blob: c8a5b0e983b4b1b54b53dbd832cd95b39247f7c4 [file] [log] [blame]
This is Info file, produced by Makeinfo version 1.68 from the
input file cpp.texi.
This file documents the GNU C Preprocessor.
Copyright 1987, 1989, 1991, 1992, 1993, 1994, 1995 Free Software
Foundation, Inc.
Permission is granted to make and distribute verbatim copies of this
manual provided the copyright notice and this permission notice are
preserved on all copies.
Permission is granted to copy and distribute modified versions of
this manual under the conditions for verbatim copying, provided also
that the entire resulting derived work is distributed under the terms
of a permission notice identical to this one.
Permission is granted to copy and distribute translations of this
manual into another language, under the above conditions for modified

File:, Node: Macro Pitfalls, Prev: Redefining, Up: Macros
Pitfalls and Subtleties of Macros
In this section we describe some special rules that apply to macros
and macro expansion, and point out certain cases in which the rules have
counterintuitive consequences that you must watch out for.
* Menu:
* Misnesting:: Macros can contain unmatched parentheses.
* Macro Parentheses:: Why apparently superfluous parentheses
may be necessary to avoid incorrect grouping.
* Swallow Semicolon:: Macros that look like functions
but expand into compound statements.
* Side Effects:: Unsafe macros that cause trouble when
arguments contain side effects.
* Self-Reference:: Macros whose definitions use the macros' own names.
* Argument Prescan:: Actual arguments are checked for macro calls
before they are substituted.
* Cascaded Macros:: Macros whose definitions use other macros.
* Newlines in Args:: Sometimes line numbers get confused.

File:, Node: Misnesting, Next: Macro Parentheses, Prev: Macro Pitfalls, Up: Macro Pitfalls
Improperly Nested Constructs
Recall that when a macro is called with arguments, the arguments are
substituted into the macro body and the result is checked, together with
the rest of the input file, for more macro calls.
It is possible to piece together a macro call coming partially from
the macro body and partially from the actual arguments. For example,
#define double(x) (2*(x))
#define call_with_1(x) x(1)
would expand `call_with_1 (double)' into `(2*(1))'.
Macro definitions do not have to have balanced parentheses. By
writing an unbalanced open parenthesis in a macro body, it is possible
to create a macro call that begins inside the macro body but ends
outside of it. For example,
#define strange(file) fprintf (file, "%s %d",
strange(stderr) p, 35)
This bizarre example expands to `fprintf (stderr, "%s %d", p, 35)'!

File:, Node: Macro Parentheses, Next: Swallow Semicolon, Prev: Misnesting, Up: Macro Pitfalls
Unintended Grouping of Arithmetic
You may have noticed that in most of the macro definition examples
shown above, each occurrence of a macro argument name had parentheses
around it. In addition, another pair of parentheses usually surround
the entire macro definition. Here is why it is best to write macros
that way.
Suppose you define a macro as follows,
#define ceil_div(x, y) (x + y - 1) / y
whose purpose is to divide, rounding up. (One use for this operation is
to compute how many `int' objects are needed to hold a certain number
of `char' objects.) Then suppose it is used as follows:
a = ceil_div (b & c, sizeof (int));
This expands into
a = (b & c + sizeof (int) - 1) / sizeof (int);
which does not do what is intended. The operator-precedence rules of C
make it equivalent to this:
a = (b & (c + sizeof (int) - 1)) / sizeof (int);
But what we want is this:
a = ((b & c) + sizeof (int) - 1)) / sizeof (int);
Defining the macro as
#define ceil_div(x, y) ((x) + (y) - 1) / (y)
provides the desired result.
Unintended grouping can result in another way. Consider `sizeof
ceil_div(1, 2)'. That has the appearance of a C expression that would
compute the size of the type of `ceil_div (1, 2)', but in fact it means
something very different. Here is what it expands to:
sizeof ((1) + (2) - 1) / (2)
This would take the size of an integer and divide it by two. The
precedence rules have put the division outside the `sizeof' when it was
intended to be inside.
Parentheses around the entire macro definition can prevent such
problems. Here, then, is the recommended way to define `ceil_div':
#define ceil_div(x, y) (((x) + (y) - 1) / (y))

File:, Node: Swallow Semicolon, Next: Side Effects, Prev: Macro Parentheses, Up: Macro Pitfalls
Swallowing the Semicolon
Often it is desirable to define a macro that expands into a compound
statement. Consider, for example, the following macro, that advances a
pointer (the argument `p' says where to find it) across whitespace
#define SKIP_SPACES (p, limit) \
{ register char *lim = (limit); \
while (p != lim) { \
if (*p++ != ' ') { \
p--; break; }}}
Here Backslash-Newline is used to split the macro definition, which must
be a single line, so that it resembles the way such C code would be
laid out if not part of a macro definition.
A call to this macro might be `SKIP_SPACES (p, lim)'. Strictly
speaking, the call expands to a compound statement, which is a complete
statement with no need for a semicolon to end it. But it looks like a
function call. So it minimizes confusion if you can use it like a
function call, writing a semicolon afterward, as in `SKIP_SPACES (p,
But this can cause trouble before `else' statements, because the
semicolon is actually a null statement. Suppose you write
if (*p != 0)
SKIP_SPACES (p, lim);
else ...
The presence of two statements--the compound statement and a null
statement--in between the `if' condition and the `else' makes invalid C
The definition of the macro `SKIP_SPACES' can be altered to solve
this problem, using a `do ... while' statement. Here is how:
#define SKIP_SPACES (p, limit) \
do { register char *lim = (limit); \
while (p != lim) { \
if (*p++ != ' ') { \
p--; break; }}} \
while (0)
Now `SKIP_SPACES (p, lim);' expands into
do {...} while (0);
which is one statement.

File:, Node: Side Effects, Next: Self-Reference, Prev: Swallow Semicolon, Up: Macro Pitfalls
Duplication of Side Effects
Many C programs define a macro `min', for "minimum", like this:
#define min(X, Y) ((X) < (Y) ? (X) : (Y))
When you use this macro with an argument containing a side effect,
as shown here,
next = min (x + y, foo (z));
it expands as follows:
next = ((x + y) < (foo (z)) ? (x + y) : (foo (z)));
where `x + y' has been substituted for `X' and `foo (z)' for `Y'.
The function `foo' is used only once in the statement as it appears
in the program, but the expression `foo (z)' has been substituted twice
into the macro expansion. As a result, `foo' might be called two times
when the statement is executed. If it has side effects or if it takes
a long time to compute, the results might not be what you intended. We
say that `min' is an "unsafe" macro.
The best solution to this problem is to define `min' in a way that
computes the value of `foo (z)' only once. The C language offers no
standard way to do this, but it can be done with GNU C extensions as
#define min(X, Y) \
({ typeof (X) __x = (X), __y = (Y); \
(__x < __y) ? __x : __y; })
If you do not wish to use GNU C extensions, the only solution is to
be careful when *using* the macro `min'. For example, you can
calculate the value of `foo (z)', save it in a variable, and use that
variable in `min':
#define min(X, Y) ((X) < (Y) ? (X) : (Y))
int tem = foo (z);
next = min (x + y, tem);
(where we assume that `foo' returns type `int').

File:, Node: Self-Reference, Next: Argument Prescan, Prev: Side Effects, Up: Macro Pitfalls
Self-Referential Macros
A "self-referential" macro is one whose name appears in its
definition. A special feature of ANSI Standard C is that the
self-reference is not considered a macro call. It is passed into the
preprocessor output unchanged.
Let's consider an example:
#define foo (4 + foo)
where `foo' is also a variable in your program.
Following the ordinary rules, each reference to `foo' will expand
into `(4 + foo)'; then this will be rescanned and will expand into `(4
+ (4 + foo))'; and so on until it causes a fatal error (memory full) in
the preprocessor.
However, the special rule about self-reference cuts this process
short after one step, at `(4 + foo)'. Therefore, this macro definition
has the possibly useful effect of causing the program to add 4 to the
value of `foo' wherever `foo' is referred to.
In most cases, it is a bad idea to take advantage of this feature. A
person reading the program who sees that `foo' is a variable will not
expect that it is a macro as well. The reader will come across the
identifier `foo' in the program and think its value should be that of
the variable `foo', whereas in fact the value is four greater.
The special rule for self-reference applies also to "indirect"
self-reference. This is the case where a macro X expands to use a
macro `y', and the expansion of `y' refers to the macro `x'. The
resulting reference to `x' comes indirectly from the expansion of `x',
so it is a self-reference and is not further expanded. Thus, after
#define x (4 + y)
#define y (2 * x)
`x' would expand into `(4 + (2 * x))'. Clear?
But suppose `y' is used elsewhere, not from the definition of `x'.
Then the use of `x' in the expansion of `y' is not a self-reference
because `x' is not "in progress". So it does expand. However, the
expansion of `x' contains a reference to `y', and that is an indirect
self-reference now because `y' is "in progress". The result is that
`y' expands to `(2 * (4 + y))'.
It is not clear that this behavior would ever be useful, but it is
specified by the ANSI C standard, so you may need to understand it.

File:, Node: Argument Prescan, Next: Cascaded Macros, Prev: Self-Reference, Up: Macro Pitfalls
Separate Expansion of Macro Arguments
We have explained that the expansion of a macro, including the
substituted actual arguments, is scanned over again for macro calls to
be expanded.
What really happens is more subtle: first each actual argument text
is scanned separately for macro calls. Then the results of this are
substituted into the macro body to produce the macro expansion, and the
macro expansion is scanned again for macros to expand.
The result is that the actual arguments are scanned *twice* to expand
macro calls in them.
Most of the time, this has no effect. If the actual argument
contained any macro calls, they are expanded during the first scan.
The result therefore contains no macro calls, so the second scan does
not change it. If the actual argument were substituted as given, with
no prescan, the single remaining scan would find the same macro calls
and produce the same results.
You might expect the double scan to change the results when a
self-referential macro is used in an actual argument of another macro
(*note Self-Reference::.): the self-referential macro would be expanded
once in the first scan, and a second time in the second scan. But this
is not what happens. The self-references that do not expand in the
first scan are marked so that they will not expand in the second scan
The prescan is not done when an argument is stringified or
concatenated. Thus,
#define str(s) #s
#define foo 4
str (foo)
expands to `"foo"'. Once more, prescan has been prevented from having
any noticeable effect.
More precisely, stringification and concatenation use the argument as
written, in un-prescanned form. The same actual argument would be used
in prescanned form if it is substituted elsewhere without
stringification or concatenation.
#define str(s) #s lose(s)
#define foo 4
str (foo)
expands to `"foo" lose(4)'.
You might now ask, "Why mention the prescan, if it makes no
difference? And why not skip it and make the preprocessor faster?"
The answer is that the prescan does make a difference in three special
* Nested calls to a macro.
* Macros that call other macros that stringify or concatenate.
* Macros whose expansions contain unshielded commas.
We say that "nested" calls to a macro occur when a macro's actual
argument contains a call to that very macro. For example, if `f' is a
macro that expects one argument, `f (f (1))' is a nested pair of calls
to `f'. The desired expansion is made by expanding `f (1)' and
substituting that into the definition of `f'. The prescan causes the
expected result to happen. Without the prescan, `f (1)' itself would
be substituted as an actual argument, and the inner use of `f' would
appear during the main scan as an indirect self-reference and would not
be expanded. Here, the prescan cancels an undesirable side effect (in
the medical, not computational, sense of the term) of the special rule
for self-referential macros.
But prescan causes trouble in certain other cases of nested macro
calls. Here is an example:
#define foo a,b
#define bar(x) lose(x)
#define lose(x) (1 + (x))
We would like `bar(foo)' to turn into `(1 + (foo))', which would then
turn into `(1 + (a,b))'. But instead, `bar(foo)' expands into
`lose(a,b)', and you get an error because `lose' requires a single
argument. In this case, the problem is easily solved by the same
parentheses that ought to be used to prevent misnesting of arithmetic
#define foo (a,b)
#define bar(x) lose((x))
The problem is more serious when the operands of the macro are not
expressions; for example, when they are statements. Then parentheses
are unacceptable because they would make for invalid C code:
#define foo { int a, b; ... }
In GNU C you can shield the commas using the `({...})' construct which
turns a compound statement into an expression:
#define foo ({ int a, b; ... })
Or you can rewrite the macro definition to avoid such commas:
#define foo { int a; int b; ... }
There is also one case where prescan is useful. It is possible to
use prescan to expand an argument and then stringify it--if you use two
levels of macros. Let's add a new macro `xstr' to the example shown
#define xstr(s) str(s)
#define str(s) #s
#define foo 4
xstr (foo)
This expands into `"4"', not `"foo"'. The reason for the difference
is that the argument of `xstr' is expanded at prescan (because `xstr'
does not specify stringification or concatenation of the argument).
The result of prescan then forms the actual argument for `str'. `str'
uses its argument without prescan because it performs stringification;
but it cannot prevent or undo the prescanning already done by `xstr'.

File:, Node: Cascaded Macros, Next: Newlines in Args, Prev: Argument Prescan, Up: Macro Pitfalls
Cascaded Use of Macros
A "cascade" of macros is when one macro's body contains a reference
to another macro. This is very common practice. For example,
#define BUFSIZE 1020
This is not at all the same as defining `TABLESIZE' to be `1020'.
The `#define' for `TABLESIZE' uses exactly the body you specify--in
this case, `BUFSIZE'--and does not check to see whether it too is the
name of a macro.
It's only when you *use* `TABLESIZE' that the result of its expansion
is checked for more macro names.
This makes a difference if you change the definition of `BUFSIZE' at
some point in the source file. `TABLESIZE', defined as shown, will
always expand using the definition of `BUFSIZE' that is currently in
#define BUFSIZE 1020
#undef BUFSIZE
#define BUFSIZE 37
Now `TABLESIZE' expands (in two stages) to `37'. (The `#undef' is to
prevent any warning about the nontrivial redefinition of `BUFSIZE'.)

File:, Node: Newlines in Args, Prev: Cascaded Macros, Up: Macro Pitfalls
Newlines in Macro Arguments
Traditional macro processing carries forward all newlines in macro
arguments into the expansion of the macro. This means that, if some of
the arguments are substituted more than once, or not at all, or out of
order, newlines can be duplicated, lost, or moved around within the
expansion. If the expansion consists of multiple statements, then the
effect is to distort the line numbers of some of these statements. The
result can be incorrect line numbers, in error messages or displayed in
a debugger.
The GNU C preprocessor operating in ANSI C mode adjusts appropriately
for multiple use of an argument--the first use expands all the
newlines, and subsequent uses of the same argument produce no newlines.
But even in this mode, it can produce incorrect line numbering if
arguments are used out of order, or not used at all.
Here is an example illustrating this problem:
#define ignore_second_arg(a,b,c) a; c
ignore_second_arg (foo (),
ignored (),
syntax error);
The syntax error triggered by the tokens `syntax error' results in an
error message citing line four, even though the statement text comes
from line five.

File:, Node: Conditionals, Next: Combining Sources, Prev: Macros, Up: Top
In a macro processor, a "conditional" is a directive that allows a
part of the program to be ignored during compilation, on some
conditions. In the C preprocessor, a conditional can test either an
arithmetic expression or whether a name is defined as a macro.
A conditional in the C preprocessor resembles in some ways an `if'
statement in C, but it is important to understand the difference between
them. The condition in an `if' statement is tested during the execution
of your program. Its purpose is to allow your program to behave
differently from run to run, depending on the data it is operating on.
The condition in a preprocessing conditional directive is tested when
your program is compiled. Its purpose is to allow different code to be
included in the program depending on the situation at the time of
* Menu:
* Uses: Conditional Uses. What conditionals are for.
* Syntax: Conditional Syntax. How conditionals are written.
* Deletion: Deleted Code. Making code into a comment.
* Macros: Conditionals-Macros. Why conditionals are used with macros.
* Assertions:: How and why to use assertions.
* Errors: #error Directive. Detecting inconsistent compilation parameters.

File:, Node: Conditional Uses, Next: Conditional Syntax, Up: Conditionals
Why Conditionals are Used
Generally there are three kinds of reason to use a conditional.
* A program may need to use different code depending on the machine
or operating system it is to run on. In some cases the code for
one operating system may be erroneous on another operating system;
for example, it might refer to library routines that do not exist
on the other system. When this happens, it is not enough to avoid
executing the invalid code: merely having it in the program makes
it impossible to link the program and run it. With a
preprocessing conditional, the offending code can be effectively
excised from the program when it is not valid.
* You may want to be able to compile the same source file into two
different programs. Sometimes the difference between the programs
is that one makes frequent time-consuming consistency checks on its
intermediate data, or prints the values of those data for
debugging, while the other does not.
* A conditional whose condition is always false is a good way to
exclude code from the program but keep it as a sort of comment for
future reference.
Most simple programs that are intended to run on only one machine
will not need to use preprocessing conditionals.

File:, Node: Conditional Syntax, Next: Deleted Code, Prev: Conditional Uses, Up: Conditionals
Syntax of Conditionals
A conditional in the C preprocessor begins with a "conditional
directive": `#if', `#ifdef' or `#ifndef'. *Note Conditionals-Macros::,
for information on `#ifdef' and `#ifndef'; only `#if' is explained here.
* Menu:
* If: #if Directive. Basic conditionals using `#if' and `#endif'.
* Else: #else Directive. Including some text if the condition fails.
* Elif: #elif Directive. Testing several alternative possibilities.

File:, Node: #if Directive, Next: #else Directive, Up: Conditional Syntax
The `#if' Directive
The `#if' directive in its simplest form consists of
#endif /* EXPRESSION */
The comment following the `#endif' is not required, but it is a good
practice because it helps people match the `#endif' to the
corresponding `#if'. Such comments should always be used, except in
short conditionals that are not nested. In fact, you can put anything
at all after the `#endif' and it will be ignored by the GNU C
preprocessor, but only comments are acceptable in ANSI Standard C.
EXPRESSION is a C expression of integer type, subject to stringent
restrictions. It may contain
* Integer constants, which are all regarded as `long' or `unsigned
* Character constants, which are interpreted according to the
character set and conventions of the machine and operating system
on which the preprocessor is running. The GNU C preprocessor uses
the C data type `char' for these character constants; therefore,
whether some character codes are negative is determined by the C
compiler used to compile the preprocessor. If it treats `char' as
signed, then character codes large enough to set the sign bit will
be considered negative; otherwise, no character code is considered
* Arithmetic operators for addition, subtraction, multiplication,
division, bitwise operations, shifts, comparisons, and logical
operations (`&&' and `||').
* Identifiers that are not macros, which are all treated as zero(!).
* Macro calls. All macro calls in the expression are expanded before
actual computation of the expression's value begins.
Note that `sizeof' operators and `enum'-type values are not allowed.
`enum'-type values, like all other identifiers that are not taken as
macro calls and expanded, are treated as zero.
The CONTROLLED TEXT inside of a conditional can include
preprocessing directives. Then the directives inside the conditional
are obeyed only if that branch of the conditional succeeds. The text
can also contain other conditional groups. However, the `#if' and
`#endif' directives must balance.

File:, Node: #else Directive, Next: #elif Directive, Prev: #if Directive, Up: Conditional Syntax
The `#else' Directive
The `#else' directive can be added to a conditional to provide
alternative text to be used if the condition is false. This is what it
looks like:
#else /* Not EXPRESSION */
#endif /* Not EXPRESSION */
If EXPRESSION is nonzero, and thus the TEXT-IF-TRUE is active, then
`#else' acts like a failing conditional and the TEXT-IF-FALSE is
ignored. Contrariwise, if the `#if' conditional fails, the
TEXT-IF-FALSE is considered included.

File:, Node: #elif Directive, Prev: #else Directive, Up: Conditional Syntax
The `#elif' Directive
One common case of nested conditionals is used to check for more
than two possible alternatives. For example, you might have
#if X == 1
#else /* X != 1 */
#if X == 2
#else /* X != 2 */
#endif /* X != 2 */
#endif /* X != 1 */
Another conditional directive, `#elif', allows this to be abbreviated
as follows:
#if X == 1
#elif X == 2
#else /* X != 2 and X != 1*/
#endif /* X != 2 and X != 1*/
`#elif' stands for "else if". Like `#else', it goes in the middle
of a `#if'-`#endif' pair and subdivides it; it does not require a
matching `#endif' of its own. Like `#if', the `#elif' directive
includes an expression to be tested.
The text following the `#elif' is processed only if the original
`#if'-condition failed and the `#elif' condition succeeds. More than
one `#elif' can go in the same `#if'-`#endif' group. Then the text
after each `#elif' is processed only if the `#elif' condition succeeds
after the original `#if' and any previous `#elif' directives within it
have failed. `#else' is equivalent to `#elif 1', and `#else' is
allowed after any number of `#elif' directives, but `#elif' may not
follow `#else'.

File:, Node: Deleted Code, Next: Conditionals-Macros, Prev: Conditional Syntax, Up: Conditionals
Keeping Deleted Code for Future Reference
If you replace or delete a part of the program but want to keep the
old code around as a comment for future reference, the easy way to do
this is to put `#if 0' before it and `#endif' after it. This is better
than using comment delimiters `/*' and `*/' since those won't work if
the code already contains comments (C comments do not nest).
This works even if the code being turned off contains conditionals,
but they must be entire conditionals (balanced `#if' and `#endif').
Conversely, do not use `#if 0' for comments which are not C code.
Use the comment delimiters `/*' and `*/' instead. The interior of `#if
0' must consist of complete tokens; in particular, singlequote
characters must balance. But comments often contain unbalanced
singlequote characters (known in English as apostrophes). These
confuse `#if 0'. They do not confuse `/*'.

File:, Node: Conditionals-Macros, Next: Assertions, Prev: Deleted Code, Up: Conditionals
Conditionals and Macros
Conditionals are useful in connection with macros or assertions,
because those are the only ways that an expression's value can vary
from one compilation to another. A `#if' directive whose expression
uses no macros or assertions is equivalent to `#if 1' or `#if 0'; you
might as well determine which one, by computing the value of the
expression yourself, and then simplify the program.
For example, here is a conditional that tests the expression
`BUFSIZE == 1020', where `BUFSIZE' must be a macro.
#if BUFSIZE == 1020
printf ("Large buffers!\n");
#endif /* BUFSIZE is large */
(Programmers often wish they could test the size of a variable or
data type in `#if', but this does not work. The preprocessor does not
understand `sizeof', or typedef names, or even the type keywords such
as `int'.)
The special operator `defined' is used in `#if' expressions to test
whether a certain name is defined as a macro. Either `defined NAME' or
`defined (NAME)' is an expression whose value is 1 if NAME is defined
as macro at the current point in the program, and 0 otherwise. For the
`defined' operator it makes no difference what the definition of the
macro is; all that matters is whether there is a definition. Thus, for
#if defined (vax) || defined (ns16000)
would succeed if either of the names `vax' and `ns16000' is defined as
a macro. You can test the same condition using assertions (*note
Assertions::.), like this:
#if #cpu (vax) || #cpu (ns16000)
If a macro is defined and later undefined with `#undef', subsequent
use of the `defined' operator returns 0, because the name is no longer
defined. If the macro is defined again with another `#define',
`defined' will recommence returning 1.
Conditionals that test whether just one name is defined are very
common, so there are two special short conditional directives for this
`#ifdef NAME'
is equivalent to `#if defined (NAME)'.
`#ifndef NAME'
is equivalent to `#if ! defined (NAME)'.
Macro definitions can vary between compilations for several reasons.
* Some macros are predefined on each kind of machine. For example,
on a Vax, the name `vax' is a predefined macro. On other
machines, it would not be defined.
* Many more macros are defined by system header files. Different
systems and machines define different macros, or give them
different values. It is useful to test these macros with
conditionals to avoid using a system feature on a machine where it
is not implemented.
* Macros are a common way of allowing users to customize a program
for different machines or applications. For example, the macro
`BUFSIZE' might be defined in a configuration file for your
program that is included as a header file in each source file. You
would use `BUFSIZE' in a preprocessing conditional in order to
generate different code depending on the chosen configuration.
* Macros can be defined or undefined with `-D' and `-U' command
options when you compile the program. You can arrange to compile
the same source file into two different programs by choosing a
macro name to specify which program you want, writing conditionals
to test whether or how this macro is defined, and then controlling
the state of the macro with compiler command options. *Note
Assertions are usually predefined, but can be defined with
preprocessor directives or command-line options.

File:, Node: Assertions, Next: #error Directive, Prev: Conditionals-Macros, Up: Conditionals
"Assertions" are a more systematic alternative to macros in writing
conditionals to test what sort of computer or system the compiled
program will run on. Assertions are usually predefined, but you can
define them with preprocessing directives or command-line options.
The macros traditionally used to describe the type of target are not
classified in any way according to which question they answer; they may
indicate a hardware architecture, a particular hardware model, an
operating system, a particular version of an operating system, or
specific configuration options. These are jumbled together in a single
namespace. In contrast, each assertion consists of a named question and
an answer. The question is usually called the "predicate". An
assertion looks like this:
You must use a properly formed identifier for PREDICATE. The value of
ANSWER can be any sequence of words; all characters are significant
except for leading and trailing whitespace, and differences in internal
whitespace sequences are ignored. Thus, `x + y' is different from
`x+y' but equivalent to `x + y'. `)' is not allowed in an answer.
Here is a conditional to test whether the answer ANSWER is asserted
for the predicate PREDICATE:
There may be more than one answer asserted for a given predicate. If
you omit the answer, you can test whether *any* answer is asserted for
Most of the time, the assertions you test will be predefined
assertions. GNU C provides three predefined predicates: `system',
`cpu', and `machine'. `system' is for assertions about the type of
software, `cpu' describes the type of computer architecture, and
`machine' gives more information about the computer. For example, on a
GNU system, the following assertions would be true:
#system (gnu)
#system (mach)
#system (mach 3)
#system (mach 3.SUBVERSION)
#system (hurd)
#system (hurd VERSION)
and perhaps others. The alternatives with more or less version
information let you ask more or less detailed questions about the type
of system software.
On a Unix system, you would find `#system (unix)' and perhaps one of:
`#system (aix)', `#system (bsd)', `#system (hpux)', `#system (lynx)',
`#system (mach)', `#system (posix)', `#system (svr3)', `#system
(svr4)', or `#system (xpg4)' with possible version numbers following.
Other values for `system' are `#system (mvs)' and `#system (vms)'.
*Portability note:* Many Unix C compilers provide only one answer
for the `system' assertion: `#system (unix)', if they support
assertions at all. This is less than useful.
An assertion with a multi-word answer is completely different from
several assertions with individual single-word answers. For example,
the presence of `system (mach 3.0)' does not mean that `system (3.0)'
is true. It also does not directly imply `system (mach)', but in GNU
C, that last will normally be asserted as well.
The current list of possible assertion values for `cpu' is: `#cpu
(a29k)', `#cpu (alpha)', `#cpu (arm)', `#cpu (clipper)', `#cpu
(convex)', `#cpu (elxsi)', `#cpu (tron)', `#cpu (h8300)', `#cpu
(i370)', `#cpu (i386)', `#cpu (i860)', `#cpu (i960)', `#cpu (m68k)',
`#cpu (m88k)', `#cpu (mips)', `#cpu (ns32k)', `#cpu (hppa)', `#cpu
(pyr)', `#cpu (ibm032)', `#cpu (rs6000)', `#cpu (sh)', `#cpu (sparc)',
`#cpu (spur)', `#cpu (tahoe)', `#cpu (vax)', `#cpu (we32000)'.
You can create assertions within a C program using `#assert', like
(Note the absence of a `#' before PREDICATE.)
Each time you do this, you assert a new true answer for PREDICATE.
Asserting one answer does not invalidate previously asserted answers;
they all remain true. The only way to remove an assertion is with
`#unassert'. `#unassert' has the same syntax as `#assert'. You can
also remove all assertions about PREDICATE like this:
#unassert PREDICATE
You can also add or cancel assertions using command options when you
run `gcc' or `cpp'. *Note Invocation::.

File:, Node: #error Directive, Prev: Assertions, Up: Conditionals
The `#error' and `#warning' Directives
The directive `#error' causes the preprocessor to report a fatal
error. The rest of the line that follows `#error' is used as the error
message. The line must consist of complete tokens.
You would use `#error' inside of a conditional that detects a
combination of parameters which you know the program does not properly
support. For example, if you know that the program will not run
properly on a Vax, you might write
#ifdef __vax__
#error "Won't work on Vaxen. See comments at get_last_object."
*Note Nonstandard Predefined::, for why this works.
If you have several configuration parameters that must be set up by
the installation in a consistent way, you can use conditionals to detect
an inconsistency and report it with `#error'. For example,
#if HASH_TABLE_SIZE % 2 == 0 || HASH_TABLE_SIZE % 3 == 0 \
|| HASH_TABLE_SIZE % 5 == 0
#error HASH_TABLE_SIZE should not be divisible by a small prime
The directive `#warning' is like the directive `#error', but causes
the preprocessor to issue a warning and continue preprocessing. The
rest of the line that follows `#warning' is used as the warning message.
You might use `#warning' in obsolete header files, with a message
directing the user to the header file which should be used instead.

File:, Node: Combining Sources, Next: Other Directives, Prev: Conditionals, Up: Top
Combining Source Files
One of the jobs of the C preprocessor is to inform the C compiler of
where each line of C code came from: which source file and which line
C code can come from multiple source files if you use `#include';
both `#include' and the use of conditionals and macros can cause the
line number of a line in the preprocessor output to be different from
the line's number in the original source file. You will appreciate the
value of making both the C compiler (in error messages) and symbolic
debuggers such as GDB use the line numbers in your source file.
The C preprocessor builds on this feature by offering a directive by
which you can control the feature explicitly. This is useful when a
file for input to the C preprocessor is the output from another program
such as the `bison' parser generator, which operates on another file
that is the true source file. Parts of the output from `bison' are
generated from scratch, other parts come from a standard parser file.
The rest are copied nearly verbatim from the source file, but their
line numbers in the `bison' output are not the same as their original
line numbers. Naturally you would like compiler error messages and
symbolic debuggers to know the original source file and line number of
each line in the `bison' input.
`bison' arranges this by writing `#line' directives into the output
file. `#line' is a directive that specifies the original line number
and source file name for subsequent input in the current preprocessor
input file. `#line' has three variants:
`#line LINENUM'
Here LINENUM is a decimal integer constant. This specifies that
the line number of the following line of input, in its original
source file, was LINENUM.
Here LINENUM is a decimal integer constant and FILENAME is a
string constant. This specifies that the following line of input
came originally from source file FILENAME and its line number there
was LINENUM. Keep in mind that FILENAME is not just a file name;
it is surrounded by doublequote characters so that it looks like a
string constant.
ANYTHING ELSE is checked for macro calls, which are expanded. The
result should be a decimal integer constant followed optionally by
a string constant, as described above.
`#line' directives alter the results of the `__FILE__' and
`__LINE__' predefined macros from that point on. *Note Standard
The output of the preprocessor (which is the input for the rest of
the compiler) contains directives that look much like `#line'
directives. They start with just `#' instead of `#line', but this is
followed by a line number and file name as in `#line'. *Note Output::.

File:, Node: Other Directives, Next: Output, Prev: Combining Sources, Up: Top
Miscellaneous Preprocessing Directives
This section describes three additional preprocessing directives.
They are not very useful, but are mentioned for completeness.
The "null directive" consists of a `#' followed by a Newline, with
only whitespace (including comments) in between. A null directive is
understood as a preprocessing directive but has no effect on the
preprocessor output. The primary significance of the existence of the
null directive is that an input line consisting of just a `#' will
produce no output, rather than a line of output containing just a `#'.
Supposedly some old C programs contain such lines.
The ANSI standard specifies that the effect of the `#pragma'
directive is implementation-defined. In the GNU C preprocessor,
`#pragma' directives are not used, except for `#pragma once' (*note
Once-Only::.). However, they are left in the preprocessor output, so
they are available to the compilation pass.
The `#ident' directive is supported for compatibility with certain
other systems. It is followed by a line of text. On some systems, the
text is copied into a special place in the object file; on most systems,
the text is ignored and this directive has no effect. Typically
`#ident' is only used in header files supplied with those systems where
it is meaningful.

File:, Node: Output, Next: Invocation, Prev: Other Directives, Up: Top
C Preprocessor Output
The output from the C preprocessor looks much like the input, except
that all preprocessing directive lines have been replaced with blank
lines and all comments with spaces. Whitespace within a line is not
altered; however, unless `-traditional' is used, spaces may be inserted
into the expansions of macro calls to prevent tokens from being
Source file name and line number information is conveyed by lines of
the form
which are inserted as needed into the middle of the input (but never
within a string or character constant). Such a line means that the
following line originated in file FILENAME at line LINENUM.
After the file name comes zero or more flags, which are `1', `2',
`3', or `4'. If there are multiple flags, spaces separate them. Here
is what the flags mean:
This indicates the start of a new file.
This indicates returning to a file (after having included another
This indicates that the following text comes from a system header
file, so certain warnings should be suppressed.
This indicates that the following text should be treated as C.