[SDK] Add Stack IDs (#269)

* Add Stack IDs

* Add memcpy test

* Add async corr id record

* Async events use `rocprofiler_async_correlation_id_t`
* Sync events use `rocprofiler_correlation_id_t`

* Update ATT to use asnyc IDs

* Review comments

source/lib/rocprofiler-sdk/pc_sampling/parser/correlation.hpp

@@ -49,8 +49,8 @@ namespace Parser
 {
 struct dispatch_correlation_ids_t
 {
-    rocprofiler_dispatch_id_t    dispatch_id;
-    rocprofiler_correlation_id_t correlation_id;
+    rocprofiler_dispatch_id_t          dispatch_id;
+    rocprofiler_async_correlation_id_t correlation_id;
 };
 
 /**
@@ -107,6 +107,14 @@ struct std::hash<Parser::DispatchPkt>
 
 namespace Parser
 {
+// 64B for performance reasons
+constexpr auto pcs_parser_sample_record_size = 64;
+static_assert(sizeof(generic_sample_t) == pcs_parser_sample_record_size);
+static_assert(sizeof(generic_sample_t) == sizeof(perf_sample_snapshot_v1));
+static_assert(sizeof(generic_sample_t) == sizeof(perf_sample_host_trap_v1));
+static_assert(sizeof(generic_sample_t) == sizeof(upcoming_samples_t));
+static_assert(sizeof(generic_sample_t) == sizeof(dispatch_pkt_id_t));
+
 /**
  * Coordinates DispatchMap and DoorBellMap to reconstruct the original correlation_id
  * from the correlation_id seen by the trap handler.

source/lib/rocprofiler-sdk/pc_sampling/parser/rocr.h

@@ -72,8 +72,8 @@ typedef struct
     uint64_t      write_index;
     uint64_t      read_index;
     /// both internal and external correlation ID.
-    rocprofiler_correlation_id_t correlation_id;
-    rocprofiler_dispatch_id_t    dispatch_id;
+    rocprofiler_async_correlation_id_t correlation_id;
+    rocprofiler_dispatch_id_t          dispatch_id;
 } dispatch_pkt_id_t;
 
 typedef struct

source/lib/rocprofiler-sdk/pc_sampling/parser/stochastic_records.h

@@ -169,10 +169,10 @@ typedef struct rocprofiler_pc_sampling_record_stochastic_v0_t
     rocprofiler_pc_t                   pc;            ///< information about sampled program counter
     uint64_t                           exec_mask;     ///< active SIMD lanes at the moment sampling
     rocprofiler_dim3_t                 workgroup_id;  ///< wave coordinates within the workgroup
-    uint32_t                     wave_count;   /// active waves on the CU at the moment of sampling
-    uint64_t                     timestamp;    ///< timestamp when sample is generated
-    uint64_t                     dispatch_id;  ///< originating kernel dispatch ID
-    rocprofiler_correlation_id_t correlation_id;
+    uint32_t wave_count;   /// active waves on the CU at the moment of sampling
+    uint64_t timestamp;    ///< timestamp when sample is generated
+    uint64_t dispatch_id;  ///< originating kernel dispatch ID
+    rocprofiler_async_correlation_id_t correlation_id;
     rocprofiler_pc_sampling_snapshot_v0_t
         snapshot;  ///< @see ::rocprofiler_pc_sampling_snapshot_v0_t
     rocprofiler_pc_sampling_memory_counters_t