Route 64 Developer's Guide
<< || toc || >>

Model

Route 64 is fully based on the abstractions that are defined in the package de.michab.simulator. The set of interfaces and classes in this package lays out a generic definition of the hardware pieces that make up a computer. The model that is implemented there assumes that a computer consists of a set of chips that are placed at certain address positions in a common memory. Communication between the chips is done through reading and writing shared memory cells. As in a real computer system the synchronisation is based on a clock that coordinates and synchronises all actions in the system.

In the emulation a computer's memory is represented by an implementation of the Memory interface. In its simplest form a piece of memory behaves rather straight: In the case of RAM (random access memory) it allows to write a value into a certain memory address (also called memory cell) and returns the very same value if the address is read again. No modification of the value is performed, the memory is just a passive information store. For ROM (read only memory) the story is even simpler: In this case the values that the memory holds are predefined and cannot be modified.

Memory layout of the Commodore 64 uses both types of memory and is a bit more complicated since the memroy is not only adressed in a linear way. A general solution to place more memory in the 64k address space that the processor could address is called memory banking. This means to have several memory levels where the level that actually is to be used was selected by an additional register. Since the 64 had 64k of RAM as well as additional ROM areas, the technique was used in the Commodore 64. The register that allowed to define the current memory configuration is part of the processor.

Each of the different chips in the emulation is represented by an implementation of the Chip interface. Each chip has a set of registers or ports. These are memory cells that are implemented by a chip and that represent the programmatic interface to control the chip's functionality. The chip registers can allow read and write operations or only one of these. For example most of the Commodore 64's video chip registers allow read and write while most of the sound chip registers are write-only. When a chip is mapped into memory then the normal memory cells are in a certain address range replaced by the chip register. The first address that is replaced by a chip register is called the base address of a chip. A consequence is that certain memory addresses represent simple passive memory cells while others actually trigger chip internal functionality.

A special kind of a chip is the processor. A processor is the driving element in a computer system. In a loop it reads a register cell, interprets its contents as an instruction and executes the instruction, which potentially leads to memory writes, memory reads, or processor internal register updates. These processor internal registers must not be mixed with the chip registers described above; the processor registers cannot be addressed (read) from the outside. Instead the processor registers are used to hold the current status of the computation like the program counter position and intermediate computation values. The model defines Processor as the interface that is used to implement processor functionality. This interface allows to plug a debugger to a processor that is then called for each computation step the processor performs and that can display the processor status and the current instructions for debug purposes.

Last modified: 2005/07/21

Comments