How to measure the time taken per call when a single or multiple process is accessing?

Issue

• How to measure the time taken per call when a single or multiple process is accessing?
• calculate time taken by per call when accessing process.
• Is there any impact if multiple processes(10+) accessing same memory(shared) location using a single call?

eg:-  Currently following GNU call is in use
Currently I am using the GNU call:

__asm __volatile ("lock; cmpxchgl %2, %1"
                         : "=a" (ret), "=m" (*mem)
                         : "r" (newval), "m" (*mem), "0" (oldval));