y are accessed with the same offset by the same thread,
|
|
// and the write to z follows the y read, y and z can share the same base address (in-place computing).
|
|
py = (VALUE_TYPE *) y;
|
|
pz = z;
|
|
if (CONCAT(VALUE_TYPE, _isNotZero(beta)))
|
|
for (int k=0; k<count; k++) {
|
|
yVal = py[i];
|
|
pz[i] = CONCAT(VALUE_TYPE, _fma)(beta, yVal, CONCAT(VALUE_TYPE, _mul) (alpha, temp[k][threadIdx.x]));
|
|
py += yPitch;
|
|
pz += zPitch;
|
|
}
|
|
else
|
|
for (int k=0; k<count; k++) {
|
|
pz[i] = CONCAT(VALUE_TYPE, _mul) (alpha, temp[k][threadIdx.x]);
|
|
pz += zPitch;
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
void
|
|
CONCAT(_,GEN_SPGPU_HELL_NAME_VANILLA(TYPE_SYMBOL))
|
|
(spgpuHandle_t handle, int count, VALUE_TYPE* z, int zPitch, const VALUE_TYPE *y, int yPitch,
|
|
VALUE_TYPE alpha, const VALUE_TYPE* cM, const int* rP, int hackSize, const int* hackOffsets,
|
|
const int* rS, const __device int* rIdx, int rows,
|
|
const VALUE_TYPE *x, int xPitch, VALUE_TYPE beta, int baseIndex)
|
|
{
|
|
dim3 block (THREAD_BLOCK, 1);
|
|
dim3 grid ((rows + THREAD_BLOCK - 1) / THREAD_BLOCK);
|
|
// Should we generalize the code to 1/2/4/8 threads per row?
|
|
// And maybe adjust THREAD_BLOCK size?
|
|
int shrMemSize,maxShmemSz;
|
|
maxShmemSz=getGPUSharedMemPerBlock();
|
|
shrMemSize=MMBSZ*THREAD_BLOCK*sizeof(VALUE_TYPE);
|
|
CONCAT(GEN_SPGPU_HELL_NAME(TYPE_SYMBOL), _krn)
|
|
<<< grid, block, shrMemSize, handle->currentStream >>> (count, z, zPitch,y, yPitch,
|
|
alpha, cM, rP, hackSize, hackOffsets, rS, rows,
|
|
x, xPitch, beta, baseIndex);
|
|
}
|
|
|
|
void
|
|
GEN_SPGPU_HELL_NAME(TYPE_SYMBOL)
|
|
(spgpuHandle_t handle,
|
|
int count,
|
|
VALUE_TYPE* z,
|
|
int zPitch,
|
|
const VALUE_TYPE *y,
|
|
int yPitch,
|
|
VALUE_TYPE alpha,
|
|
const VALUE_TYPE* cM,
|
|
const int* rP,
|
|
int hackSize,
|
|
const __device int* hackOffsets,
|
|
const __device int* rS,
|
|
const __device int* rIdx,
|
|
int rows,
|
|
const VALUE_TYPE *x,
|
|
int xPitch,
|
|
VALUE_TYPE beta,
|
|
int baseIndex)
|
|
{
|
|
VALUE_TYPE *px,*py, *pz;
|
|
int cnt;
|
|
int maxNForACall = max(handle->maxGridSizeX, THREAD_BLOCK*handle->maxGridSizeX);
|
|
|
|
// maxNForACall should be a multiple of hackSize
|
|
maxNForACall = (maxNForACall/hackSize)*hackSize;
|
|
int maxShmemSz;
|
|
maxShmemSz=getGPUSharedMemPerBlock();
|
|
//fprintf(stderr,"MaxSHmemSz %d \n",maxShmemSz);
|
|
while (rows > maxNForACall) {//managing large vectors
|
|
cnt = count;
|
|
px = (VALUE_TYPE *) x;
|
|
py = (VALUE_TYPE *) y;
|
|
pz = (VALUE_TYPE *) z;
|
|
while (cnt > MMBSZ) {
|
|
//fprintf(stderr,"counts %d %d %d : pointers: %p %p %p\n",rows,cnt,MMBSZ,px,py,pz);
|
|
CONCAT(_,GEN_SPGPU_HELL_NAME_VANILLA(TYPE_SYMBOL)) (handle, MMBSZ, pz, zPitch,
|
|
py, yPitch,
|
|
alpha, cM, rP,
|
|
hackSize, hackOffsets,
|
|
rS, rIdx,
|
|
maxNForACall,
|
|
px, xPitch, beta, baseIndex);
|
|
px += xPitch*MMBSZ;
|
|
py += yPitch*MMBSZ;
|
|
pz += zPitch*MMBSZ;
|
|
cnt -= MMBSZ;
|
|
}
|
|
if (cnt >0) {
|
|
CONCAT(_,GEN_SPGPU_HELL_NAME_VANILLA(TYPE_SYMBOL)) (handle, cnt, pz, zPitch,
|
|
py, yPitch,
|
|
alpha, cM, rP,
|
|
hackSize, hackOffsets,
|
|
rS, rIdx,
|
|
maxNForACall,
|
|
px, xPitch, beta, baseIndex);
|
|
}
|
|
|
|
y = y + maxNForACall;
|
|
z = z + maxNForACall;
|
|
hackOffsets = hackOffsets + maxNForACall/hackSize;
|
|
rS = rS + maxNForACall;
|
|
|
|
rows -= maxNForACall;
|
|
}
|
|
cnt = count;
|
|
px = (VALUE_TYPE *) x;
|
|
py = (VALUE_TYPE *) y;
|
|
pz = (VALUE_TYPE *) z;
|
|
while (cnt > MMBSZ) {
|
|
//fprintf(stderr,"counts %d %d %d : pointers: %p %p %p\n",rows,cnt,MMBSZ,px,py,pz);
|
|
CONCAT(_,GEN_SPGPU_HELL_NAME_VANILLA(TYPE_SYMBOL)) (handle, MMBSZ, pz, zPitch, py, yPitch,
|
|
alpha, cM, rP, hackSize, hackOffsets,
|
|
rS, rIdx, rows,
|
|
px, xPitch, beta, baseIndex);
|
|
px += xPitch*MMBSZ;
|
|
py += yPitch*MMBSZ;
|
|
pz += zPitch*MMBSZ;
|
|
cnt -= MMBSZ;
|
|
}
|
|
if (cnt >0) {
|
|
CONCAT(_,GEN_SPGPU_HELL_NAME_VANILLA(TYPE_SYMBOL)) (handle, cnt, pz, zPitch,
|
|
py, yPitch,
|
|
alpha, cM, rP,
|
|
hackSize, hackOffsets,
|
|
rS, rIdx,
|
|
rows,
|
|
px, xPitch, beta, baseIndex);
|
|
}
|
|
|
|
|
|
cudaCheckError("CUDA error on hell_spmm");
|
|
}
|
|
|
|
#endif
|
|
|