GitHub - akritibhat/malloc-library: Malloc Library for multi threaded programs using buddy allocation technique

akritibhat / malloc-library Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Malloc Library for multi threaded programs using buddy allocation technique

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
.malloc.c.swp		.malloc.c.swp
Makefile		Makefile
MyNode.h		MyNode.h
README		README
arenaliblist.c		arenaliblist.c
arr.c		arr.c
calloc.c		calloc.c
freearena.c		freearena.c
libmalloc.so		libmalloc.so
mallib.c		mallib.c
mallib.o		mallib.o
mallocarena.c		mallocarena.c
mallocarenalib.c		mallocarenalib.c
memalign.c		memalign.c
posixmemalign.c		posixmemalign.c
realloc.c		realloc.c
reallocarray.c		reallocarray.c
t-test1		t-test1
t-test1.c		t-test1.c
t-test1.o		t-test1.o
test		test
tt1		tt1

Repository files navigation

Data Structure Used: Linked List

Created a linked list with Node : MyNode
MyNode contains the meta data about memory blocks
  1.freeMem- indicator if the memory is free
	2. size- size of the block
	3. occupied- memory used from the block (size-occupied gives us internal fragmentation)
	4. *sb_pointer- address of the memory block
	5. *next - pointer to the next node

Implemented buddy allocation.

When first malloc request is made , Arenas are initialised which are equal to the number of cores.
This makes it faster than HW3.

Used structure ArenaInfo to maintain Arena Data.

ArenaInfo {
   int numberOfArena; 
   int mallocReq;  
   int freeReq;  
   int arenaSize;     
   int usedBlocks;	
   int freeBlocks;
   struct arenaInfo *next;	
   pthread_mutex_t arrlock;
   struct MyNode* ghead;
   struct MyNode* endNode;
};

If arena is already initialised for a thread, using __thread, we make that arena available to the thread.
Arena locks are used to avoid collision in same arena. 
Locks are also placed around sbrk.