On Sat, Apr 19, 2014 at 06:55:04PM -0400, Derek Buitenhuis wrote:
> From: Kostya Shishkov <[email protected]>
>
> Signed-off-by: Derek Buitenhuis <[email protected]>
> ---
> The static inline version is just as fast as the macro templaed one,
> and far less ugly. I tested.
How does one test ugliness?
> ---
> configure | 2 +-
> libavcodec/fic.c | 63
> ++++++++++++++++++++++++++++++++++++++++++++++----------
> 2 files changed, 53 insertions(+), 12 deletions(-)
>
> diff --git a/configure b/configure
> index 63a9e72..65a9cf8 100755
> --- a/configure
> +++ b/configure
> @@ -1737,7 +1737,7 @@ ffv1_decoder_select="golomb rangecoder"
> ffv1_encoder_select="rangecoder"
> ffvhuff_decoder_select="huffyuv_decoder"
> ffvhuff_encoder_select="huffyuv_encoder"
> -fic_decoder_select="dsputil golomb"
> +fic_decoder_select="golomb"
> flac_decoder_select="golomb"
> flac_encoder_select="dsputil golomb lpc"
> flashsv_decoder_deps="zlib"
> diff --git a/libavcodec/fic.c b/libavcodec/fic.c
> index df03437..2d9da65 100644
> --- a/libavcodec/fic.c
> +++ b/libavcodec/fic.c
> @@ -24,7 +24,6 @@
> #include "libavutil/common.h"
> #include "avcodec.h"
> #include "internal.h"
> -#include "dsputil.h"
> #include "get_bits.h"
> #include "golomb.h"
>
> @@ -40,9 +39,6 @@ typedef struct FICContext {
> AVCodecContext *avctx;
> AVFrame *frame;
>
> - DSPContext dsp;
> - ScanTable scantable;
> -
> FICThreadContext *slice_data;
> int slice_data_size;
>
> @@ -80,6 +76,55 @@ static const uint8_t fic_header[7] = { 0, 0, 1, 'F', 'I',
> 'C', 'V' };
>
> #define FIC_HEADER_SIZE 27
>
> +static av_always_inline void fic_idct(int16_t *blk, int step, int shift)
> +{
> + const int t0 = 27246 * blk[3 * step] + 18405 * blk[5 * step];
> + const int t1 = 27246 * blk[5 * step] - 18405 * blk[3 * step];
> + const int t2 = 6393 * blk[7 * step] + 32139 * blk[1 * step];
> + const int t3 = 6393 * blk[1 * step] - 32139 * blk[7 * step];
> + const int t4 = 5793 * (t2 + t0 + 0x800 >> 12);
> + const int t5 = 5793 * (t3 + t1 + 0x800 >> 12);
> + const int t6 = t2 - t0;
> + const int t7 = t3 - t1;
> + const int t8 = 17734 * blk[2 * step] - 42813 * blk[6 * step];
> + const int t9 = 17734 * blk[6 * step] + 42814 * blk[2 * step];
> + const int tA = (blk[0 * step] - blk[4 * step] << 15) + (1 << shift - 1);
> + const int tB = (blk[0 * step] + blk[4 * step] << 15) + (1 << shift - 1);
> + blk[0 * step] = ( t4 + t9 + tB) >> shift;
> + blk[1 * step] = ( t6 + t7 + t8 + tA) >> shift;
> + blk[2 * step] = ( t6 - t7 - t8 + tA) >> shift;
> + blk[3 * step] = ( t5 - t9 + tB) >> shift;
> + blk[4 * step] = ( -t5 - t9 + tB) >> shift;
> + blk[5 * step] = (-(t6 - t7) - t8 + tA) >> shift;
> + blk[6 * step] = (-(t6 + t7) + t8 + tA) >> shift;
> + blk[7 * step] = ( -t4 + t9 + tB) >> shift;
> +}
> +
> +static void fic_idct_put(uint8_t *dst, int stride, int16_t *block)
> +{
> + int i, j;
> + int16_t *ptr;
> +
> + ptr = block;
> + for (i = 0; i < 8; i++) {
> + fic_idct(ptr, 8, 13);
> + ptr++;
> + }
> +
> + ptr = block;
> + for (i = 0; i < 8; i++) {
> + fic_idct(ptr, 1, 20);
> + ptr += 8;
> + }
> +
> + ptr = block;
> + for (j = 0; j < 8; j++) {
> + for (i = 0; i < 8; i++)
> + dst[i] = av_clip_uint8(ptr[i]);
> + dst += stride;
> + ptr += 8;
> + }
> +}
> static int fic_decode_block(FICContext *ctx, GetBitContext *gb,
> uint8_t *dst, int stride, int16_t *block)
> {
> @@ -94,16 +139,16 @@ static int fic_decode_block(FICContext *ctx,
> GetBitContext *gb,
> return 0;
> }
>
> - ctx->dsp.clear_block(block);
> + memset(block, 0, sizeof(int16_t) * 64);
sizeof(*block) please
> num_coeff = get_bits(gb, 7);
> if (num_coeff > 64)
> return AVERROR_INVALIDDATA;
>
> for (i = 0; i < num_coeff; i++)
> - block[ctx->scantable.permutated[i]] = get_se_golomb(gb) *
> ctx->qmat[i];
> + block[ff_zigzag_direct[i]] = get_se_golomb(gb) * ctx->qmat[i];
>
> - ctx->dsp.idct_put(dst, stride, block);
> + fic_idct_put(dst, stride, block);
>
> return 0;
> }
> @@ -283,10 +328,6 @@ static av_cold int fic_decode_init(AVCodecContext *avctx)
> if (!ctx->frame)
> return AVERROR(ENOMEM);
>
> - ff_dsputil_init(&ctx->dsp, avctx);
> -
> - ff_init_scantable(ctx->dsp.idct_permutation, &ctx->scantable,
> ff_zigzag_direct);
> -
> return 0;
> }
>
> --
other than that LGTM
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
