Workflow, submodules, and thread info Updates (#352)

* Update CI workflows

- use node20 workflow packages

* Update tests/source/CMakeLists.txt

- Use OMNITRACE_TRACE and OMNTRACE_PROFILE instead of perfetto/timemory

* Update timemory submodule

- argparse: requires -> required
- parse callbacks

* Update thread_info.cpp

- fix causal::delay::get_local usage

* Update timemory submodule

* Update kokkos submodule

- release 3.7.02

* Revert opensuse.yml and ubuntu-bionic.yml to use node16 workflows

* Update docs.yml

.github/workflows/docs.yml

@@ -49,7 +49,7 @@ jobs:
           conda activate omnitrace-docs
           ./update-docs.sh
       - name: Upload artifact
-        uses: actions/upload-pages-artifact@v1
+        uses: actions/upload-pages-artifact@v3
         with:
           path: ./docs
 
@@ -67,4 +67,4 @@ jobs:
     steps:
       - name: Deploy to GitHub Pages
         id: deployment
-        uses: actions/deploy-pages@v1
+        uses: actions/deploy-pages@v4

source/bin/omnitrace-sample/impl.cpp

@@ -562,7 +562,7 @@ parse_args(int argc, char** argv, std::vector<char*>& _env)
     parser.add_argument({ "--profile-format" }, "Data formats for profiling results")
         .min_count(1)
         .max_count(3)
-        .requires({ "profile|flat-profile" })
+        .required({ "profile|flat-profile" })
         .choices({ "text", "json", "console" })
         .action([&](parser_t& p) {
             auto _v = p.get<std::set<std::string>>("profile");
@@ -624,7 +624,7 @@ parse_args(int argc, char** argv, std::vector<char*>& _env)
         .add_argument({ "--cpus" },
                       "CPU IDs for frequency sampling. Supports integers and/or ranges")
         .dtype("int or range")
-        .requires({ "host" })
+        .required({ "host" })
         .action([&](parser_t& p) {
             update_env(
                 _env, "OMNITRACE_SAMPLING_CPUS",
@@ -634,7 +634,7 @@ parse_args(int argc, char** argv, std::vector<char*>& _env)
         .add_argument({ "--gpus" },
                       "GPU IDs for SMI queries. Supports integers and/or ranges")
         .dtype("int or range")
-        .requires({ "device" })
+        .required({ "device" })
         .action([&](parser_t& p) {
             update_env(
                 _env, "OMNITRACE_SAMPLING_GPUS",
@@ -709,7 +709,7 @@ parse_args(int argc, char** argv, std::vector<char*>& _env)
 
     parser.add_argument({ "--realtime" }, _realtime_desc)
         .min_count(0)
-        .requires(std::move(_realtime_reqs))
+        .required(std::move(_realtime_reqs))
         .action([&](parser_t& p) {
             auto _v = p.get<std::deque<std::string>>("realtime");
             update_env(_env, "OMNITRACE_SAMPLING_REALTIME", true);

source/lib/core/argparse.cpp

@@ -870,7 +870,7 @@ add_core_arguments(parser_t& _parser, parser_data& _data)
             .min_count(1)
             .max_count(3)
             .dtype("string")
-            .requires({ "profile|flat-profile" })
+            .required({ "profile|flat-profile" })
             .choices({ "text", "json", "console" })
             .action([&](parser_t& p) {
                 auto _v = p.get<strset_t>("profile-format");
@@ -976,7 +976,7 @@ add_core_arguments(parser_t& _parser, parser_data& _data)
                 { "--cpus" },
                 "CPU IDs for frequency sampling. Supports integers and/or ranges")
             .dtype("int and/or range")
-            .requires({ "host" })
+            .required({ "host" })
             .action([&](parser_t& p) {
                 update_env(_data, "OMNITRACE_SAMPLING_CPUS",
                            join(array_config_t{ "," }, p.get<strvec_t>("cpus")));
@@ -992,7 +992,7 @@ add_core_arguments(parser_t& _parser, parser_data& _data)
             .add_argument({ "--gpus" },
                           "GPU IDs for SMI queries. Supports integers and/or ranges")
             .dtype("int and/or range")
-            .requires({ "device" })
+            .required({ "device" })
             .action([&](parser_t& p) {
                 update_env(_data, "OMNITRACE_SAMPLING_GPUS",
                            join(array_config_t{ "," }, p.get<strvec_t>("gpus")));
@@ -1117,7 +1117,7 @@ add_core_arguments(parser_t& _parser, parser_data& _data)
         _parser.add_argument({ "--sample-realtime" }, _realtime_desc)
             .min_count(0)
             .dtype("[freq] [delay] [tids...]")
-            .requires(std::move(_realtime_reqs))
+            .required(std::move(_realtime_reqs))
             .action([&](parser_t& p) {
                 auto _v = p.get<std::deque<std::string>>("sample-realtime");
                 update_env(_data, "OMNITRACE_SAMPLING_REALTIME", true);

source/lib/omnitrace/library/thread_info.cpp

@@ -35,6 +35,8 @@
 #include <timemory/components/timing/backends.hpp>
 #include <timemory/process/threading.hpp>
 
+#include <cstdint>
+
 namespace omnitrace
 {
 namespace
@@ -105,8 +107,9 @@ init_index_data(int64_t _tid, bool _offset = false)
     return itr;
 }
 
-const auto unknown_thread   = std::optional<thread_info>{};
+thread_local int64_t offset_causal_count = 0;
-int64_t    peak_num_threads = max_supported_threads;
+const auto           unknown_thread      = std::optional<thread_info>{};
+int64_t              peak_num_threads    = max_supported_threads;
 }  // namespace
 
 std::string
@@ -187,8 +190,13 @@ thread_info::init(bool _offset)
         _info                 = thread_info{};
         _info->is_offset      = threading::offset_this_id();
         _info->index_data     = init_index_data(_tid, _info->is_offset);
-        _info->causal_count   = &causal::delay::get_local();
         _info->lifetime.first = tim::get_clock_real_now<uint64_t, std::nano>();
+
+        const auto _sequent_tid = _info->index_data->sequent_value;
+        _info->causal_count     = (!_info->is_offset && _sequent_tid < peak_num_threads)
+                                      ? &causal::delay::get_local(_sequent_tid)
+                                      : &offset_causal_count;
+
         if(_info->is_offset) set_thread_state(ThreadState::Disabled);
     }
 

tests/source/CMakeLists.txt

@@ -9,7 +9,7 @@ target_compile_definitions(thread-limit PRIVATE MAX_THREADS=${OMNITRACE_MAX_THRE
 target_link_libraries(thread-limit PRIVATE Threads::Threads tests-compile-options)
 
 set(_thread_limit_environment
-    "${_base_environment}" "OMNITRACE_USE_PERFETTO=ON" "OMNITRACE_USE_TIMEMORY=ON"
+    "${_base_environment}" "OMNITRACE_TRACE=ON" "OMNITRACE_PROFILE=ON"
     "OMNITRACE_COUT_OUTPUT=ON" "OMNITRACE_USE_SAMPLING=ON" "OMNITRACE_SAMPLING_FREQ=250"
     "OMNITRACE_VERBOSE=2" "OMNITRACE_TIMEMORY_COMPONENTS=wall_clock,peak_rss,page_rss")