TMS320C6748: SYSBIOS performance after porting

Hi Team,

Can you please help with the issue below? 

A customer ported TI-RTOS to their project and encountered a problem using code like this, running with sysbios and running with sysbios (even before bios_start()), the time takes is very different:
1.    i = 4560;

2.    while(i)

3.    { i--; };

In the bare metal state, each cycle only needs 9 clock cycles ticks, but under sysbios, even before the OS is initialized, each cycle also needs more than 340 clock cycles ticks. why?

let me know if you need more information.

Regards,

Marvin