Determining the Size of a Stack

The size of the stack required by the task is application specific. When sizing the stack, however, one must account for the nesting of all the functions called by the task, the number of local variables to be allocated by all functions called by the task, and the stack requirements for all nested interrupt service routines. In addition, the stack must be able to store all CPU registers and possibly Floating-Point Unit (FPU) registers if the processor has a FPU. In addition, as a general rule in embedded systems, avoid writing recursive code.

It is possible to manually figure out the stack space needed by adding all the memory required by all function call nesting (1 pointer each function call for the return address), plus all the memory required by all the arguments passed in those function calls, plus storage for a full CPU context (depends on the CPU), plus another full CPU context for each nested ISRs (if the CPU doesn’t have a separate stack to handle ISRs), plus whatever stack space is needed by those ISRs. Adding all this up is a tedious chore and the resulting number is a minimum requirement. Most likely you would not make the stack size that precise in order to account for “surprises.” The number arrived at should probably be multiplied by some safety factor, possibly 1.5 to 2.0. This calculation assumes that the exact path of the code is known at all times, which is not always possible. Specifically, when calling a function such as printf() or some other library function, it might be difficult or nearly impossible to even guess just how much stack space printf() will require. In this case, start with a fairly large stack space and monitor the stack usage at run-time to see just how much stack space is actually used after the application runs for a while.

There are really cool and clever compilers/linkers that provide this information in a link map. For each function, the link map indicates the worst-case stack usage. This feature clearly enables you to better evaluate stack usage for each task. It is still necessary to add the stack space for a full CPU context plus, another full CPU context for each nested ISR (if the CPU does not have a separate stack to handle ISRs), plus whatever stack space is needed by those ISRs. Again, allow for a safety net and multiply this value by some factor.

Always monitor stack usage at run-time while developing and testing the product as stack overflows occur often and can lead to some curious behaviors. In fact, whenever someone mentions that his or her application behaves “strangely,” insufficient stack size is the first thing that comes to mind.