Re: vasnprintf: Reject a width > INT_MAX

I did: > vasnprintf: Reject a width > INT_MAX. > * lib/vasnprintf.c (VASNPRINTF): If a width is > INT_MAX, fail with > EOVERFLOW. This causes a test failure on Cygwin. Namely, Cygwin has vasnprintf, but with this argument list "x%030dy\n"

vasnprintf: Reject a width > INT_MAX

with ERANGE on MSVC.) See the attached test program. * It's consistent with passing the width as an 'int' argument: "%*d". It would be odd if an application could use the 'int' argument for some widths and had to use a dynamically created format string fo