---Dinero III by Mark D. Hill.
---Version 3.3, Released May 1989.
---Execution begins.
CMDLINE: dinero -b16 -i8k -d8k
CACHE (bytes): blocksize=16, sub-blocksize=0, Usize=0, Dsize=8192, Isize=8192.
POLICIES: assoc=1-way, replacement=l, fetch=d(1,0), write=c, allocate=w.
CTRL: debug=0, output=0, skipcount=0, maxcount=10000000, Q=0.
---Simulation begins.
---Simulation complete.
Metrics Access Type:
(totals,fraction) Total Instrn Data Read Write Misc
----------------- ------ ------ ------ ------ ------ ------
Demand Fetches 0 0 0 0 0 0
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Demand Misses 0 0 0 0 0 0
0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Words From Memory 0
( / Demand Fetches) 0.0000
Words Copied-Back 0
( / Demand Writes) 0.0000
Total Traffic (words) 0
( / Demand Fetches) 0.0000
---Execution complete.
The six columns of numbers in the upper portion of the output represent the absolute counts and relative frequencies of accesses and misses broken down into access type. There are two basic access types: Instrn, meaning instruction references, and Data meaning of course data references. The sum of these two is shown in the first column, labelled Total. The data references are further broken down into Read and Write references in columns four and five respectively. Column six counts any references that are not recognised, and can be safely ignored at the moment.
There are four rows of numbers. Rows 1 and 3 represent absolute event counts, whereas rows 2 and 4 represent relative frequencies. Thus, the relative instruction and data frequencies in row 2 ought to add up to 1. Rows 1 and 2 refer to the accesses made by the processor, and rows 3 and 4 refer to the caches misses which resulted from those accesses. Consequently, column 3 from row four is the total data misses and column 3 from fow 2 is the total data accesses. Naturally, the data miss rate is the former divided by the latter.