Hi,
On Sep 8, 2012 8:42 AM, "Hendrik Leppkes" <h.lepp...@gmail.com> wrote:
>
> Currently, implicit function declarations are only considered an error
during building,
> but not during configure, which can result in configure mis-detecting the
availability
> of some functions (most notably "log2" on amd64).
>
> Instead, treat it as an error condition in all stages of building, for a
consistent
> behaviour between configure and the actual build process.
> ---
> configure | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/configure b/configure
> index 1fe505c..7ca262f 100755
> --- a/configure
> +++ b/configure
> @@ -2104,7 +2104,7 @@ msvc_flags(){
> -g) echo -Z7 ;;
> -Wall) echo -W4 -wd4244 -wd4127 -wd4018
-wd4389 \
> -wd4146 -wd4057 -wd4204 -wd4706
-wd4305 \
> - -wd4152 -wd4324 -we4013 -wd4100
-wd4214 \
> + -wd4152 -wd4324 -wd4100 -wd4214 \
> -wd4996 ;;
> -std=c99) ;;
> -fno-math-errno) ;;
> @@ -2322,7 +2322,7 @@ probe_cc(){
> _ld_lib='lib%.a'
> _ld_path='-libpath:'
> _flags='-nologo'
> - _cflags='-D_USE_MATH_DEFINES -Dinline=__inline -FIstdlib.h
-Dstrtoll=_strtoi64'
> + _cflags='-D_USE_MATH_DEFINES -Dinline=__inline -FIstdlib.h
-Dstrtoll=_strtoi64 -we4013'
> if [ $pfx = hostcc ]; then
> _cflags="$cflags -Dsnprintf=_snprintf"
> fi
This sounds like the wrong fix tbh. The first chunk looks OK but wouldn't
other systems, eg old GCC versions, be similarly affected? Shouldn't the
true fix live elsewhere, ie fix the log2() check?
Ronald
_______________________________________________
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
