Martin Sebor wrote: > >faridz wrote: >> >> >> URL: http://svn.apache.org/viewvc?rev=693424&view=rev >> Log: >> 2008-09-09 Farid Zaripov <[EMAIL PROTECTED]> >> >> * tests/regress/18.c.limits.stdcxx-988.cpp: Resolved compilation >> error on MSVC and ICC/Windows. >> >> Modified: >> stdcxx/branches/4.2.x/tests/regress/18.c.limits.stdcxx-988.cpp >> >> Modified: >stdcxx/branches/4.2.x/tests/regress/18.c.limits.stdcxx-988.cpp >> URL: >> >http://svn.apache.org/viewvc/stdcxx/branches/4.2.x/tests/regres >s/18.c.limits.stdcxx-988.cpp?rev=693424&r1=693423&r2=693424&view=diff >> >=============================================================== >=============== >> --- stdcxx/branches/4.2.x/tests/regress/18.c.limits.stdcxx-988.cpp >> (original) >> +++ >stdcxx/branches/4.2.x/tests/regress/18.c.limits.stdcxx-988.cpp Tue Sep >> 9 03:35:59 2008 >> @@ -138,14 +138,14 @@ >> assert (SCHAR_MAX == +127); >> assert (SHRT_MAX >= +32767); >> >> -#ifndef _RWSTD_EDG_ECCP >> +#if !defined (_RWSTD_EDG_ECCP) && !defined (_MSC_VER) >> >> > >_RWSTD_EDG_ECCP should only be #defined when we're using the EDG eccp >vanilla demo, but not when we're using a compiler that sits on top of >the EDG front end such as Intel C++. If this condition evaluates to >true we have a configuration problem...
The seems valid to me. The conditional is only true for compilers other than EDG and Microsoft. > >Martin > >> >> // avoid exercising with the EDG eccp front end (strict mode) >> assert (LONG_BIT >= 32); >> assert (SSIZE_MAX >= _POSIX_SSIZE_MAX); >> assert (WORD_BIT >= 32); >> >> -#endif // EDG eccp >> +#endif // EDG eccp || MSVC || ICC/Windows >> >> assert (UCHAR_MAX >= 255); >> assert (UINT_MAX >= 4294967295); >> >> >> >> > >
