[libc-commits] [libc] [libc][complex] fix compiler support matrix for cfloat128 (PR #122593)

[via libc-commits]

================
@@ -18,25 +18,19 @@
 //
 // TODO: Update the complex variant of C23 `_Float128` type detection again when
 // clang supports it.
-#if defined(__STDC_IEC_60559_COMPLEX__) && !defined(__clang__)
-#if !defined(__cplusplus)
-#define LIBC_TYPES_HAS_CFLOAT128
-typedef _Complex _Float128 cfloat128;
-#elif defined(__GNUC__) && __GNUC__ >= 13
-#define LIBC_TYPES_HAS_CFLOAT128
-typedef _Complex _Float128 cfloat128;
-#endif
-#elif __clang_major__ >= 11 &&                                                 \
+#if defined(__clang__) && (__clang_major__ >= 11) &&                           \
----------------
lntue wrote:

Now that I think about it, probably we should move the long double to a separate clause:
```
#ifdef __clang__
...
#elif defined (__GNUC__)
...
#endif

#if !defined(LIBC_TYPES_HAS_CFLOAT128) && (LDBL_MANT_DIG == 113)
...
#endif
```

https://github.com/llvm/llvm-project/pull/122593