Chiebot-Mirror
/
opencv
8
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge pull request #20661 from alalek:dnn_ocl_fix_gemm_like_kernel

Browse Source
pull/20687/head
Alexander Alekhin 4 years ago
parent d31b93b513 35e824c287
commit 6ace801418
2 changed files with 80 additions and 72 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 73
      modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
  2. 79
      modules/dnn/src/opencl/conv_layer_spatial.cl

73
modules/dnn/src/ocl4dnn/src/ocl4dnn_conv_spatial.cpp
Unescape View File

@ -945,9 +945,11 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
} else if (config->kernelType == KERNEL_TYPE_GEMM_LIKE) { } else if (config->kernelType == KERNEL_TYPE_GEMM_LIKE) {
if (!swizzleWeight(weight, config->workItem_output[1], true)) if (!swizzleWeight(weight, config->workItem_output[1], true))
return false; return false;
#if 0
size_t total_bottom_size = bottom_dim_ * numImages; size_t total_bottom_size = bottom_dim_ * numImages;
size_t total_kernel_size = kernel_h_ * kernel_w_ * channels_ * M_; size_t total_kernel_size = kernel_h_ * kernel_w_ * channels_ * M_;
size_t total_bias_size = M_ * group_; size_t total_bias_size = M_ * group_;
#endif
size_t total_top_size = top_dim_ * numImages; size_t total_top_size = top_dim_ * numImages;
for (int32_t g = 0; g < group_; ++g) { for (int32_t g = 0; g < group_; ++g) {
bias_offset = M_ * g; bias_offset = M_ * g;
@ -960,72 +962,25 @@ bool OCL4DNNConvSpatial<float>::convolve(const UMat &bottom, UMat &top,
return false; return false;
cl_uint argIdx = 0; cl_uint argIdx = 0;
setFusionArg(fused_activ_, fused_eltwise_, -1, kernel, argIdx); setFusionArg(fused_activ_, fused_eltwise_, output_image_offset, kernel, argIdx);
UMat img_buffer;
if (image_offset)
{
CreateSubBuffer(bottom, img_buffer, image_offset,
total_bottom_size - image_offset, false);
if (img_buffer.empty())
return false;
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(img_buffer));
}
else
{
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bottom));
}
UMat kernel_buffer; kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bottom));
if (kernel_offset) kernel.set(argIdx++, (int)image_offset);
{ kernel.set(argIdx++, (int)(bottom.total() - image_offset));
CreateSubBuffer(swizzled_weights_umat, kernel_buffer, kernel_offset,
total_kernel_size - kernel_offset, false);
if (kernel_buffer.empty())
return false;
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(kernel_buffer)); kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(swizzled_weights_umat));
} kernel.set(argIdx++, (int)kernel_offset);
else kernel.set(argIdx++, (int)(swizzled_weights_umat.total() - kernel_offset));
{
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(swizzled_weights_umat));
}
UMat bias_buffer;
if (bias_term_) if (bias_term_)
{ {
if (bias_offset) kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bias));
{ kernel.set(argIdx++, (int)bias_offset);
CreateSubBuffer(bias, bias_buffer, bias_offset,
total_bias_size - bias_offset, false);
if (bias_buffer.empty())
return false;
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bias_buffer));
}
else
{
kernel.set(argIdx++, ocl::KernelArg::PtrReadOnly(bias));
}
} }
UMat out_buffer; kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
if (output_image_offset) kernel.set(argIdx++, (int)(top.offset / element_size) + output_image_offset);
{ kernel.set(argIdx++, (int)total_top_size - (int)(top.offset / element_size));
CreateSubBuffer(top, out_buffer, output_image_offset,
total_top_size - output_image_offset, true);
if (out_buffer.empty())
return false;
kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(out_buffer));
kernel.set(argIdx++, (int)(out_buffer.offset / element_size));
}
else
{
kernel.set(argIdx++, ocl::KernelArg::PtrWriteOnly(top));
kernel.set(argIdx++, (int)(top.offset / element_size));
}
kernel.set(argIdx++, (uint16_t)width_); kernel.set(argIdx++, (uint16_t)width_);
kernel.set(argIdx++, (uint16_t)height_); kernel.set(argIdx++, (uint16_t)height_);

79
modules/dnn/src/opencl/conv_layer_spatial.cl
Unescape View File

@ -401,13 +401,12 @@ typedef struct half0 { half s0; } half0; //never used but makes compiler happy.
#define ROW_PITCH input_width #define ROW_PITCH input_width
#define GEMM_LIKE_KERNEL_ARGS \ #define GEMM_LIKE_KERNEL_ARGS \
ELTWISE_DATA_ARG \ ELTWISE_DATA_ARG_WITH_OFFSET \
FUSED_ARG \ FUSED_ARG \
const __global Dtype *src0, \ const __global Dtype *src0_ptr, const unsigned int src0_offset, const unsigned int src0_limit, \
const __global Dtype *src1, \ const __global Dtype *src1_ptr, const unsigned int src1_offset, const unsigned int src1_limit, \
BIAS_KERNEL_ARG \ BIAS_KERNEL_ARG_WITH_OFFSET \
__global Dtype *dst_base, \ __global Dtype *dst_base, const unsigned int dst_offset, const unsigned int dst_limit, \
const int dst_offset, \
const ushort input_width, \ const ushort input_width, \
const ushort input_height, \ const ushort input_height, \
const ushort output_width, \ const ushort output_width, \
@ -437,7 +436,17 @@ typedef struct half0 { half s0; } half0; //never used but makes compiler happy.
__attribute__((intel_reqd_sub_group_size(8))) __attribute__((intel_reqd_sub_group_size(8)))
__kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS) __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
{ {
const __global Dtype *src0 = src0_ptr + src0_offset;
const __global Dtype *src1 = src1_ptr + src1_offset;
#if APPLY_BIAS
__global Dtype* biases_base = biases_base_ptr + biases_base_offset;
#endif
__global Dtype *dst = dst_base + dst_offset; __global Dtype *dst = dst_base + dst_offset;
#ifdef FUSED_CONV_ELTWISE
__global Dtype* eltwise_data = eltwise_ptr + eltwise_offset;
#endif
const int group_x = get_group_id(0); const int group_x = get_group_id(0);
const int group_y = get_group_id(1); const int group_y = get_group_id(1);
const int global_x = get_global_id(0); const int global_x = get_global_id(0);
@ -460,6 +469,14 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
} }
typedef CAT( Dtype, KERNEL_WIDTH ) Dtype_t; typedef CAT( Dtype, KERNEL_WIDTH ) Dtype_t;
// U_GEMM_LIKE_CONV_k11x11_cn3_g1_s4x4_d1x1_b1_in240x240_p0x0_num1_M96_activ1_eltwise0_FP32_5_1_8_32_SIMD8 doesn't run properly (src0_read out of bounds)
// Test: DNNTestNetwork.AlexNet/0 (to run all kernels use OPENCV_OCL4DNN_FORCE_AUTO_TUNING=1)
#if 0 // INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0
#define OPTIMIZE_READ 1
#else
#define OPTIMIZE_READ 0
#endif
// True for all threads if filter_width is multiple of TILE_N // True for all threads if filter_width is multiple of TILE_N
// else, true for all but right-most column of threads. // else, true for all but right-most column of threads.
if( TILE_N_LAST == 0 || global_x < WIDTH1 / TILE_N ) if( TILE_N_LAST == 0 || global_x < WIDTH1 / TILE_N )
@ -476,7 +493,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// atile is M rows x K columns. // atile is M rows x K columns.
int curr_x = ( global_y % output_width ) * STRIDE_X; int curr_x = ( global_y % output_width ) * STRIDE_X;
int curr_y = ( global_y / output_width ) * STRIDE_Y; int curr_y = ( global_y / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0 #if !OPTIMIZE_READ
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
const __global Dtype *src0_read = src0 const __global Dtype *src0_read = src0
@ -496,7 +513,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
do do
{ {
int patch_row = 0; int patch_row = 0;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0 #if !OPTIMIZE_READ
curr_y = saved_y; curr_y = saved_y;
#endif #endif
@ -514,11 +531,17 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// ... // ...
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0 #if OPTIMIZE_READ
#if KERNEL_WIDTH == 3 #if KERNEL_WIDTH == 3
Dtype_t blockA00 = vload3(0, src0_read); Dtype_t blockA00 = vload3(0, src0_read);
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#else #else
#if 0 // debug
if ((int)(src0_read - src0) >= src0_limit - KERNEL_WIDTH)
{
printf("CATCH: src0_read-src0: %d limit=%d curr_y,curr_x=%d,%d\n", (int)(src0_read - src0), src0_limit, curr_y, curr_x);
}
#endif
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ]; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ];
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#endif #endif
@ -639,7 +662,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
// atile is M rows x K columns. // atile is M rows x K columns.
int curr_x = ( global_y % output_width ) * STRIDE_X; int curr_x = ( global_y % output_width ) * STRIDE_X;
int curr_y = ( global_y / output_width ) * STRIDE_Y; int curr_y = ( global_y / output_width ) * STRIDE_Y;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0 #if !OPTIMIZE_READ
int saved_y = curr_y; int saved_y = curr_y;
#endif #endif
const __global Dtype *src0_read = src0 const __global Dtype *src0_read = src0
@ -659,14 +682,14 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
do do
{ {
int patch_row = 0; int patch_row = 0;
#if INPUT_PAD_H != 0 || INPUT_PAD_W != 0 || DILATION_X != 1 || DILATION_Y != 1 || INPUT_PAD_BOTTOM != 0 || INPUT_PAD_RIGHT != 0 #if !OPTIMIZE_READ
curr_y = saved_y; curr_y = saved_y;
#endif #endif
do do
{ {
// Load atile and interleaved btile. // Load atile and interleaved btile.
const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1; const bool kernel_width_is_odd = KERNEL_WIDTH % 2 == 1;
#if INPUT_PAD_W == 0 && INPUT_PAD_H == 0 && DILATION_X == 1 && DILATION_Y == 1 && INPUT_PAD_BOTTOM == 0 && INPUT_PAD_RIGHT == 0 #if OPTIMIZE_READ
Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ]; Dtype_t blockA00 = ( (const __global Dtype_t*)src0_read )[ 0 ];
Dtype* pblockA00 = (Dtype*)(&blockA00); Dtype* pblockA00 = (Dtype*)(&blockA00);
#else #else
@ -803,7 +826,7 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
} }
} }
} }
#endif #endif // TILE_N_LAST > 0
} }
#endif #endif
#ifdef GEMM_LIKE_CONV_32_2 #ifdef GEMM_LIKE_CONV_32_2
@ -826,7 +849,17 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
__attribute__((intel_reqd_sub_group_size(8))) __attribute__((intel_reqd_sub_group_size(8)))
__kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS) __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
{ {
const __global Dtype *src0 = src0_ptr + src0_offset;
const __global Dtype *src1 = src1_ptr + src1_offset;
#if APPLY_BIAS
__global Dtype* biases_base = biases_base_ptr + biases_base_offset;
#endif
__global Dtype *dst = dst_base + dst_offset; __global Dtype *dst = dst_base + dst_offset;
#ifdef FUSED_CONV_ELTWISE
__global Dtype* eltwise_data = eltwise_ptr + eltwise_offset;
#endif
const int group_x = get_group_id(0); const int group_x = get_group_id(0);
const int group_y = get_group_id(1); const int group_y = get_group_id(1);
const int global_x = get_global_id(0); const int global_x = get_global_id(0);
@ -1388,7 +1421,17 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
__attribute__((intel_reqd_sub_group_size(16))) __attribute__((intel_reqd_sub_group_size(16)))
__kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS) __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
{ {
const __global Dtype *src0 = src0_ptr + src0_offset;
const __global Dtype *src1 = src1_ptr + src1_offset;
#if APPLY_BIAS
__global Dtype* biases_base = biases_base_ptr + biases_base_offset;
#endif
__global Dtype *dst = dst_base + dst_offset; __global Dtype *dst = dst_base + dst_offset;
#ifdef FUSED_CONV_ELTWISE
__global Dtype* eltwise_data = eltwise_ptr + eltwise_offset;
#endif
const int group_x = get_group_id(0); const int group_x = get_group_id(0);
const int group_y = get_group_id(1); const int group_y = get_group_id(1);
const int global_x = get_global_id(0); const int global_x = get_global_id(0);
@ -1574,7 +1617,17 @@ __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
__attribute__((intel_reqd_sub_group_size(16))) __attribute__((intel_reqd_sub_group_size(16)))
__kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS) __kernel void Conv_Interleaved(GEMM_LIKE_KERNEL_ARGS)
{ {
const __global Dtype *src0 = src0_ptr + src0_offset;
const __global Dtype *src1 = src1_ptr + src1_offset;
#if APPLY_BIAS
__global Dtype* biases_base = biases_base_ptr + biases_base_offset;
#endif
__global Dtype *dst = dst_base + dst_offset; __global Dtype *dst = dst_base + dst_offset;
#ifdef FUSED_CONV_ELTWISE
__global Dtype* eltwise_data = eltwise_ptr + eltwise_offset;
#endif
const int group_x = get_group_id(0); const int group_x = get_group_id(0);
const int group_y = get_group_id(1); const int group_y = get_group_id(1);
const int global_x = get_global_id(0); const int global_x = get_global_id(0);

Write Preview
Loading…
Cancel
Save