We conducted a full attack on Kalyna 128/128 to demonstrate that our design is practical. 
The random key we choose leads to the worst case of reovering the full key.

We have overall 69248 candidate odd bytes.
We group these candidates to 4329 groups and each group contains 16 candidates, except the final group only contain one candidate. 
The correct guess of 4 odd bytes is among these candidates, so we don't have preknowledege on when the program will test the correct guess.

The whole test was run on the PHOENIX HPC.
In each folder, there is a test_input.txt, 16 output_result, and one output given by PHOENIX HPC.

The test_input.txt contains the candidates of 4 odd bytes. 
In output_results, you will find the tested odd bytes, the execution time and whether we found the key or not.
The execution time in output_results record the time cost of finding the whole key based on the guess of 4 odd bytes. 

The output given by PHOENIX HPC is more comprehensive.
It includes the start_time and end_time of the program.
It also includes the wall time and CPU-time of trying all 16 odd bytes candidates. 
Note that although we have 2 cores, the program only runs on one core. 
Except the time information, it also has the memory usage information. 

Information on the correct guess and its output is in test_folder3928.