Data types (x86-64)

The C language standard does not specify in absolute terms the size of its primitive data types. The information in this chapter applies specifically to the following software and hardware:

Compiler: GNU Compiler Collection
Operating system: Linux
Architecture: x86-64

The code to generate the information in this chapter can be obtained here.

Integers

It is advisable to use stdint.h as a replacement for the sake of specificity.

Signed

Name	Range		Size	Literal	Constants		Format specifier
Name	Range		Size	Literal	Minimum	Maximum	Format specifier
`signed char`	−128	127	1 byte 8 bits	`(name)value`	`SCHAR_MIN`	`SCHAR_MAX`	`%hhi`
`short` `short int` `signed short` `signed short int`	−32,768	32,767	2 bytes 16 bits	`(name)value`	`SHRT_MIN`	`SHRT_MAX`	`%hi`
`int` `signed` `signed int`	−2,147,483,648	2,147,483,647	4 bytes 32 bits	`value`	`INT_MIN`	`INT_MAX`	`%i`
`long` `long int` `signed long` `signed long int`	−9,223,372,036,854,775,808	9,223,372,036,854,775,807	8 bytes 64 bits	Value suffix `l` or `L`	`LONG_MIN`	`LONG_MAX`	`%li`
`long long` `long long int` `signed long long` `signed long long int`	−9,223,372,036,854,775,808	9,223,372,036,854,775,807	8 bytes 64 bits	Value suffix `ll` or `LL`	`LLONG_MIN`	`LLONG_MAX`	`%lli`

Unsigned

Name	Range		Size	Literal	Constants		Format specifier
Name	Range		Size	Literal	Minimum	Maximum	Format specifier
`unsigned char`	0	255	1 byte 8 bits	`(name)value`	None	`UCHAR_MAX`	`%hhu`
`unsigned short` `unsigned short int`	0	65,535	2 bytes 16 bits	`(name)value`	None	`USHRT_MAX`	`%hu`
`unsigned` `unsigned int`	0	4,294,967,295	4 bytes 32 bits	Value suffix `u` or `U`	None	`UINT_MAX`	`%u`
`unsigned long` `unsigned long int`	0	18,446,744,073,709,551,615	8 bytes 64 bits	Value suffix `ul` or `UL`	None	`ULONG_MAX`	`%lu`
`unsigned long long` `unsigned long long int`	0	18,446,744,073,709,551,615	8 bytes 64 bits	Value suffix `ull` or `ULL`	None	`ULLONG_MAX`	`%llu`

Floating-point numbers

It is not advisable to use float_t and double_t which are declared in math.h as replacements.

Name	Range		Size	Literal	Constants		Format specifier
Name	Range		Size	Literal	Minimum	Maximum	`printf`¹	`scanf`²
`float`	1.175494E−38	3.402823E+38	4 bytes 32 bits	Value suffix `f` or `F`	`FLT_MIN`	`FLT_MAX`	None³	`%f`
`double`	2.225074E−308	1.797693E+308	8 bytes 64 bits	`value`	`DBL_MIN`	`DBL_MAX`	`%f`	`%lf`
`long double`	3.362103E−4932	1.189731E+4932	16 bytes 128 bits	Value suffix `l` or `L`	`LDBL_MIN`	`LDBL_MIN`	`%Lf`	`%Lf`

¹ Also fprintf, snprintf, sprintf, vfprintf, vprintf, vsnprintf and vsprintf.
² Also fscanf, sscanf, vfscanf, vscanf and vsscanf.
³ float is promoted to double (%f).

Character

The character char is an integer data type but is primarily used to represent typographical symbols.

Name	Range		Size	Literal	Constants		Format specifier
Name	Range		Size	Literal	Minimum	Maximum	Format specifier
`char`	−128	127	1 byte 8 bits	`'character_value'` or `integral_value`	`CHAR_MIN`	`CHAR_MAX`	None^*

^* char is the same as signed char (%hhi). Typographical representation: %c for character and %s for string (pointer to char).

Boolean

The boolean _Bool is an integer data type but the compiler will not allow it to be set to any value outside its range. It is advisable to use stdbool.h as a replacement for the sake of clarity.

Name	Range		Size	Literal	Format specifier
`_Bool`	0	1	1 byte 8 bits	`value`	None*

^* _Bool is promoted to int (%i).

Miscellaneous

stdbool.h

The stdbool.h header declares macros for the _Bool data type.

Name	Range		Size	Literal	Format specifier	Primitive equivalent
`bool`	`false`	`true`	1 byte 8 bits	`value`	None*	`_Bool`

^* bool is promoted to int (%i).

stddef.h

The stddef.h header declares a type definition for the return type of the sizeof operator.

Name	Range		Size	Literal	Constants		Format specifier	Primitive equivalent
Name	Range		Size	Literal	Minimum	Maximum	Format specifier	Primitive equivalent
`size_t`	0	18,446,744,073,709,551,615	8 bytes 64 bits	Value suffix `ul` or `UL`	None	`SIZE_MAX`	`%zu`	`unsigned long` `unsigned long int`

stdint.h

The stdint.h header declares type definitions and macros for the integer primitives.

Signed

Name	Range		Size	Literal macro	Constants		Format specifier macro		Primitive equivalent
Name	Range		Size	Literal macro	Minimum	Maximum	`printf`¹	`scanf`²	Primitive equivalent
`int8_t`	−128	127	1 byte 8 bits	`INT8_C(value)`	`INT8_MIN`	`INT8_MAX`	`PRId8` / `PRIi8`	`SCNi8`	`signed char`
`int16_t`	−32,768	32,767	2 bytes 16 bits	`INT16_C(value)`	`INT16_MIN`	`INT16_MAX`	`PRId16` / `PRIi16`	`SCNi16`	`short` `short int` `signed short` `signed short int`
`int32_t`	−2,147,483,648	2,147,483,647	4 bytes 32 bits	`INT32_C(value)`	`INT32_MIN`	`INT32_MAX`	`PRId32` / `PRIi32`	`SCNi32`	`int` `signed` `signed int`
`int64_t`	−9,223,372,036,854,775,808	9,223,372,036,854,775,807	8 bytes 64 bits	`INT64_C(value)`	`INT64_MIN`	`INT64_MAX`	`PRId64` / `PRIi64`	`SCNi64`	`long` `long int` `signed long` `signed long int`

¹ Also fprintf, snprintf, sprintf, vfprintf, vprintf, vsnprintf and vsprintf.
² Also fscanf, sscanf, vfscanf, vscanf and vsscanf.

Unsigned

Name	Range		Size	Literal macro	Constants		Format specifier macro		Primitive equivalent
Name	Range		Size	Literal macro	Minimum	Maximum	`printf`¹	`scanf`²	Primitive equivalent
`uint8_t`	0	255	1 byte 8 bits	`UINT8_C(value)`	None	`UINT8_MAX`	`PRIu8`	`SCNu8`	`unsigned char`
`uint16_t`	0	65,535	2 bytes 16 bits	`UINT16_C(value)`	None	`UINT16_MAX`	`PRIu16`	`SCNu16`	`unsigned short` `unsigned short int`
`uint32_t`	0	4,294,967,295	4 bytes 32 bits	`UINT32_C(value)`	None	`UINT32_MAX`	`PRIu32`	`SCNu32`	`unsigned` `unsigned int`
`uint64_t`	0	18,446,744,073,709,551,615	8 bytes 64 bits	`UINT64_C(value)`	None	`UINT64_MAX`	`PRIu64`	`SCNu64`	`unsigned long` `unsigned long int`

¹ Also fprintf, snprintf, sprintf, vfprintf, vprintf, vsnprintf and vsprintf.
² Also fscanf, sscanf, vfscanf, vscanf and vsscanf.