EE 3755 Homework 4 Due: 12 November 2001
See http://www.ece.lsu.edu/ee3755/ref.html for documentation on MIPS and the SPIM
MIPS simulator. Alternate instructions can be found in Appendix A of the Patterson & Hennessy
text. The links are clickable when this assignment is viewed with Acrobat Reader.
Copy files /home/classes/ee3755/com/s/hw04.s (the solution template) and
/home/classes/ee3755/com/s/hw04p0.s into a subdirectory named hw in your class account. File
hw04p0.s contains instructions for running the SPIM simulator. When these instructions are
followed the SPIM simulator is run with the -notrap and -delayed_branches switches, and a
version of the simulator implementing the clz instruction is used. If you choose to use another
installation of the simulator be sure to start it using those switches and replace the clz instructions
in the solution template with a call to the completed countlz routine (see Problem 1).
Problem 0: This solution to this problem will not be collected graded, but do it anyway! File
hw04p0.s contains a buggy MIPS procedure, pop, that’s supposed to find the population of an
integer. (An integer’s population is the number of 1’s in its binary representation.) The file
also contains a startup routine that runs pop on three different integers, moving the incorrectly
computed population into registers s0,s1,ands2.
(a) Following the instructions in the file, run the code and verify that it doesn’t work.
(b) Debug the pop count module.
Problem 1: Write a MIPS assembly language procedure that counts the number of leading zeros
of the value in a0 and puts that count into v0. (Remember that MIPS registers are 32 bits.) For
example, if 1 (written another way 000000000000000000000000000000012)isina0 then 31 should
be put in v0.Ifffff16 is in a0 then 16 should be put in v0. Put the code in the appropriate place
in the solution template after the line labeled countlz in hw04.s. The template is set up with a
routine to test countlz. Do not use the clz instruction.
Problem 2: Write a MIPS assembly language procedure that converts a 32-bit signed integer
in register a0 into an IEEE 754 single-precision floating point number and puts it in v0. The
procedure must round the integer towards zero (for most solutions, this is equivalent to not doing
any rounding at all). Put the code in the appropriate place in the solution template: after the
line with label itos in hw04.s. The template is set up with code to test pops. Do not use MIPS’
floating-point instructions.
1