This is the code that generated the various matrix multiplication 
performance charts in Introduction to Computer Systems.

Files:
mm.{c,h}        - regular matmult
bmm.c           - blocked matmult        
clock.{c,h}	- routines for accessing the Pentium cycle timers
fcycmm.{c,h}    - routines that use the K-best scheme for estimating the
                  number of cycles required by a function with 4 args.

To compile:
	linux> make clean
	linux> make

To run:
	linux> mm
	linux> bmm