Download Memory Allocation Lab: Implementing Fit Algorithms - Prof. Olmsted and more Exercises Computer Science in PDF only on Docsity!
Memory Allocation Lab
This lab project addresses the various memory-management schemes in an operating system. Reference Textbook : Garrido, José M., and Richard Schlesinger. Principles of modern operating systems. Jones & Bartlett Learning, (any edition 2 - 4) We model the memory managed in an operating system as a list of memory blocks. Each block of memory is modeled by a data structure of MEMORY_BLOCK defined below: MEMORY_BLOCK = { start_address: 0 end_address: 0 segment_size: 0 process_id: 0; //0 indicates a free block } Thus, the operating system code can set up memory blocks for the available memory and keep track of which blocks are allocated to which processes. In particular, note that for each memory block, the system tracks the starting and ending addresses, along with the size of the block and the process to which the block is currently allocated. For free blocks, the process is indicated as
The system manages the memory as an array of memory blocks, stored in an array. As memory is allocated and released, the list of blocks in the memory map changes. In this lab project, we address the various ways in which memory can be allocated (based on different schemes - Best Fit, First Fit, Next Fit, Worst (Largest) Fit). In addition, we also develop a method to manage the release of memory blocks by processes. For the purposes of these labs we define the NULLBLOCK as [start_address: 0, end_address: 0, segment_size: 0, process_id: 0]
Implement the following five methods in a file called memory.py. A starting place for the file would be: def best_fit_allocate(request_size, memory_map,process_id): pass def first_fit_allocate(request_size, memory_map,process_id): pass def worst_fit_allocate(request_size, memory_map,process_id): pass def next_fit_allocate(request_size, memory_map, process_id, last_address): pass def release_memory(freed_block, memory_map): pass
(a list of MEMORY_BLOCK dictionaries). It finds the first (lowest starting address) free memory block whose size is at least as large as the requested size. If the free block found is exactly of the same size as the requested size, the method updates the process id to allocate it and returns this memory block. If the free block found is larger than the requested size, the block is split into two pieces - the first piece allocated and the second piece becoming a free block in the memory map. Thus, the method may alter the memory map appropriately. Note that if there is no free block of memory (in the memory map) that is at least as large as the requested size, the method returns the NULLBLOCK. The signature of the method is as follows: first_fit_allocate(request_size, memory_map, process_id); A sample execution input and output: input/output parameter value input memory_map [start_address: 0, end_address: 1023, segment_size: 1024, process_id: 0] input request_size 10 input process_id 32 output memory_map [start_address: 0, end_address: 9, segment_size: 10, process_id: 32], [start_address: 10, end_address: 1023, segment_size: 1014, process_id: 0] output MEMORY_BLOCK [start_address: 0, end_address: 9, segment_size: 10, process_id: 32] Please refer to Section 3.2 of the Modern Operating Systems book for a detailed discussion of the First Fit algorithm.
worst_fit_allocate: This method allocates memory according to the Worst Fit scheme. The method is given the process id of the requesting process, size of the memory being requested, and the memory map (a list of MEMORY_BLOCK dictionaries). It finds the candidate memory blocks that can be allocated and chooses the largest among these blocks. If the free block found is exactly of the same size as the requested size, the method updates the process id to allocate it and returns this memory block. If the free block found is larger than the requested size, the block is split into two pieces - the first piece allocated and the second piece becoming a free block in the memory map. Thus, the method may alter the memory map appropriately. Note that if there is no free block of memory (in the memory map) that is at least as large as the requested size, the method returns the NULLBLOCK. The signature of the method is as follows: worst_fit_allocate(request_size, memory_map],process_id); A sample execution input and output: input/output parameter value input memory_map [start_address: 0, end_address: 1023, segment_size: 1024, process_id: 0] input map_cnt 1 input request_size 10 input process_id 32 output memory_map [start_address: 0, end_address: 9, segment_size: 10, process_id: 32], [start_address: 10, end_address: 1023, segment_size: 1014, process_id: 0] output map_cnt 2 output MEMORY_BLOCK [start_address: 0, end_address: 9, segment_size: 10, process_id: 32] Please refer to Section 3.2 of the Modern Operating Systems book for a detailed discussion of the Worst Fit algorithm.
release_memory: This method releases a memory block. Accordingly, it modifies the memory map passed in. Specifically, it marks the released block of memory as free and then it merges that block with adjacent free blocks if any. That is, if the memory block adjacent to the newly released block is free, the memory map is altered to reduce the number of memory blocks by one and the ending address (and the size) of the previous free block extended. Note that the method does not have any explicit return value and instead modifies the memory map passed in. The signature of the method is as follows: release_memory(freed_block, memory_map); A sample execution input and output: input/output parameter value input memory_map [start_address: 0, end_address: 7, segment_size: 8, process_id: 12], [start_address: 8, end_address: 15, segment_size: 8, process_id: 0], [start_address: 16, end_address: 23, segment_size: 8, process_id: 13], [start_address: 24, end_address: 27, segment_size: 4, process_id: 0], [start_address: 28, end_address: 29, segment_size: 2, process_id: 11] input freed_block [start_address: 16, end_address: 23, segment_size: 8, process_id: 13] output memory_map [start_address: 0, end_address: 7, segment_size: 8, process_id: 12], [start_address: 8, end_address: 27, segment_size: 20, process_id: 0], [start_address: 28, end_address: 29, segment_size: 2, process_id: 11]
