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

Modified perfetto output for HIP stream display (#431)

Browse Source 
* Modified perfetto output for HIP stream display

* Moved stream_map file location and changed perfetto output names Private_Segment_Size and Group_Segment_Size to Scratch_Size and LDS_Block_Size respectively

* Used const_cast to remove const modifier on void*

* Reverted stream_map changes, now using tool_metadata map to track mapping between stream ptrs and stream IDs

* Removed buffer tracing args in perfetto, added tool_...hip buffer record struct that stores the HIP stream ID for display purposes

* Updated rocpd perfetto.cpp to reflect stream changes. Still need to add vgpr values and stream ID for HIP API

* Changes pass-by const reference to pass-by const value
This commit is contained in:
Trowbridge, Ian authored and GitHub committed 2025-06-26 14:22:50 -05:00
1 parent 8a461afe20
commit 1f8b8c5e9f
16 files changed
+126 -90
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

source/lib/rocprofiler-sdk-tool/tool.cpp
+8 -3
View File
@@ -615,6 +615,7 @@ hip_stream_display_callback(rocprofiler_callback_tracing_record_t record,
auto* stream_handle_data =
static_cast<rocprofiler_callback_tracing_hip_stream_data_t*>(record.payload);
auto stream_id = stream_handle_data->stream_id;
// STREAM_HANDLE_CREATE and DESTROY are no-ops
if(record.operation == ROCPROFILER_HIP_STREAM_CREATE)
{
@@ -1067,7 +1068,10 @@ buffered_tracing_callback(rocprofiler_context_id_t /*context*/,
auto* record =
static_cast<rocprofiler_buffer_tracing_hip_api_ext_record_t*>(header->payload);
tool::write_ring_buffer(*record, domain_type::HIP);
auto stream_id = get_stream_id(record);
tool::write_ring_buffer(
tool::tool_buffer_tracing_hip_api_ext_record_t{*record, stream_id},
domain_type::HIP);
}
else if(header->kind == ROCPROFILER_BUFFER_TRACING_RCCL_API)
{
@@ -2023,10 +2027,11 @@ tool_init(rocprofiler_client_finalize_t fini_func, void* tool_data)
tool::get_config().benchmark_mode != tool::config::benchmark::execution_profile)
{
auto external_corr_id_request_kinds =
std::array<rocprofiler_external_correlation_id_request_kind_t, 3>{
std::array<rocprofiler_external_correlation_id_request_kind_t, 4>{
ROCPROFILER_EXTERNAL_CORRELATION_REQUEST_KERNEL_DISPATCH,
ROCPROFILER_EXTERNAL_CORRELATION_REQUEST_MEMORY_COPY,
ROCPROFILER_EXTERNAL_CORRELATION_REQUEST_MEMORY_ALLOCATION};
ROCPROFILER_EXTERNAL_CORRELATION_REQUEST_MEMORY_ALLOCATION,
ROCPROFILER_EXTERNAL_CORRELATION_REQUEST_HIP_RUNTIME_API};
ROCPROFILER_CALL(rocprofiler_configure_external_correlation_id_request_service(
get_client_ctx(),