================ @@ -665,28 +665,30 @@ KEYWORD(__kindof , KEYOBJC) // Alternate spelling for various tokens. There are GCC extensions in all // languages, but should not be disabled in strict conformance mode. -ALIAS("__alignof__" , __alignof , KEYALL) -ALIAS("__asm" , asm , KEYALL) -ALIAS("__asm__" , asm , KEYALL) -ALIAS("__attribute__", __attribute, KEYALL) -ALIAS("__complex" , _Complex , KEYALL) -ALIAS("__complex__" , _Complex , KEYALL) -ALIAS("__const" , const , KEYALL) -ALIAS("__const__" , const , KEYALL) -ALIAS("__decltype" , decltype , KEYCXX) -ALIAS("__imag__" , __imag , KEYALL) -ALIAS("__inline" , inline , KEYALL) -ALIAS("__inline__" , inline , KEYALL) -ALIAS("__nullptr" , nullptr , KEYCXX) -ALIAS("__real__" , __real , KEYALL) -ALIAS("__restrict" , restrict , KEYALL) -ALIAS("__restrict__" , restrict , KEYALL) -ALIAS("__signed" , signed , KEYALL) -ALIAS("__signed__" , signed , KEYALL) -ALIAS("__typeof" , typeof , KEYALL) -ALIAS("__typeof__" , typeof , KEYALL) -ALIAS("__volatile" , volatile , KEYALL) -ALIAS("__volatile__" , volatile , KEYALL) +ALIAS("__alignof__" , __alignof , KEYALL) +ALIAS("__asm" , asm , KEYALL) +ALIAS("__asm__" , asm , KEYALL) +ALIAS("__attribute__" , __attribute , KEYALL) +ALIAS("__complex" , _Complex , KEYALL) +ALIAS("__complex__" , _Complex , KEYALL) +ALIAS("__const" , const , KEYALL) +ALIAS("__const__" , const , KEYALL) +ALIAS("__decltype" , decltype , KEYCXX) +ALIAS("__imag__" , __imag , KEYALL) +ALIAS("__inline" , inline , KEYALL) +ALIAS("__inline__" , inline , KEYALL) +ALIAS("__nullptr" , nullptr , KEYCXX) +ALIAS("__real__" , __real , KEYALL) +ALIAS("__restrict" , restrict , KEYALL) +ALIAS("__restrict__" , restrict , KEYALL) +ALIAS("__signed" , signed , KEYALL) +ALIAS("__signed__" , signed , KEYALL) +ALIAS("__typeof" , typeof , KEYALL) +ALIAS("__typeof__" , typeof , KEYALL) +ALIAS("__typeof_unqual" , typeof_unqual, KEYNOCXX) +ALIAS("__typeof_unqual__", typeof_unqual, KEYNOCXX) ---------------- h-vetinari wrote:
The C++ rejection should be questioned IMO. To my understanding, once upon a time C++ went with `decltype` to not steal `typeof` from C, waiting for WG14 to standardize it (and once that happened - admittedly decades later -, it was [explicitly](https://www.open-std.org/jtc1/sc22/wg14/www/docs/n2930.pdf) made more compatible with C++ during that process). Now that it's finally there, clang could reasonably expose this in all C/C++ modes, also without the leading underscores. https://github.com/llvm/llvm-project/pull/87392 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits