Measuring data transfer and execution time

I was wondering if there was a way to measure the time it took to transfer my data to the device, the time the kernel took to do its job and the time it took to send the results back?  Currently I am invoking my kernels like this:

queue.putReadBuffer(...).put1DKernel(..).putWriteBuffer(..)

Thank you in advance!

I have figured out how to put my command queue into PROFILING_MODE, but I still dont know how to create a CLEvent object (it does not have a constructor). Lets say I wanted to measure how long the kernel execution takes...how do I initialize a CLEvent for it...