On 04/04/2012 09:14 AM, Christophe Gisquet wrote:
> Commit 356ee8d caused the initial inversion.
> ---
> libavcodec/x86/dsputil_mmx.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/libavcodec/x86/dsputil_mmx.c b/libavcodec/x86/dsputil_mmx.c
> index 207202c..adfc268 100644
> --- a/libavcodec/x86/dsputil_mmx.c
> +++ b/libavcodec/x86/dsputil_mmx.c
> @@ -2977,7 +2977,7 @@ static void dsputil_init_sse2(DSPContext *c,
> AVCodecContext *avctx,
> const int bit_depth = avctx->bits_per_raw_sample;
> const int high_bit_depth = bit_depth > 8;
>
> - if (mm_flags & AV_CPU_FLAG_3DNOW) {
> + if (!(mm_flags & AV_CPU_FLAG_3DNOW)) {
> // these functions are slower than mmx on AMD, but faster on Intel
> if (!high_bit_depth) {
> c->put_pixels_tab[0][0] = put_pixels16_sse2;
Basically ok for the revert.
However, this looks to me like a situation to use the sse2slow flag
instead of the 3dnow flag. Anyone care to test on a modern AMD cpu?
-Justin
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
