forked from Github_Repos/cvw
1: It simplifies the interactions between the caches and the hptw. 2: instruction page faults are fetched 3 times, caching them in the ITLB speeds up this process. There are two downsides. 1: Pollute the TLBs with not very relavent translations 2: Have to compute the misalignment. This can be cached in the TLB which only costs 1 flip flop for each TLB line. |
||
|---|---|---|
| .. | ||
| cache | ||
| ebu | ||
| fpu | ||
| generic | ||
| hazard | ||
| ieu | ||
| ifu | ||
| lsu | ||
| mmu | ||
| muldiv | ||
| privileged | ||
| sdc | ||
| uncore | ||
| wally | ||