AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Generic comm_data component (#132)

Browse Source 
* Generic comm_data component

- moved rccl_comm_data to comm_data
- comm_data includes communication data for MPI

* fix timemory include with quotes

* Only support MPI comm data with full MPI support

* Increase timeouts + kill perfetto

* Update timemory submodule

* Fix missing command killall

* set +e in Kill Perfetto workflow step

* Updated MPI example to include MPI_Send and MPI_Recv calls

* Update timemory submodule with storage merge fix

* Perfetto comm data

- tracing::now<T>() function

* Fix timemory header include
This commit is contained in:
Jonathan R. Madsen
2022-08-25 19:48:10 -05:00
committed by GitHub
parent a1afd69a02
commit 0dd8f52292
15 changed files with 860 additions and 248 deletions
Download Patch File Download Diff File Expand all files Collapse all files
source/lib/omnitrace/library/rcclp.cpp
+12 -9
View File
@@ -51,10 +51,7 @@ namespace rcclp
{
void
configure()
{
comp::rccl_data_tracker_t::label() = "rccl_comm_data";
comp::rccl_data_tracker_t::description() = "Tracks RCCL communication data";
}
{}
void
setup()
@@ -67,11 +64,17 @@ setup()
if(!librccl_handle) fprintf(stderr, "%s\n", dlerror());
dlerror(); // Clear any existing error
auto _data = tim::get_env("OMNITRACE_RCCLP_COMM_DATA", true);
if(_data)
comp::rccl_toolset_t::get_initializer() = [](comp::rccl_toolset_t& cb) {
cb.initialize<comp::rccl_comm_data>();
};
auto _use_data = tim::get_env("OMNITRACE_RCCLP_COMM_DATA", get_use_timemory());
if(!get_use_timemory())
{
trait::runtime_enabled<comp::comm_data>::set(false);
trait::runtime_enabled<comp::comm_data_tracker_t>::set(false);
}
else
{
trait::runtime_enabled<comp::comm_data>::set(_use_data);
trait::runtime_enabled<comp::comm_data_tracker_t>::set(_use_data);
}
comp::configure_rcclp();
global_id = comp::activate_rcclp();