1. Rensselaer Polytechnic Institute CSCI-4210 – Operating Systems David Goldschmidt, Ph.D.

2. very small very large very fast very slow volatile non-volatile

3.  Locations in memory are identified by memory addresses  When compiled, programs consist of relocatable code  Other compiled modules also consist of relocatable code symbolic addresses in source code relative addresses in object code

4.  At load time, any additional libraries also consist of relocatable code physical addresses generated by loader

5.  At run time, memory addresses of all object files are mapped to a single memory space in physical memory

6.  A pair of base and limit registers define the logical address space  Also known as relocation registers

7.  Variable-length or dynamic partitions:  When a new process enters the system, the process is allocated to a single contiguous block  The operating system maintains a list of allocated partitions and free partitions OS Process 5 Process 8 Process 2 OS Process 5 Process 2 OS Process 5 Process 2 Process 9 OS Process 5 Process 9 Process 2 Process 1

8.  How can we place new process P i in memory?  First-fit algorithm: allocate the first free block that’s large enough to accommodate P i  Best-fit algorithm: allocate the smallest free block that’s large enough to accommodate P i  Next-fit algorithm: allocate the next free block, searching from last allocated block  Worst-fit algorithm: allocate the largest free block that’s large enough to accommodate P i

9.  Memory is wasted due to fragmentation, which can cause performance issues  Internal fragmentation is wasted memory within a partition or process memory  External fragmentation can reduce the number of runnable processes ▪ Total memory space exists to satisfy a memory request, but memory is not contiguous OS Process 5 Process 8 Process 2 Process 3 Process 6 Process 12 Process 7 Process 9

10.  A noncontiguous memory allocation scheme avoids the external fragmentation problem  Slice up physical memory into fixed-sized blocks called frames ▪ Sizes typically range from 2 9 to 2 14  Slice up logical memory into fixed-sized blocks called pages  Allocate pages into frames ▪ Note that frame size equals page size

11.  When a process of size n pages is ready to run, operating system finds n free frames  The OS keeps track of pages via a page table main memory process P i == in use == free

12.  Page tables map logical memory addresses to physical memory addresses

13.  Example process P i needs 16MB of logical memory  Page size is 4MB  Logical memory is mapped to a 32MB physical memory  Frame size is 4MB binary 0 ==> 000000 4 ==> 000100 8 ==> 001000 12 ==> 001100 16 ==> 010000 20 ==> 010100 24 ==> 011000 28 ==> 011100

15.  Every logical address is sliced into two distinct components:  Page number (p): used as an index into the page table to obtain the base physical memory address  Page offset (d): combined with the base address to identify the physical memory address page number page offset p d

16.  Covers a logical address space of size 2 m with page size 2 n page number page offset p d (m – n)(n)

18. 1 2  The page table is in main memory  Every memory access request actually requires two memory accesses:

19.  Use page table caching at the hardware level to speed address translation  Hardware-level translation look-aside buffer (TLB)

20.  Given:  Memory access time is 100 nanoseconds  TLB access time is 20 nanoseconds  TLB hit ratio is 80%  The effective memory-access time (EMAT) is  0.80 x 120 ns + 0.20 x 220 ns = 140 ns  What is the effective memory-access time given a hit ratio of 99%? 50%?

21.  For large page tables, use multiple page table levels  Slice up the logical address into multiple page indicators

22.  Processes in the ready queue have memory images waiting on disk  Processes are swapped in and out of memory  Can suffer from slow data transfer times