your c is not standard c · Listed Articles

An incomplete list of incompatibilities between C as written in the ISO standard and C as it exists for people using gcc/clang with glibc/musl, covering the asm keyword, FLT_ROUNDS, standard floating-point pragmas, and other places where real toolchains are not conformant by default.

"The committee has created an unreal language that no one can or will actually use."

Dennis Ritchie (out of context)

this is an incomplete list of some incompatibilities between c as it's written in the standard and c as it exists in the real world for people who use gcc/clang + glibc/musl.

`asm`

by default, gcc and clang recognize the asm keyword as an extension. this is not strictly conformant, because in standard c, asm is a valid non-reserved identifier. so code which conforms to the standard may not compile with gcc/clang with their default settings.

and like, whatever, that's fine, that's why -std exists, but 1. it's notable that compilers aren't conformant by default, and 2. why doesn't the standard just make asm a reserved word? c++ does that; i'm not sure why c doesn't.

`FLT_ROUNDS`

the standard specifies the FLT_ROUNDS macro in float.h as follows:

5.3.5.3.3p23:

The rounding mode for floating-point addition for standard floating types is characterized by the implementation-defined value of FLT_ROUNDS. Evaluation of FLT_ROUNDS correctly reflects any execution-time change of rounding mode through the function fesetround in <fenv.h>.

notably, the standard requires that the value dynamically changes with the runtime floating environment. gcc, however, always defines FLT_ROUNDS to 1 (rounds to nearest), regardless of the runtime rounding mode:

/* Addition rounds to 0: zero, 1: nearest, 2: +inf, 3: -inf, -1: unknown.  */
/* ??? This is supposed to change with calls to fesetround in <fenv.h>.  */
#undef FLT_ROUNDS
#define FLT_ROUNDS 1

clang defines a builtin for this (__builtin_flt_rounds()), but gcc doesn't, for some reason.

standard pragmas

at some point in the future i'm gonna make an entire blog post dedicated to just this, because i kinda find it fascinating. but the super duper short version is this:

the c standard specifies five standard pragmas to control the behavior of floating point arithmetic:

// added in c99:
// switch can be ON, OFF, or DEFAULT
#pragma STDC FP_CONTACT ON
#pragma STDC FENV_ACCESS ON
#pragma STDC CX_LIMITED_RANGE ON

// added in c23:
// direction can be FE_DYNAMIC, FE_DOWNWARD, FE_TONEAREST,
// FE_TONEARESTFROMZERO, FE_TOWARDZERO, FE_UPWARD, or some
// other implementation-defined value
#pragma STDC FENV_ROUND FE_DYNAMIC
#pragma STDC FENV_DEC_ROUND FE_DYNAMIC

however, over 27 years after c99 was standardized, gcc still doesn't support any of the standard pragmas. they're just no-ops in gcc.

some (but not all) of the functionality can be replicated with command-line flags, but those operate on the entire translation unit, whereas the pragmas can apply to individual functions and blocks.

clang mostly supports the c99 pragmas, but still doesn't support the c23 pragmas.

there's actually so much more to say about these. i will be writing more about them at some point, but rn i don't really have the motivation cuz it's such a big thing lol

`initstate` error behavior

this one is a posix extension, so not necessarily "standard c", more like "posix c", but it's still a documented incompatibility with a standard: in glibc, initstate doesn't return NULL on error, despite posix requiring that it does:

// EINVAL: state array less than 8 bytes
char *state = initstate(67, (char [1]){}, 1);
assert(errno == EINVAL); // ok
assert(state == NULL); // fails, despite posix requiring it to succeed

the man page initstate(3) very unhelpfully first describes the posix behavior in the RETURN VALUE section, as though it's the actual behavior:

The initstate() function returns a pointer to the previous state array. On failure, it returns NULL, and errno is set to indicate the error.

only at the very bottom in the BUGS section does it say that glibc behaves differently:

According to POSIX, initstate() should return NULL on error. In the glibc implementation, errno is (as specified) set on error, but the function does not return NULL.

qualifiers in function return types

the standard requires that qualifiers in function return types are discarded:

If, in the declaration "T D1", D1 has the form
D ( parameter-type-listopt ) attribute-specifier-sequenceopt
and the type specified for ident in the declaration "T D" is "derived-declarator-type-list T", then the type specified for ident is "derived-declarator-type-list function returning the unqualified, non-atomic version of T". The optional attribute specifier appertains to the function type.

but clang treats them as significant:

$ cat t.c
int f();
const int f();

$ clang t.c
t.c:2:11: error: conflicting types for 'f'
    2 | const int f();
      |           ^
t.c:1:5: note: previous declaration is here
    1 | int f();
      |     ^
1 error generated.

the two declarations should be identical, but clang thinks they're incompatible.

outdated `va_start` definition on musl

for some reason, musl provides its own versions of some headers which the compiler already provides itself. one of these is stdarg.h, which is required to be supported by freestanding implementations and so is virtually guaranteed to already be provided by the compiler. but because it's provided by musl, compilers will use that header instead of the builtin one.

this is a problem because musl's header is strictly worse than the builtin header. here's how musl defines va_start:

#define va_start(v,l)   __builtin_va_start(v,l)

this definition is incorrect, because c23 allows the second argument to be omitted. it should be defined like this:

#if __STDC_VERSION__ >= 202311L
#define va_start(...) __builtin_c23_va_start(__VA_ARGS__)
#else
#define va_start(v,l) __builtin_va_start(v,l)
#endif

the result is that, when compiling for musl, c23's single-argument va_start can't be used, even if it's supported by the compiler.