Assignment 5: Allocation Lab (due on Tue, Apr 23, 2024 at 11:59pm)

Introduction

In this lab, you will implement a dynamic storage allocator for C programs, i.e., your own version of the malloc, free and realloc routines! You are encouraged to explore the design space creatively and implement an allocator that is correct, efficient and fast.

What to Implement

You will find all the files needed for this project inside a new GitHub repository shared with you.

  • You need to implement the functions in mm_block.c, mm_list.c, mm.c.
  • We recommend implementing the functions in this order, to simplify unit testing (mm.c depends on the other units).
  • Check the instructions in README.md on how to run/debug unit tests.
  • The mtest.c program allows you to evaluate the performance of your solution. To do that, you can run ./grade.

The public API of your dynamic storage allocator will consist of the following four functions, which are declared in mm.h and defined in mm.c.

 int  mm_init(void);
void *mm_malloc(size_t size);
void  mm_free(void *ptr);
void *mm_realloc(void *ptr, size_t size);
  • int mm_init(void): Before calling mm_malloc, mm_realloc or mm_free, the application program (i.e., the trace-driven testing program that we will use to evaluate your implementation) calls mm_init to perform any necessary initialization, such as allocating the initial heap area. The return value should be -1 if there was a problem in performing the initialization, 0 otherwise.
  • void *mm_malloc(size_t size): The mm_malloc routine returns a pointer to an allocated block payload of at least size bytes. The entire allocated block should lie within the heap region and should not overlap with any other allocated chunk. In addition, your mm_malloc implementation should always return 8-byte aligned pointers (similarly to libc’s malloc).
  • void mm_free(void *ptr): The mm_free routine frees the block pointed by ptr. This routine is only guaranteed to work when the passed pointer was returned by an earlier call to mm_malloc or mm_realloc and has not yet been freed.
  • void *mm_realloc(void *ptr, size_t size): The mm_realloc routine returns a pointer to an allocated region of at least size bytes with the following constraints.
    • If ptr is NULL, the call is equivalent to mm_malloc(size);
    • If size is equal to zero, the call is equivalent to mm_free(ptr);

    • If ptr is not NULL, it must have been returned by an earlier call to mm_malloc or mm_realloc. The call to mm_realloc changes the size of the memory block pointed to by ptr (the old block) to size bytes and returns the address of the new block. Note that the address of the new block might be the same as the old block, or it might be different, depending on your implementation, on the amount of internal fragmentation in the old block, and on the size of the realloc request.

      The contents of the new block are the same as those of the old ptr block, up to the minimum of the old and new sizes. Everything else is uninitialized. For example, if the old block is 8 bytes and the new block is 12 bytes, then the first 8 bytes of the new block are identical to the first 8 bytes of the old block and the last 4 bytes are uninitialized. Similarly, if the old block is 8 bytes and the new block is 4 bytes, then the contents of the new block are identical to the first 4 bytes of the old block.

These semantics match the semantics of the corresponding malloc, realloc, and free routines of libc.

The included mm_block.c, mm_list.c, mm.c files define the internal API of a solution using explicit free lists. The unit tests test/test_mm_block.c, test/test_mm.c, test/test_mm_list.c check your implementation of this internal API on simple scenarios.

Of course, you can change this internal API to use any data structure of your choice to manage your heap! For example, instead of one explicit free list, you could use segregated lists, or even balanced trees :-)

Heap Consistency Checker

Dynamic memory allocators are notoriously tricky to program correctly and efficiently. They are difficult to program correctly because they involve a lot of untyped pointer manipulation. You will find it very helpful to write a heap checker that scans the heap and checks it for consistency.

Some examples of what a heap checker might check are:

  • Is every block in the free list marked as free?
  • Are there any contiguous free blocks that somehow escaped coalescing?
  • Is every free block actually in the free list?
  • Do the pointers in the free list point to valid free blocks?
  • Do any allocated blocks overlap?
  • Do the pointers in a heap block point to valid heap addresses?

Your heap checker will consist of the function int mm_check(void) in mm.c. It will check any invariants or consistency conditions you consider prudent. It returns a nonzero value if and only if your heap is consistent. You are not limited to the listed suggestions nor are you required to check all of them. You are encouraged to print out error messages when mm_check fails.

This consistency checker is for your own debugging during development. When you submit mm.c, make sure to remove any calls to mm_check as they will slow down your throughput. Even better, you can guard calls to mm_check using #ifdef DEBUG … #endif.

Support Routines

The memlib.c package simulates the memory system for your dynamic memory allocator. You can invoke the following functions in memlib.c:

  • void *mem_sbrk(int incr): Expands the heap by incr bytes, where incr is a positive non-zero integer and returns a generic pointer to the first byte of the newly allocated heap area. The semantics are identical to the Unix sbrk function, except that mem_sbrk accepts only a positive non-zero integer argument.
  • void *mem_heap_lo(void): Returns a generic pointer to the first byte in the heap.
  • void *mem_heap_hi(void): Returns a generic pointer to the last byte in the heap.
  • size_t mem_heapsize(void): Returns the current size of the heap in bytes.

The Trace-driven Testing Program

The testing program mtest.c in your repository tests your mm.c package for correctness, space utilization, and throughput. The testing program runs a set of trace files that are included in the traces folder. Each trace file contains a sequence of allocate, reallocate, and free directions that instruct the mtest program to call your mm_malloc, mm_realloc, and mm_free routines in some sequence. The mtest program and the trace files are the same ones we will use when we grade your submission.

The mtest.c program accepts the following command-line arguments:

  • -f <tracefile>: Use one particular tracefile for testing instead of the default set of trace files.
  • -r <num>: Repeat throughput measurements <num> times (3 by default).
  • -h: Print a summary of the command line arguments.

Programming Rules

  • You should not change any of the public interfaces in mm.c (i.e., the non-static functions).
  • You should not invoke any memory-management related library calls or system calls. This excludes the use of malloc, calloc, free, realloc, sbrk, brk or any variants of these calls in your code.
  • As a consequence, variable-size data structures should be allocated in the heap itself.
  • Like the libc malloc package, your allocator must always return pointers that are aligned at 8-byte boundaries. The grading program mtest will enforce.

Evaluation

You will receive zero points if you break any of the rules or your code is buggy and crashes mtest. We will measure the performance of your solution through:

  • Space utilization $U$: The peak ratio between the aggregate amount of memory used by mtest (i.e., allocated via mm_malloc or mm_realloc but not yet freed via mm_free) and the size of the heap used by your allocator. The optimal ratio equals to 1. You should find good policies to minimize fragmentation in order to make this ratio as close as possible to the optimal one (analyze the traces!).
  • Throughput $T$: The average number of operations completed per second.

The mtest program summarizes the performance of your allocator by computing a performance index $P$, which is a weighted sum of the space utilization and throughput:

\[P = w\min\left(1, \frac{U}{0.95}\right) + (1-w) \min\left(1, \frac{T}{0.9\times T_{libc}}\right)\]

where $U$ is your space utilization, $T$ is your throughput, and $T_{libc}$ is the throughput of libc’s malloc on the testing traces. The performance index favors space utilization over throughput, with a weighting factor of $w = 0.6$.

Observing that both memory and CPU cycles are expensive system resources, we adopt this formula to encourage balanced optimization of both memory utilization and throughput. Ideally, the performance index will reach $P = w + (1-w) = 1$ or $100\%.$ Since each metric will contribute at most $w$ and $1-w$ to the performance index, respectively, you should not go to extremes to optimize either memory utilization or throughput only. To receive a good score, you must achieve a balance between utilization and throughput.

The mtest program will list the number of trace files your allocator passes and your performance index out of 100.

If you obtain utilization of $95\%$ and reach $90\%$ of libc’s malloc throughput, you will receive full credit for this assignment.

Handin Instructions

For on-time submissions, ensure that the version of your files that you want to be graded is pushed to GitHub. We will use push times to check for late days.

Hints

  • Use the mtest -f option. During initial development, using tiny trace files will simplify debugging and testing. We have included two such trace files (short{1,2}-bal.rep) that you can use for initial debugging.
  • Debug unit tests that fail. A debugger will help you isolate and identify out of bounds memory references. Check the instructions in README.md on how to debug unit tests.
  • Understand every line of given functions. The textbook also has a similar example of a simple allocator based on an implicit free list.
  • Do your implementation in stages. The first 9 traces contain requests to malloc and free. The last 2 traces contain requests for realloc, malloc, and free. We recommend that you start by getting your malloc and free routines working correctly and efficiently on the first 9 traces. Only then should you turn your attention to the realloc implementation. For starters, build realloc on top of your existing malloc and free implementations. But to get really good performance, you will need to build a standalone realloc.

And, most importantly… Start early! It is possible to write an efficient malloc package in about 450 lines of code. However, it will be some of the most difficult and sophisticated code you have written so far in your career. So, start early, and good luck!

Acknowledgements. This lab was initially developed by the authors of the course textbook and their staff. It has been customized for use by this course.