# C99

> C++ Atlas: `cppatlas.com/w/c/99` | Source: https://en.cppreference.com/w/c/99 | Domain: Standards & support

ISO/IEC 9899:1999, a.k.a. C99, is a previous revision of the C standard.

## Official Specification Summary (cppreference)

**ISO/IEC 9899:1999**, a.k.a. **C99**, is a previous revision of the C standard.

> Note: This section is incomplete Reason: core/library features, TC1~TC3, DRs

### Obsolete

#### Removed

- Implicit `int` in declarations
- Implicit function declarations

#### Deprecated

- `gets()`

### New language features

- Universal-character-names in identifiers
- Increased translation limits
- `//` comments
- `restrict` pointers
- Enhanced arithmetic types
  - `_Bool`
  - `long long` and `unsigned long long`
  - Extended integer types
  - `_Complex` types (`float _Complex`, `double _Complex`, and `long double _Complex`)
  - `_Imaginary` types (`float _Imaginary`, `double _Imaginary`, and `long double _Imaginary`)
- Flexible array members
- Variable-length array (VLA) types and variably-modified (VM) types
- Improvements of *braced-init-list* for array, struct and union types
  - Non-constant initializers
  - Designated initializers
- Idempotent cvr-qualifiers
- Trailing comma in *enumerator-list*
- Hexadecimal floating constants
- Compound literals
- Floating-point environment
- Requiring truncation for divisions of signed integer types
- Implicit `return 0;` in the `main()` function
- Declarations and statements in mixed order
- *init-statement* in `for` loops
- `inline` functions
- Predefined variable `__func__`
- Cvr-qualifiers and `static` in `[]` within function declarations
- Variadic macros
- `_Pragma` preprocessor operator
- Standard pragmas for floating-point evaluation
  - `STDC` `FENV_ACCESS`
  - `STDC` `FP_CONTRACT`
  - `STDC` `CX_LIMITED_RANGE`

#### Feature test macros for optional features

- `__STDC_IEC_559__`
  - Indicates IEEE-754 binary floating-point arithmetic and required math functions are supported.
- `__STDC_IEC_559_COMPLEX__`
  - Indicates IEEE-754 complex arithmetic and required math functions are supported.
- `__STDC_HOSTED__`
  - Indicates that the implementation is hosted.
- `__STDC_ISO_10646__`
  - Indicates that Unicode is used by the wide literal encodings and expands to the latest supported revision.
- `__STDC_MB_MIGHT_NEQ_WC__`
  - Indicates that there are some characters in the basic character set having different code unit values in ordinary and wide literal encodings.

### New library features

#### New headers

- `<complex.h>`
- `<fenv.h>`
- `<inttypes.h>`
- `<stdbool.h>`
- `<stdint.h>`
- `<tgmath.h>`

#### Library features

- Aliases for integer types
  - Integer types with exact width (`intN_t` and `uintN_t`)
  - Fastest integer types with at least given width (`int_fastN_t` and `uint_fastN_t`)
  - Smallest integer types with at least given width (`int_leastN_t` and `uint_leastN_t`)
  - Integer types capable for cast between object pointers (`intptr_t` and `uintptr_t`)
  - Integer types with maximum width (`intmax_t` and `uintmax_t`)
- Operations on `long long` and `intmax_t`
  - `llabs()`
  - `imaxabs()`
  - `lldiv()`
  - `imaxdiv()`
- Floating-point environment access
- Extended floating-point math functions
  - New floating-point math functions
  - -`f` and -`l` variants for existing and new floating-point math functions
  - Math error handling
- Complex functions
- `_Exit()`
- Formatting support for `long long`, `unsigned long long`, `intmax_t`, and `uintmax_t`
  - `atoll()`
  - `strtoimax()`
  - `strtoll()`
  - `strtoull`
  - `strtoumax()`
  - `wcstoimax()`
  - `wcstoll()`
  - `wcstoull()`
  - `wcstoumax()`
- `isblank()` and `iswblank()`
- `snprintf()` and `vsnprintf()`
- `vfscanf()` and `vfwscanf()` function families
- Extensions for `fscanf()` and `fprintf()` function families
  - `ll` length modifier for `long long` and `unsigned long long`
  - `hh` length modifier for `signed char` and `unsigned char`
  - `l` length modifier for `double`
  - `z` length modifier for `size_t` and its signed version
  - `t` length modifier for `ptrdiff_t` and its unsigned version
  - `j` length modifier for `intmax_t` and `uintmax_t`
  - `a` conversion specifier for floating-point types
- Numeric limit macros for `long long` and `unsigned long long`
- Numeric limit macros corresponding to existing and new aliases for integer types
- Format string macros for integer types
- `va_copy`
- Type-generic math macros
- Floating-point comparison macros
- Floating-point classification macros
- Compatibility macros for `_Bool` (`bool`, `true`, and `false`)

### Defect reports

Template:c/language/history/DR99

### Compiler support

#### C99 core language features

> Note: This section is incomplete Reason: needs to list C compilers, verification

| C99 feature | Paper(s) | GCC | Clang | MSVC | Apple Clang | EDG eccp | Intel C++ | Nvidia HPC C++ (ex PGI)* | Nvidia nvcc | Cray |   |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Universal-character-names in identifiers |   | 3.1 | Yes | Yes |   |   |   |   |   |   |   |
| Increased translation limits | N590 | 0.9 | N/A |   |   |   |   |   |   |   |   |
| `//` comments | N644 | 2.7 | Yes | Yes |   |   |   |   |   |   |   |
| `restrict` pointers | N448 | 2.95 | Yes | partial* |   |   |   |   |   |   |   |
| Enhanced arithmetic types | N815 N601 N620 N638 N657 N694 N809 | Yes | partial | Maybe |   |   |   |   |   |   |   |
| Flexible array members |   | 3.0 | Yes | Yes |   |   |   |   |   |   |   |
| Variable-length array (VLA) types | N683 | 0.9 | Yes |   |   |   |   |   |   |   |   |
| Variably-modified (VM) types | N2778 | N/A | Yes |   |   |   |   |   |   |   |   |
| Designated initializers | N494 | 3.0 | Yes | Yes |   |   |   |   |   |   |   |
| Non-constant initializers |   | 1.21 | N/A |   |   |   |   |   |   |   |   |
| Idempotent cvr-qualifiers | N505 | 3.0 | N/A |   |   |   |   |   |   |   |   |
| Trailing comma in *enumerator-list* |   | 0.9 | Yes | Yes |   |   |   |   |   |   |   |
| Hexadecimal floating constants | N308 | 2.8 | Yes | Yes |   |   |   |   |   |   |   |
| Compound literals | N716 | 3.1 | Yes | Yes |   |   |   |   |   |   |   |
| Floating-point environment |   | partial | partial |   |   |   |   |   |   |   |   |
| Requiring truncation for divisions of signed integer types | N617 | 0.9 | N/A |   |   |   |   |   |   |   |   |
| Implicit `return 0;` in the `main()` function |   | Yes | Yes | Yes |   |   |   |   |   |   |   |
| Declarations and statements in mixed order | N740 | 3.0 | Yes | Yes |   |   |   |   |   |   |   |
| *init-statement* in `for` loops |   | Yes | Yes | Yes |   |   |   |   |   |   |   |
| `inline` functions | N741 | 4.3 | Yes | Yes |   |   |   |   |   |   |   |
| Predefined variable `__func__` | N611 | 2.95 | Yes | Yes |   |   |   |   |   |   |   |
| Cvr-qualifiers and `static` in `[]` within function declarations |   | 3.1 | Yes |   |   |   |   |   |   |   |   |
| Variadic macros | N707 | 2.95 | Yes | Yes |   |   |   |   |   |   |   |
| `_Pragma` preprocessor operator | N634 | 3.0 | Yes | partial* |   |   |   |   |   |   |   |
| Standard pragmas for floating-point evaluation | N631 N696 | No | No |   |   |   |   |   |   |   |   |
| C99 feature | Paper(s) | GCC | Clang | MSVC | Apple Clang | EDG eccp | Intel C++ | Nvidia HPC C++ (ex PGI)* | Nvidia nvcc | Cray |   |

---
Attribution: content from https://en.cppreference.com/w/c/99 (cppreference.com), licensed under CC-BY-SA 3.0 and GFDL. Modified: converted to markdown, study notes added. Offline snapshot of 9 February 2025.