Transcript Lecture 26

ITEC 352
Lecture 25
Memory(3)
Review
• Questions
• RAM
– What is the difference between register
memory, cache memory, and main memory?
– What connects the different types of memory
together?
– What are some of the different ways to handle
what is in the cache?
Memory (3)
Objectives
• Cache memory
• Intro to virtual memory
Memory (3)
Performanc
e of a cache
Memory address
range
Cache
• Assume a program
(instructions) is loaded
into the memory from
address 0 to 57. Also
assume: cache access
time: 80 ns and memory
access time: 2500 ns.
• What is the performance
when using a cache.
0 - 16
17 - 33
34-40
41-57
Memory (3)
Memory
Performanc
e of a cache
Memory address
range
Cache
• What is the memory
address of the first
instruction accessed?
– Address: 0
0 - 16
17 - 33
• Is this in the cache?
– Initially, cache is empty.
Hence, no!
34-40
• So what do you do?
– Depends on the policy
Memory (3)
41-57
Memory
Performanc
e of a cache
Memory address
range
Cache
•
•
As the data is not in the
cache, the block 0-16 is
loaded into the slot1 of the
cache.
The next instruction is at
address 1.
– Is this in the cache?
– Yes – so the next 15
instructions are in the
cache.
0 - 16
17 - 33
34-40
41-57
Memory (3)
Memory
Performanc
e of cache
• Hence:
Event Location Time
1 miss 0
2500 ns
15 hits 1-16
80ns X 15
…
Memory (3)
Hit Ratios and Effective Access Times
Hit ratio and effective access time for single level cache:
Hit ratios and effective access time for multi-level cache:
Memory (3)
Multilevel
caches
• As size of ICs have increased, packing
density also has increased.
– Multilevel caches have been developed
• Fastest level L1 is on the chip.
– Usually data and instructions are kept
separate on this cache. Called split cache.
• Level L2 and L3 are slower than L1 and
are unified caches.
Memory (3)
Question
• Lets say there are 10000 memory references to
execute a process.
– 90 cause L1 misses and of these 10 cause L2
misses.
– Let L1 hit time: 5 ns (this is the time to access a
memory location if it is in the cache).
– Let L2 hit time: 20 ns
– Let the L2 miss time: 100 ns (time to access memory
in the main memory).
• What is the effective access time to access a
memory?
Memory (3)
Summary
so far …
• We have seen different types of memory elements and
gone up the hierarchy of memory.
– We have developed RAM, ROM, Registers and looked at
Caches.
• Next: we will see how programs that we develop are
allocated memory.
• Some terminology:
– Process: any program in execution is called a process.
• E.g., A java program that you write is simply a program, unless you execute
it. During its execution it becomes a process.
• There can be multiple processes of the same program.
Memory (3)
Memory
Allocation
• Two problems in memory allocation.
– Processes must not have conflicts in memory
they use. (must have separate memory).
• E.g., you wouldn’t want memory used by your process to
be overwritten by that of another user.
• Solution: Relocatable Code
– Small amount of physical memory (RAM) must be
used to execute a large number of processes
– (or execute large processes each of which
exceed the size of RAM)
– Solution: Virtual Memory.
Memory (3)
Relocatable code
and
Virtual Memory
• Relocatable assembly code: processes must be able to reside
in any portion of the physical memory.
– Why ? Consider an operating system, where all executions of
Microsoft Word must only reside at address 0x800000. What
would happen ?
– Hence, compiled and assembled programs do not have fixed
addresses allocated to them.
 Virtual memory: use hard disk as an extension for RAM.
Intuition: processes are allocated memory in a special
portion of the hard disk. They are loaded into RAM only
when they are executing on the CPU
Memory (3)
Relocatable
code
• Program binary code, such that the
addresses of the process address space
can be changed dynamically.
• I.e. it doesn’t map to main memory
• It has it’s own memory space
• Operating system replaces with what it
deems appropriate
Memory (3)
Relocatable
Code
#include <…>
…
int main() {
int x = 10;
return x;
}
Compiler + <main+0>: push %ebp BINARY CODE
assembler
0x0804867c <main+0>:
0x0804867d <main+1>:
0x0804867f <main+3>:
…
0x0804869c <main+32>:
0x0804869d <main+33>:
Memory (3)
<main+1>:
<main+3>:
mov %esp,%ebp
sub $0x8,%esp
…
<main+32>: ret
<main+33>: nop
push %ebp
mov %esp,%ebp
sub $0x8,%esp
ret
nop
Compiler generates relocatable code:
e.g., return instruction is 32 bytes away
from beginning of code segment
When process is created. The process
is given addresses by OS Address
called as logical address
Address binding: binding each
logical address to a location in
physical memory
Relocatable
Code
#include <…>
…
int main() {
int x = 10;
return x;
}
Compiler + <main+0>: push %ebp BINARY CODE
assembler
0x0804867c <main+0>:
0x0804867d <main+1>:
0x0804867f <main+3>:
…
0x0804869c <main+32>:
0x0804869d <main+33>:
Memory (3)
<main+1>:
<main+3>:
mov %esp,%ebp
sub $0x8,%esp
…
<main+32>: ret
<main+33>: nop
push %ebp
mov %esp,%ebp
sub $0x8,%esp
ret
nop
Compiler generates relocatable code:
e.g., return instruction is 32 bytes away
from beginning of code segment
When process is created. The process
is given addresses by CPU. Address
called as logical address
Address binding: binding each
logical address to a location in
physical memory
Using a HD
as memory
© Image from Silberschatz and Galvin
Memory (3)
Three types of memory addresses
C program
Binary code
(relocatable addresses)\
Aka linear addresses
Compiler/Assembler
Three types of addresses: linear address (addresses in
compiled binaries), logical address (address in the virtual
memory) and physical address (address in the RAM
when a program is loaded).
Logical Address
Bound to Physical memory
address
Loader
When
program is
Executed.
CPU generated Logical Addresses
Also called Virtual address
Dynamic Relocation using relocation register: mapping logical to physical
address.
© Image from Silberschatz and Galvin
Virtual
Memory (2)
• When a process is executed, the OS allocates memory for the
process on the disk (usually 4 GB of memory is allocated).
– This includes a code segment, data segment and program stack.
– In UNIX, we usually call this portion of the disk as swap partition.
• Advantage: Process memory is not tied to the amount of RAM.
• However, for a CPU to execute a process, the process must be in
the physical memory.
– Hence, OSes provide methods to map process address space on
virtual memory (called logical memory) to physical memory
(memory addresses in RAM).
Memory (3)
Summary
• Cache access
• Virtual memory
Memory (3)