valgrind and matmul

Valgrind thing

Worksheet: look at inside loop for stride-1 vs stride-n

Valgrind:  compile and time program, then

$ valgrind --tool=cachegrind ./matmul 100

this prints the output to a file called cachegrind.out.##### (the pid)
so then

$ cg_annotate cachegrind.out.#####

this will make a dope ass table and we want to look at the value under D1mr in the row for function matmul_a

Cool and good