From 8436251a05ec0cfa906089ee022db17498e6f65f Mon Sep 17 00:00:00 2001
From: AlexeyAB <alexeyab84@gmail.com>
Date: Tue, 21 Aug 2018 17:44:18 +0000
Subject: [PATCH] maxpool_layer on CPU uses AVX2 and OpenMP
---
src/gemm.h | 5 +
src/gemm.c | 142 ++++++++++++++++++++++++++++++++++++++++++++++-
src/maxpool_layer.c | 6 ++
3 files changed, 149 insertions(+), 4 deletions(-)
diff --git a/src/gemm.c b/src/gemm.c
index 6a34da5..7abfe25 100644
--- a/src/gemm.c
+++ b/src/gemm.c
@@ -5,6 +5,7 @@
#include <stdlib.h>
#include <stdio.h>
#include <math.h>
+#include <float.h>
#if defined(_OPENMP)
#include <omp.h>
@@ -594,7 +595,7 @@
static int max_num_threads = 0;
if (max_num_threads == 0) {
max_num_threads = omp_get_max_threads();
- omp_set_num_threads(4);// max_num_threads / 2);
+ omp_set_num_threads( max_num_threads / 2);
}
#endif
@@ -1167,6 +1168,100 @@
}
+void forward_maxpool_layer_avx(float *src, float *dst, int *indexes, int size, int w, int h, int out_w, int out_h, int c,
+ int pad, int stride, int batch)
+{
+
+ int w_offset = -pad / 2;
+ int h_offset = -pad / 2;
+ int b, k;
+
+ for (b = 0; b < batch; ++b) {
+ #pragma omp parallel for
+ for (k = 0; k < c; ++k) {
+ int i, j, m, n;
+ for (i = 0; i < out_h; ++i) {
+ //for (j = 0; j < out_w; ++j) {
+ j = 0;
+
+ if(stride == 1 && is_avx() == 1) {
+ for (j = 0; j < out_w - 8 - (size - 1); j += 8) {
+ int out_index = j + out_w*(i + out_h*(k + c*b));
+ __m256 max256 = _mm256_set1_ps(-FLT_MAX);
+ for (n = 0; n < size; ++n) {
+ for (m = 0; m < size; ++m) {
+ int cur_h = h_offset + i*stride + n;
+ int cur_w = w_offset + j*stride + m;
+ int index = cur_w + w*(cur_h + h*(k + b*c));
+ int valid = (cur_h >= 0 && cur_h < h &&
+ cur_w >= 0 && cur_w < w);
+ if (!valid) continue;
+
+ __m256 src256 = _mm256_loadu_ps(&src[index]);
+ max256 = _mm256_max_ps(src256, max256);
+ }
+ }
+ _mm256_storeu_ps(&dst[out_index], max256);
+
+ }
+ }
+ else if (size == 2 && stride == 2 && is_avx() == 1) {
+ for (j = 0; j < out_w - 4; j += 4) {
+ int out_index = j + out_w*(i + out_h*(k + c*b));
+ float max = -FLT_MAX;
+ int max_i = -1;
+ __m128 max128 = _mm_set1_ps(-FLT_MAX);
+
+ for (n = 0; n < size; ++n) {
+ //for (m = 0; m < size; ++m)
+ m = 0;
+ {
+ int cur_h = h_offset + i*stride + n;
+ int cur_w = w_offset + j*stride + m;
+ int index = cur_w + w*(cur_h + h*(k + b*c));
+ int valid = (cur_h >= 0 && cur_h < h &&
+ cur_w >= 0 && cur_w < w);
+ if (!valid) continue;
+
+ __m256 src256 = _mm256_loadu_ps(&src[index]);
+ __m256 src256_2 = _mm256_permute_ps(src256, (1 << 0) | (3 << 4));
+ __m256 max256 = _mm256_max_ps(src256, src256_2);
+
+ __m128 src128_0 = _mm256_extractf128_ps(max256, 0);
+ __m128 src128_1 = _mm256_extractf128_ps(max256, 1);
+ __m128 src128 = _mm_shuffle_ps(src128_0, src128_1, (2 << 2) | (2 << 6));
+
+ max128 = _mm_max_ps(src128, max128);
+ }
+ }
+ _mm_storeu_ps(&dst[out_index], max128);
+ }
+ }
+
+ for (; j < out_w; ++j) {
+ int out_index = j + out_w*(i + out_h*(k + c*b));
+ float max = -FLT_MAX;
+ int max_i = -1;
+ for (n = 0; n < size; ++n) {
+ for (m = 0; m < size; ++m) {
+ int cur_h = h_offset + i*stride + n;
+ int cur_w = w_offset + j*stride + m;
+ int index = cur_w + w*(cur_h + h*(k + b*c));
+ int valid = (cur_h >= 0 && cur_h < h &&
+ cur_w >= 0 && cur_w < w);
+ float val = (valid != 0) ? src[index] : -FLT_MAX;
+ max_i = (val > max) ? index : max_i;
+ max = (val > max) ? val : max;
+ }
+ }
+ dst[out_index] = max;
+ indexes[out_index] = max_i;
+ }
+ }
+ }
+ }
+}
+
#else
void gemm_nn(int M, int N, int K, float ALPHA,
@@ -1283,6 +1378,8 @@
int channels, int height, int width,
int ksize, int stride, int pad, float* data_col)
{
+ im2col_cpu(data_im, channels, height, width, ksize, stride, pad, data_col);
+ return;
int c, h, w;
int height_col = (height + 2 * pad - ksize) / stride + 1;
@@ -1304,7 +1401,7 @@
int col_index = (c * height_col + h) * width_col + w;
data_col[col_index] = data_im[im_col + width*(im_row + height*c_im)];
- }
+ }
for (; w < width_col - pad; ++w) {
int im_row = h_offset + h - pad;
@@ -1313,7 +1410,7 @@
data_col[col_index] = data_im[im_col + width*(im_row + height*c_im)];
}
-}
+ }
{
w = 0;
@@ -1445,7 +1542,44 @@
}
}
}
-#endif // __x86_64
+
+void forward_maxpool_layer_avx(float *src, float *dst, int *indexes, int size, int w, int h, int out_w, int out_h, int c,
+ int pad, int stride, int batch)
+{
+ int b, k;
+ int w_offset = -pad / 2;
+ int h_offset = -pad / 2;
+
+ for (b = 0; b < batch; ++b) {
+ #pragma omp parallel for
+ for (k = 0; k < c; ++k) {
+ int i, j, m, n;
+ for (i = 0; i < out_h; ++i) {
+ for (j = 0; j < out_w; ++j) {
+ int out_index = j + out_w*(i + out_h*(k + c*b));
+ float max = -FLT_MAX;
+ int max_i = -1;
+ for (n = 0; n < size; ++n) {
+ for (m = 0; m < size; ++m) {
+ int cur_h = h_offset + i*stride + n;
+ int cur_w = w_offset + j*stride + m;
+ int index = cur_w + w*(cur_h + h*(k + b*c));
+ int valid = (cur_h >= 0 && cur_h < h &&
+ cur_w >= 0 && cur_w < w);
+ float val = (valid != 0) ? src[index] : -FLT_MAX;
+ max_i = (val > max) ? index : max_i;
+ max = (val > max) ? val : max;
+ }
+ }
+ dst[out_index] = max;
+ indexes[out_index] = max_i;
+ }
+ }
+ }
+ }
+}
+
+#endif // AVX
void gemm_nt(int M, int N, int K, float ALPHA,
float *A, int lda,
diff --git a/src/gemm.h b/src/gemm.h
index 1472cca..3cdca99 100644
--- a/src/gemm.h
+++ b/src/gemm.h
@@ -46,6 +46,11 @@
float *B, int ldb,
float *C, int ldc);
+
+void forward_maxpool_layer_avx(float *src, float *dst, int *indexes, int size, int w, int h, int out_w, int out_h, int c,
+ int pad, int stride, int batch);
+
+
void gemm(int TA, int TB, int M, int N, int K, float ALPHA,
float *A, int lda,
float *B, int ldb,
diff --git a/src/maxpool_layer.c b/src/maxpool_layer.c
index c543b52..5ff7e9d 100644
--- a/src/maxpool_layer.c
+++ b/src/maxpool_layer.c
@@ -1,5 +1,6 @@
#include "maxpool_layer.h"
#include "cuda.h"
+#include "gemm.h"
#include <stdio.h>
image get_maxpool_image(maxpool_layer l)
@@ -79,6 +80,11 @@
void forward_maxpool_layer(const maxpool_layer l, network_state state)
{
+ if (!state.train) {
+ forward_maxpool_layer_avx(state.input, l.output, l.indexes, l.size, l.w, l.h, l.out_w, l.out_h, l.c, l.pad, l.stride, l.batch);
+ return;
+ }
+
int b,i,j,k,m,n;
int w_offset = -l.pad / 2;
int h_offset = -l.pad / 2;
--
Gitblit v1.10.0