Princeton University
COS 217: Introduction to Programming Systems

Assignment 6: A Heap Manager Module


Purpose

The purpose of this assignment is to help you understand how dynamic memory management works in C. It also will give you more opportunity to use the GNU/Unix programming tools, especially bash, emacs, gcc217, gdb, and make.


Rules

You may work with one partner on this assignment. You need not work with a partner, but we strongly prefer that you do. You must notify your preceptor of who you will be working with, no later than Sunday 4/17

. Your preceptor will help you find a partner if you cannot find one yourself.

If you work with a partner, then only one of the partners should submit files. Your readme file, your makefile, and your source code files should contain your name and your partner's name.

If you are an undergraduate student, then your partner should be another undergraduate student. If you are a graduate student, then your partner should be another graduate student.

Your partner should be from your precept. You may work with a partner from another precept only in the case of extraordinary circumstances, and with the permission of the pertinent preceptors. If you wish to work with a partner from another precept, then you must obtain permission from the pertinent preceptors no later than Sunday 4/17.

"Checking invariants" (as described below) is the "on your own" part of this assignment. That is, when doing that part you may consult with your partner, but must not consult with the course instructors, lab TAs, listserv, etc., except perhaps to clarify requirements. That part is worth 15% of this assignment.


Background

A standard C programming environment contains four functions that allow management of the runtime heap: malloc, free, calloc, and realloc. Those heap management functions are used heavily in many C programs.

Section 8.7 of the book The C Programming Language (Kernighan and Ritchie) shows an implementation of the malloc and free functions. That book section is on electronic reserve at Princeton's library; you can access it through Princeton's Blackboard system (http://blackboard.princeton.edu) by selecting the COS 217 course and clicking on E-Reserves. The key data structure in that implementation is a circular singly-linked list; each free memory "chunk" is stored in that list. Each memory chunk contains a header which specifies its size and, if free, the address of the next chunk in the list. Although elegant in its simplicity, that implementation can be inefficient.

The web page http://gee.cs.oswego.edu/dl/html/malloc.html (Doug Lea) describes how one can enhance such an implementation so it is more efficient. The key data structure is an array of non-circular doubly-linked lists, that is, an array of "bins." Each bin contains all free chunks of a prescribed size. The use of multiple bins instead of a single linked list allows malloc to be more efficient.

Moreover, each memory chunk contains both a header and a footer. The header contains three fields: the size of the chunk, an indication of whether the chunk is free, and, if free, a pointer to the next free chunk in its bin. The footer contains two fields: the size of the chunk, and, if free, a pointer to the previous free chunk in its bin. That chunk structure allows free to be more efficient.

A more thorough description of the pertinent data structures and algorithms will be provided in lectures and precepts.


Your Task

You are given the interface of a HeapMgr (heap manager) module in a file named heapmgr.h. It declares two functions:

void *HeapMgr_malloc(size_t uiSize);
void  HeapMgr_free(void *pv);

You also are given three implementations of the HeapMgr module:

Your task is to create two additional implementations of the HeapMgr module. Your first implementation, heapmgr1.c, should enhance heapmgrbase.c so it is reasonably efficient. To do that it should use a single doubly-linked list and chunks that contains headers and footers (as described above, in lectures, and in precepts).

If designed properly, heapmgr1.c will be reasonably efficient in most cases. However, heapmgr1.c is subject to poor worst-case behavior. Your second implementation, heapmgr2.c, should enhance heapmgr1.c so the worst-case behavior is not poor. To do that it should use multiple doubly-linked lists, alias bins (as described above, in lectures, and in precepts).

Your HeapMgr implementations should not call the standard malloc, free, calloc, or realloc functions.

Your HeapMgr implementations should thoroughly validate function parameters by calling the standard assert macro.

Your HeapMgr implementations should check invariants by:


Logistics

Develop on hats, using emacs to create source code and gdb to debug.

The directory /u/cos217/Assignment6 contains files that you will find useful:

The testheapmgr program requires three command-line arguments. The first should be any one of seven strings, as shown in the following table, indicating which of seven tests the program should run:

Argument Test Performed
LifoFixed LIFO with fixed size chunks
FifoFixed FIFO with fixed size chunks
LifoRandom LIFO with random size chunks
FifoRandom FIFO with random size chunks
RandomFixed Random order with fixed size chunks
RandomRandom Random order with random size chunks
Worst Worst case order for a heap manager implemented using a single linked list

The second command-line argument is the number of calls of HeapMgr_malloc and HeapMgr_free that the program should execute. The third command-line argument is the (maximum) size, in bytes, of each memory chunk that the program should allocate and free.

Immediately before termination testheapmgr prints to stdout an indication of how much CPU time and heap memory it consumed. See the testheapmgr.c file for more details.

To test your HeapMgr implementations, you should build two programs using these gcc217 commands:

gcc217 testheapmgr.c heapmgr1.c chunk.c -o testheapmgr1
gcc217 testheapmgr.c heapmgr2.c chunk.c -o testheapmgr2

To collect timing statistics, you should build five programs using these gcc217 commands:

gcc217 -O3 -D NDEBUG testheapmgr.c heapmgrgnu.c -o testheapmgrgnu
gcc217 -O3 -D NDEBUG testheapmgr.c heapmgrkr.c -o testheapmgrkr
gcc217 -O3 -D NDEBUG testheapmgr.c heapmgrbase.c chunkbase.c -o testheapmgrbase
gcc217 -O3 -D NDEBUG testheapmgr.c heapmgr1.c chunk.c -o testheapmgr1
gcc217 -O3 -D NDEBUG testheapmgr.c heapmgr2.c chunk.c -o testheapmgr2

The -O3 (that's uppercase "oh", followed by the number "3") argument commands gcc to optimize the machine language code that it produces. When given the -O3 argument, gcc spends more time compiling your code so, subsequently, the computer spends less time executing your code. The -D NDEBUG argument commands gcc to define the NDEBUG macro, just as if the preprocessor directive #define NDEBUG appeared in the specified .c file(s). Defining the NDEBUG macro disables the calls of the assert macro within the HeapMgr implementations. Doing so also disables code within testheapmgr.c that performs (very time consuming) checks of memory contents.

Create additional test programs as you deem necessary. You need not submit your additional test programs.

Create a makefile. The first dependency rule of the makefile should build five executable files: testheapmgrgnu, testheapmgrkr, testheapmgrbase, testheapmgr1, and testheapmgr2. That is, the first dependency rule of your makefile should be:

all: testheapmgrgnu testheapmgrkr testheapmgrbase testheapmgr1 testheapmgr2

The makefile that you submit should:

We recommend that you create your makefile early in your development process. Doing so will allow you to use and test your makefile during development.

Create a readme text file that contains:

Submit your work electronically on hats via the command:

submit 6 heapmgr1.c heapmgr2.c readme makefile

Grading

We will grade your work on quality from the user's point of view and from the programmer's point of view. To encourage good coding practices, we will deduct points if gcc217 generates warning messages. We also will deduct points if splint generates warning messages that are not explained in your readme file. But see the next section of this document regarding splint warnings.

From the user's point of view, your module has quality if it behaves as it should. The correct behavior of the HeapMgr module is defined by the previous sections of this assignment specification. From the programmer's point of view, your module has quality if it is well styled and thereby simple to maintain. See the specifications of previous assignments for guidelines concerning style. Specifically, function modularity will a prominent part of your grade.


Splint Warnings

splint generates these warnings when it analyzes the testheapmgr.c file:

testheapmgr.c: (in function main)
testheapmgr.c:118:21: Unrecognized identifier: sbrk
  Identifier used in code has not been declared. (Use -unrecog to inhibit
  warning)
testheapmgr.c: (in function setCpuLimit)
testheapmgr.c:223:4: Variable sRlimit used before definition
  An rvalue is used that may not be initialized to a value on some execution
  path. (Use -usedef to inhibit warning)
testheapmgr.c:225:4: Unrecognized identifier: setrlimit
testheapmgr.c:225:14: Unrecognized identifier: RLIMIT_CPU

Don't be concerned about those warnings; you need not explain them in your readme file. splint generates them because it ignores the #include of the unistd.h file (and all such Unix-specific header files). The setrlimit and sbrk functions are declared in that file, and the struct rlimit type and RLIMIT_CPU constant are defined in that file.

Similarly, splint may generate a warning for each call of brk or sbrk in your your heapmgr1.c and heapmgr2.c files. Don't be concerned about those messages. You need not explain them in your readme file.