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

HIP Streams to Queues Translation (#235)

Browse Source 
* rocprofiler_stream_id_t: opaque handle for a stream

- e.g. HIP stream
- the same HIP stream may map to different HSA queues at different points in the application
- added to:
  - rocprofiler_buffer_tracing_hip_api_record_t
  - rocprofiler_buffer_tracing_memory_copy_record_t
  - rocprofiler_callback_tracing_hip_api_data_t
  - rocprofiler_callback_tracing_memory_copy_data_t
---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
Co-authored-by: Mark Meserve <mark.meserve@amd.com>
Co-authored-by: Elwazir, Ammar <Ammar.Elwazir@amd.com>
Co-authored-by: Ammar ELWazir <aelwazir@amd.com>
Co-authored-by: Jakaraddi, Manjunath <Manjunath.Jakaraddi@amd.com>
Co-authored-by: Bhardwaj, Gopesh <Gopesh.Bhardwaj@amd.com>
Co-authored-by: Nagaraj, Sriraksha <Sriraksha.Nagaraj@amd.com>
Co-authored-by: U, Srihari <Srihari.U@amd.com>
Co-authored-by: Madsen, Jonathan <Jonathan.Madsen@amd.com>
Co-authored-by: Welton, Benjamin <Benjamin.Welton@amd.com>
Co-authored-by: Benjamin Welton <ben@amd.com>
Co-authored-by: Indic, Vladimir <Vladimir.Indic@amd.com>
Co-authored-by: Benjamin Welton <bewelton@amd.com>
This commit is contained in:
Trowbridge, Ian
2025-03-14 04:45:13 -05:00
committed by GitHub
parent aa88dd44c7
commit ccd1e54293
50 changed files with 1812 additions and 220 deletions
Download Patch File Download Diff File Expand all files Collapse all files
source/lib/rocprofiler-sdk/hsa/queue_controller.cpp
+27 -25
View File
@@ -274,31 +274,7 @@ QueueController::init(CoreApiTable& core_table, AmdExtTable& ext_table)
}
}
auto enable_intercepter = false;
for(const auto& itr : context::get_registered_contexts())
{
constexpr auto expected_context_size = 216UL;
static_assert(
sizeof(context::context) == expected_context_size,
"If you added a new field to context struct, make sure there is a check here if it "
"requires queue interception. Once you have done so, increment expected_context_size");
bool has_kernel_tracing = itr->is_tracing(ROCPROFILER_CALLBACK_TRACING_KERNEL_DISPATCH) ||
itr->is_tracing(ROCPROFILER_BUFFER_TRACING_KERNEL_DISPATCH);
bool has_scratch_reporting = itr->is_tracing(ROCPROFILER_CALLBACK_TRACING_SCRATCH_MEMORY) ||
itr->is_tracing(ROCPROFILER_BUFFER_TRACING_SCRATCH_MEMORY);
if(itr->counter_collection || itr->pc_sampler || has_kernel_tracing ||
has_scratch_reporting || itr->device_counter_collection || itr->agent_thread_trace ||
itr->dispatch_thread_trace)
{
enable_intercepter = true;
break;
}
}
if(enable_intercepter)
if(enable_queue_intercept())
{
core_table.hsa_queue_create_fn = hsa::create_queue;
core_table.hsa_queue_destroy_fn = hsa::destroy_queue;
@@ -475,6 +451,32 @@ get_queue_controller()
return controller;
}
bool
enable_queue_intercept()
{
for(const auto& itr : context::get_registered_contexts())
{
constexpr auto expected_context_size = 216UL;
static_assert(
sizeof(context::context) == expected_context_size,
"If you added a new field to context struct, make sure there is a check here if it "
"requires queue interception. Once you have done so, increment expected_context_size");
bool has_kernel_tracing = itr->is_tracing(ROCPROFILER_CALLBACK_TRACING_KERNEL_DISPATCH) ||
itr->is_tracing(ROCPROFILER_BUFFER_TRACING_KERNEL_DISPATCH);
bool has_scratch_reporting = itr->is_tracing(ROCPROFILER_CALLBACK_TRACING_SCRATCH_MEMORY) ||
itr->is_tracing(ROCPROFILER_BUFFER_TRACING_SCRATCH_MEMORY);
if(itr->counter_collection || itr->pc_sampler || has_kernel_tracing ||
has_scratch_reporting || itr->device_counter_collection || itr->agent_thread_trace ||
itr->dispatch_thread_trace)
return true;
}
return false;
}
void
queue_controller_init(HsaApiTable* table)
{