Cache learning

1
Not only cache It is divided into instructions and data
mmu Also divided into instructions Follow data

2
cache Is based on The principle of program locality
Reference resources ： What is the principle of program locality

3 canche Usage flow
a In limine cache Read a Address data , No desired data , It's called cache miss
b cpu Send the address to sdram, Read into one cacheline be called cache fill
c The program reads the address again a The data of , from cache Up lookup a The data of , Successfully found , It's called cache hit

4 cache line
Computer cache Cache as well as Cache Line Detailed explanation

5 write buffer yes cache Reverse operation .

6 cache Follow mmu equally , Focused on the cpu Inside , The address is not on the system bus .

7 cache Capacity in 1k To 512k Word is the optimal solution .

8 cache Solutions to consistency problems are divided into software and chip methods
Software solution ：
a Set the storage area where consistency problems may occur to non cache, But it may cause system performance degradation .
b In critical code areas where consistency problems may occur , Insert cache Control command , Achieve consistency