"Ronald S. Bultje" <[email protected]> writes:
> From: "Ronald S. Bultje" <[email protected]>
>
> ---
> configure | 9 +++++++++
> libavutil/internal.h | 6 +++++-
> 2 files changed, 14 insertions(+), 1 deletion(-)
>
> diff --git a/configure b/configure
> index 351d8a0..4fc20a2 100755
> --- a/configure
> +++ b/configure
> @@ -1118,7 +1118,9 @@ HAVE_LIST="
> MapViewOfFile
> memalign
> mkstemp
> + mm_empty
> mmap
> + mmintrin_h
> nanosleep
> netinet_sctp_h
> poll_h
> @@ -2646,6 +2648,13 @@ check_cc <<EOF && enable inline_asm
> void foo(void) { __asm__ volatile ("" ::); }
> EOF
>
> +if check_header mmintrin.h; then
> + check_cc <<EOF && enable mm_empty
> +#include <mmintrin.h>
> +int main (void) { _mm_empty(); return 0; }
> +EOF
> +fi
The previous patch used intrin.h, now it's mmintrin.h. Please explain.
Looking around my hard drive, most x86 compilers seem to have a
compatible mmintrin.h, so I guess this is better.
> _restrict=
> for restrict_keyword in restrict __restrict__ __restrict; do
> check_cc <<EOF && _restrict=$restrict_keyword && break
> diff --git a/libavutil/internal.h b/libavutil/internal.h
> index 41e8a9a..81426af 100644
> --- a/libavutil/internal.h
> +++ b/libavutil/internal.h
> @@ -231,7 +231,7 @@ struct AVDictionary {
> # define ONLY_IF_THREADS_ENABLED(x) NULL
> #endif
>
> -#if HAVE_MMX
> +#if HAVE_MMX && HAVE_INLINE_ASM
> /**
> * Empty mmx state.
> * this must be called between any dsp function and float/double code.
> @@ -241,6 +241,10 @@ static av_always_inline void emms_c(void)
> {
> __asm__ volatile ("emms" ::: "memory");
> }
> +#elif HAVE_MMX && HAVE_MM_EMPTY
> +#include <mmintrin.h>
> +
> +#define emms_c _mm_empty
> #else /* HAVE_MMX */
> #define emms_c()
> #endif /* HAVE_MMX */
> --
This is still almost impossible to read. Something like this would be
clearer:
#elif HAVE_MMX && HAVE_MM_EMPTY
# include <mmintrin.h>
# define emms_c() _mm_empty()
#else
--
Måns Rullgård
[email protected]
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
