Enable queue interception with scratch reporting (#1069)

* Enable queue interception with scratch reporting

Scratch reporting reports agent ID in buffer and callback records, but
HSA runtime provides only queue ID in the scratch callback.

This change enables queue interception when scratch reporting is requested

* Validation test for rocprofv3 + scratch-memory-trace

* Simplify checks for whether context is tracing a domain

* Update changelog

---------

Co-authored-by: Jonathan R. Madsen <jonathanrmadsen@gmail.com>
This commit is contained in:
Mythreya
2024-09-12 16:26:34 -07:00
کامیت شده توسط GitHub
والد 6098d52335
کامیت efbe4ea0a2
9فایلهای تغییر یافته به همراه181 افزوده شده و 6 حذف شده
@@ -270,14 +270,15 @@ QueueController::init(CoreApiTable& core_table, AmdExtTable& ext_table)
"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->callback_tracer &&
itr->callback_tracer->domains(ROCPROFILER_CALLBACK_TRACING_KERNEL_DISPATCH)) ||
(itr->buffered_tracer &&
itr->buffered_tracer->domains(ROCPROFILER_BUFFER_TRACING_KERNEL_DISPATCH));
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 ||
itr->agent_counter_collection || itr->agent_thread_trace || itr->dispatch_thread_trace)
has_scratch_reporting || itr->agent_counter_collection || itr->agent_thread_trace ||
itr->dispatch_thread_trace)
{
enable_intercepter = true;
break;