m




 

 


 




5.4. STORAGE MANAGEMENT
— Different features in a language causes different
storage management techniques to be used.
J ?ORTRAN: no recursive calls, no dynamic storage
management.
J Pascal: stack-based storage management.
J LISP: garbage collection.
Language implementers decide about the details.
Programmers don·t know about it.

0
MAJOR RUN-TIME ELEMENTS REQUIRING
STORAGE
— Data
— operations

^
DATA AND PROGRAM REQUIRING STORAGE
— Code segments for translated user programs.
— System run-time programs.
J Supporting user programs.
J Like library routines, software interpreters or
translator, storage management routines.
— User-defined data structures and constants.
— Subprogram return points.

R
— Referencing environments.
J Identifier associations (LISP A-list)
— Temporaries in expression evaluation.
J Recursive function calls make it a lot.
— Temporaries in parameter transmission.
J Resulting values for evaluation of actual parameters
are stored in temporaries until the total evaluation is
completed.

@
— Input-output buffers.
J Temporary storage areas used between the time of
the actual physical transfer of the data to or from
external storage and the program-initiated input /
output operation.
— Miscellaneous system data.
J System data like tables, status information for input-
output, ...

Ñ
MAJOR OPERATIONS REQUIRING STORAGE
— Subprogram call and return operations.
J Activation record,
J local referencing environment, «

— Data structure creation and destruction

— Component insertion and deletion operations.

ÿ
PROGRAMMER- AND SYSTEM-
CONTROLLED STORAGE MANAGEMENT
— Programmer control of storage management
J place a large and often undesirable burden on the
programmer,
J may interfere with the necessary system-controlled
storage management.
— Programmer can cause dangling references and
garbage.


— èhat language?
J Protection for the programmer by using a language
with strong typing and effective storage management
features,
J decrease in performance
OR
J performance
J more risk in having errors and fail during execution.

D
STORAGE MANAGEMENT PHASES
— Initial allocation
— Recovery

— Compaction and reuse

r
STATIC STORAGE MANAGEMENT
— Simplest
— static allocation

— no run-time storage management

— no concern for recovery and reuse

— efficient

— in COBOL and ?ORTRAN

rr
— In ?ORTRAN
J each subprogram is compiled separately,
J the code segment includes an activation record
— compiled program,
— its data areas,

— return point location,

— miscellaneous items of system data.

r0
STACK-BASED STORAGE MANAGEMENT
— Simplest run-time storage management
technique.
— Based on the nested last in first out structure in
subprograms calls and returns.
— Automatic compaction.

— In Pascal : a single central stack of activation

r^
HEAP STORAGE MANAGEMENT: ?IXED-
SIZE ELEMENTS
— A heap is a block of storage within which pieces
are allocated and freed in some relatively
unstructured manner.
— Need for heap , when a language permits storage
to be allocated and freed at execution time.
— ?ixed size elements allocated => no need for
compaction.

rR
RECOVERY
The problem: identification of reusable element,
solutions:
— Explicit return by programmer or system.
J Natural, but cause garbage and dangling reference.
P.226
— Reference counts.
J Cost of maintaining. P.227
J popular with parallel processing systems.

— Garbage collection.

r@
GARBAGE COLLECTION
— Dangling references more dangorous
— Two stages
J Mark
— garbage collection bit, set off if it is active.
J Sweep
— links the ´onµ elements to the free list.
èhen is a heap element active?
— There is a pointer to it from
— outside the heap

— another active heap element

rÑ
— Three critical assumptions
J any active element must be reachable by a chain of
pointers beginning outside the heap.
J It must be possible to identify every pointer outside
the heap that points to an element inside the heap.
J It must be possible to identify within any active heap
element the fields that contain pointers to other heap
elements.
P. 231

rÿ
HEAP STORAGE MANAGEMENT:
VARIABLE-SIZE ELEMENTS
— More difficult
— if space for programmer defined data structures
is sequential, like arrays or activation records.
— Major difficulty : reuse of recovered space.

r
— Initial allocation and reuse.
— reuse directly from a free-space list.
J ?irst-fit method
J best-fit method
keeping free-space list in size order.
— Recovery with variable-size blocks.
J In first word of each block: a length indicator.
— Compaction and memory fragmentation problem.

rD
— Compaction approaches:
J Partial compaction
— only adjacent free blocks
J ?ull compaction
— active blocks may be shifted

0
  


—  ! 
—  !!"! 
— ! ! 
!! 
— #!$ ! 
!! 

 %
  

—m



 




— 


 


— 

 




 
—  



 
—  



 %
  
— §


   
— §





 
—     


—  
 

—  




 
 
 
 

   




— 



 



  
 
—m  
 

 

 
 

   

m    

M

 
M
  

 


 



M
     









M 
 



 


 

 m  



  
  


  

 

   

   

 



m  m &
#  



6










 
   

  


#

m'
#  



§!




 
   

 









 
! 

"  
! 

HEAP COMPACTION

  

 

 
HEAP COMPACTION