SCIENTIFIC PROGRAMMING

Editor: Paul F. Dubois, [email protected]

THE INSIDE STORY ON

SHARED LIBRARIES AND DYNAMIC LOADING
By David M. Beazley, Brian D. Ward, and Ian R. Cooke

RADITIONALLY, DEVELOPERS HAVE BUILT

SCIENTIFIC SOFTWARE AS STAND-ALONE

APPLICATIONS WRITTEN IN A SINGLE LANGUAGE

machine code instructions of the program, a data section with

the global variable x, and a read-only section with the
string literal Hello World, x = %d\n. Additionally, the
object le contains a symbol table for all the identiers that
appear in the source code. An easy way to view the symbol
table is with the Unix command nmfor example,

SUCH AS FORTRAN, C, OR C++. HOWEVER, MANY

scientists are starting to build their applications as extensions

to scripting language interpreters or component frameworks.
This often involves shared libraries and dynamically loadable modules. However, the inner workings of shared libraries and dynamic loading are some of the least understood
and most mysterious areas of software development.
In this installment of Scientic Programming, we tour the
inner workings of linkers, shared libraries, and dynamically
loadable extension modules. Rather than simply providing a
tutorial on creating shared libraries on different platforms, we
want to provide an overview of how shared libraries work and
how to use them to build extensible systems. For illustration,
we use a few examples in C/C++ using the gcc compiler on
GNU-Linux-i386. However, the concepts generally apply to
other programming languages and operating systems.
Compilers and object files

When you build a program, the compiler converts source

files to object files. Each object file contains the machine
code instructions corresponding to the statements and declarations in the source program. However, closer examination reveals that object les are broken into a collection
of sections corresponding to different parts of the source
program. For example, the C program
#include <stdio.h>
int x = 42;
int main() {
printf(Hello World, x = %d\n, x);
}

produces an object file that contains a text section with the

90

$ nm hello.o
00000000 T main
U printf
00000000 D x

For symbols such as x and main, the symbol table simply

contains an offset indicating the symbols position relative
to the beginning of its corresponding section (in this case,
main is the rst function in the text section, and x is the rst
variable in the data section). For other symbols such as
printf, the symbol is marked as undened, meaning that
it was used but not dened in the source program.
Linkers and linking

To build an executable file, the linker (for example, ld)

collects object les and libraries. The linkers primary function is to bind symbolic names to memory addresses. To do
this, it rst scans the object les and concatenates the object
le sections to form one large le (the text sections of all object les are concatenated, the data sections are concatenated,
and so on). Then, it makes a second pass on the resulting le
to bind symbol names to real memory addresses. To complete the second pass, each object le contains a relocation
list, which contains symbol names and offsets within the object le that must be patched. For example, the relocation list
for the earlier example looks something like this:
$ objdump -r hello.o
hello.o: le format elf32-i386
RELOCATION RECORDS FOR [.text]:
OFFSET
TYPE
VALUE
0000000a
R_386_32
x

COMPUTING IN SCIENCE & ENGINEERING

undefined, the linker usually replaces its value with 0. So,

this technique can be a useful way for a program to invoke
optional code that does not require recompiling the entire
Static libraries
application (contrast this to enabling optional features with
To improve modularity and reusability, programming li- a preprocessor macro).
braries usually include commonly used functions. The traAlthough static libraries are easy to create and use, they
ditional library is an archive (.a le), created like this:
present a number of software maintenance and resource utilization problems. For example, when the linker includes a
$ ar cr libfoo.a foo.o bar.o spam.o...
static library in a program, it copies data from the library to
the target program. If patching the library is ever necessary,
The resulting libfoo.a le is known as
everything linked against that library must
a static library. An archives structure is
be rebuilt for the changes to take effect.
nothing more than a collection of raw obAlso, copying library contents into the tarMany compilers
ject les strung together along with a table
get program wastes disk space and memof contents for fast symbol access. (On
oryespecially for commonly used liprovide a pragma for
older systems, it is sometimes necessary to
braries such as the C library. For example,
manually construct the table of contents
if every program on a Unix machine indeclaring certain
using a utility such as the Unix ranlib
cluded its own copy of the C library, the
command.)
size of these programs would increase drasymbols as weak.
When a static library is included during
matically. Moreover, with a large number
program linking, the linker makes a pass
of active programs, a considerable amount
through the library and adds all the code
of system memory goes to storing these
and data corresponding to symbols used in the source pro- copies of library functions.
gram. The linker ignores unreferenced library symbols and
aborts with an error when it encounters a redened symbol. Shared libraries
An often-overlooked aspect of linking is that many compilTo address the maintenance and resource problems with staers provide a pragma for declaring certain symbols as weak. tic libraries, most modern systems now use shared libraries or
For example, the following code declares a function that the dynamic link libraries (DLLs). The primary difference between
linker will include only if its not already dened elsewhere.
static and shared libraries is that using shared libraries delays
the actual task of linking to runtime, where it is performed by
#pragma weak foo
a special dynamic linkerloader. So, a program and its libraries
/* Only included by linker if not already dened */
remain decoupled until the program actually runs.
void foo() {
Runtime linking allows easier library maintenance. For
...
instance, if a bug appears in a common library, such as the C
}
library, you can patch and update the library without reAlternatively, you can use the weak pragma to force the compiling or relinking any applicationsthey simply use
linker to ignore unresolved symbols. For example, if you the new library the next time they execute. A more subtle aswrite the program
pect of shared libraries is that they let the operating system
make a number of signicant memory optimizations. Specif#pragma weak debug
ically, because libraries mostly consist of executable instrucextern void debug(void);
tions and this code is normally not self-modifying, the opvoid (*debugfunc)(void) = debug;
erating system can arrange to place library code in read-only
int main() {
memory regions shared among processes (using page-sharprintf(Hello World\n);
if (debugfunc) (*debugfunc)();
ing and other virtual memory techniques). So, if hundreds
}
of programs are running and each program includes the
same library, the operating system can load a single shared
the program compiles and links whether or not debug() is copy of the librarys instructions into physical memory. This
actually dened in any object le. When the symbol remains reduces memory use and improves system performance.
00000010
00000015

R_386_32
R_386_PC32

SEPTEMBER/OCTOBER 2001

.rodata
printf

91

SCIENTIFIC PROGRAMMING

Cafe Dubois
The Times, They Are a Changin

Twenty years of schoolin and they put you on the day shift.
Bob Dylan
This summer marks my 25th year at Lawrence Livermore
National Laboratory, all of it on the day shift. LLNL is a
good place to work if you are someone like me who likes to
try new areas, because you can do it without moving to a
new company.
When my daughter was in the fifth grade, she came to
Take Your Daughter to Work Day, and afterwards told me,
referring to the system of community bicycles that you can
ride around on, The Lab is the greatest place in the world
to work. They have free bikes and the food at the cafeteria
is yummy! After that day she paid a lot of attention to her
math and science. Free bikes and yummy food is a lot of
motivation. Shes off to college this year, and I will miss her.
We technical types live in such a constant state of
change, and it is so hard to take the time to keep up. For
each of us, the time will come when we have learned our
last new thing, when we tell ourselves something is not
worth learning when the truth is we just cant take the pain
anymore. So, when I decide not to learn something these
days, I worry about my decision. Was that the one? Is it already too late?
Was it Java Beans? I sure hope it wasnt Java Beans. What
an ignominious end that would be.
F90 pointers
In my article on Fortran 90s space provisions, I didnt
have space to discuss pointers. One reader wrote me about
having performance problems allocating and deallocating
a lot of small objects. So, here is a simple small object
cache module that will give you the idea of how to use
pointers. In this module, one-dimensional objects of size N
or smaller can be allocated by handing out columns of a
fixed cache. The free slots are kept track of through a simple linked list. If the cache fills up, we go to the heap:
module soc
! Allocate memory of size <= N from a xed block.
private
public get, release, init_soc

On most systems, the static linker handles both static and

shared libraries. For example, consider a simple program
linked against a few different libraries:
$ gcc hello.c -lpthread -lm

If the libraries -lpthread and -lm have been compiled as

shared libraries (usually indicated by a .so sufx on the ac-

92

Paul in Paris, considering how life imitates art.

integer, parameter:: N=16, M=100

real, target:: cache(N, M)
integer::links(M), rst
contains
subroutine init_soc ()
integer i
do i = 1, M-1
links(i) = i + 1
enddo
links(M) = -1
rst = 1
end subroutine init_soc
function get(s)
integer, intent(in):: s
real, pointer:: get(:)
integer k
if (s > N) then
allocate(get(s))
return
endif
if (rst == -1) then
allocate(get(s))

tual library le), the static linker checks for unresolved symbols and reports errors as usual. However, rather than copying the contents of the libraries into the target executable,
the linker simply records the names of the libraries in a list
in the executable. You can view the contents of the library
dependency list with a command such as ldd:
ldd a.out

COMPUTING IN SCIENCE & ENGINEERING

return
endif
k = rst
rst = links(k)
get => cache(1:s, k)
return
end function get
subroutine release(x)
real, pointer:: x(:)
integer i
if (size(x) > N) then
deallocate(x)
return
endif
do i = 1, M
if (associated(x, cache(1:size(x), i))) then
links(i) = rst
rst = i
return
endif
enddo
deallocate(x)
end subroutine release
end module soc
program socexample
use soc
real, pointer:: x1(:), x2(:), x3(:)
integer i
call init_soc ()
x1 => get(3)
x2 => get(3)
x3 => get(20)
x3 = (/ (i/2., i=1, 20) /)
do i = 1, 3
x1(i) = i
x2(i) = -i
enddo
print *, x1+x2
print *, x3
call release(x2)
call release(x1)
call release(x3)
end program socexample
The input queue is low just now and Id love to hear from
authors about proposed articles. Just email me at paul@
pfdubois.com. And remember, if its Java Beans you want, it
aint me youre lookin for, babe.

libpthread.so.0 => /lib/libpthread.so.0 (0x40017000)

libm.so.6 => /lib/libm.so.6 (0x40028000)
libc.so.6 => /lib/libc.so.6 (0x40044000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)

When binding symbols at runtime, the dynamic linker

searches libraries in the same order as they were specied on
the link line and uses the first definition of the symbol en-

SEPTEMBER/OCTOBER 2001

countered. If more than one library happens to define the

same symbol, only the first definition applies. Duplicate
symbols normally dont occur, because the static linker scans
all the libraries and reports an error if duplicate symbols are
defined. However, duplicate symbol names might exist if
they are weakly defined, if an update to an existing shared
library introduces new names that conflict with other libraries, or if a setting of the LD_LIBRARY_PATH variable
subverts the load path (described later).
By default, many systems export all the globally defined
symbols in a library (anything accessible by using an extern specier in C/C++). However, on certain platforms,
the list of exported symbols is more tightly controlled with
export lists, special linker options, or compiler extensions.
When these extensions are required, the dynamic linker will
bind only to symbols that are explicitly exported. For example, on Windows, exported library symbols must be declared using compiler-specic code such as this:
__ declspec(dllexport) extern void foo(void);

An interesting aspect of shared libraries is that the linking process happens at each program invocation. To minimize this performance overhead, shared libraries use both
indirection tables and lazy symbol binding. That is, the location
of external symbols actually refers to table entries, which remain unbound until the application actually needs them.
This reduces startup time because most applications use
only a small subset of library functions.
To implement lazy symbol binding, the static linker creates
a jump table known as a procedure-linking table and includes it
as part of the nal executable. Next, the linker resolves all unresolved function references by making them point directly
to a specic PLT entry. So, executable programs created by
the static linker have an internal structure similar to that in
Figure 1. To make lazy symbol binding work at runtime, the
dynamic linker simply clears all the PLT entries and sets them
to point to a special symbol-binding function inside the dynamic library loader. The neat part about this trick is that as
each library function is used for the rst time, the dynamic
linker regains control of the process and performs all the necessary symbol bindings. After it locates a symbol, the linker
simply overwrites the corresponding PLT entry so that subsequent calls to the same function transfer control directly to
the function instead of calling the dynamic linker again. Figure 2 illustrates an overview of this process.
Although symbol binding is normally transparent to users,
you can watch it by setting the LD_DEBUG environment
variable to the value bindings before starting your program.

93

SCIENTIFIC PROGRAMMING
Figure 1. The internal structure of
an executable linked with shared
libraries. External library calls
point to procedure-linking table
entries, which remain unresolved
until the dynamic linker fills them
in at runtime.

Source

foo() {
...
malloc();
...
print();
...
}

Executable
Compile/link

foo() {
...
call malloc()
...
call printf
...
}
PLT

malloc: jmp ???

printf: jmp ???

Executable

Dynamic linker

foo()
...
call malloc
...
call printf
...

ld.so.1
bindsymbol:
Shared library

libc.so
PLT

malloc: jmp x
printf: jmp ???

malloc:

Figure 2. The dynamic binding of library symbols in shared

libraries: malloc is bound to the C library, and printf has
not yet been used and is bound to the dynamic linker.

An interesting experiment is to watch the dynamic symbol

binding for interactive applications such as an interpreter or
a browserfor example,
$ LD_DEBUG=bindings python

If you really enjoy copious debugging information, you can

set LD_DEBUG=bindings in the shell and start running a
few programs.
Library loading

When a program linked with shared libraries runs, program

execution does not immediately start with that programs rst
statement. Instead, the operating system loads and executes the
dynamic linker (usually called ld.so), which then scans the
list of library names embedded in the executable. These library
names are never encoded with absolute pathnames. Instead, the
list has only simple names such as libpthread.so.0,
libm.so.6, and libc.so.6, where the last digit is a library
version number. To locate the libraries, the dynamic linker uses
a congurable library search path. This paths default value is
normally stored in a system conguration le such as
/etc/ld.so.conf. Additionally, other library search directories might be embedded in the executable or specied by the
user in the LD_LIBRARY_PATH environment variable.
Libraries always load in the order in which they were
linked. Therefore, if you linked your program like this,

94

Unresolved
Unresolved

$ cc $SRCS -lfoo -lbar -lsocket -lm -lc

the dynamic linker loads the libraries in the order libfoo.so, libbar.so, libsocket.so, libm.so, and so
forth. If a shared library includes additional library dependencies, those libraries are appended to the end of the library
list during loading. For example, if the library libbar.so
depends on an additional library libspam.so, that library
loads after all the other libraries on the link line (such as
those after libc.so). To be more precise, the library loading order is determined by a breadth-rst traversal of library
dependencies starting with the libraries directly linked to the
executable. Internally, the loader keeps track of the libraries
by placing them on a linked list known as a linkchain. Moreover, the loader guarantees that no library is ever loaded
more than once (previously loaded copies are used when a
library is repeated).
Because directory traversal is relatively slow, the loader does
not look at the directories in /etc/ld.so.conf every time
it runs to nd library les, but consults a cache le instead. Normally named /etc/ld.so.cache, this cache le is a table
that matches library names to full pathnames. If you add a new
shared library to a library directory listed in /etc/ld.
so.conf, you must rebuild the cache le with the ldcong
command, or the loader wont nd it. The -v option produces
verbose detail, including any new or altered libraries.
If the dynamic loader cant find a library in the cache, it
often makes a last-ditch effort with a manual search of system library directories such as /lib and /usr/lib before
it gives up and returns an error. This behavior depends on
the operating system.
You can obtain detailed information about how the dynamic linker loads libraries by setting the LD_DEBUG environment variable to libsfor example,
$ LD_DEBUG=libs a.out

When users rst work with shared libraries, they commonly

experience error messages related to missing libraries. For example, if you create your own shared library such as this,
$ cc -shared $OBJS -o libfoo.so

COMPUTING IN SCIENCE & ENGINEERING

and link an executable with the library, you might get the
following error when you try to run your program:
./a.out: error in loading shared libraries: libfoo.so:
cannot open shared object le: No such le or directory

This error usually occurs when the shared library is placed in

a nonstandard location (not dened in ld.so.conf). A common hack to x this problem is to set the LD_LIBRARY_PATH
environment variable. However, setting LD_LIBRARY_PATH
is nearly always a bad idea. One common mistake is to set it in
your default user environment. Because there is no cache for
this user-dened variable, ld.so must search through each
entry of every directory in this path for a library before it looks
anywhere else. So, virtually every command you run makes
ld.so do more work than it really should whenever it needs
to access a shared library.
A more serious problem is that this approach invites library clashes. A classic example of this was in SunOS 4,
which shipped with the default user environments LD_
LIBRARY_PATH set to /usr/openwin/lib. Because
SunOS 4 had many incompatibilities with several packages,
systems administrators often had to compile the MIT X distribution. However, users ended up with the libraries specied by the environment variables and more warnings about
older versions of libraries than expected (and if they were
lucky, had their programs crash).
A better solution to the library path problem is to embed
customized search paths in the executable itself using special
linker options such as -R or -Wl, -rpathfor example,
$ cc $SRCS Wl,-rpath=/home/beazley/libs \
-L/home/beazleys/libs -lfoo

In this case, the program will nd libfoo.so in the

proper directory without having to set any special environment variables. Even with this approach, managing shared libraries is tricky. If you simply put your custom shared libraries
in a place such as /usr/local/lib, you might have problems if the API changes when you upgrade the library. On the
other hand, if you put them somewhere else, a user might not
be able to nd them. Some library packages such as gtk+ come
with a command that, with certain ags, spits back the linker
options you need for shared libraries (you embed the command in your congure script or Makele).
Library initialization and finalization

As libraries are loaded, the system must occasionally per-

SEPTEMBER/OCTOBER 2001

form certain preliminary steps. For example, if a C++ program has any statically constructed objects, they must be initialized before program startup. For example,
class Foo {
public:
Foo();
~Foo();
...
}
/* Statically initialized object */
Foo f;

To handle this situation, the dynamic linker looks for a

special _init() function or init section in each loaded
library. The compiler creates the contents of _init()and
contains the startup code needed to initialize objects and
other parts of the runtime environment. The invocation of
_init() functions follows in the reverse order of library
loading (for example, the rst library loaded is the last to call
_init()). This reversed ordering is necessary because libraries appearing earlier on the link line might depend on
functions used in libraries appearing later.
When libraries unload at program termination, the dynamic linker looks for special _ni() functions and invokes
them in the opposite order as the _init() functions. The
role of _fini() is to destroy objects and perform other
kinds of cleanup.
If youre curious, you can also trace debugging information
about the invocation of _init() and _ni() functions by
setting the environment variable LD_DEBUG to libs.
Dynamic loading

So far, our discussion has focused primarily on the underlying implementation of shared libraries and how they are
organized to support runtime linking of programs. An added
feature of the dynamic linker is an API for accessing symbols
and loading new libraries at runtime (dynamic loading). The
dynamic loading mechanism is a critical part of many extensible systems, including scripting language interpreters.
Dynamic loading is usually managed by three functions
exposed by the dynamic linker: dlopen() (which loads a
new shared library), dlsym() (which looks up a specific
symbol in the library), and dlclose() (which unloads the
library and removes it from memory)for example,
void *handle;
void (*foo)(void);

95

SCIENTIFIC PROGRAMMING
Figure 3. Linkchains created by
dynamic loading. Each dynamically
loadable module goes on a new
linkchain but can still bind to symbols
defined in the primary executable.

Dynamic module 1

mod1.so

libd.so

Dynamic module 2

Executable

mod2.so

a.out

liba.so

libb.so

libc.so

Dynamic module 3

mod3.so

libe.so

handle = dlopen(foo.so, RTLD_NOW | RTLD_LOCAL);

foo = (void (*)(void)) dlsym(handle,foo);
if (foo) foo();
...
dlclose(handle);

Unlike standard linking, dynamically loaded modules are

completely decoupled from the underlying application. The dynamic loader does not use them to resolve any unbound symbols
or any other part of the normal application-linking process.
The dynamic linker loads a module and all of its required libraries using the exact same procedure as before (that is, it loads
libraries in the order specied on the link line, _init() functions are invoked, and so forth). Again, the linker keeps track
of previously loaded libraries and will not reload shared libraries
that are already present. The main difference is that when modules are loaded, they normally go on private linkchains. This results in a tree of library dependencies (see Figure 3).
As symbols are bound in each dynamically loaded module, the linker searches libraries starting with the module itself and the list of libraries in its corresponding linkchain.
Additionally, the linker searches for symbols in the main executable and all its libraries.
One consequence of the tree structure is that each dynamically loaded module gets its own private symbol namespace. So, if a symbol name is dened in more than one loadable module, those symbols remain distinct and do not clash
with each other during execution. Similarly, unless the program gives special options to the dynamic loader, it doesnt
use symbols dened in previously loaded modules to resolve
symbols in newly loaded modules. For users of scripting languages, this explains why everything still works even when
two extension modules appear to have a namespace clash. It
also explains why extension modules cant dynamically bind
to symbols dened in other dynamically loaded modules.
Filters

Dynamic linking enables modes of linking that are not easily achieved with static libraries. One such example is the implementation of lters that let you alter the behavior of common library functions. For example, suppose you want to
track calls to dynamic memory-allocation functions malloc() and free(). One way to do this with dynamic linking is to write a simple library as in Figure 4. In the gure, we
implemented replacements for the standard memory man-

96

libf.so

agement functions. However, these replacements use dlsym() to search for the real implementation of malloc()
and free() farther down the linkchain.
To use this lter, the library can be included on the linkline
like a normal library. Alternatively, the library can be preloaded
by supplying a special environment variable to the dynamic
linker. The following command illustrates library preloading by
running Python with our memory-allocation lter enabled:
$ LD_PRELOAD=./libmlter.so python

Depending on how you tweak the linker, the user can place
lters almost anywhere on the linkchain. For instance, if the
lter functions were linked to a dynamically loadable module,
only the memory allocations performed by the module pass
through the lters; all other allocations remain unaffected.
Constructing shared libraries

Lets examine a few problematic aspects of shared library

construction. First, when shared libraries and dynamically
loadable modules are compiled, it is fairly common to see special compiler options for creating position-independent code (fpic, -FPIC, -KPIC, and so on). When these options are
present during compilation, the compiler generates code that
accesses symbols through a collection of indirection registers
and global offset tables. The primary benet of PIC code is
that the text segment of the resulting library (containing machine instructions) can quickly relocate to any memory address
without code patches. This improves program startup time
and is important for libraries that large numbers of running
processes must share. (PIC lets the operating system relocate
libraries to different virtual memory regions without modifying the library instructions.) The downside to PIC is a modest degradation in application performance (often 5 to 15 percent). A somewhat overlooked point in many shared library
examples is that shared libraries and dynamically loadable
modules generally do not require PIC code. So, if performance
is important for a library or dynamically loadable module, you
can compile it as non-PIC code. The primary downside to
compiling the module as non-PIC is that loading time increases because the dynamic linker must make a large number
of code patches when binding symbols.
A second problem with shared library construction concerns
the inclusion of static libraries when building dynamically
loadable modules. When the user or programmer works with

COMPUTING IN SCIENCE & ENGINEERING

/* libmlter.c */
#include <dlfcn.h>
#include <stdio.h>
#include <assert.h>
typedef void *(*malloc_t)(size_t nbytes);
void *malloc(size_t nbytes) {
void *r;
static malloc_t real_malloc = 0;
if (!real_malloc) {
real_malloc = (malloc_t) dlsym(RTLD_NEXT, malloc);
assert(real_malloc);
}
r = (*real_malloc)(nbytes);
printf(malloc %d bytes at %x\n, nbytes, r);
return r;
}

scripting language interpreters and other extensible systems, it is common to create various
types of modules for different parts of an application. However, if parts of the application are
built as static libraries (.a les), the linking
process does not do what you might expect.
Specically, when a static library is linked into
a shared module, the relevant parts of the static
library are simply copied into the resulting
typedef void (*free_t)(void *ptr);
shared library object. For multiple modules
void free(void *ptr) {
linked in this manner, each module ends up
static free_t real_free = 0;
with its own private copy of the static library.
if (!real_free) {
When these modules load, all private library
real_free = (free_t) dlsym(RTLD_NEXT, free);
copies remain isolated from each other, owing
assert(real_free);
}
to the way in which symbols are bound in dyprintf(free %x\n, ptr);
namic modules (as described earlier). The end
(*real_free)(ptr);
result is grossly erratic program behavior. For
}
instance, changes to global variables dont seem
to affect other modules, functions seem to operate on different sets of data, and so on. To x
Figure 4. Filters for malloc and free. The filters look up the real
this problem, convert static libraries to shared implementation using dlsym() and print debugging information.
libraries. Then, when multiple dynamic modules link again to the library, the dynamic linker
will guarantee that only one copy of the library loads into 3. R.A. Gingell et al., Shared Libraries in SunOS, Usenix, San Diego, Calif.,
1987.
memory.

4. J.Q. Arnold, Shared Libraries on UNIX System V, Usenix, San Diego, Calif.,
1986.

A few final words and further reading

As extensible systems become more popular in scientic

software, scientists must have a rm understanding of how
shared libraries and dynamically loadable modules interact
with each other and t into the overall picture of large applications. We have described many of the general concepts
and techniques used behind the scenes on these systems. Although our discussion focused primarily on Unix, high-level
concepts apply to all systems that use shared libraries and
dynamic linking.
You can nd further information about linkers, loaders, and
dynamic loading elsewhere.1 Advanced texts on programming
languages and compilers often contain general information
about PLTs, lazy binding, and shared library implementation.2
Several research papers on operating systems contain the specic implementation details of dynamic linking.3,4

David Beazley is an assistant professor in the Department of Computer

Science at the University of Chicago. He created the Simplified Wrapper
and Interface Generator, a popular tool for creating scripting language
extension modules, and wrote the Python Essential Reference (New Riders Publishing, 1999). Contact him at the Dept. of Computer Science,
Univ. of Chicago, Chicago, IL 60637; [email protected].
Brian Ward is a PhD candidate in the Department of Computer Science
at the University of Chicago. He wrote the Linux Kernel HOWTO, The Linux
Problem Solver (No Starch Press, 2000), and The Book of VMware (No Starch
Press, forthcoming). Contact him at the Dept. of Computer Science, Univ.
of Chicago, Chicago, IL 60637; [email protected].

References

Ian Cooke is a PhD student in the Department of Computer Science at

1. J.R. Levine, Linkers & Loaders, Morgan Kaufmann, San Francisco, 2000.

the University of Chicago. Contact him at the Dept. of Computer Sci-

2. M.L. Scott, Programming Language Pragmatics, Morgan Kaufmann, San

Francisco, 2000.

ence, Univ. of Chicago, Chicago, IL 60637; [email protected].

SEPTEMBER/OCTOBER 2001

97