Support for tracing mutex locking (#52)

* Parallel overhead example with locks

* Support tracing mutex locking + more

- support wrapping pthread_mutex_lock
- support wrapping pthread_mutex_unlock
- support wrapping pthread_mutex_trylock
- get_perfetto_combined_traces setting
- OMNITRACE_TRACE_THREAD_LOCKS option
- ThreadState
- critical trace includes queue id
- enabled/disabled settings in timemory
- fix OMNITRACE_TIMEMORY_COMPONENTS
- fix reading config
- fix setting categories
- applied ThreadState::Internal in various places
- utility::get_filled_array
- utility::get_reserved_vector
- utility::get_thread_index
- fork_gotcha messages about forks
- split out some pthread_gotcha functionality into pthread_create_gotcha
- handle queue id in roctracer callbacks

* Update timemory and PTL submodules

* Misc CMake updates

- Includes fix to omnitrace-static-lib{gcc,stdcxx}

* Misc cleanup to pthread_mutex_gotcha and backtrace

* Fix to duplicate field in module_function json

* Improvement to debug messages

* omnitrace-dl and common improvements

- tweak to delimit
- common::ignore message
- common::join quoting of strings
- omnitrace_set_env ignores if inited and active
- omnitrace_set_mpi ignores if inited and active

* nsync for transpose example

* Fix to thread_deleter<void> functor invoke

* Fix thread state and HIP stream enums

[ROCm/rocprofiler-systems commit: b208047741]
This commit is contained in:
Jonathan R. Madsen
2022-05-08 04:40:10 -05:00
کامیت شده توسط GitHub
والد 0094a471fd
کامیت 0d5f0fb9cf
50فایلهای تغییر یافته به همراه1736 افزوده شده و 483 حذف شده
@@ -439,7 +439,9 @@ main(int argc, char** argv)
parser.add_argument({ "" }, "");
parser.add_argument({ "[VIEW OPTIONS]" }, "");
parser.add_argument({ "-A", "--available" }, "Only display available components")
parser
.add_argument({ "-A", "--available" },
"Only display available components/settings/hw-counters")
.max_count(1)
.action([](parser_t& p) { available_only = p.get<bool>("available"); });
parser
@@ -892,8 +894,8 @@ write_settings_info(std::ostream& os, const array_t<bool, N>& opts,
_setting_output.end());
// patch up the categories
str_set_t _not_in_category_view{};
auto _settings = tim::settings::shared_instance();
auto _not_in_category_view = str_set_t{};
auto _settings = tim::settings::shared_instance();
for(auto& itr : _setting_output)
{
auto _name = itr.find("environ")->second;
@@ -942,6 +944,19 @@ write_settings_info(std::ostream& os, const array_t<bool, N>& opts,
}),
_setting_output.end());
if(available_only)
{
_setting_output.erase(
std::remove_if(_setting_output.begin(), _setting_output.end(),
[&_settings](const auto& itr) {
auto iitr = _settings->find(itr.at("environ"));
if(iitr != _settings->end())
return (iitr->second->get_enabled() == false);
return true;
}),
_setting_output.end());
}
if(alphabetical)
{
std::sort(_setting_output.begin(), _setting_output.end(),