Skip to content

graph: backend: dnnl: executables: add exec profile verbose for gen_index #4849

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

Open
TaoLv wants to merge 1 commit into
base: main
Choose a base branch
from
Open

graph: backend: dnnl: executables: add exec profile verbose for gen_index #4849

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
61 changes: 58 additions & 3 deletions src/graph/backend/dnnl/executables/gen_index.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,9 @@ genindex_executable_t::genindex_executable_t(std::shared_ptr<op_t> &op,
output_dims_[i] = output_lt.dims[i];
output_strides_[i] = output_lt.layout.strides[i];
}
info_ = std::string(dnnl_engine_kind2str(
static_cast<dnnl_engine_kind_t>(p_engine.get_kind())))
+ "," + op->str();
#if DNNL_GPU_RUNTIME != DNNL_RUNTIME_NONE \
&& DNNL_GPU_VENDOR == DNNL_VENDOR_INTEL
if (p_engine.get_kind() == engine::kind::gpu) {
Expand All @@ -59,7 +62,7 @@ genindex_executable_t::genindex_executable_t(std::shared_ptr<op_t> &op,
#endif
}

void genindex_executable_t::execute(const stream &stream,
void genindex_executable_t::execute_impl(const stream &stream,
const std::unordered_map<int, memory> &args) const {
const auto &it_dst = args.find(DNNL_ARG_DST);
if (it_dst == args.end()) return;
Expand All @@ -79,8 +82,24 @@ void genindex_executable_t::execute(const stream &stream,
stream.get()->after_exec_hook();
}

void genindex_executable_t::execute(const stream &stream,
const std::unordered_map<int, memory> &args) const {
if (get_verbose(dnnl::impl::verbose_t::exec_profile,
dnnl::impl::component_t::graph)) {
stream.get()->wait();
double start_ms = dnnl::impl::get_msec();
execute_impl(stream, args);
stream.get()->wait();
double duration_ms = dnnl::impl::get_msec() - start_ms;
VPROF(start_ms, graph, exec, VERBOSE_profile, info_.c_str(),
duration_ms);
} else {
execute_impl(stream, args);
}
}

#ifdef DNNL_WITH_SYCL
::sycl::event genindex_executable_t::execute_sycl(const stream &stream,
::sycl::event genindex_executable_t::execute_sycl_impl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<::sycl::event> &deps) const {
if (stream.get_engine().get_kind() == engine::kind::cpu) {
Expand Down Expand Up @@ -127,10 +146,28 @@ ::sycl::event genindex_executable_t::execute_sycl(const stream &stream,
throw std::runtime_error("Unimplement");
#endif
}

::sycl::event genindex_executable_t::execute_sycl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<::sycl::event> &deps) const {
if (get_verbose(dnnl::impl::verbose_t::exec_profile,
dnnl::impl::component_t::graph)) {
stream.get()->wait();
double start_ms = dnnl::impl::get_msec();
execute_sycl_impl(stream, args, deps);
stream.get()->wait();
double duration_ms = dnnl::impl::get_msec() - start_ms;
VPROF(start_ms, graph, exec, VERBOSE_profile, info_.c_str(),
duration_ms);
return {}; // no event returned in profiling mode
} else {
return execute_sycl_impl(stream, args, deps);
}
}
#endif

#if DNNL_GPU_RUNTIME == DNNL_RUNTIME_OCL
cl_event genindex_executable_t::execute_ocl(const stream &stream,
cl_event genindex_executable_t::execute_ocl_impl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<cl_event> &deps) const {
#if DNNL_GPU_VENDOR == DNNL_VENDOR_INTEL
Expand Down Expand Up @@ -174,6 +211,24 @@ cl_event genindex_executable_t::execute_ocl(const stream &stream,
throw std::runtime_error("Unimplement");
#endif
}

cl_event genindex_executable_t::execute_ocl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<cl_event> &deps) const {
if (get_verbose(dnnl::impl::verbose_t::exec_profile,
dnnl::impl::component_t::graph)) {
stream.get()->wait();
double start_ms = dnnl::impl::get_msec();
execute_ocl_impl(stream, args, deps);
stream.get()->wait();
double duration_ms = dnnl::impl::get_msec() - start_ms;
VPROF(start_ms, graph, exec, VERBOSE_profile, info_.c_str(),
duration_ms);
return nullptr; // no event returned in profiling mode
} else {
return execute_ocl_impl(stream, args, deps);
}
}
#endif

arg_indices_t genindex_executable_t::get_arg_indices(const op_t *op) {
Expand Down
9 changes: 9 additions & 0 deletions src/graph/backend/dnnl/executables/gen_index.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,22 +38,31 @@ struct genindex_executable_t : public op_executable_t {

void execute(const stream &stream,
const std::unordered_map<int, memory> &args) const override;
void execute_impl(const stream &stream,
const std::unordered_map<int, memory> &args) const;

#ifdef DNNL_WITH_SYCL
::sycl::event execute_sycl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<::sycl::event> &deps) const override;
::sycl::event execute_sycl_impl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<::sycl::event> &deps) const;
#endif

#if DNNL_GPU_RUNTIME == DNNL_RUNTIME_OCL
cl_event execute_ocl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<cl_event> &deps) const override;
cl_event execute_ocl_impl(const stream &stream,
const std::unordered_map<int, memory> &args,
const std::vector<cl_event> &deps) const;
#endif

private:
int axis_, nelems_, ndims_;
dims_t output_dims_, output_strides_;
std::string info_;

#if (DNNL_GPU_RUNTIME != DNNL_RUNTIME_NONE) \
&& (DNNL_GPU_VENDOR == DNNL_VENDOR_INTEL)
Expand Down
Loading