Visible to Intel only — GUID: GUID-35D7F3BA-9CD9-4FD3-B232-E6FFE37674A6
Introduction to Intel® Graphics Performance Analyzers
Game Optimization Methodology
Install and Launch Intel® GPA
Enable GPU Metrics Collection
Launch Application for Analysis
Configure Analysis Settings
Analyze Desktop Graphics Applications
Quick Start with Profiling Unreal Engine* Games
Instrumentation and Tracing Technology API Support
Intel® GPA Reference
Notices and Disclaimers
Visible to Intel only — GUID: GUID-35D7F3BA-9CD9-4FD3-B232-E6FFE37674A6
Instrument Your Application
Graphics Trace Analyzer is an instrumentation-based tool. It obtains its data from gpa_trace files, which are generated by Intel® GPA while profiling an application that has been instrumented with Instrumentation and Tracing Technology API (ITT API) calls.
To get the most out of the ITT APIs, you need to add API calls in your code to designate logical tasks. This will help you visualize the relationship between tasks in your code, including when they start and end, relative to other CPU and GPU tasks.
At the highest level a task is a logical group of work executing on a specific thread, and may correspond to any grouping of code within your program that you consider important. You can mark up your code by identifying the beginning and end of each logical execution chunk.
To resolve the majority of performance bottlenecks, the following API calls are enough
- __itt_domain_create() creates a domain required in most ITT API calls. You need to define at least one domain.
- __itt_string_handle_create() creates string handles for identifying your tasks. String handles are more efficient for identifying traces than strings.
- __itt_task_begin() marks the beginning of a task.
Example
The following sample shows how four basic ITT API functions are used in a multi threaded application:
#include <windows.h>
#include <ittnotify.h>
// Forward declaration of a thread function.
DWORD WINAPI workerthread(LPVOID);
bool g_done = false;
// Create a domain that is visible globally: we will use it in our example.
__itt_domain* domain = __itt_domain_create("Example.Domain.Global");
// Create string handles which associates with the "main" task.
__itt_string_handle* handle_main = __itt_string_handle_create("main");
__itt_string_handle* handle_createthread = __itt_string_handle_create("CreateThread");
void main(int, char* argv[])
{
// Create a task associated with the "main" routine.
__itt_task_begin(domain, __itt_null, __itt_null, handle_main);
// Now we'll create 4 worker threads
for (int i = 0; i < 4; i++)
{
// We might be curious about the cost of CreateThread. We add tracing to do the measurement.
__itt_task_begin(domain, __itt_null, __itt_null, handle_createthread);
::CreateThread(NULL, 0, workerthread, (LPVOID)i, 0, NULL);
__itt_task_end(domain);
}
// Wait a while,...
::Sleep(5000);
g_done = true;
// Mark the end of the main task
__itt_task_end(domain);
}
// Create string handle for the work task.
__itt_string_handle* handle_work = __itt_string_handle_create("work");
DWORD WINAPI workerthread(LPVOID data)
{
// Set the name of this thread so it shows up in the UI as something meaningful
char threadname[32];
wsprintf(threadname, "Worker Thread %d", data);
__itt_thread_set_name(threadname);
// Each worker thread does some number of "work" tasks
while(!g_done)
{
__itt_task_begin(domain, __itt_null, __itt_null, handle_work);
::Sleep(150);
__itt_task_end(domain);
}
return 0;
}
Parent topic: Instrumentation and Tracing Technology API Support