Daniil Gusev <[email protected]> wrote: > The compiler itself builds and tests pass, so the warning does not appear to > be critical. Could you please advise on the best way to handle this warning?
0. btw: you can disable inclusion of bcheck.c by an option in the configure script 1. fix it by not defining wchar_t et al in tcc - it's the job of the C library (see also https://pubs.opengroup.org/onlinepubs/9799919799/) - imo it's strange for tcc to do this job instead > Currently, the following block exists in `tccdefs.h`: > > #if defined _WIN32 > #define __WCHAR_TYPE__ unsigned short > #define __WINT_TYPE__ unsigned short > #elif defined __linux__ > #define __WCHAR_TYPE__ int > #define __WINT_TYPE__ unsigned int > #else > #define __WCHAR_TYPE__ int you could do the following instead (idk about win32, hence it stays) > #if defined _WIN32 > #define __WCHAR_TYPE__ unsigned short > #define __WINT_TYPE__ unsigned short after "git grep WCHAR", I'm even more convinced to just delete it - let the C library do its job on POSIX-conformant platforms, let w32 have its special sauce where do you get wchar_t from instead? stddef.h from libc first, if possible - tcc's stddef.h is usually wronger than libc since this is quite the job, you should probably just live with a downstream patch for now
signature.asc
Description: PGP signature
_______________________________________________ Tinycc-devel mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/tinycc-devel
