Support sampling duration, sampling TIDs (#142)

- Sampling duration config values
  - OMNITRACE_SAMPLING_DURATION
  - OMNITRACE_PROCESS_SAMPLING_DURATION
  - Disables sampling after this time (in seconds) has elapsed 
- Sampling thread-id config values
  - OMNITRACE_SAMPLING_TIDS
  - OMNITRACE_SAMPLING_CPUTIME_TIDS
  - OMNITRACE_SAMPLING_REALTIME_TIDS
  - Allows user to select certain threads for sampling
- Miscellaneous
  - Tweaked the finalization verbosity messages
  - moved sampling-on-child-threads into runtime.hpp and runtime.cpp
  - fixed submodule dyninst header install

[ROCm/rocprofiler-systems commit: e67afd33eb]

projects/rocprofiler-systems/source/lib/omnitrace/library/components/pthread_create_gotcha.cpp

@@ -22,7 +22,6 @@
 
 #include "library/components/pthread_create_gotcha.hpp"
 #include "library/components/category_region.hpp"
-#include "library/components/pthread_gotcha.hpp"
 #include "library/components/roctracer.hpp"
 #include "library/config.hpp"
 #include "library/debug.hpp"
@@ -213,9 +212,8 @@ pthread_create_gotcha::wrapper::operator()() const
         if(m_enable_sampling)
         {
             _is_sampling = true;
-            pthread_gotcha::push_enable_sampling_on_child_threads(false);
+            OMNITRACE_SCOPED_SAMPLING_ON_CHILD_THREADS(false);
             _signals = sampling::setup();
-            pthread_gotcha::pop_enable_sampling_on_child_threads();
             sampling::unblock_signals();
         }
     }
@@ -336,7 +334,7 @@ pthread_create_gotcha::operator()(pthread_t* thread, const pthread_attr_t* attr,
     auto        _active       = (get_state() == ::omnitrace::State::Active && !_disabled);
     auto        _coverage     = (get_mode() == Mode::Coverage);
     auto        _use_sampling = get_use_sampling();
-    auto        _sample_child = pthread_gotcha::sampling_enabled_on_child_threads();
+    auto        _sample_child = sampling_enabled_on_child_threads();
     auto        _tid          = utility::get_thread_index();
     auto        _use_bundle   = (_active && !_coverage);
     const auto& _info = thread_info::init(!_active || !_sample_child || _disabled);

projects/rocprofiler-systems/source/lib/omnitrace/library/components/pthread_gotcha.cpp

@@ -69,14 +69,6 @@ namespace
 using bundle_t = tim::lightweight_tuple<component::pthread_create_gotcha_t,
                                         component::pthread_mutex_gotcha_t>;
 
-auto&
-get_sampling_on_child_threads_history(int64_t _idx = utility::get_thread_index())
-{
-    static auto _v = utility::get_filled_array<OMNITRACE_MAX_THREADS>(
-        []() { return utility::get_reserved_vector<bool>(32); });
-    return _v.at(_idx);
-}
-
 auto&
 get_bundle()
 {
@@ -112,51 +104,6 @@ pthread_gotcha::shutdown()
     }
 }
 
-bool
-pthread_gotcha::sampling_enabled_on_child_threads()
-{
-    return sampling_on_child_threads();
-}
-
-bool
-pthread_gotcha::push_enable_sampling_on_child_threads(bool _v)
-{
-    bool _last                  = sampling_on_child_threads();
-    sampling_on_child_threads() = _v;
-    auto& _hist                 = get_sampling_on_child_threads_history();
-    _hist.emplace_back(_last);
-    return _last;
-}
-
-bool
-pthread_gotcha::pop_enable_sampling_on_child_threads()
-{
-    auto& _hist = get_sampling_on_child_threads_history();
-    if(!_hist.empty())
-    {
-        bool _restored = _hist.back();
-        _hist.pop_back();
-        sampling_on_child_threads() = _restored;
-    }
-    return sampling_on_child_threads();
-}
-
-void
-pthread_gotcha::set_sampling_on_all_future_threads(bool _v)
-{
-    for(size_t i = 0; i < max_supported_threads; ++i)
-        get_sampling_on_child_threads_history(i).emplace_back(_v);
-}
-
-bool&
-pthread_gotcha::sampling_on_child_threads()
-{
-    static thread_local bool _v = get_sampling_on_child_threads_history().empty()
-                                      ? false
-                                      : get_sampling_on_child_threads_history().back();
-    return _v;
-}
-
 void
 pthread_gotcha::start()
 {

projects/rocprofiler-systems/source/lib/omnitrace/library/components/pthread_gotcha.hpp

@@ -42,22 +42,7 @@ struct pthread_gotcha : tim::component::base<pthread_gotcha, void>
     static void configure();
     static void shutdown();
 
-    // query current value
-    static bool sampling_enabled_on_child_threads();
-
-    // use this to disable sampling in a region (e.g. right before thread creation)
-    static bool push_enable_sampling_on_child_threads(bool _v);
-
-    // use this to restore previous setting
-    static bool pop_enable_sampling_on_child_threads();
-
-    // make sure every newly created thead starts with this value
-    static void set_sampling_on_all_future_threads(bool _v);
-
     static void start();
     static void stop();
-
-private:
-    static bool& sampling_on_child_threads();
 };
 }  // namespace omnitrace

projects/rocprofiler-systems/source/lib/omnitrace/library/components/pthread_mutex_gotcha.cpp

@@ -22,7 +22,6 @@
 
 #include "library/components/pthread_mutex_gotcha.hpp"
 #include "library/components/category_region.hpp"
-#include "library/components/pthread_gotcha.hpp"
 #include "library/config.hpp"
 #include "library/critical_trace.hpp"
 #include "library/debug.hpp"
@@ -293,7 +292,7 @@ pthread_mutex_gotcha::is_disabled()
 {
     return (get_state() != ::omnitrace::State::Active ||
             get_thread_state() != ThreadState::Enabled ||
-            (get_use_sampling() && !pthread_gotcha::sampling_enabled_on_child_threads()));
+            (get_use_sampling() && !sampling_enabled_on_child_threads()));
 }
 }  // namespace component
 }  // namespace omnitrace

projects/rocprofiler-systems/source/lib/omnitrace/library/components/roctracer.cpp

@@ -22,13 +22,13 @@
 
 #include "library/components/roctracer.hpp"
 #include "library/common.hpp"
-#include "library/components/pthread_gotcha.hpp"
 #include "library/config.hpp"
 #include "library/debug.hpp"
 #include "library/defines.hpp"
 #include "library/dynamic_library.hpp"
 #include "library/redirect.hpp"
 #include "library/roctracer.hpp"
+#include "library/runtime.hpp"
 #include "library/sampling.hpp"
 #include "library/thread_data.hpp"
 
@@ -121,7 +121,7 @@ roctracer::setup()
     roctracer_is_setup() = true;
 
     OMNITRACE_VERBOSE_F(1, "setting up roctracer...\n");
-    pthread_gotcha::push_enable_sampling_on_child_threads(false);
+    OMNITRACE_SCOPED_SAMPLING_ON_CHILD_THREADS(false);
 
     dynamic_library _amdhip64{ "OMNITRACE_ROCTRACER_LIBAMDHIP64",
                                find_library_path("libamdhip64.so",
@@ -169,8 +169,6 @@ roctracer::setup()
     for(auto& itr : roctracer_setup_routines())
         itr.second();
 
-    pthread_gotcha::pop_enable_sampling_on_child_threads();
-
     OMNITRACE_VERBOSE_F(1, "roctracer is setup\n");
 }