Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Cherry-Pick][OpenCL] Fix OpenCL opt #9630

Merged
merged 3 commits into from
Nov 4, 2022
Merged

[Cherry-Pick][OpenCL] Fix OpenCL opt #9630

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2742f22
fix opencl opt test=develop
sprouteer Nov 2, 2022
92a75f2
fix io_copy bug test=develop
sprouteer Nov 2, 2022
32c19c1
fix opencl mem object config test=develop
sprouteer Nov 3, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions lite/api/tools/opt_base.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -116,6 +116,8 @@ void OptBase::SetValidPlaces(const std::string& valid_places) {
Place{TARGET(kOpenCL), PRECISION(kFP16), DATALAYOUT(kImageDefault)});
valid_places_.emplace_back(
Place{TARGET(kOpenCL), PRECISION(kFP16), DATALAYOUT(kImageFolder)});
valid_places_.emplace_back(
Place{TARGET(kOpenCL), PRECISION(kFP16), DATALAYOUT(kNCHW)});
valid_places_.emplace_back(
Place{TARGET(kOpenCL), PRECISION(kFloat), DATALAYOUT(kNCHW)});
valid_places_.emplace_back(
Expand Down
19 changes: 18 additions & 1 deletion lite/core/optimizer/mir/opencl_memory_object_config_pass.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -135,7 +135,20 @@ void OpenCLMemoryObjectConfigPass::CorrectArgumentPlace(SSAGraph* graph) {
};

// 1. op unsupport persistable
if (op_type == "reshape2" || op_type == "unsqueeze2") {
const std::vector<std::string> op_cases_unsupport_persistable{
"elementwise_floordiv",
"elementwise_add",
"elementwise_sub",
"elementwise_mul",
"elementwise_div",
"elementwise_pow",
"elementwise_mod",
"reshape2",
"unsqueeze2",
"split"};
if (std::find(op_cases_unsupport_persistable.begin(),
op_cases_unsupport_persistable.end(),
op_type) != op_cases_unsupport_persistable.end()) {
for (std::list<Node*>::iterator i = x->inlinks.begin();
i != x->inlinks.end();
++i) {
Expand Down Expand Up @@ -287,6 +300,10 @@ void OpenCLMemoryObjectConfigPass::CorrectArgumentPlace(SSAGraph* graph) {
}
}
}

// 7. reshape change target
if (op_type == "reshape" || op_type == "reshape2")
change_image2d_to_buffer = true;
}

if (change_image2d_to_cpu) {
Expand Down
6 changes: 4 additions & 2 deletions lite/kernels/opencl/io_copy_buffer_compute.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -104,7 +104,8 @@ class IoCopyHostToOpenCLCompute
CHECK(param.x->target() == TARGET(kHost) ||
param.x->target() == TARGET(kARM));

auto mem_size = param.x->memory_size();
auto mem_size = param.x->dims().production() *
PrecisionTypeLength(param.x->precision());
#ifdef LITE_WITH_LOG
VLOG(2) << "param.x->memory_size():" << mem_size;
VLOG(2) << "param.x->dims().size():" << param.x->dims().size();
Expand Down Expand Up @@ -212,7 +213,8 @@ class IoCopykOpenCLToHostCompute
void Run() override {
auto& param = Param<operators::IoCopyParam>();
CHECK(param.x->target() == TARGET(kOpenCL));
auto mem_size = param.x->memory_size();
auto mem_size = param.x->dims().production() *
PrecisionTypeLength(param.x->precision());
const cl::Buffer* x_ptr;
if (param.process_type == 1) {
x_ptr = param.x->data<uint8_t, cl::Buffer>();
Expand Down