Tizen Native API 5.5
These functions allow you to add a benchmark framework to a project for timing critical parts and detecting slow parts of code. It is used in Eina to compare the time used by eina, glib, evas and ecore data types.
To use the benchmark module, Eina must be initialized with eina_init() and later shut down with eina_shutdown(). A benchmark is created with eina_benchmark_new() and freed with eina_benchmark_free().
eina_benchmark_register() adds a test to a benchmark. That test can be run a certain amount of times. Adding more than one test to be executed allows the comparison between several parts of a program, or different implementations.
eina_benchmark_run() runs all the tests registered with eina_benchmark_register(). The amount of time of each test is written in a gnuplot file.
For more information, you can look at the Benchmark Tutorial.
|Eina_Benchmark *||eina_benchmark_new (const char *name, const char *run)|
|Creates a new array. |
|void||eina_benchmark_free (Eina_Benchmark *bench)|
|Frees a benchmark object. |
|Eina_Bool||eina_benchmark_register (Eina_Benchmark *bench, const char *name, Eina_Benchmark_Specimens bench_cb, int count_start, int count_end, int count_step)|
|Adds a test to a benchmark. |
|Eina_Array *||eina_benchmark_run (Eina_Benchmark *bench)|
|Runs the benchmark's registered tests. |
|typedef struct _Eina_Benchmark||Eina_Benchmark|
|typedef void(*||Eina_Benchmark_Specimens )(int request)|
|Definition for the cast to an Eina_Benchmark_Specimens. |
Definition for the cast to an Eina_Benchmark_Specimens.
[in] function The function to cast.
This macro casts
function to Eina_Benchmark_Specimens.
Type for a benchmark.
Type for a test function to be called when running a benchmark.
|void eina_benchmark_free||(||Eina_Benchmark *||bench||)|
Frees a benchmark object.
[in,out] bench The benchmark to free.
This function removes all the benchmark tests that have been registered and frees
NULL, this function returns immediately.
- Since :
|Eina_Benchmark* eina_benchmark_new||(||const char *||name,|
|const char *||run|
Creates a new array.
[in] name The name of the benchmark. [in] run The name of the run.
- A valid benchmark on success, or
NULLon memory allocation failure.
This function creates a new benchmark.
run are used to name the gnuplot file that eina_benchmark_run() will create.
When the new module is not needed anymore, use eina_benchmark_free() to free the allocated memory.
- Since :
|Eina_Bool eina_benchmark_register||(||Eina_Benchmark *||bench,|
|const char *||name,|
Adds a test to a benchmark.
[in,out] bench The benchmark. [in] name The name of the test. [in] bench_cb The test function to be called. [in] count_start The start data to be passed to
[in] count_end The end data to be passed to
[in] count_step The step data to be passed to
- EINA_FALSE on failure, EINA_TRUE otherwise.
This function adds the test named
bench_cb is the function called when the test is executed. That test can be executed a certain number of times, from
count_end, with a step increment of
count_step. This counter is passed to
bench_cb when eina_benchmark_run() is called.
count_step is 0, this function returns immediately and does not add any tests to the benchmark.
- Since :
|Eina_Array* eina_benchmark_run||(||Eina_Benchmark *||bench||)|
Runs the benchmark's registered tests.
[in,out] bench The benchmark.
- A list of gnuplot filenames for the test results, or
This function runs all the tests that have been registered with eina_benchmark_register() and saves the result in gnuplot input files. The filenames have the following format:
Where [name] and [run] are the values passed to eina_benchmark_new() when registering the test.
Each registered test is executed and timed. The time is written to the gnuplot file. The number of times each test is executed is controlled by the parameters passed to eina_benchmark_register().
- Since :