Hardware Cache (sRAM)

Introduced due to clock speed discrepancy in DRAM (dynamic RAM main memory) and CPU clock
Based on the locality principle
- It states due to cyclic structure of programs and data packed into linear arrays addresses close by have a high probability of being used near in the future
Thus the concept of static RAM (sRAM) is implemented as on chip memory
The cache is subdivided into subsets of lines. There are several ways of mapping
- Direct mapping. A line in main memory is stored in exact same location in cache
- Fully associative. A line in main memory can be stored in any location in cache
- N-Way Associative. A line in main memory can be stored in N number of places on cache
CD flag of cr0 processor register determines whether caching is enabled (1) or not
NW flag of cr0 determines write-through or write-back
Pentium cache allows each page frame to have its own cache management policy
- Therefore the page directory and page table entry has 2 extra flags
  - PCD: Page cache disable
  - PWT: Page write-through
- Linux enables page cache always and always uses write-back

Cache Unit: Two Parts

Cache Controller: Essentially the Page directory of cache. Contains a few flags and a tag. Which is a way of representing the requested memory location (called a tag).
- Data stored comes in 3 parts in order tag, cache controller subset index, offset within the line
sRAM (hardware cache): The actual data stored

Cache Process

CPU attempts to access RAM. CPU extracts subset index from physical address.
CPU compares the tags of all lines in the subset with the high order bits of physical address
If a line with the same tag is found CPU has a cache hit (other wise cache miss)

Cache Hits and Misses

Cache Hit: When searching for a portion of RAM the CPU finds it in the cache.
- For reads, the data is extracted from sRAM and put in a register dRAM is not accessed
- For write the CPU uses one of two strategies write-through and write-back
  - Write-through: The controller always writes into both RAM and the cache line, effectively switching off the cache for write operations
  - Write-back: More immediate efficiency, only the cache line is updated and the contents of the RAM are left unchanged. RAM is only written to when the CPU executes a FLUSH command usually after cache misses
Cache Miss: When the CPU fails to do the above
- Cache line will be written into RAM and if neccesary correct line will be fetched from main memory.

Multiprocessor System Cache

Each CPU has its own cache
If two or more CPUs have the same memory on cache an update of shared cache memory must be updated on all CPUs with that memory
- This is done using cache snooping

Levels of Cache

L1 is the fastest cache and often the smallest L2, L3 and onwards are slower and often larger than L1
Linux assumes only one cache