@@ -31,10 +31,10 @@

 #include <math.h>

-#include <libavutil/imgutils.h>

 #include <libavutil/opt.h>

 #include <libavcodec/avcodec.h>

 #include <libavutil/audioconvert.h>

+#include <libavutil/imgutils.h>

 #include <libavutil/mathematics.h>

 #include <libavutil/samplefmt.h>

@@ -315,11 +315,11 @@ static void video_encode_example(const char *filename, int codec_id)

 {

     AVCodec *codec;

     AVCodecContext *c= NULL;

-    int i, out_size, x, y, outbuf_size;

+    int i, ret, x, y, got_output;

     FILE *f;

     AVFrame *picture;

-    uint8_t *outbuf;

-    int had_output=0;

+    AVPacket pkt;

+    uint8_t endcode[] = { 0, 0, 1, 0xb7 };

     printf("Encode video file %s\n", filename);

@@ -359,17 +359,25 @@ static void video_encode_example(const char *filename, int codec_id)

         exit(1);

     }

-    /* alloc image and output buffer */

-    outbuf_size = 100000 + 12*c->width*c->height;

-    outbuf = malloc(outbuf_size);

-

     /* the image can be allocated by any means and av_image_alloc() is

      * just the most convenient way if av_malloc() is to be used */

-    av_image_alloc(picture->data, picture->linesize,

-                   c->width, c->height, c->pix_fmt, 1);

+    ret = av_image_alloc(picture->data, picture->linesize, c->width, c->height,

+                         c->pix_fmt, 32);

+    if (ret < 0) {

+        fprintf(stderr, "could not alloc raw picture buffer\n");

+        exit(1);

+    }

+

+    picture->format = c->pix_fmt;

+    picture->width  = c->width;

+    picture->height = c->height;

     /* encode 1 second of video */

     for(i=0;i<25;i++) {

+        av_init_packet(&pkt);

+        pkt.data = NULL;    // packet data will be allocated by the encoder

+        pkt.size = 0;

+

         fflush(stdout);

         /* prepare a dummy image */

         /* Y */

@@ -387,35 +395,46 @@ static void video_encode_example(const char *filename, int codec_id)

             }

         }

+        picture->pts = i;

+

         /* encode the image */

-        out_size = avcodec_encode_video(c, outbuf, outbuf_size, picture);

-        had_output |= out_size;

-        printf("encoding frame %3d (size=%5d)\n", i, out_size);

-        fwrite(outbuf, 1, out_size, f);

+        ret = avcodec_encode_video2(c, &pkt, picture, &got_output);

+        if (ret < 0) {

+            fprintf(stderr, "error encoding frame\n");

+            exit(1);

+        }

+

+        if (got_output) {

+            printf("encoding frame %3d (size=%5d)\n", i, pkt.size);

+            fwrite(pkt.data, 1, pkt.size, f);

+            av_free_packet(&pkt);

+        }

     }

     /* get the delayed frames */

-    for(; out_size || !had_output; i++) {

+    for (got_output = 1; got_output; i++) {

         fflush(stdout);

-        out_size = avcodec_encode_video(c, outbuf, outbuf_size, NULL);

-        had_output |= out_size;

-        printf("write frame %3d (size=%5d)\n", i, out_size);

-        fwrite(outbuf, 1, out_size, f);

+        ret = avcodec_encode_video2(c, &pkt, NULL, &got_output);

+        if (ret < 0) {

+            fprintf(stderr, "error encoding frame\n");

+            exit(1);

+        }

+

+        if (got_output) {

+            printf("write frame %3d (size=%5d)\n", i, pkt.size);

+            fwrite(pkt.data, 1, pkt.size, f);

+            av_free_packet(&pkt);

+        }

     }

     /* add sequence end code to have a real mpeg file */

-    outbuf[0] = 0x00;

-    outbuf[1] = 0x00;

-    outbuf[2] = 0x01;

-    outbuf[3] = 0xb7;

-    fwrite(outbuf, 1, 4, f);

+    fwrite(endcode, 1, sizeof(endcode), f);

     fclose(f);

-    free(outbuf);

     avcodec_close(c);

     av_free(c);

-    av_free(picture->data[0]);

+    av_freep(&picture->data[0]);

     av_free(picture);

     printf("\n");

 }

@@ -272,6 +272,12 @@ Name of live stream to subscribe to. By default no value will be sent.

 It is only sent if the option is specified or if rtmp_live

 is set to live.

+@item rtmp_swfhash

+SHA256 hash of the decompressed SWF file (32 bytes).

+

+@item rtmp_swfsize

+Size of the decompressed SWF file, required for SWFVerification.

+

 @item rtmp_swfurl

 URL of the SWF player for the media. By default no value will be sent.

@@ -406,10 +406,15 @@ static int encode_slice(AVCodecContext *avctx, const AVFrame *pic,

     int total_size = 0;

     const uint16_t *src;

     int slice_width_factor = av_log2(mbs_per_slice);

-    int num_cblocks, pwidth, linesize, line_offset;

+    int num_cblocks, pwidth, linesize, line_add;

     int plane_factor, is_chroma;

     uint16_t *qmat;

+    if (ctx->pictures_per_frame == 1)

+        line_add = 0;

+    else

+        line_add = ctx->cur_picture_idx ^ !pic->top_field_first;

+

     if (ctx->force_quant) {

         qmat = ctx->quants[0];

     } else if (quant < MAX_STORED_Q) {

@@ -437,15 +442,14 @@ static int encode_slice(AVCodecContext *avctx, const AVFrame *pic,

             pwidth      = avctx->width >> 1;

         }

-        line_offset = ((ctx->cur_picture_idx ^ !pic->top_field_first) &

-                       (ctx->pictures_per_frame - 1)) * pic->linesize[i];

         linesize = pic->linesize[i] * ctx->pictures_per_frame;

-        src = (const uint16_t*)(pic->data[i] + yp * linesize + line_offset) + xp;

+        src = (const uint16_t*)(pic->data[i] + yp * linesize +

+                                line_add * pic->linesize[i]) + xp;

         get_slice_data(ctx, src, linesize, xp, yp,

                        pwidth, avctx->height / ctx->pictures_per_frame,

-                       ctx->blocks[0], ctx->emu_buf, mbs_per_slice,

-                       num_cblocks, is_chroma);

+                       ctx->blocks[0], ctx->emu_buf,

+                       mbs_per_slice, num_cblocks, is_chroma);

         sizes[i] = encode_slice_plane(ctx, pb, src, linesize,

                                       mbs_per_slice, ctx->blocks[0],

                                       num_cblocks, plane_factor,

@@ -579,8 +583,12 @@ static int find_slice_quant(AVCodecContext *avctx, const AVFrame *pic,

     int slice_bits[TRELLIS_WIDTH], slice_score[TRELLIS_WIDTH];

     int overquant;

     uint16_t *qmat;

-    int linesize[4], line_offset;

+    int linesize[4], line_add;

+    if (ctx->pictures_per_frame == 1)

+        line_add = 0;

+    else

+        line_add = ctx->cur_picture_idx ^ !pic->top_field_first;

     mbs = x + mbs_per_slice;

     for (i = 0; i < ctx->num_planes; i++) {

@@ -600,15 +608,14 @@ static int find_slice_quant(AVCodecContext *avctx, const AVFrame *pic,

             pwidth         = avctx->width >> 1;

         }

-        line_offset = ((ctx->cur_picture_idx ^ !pic->top_field_first) &

-                       (ctx->pictures_per_frame - 1)) * pic->linesize[i];

         linesize[i] = pic->linesize[i] * ctx->pictures_per_frame;

-        src = (const uint16_t*)(pic->data[i] + yp * linesize[i] + line_offset) + xp;

+        src = (const uint16_t*)(pic->data[i] + yp * linesize[i] +

+                                line_add * pic->linesize[i]) + xp;

         get_slice_data(ctx, src, linesize[i], xp, yp,

                        pwidth, avctx->height / ctx->pictures_per_frame,

-                       td->blocks[i], td->emu_buf, mbs_per_slice,

-                       num_cblocks[i], is_chroma[i]);

+                       td->blocks[i], td->emu_buf,

+                       mbs_per_slice, num_cblocks[i], is_chroma[i]);

     }

     for (q = min_quant; q < max_quant + 2; q++) {

@@ -767,9 +774,8 @@ static int encode_frame(AVCodecContext *avctx, AVPacket *pkt,

     bytestream_put_be16  (&buf, avctx->height);

     frame_flags = ctx->chroma_factor << 6;

-    if (avctx->flags & CODEC_FLAG_INTERLACED_DCT) {

+    if (avctx->flags & CODEC_FLAG_INTERLACED_DCT)

         frame_flags |= pic->top_field_first ? 0x04 : 0x08;

-    }

     bytestream_put_byte  (&buf, frame_flags);

     bytestream_put_byte  (&buf, 0);             // reserved

@@ -791,7 +797,9 @@ static int encode_frame(AVCodecContext *avctx, AVPacket *pkt,

     }

     bytestream_put_be16  (&tmp, buf - orig_buf); // write back frame header size

-    for (ctx->cur_picture_idx = 0; ctx->cur_picture_idx < ctx->pictures_per_frame; ++ctx->cur_picture_idx) {

+    for (ctx->cur_picture_idx = 0;

+         ctx->cur_picture_idx < ctx->pictures_per_frame;

+         ctx->cur_picture_idx++) {

         // picture header

         picture_size_pos = buf + 1;

         bytestream_put_byte  (&buf, 0x40);          // picture header size (in bits)

@@ -845,7 +853,6 @@ static int encode_frame(AVCodecContext *avctx, AVPacket *pkt,

     frame_size = buf - orig_buf;

     bytestream_put_be32(&orig_buf, frame_size);

-

     pkt->size   = frame_size;

     pkt->flags |= AV_PKT_FLAG_KEY;

     *got_packet = 1;

@@ -927,7 +934,8 @@ static av_cold int encode_init(AVCodecContext *avctx)

     if (!ctx->force_quant) {

         if (!ctx->bits_per_mb) {

             for (i = 0; i < NUM_MB_LIMITS - 1; i++)

-                if (prores_mb_limits[i] >= ctx->mb_width * ctx->mb_height * ctx->pictures_per_frame)

+                if (prores_mb_limits[i] >= ctx->mb_width * ctx->mb_height *

+                                           ctx->pictures_per_frame)

                     break;

             ctx->bits_per_mb   = ctx->profile_info->br_tab[i];

         } else if (ctx->bits_per_mb < 128) {

@@ -991,12 +999,15 @@ static av_cold int encode_init(AVCodecContext *avctx)

     ctx->frame_size_upper_bound = ctx->pictures_per_frame *

                                   ctx->slices_per_picture *

                                   (2 + 2 * ctx->num_planes +

-                                   (mps * ctx->bits_per_mb) / 8) + 200;

+                                   (mps * ctx->bits_per_mb) / 8)

+                                  + 200;

     avctx->codec_tag   = ctx->profile_info->tag;

-    av_log(avctx, AV_LOG_DEBUG, "profile %d, %d slices/pic, %d pics/frame, %d bits per MB\n",

-           ctx->profile, ctx->slices_per_picture, ctx->pictures_per_frame, ctx->bits_per_mb);

+    av_log(avctx, AV_LOG_DEBUG,

+           "profile %d, %d slices, interlacing: %s, %d bits per MB\n",

+           ctx->profile, ctx->slices_per_picture * ctx->pictures_per_frame,

+           interlaced ? "yes" : "no", ctx->bits_per_mb);

     av_log(avctx, AV_LOG_DEBUG, "frame size upper bound: %d\n",

            ctx->frame_size_upper_bound);

@@ -40,6 +40,8 @@

 #include <process.h>

 #include "libavutil/common.h"

+#include "libavutil/internal.h"

+#include "libavutil/mem.h"

 typedef struct {

     void *handle;

@@ -5,7 +5,7 @@ OBJS-$(CONFIG_XMM_CLOBBER_TEST)        += x86/w64xmmtest.o

 MMX-OBJS                               += x86/dsputil_mmx.o             \

                                           x86/fdct_mmx.o                \

-                                          x86/fmtconvert_mmx.o          \

+                                          x86/fmtconvert_init.o         \

                                           x86/idct_mmx_xvid.o           \

                                           x86/idct_sse2_xvid.o          \

                                           x86/motion_est_mmx.o          \

@@ -13,15 +13,15 @@ MMX-OBJS                               += x86/dsputil_mmx.o             \

                                           x86/simple_idct_mmx.o         \

 MMX-OBJS-$(CONFIG_AAC_DECODER)         += x86/sbrdsp_init.o

-MMX-OBJS-$(CONFIG_AC3DSP)              += x86/ac3dsp_mmx.o

+MMX-OBJS-$(CONFIG_AC3DSP)              += x86/ac3dsp_init.o

 MMX-OBJS-$(CONFIG_CAVS_DECODER)        += x86/cavsdsp_mmx.o

 MMX-OBJS-$(CONFIG_DNXHD_ENCODER)       += x86/dnxhd_mmx.o

 MMX-OBJS-$(CONFIG_DWT)                 += x86/snowdsp_mmx.o \

                                           x86/dwt.o

 MMX-OBJS-$(CONFIG_ENCODERS)            += x86/dsputilenc_mmx.o

-MMX-OBJS-$(CONFIG_FFT)                 += x86/fft.o

+MMX-OBJS-$(CONFIG_FFT)                 += x86/fft_init.o

 MMX-OBJS-$(CONFIG_GPL)                 += x86/idct_mmx.o

-MMX-OBJS-$(CONFIG_H264DSP)             += x86/h264dsp_mmx.o

+MMX-OBJS-$(CONFIG_H264DSP)             += x86/h264dsp_init.o

 MMX-OBJS-$(CONFIG_H264PRED)            += x86/h264_intrapred_init.o

 MMX-OBJS-$(CONFIG_LPC)                 += x86/lpc_mmx.o

 MMX-OBJS-$(CONFIG_MPEGAUDIODSP)        += x86/mpegaudiodec_mmx.o

new file mode 100644

@@ -0,0 +1,93 @@

+/*

+ * x86-optimized AC-3 DSP utils

+ * Copyright (c) 2011 Justin Ruggles

+ *

+ * This file is part of FFmpeg.

+ *

+ * FFmpeg is free software; you can redistribute it and/or

+ * modify it under the terms of the GNU Lesser General Public

+ * License as published by the Free Software Foundation; either

+ * version 2.1 of the License, or (at your option) any later version.

+ *

+ * FFmpeg is distributed in the hope that it will be useful,

+ * but WITHOUT ANY WARRANTY; without even the implied warranty of

+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

+ * Lesser General Public License for more details.

+ *

+ * You should have received a copy of the GNU Lesser General Public

+ * License along with FFmpeg; if not, write to the Free Software

+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

+ */

+

+#include "libavutil/x86/asm.h"

+#include "dsputil_mmx.h"

+#include "libavcodec/ac3dsp.h"

+

+extern void ff_ac3_exponent_min_mmx   (uint8_t *exp, int num_reuse_blocks, int nb_coefs);

+extern void ff_ac3_exponent_min_mmxext(uint8_t *exp, int num_reuse_blocks, int nb_coefs);

+extern void ff_ac3_exponent_min_sse2  (uint8_t *exp, int num_reuse_blocks, int nb_coefs);

+

+extern int ff_ac3_max_msb_abs_int16_mmx  (const int16_t *src, int len);

+extern int ff_ac3_max_msb_abs_int16_mmx2 (const int16_t *src, int len);

+extern int ff_ac3_max_msb_abs_int16_sse2 (const int16_t *src, int len);

+extern int ff_ac3_max_msb_abs_int16_ssse3(const int16_t *src, int len);

+

+extern void ff_ac3_lshift_int16_mmx (int16_t *src, unsigned int len, unsigned int shift);

+extern void ff_ac3_lshift_int16_sse2(int16_t *src, unsigned int len, unsigned int shift);

+

+extern void ff_ac3_rshift_int32_mmx (int32_t *src, unsigned int len, unsigned int shift);

+extern void ff_ac3_rshift_int32_sse2(int32_t *src, unsigned int len, unsigned int shift);

+

+extern void ff_float_to_fixed24_3dnow(int32_t *dst, const float *src, unsigned int len);

+extern void ff_float_to_fixed24_sse  (int32_t *dst, const float *src, unsigned int len);

+extern void ff_float_to_fixed24_sse2 (int32_t *dst, const float *src, unsigned int len);

+

+extern int ff_ac3_compute_mantissa_size_sse2(uint16_t mant_cnt[6][16]);

+

+extern void ff_ac3_extract_exponents_3dnow(uint8_t *exp, int32_t *coef, int nb_coefs);

+extern void ff_ac3_extract_exponents_sse2 (uint8_t *exp, int32_t *coef, int nb_coefs);

+extern void ff_ac3_extract_exponents_ssse3(uint8_t *exp, int32_t *coef, int nb_coefs);

+

+av_cold void ff_ac3dsp_init_x86(AC3DSPContext *c, int bit_exact)

+{

+#if HAVE_YASM

+    int mm_flags = av_get_cpu_flags();

+

+    if (mm_flags & AV_CPU_FLAG_MMX) {

+        c->ac3_exponent_min = ff_ac3_exponent_min_mmx;

+        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_mmx;

+        c->ac3_lshift_int16 = ff_ac3_lshift_int16_mmx;

+        c->ac3_rshift_int32 = ff_ac3_rshift_int32_mmx;

+    }

+    if (mm_flags & AV_CPU_FLAG_3DNOW && HAVE_AMD3DNOW) {

+        c->extract_exponents = ff_ac3_extract_exponents_3dnow;

+        if (!bit_exact) {

+            c->float_to_fixed24 = ff_float_to_fixed24_3dnow;

+        }

+    }

+    if (mm_flags & AV_CPU_FLAG_MMXEXT && HAVE_MMXEXT) {

+        c->ac3_exponent_min = ff_ac3_exponent_min_mmxext;

+        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_mmx2;

+    }

+    if (mm_flags & AV_CPU_FLAG_SSE && HAVE_SSE) {

+        c->float_to_fixed24 = ff_float_to_fixed24_sse;

+    }

+    if (mm_flags & AV_CPU_FLAG_SSE2 && HAVE_SSE) {

+        c->ac3_exponent_min = ff_ac3_exponent_min_sse2;

+        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_sse2;

+        c->float_to_fixed24 = ff_float_to_fixed24_sse2;

+        c->compute_mantissa_size = ff_ac3_compute_mantissa_size_sse2;

+        c->extract_exponents = ff_ac3_extract_exponents_sse2;

+        if (!(mm_flags & AV_CPU_FLAG_SSE2SLOW)) {

+            c->ac3_lshift_int16 = ff_ac3_lshift_int16_sse2;

+            c->ac3_rshift_int32 = ff_ac3_rshift_int32_sse2;

+        }

+    }

+    if (mm_flags & AV_CPU_FLAG_SSSE3 && HAVE_SSSE3) {

+        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_ssse3;

+        if (!(mm_flags & AV_CPU_FLAG_ATOM)) {

+            c->extract_exponents = ff_ac3_extract_exponents_ssse3;

+        }

+    }

+#endif

+}

deleted file mode 100644

@@ -1,93 +0,0 @@

-/*

- * x86-optimized AC-3 DSP utils

- * Copyright (c) 2011 Justin Ruggles

- *

- * This file is part of FFmpeg.

- *

- * FFmpeg is free software; you can redistribute it and/or

- * modify it under the terms of the GNU Lesser General Public

- * License as published by the Free Software Foundation; either

- * version 2.1 of the License, or (at your option) any later version.

- *

- * FFmpeg is distributed in the hope that it will be useful,

- * but WITHOUT ANY WARRANTY; without even the implied warranty of

- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

- * Lesser General Public License for more details.

- *

- * You should have received a copy of the GNU Lesser General Public

- * License along with FFmpeg; if not, write to the Free Software

- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

- */

-

-#include "libavutil/x86/asm.h"

-#include "dsputil_mmx.h"

-#include "libavcodec/ac3dsp.h"

-

-extern void ff_ac3_exponent_min_mmx   (uint8_t *exp, int num_reuse_blocks, int nb_coefs);

-extern void ff_ac3_exponent_min_mmxext(uint8_t *exp, int num_reuse_blocks, int nb_coefs);

-extern void ff_ac3_exponent_min_sse2  (uint8_t *exp, int num_reuse_blocks, int nb_coefs);

-

-extern int ff_ac3_max_msb_abs_int16_mmx  (const int16_t *src, int len);

-extern int ff_ac3_max_msb_abs_int16_mmx2 (const int16_t *src, int len);

-extern int ff_ac3_max_msb_abs_int16_sse2 (const int16_t *src, int len);

-extern int ff_ac3_max_msb_abs_int16_ssse3(const int16_t *src, int len);

-

-extern void ff_ac3_lshift_int16_mmx (int16_t *src, unsigned int len, unsigned int shift);

-extern void ff_ac3_lshift_int16_sse2(int16_t *src, unsigned int len, unsigned int shift);

-

-extern void ff_ac3_rshift_int32_mmx (int32_t *src, unsigned int len, unsigned int shift);

-extern void ff_ac3_rshift_int32_sse2(int32_t *src, unsigned int len, unsigned int shift);

-

-extern void ff_float_to_fixed24_3dnow(int32_t *dst, const float *src, unsigned int len);

-extern void ff_float_to_fixed24_sse  (int32_t *dst, const float *src, unsigned int len);

-extern void ff_float_to_fixed24_sse2 (int32_t *dst, const float *src, unsigned int len);

-

-extern int ff_ac3_compute_mantissa_size_sse2(uint16_t mant_cnt[6][16]);

-

-extern void ff_ac3_extract_exponents_3dnow(uint8_t *exp, int32_t *coef, int nb_coefs);

-extern void ff_ac3_extract_exponents_sse2 (uint8_t *exp, int32_t *coef, int nb_coefs);

-extern void ff_ac3_extract_exponents_ssse3(uint8_t *exp, int32_t *coef, int nb_coefs);

-

-av_cold void ff_ac3dsp_init_x86(AC3DSPContext *c, int bit_exact)

-{

-#if HAVE_YASM

-    int mm_flags = av_get_cpu_flags();

-

-    if (mm_flags & AV_CPU_FLAG_MMX) {

-        c->ac3_exponent_min = ff_ac3_exponent_min_mmx;

-        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_mmx;

-        c->ac3_lshift_int16 = ff_ac3_lshift_int16_mmx;

-        c->ac3_rshift_int32 = ff_ac3_rshift_int32_mmx;

-    }

-    if (mm_flags & AV_CPU_FLAG_3DNOW && HAVE_AMD3DNOW) {

-        c->extract_exponents = ff_ac3_extract_exponents_3dnow;

-        if (!bit_exact) {

-            c->float_to_fixed24 = ff_float_to_fixed24_3dnow;

-        }

-    }

-    if (mm_flags & AV_CPU_FLAG_MMXEXT && HAVE_MMXEXT) {

-        c->ac3_exponent_min = ff_ac3_exponent_min_mmxext;

-        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_mmx2;

-    }

-    if (mm_flags & AV_CPU_FLAG_SSE && HAVE_SSE) {

-        c->float_to_fixed24 = ff_float_to_fixed24_sse;

-    }

-    if (mm_flags & AV_CPU_FLAG_SSE2 && HAVE_SSE) {

-        c->ac3_exponent_min = ff_ac3_exponent_min_sse2;

-        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_sse2;

-        c->float_to_fixed24 = ff_float_to_fixed24_sse2;

-        c->compute_mantissa_size = ff_ac3_compute_mantissa_size_sse2;

-        c->extract_exponents = ff_ac3_extract_exponents_sse2;

-        if (!(mm_flags & AV_CPU_FLAG_SSE2SLOW)) {

-            c->ac3_lshift_int16 = ff_ac3_lshift_int16_sse2;

-            c->ac3_rshift_int32 = ff_ac3_rshift_int32_sse2;

-        }

-    }

-    if (mm_flags & AV_CPU_FLAG_SSSE3 && HAVE_SSSE3) {

-        c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_ssse3;

-        if (!(mm_flags & AV_CPU_FLAG_ATOM)) {

-            c->extract_exponents = ff_ac3_extract_exponents_ssse3;

-        }

-    }

-#endif

-}

deleted file mode 100644

@@ -1,72 +0,0 @@

-/*

- * This file is part of FFmpeg.

- *

- * FFmpeg is free software; you can redistribute it and/or

- * modify it under the terms of the GNU Lesser General Public

- * License as published by the Free Software Foundation; either

- * version 2.1 of the License, or (at your option) any later version.

- *

- * FFmpeg is distributed in the hope that it will be useful,

- * but WITHOUT ANY WARRANTY; without even the implied warranty of

- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

- * Lesser General Public License for more details.

- *

- * You should have received a copy of the GNU Lesser General Public

- * License along with FFmpeg; if not, write to the Free Software

- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

- */

-

-#include "libavutil/cpu.h"

-#include "libavcodec/dsputil.h"

-#include "libavcodec/dct.h"

-#include "fft.h"

-

-av_cold void ff_fft_init_mmx(FFTContext *s)

-{

-#if HAVE_YASM

-    int has_vectors = av_get_cpu_flags();

-#if ARCH_X86_32

-    if (has_vectors & AV_CPU_FLAG_3DNOW && HAVE_AMD3DNOW) {

-        /* 3DNow! for K6-2/3 */

-        s->imdct_calc = ff_imdct_calc_3dnow;

-        s->imdct_half = ff_imdct_half_3dnow;

-        s->fft_calc   = ff_fft_calc_3dnow;

-    }

-    if (has_vectors & AV_CPU_FLAG_3DNOWEXT && HAVE_AMD3DNOWEXT) {

-        /* 3DNowEx for K7 */

-        s->imdct_calc = ff_imdct_calc_3dnowext;

-        s->imdct_half = ff_imdct_half_3dnowext;

-        s->fft_calc   = ff_fft_calc_3dnowext;

-    }

-#endif

-    if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE) {

-        /* SSE for P3/P4/K8 */

-        s->imdct_calc  = ff_imdct_calc_sse;

-        s->imdct_half  = ff_imdct_half_sse;

-        s->fft_permute = ff_fft_permute_sse;

-        s->fft_calc    = ff_fft_calc_sse;

-        s->fft_permutation = FF_FFT_PERM_SWAP_LSBS;

-    }

-    if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX && s->nbits >= 5) {

-        /* AVX for SB */

-        s->imdct_half      = ff_imdct_half_avx;

-        s->fft_calc        = ff_fft_calc_avx;

-        s->fft_permutation = FF_FFT_PERM_AVX;

-    }

-#endif

-}

-

-#if CONFIG_DCT

-av_cold void ff_dct_init_mmx(DCTContext *s)

-{

-#if HAVE_YASM

-    int has_vectors = av_get_cpu_flags();

-    if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE)

-        s->dct32 = ff_dct32_float_sse;

-    if (has_vectors & AV_CPU_FLAG_SSE2 && HAVE_SSE)

-        s->dct32 = ff_dct32_float_sse2;

-    if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX)

-        s->dct32 = ff_dct32_float_avx;

-#endif

-}

-#endif

new file mode 100644

@@ -0,0 +1,72 @@

+/*

+ * This file is part of FFmpeg.

+ *

+ * FFmpeg is free software; you can redistribute it and/or

+ * modify it under the terms of the GNU Lesser General Public

+ * License as published by the Free Software Foundation; either

+ * version 2.1 of the License, or (at your option) any later version.

+ *

+ * FFmpeg is distributed in the hope that it will be useful,

+ * but WITHOUT ANY WARRANTY; without even the implied warranty of

+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

+ * Lesser General Public License for more details.

+ *

+ * You should have received a copy of the GNU Lesser General Public

+ * License along with FFmpeg; if not, write to the Free Software

+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

+ */

+

+#include "libavutil/cpu.h"

+#include "libavcodec/dsputil.h"

+#include "libavcodec/dct.h"

+#include "fft.h"

+

+av_cold void ff_fft_init_mmx(FFTContext *s)

+{

+#if HAVE_YASM

+    int has_vectors = av_get_cpu_flags();

+#if ARCH_X86_32

+    if (has_vectors & AV_CPU_FLAG_3DNOW && HAVE_AMD3DNOW) {

+        /* 3DNow! for K6-2/3 */

+        s->imdct_calc = ff_imdct_calc_3dnow;

+        s->imdct_half = ff_imdct_half_3dnow;

+        s->fft_calc   = ff_fft_calc_3dnow;

+    }

+    if (has_vectors & AV_CPU_FLAG_3DNOWEXT && HAVE_AMD3DNOWEXT) {

+        /* 3DNowEx for K7 */

+        s->imdct_calc = ff_imdct_calc_3dnowext;

+        s->imdct_half = ff_imdct_half_3dnowext;

+        s->fft_calc   = ff_fft_calc_3dnowext;

+    }

+#endif

+    if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE) {

+        /* SSE for P3/P4/K8 */

+        s->imdct_calc  = ff_imdct_calc_sse;

+        s->imdct_half  = ff_imdct_half_sse;

+        s->fft_permute = ff_fft_permute_sse;

+        s->fft_calc    = ff_fft_calc_sse;

+        s->fft_permutation = FF_FFT_PERM_SWAP_LSBS;

+    }

+    if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX && s->nbits >= 5) {

+        /* AVX for SB */

+        s->imdct_half      = ff_imdct_half_avx;

+        s->fft_calc        = ff_fft_calc_avx;

+        s->fft_permutation = FF_FFT_PERM_AVX;

+    }

+#endif

+}

+

+#if CONFIG_DCT

+av_cold void ff_dct_init_mmx(DCTContext *s)

+{

+#if HAVE_YASM

+    int has_vectors = av_get_cpu_flags();

+    if (has_vectors & AV_CPU_FLAG_SSE && HAVE_SSE)

+        s->dct32 = ff_dct32_float_sse;

+    if (has_vectors & AV_CPU_FLAG_SSE2 && HAVE_SSE)

+        s->dct32 = ff_dct32_float_sse2;

+    if (has_vectors & AV_CPU_FLAG_AVX && HAVE_AVX)

+        s->dct32 = ff_dct32_float_avx;

+#endif

+}

+#endif

new file mode 100644

@@ -0,0 +1,147 @@

+/*

+ * Format Conversion Utils

+ * Copyright (c) 2000, 2001 Fabrice Bellard

+ * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>

+ *

+ * This file is part of FFmpeg.

+ *

+ * FFmpeg is free software; you can redistribute it and/or

+ * modify it under the terms of the GNU Lesser General Public

+ * License as published by the Free Software Foundation; either

+ * version 2.1 of the License, or (at your option) any later version.

+ *

+ * FFmpeg is distributed in the hope that it will be useful,

+ * but WITHOUT ANY WARRANTY; without even the implied warranty of

+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

+ * Lesser General Public License for more details.

+ *

+ * You should have received a copy of the GNU Lesser General Public

+ * License along with FFmpeg; if not, write to the Free Software

+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

+ *

+ * MMX optimization by Nick Kurshev <nickols_k@mail.ru>

+ */

+

+#include "libavutil/cpu.h"

+#include "libavutil/x86/asm.h"

+#include "libavcodec/fmtconvert.h"

+#include "libavcodec/dsputil.h"

+

+#if HAVE_YASM

+

+void ff_int32_to_float_fmul_scalar_sse (float *dst, const int *src, float mul, int len);

+void ff_int32_to_float_fmul_scalar_sse2(float *dst, const int *src, float mul, int len);

+

+void ff_float_to_int16_3dnow(int16_t *dst, const float *src, long len);

+void ff_float_to_int16_sse  (int16_t *dst, const float *src, long len);

+void ff_float_to_int16_sse2 (int16_t *dst, const float *src, long len);

+

+void ff_float_to_int16_step_3dnow(int16_t *dst, const float *src, long len, long step);

+void ff_float_to_int16_step_sse  (int16_t *dst, const float *src, long len, long step);

+void ff_float_to_int16_step_sse2 (int16_t *dst, const float *src, long len, long step);

+

+void ff_float_to_int16_interleave2_3dnow(int16_t *dst, const float **src, long len);

+void ff_float_to_int16_interleave2_sse  (int16_t *dst, const float **src, long len);

+void ff_float_to_int16_interleave2_sse2 (int16_t *dst, const float **src, long len);

+

+void ff_float_to_int16_interleave6_sse(int16_t *dst, const float **src, int len);

+void ff_float_to_int16_interleave6_3dnow(int16_t *dst, const float **src, int len);

+void ff_float_to_int16_interleave6_3dnowext(int16_t *dst, const float **src, int len);

+

+#define ff_float_to_int16_interleave6_sse2 ff_float_to_int16_interleave6_sse

+

+#define FLOAT_TO_INT16_INTERLEAVE(cpu) \

+/* gcc pessimizes register allocation if this is in the same function as float_to_int16_interleave_sse2*/\

+static av_noinline void float_to_int16_interleave_misc_##cpu(int16_t *dst, const float **src, long len, int channels){\

+    int c;\

+    for(c=0; c<channels; c++){\

+        ff_float_to_int16_step_##cpu(dst+c, src[c], len, channels);\

+    }\

+}\

+\

+static void float_to_int16_interleave_##cpu(int16_t *dst, const float **src, long len, int channels){\

+    if(channels==1)\

+        ff_float_to_int16_##cpu(dst, src[0], len);\

+    else if(channels==2){\

+        ff_float_to_int16_interleave2_##cpu(dst, src, len);\

+    }else if(channels==6){\

+        ff_float_to_int16_interleave6_##cpu(dst, src, len);\

+    }else\

+        float_to_int16_interleave_misc_##cpu(dst, src, len, channels);\

+}

+

+FLOAT_TO_INT16_INTERLEAVE(3dnow)

+FLOAT_TO_INT16_INTERLEAVE(sse)

+FLOAT_TO_INT16_INTERLEAVE(sse2)

+

+static void float_to_int16_interleave_3dnowext(int16_t *dst, const float **src,

+                                               long len, int channels)

+{

+    if(channels==6)

+        ff_float_to_int16_interleave6_3dnowext(dst, src, len);

+    else

+        float_to_int16_interleave_3dnow(dst, src, len, channels);

+}

+

+void ff_float_interleave2_mmx(float *dst, const float **src, unsigned int len);

+void ff_float_interleave2_sse(float *dst, const float **src, unsigned int len);

+

+void ff_float_interleave6_mmx(float *dst, const float **src, unsigned int len);

+void ff_float_interleave6_sse(float *dst, const float **src, unsigned int len);

+

+static void float_interleave_mmx(float *dst, const float **src,

+                                 unsigned int len, int channels)

+{

+    if (channels == 2) {

+        ff_float_interleave2_mmx(dst, src, len);

+    } else if (channels == 6)

+        ff_float_interleave6_mmx(dst, src, len);

+    else

+        ff_float_interleave_c(dst, src, len, channels);

+}

+

+static void float_interleave_sse(float *dst, const float **src,

+                                 unsigned int len, int channels)

+{

+    if (channels == 2) {

+        ff_float_interleave2_sse(dst, src, len);

+    } else if (channels == 6)

+        ff_float_interleave6_sse(dst, src, len);

+    else

+        ff_float_interleave_c(dst, src, len, channels);

+}

+#endif

+

+void ff_fmt_convert_init_x86(FmtConvertContext *c, AVCodecContext *avctx)

+{

+#if HAVE_YASM

+    int mm_flags = av_get_cpu_flags();

+

+    if (mm_flags & AV_CPU_FLAG_MMX) {

+        c->float_interleave = float_interleave_mmx;

+

+        if (HAVE_AMD3DNOW && mm_flags & AV_CPU_FLAG_3DNOW) {

+            if(!(avctx->flags & CODEC_FLAG_BITEXACT)){

+                c->float_to_int16 = ff_float_to_int16_3dnow;

+                c->float_to_int16_interleave = float_to_int16_interleave_3dnow;

+            }

+        }

+        if (HAVE_AMD3DNOWEXT && mm_flags & AV_CPU_FLAG_3DNOWEXT) {

+            if(!(avctx->flags & CODEC_FLAG_BITEXACT)){

+                c->float_to_int16_interleave = float_to_int16_interleave_3dnowext;

+            }

+        }

+        if (HAVE_SSE && mm_flags & AV_CPU_FLAG_SSE) {

+            c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_sse;

+            c->float_to_int16 = ff_float_to_int16_sse;

+            c->float_to_int16_interleave = float_to_int16_interleave_sse;

+            c->float_interleave = float_interleave_sse;

+        }

+        if (HAVE_SSE && mm_flags & AV_CPU_FLAG_SSE2) {

+            c->int32_to_float_fmul_scalar = ff_int32_to_float_fmul_scalar_sse2;

+            c->float_to_int16 = ff_float_to_int16_sse2;

+            c->float_to_int16_interleave = float_to_int16_interleave_sse2;

+        }

+    }

+#endif

+}

deleted file mode 100644

@@ -1,147 +0,0 @@

-/*

- * Format Conversion Utils

- * Copyright (c) 2000, 2001 Fabrice Bellard

- * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>

- *

- * This file is part of FFmpeg.

- *

- * FFmpeg is free software; you can redistribute it and/or

- * modify it under the terms of the GNU Lesser General Public

- * License as published by the Free Software Foundation; either

- * version 2.1 of the License, or (at your option) any later version.

- *

- * FFmpeg is distributed in the hope that it will be useful,

- * but WITHOUT ANY WARRANTY; without even the implied warranty of

- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU

- * Lesser General Public License for more details.

- *

- * You should have received a copy of the GNU Lesser General Public

- * License along with FFmpeg; if not, write to the Free Software

- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

- *

- * MMX optimization by Nick Kurshev <nickols_k@mail.ru>

- */

-

-#include "libavutil/cpu.h"

-#include "libavutil/x86/asm.h"

-#include "libavcodec/fmtconvert.h"

-#include "libavcodec/dsputil.h"

-

-#if HAVE_YASM

-

-void ff_int32_to_float_fmul_scalar_sse (float *dst, const int *src, float mul, int len);

-void ff_int32_to_float_fmul_scalar_sse2(float *dst, const int *src, float mul, int len);

-

-void ff_float_to_int16_3dnow(int16_t *dst, const float *src, long len);

-void ff_float_to_int16_sse  (int16_t *dst, const float *src, long len);

-void ff_float_to_int16_sse2 (int16_t *dst, const float *src, long len);

-

-void ff_float_to_int16_step_3dnow(int16_t *dst, const float *src, long len, long step);

-void ff_float_to_int16_step_sse  (int16_t *dst, const float *src, long len, long step);

-void ff_float_to_int16_step_sse2 (int16_t *dst, const float *src, long len, long step);

-

-void ff_float_to_int16_interleave2_3dnow(int16_t *dst, const float **src, long len);

-void ff_float_to_int16_interleave2_sse  (int16_t *dst, const float **src, long len);

-void ff_float_to_int16_interleave2_sse2 (int16_t *dst, const float **src, long len);

-

-void ff_float_to_int16_interleave6_sse(int16_t *dst, const float **src, int len);

-void ff_float_to_int16_interleave6_3dnow(int16_t *dst, const float **src, int len);

-void ff_float_to_int16_interleave6_3dnowext(int16_t *dst, const float **src, int len);

-

-#define ff_float_to_int16_interleave6_sse2 ff_float_to_int16_interleave6_sse

-

-#define FLOAT_TO_INT16_INTERLEAVE(cpu) \

-/* gcc pessimizes register allocation if this is in the same function as float_to_int16_interleave_sse2*/\

-static av_noinline void float_to_int16_interleave_misc_##cpu(int16_t *dst, const float **src, long len, int channels){\

-    int c;\

-    for(c=0; c<channels; c++){\

-        ff_float_to_int16_step_##cpu(dst+c, src[c], len, channels);\

-    }\

-}\

-\

-static void float_to_int16_interleave_##cpu(int16_t *dst, const float **src, long len, int channels){\

-    if(channels==1)\

-        ff_float_to_int16_##cpu(dst, src[0], len);\

-    else if(channels==2){\

-        ff_float_to_int16_interleave2_##cpu(dst, src, len);\

-    }else if(channels==6){\

-        ff_float_to_int16_interleave6_##cpu(dst, src, len);\

-    }else\

-        float_to_int16_interleave_misc_##cpu(dst, src, len, channels);\

-}

-

-FLOAT_TO_INT16_INTERLEAVE(3dnow)

-FLOAT_TO_INT16_INTERLEAVE(sse)

-FLOAT_TO_INT16_INTERLEAVE(sse2)

-