Re: AC_HEADER_ASSERT: don't say assertions are disabled when they're not

[Eric Blake]

Jim Meyering <jim <at> meyering.net> writes:

> 
> I was surprised to see assertions being turned off by default:
>     checking whether to enable assertions... no <-- wrong! default is to 
enable

Aaargh.  So much for my careful review.

> +     AS_IF(dnl
> +      [test "x$enableval" = xno],
> +        [AC_DEFINE([NDEBUG], [1],
> +          [Define to 1 if assertions should be disabled.])],
>        [test "x$enableval" != xyes],
> -      [AC_MSG_WARN([invalid argument supplied to --enable-assert])
> -      ac_enable_assert=no])],
> -    [ac_enable_assert=no])
> +        [AC_MSG_WARN([invalid argument supplied to --enable-assert])
> +        ac_enable_assert=no])],

This line also needs to be ac_enable_assert=yes.  In short, a warning is 
printed, and _assertions remain enabled_, if you did --enable-assert=oops

> +    [ac_enable_assert=yes])
>    AC_MSG_RESULT([$ac_enable_assert])

Other than that, yes, please apply and please sync to gnulib.

-- 
Eric Blake