I have the following code for a one dimensional array:
#include <stdio.h>
int main()
{
static int b[4] = {11, 12, 13, 14};
int (*p)[4];
p = b;
printf("%d \n", *(p + 1));
return 0;
}
Even though I consider "b (the array name)" as a pointer pointing to a one dimensional array, I got a compiling error as
'=': cannot convert from 'int [4]' to 'int (*)[4]'
However, if I change b array into a two dimensional array "a (the array name)", everything works fine. Does this mean that, in the usage of "int (*p)[4];", "*p" has to represent a[] as in the following:
static int a[3][4] = { {1, 2, 3, 4}, {5, 6, 7, 8}, {9, 10, 11, 12} };
int (*p)[4];
p = a;
As a result, "int (*p)[4]" only provides the flexibility on the number of rows of a two dimensional array.
Any insights on this problem?
Arrays naturally decay to pointers to their first elements, depending on context. That is, when such a decay happen then plain b is the same as &b[0], which have the type int *. Since the types of p and b (or &b[0]) are different you get an error.
As for a it's the same thing here, it decays to a pointer to its first element, i.e. a is the same as &a[0]. But since a[0] is an array of 4 elements, then &a[0] is a pointer to an array of four elements, or int (*)[4]. Which is also the type of p in the second example.
If you have an object of some type T like
T a;
then declaration of a pointer to the object will look like
T *p = &a;
Your array b has the type int[4]. So a pointer to the array will look like
int ( *p )[4] = &b;
To output the second element of the array using the pointer you should write
printf("%d \n", *( *p + 1 ) );
Thus your compiler issued the error message
cannot convert from 'int [4]' to 'int (*)[4]
because instead of writing at least
int ( *p )[4] = &b;
you wrote
int ( *p )[4] = b;
On the other hand, an array designator used in expressions with rare exceptions is implicitly converted to pointer to its first element. For example in this declaration
int *p = b;
the array b used as an initializer is converted to pointer to its firs element. The above declaration is equivalent to
int *p = &b[0];
or that is the same
int *p = b + 0;
Using this pointer you can call the function printf like
printf("%d \n", *(p + 1));
If you have a two-dimensional array as
int a[3][4];
then used in expressions it is converted to pointer to its first element that has the type int[4]. So you may write
int ( *p )[4] = a;
If you want to declare a pointer to the whole array as a single object you can write
int ( *p )[3][4] = &a;
a pointer pointing to a one dimensional array,
No, it points directly to the first element. Likewise:
int *p = b;
is enough.
The number 4 is not really part of any type here;
static int b[] = {11, 12, 13, 14};
It can be left out in the declaration. (Because it is the first dimension unless you make it 2D)
This (from AA)
int (*p)[4] = &b;
...
printf("%d \n", *( *p + 1 ) );
is just a obfuscated and overtyped version of:
int (*p)[] = &b;
...
printf("%d \n", (*p)[1] );
This replaces b with (*p), normally not what you want.
As I was going though some code examples, I encountered this way of declaration:
int (*arr1)[10]
I am aware of another way of array declaration in C:
int arr2[10]
Is int (*arr1)[10] doing the same as int arr2[10]?
The differences are as follows:
int (*arr1)[10]; // arr1 is a pointer to a 10-element array of int
int *arr2[10]; // arr2 is a 10-element array of pointer to int
This is important - despite the [10] at the end, arr1 is not an array; it only stores a single pointer value, which is the address of some 10-element array of int.
Pointers to arrays crop up in two main circumstances - when an N-dimensional array expression "decays" to a pointer expression, and when we're allocating memory for an N-dimensional array.
First, the decay rule - except when it is the operand of the sizeof or the unary & operators, or is a string literal used to initialize a character array in a declaration, an expression of type "N-element array of T" will be converted, or "decay", to an expression of type "pointer to T" and the value of the expression will be the address of the first element of the array. You've probably seen this before with code like:
int arr[10];
int *p = arr; // arr "decays" to a pointer to its first element
The expression arr "decays" from type "10-element array of int" to type "pointer to int".
The exact same thing happens with 2D arrays:
int arr[5][10];
int (*p)[10] = arr;
The expression arr "decays" from type "5-element array of 10-element array of int" to type "pointer to 10-element array of int".
You can also obtain a pointer to a 1D array directly with the unary & operator:
int arr[10];
int (*p)[10] = &arr; // note presence of & operator here
but that's not very common.
Another use for pointers to arrays is when we want to dynamically allocate a contiguous 2D (or higher-dimensioned) array:
int (*p)[10] = malloc( sizeof *p * 5 ); // allocates space for a 5x10 array of int
The type of the expression *p is "10-element array of int", or int [10], so sizeof *p is equivalent to sizeof (int [10]), so the malloc call sets aside enough space for 5 10-element arrays of int. You would index into this array like any other 2D array: p[i][j] = some_value();. The rows of this array are contiguous in memory and all the same length, and it only requires a single call to free to deallocate.
You've probably seen code that dynamically allocates 2D arrays in multiple steps - first you allocate an array of pointers, then for each of those pointers you allocate an array of the target type, like so:
int **p = malloc( sizeof *p * R ); // allocates an R-element array of pointer to int
if ( p )
{
for ( size_t i = 0; i < R; i++ )
{
p[i] = malloc( sizeof *p[i] * C ); // allocates a C-element array of int
}
}
The difference here is that in the second case, the rows are not adjacent in memory - it's not a single, continuous block. Also, different rows may be different lengths (what's sometimes called a "jagged" array). You also have to free each row separately before freeing p.
The basic rules of pointer declarations are:
T *p; // p is a pointer to T
T *a[N]; // a is an array of pointer to T
T *f(); // f is a function returning pointer to T
T (*a)[N]; // a is a pointer to an N-element array of T
T (*f)(); // f is a pointer to a function returning T
T const *p; // p is a non-const pointer to const T - you can modify p to point
// to a different object, but you cannot modify the thing p points to
const T *p; // same as above
T * const p; // p is a const pointer to non-const T - you can modify what p
// p points to, but you can't modify p to point to a different object
// (which means p needs to be initialized to a valid pointer value
// when it's declared).
T const * const p; // p is a const pointer to const T - you can't update either
// p or *p
const T * const p; // same as above.
If in this specification
int (*arr1)[10]
you will remove the term in the parentheses you will get int[10]. The term in the parentheses denotes a pointer. So you have a pointer to an array of 10 elements of the type int. For example
int a[10];
int ( *arr )[10] = &a;
Dereferencing the pointer you will get the array (the object of the array type pointed to by the pointer) itself.
Here is a demonstrative program.
#include <stdio.h>
int main(void)
{
enum { N = 10 };
int a[N] = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 };
int ( *arr )[N] = &a;
for ( size_t i = 0; i < N; i++ )
{
printf( "%d ", ( *arr )[i] );
}
putchar( '\n' );
return 0;
}
The program output is
0 1 2 3 4 5 6 7 8 9
Instead of these declarations you could use declarations based on a typedef like
enum { N = 10 };
typedef int Array[N];
Array a = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 };
Array *arr = &a;
I tried to assign two fixed-size arrays to an array of pointers to them, but the compiler warns me and I don't understand why.
int A[5][5];
int B[5][5];
int*** C = {&A, &B};
This code compiles with the following warning:
warning: initialization from incompatible pointer type [enabled by default]
If I run the code, it will raise a segmentation fault. However, if I dynamically allocate A and B, it works just fine. Why is this?
If you want a declaration of C that fits the existing declarations of A and B you need to do it like this:
int A[5][5];
int B[5][5];
int (*C[])[5][5] = {&A, &B};
The type of C is read as "C is an array of pointers to int [5][5] arrays". Since you can't assign an entire array, you need to assign a pointer to the array.
With this declaration, (*C[0])[1][2] is accessing the same memory location as A[1][2].
If you want cleaner syntax like C[0][1][2], then you would need to do what others have stated and allocate the memory dynamically:
int **A;
int **B;
// allocate memory for A and each A[i]
// allocate memory for B and each B[i]
int **C[] = {A, B};
You could also do this using the syntax suggested by Vlad from Moscow:
int A[5][5];
int B[5][5];
int (*C[])[5] = {A, B};
This declaration of C is read as "C is an array of pointers to int [5] arrays". In this case, each array element of C is of type int (*)[5], and array of type int [5][5] can decay to this type.
Now, you can use C[0][1][2] to access the same memory location as A[1][2].
This logic can be expanded to higher dimensions as well:
int A[5][5][3];
int B[5][5][3];
int (*C[])[5][3] = {A, B};
Unfortunately there's a lot of crappy books/tutorials/teachers out there who will teach you wrong things....
Forget about pointer-to-pointers, they have nothing to do with arrays. Period.
Also as a rule of thumb: whenever you find yourself using more than 2 levels of indirection, it most likely means that your program design is fundamentally flawed and needs to be remade from scratch.
To do this correctly, you would have to do like this:
A pointer to an array int [5][5] is called array pointer and is declared as int(*)[5][5]. Example:
int A[5][5];
int (*ptr)[5][5] = &A;
If you want an array of array pointers, it would be type int(*[])[5][5]. Example:
int A[5][5];
int B[5][5];
int (*arr[2])[5][5] = {&A, &B};
As you can tell this code looks needlessly complicated - and it is. It will be a pain to access the individual items, since you will have to type (*arr[x])[y][z]. Meaning: "in the array of array pointers take array pointer number x, take the contents that it points at - which is a 2D array - then take item of index [y][z] in that array".
Inventing such constructs is just madness and nothing I would recommend. I suppose the code can be simplified by working with a plain array pointer:
int A[5][5];
int B[5][5];
int (*arr[2])[5][5] = {&A, &B};
int (*ptr)[5][5] = arr[0];
...
ptr[x][y][z] = 0;
However, this is still somewhat complicated code. Consider a different design entirely! Examples:
Make a 3D array.
Make a struct containing a 2D array, then create an array of such structs.
There is a lot wrong with the line
int*** C = {&A, &B};
You're declaring a single pointer C, but you're telling it to point to multiple objects; that won't work. What you need to do is declare C as an array of pointers to those arrays.
The types of both &A and &B are int (*)[5][5], or "pointer to 5-element array of 5-element array of int"; thus, the type of C needs to be "array of pointer to 5-element array of 5-element array of int", or
int (*C[2])[5][5] = { &A, &B };
which reads as
C -- C is a
C[2] -- 2-element array of
*C[2] -- pointers to
(*C[2])[5] -- 5-element arrays of
(*C[2])[5][5] -- 5-element arrays of
int (*C[2])[5][5] -- int
Yuck. That's pretty damned ugly. It gets even uglier if you want to access an element of either A or B through C:
int x = (*C[0])[i][j]; // x = A[i][j]
int y = (*C[1])[i][j]; // y = B[i][j]
We have to explicitly dereference C[i] before we can index into the array it points to, and since the subscript operator [] has higher precedence than the unary * operator, we need to group *C[0] in parens.
We can clean this up a little bit. Except when it is the operand of the sizeof or unary & operators (or is a string literal being used to initialize another array in a declaration), an expression of type "N-element array of T" will be converted ("decay") to an expression of type "pointer to T", and the value of the expression will be the address of the first element of the array.
The expressions A and B have type int [5][5], or "5-element array of 5-element array of int". By the rule above, both expressions "decay" to expressions of type "pointer to 5-element array of int", or int (*)[5]. If we initialize the array with A and B instead of &A and &B, then we need an array of pointers to 5-element arrays of int, or
int (*C[2])[5] = { A, B };
Okay, that's still pretty eye-stabby, but that's as clean as this is going to get without typedefs.
So how do we access elements of A and B through C?
Remember that the array subscript operation a[i] is defined as *(a + i); that is, given a base address a, offset i elements (not bytes)1 from that address and dereference the result. This means that
*a == *(a + 0) == a[0]
Thus,
*C[i] == *(C[i] + 0) == C[i][0]
Putting this all together:
C[0] == A // int [5][5], decays to int (*)[5]
C[1] == B // int [5][5], decays to int (*)[5]
*C[0] == C[0][0] == A[0] // int [5], decays to int *
*C[1] == C[1][0] == B[0] // int [5], decays to int *
C[0][i] == A[i] // int [5], decays to int *
C[1][i] == B[i] // int [5], decays to int *
C[0][i][j] == A[i][j] // int
C[1][i][j] == B[i][j] // int
We can index C as though it were a 3D array of int, which is a bit cleaner than (*C[i)[j][k].
This table may also be useful:
Expression Type "Decays" to Value
---------- ---- ----------- -----
A int [5][5] int (*)[5] Address of A[0]
&A int (*)[5][5] Address of A
*A int [5] int * Value of A[0] (address of A[0][0])
A[i] int [5] int * Value of A[i] (address of A[i][0])
&A[i] int (*)[5] Address of A[i]
*A[i] int Value of A[i][0]
A[i][j] int Value of A[i][j]
Note that A, &A, A[0], &A[0], and &A[0][0] all yield the same value (the address of an array and the address of the first element of the array are always the same), but the types are different, as shown in the table above.
Pointer arithmetic takes the size of the pointed-to type into account; if p contains the address of an int object, then p+1 yields the address of the next int object, which may be 2 to 4 bytes away.
A common misconception among C beginners is that they just assume pointers and arrays are equivalent. That's completely wrong.
Confusion comes to beginners when they see the code like
int a1[] = {1,2,3,4,5};
int *p1 = a1; // Beginners intuition: If 'p1' is a pointer and 'a1' can be assigned
// to it then arrays are pointers and pointers are arrays.
p1[1] = 0; // Oh! I was right
a1[3] = 0; // Bruce Wayne is the Batman! Yeah.
Now, it is verified by the beginners that arrays are pointers and pointers are arrays so they do such experiments:
int a2[][5] = {{0}};
int **p2 = a2;
And then a warning pops up about incompatible pointer assignment then they think: "Oh my God! Why has this array become Harvey Dent?".
Some even goes to one step ahead
int a3[][5][10] = {{{0}}};
int ***p3 = a3; // "?"
and then Riddler comes to their nightmare of array-pointer equivalence.
Always remember that arrays are not pointers and vice-versa. An array is a data type and a pointer is another data type (which is not array type). This has been addressed several years ago in the C-FAQ:
Saying that arrays and pointers are "equivalent" means neither that they are identical nor even interchangeable. What it means is that array and pointer arithmetic is defined such that a pointer can be conveniently used to access an array or to simulate an array. In other words, as Wayne Throop has put it, it's "pointer arithmetic and array indexing [that] are equivalent in C, pointers and arrays are different.")
Now always remember few important rules for array to avoid this kind of confusion:
Arrays are not pointers. Pointers are not arrays.
Arrays are converted to pointer to their first element when used in an expression except when an operand of sizeof and & operator.
It's the pointer arithmetic and array indexing that are same.
Pointers and arrays are different.
Did I say "pointers are not arrays and vice-versa".
Now you have the rules, you can conclude that in
int a1[] = {1,2,3,4,5};
int *p1 = a1;
a1 is an array and in the declaration int *p1 = a1; it converted to pointer to its first element. Its elements are of type int then pointer to its first element would be of type int * which is compatible to p1.
In
int a2[][5] = {{0}};
int **p2 = a2;
a2 is an array and in int **p2 = a2; it decays to pointer to its first element. Its elements are of type int[5] (a 2D array is an array of 1D arrays), so a pointer to its first element would be of type int(*)[5] (pointer to array) which is incompatible with type int **. It should be
int (*p2)[5] = a2;
Similarly for
int a3[][5][10] = {{{0}}};
int ***p3 = a3;
elements of a3 is of type int [5][10] and pointer to its first element would be of type int (*)[5][10], but p3 is of int *** type, so to make them compatible, it should be
int (*p3)[5][10] = a3;
Now coming to your snippet
int A[5][5];
int B[5][5];
int*** C = {&A, &B};
&A and &B are of type int(*)[5][5]. C is of type int***, it's not an array. Since you want to make C to hold the address of both the arrays A and B, you need to declare C as an array of two int(*)[5][5] type elements. This should be done as
int (*C[2])[5][5] = {&A, &B};
However, if I dynamically allocate A and B it works just fine. Why is this?
In that case you must have declared A and B as int **. In this case both are pointers, not arrays. C is of type int ***, so it can hold an address of int** type data. Note that in this case the declaration int*** C = {&A, &B}; should be
int*** C = &A;
In case of int*** C = {&A, &B};, the behavior of program would be either undefined or implementation defined.
C11: 5.1.1.3 (P1):
A conforming implementation shall produce at least one diagnostic message (identified in an implementation-defined manner) if a preprocessing translation unit or translation unit contains a violation of any syntax rule or constraint, even if the behavior is also explicitly specified as undefined or implementation-defined
Read this post for further explanation.
Arrays are not the same thing as multi-dimensional pointers in C. The name of the array gets interpreted as the address of the buffer that contains it in most cases, regardless of how you index it. If A is declared as int A[5][5], then A will usually mean the address of the first element, i.e., it is interpreted effectively as an int * (actually int *[5]), not an int ** at all. The computation of the address just happens to require two elements: A[x][y] = A + x + 5 * y. This is a convenience for doing A[x + 5 * y], it does not promote A to multidimensional buffer.
If you want multi-dimensional pointers in C, you can do that too. The syntax would be very similar, but it requires a bit more set up. There are a couple of common ways of doing it.
With a single buffer:
int **A = malloc(5 * sizeof(int *));
A[0] = malloc(5 * 5 * sizeof(int));
int i;
for(i = 1; i < 5; i++) {
A[i] = A[0] + 5 * i;
}
With a separate buffer for each row:
int **A = malloc(5 * sizeof(int *));
int i;
for(i = 0; i < 5; i++) {
A[i] = malloc(5 * sizeof(int));
}
You are being confused by the equivalence of arrays and pointers.
When you declare an array like A[5][5], because you have declared both dimensions, C will allocate memory for 25 objects contiguously. That is, memory will be allocated like this:
A00, A01, ... A04, A10, A11, ..., A14, A20, ..., A24, ...
The resulting object, A, is a pointer to the start of this block of memory. It is of type int *, not int **.
If you want a vector of pointers to arrays, you want to declare your variables as:
int *A[5], *B[5];
That would give you:
A0, A1, A2, A3, A4
all of type int*, which you would have to fill using malloc() or whatever.
Alternatively, you could declare C as int **C.
Although arrays and pointers are closely associated, they are not at all the same thing. People are sometimes confused about this because in most contexts, array values decay to pointers, and because array notation can be used in function prototypes to declare parameters that are in fact pointers. Additionally, what many people think of as array indexing notation really performs a combination of pointer arithmetic and dereferencing, so that it works equally well for pointer values and for array values (because array values decay to pointers).
Given the declaration
int A[5][5];
Variable A designates an array of five arrays of five int. This decays, where it decays, to a pointer of type int (*)[5] -- that is, a pointer to an array of 5 int. A pointer to the whole multi-dimensional array, on the other hand, has type int (*)[5][5] (pointer to array of 5 arrays of 5 int), which is altogether different from int *** (pointer to pointer to pointer to int). If you want to declare a pointer to a multi-dimensional array such as these then you could do it like this:
int A[5][5];
int B[5][5];
int (*C)[5][5] = &A;
If you want to declare an array of such pointers then you could do this:
int (*D[2])[5][5] = { &A, &B };
Added:
These distinctions come into play in various ways, some of the more important being the contexts where array values do not decays to pointers, and contexts related to those. One of the most significant of these is when a value is the operand of the sizeof operator. Given the above declarations, all of the following relational expressions evaluate to 1 (true):
sizeof(A) == 5 * 5 * sizeof(int)
sizeof(A[0]) == 5 * sizeof(int)
sizeof(A[0][4]) == sizeof(int)
sizeof(D[1]) == sizeof(C)
sizeof(*C) == sizeof(A)
Additionally, it is likely, but not guaranteed, that these relational expressions evaluate to 1:
sizeof(C) == sizeof(void *)
sizeof(D) == 2 * sizeof(void *)
This is fundamental to how array indexing works, and essential to understand when you are allocating memory.
Either you should declare the third array like
int A[5][5];
int B[5][5];
int ( *C[] )[N][N] = { &A, &B };
that is as an array of pointers to two-dimensional arrays.
For example
#include <stdio.h>
#define N 5
void output( int ( *a )[N][N] )
{
for ( size_t i = 0; i < N; i++ )
{
for ( size_t j = 0; j < N; j++ ) printf( "%2d ", ( *a )[i][j] );
printf( "\n" );
}
}
int main( void )
{
int A[N][N] =
{
{ 1, 2, 3, 4, 5 },
{ 6, 7, 8, 9, 10 },
{ 11, 12, 13, 14, 15 },
{ 16, 17, 18, 19, 20 },
{ 21, 22, 23, 24, 25 }
};
int B[N][N] =
{
{ 25, 24, 23, 22, 21 },
{ 20, 19, 18, 17, 16 },
{ 15, 14, 13, 12, 11 },
{ 10, 9, 8, 7, 6 },
{ 5, 4, 3, 2, 1 }
};
/*
typedef int ( *T )[N][N];
T C[] = { &A, &B };
*/
int ( *C[] )[N][N] = { &A, &B };
output( C[0] );
printf( "\n" );
output( C[1] );
printf( "\n" );
}
The program output is
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
16 17 18 19 20
21 22 23 24 25
25 24 23 22 21
20 19 18 17 16
15 14 13 12 11
10 9 8 7 6
5 4 3 2 1
or like
int A[5][5];
int B[5][5];
int ( *C[] )[N] = { A, B };
that is as an array of pointers to the first elements of two-dimensional arrays.
For example
#include <stdio.h>
#define N 5
void output( int ( *a )[N] )
{
for ( size_t i = 0; i < N; i++ )
{
for ( size_t j = 0; j < N; j++ ) printf( "%2d ", a[i][j] );
printf( "\n" );
}
}
int main( void )
{
int A[N][N] =
{
{ 1, 2, 3, 4, 5 },
{ 6, 7, 8, 9, 10 },
{ 11, 12, 13, 14, 15 },
{ 16, 17, 18, 19, 20 },
{ 21, 22, 23, 24, 25 }
};
int B[N][N] =
{
{ 25, 24, 23, 22, 21 },
{ 20, 19, 18, 17, 16 },
{ 15, 14, 13, 12, 11 },
{ 10, 9, 8, 7, 6 },
{ 5, 4, 3, 2, 1 }
};
/*
typedef int ( *T )[N];
T C[] = { A, B };
*/
int ( *C[] )[N] = { A, B };
output( C[0] );
printf( "\n" );
output( C[1] );
printf( "\n" );
}
The program output is the same as above
1 2 3 4 5
6 7 8 9 10
11 12 13 14 15
16 17 18 19 20
21 22 23 24 25
25 24 23 22 21
20 19 18 17 16
15 14 13 12 11
10 9 8 7 6
5 4 3 2 1
depending on how you are going to use the third array.
Using typedefs (shown in the demonstrative program as commented) ssimplifies the arrays' definitions.
As for this declaration
int*** C = {&A, &B};
then in the left side there is declared a pointer of type int *** that is a scalar object while in the right side there is a list of initializers that have different type int ( * )[N][N].
So the compiler issues a message.
I am a great believer in using typedef:
#define SIZE 5
typedef int OneD[SIZE]; // OneD is a one-dimensional array of ints
typedef OneD TwoD[SIZE]; // TwoD is a one-dimensional array of OneD's
// So it's a two-dimensional array of ints!
TwoD a;
TwoD b;
TwoD *c[] = { &a, &b, 0 }; // c is a one-dimensional array of pointers to TwoD's
// That does NOT make it a three-dimensional array!
int main() {
for (int i = 0; c[i] != 0; ++i) { // Test contents of c to not go too far!
for (int j = 0; j < SIZE; ++j) {
for (int k = 0; k < SIZE; ++k) {
// c[i][j][k] = 0; // Error! This proves it's not a 3D array!
(*c[i])[j][k] = 0; // You need to dereference the entry in c first
} // for
} // for
} // for
return 0;
} // main()
I was reading through some lecture notes that in order for a pointer to reference a 2D array, it has to be given the address of the first element.
int a[10][10];
int *p = &a[0][0];
I've never tried this, so I was curious why isn't it enough to assign the array itself to the pointer, just as we do in a 1D case.
int a[10][10];
int *p = a;
The array is kept in an uninterrupted 'line' of memory anyway, and 2D arrays only have a different type, but the same structure as 1D arrays.
By doing this
int *p = &a[0][0];
I don't see how we give the pointer any more information than by doing this
int *p = a;
Or maybe all arrays regardless of their number of dimensions have the same type, the only difference being that multidimensional arrays store their extra dimensions before their first element and we need to jump over those memory spaces which remember sizes of an array's dimensions?
First, some background:
Except when it is the operand of the sizeof or unary & operators, or is a string literal being used to initialize another array in a declaration, an expression of type "N-element array of T" will be converted ("decay") to an expression of type "pointer to T", and the value of the expression will be the address of the first element of the array.
Given the declaration
int a[10][10];
the expression a has type "10-element array of 10-element array of int". Unless this expression is the operand of the sizeof or unary & operators, it will be converted to an expression of type "pointer to 10-element array of int", or int (*)[10].
Given that declaration, all of the following are true:
Expression Type Decays to
---------- ---- ---------
a int [10][10] int (*)[10]
&a int (*)[10][10]
*a int [10] int *
a[i] int [10] int *
&a[i] int (*)[10]
*a[i] int
a[i][j] int
&a[i][j] int *
Also,
sizeof a == sizeof (int) * 10 * 10
sizeof &a == sizeof (int (*)[10][10])
sizeof *a == sizeof (int) * 10
sizeof a[i] == sizeof (int) * 10
sizeof &a[i] == sizeof (int (*)[10] )
sizeof *a[i] == sizeif (int)
sizeof a[i][j] == sizeof (int)
sizeof &a[i][j] == sizeof (int *)
Note that the different pointer types int (*)[10][10], int (*)[10], and int * don't have to be the same size or have the same representation, although on the platforms I'm familiar with they do.
The address of the first element of the array is the same as the address of the array itself; thus, all of a, &a, a[0], &a[0], and &a[0][0] will yield the same value, but the types will be different (as shown in the table above).
So, assume we add the following declarations:
int *p0 = &a[0][0]; // equivalent to int *p0 = a[0];
int (*p1)[10] = &a[0]; // equivalent to int (*p1)[10] = a;
int (*p2)[10][10] = &a;
All of p0, p1, and p2 initially have the same value, which is the address of the first element in a; however, because of the different pointer types, the results operations involving pointer arithmetic will be different. The expression p0 + 1 will yield the address of the next int object (&a[0][1]). The expression p1 + 1 will yield the address of the next 10-element array of int (&a[1][0]). And finally, the expression p2 + 1 will yield the address of the next 10-element array of 10-element array of int (effectively, &a[11][0]).
Note the types of p1 and p2; neither is a simple int *, because the expressions being used to initialize them are not that type (refer to the first table).
Note the pattern; for an array type, the simpler the expression, the more complicated the corresponding type will be. The expression a does not refer to a single int object; it refers to a 10x10 array of int objects, so when it appears in an expression, it is treated as a pointer to an array of integers, not a pointer to a single integer.
The compiler knows that "a" is a pointer to ten integers. If you don't declare the dimensions, then the compiler sees the new pointer as a pointer to an unknown number of integers. This will work in your case, but it will generate a compiler warning because the compiler sees them as incompatible pointers. The syntax for what you are trying to do (without generating a compiler warning) is:
int a[10][10];
int *p1 = &a[0][0];
int (*p2)[10] = a;
printf("p1: %p p2: %p\n", p1, p2);
One reason this is important is pointer arithmetic:
p1++; //move forward sizeof(int) bytes
p2++; //move forward sizeof(int) * 10 bytes
You understanding is close, the difference is the type information. Pointer does has its type. For example int* p, the pointer type is int*, as int a[10][10], the corresponding pointer type is int *[10][10].
In your example, p and a do point to the same address, but they're different type, which matters when perform arithmetic operation on them.
Here's an example from this URL
Suppose now that we define three pointers :
char *mychar;
short *myshort;
long *mylong;
and that we know that they point to the memory locations 1000, 2000, and 3000, respectively.
Therefore, if we write:
++mychar;
++myshort;
++mylong;
mychar, as one would expect, would contain the value 1001. But not so obviously, myshort would contain the value 2002, and mylong would contain 3004, even though they have each been incremented only once. The reason is that, when adding one to a pointer, the pointer is made to point to the following element of the same type, and, therefore, the size in bytes of the type it points to is added to the pointer.
You are right, you can assign the array itself to the pointer:
int a[10][10] = {[0][0]=6,[0][1]=1,[1][0]=10,[1][1]=11};
int b[10][10][10] = {[0][0][0]=8,[0][0][1]=1,[0][1][0]=10,[1][0][0]=100};
int *p, *q, *r, *s;
p = &a[0][0];
q = a; // what you are saying
r = &b[0][0][0];
s = b; // what you are saying
printf("p= %p,*p= %d\n",p,*p);
printf("q= %p,*q= %d\n",q,*q);
printf("r= %p,*r= %d\n",r,*r);
printf("s= %p,*s= %d\n",s,*s);
And the output is:
p= 0xbfdd2eb0,*p= 6
q= 0xbfdd2eb0,*q= 6
r= 0xbfdd3040,*r= 8
s= 0xbfdd3040,*s= 8
They point to the same address, regardless of the dimension of the matrix. So, what you are saying is right.
Well in 2D array, the outcome of *a and a is the same, they all point to the first address of this 2D array!
But if you want to define a pointer to point to this array, you could use int (*ptr)[10] for example.
You are right, 1D and 2D share the same structure, but 2D has some additional manipulation on pointers like above.
So all in all, in 2D array, a, *a and &a[0][0] prints the same address, but their usages may vary.
Like this:
#include<stdio.h>
int main() {
int a[10][10];
int *pa1 = &a[0][0];
int *pa2 = *a;
printf("pa1 is %p\n", pa1);
printf("pa2 is %p\n", pa2);
printf("Address of a is %p\n", a);
// pointer to array
int (*pa3)[10];
pa3 = a;
printf("pa3 is %p\n", pa3);
return 0;
}
They print the same address.