Re: math.in.h problem

[Eric Blake]

On 07/16/2013 06:04 PM, Jake Aleemi wrote:
> I noticed in lines 157-158 of /gnulib/lib/math.in.h,
> there was a multiline comment end tag after the text after
> the 'else' preprocessor directive
> 
> # else Pull in a function that fixes the 'int' to 'long double'
> conversion 
>    of glibc 2.7.  */ <--???
> 
> this can be fixed by adding a multiline comment tag after 'else'
> and before 'Pull', e.g.:
> 
> # else /* Pull in a function that fixes the 'int' to 'long double'
> conversion
>    of glibc 2.7.  */ <--that's better!

Thanks for the report.  However, I can't reproduce your problem; line
157 of gnulib.git's math.in.h is:


/* HUGE_VALF is a 'float' Infinity.  */
#ifndef HUGE_VALF
# if defined _MSC_VER
/* The Microsoft MSVC 9 compiler chokes on the expression 1.0f / 0.0f.  */
#  define HUGE_VALF (1e25f * 1e25f)

and the comment in question (present since commit fddec5c in Sep 2011)
was formatted in gnulib as:

#if @REPLACE_ITOLD@
/* Pull in a function that fixes the 'int' to 'long double' conversion
   of glibc 2.7.  */
_GL_EXTERN_C void _Qp_itoq (long double *, int);

If you are still experiencing a problem, you'll need to give us more
details about what version you are testing, and more context about what
you are trying to patch.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

signature.asc
Description: OpenPGP digital signature