Re: MirBSD mbtowc bug? failure on test-wcrtomb

[Bruno Haible]

Hi Eric,

> Bruno, this one probably needs your help for a gnulib workaround.  On 
> MirBSD 10 (20100815), wcrtomb is found but not replaced; leading to 
> failure in the unit test:
> 
> ac_cv_func_wcrtomb=yes
> ac_cv_have_decl_wcrtomb=yes
> gl_cv_func_wcrtomb_retval=yes
> gl_cv_have_raw_decl_wcrtomb=yes
> GNULIB_WCRTOMB='1'
> HAVE_WCRTOMB='1'
> REPLACE_WCRTOMB='0'
> 
> ../../gltests/test-wcrtomb.c:43: assertion failed
> Abort trap (core dumped)
> FAIL: test-wcrtomb.sh

Can you please recompile the test with -g -O0 and then show a gdb stack
trace?

> Failing on 'LC_ALL=ja_JP ./test-wcrtomb 3' at this line:
> 
>    wc = (wchar_t) 0xBADFACE;
>    iret = mbtowc (&wc, s, n);
>    ASSERT (iret == n);
> 
> with n==2, iret==1, s=="\303\274\303\237er".

This makes no sense to me: LC_ALL=ja_JP and the argument "3" are indicating
that a locale with EUC-JP encoding is being tested. But the string
s=="\303\274\303\237er" indicates an argument "2" and that an UTF-8 locale
is being tested. Has the test program been miscompiled or what??

Bruno