1
This directory contains mpn functions optimized for MIPS3. Example of
2
processors that implement MIPS3 are R4000, R4400, R4600, R4700, and R8000.
4
RELEVANT OPTIMIZATION ISSUES
6
1. On the R4000 and R4400, branches, both the plain and the "likely" ones,
7
take 3 cycles to execute. (The fastest possible loop will take 4 cycles,
8
because of the delay insn.)
10
On the R4600, branches takes a single cycle
12
On the R8000, branches often take no noticable cycles, as they are
13
executed in a separate function unit..
15
2. The R4000 and R4400 have a load latency of 4 cycles.
17
3. On the R4000 and R4400, multiplies take a data-dependent number of
18
cycles, contrary to the SGI documentation. There seem to be 3 or 4