[MS][LITE] fullconnection matmul A B matrix const node bug

This commit is contained in:
ling 2020-08-17 18:55:16 +08:00
parent b32c5c551e
commit 7aab3f07b4
5 changed files with 96 additions and 13 deletions

View File

@ -23,6 +23,11 @@ using mindspore::lite::RET_OK;
namespace mindspore::kernel { namespace mindspore::kernel {
FullconnectionCPUKernel::~FullconnectionCPUKernel() { FullconnectionCPUKernel::~FullconnectionCPUKernel() {
FreeBuf();
return;
}
void FullconnectionCPUKernel::FreeBuf() {
if (a_c8_ptr_ != nullptr) { if (a_c8_ptr_ != nullptr) {
free(a_c8_ptr_); free(a_c8_ptr_);
a_c8_ptr_ = nullptr; a_c8_ptr_ = nullptr;
@ -41,7 +46,11 @@ FullconnectionCPUKernel::~FullconnectionCPUKernel() {
} }
} }
int FullconnectionCPUKernel::ReSize() { return RET_OK; } int FullconnectionCPUKernel::ReSize() {
FreeBuf();
Init();
return RET_OK;
}
int FullconnectionCPUKernel::Init() { int FullconnectionCPUKernel::Init() {
if (context_->infer_shape_interrupt_ && !context_->running_) { if (context_->infer_shape_interrupt_ && !context_->running_) {
@ -75,16 +84,44 @@ int FullconnectionCPUKernel::Init() {
return RET_MEMORY_FAILED; return RET_MEMORY_FAILED;
} }
memset(b_r8_ptr_, 0, fc_param_->col_8_ * fc_param_->deep_ * sizeof(float)); memset(b_r8_ptr_, 0, fc_param_->col_8_ * fc_param_->deep_ * sizeof(float));
RowMajor2Col8Major(reinterpret_cast<float *>(in_tensors_[1]->Data()), b_r8_ptr_, fc_param_->col_, fc_param_->deep_);
c_r8x8_ptr_ = reinterpret_cast<float *>(malloc(fc_param_->row_8_ * fc_param_->col_8_ * sizeof(float))); c_r8x8_ptr_ = reinterpret_cast<float *>(malloc(fc_param_->row_8_ * fc_param_->col_8_ * sizeof(float)));
if (c_r8x8_ptr_ == nullptr) { if (c_r8x8_ptr_ == nullptr) {
return RET_MEMORY_FAILED; return RET_MEMORY_FAILED;
} }
memset(c_r8x8_ptr_, 0, fc_param_->row_8_ * fc_param_->col_8_ * sizeof(float)); memset(c_r8x8_ptr_, 0, fc_param_->row_8_ * fc_param_->col_8_ * sizeof(float));
fc_param_->a_const_ = false;
fc_param_->b_const_ = false;
InitMatrixA(reinterpret_cast<float *>(in_tensors_[0]->Data()), a_c8_ptr_);
InitMatrixB(reinterpret_cast<float *>(in_tensors_[1]->Data()), b_r8_ptr_);
return RET_OK; return RET_OK;
} }
void FullconnectionCPUKernel::InitMatrixA(float *src_ptr, float *dst_ptr) {
if (fc_param_->a_const_ == true) {
return;
}
if (src_ptr == nullptr) {
return;
}
fc_param_->a_const_ = true;
RowMajor2Col8Major(src_ptr, a_c8_ptr_, fc_param_->row_, fc_param_->deep_);
return;
}
void FullconnectionCPUKernel::InitMatrixB(float *src_ptr, float *dst_ptr) {
if (fc_param_->b_const_ == true) {
return;
}
if (src_ptr == nullptr) {
return;
}
fc_param_->b_const_ = true;
RowMajor2Col8Major(src_ptr, dst_ptr, fc_param_->col_, fc_param_->deep_);
return;
}
int FcFp32MatmulRun(int task_id, LiteParallelGroupEnv *penv, void *cdata) { int FcFp32MatmulRun(int task_id, LiteParallelGroupEnv *penv, void *cdata) {
auto fc = reinterpret_cast<FullconnectionCPUKernel *>(cdata); auto fc = reinterpret_cast<FullconnectionCPUKernel *>(cdata);
auto error_code = fc->DoMatmul(task_id); auto error_code = fc->DoMatmul(task_id);
@ -115,9 +152,11 @@ int FullconnectionCPUKernel::Run() {
return prepare_ret; return prepare_ret;
} }
auto a_ptr = reinterpret_cast<float *>(in_tensors_.at(0)->Data()); auto a_ptr = reinterpret_cast<float *>(in_tensors_.at(0)->Data());
auto b_ptr = reinterpret_cast<float *>(in_tensors_.at(1)->Data());
auto output_ptr = reinterpret_cast<float *>(out_tensors_.at(0)->Data()); auto output_ptr = reinterpret_cast<float *>(out_tensors_.at(0)->Data());
RowMajor2Col8Major(a_ptr, a_c8_ptr_, fc_param_->row_, fc_param_->deep_); InitMatrixA(a_ptr, a_c8_ptr_);
InitMatrixB(b_ptr, b_r8_ptr_);
LiteBackendParallelLaunch(FcFp32MatmulRun, this, thread_count_); LiteBackendParallelLaunch(FcFp32MatmulRun, this, thread_count_);

View File

@ -40,6 +40,11 @@ class FullconnectionCPUKernel : public FullconnectionBaseCPUKernel {
public: public:
int DoMatmul(int task_id); int DoMatmul(int task_id);
void FreeBuf();
private:
void InitMatrixA(float *src_ptr, float *dst_ptr);
void InitMatrixB(float *src_ptr, float *dst_ptr);
private: private:
float *a_c8_ptr_; float *a_c8_ptr_;

View File

@ -78,6 +78,11 @@ int MatmulCPUKernel::Init() {
} }
memset(c_r8x8_ptr_, 0, params_->row_8_ * params_->col_8_ * sizeof(float)); memset(c_r8x8_ptr_, 0, params_->row_8_ * params_->col_8_ * sizeof(float));
params_->a_const_ = false;
params_->b_const_ = false;
InitMatrixA(reinterpret_cast<float *>(in_tensors_[0]->Data()), a_c8_ptr_);
InitMatrixB(reinterpret_cast<float *>(in_tensors_[1]->Data()), b_r8_ptr_);
if (in_tensors_.size() == 3) { if (in_tensors_.size() == 3) {
bias_ptr_ = reinterpret_cast<float *>(malloc(params_->col_8_ * sizeof(float))); bias_ptr_ = reinterpret_cast<float *>(malloc(params_->col_8_ * sizeof(float)));
memset(bias_ptr_, 0, params_->col_8_ * sizeof(float)); memset(bias_ptr_, 0, params_->col_8_ * sizeof(float));
@ -89,6 +94,40 @@ int MatmulCPUKernel::Init() {
return RET_OK; return RET_OK;
} }
void MatmulCPUKernel::InitMatrixA(float *src_ptr, float *dst_ptr) {
if (params_->a_const_ == true) {
return;
}
if (src_ptr == nullptr) {
return;
}
params_->a_const_ = true;
if (params_->a_transpose_) {
RowMajor2Row8Major(src_ptr, dst_ptr, params_->deep_, params_->row_);
} else {
RowMajor2Col8Major(src_ptr, a_c8_ptr_, params_->row_, params_->deep_);
}
return;
}
void MatmulCPUKernel::InitMatrixB(float *src_ptr, float *dst_ptr) {
if (params_->b_const_ == true) {
return;
}
if (src_ptr == nullptr) {
return;
}
params_->b_const_ = true;
if (params_->b_transpose_) {
RowMajor2Col8Major(src_ptr, dst_ptr, params_->col_, params_->deep_);
} else {
RowMajor2Row8Major(src_ptr, dst_ptr, params_->deep_, params_->col_);
}
return;
}
int MatmulCPUKernel::RunImpl(int task_id) { int MatmulCPUKernel::RunImpl(int task_id) {
int cur_oc = MSMIN(thread_stride_, UP_DIV(params_->col_8_, 8) - task_id * thread_stride_); int cur_oc = MSMIN(thread_stride_, UP_DIV(params_->col_8_, 8) - task_id * thread_stride_);
if (cur_oc <= 0) { if (cur_oc <= 0) {
@ -131,16 +170,10 @@ int MatmulCPUKernel::Run() {
auto cur_a_ptr = a_ptr + i * a_stride; auto cur_a_ptr = a_ptr + i * a_stride;
auto cur_b_ptr = b_ptr + i * b_stride; auto cur_b_ptr = b_ptr + i * b_stride;
auto cur_c_ptr = c_ptr + i * c_stride; auto cur_c_ptr = c_ptr + i * c_stride;
if (params_->a_transpose_) {
RowMajor2Row8Major(cur_a_ptr, a_c8_ptr_, params_->deep_, params_->row_); InitMatrixA(cur_a_ptr, a_c8_ptr_);
} else { InitMatrixB(cur_b_ptr, b_r8_ptr_);
RowMajor2Col8Major(cur_a_ptr, a_c8_ptr_, params_->row_, params_->deep_);
}
if (params_->b_transpose_) {
RowMajor2Col8Major(cur_b_ptr, b_r8_ptr_, params_->col_, params_->deep_);
} else {
RowMajor2Row8Major(cur_b_ptr, b_r8_ptr_, params_->deep_, params_->col_);
}
LiteBackendParallelLaunch(MatmulFloatRun, this, thread_count_); LiteBackendParallelLaunch(MatmulFloatRun, this, thread_count_);
Row8x8Major2RowMajor(c_r8x8_ptr_, cur_c_ptr, params_->row_, params_->col_, params_->col_); Row8x8Major2RowMajor(c_r8x8_ptr_, cur_c_ptr, params_->row_, params_->col_, params_->col_);
} }

View File

@ -35,6 +35,10 @@ class MatmulCPUKernel : public MatmulBaseCPUKernel {
int Run() override; int Run() override;
int RunImpl(int task_id); int RunImpl(int task_id);
private:
void InitMatrixA(float *src_ptr, float *dst_ptr);
void InitMatrixB(float *src_ptr, float *dst_ptr);
private: private:
float *a_c8_ptr_; float *a_c8_ptr_;
float *b_r8_ptr_; float *b_r8_ptr_;

View File

@ -33,6 +33,8 @@ typedef struct MatMulParameter {
int batch; int batch;
bool a_transpose_; /* false : row-major */ bool a_transpose_; /* false : row-major */
bool b_transpose_; /* true : col-major */ bool b_transpose_; /* true : col-major */
bool a_const_;
bool b_const_;
ActType act_type_; ActType act_type_;
} MatMulParameter; } MatMulParameter;