24.4. Micro Program Execution Flow

So, basically a full program which is one of the main emphasis, how to see a full program which is written or and or the idea we will get that how one or after the macro instructions are executed in terms of a micro instructions when it is a micro program concept. So, a full program is written in terms of machine instruction that is macro program how it is executed. For each machine instruction when it is in the decoding cycle based on the Op-Code, the corresponding micro program is loaded into the memory. For first phase they have means very simple that is fetch.

This chunk discusses how a full program is executed via micro instructions after it's broken down from macro instructions. Each macro instruction corresponds to a micro program which contains the necessary steps to follow during the execution phase. Initially, the first step is the fetch operation, where the necessary data is gathered for processing.

Think of a chef preparing a dish. The overall recipe represents the full program (macro instruction), and the individual steps within that recipe correspond to the micro instructions. The first step is like fetching all the ingredients (fetch operation) needed to begin cooking.

So, already we have seen that for each fetch, the corresponding three micro microinstruction are executed, after that basically it waits; that means what? So, whenever a new instruction has to be executed, the micro program PC is pointed to the instruction which corresponds to basically your fetch.

This chunk covers the execution sequence during the fetch phase, where three specific micro instructions are operated to gather the necessary data for the next steps. Once the fetch is completed, the micro program counter (MPC) waits for the next instruction to be decoded.

Imagine the chef has gathered all ingredients (the three micro instructions), and now is simply waiting for a recipe to continue cooking. The chef cannot proceed until they decide what to cook next, similar to how the MPC waits for the next instruction.

Then what happens? Then the instruction register actually decode what the exact instruction is. If it is add, if it is store, if it is load accordingly the MPC will be pointed to different part. Maybe this part of the memory corresponds to add. So, it will jump over here.

When the fetch phase is completed, the instruction register decodes the specific nature of the instruction (e.g., add, store, load). Based on this decoded instruction, the MPC will redirect to distinct areas in memory mapped to execute these instructions.

Consider the chef interpreting a recipe's instructions (decoding). If the recipe specifies 'sauté vegetables', the chef will know to reach for the frying pan instead of a baking dish. The MPC is similarly making decisions based on what it has decoded.

Then, it starts actually the real execution of this instruction in terms of its micro program. So, this is one thing we are going to see for some temp template instructions. The micro program counter fetches each micro instruction in sequence and then generate the required signal or if required jumps will also be there.

Once the instruction has been decoded, the real execution begins, where the MPC fetches and executes each micro instruction in sequence. This process may involve various jumps based on results or conditions set within the instructions passed earlier.

Envision the chef following the recipe step by step, sautéing and adding ingredients as specified. If a condition arises, like needing more salt, the chef might skip back to a previous step to correct it or make adjustments, much like the MPC does with its jumps.

So, one the present micro program is called is complete, the next micro program is loaded based on the new instruction. So, once this add has been done then again automatically it finishes, then automatically the micro program will start pointing to this part, which is actually again fetch.

After completing the current instruction, the MPC automatically prepares to fetch the next instruction, effectively resetting the process. This cyclical nature is vital for efficient program execution, allowing the micro program to adapt to continuous input.

Just as a chef finishes one dish and immediately begins to prepare the next, the MPC seamlessly transitions to fetching new instructions without unnecessary delays or interruptions, ensuring a steady workflow.

So, what is the solution? That basically because there are so many different type of addressing modes etcetera. And if you write different micro codes for each addressing modes, each type of instruction type so it will be a huge memory, but you can do it. It will solve the problem; depending on the IR decoding you can directly go to that instruction and execute it, but that is actually going to be a very very unoptimized solution.

This chunk identifies the issue of memory efficiency in writing a separate micro program for every instruction. Instead, a more effective solution is suggested: writing common programs and introducing branching for variations, which reduces memory usage and optimizes execution.

Imagine a chef who maintains a single sauce base for multiple dishes instead of creating a unique sauce for every recipe. This approach saves time and ingredients, showcasing the optimization method being discussed.

So, therefore, basically we can write a single micro program for add and sub with just a branching instruction. That if we find out the instruction decoder is saying for the sub, then slight the slight change will go to one part and if it is add you just go to one part then the other part can be common, so that way we will actually save in the micro program memory size.

Here, it's explained that branching instructions allow us to reuse code for similar instructions (like addition and subtraction) by redirecting to particular segments only when necessary. This optimization makes the program memory smaller and faster.

Think of it as a common portion of a song where you only change the verses for different occasions. Instead of writing a new song every time, you exploit the common melody and just adjust the lyrics where needed.

--