I am trying to find out a way to increase the computation time of a function to 1 second without using the sleep function in xilinx microblaze, using the xilkernel.
Hence, may i know how many iterations do i need to do in a simple for loop to increase the computation time to 1 second?
You can't do this reliably and accurately. If you want do a bodge like this, you'll have to calibrate it yourself for your particular system as Microblaze is so configurable, there isn't one right answer. The bodgy way is:
Set up a GPIO peripheral, set one of the pins to '1', run a loop of 1000 iterations (make sure the compiler doesn't optimise it away!) set the pins to '0'. Hang a scope off that pin (you're doing work on embedded systems, you do have a scope, right?) and see how long it takes to run the loop.
But the right way to do it is to use a hardware timer peripheral. Even at a very simple level, you could clear the timer at the start of the function, then poll it at the end until it reaches whatever value corresponds to 1 second. This will still have some imperfections, but given that you haven't specified how close to 1 sec you need to be, it is probably adequate.