Merge remote-tracking branch 'qatar/master'

* qatar/master:
  rv30: return AVERROR(EINVAL) instead of EINVAL
  build: add -L flags before existing LDFLAGS
  simple_idct: whitespace cosmetics
  simple_idct: make repeated code a macro
  dsputil: remove huge #if 0 block
  simple_idct: change 10-bit add/put stride from pixels to bytes
  dsputil: allow 9/10-bit functions for non-h264 codecs
  dnxhd: rename some data tables
  dnxhdenc: remove inline from function only called through pointer
  dnxhdenc: whitespace cosmetics
  swscale: mark YUV422P10(LE,BE) as supported for output
  configure: add -xc99 to LDFLAGS for Sun CC
  Remove unused and non-compiling vestigial g729 decoder
  Remove unused code under G729_BITEXACT #ifdef.
  mpegvideo: fix invalid picture unreferencing.
  dsputil: Remove extra blank line at end.
  dsputil: Replace a LONG_MAX check with HAVE_FAST_64BIT.
  simple_idct: add 10-bit version

Conflicts:
	Makefile
	libavcodec/g729data.h
	libavcodec/g729dec.c
	libavcodec/rv30.c
	tests/ref/lavfi/pixdesc
	tests/ref/lavfi/pixfmts_copy
	tests/ref/lavfi/pixfmts_null
	tests/ref/lavfi/pixfmts_scale
	tests/ref/lavfi/pixfmts_vflip
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Makefile

@@ -27,8 +27,6 @@ ALLPROGS    = $(BASENAMES:%=%$(EXESUF))
 ALLPROGS_G  = $(BASENAMES:%=%_g$(EXESUF))
 ALLMANPAGES = $(BASENAMES:%=%.1)
 
-ALLFFLIBS = avcodec avdevice avfilter avformat avutil postproc swscale
-
 FFLIBS-$(CONFIG_AVDEVICE) += avdevice
 FFLIBS-$(CONFIG_AVFILTER) += avfilter
 FFLIBS-$(CONFIG_AVFORMAT) += avformat

common.mak

@@ -20,6 +20,8 @@ $(foreach VAR,$(SILENT),$(eval override $(VAR) = @$($(VAR))))
 $(eval INSTALL = @$(call ECHO,INSTALL,$$(^:$(SRC_DIR)/%=%)); $(INSTALL))
 endif
 
+ALLFFLIBS = avcodec avdevice avfilter avformat avutil postproc swscale
+
 # NASM requires -I path terminated with /
 IFLAGS     := -I. -I$(SRC_PATH)/
 CPPFLAGS   := $(IFLAGS) $(CPPFLAGS)
@@ -27,7 +29,7 @@ CFLAGS     += $(ECFLAGS)
 CCFLAGS     = $(CFLAGS)
 YASMFLAGS  += $(IFLAGS) -Pconfig.asm
 HOSTCFLAGS += $(IFLAGS)
-LDFLAGS    += $(ALLFFLIBS:%=-Llib%)
+LDFLAGS    := $(ALLFFLIBS:%=-Llib%) $(LDFLAGS)
 
 define COMPILE
        $($(1)DEP)

configure

@@ -2049,6 +2049,7 @@ elif $cc -V 2>&1 | grep -q Sun; then
     cc_ident=$($cc -V 2>&1 | head -n1 | cut -d' ' -f 2-)
     DEPEND_CMD='$(DEPCC) $(DEPFLAGS) $< | sed -e "1s,^.*: ,$@: ," -e "\$$!s,\$$, \\\," -e "1!s,^.*: , ," > $(@:.o=.d)'
     DEPFLAGS='$(CPPFLAGS) $(CFLAGS) -xM1'
+    add_ldflags -xc99
     speed_cflags='-O5'
     size_cflags='-O5 -xspace'
     filter_cflags=suncc_flags

libavcodec/acelp_pitch_delay.c

@@ -105,20 +105,9 @@ int16_t ff_acelp_decode_gain_code(
     for(i=0; i<ma_pred_order; i++)
         mr_energy += quant_energy[i] * ma_prediction_coeff[i];
 
-#ifdef G729_BITEXACT
-    mr_energy += (((-6165LL * ff_log2(dsp->scalarproduct_int16(fc_v, fc_v, subframe_size, 0))) >> 3) & ~0x3ff);
-
-    mr_energy = (5439 * (mr_energy >> 15)) >> 8;           // (0.15) = (0.15) * (7.23)
-
-    return bidir_sal(
-               ((ff_exp2(mr_energy & 0x7fff) + 16) >> 5) * (gain_corr_factor >> 1),
-               (mr_energy >> 15) - 25
-           );
-#else
     mr_energy = gain_corr_factor * exp(M_LN10 / (20 << 23) * mr_energy) /
                 sqrt(dsp->scalarproduct_int16(fc_v, fc_v, subframe_size, 0));
     return mr_energy >> 12;
-#endif
 }
 
 float ff_amr_set_fixed_gain(float fixed_gain_factor, float fixed_mean_energy,

libavcodec/acelp_vectors.c

@@ -48,26 +48,6 @@ const uint8_t ff_fc_2pulses_9bits_track1_gray[16] =
   28, 26,
 };
 
-const uint8_t ff_fc_2pulses_9bits_track2_gray[32] =
-{
-  0,  2,
-  5,  4,
-  12, 10,
-  7,  9,
-  25, 24,
-  20, 22,
-  14, 15,
-  19, 17,
-  36, 31,
-  21, 26,
-  1,  6,
-  16, 11,
-  27, 29,
-  32, 30,
-  39, 37,
-  34, 35,
-};
-
 const uint8_t ff_fc_4pulses_8bits_tracks_13[16] =
 {
   0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75,

libavcodec/acelp_vectors.h

@@ -81,37 +81,6 @@ extern const uint8_t ff_fc_4pulses_8bits_track_4[32];
 extern const uint8_t ff_fc_2pulses_9bits_track1[16];
 extern const uint8_t ff_fc_2pulses_9bits_track1_gray[16];
 
-/**
- * Track|Pulse|        Positions
- * -----------------------------------------
- *  2   | 1   | 0, 7, 14, 20, 27, 34,  1, 21
- *      |     | 2, 9, 15, 22, 29, 35,  6, 26
- *      |     | 4,10, 17, 24, 30, 37, 11, 31
- *      |     | 5,12, 19, 25, 32, 39, 16, 36
- * -----------------------------------------
- *
- * @remark Track in the table should be read top-to-bottom, left-to-right.
- *
- * @note (EE.1) This table (from the reference code) does not comply with
- *              the specification.
- *              The specification contains the following table:
- *
- * Track|Pulse|        Positions
- * -----------------------------------------
- *  2   | 1   | 0, 5, 10, 15, 20, 25, 30, 35
- *      |     | 1, 6, 11, 16, 21, 26, 31, 36
- *      |     | 2, 7, 12, 17, 22, 27, 32, 37
- *      |     | 4, 9, 14, 19, 24, 29, 34, 39
- *
- * -----------------------------------------
- *
- * @note (EE.2) Reference G.729D code also uses gray decoding for each
- *              pulse index before looking up the value in the table.
- *
- * Used in G.729 @@6.4k (with gray coding)
- */
-extern const uint8_t ff_fc_2pulses_9bits_track2_gray[32];
-
 /**
  * b60 hamming windowed sinc function coefficients
  */

libavcodec/alpha/dsputil_alpha.c

@@ -335,7 +335,7 @@ void dsputil_init_alpha(DSPContext* c, AVCodecContext *avctx)
     put_pixels_clamped_axp_p = c->put_pixels_clamped;
     add_pixels_clamped_axp_p = c->add_pixels_clamped;
 
-    if (!avctx->lowres &&
+    if (!avctx->lowres && avctx->bits_per_raw_sample <= 8 &&
         (avctx->idct_algo == FF_IDCT_AUTO ||
          avctx->idct_algo == FF_IDCT_SIMPLEALPHA)) {
         c->idct_put = ff_simple_idct_put_axp;

libavcodec/amrnbdec.c

@@ -83,7 +83,7 @@
 /** Maximum sharpening factor
  *
  * The specification says 0.8, which should be 13107, but the reference C code
- * uses 13017 instead. (Amusingly the same applies to SHARP_MAX in g729dec.c.)
+ * uses 13017 instead. (Amusingly the same applies to SHARP_MAX in bitexact G.729.)
  */
 #define SHARP_MAX 0.79449462890625
 

libavcodec/arm/dsputil_init_arm.c

@@ -80,7 +80,7 @@ void dsputil_init_arm(DSPContext* c, AVCodecContext *avctx)
     ff_put_pixels_clamped = c->put_pixels_clamped;
     ff_add_pixels_clamped = c->add_pixels_clamped;
 
-    if (!avctx->lowres) {
+    if (!avctx->lowres && avctx->bits_per_raw_sample <= 8) {
         if(avctx->idct_algo == FF_IDCT_AUTO ||
            avctx->idct_algo == FF_IDCT_ARM){
             c->idct_put              = j_rev_dct_arm_put;

libavcodec/arm/dsputil_init_armv5te.c

@@ -29,8 +29,9 @@ void ff_prefetch_arm(void *mem, int stride, int h);
 
 void av_cold ff_dsputil_init_armv5te(DSPContext* c, AVCodecContext *avctx)
 {
-    if (!avctx->lowres && (avctx->idct_algo == FF_IDCT_AUTO ||
-                           avctx->idct_algo == FF_IDCT_SIMPLEARMV5TE)) {
+    if (!avctx->lowres && avctx->bits_per_raw_sample <= 8 &&
+        (avctx->idct_algo == FF_IDCT_AUTO ||
+         avctx->idct_algo == FF_IDCT_SIMPLEARMV5TE)) {
         c->idct_put              = ff_simple_idct_put_armv5te;
         c->idct_add              = ff_simple_idct_add_armv5te;
         c->idct                  = ff_simple_idct_armv5te;

libavcodec/arm/dsputil_init_armv6.c

@@ -74,8 +74,9 @@ void av_cold ff_dsputil_init_armv6(DSPContext* c, AVCodecContext *avctx)
 {
     const int high_bit_depth = avctx->codec_id == CODEC_ID_H264 && avctx->bits_per_raw_sample > 8;
 
-    if (!avctx->lowres && (avctx->idct_algo == FF_IDCT_AUTO ||
-                           avctx->idct_algo == FF_IDCT_SIMPLEARMV6)) {
+    if (!avctx->lowres && avctx->bits_per_raw_sample <= 8 &&
+        (avctx->idct_algo == FF_IDCT_AUTO ||
+         avctx->idct_algo == FF_IDCT_SIMPLEARMV6)) {
         c->idct_put              = ff_simple_idct_put_armv6;
         c->idct_add              = ff_simple_idct_add_armv6;
         c->idct                  = ff_simple_idct_armv6;

libavcodec/arm/dsputil_init_neon.c

@@ -177,7 +177,7 @@ void ff_dsputil_init_neon(DSPContext *c, AVCodecContext *avctx)
 {
     const int high_bit_depth = avctx->codec_id == CODEC_ID_H264 && avctx->bits_per_raw_sample > 8;
 
-    if (!avctx->lowres) {
+    if (!avctx->lowres && avctx->bits_per_raw_sample <= 8) {
         if (avctx->idct_algo == FF_IDCT_AUTO ||
             avctx->idct_algo == FF_IDCT_SIMPLENEON) {
             c->idct_put              = ff_simple_idct_put_neon;

libavcodec/bfin/dsputil_bfin.c

@@ -256,16 +256,18 @@ void dsputil_init_bfin( DSPContext* c, AVCodecContext *avctx )
     if (avctx->dct_algo == FF_DCT_AUTO)
         c->fdct               = ff_bfin_fdct;
 
-    if (avctx->idct_algo==FF_IDCT_VP3) {
-        c->idct_permutation_type = FF_NO_IDCT_PERM;
-        c->idct               = ff_bfin_vp3_idct;
-        c->idct_add           = ff_bfin_vp3_idct_add;
-        c->idct_put           = ff_bfin_vp3_idct_put;
-    } else if (avctx->idct_algo == FF_IDCT_AUTO) {
-        c->idct_permutation_type = FF_NO_IDCT_PERM;
-        c->idct               = ff_bfin_idct;
-        c->idct_add           = bfin_idct_add;
-        c->idct_put           = bfin_idct_put;
+    if (avctx->bits_per_raw_sample <= 8) {
+        if (avctx->idct_algo == FF_IDCT_VP3) {
+            c->idct_permutation_type = FF_NO_IDCT_PERM;
+            c->idct                  = ff_bfin_vp3_idct;
+            c->idct_add              = ff_bfin_vp3_idct_add;
+            c->idct_put              = ff_bfin_vp3_idct_put;
+        } else if (avctx->idct_algo == FF_IDCT_AUTO) {
+            c->idct_permutation_type = FF_NO_IDCT_PERM;
+            c->idct                  = ff_bfin_idct;
+            c->idct_add              = bfin_idct_add;
+            c->idct_put              = bfin_idct_put;
+        }
     }
 }
 

libavcodec/celp_math.c

@@ -27,82 +27,6 @@
 #include "avcodec.h"
 #include "celp_math.h"
 
-#ifdef G729_BITEXACT
-/**
- * Cosine table: base_cos[i] = (1<<15) * cos(i*PI/64)
- */
-static const int16_t base_cos[64] =
-{
-  32767,  32729,  32610,  32413,  32138,  31786,  31357,  30853,
-  30274,  29622,  28899,  28106,  27246,  26320,  25330,  24279,
-  23170,  22006,  20788,  19520,  18205,  16846,  15447,  14010,
-  12540,  11039,   9512,   7962,   6393,   4808,   3212,   1608,
-      0,  -1608,  -3212,  -4808,  -6393,  -7962,  -9512, -11039,
- -12540, -14010, -15447, -16846, -18205, -19520, -20788, -22006,
- -23170, -24279, -25330, -26320, -27246, -28106, -28899, -29622,
- -30274, -30853, -31357, -31786, -32138, -32413, -32610, -32729
-};
-
-/**
- * Slope used to compute cos(x)
- *
- * cos(ind*64+offset) = base_cos[ind]+offset*slope_cos[ind]
- * values multiplied by 1<<19
- */
-static const int16_t slope_cos[64] =
-{
-   -632,  -1893,  -3150,  -4399,  -5638,  -6863,  -8072,  -9261,
- -10428, -11570, -12684, -13767, -14817, -15832, -16808, -17744,
- -18637, -19486, -20287, -21039, -21741, -22390, -22986, -23526,
- -24009, -24435, -24801, -25108, -25354, -25540, -25664, -25726,
- -25726, -25664, -25540, -25354, -25108, -24801, -24435, -24009,
- -23526, -22986, -22390, -21741, -21039, -20287, -19486, -18637,
- -17744, -16808, -15832, -14817, -13767, -12684, -11570, -10428,
-  -9261,  -8072,  -6863,  -5638,  -4399,  -3150,  -1893,   -632
-};
-
-/**
- * Table used to compute exp2(x)
- *
- * tab_exp2[i] = (1<<14) * exp2(i/32) = 2^(i/32) i=0..32
- */
-static const uint16_t tab_exp2[33] =
-{
-  16384, 16743, 17109, 17484, 17867, 18258, 18658, 19066, 19484, 19911,
-  20347, 20792, 21247, 21713, 22188, 22674, 23170, 23678, 24196, 24726,
-  25268, 25821, 26386, 26964, 27554, 28158, 28774, 29405, 30048, 30706,
-  31379, 32066, 32767
-};
-
-int16_t ff_cos(uint16_t arg)
-{
-    uint8_t offset= arg;
-    uint8_t ind = arg >> 8;
-
-    assert(arg < 0x4000);
-
-    return FFMAX(base_cos[ind] + ((slope_cos[ind] * offset) >> 12), -0x8000);
-}
-
-int ff_exp2(uint16_t power)
-{
-    uint16_t frac_x0;
-    uint16_t frac_dx;
-    int result;
-
-    assert(power <= 0x7fff);
-
-    frac_x0 = power >> 10;
-    frac_dx = (power & 0x03ff) << 5;
-
-    result = tab_exp2[frac_x0] << 15;
-    result += frac_dx * (tab_exp2[frac_x0+1] - tab_exp2[frac_x0]);
-
-    return result >> 10;
-}
-
-#else // G729_BITEXACT
-
 /**
  * Cosine table: base_cos[i] = (1<<15) * cos(i*PI/64)
  */
@@ -154,8 +78,6 @@ int ff_exp2(uint16_t power)
     return result + ((result*(power&31)*89)>>22);
 }
 
-#endif // else G729_BITEXACT
-
 /**
  * Table used to compute log2(x)
  *
@@ -163,17 +85,10 @@ int ff_exp2(uint16_t power)
  */
 static const uint16_t tab_log2[33] =
 {
-#ifdef G729_BITEXACT
-      0,   1455,   2866,   4236,   5568,   6863,   8124,   9352,
-  10549,  11716,  12855,  13967,  15054,  16117,  17156,  18172,
-  19167,  20142,  21097,  22033,  22951,  23852,  24735,  25603,
-  26455,  27291,  28113,  28922,  29716,  30497,  31266,  32023,  32767,
-#else
       4,   1459,   2870,   4240,   5572,   6867,   8127,   9355,
   10552,  11719,  12858,  13971,  15057,  16120,  17158,  18175,
   19170,  20145,  21100,  22036,  22954,  23854,  24738,  25605,
   26457,  27294,  28116,  28924,  29719,  30500,  31269,  32025,  32769,
-#endif
 };
 
 int ff_log2(uint32_t value)

libavcodec/dct-test.c

@@ -111,7 +111,7 @@ static const struct algo idct_tab[] = {
     { "FAANI",          ff_faanidct,           NO_PERM  },
     { "REF-DBL",        ff_ref_idct,           NO_PERM  },
     { "INT",            j_rev_dct,             MMX_PERM },
-    { "SIMPLE-C",       ff_simple_idct,        NO_PERM  },
+    { "SIMPLE-C",       ff_simple_idct_8,      NO_PERM  },
 
 #if HAVE_MMX
 #if CONFIG_GPL

libavcodec/dnxhdenc.c

@@ -41,7 +41,7 @@ static const AVClass class = { "dnxhd", av_default_item_name, options, LIBAVUTIL
 
 #define LAMBDA_FRAC_BITS 10
 
-static av_always_inline void dnxhd_get_pixels_8x4(DCTELEM *restrict block, const uint8_t *pixels, int line_size)
+static void dnxhd_get_pixels_8x4(DCTELEM *restrict block, const uint8_t *pixels, int line_size)
 {
     int i;
     for (i = 0; i < 4; i++) {
@@ -52,10 +52,10 @@ static av_always_inline void dnxhd_get_pixels_8x4(DCTELEM *restrict block, const
         pixels += line_size;
         block += 8;
     }
-    memcpy(block   , block- 8, sizeof(*block)*8);
-    memcpy(block+ 8, block-16, sizeof(*block)*8);
-    memcpy(block+16, block-24, sizeof(*block)*8);
-    memcpy(block+24, block-32, sizeof(*block)*8);
+    memcpy(block,      block -  8, sizeof(*block) * 8);
+    memcpy(block +  8, block - 16, sizeof(*block) * 8);
+    memcpy(block + 16, block - 24, sizeof(*block) * 8);
+    memcpy(block + 24, block - 32, sizeof(*block) * 8);
 }
 
 static int dnxhd_init_vlc(DNXHDEncContext *ctx)
@@ -64,9 +64,9 @@ static int dnxhd_init_vlc(DNXHDEncContext *ctx)
     int max_level = 1<<(ctx->cid_table->bit_depth+2);
 
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->vlc_codes, max_level*4*sizeof(*ctx->vlc_codes), fail);
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->vlc_bits , max_level*4*sizeof(*ctx->vlc_bits ), fail);
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->run_codes, 63*2                               , fail);
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->run_bits , 63                                 , fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->vlc_bits,  max_level*4*sizeof(*ctx->vlc_bits) , fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->run_codes, 63*2,                                fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->run_bits,  63,                                  fail);
 
     ctx->vlc_codes += max_level*2;
     ctx->vlc_bits  += max_level*2;
@@ -119,8 +119,8 @@ static int dnxhd_init_qmat(DNXHDEncContext *ctx, int lbias, int cbias)
     uint16_t weight_matrix[64] = {1,}; // convert_matrix needs uint16_t*
     int qscale, i;
 
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_l,   (ctx->m.avctx->qmax+1) * 64 *     sizeof(int)     , fail);
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_c,   (ctx->m.avctx->qmax+1) * 64 *     sizeof(int)     , fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_l,   (ctx->m.avctx->qmax+1) * 64 *     sizeof(int),      fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_c,   (ctx->m.avctx->qmax+1) * 64 *     sizeof(int),      fail);
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_l16, (ctx->m.avctx->qmax+1) * 64 * 2 * sizeof(uint16_t), fail);
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->qmatrix_c16, (ctx->m.avctx->qmax+1) * 64 * 2 * sizeof(uint16_t), fail);
 
@@ -218,7 +218,7 @@ static int dnxhd_encode_init(AVCodecContext *avctx)
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->slice_size, ctx->m.mb_height*sizeof(uint32_t), fail);
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->slice_offs, ctx->m.mb_height*sizeof(uint32_t), fail);
     FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->mb_bits,    ctx->m.mb_num   *sizeof(uint16_t), fail);
-    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->mb_qscale,  ctx->m.mb_num   *sizeof(uint8_t) , fail);
+    FF_ALLOCZ_OR_GOTO(ctx->m.avctx, ctx->mb_qscale,  ctx->m.mb_num   *sizeof(uint8_t),  fail);
 
     ctx->frame.key_frame = 1;
     ctx->frame.pict_type = AV_PICTURE_TYPE_I;
@@ -341,7 +341,7 @@ static av_always_inline int dnxhd_ssd_block(DCTELEM *qblock, DCTELEM *block)
     int score = 0;
     int i;
     for (i = 0; i < 64; i++)
-        score += (block[i]-qblock[i])*(block[i]-qblock[i]);
+        score += (block[i] - qblock[i]) * (block[i] - qblock[i]);
     return score;
 }
 
@@ -369,26 +369,28 @@ static av_always_inline void dnxhd_get_blocks(DNXHDEncContext *ctx, int mb_x, in
     const uint8_t *ptr_v = ctx->thread[0]->src[2] + ((mb_y << 4) * ctx->m.uvlinesize) + (mb_x << 3);
     DSPContext *dsp = &ctx->m.dsp;
 
-    dsp->get_pixels(ctx->blocks[0], ptr_y    , ctx->m.linesize);
+    dsp->get_pixels(ctx->blocks[0], ptr_y,     ctx->m.linesize);
     dsp->get_pixels(ctx->blocks[1], ptr_y + 8, ctx->m.linesize);
-    dsp->get_pixels(ctx->blocks[2], ptr_u    , ctx->m.uvlinesize);
-    dsp->get_pixels(ctx->blocks[3], ptr_v    , ctx->m.uvlinesize);
+    dsp->get_pixels(ctx->blocks[2], ptr_u,     ctx->m.uvlinesize);
+    dsp->get_pixels(ctx->blocks[3], ptr_v,     ctx->m.uvlinesize);
 
     if (mb_y+1 == ctx->m.mb_height && ctx->m.avctx->height == 1080) {
         if (ctx->interlaced) {
-            ctx->get_pixels_8x4_sym(ctx->blocks[4], ptr_y + ctx->dct_y_offset    , ctx->m.linesize);
+            ctx->get_pixels_8x4_sym(ctx->blocks[4], ptr_y + ctx->dct_y_offset,     ctx->m.linesize);
             ctx->get_pixels_8x4_sym(ctx->blocks[5], ptr_y + ctx->dct_y_offset + 8, ctx->m.linesize);
-            ctx->get_pixels_8x4_sym(ctx->blocks[6], ptr_u + ctx->dct_uv_offset   , ctx->m.uvlinesize);
-            ctx->get_pixels_8x4_sym(ctx->blocks[7], ptr_v + ctx->dct_uv_offset   , ctx->m.uvlinesize);
+            ctx->get_pixels_8x4_sym(ctx->blocks[6], ptr_u + ctx->dct_uv_offset,    ctx->m.uvlinesize);
+            ctx->get_pixels_8x4_sym(ctx->blocks[7], ptr_v + ctx->dct_uv_offset,    ctx->m.uvlinesize);
         } else {
-            dsp->clear_block(ctx->blocks[4]); dsp->clear_block(ctx->blocks[5]);
-            dsp->clear_block(ctx->blocks[6]); dsp->clear_block(ctx->blocks[7]);
+            dsp->clear_block(ctx->blocks[4]);
+            dsp->clear_block(ctx->blocks[5]);
+            dsp->clear_block(ctx->blocks[6]);
+            dsp->clear_block(ctx->blocks[7]);
         }
     } else {
-        dsp->get_pixels(ctx->blocks[4], ptr_y + ctx->dct_y_offset    , ctx->m.linesize);
+        dsp->get_pixels(ctx->blocks[4], ptr_y + ctx->dct_y_offset,     ctx->m.linesize);
         dsp->get_pixels(ctx->blocks[5], ptr_y + ctx->dct_y_offset + 8, ctx->m.linesize);
-        dsp->get_pixels(ctx->blocks[6], ptr_u + ctx->dct_uv_offset   , ctx->m.uvlinesize);
-        dsp->get_pixels(ctx->blocks[7], ptr_v + ctx->dct_uv_offset   , ctx->m.uvlinesize);
+        dsp->get_pixels(ctx->blocks[6], ptr_u + ctx->dct_uv_offset,    ctx->m.uvlinesize);
+        dsp->get_pixels(ctx->blocks[7], ptr_v + ctx->dct_uv_offset,    ctx->m.uvlinesize);
     }
 }
 
@@ -496,14 +498,14 @@ static void dnxhd_setup_threads_slices(DNXHDEncContext *ctx)
     for (mb_y = 0; mb_y < ctx->m.mb_height; mb_y++) {
         int thread_size;
         ctx->slice_offs[mb_y] = offset;
-            ctx->slice_size[mb_y] = 0;
-            for (mb_x = 0; mb_x < ctx->m.mb_width; mb_x++) {
-                unsigned mb = mb_y * ctx->m.mb_width + mb_x;
-                ctx->slice_size[mb_y] += ctx->mb_bits[mb];
-            }
-            ctx->slice_size[mb_y] = (ctx->slice_size[mb_y]+31)&~31;
-            ctx->slice_size[mb_y] >>= 3;
-            thread_size = ctx->slice_size[mb_y];
+        ctx->slice_size[mb_y] = 0;
+        for (mb_x = 0; mb_x < ctx->m.mb_width; mb_x++) {
+            unsigned mb = mb_y * ctx->m.mb_width + mb_x;
+            ctx->slice_size[mb_y] += ctx->mb_bits[mb];
+        }
+        ctx->slice_size[mb_y] = (ctx->slice_size[mb_y]+31)&~31;
+        ctx->slice_size[mb_y] >>= 3;
+        thread_size = ctx->slice_size[mb_y];
         offset += thread_size;
     }
 }

libavcodec/dsputil.c

@@ -184,7 +184,7 @@ static int pix_norm1_c(uint8_t * pix, int line_size)
             s += sq[pix[6]];
             s += sq[pix[7]];
 #else
-#if LONG_MAX > 2147483647
+#if HAVE_FAST_64BIT
             register uint64_t x=*(uint64_t*)pix;
             s += sq[x&0xff];
             s += sq[(x>>8)&0xff];
@@ -2225,7 +2225,7 @@ static int quant_psnr8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *s
 
     s->block_last_index[0/*FIXME*/]= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i);
     s->dct_unquantize_inter(s, temp, 0, s->qscale);
-    ff_simple_idct(temp); //FIXME
+    ff_simple_idct_8(temp); //FIXME
 
     for(i=0; i<64; i++)
         sum+= (temp[i]-bak[i])*(temp[i]-bak[i]);
@@ -2866,6 +2866,12 @@ av_cold void dsputil_init(DSPContext* c, AVCodecContext *avctx)
         c->idct    = j_rev_dct1;
         c->idct_permutation_type= FF_NO_IDCT_PERM;
     }else{
+        if (avctx->bits_per_raw_sample == 10) {
+            c->idct_put              = ff_simple_idct_put_10;
+            c->idct_add              = ff_simple_idct_add_10;
+            c->idct                  = ff_simple_idct_10;
+            c->idct_permutation_type = FF_NO_IDCT_PERM;
+        } else {
         if(avctx->idct_algo==FF_IDCT_INT){
             c->idct_put= ff_jref_idct_put;
             c->idct_add= ff_jref_idct_add;
@@ -2896,11 +2902,12 @@ av_cold void dsputil_init(DSPContext* c, AVCodecContext *avctx)
             c->idct_put = ff_bink_idct_put_c;
             c->idct_permutation_type = FF_NO_IDCT_PERM;
         }else{ //accurate/default
-            c->idct_put= ff_simple_idct_put;
-            c->idct_add= ff_simple_idct_add;
-            c->idct    = ff_simple_idct;
+            c->idct_put = ff_simple_idct_put_8;
+            c->idct_add = ff_simple_idct_add_8;
+            c->idct     = ff_simple_idct_8;
             c->idct_permutation_type= FF_NO_IDCT_PERM;
         }
+        }
     }
 
     c->get_pixels = get_pixels_c;
@@ -3169,21 +3176,18 @@ av_cold void dsputil_init(DSPContext* c, AVCodecContext *avctx)
     dspfunc2(avg_h264_qpel, 1,  8, depth);\
     dspfunc2(avg_h264_qpel, 2,  4, depth);
 
-    if (avctx->codec_id != CODEC_ID_H264 || avctx->bits_per_raw_sample == 8) {
-        BIT_DEPTH_FUNCS(8)
-    } else {
-        switch (avctx->bits_per_raw_sample) {
-            case 9:
-                BIT_DEPTH_FUNCS(9)
-                break;
-            case 10:
-                BIT_DEPTH_FUNCS(10)
-                break;
-            default:
-                av_log(avctx, AV_LOG_DEBUG, "Unsupported bit depth: %d\n", avctx->bits_per_raw_sample);
-                BIT_DEPTH_FUNCS(8)
-                break;
-        }
+    switch (avctx->bits_per_raw_sample) {
+    case 9:
+        BIT_DEPTH_FUNCS(9);
+        break;
+    case 10:
+        BIT_DEPTH_FUNCS(10);
+        break;
+    default:
+        av_log(avctx, AV_LOG_DEBUG, "Unsupported bit depth: %d\n", avctx->bits_per_raw_sample);
+    case 8:
+        BIT_DEPTH_FUNCS(8);
+        break;
     }
 
 
@@ -3243,4 +3247,3 @@ av_cold void dsputil_init(DSPContext* c, AVCodecContext *avctx)
         av_log(avctx, AV_LOG_ERROR, "Internal error, IDCT permutation not set\n");
     }
 }
-

libavcodec/dsputil_template.c

@@ -230,150 +230,6 @@ static void FUNCC(add_pixels4)(uint8_t *restrict p_pixels, DCTELEM *p_block, int
     }
 }
 
-#if 0
-
-#define PIXOP2(OPNAME, OP) \
-static void OPNAME ## _pixels(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        OP(*((uint64_t*)block), AV_RN64(pixels));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_x2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        OP(*((uint64_t*)block), (a&b) + (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_x2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        OP(*((uint64_t*)block), (a|b) - (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_y2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels          );\
-        const uint64_t b= AV_RN64(pixels+line_size);\
-        OP(*((uint64_t*)block), (a&b) + (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_y2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-    int i;\
-    for(i=0; i<h; i++){\
-        const uint64_t a= AV_RN64(pixels          );\
-        const uint64_t b= AV_RN64(pixels+line_size);\
-        OP(*((uint64_t*)block), (a|b) - (((a^b)&0xFEFEFEFEFEFEFEFEULL)>>1));\
-        pixels+=line_size;\
-        block +=line_size;\
-    }\
-}\
-\
-static void OPNAME ## _pixels_xy2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-        int i;\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        uint64_t l0=  (a&0x0303030303030303ULL)\
-                    + (b&0x0303030303030303ULL)\
-                    + 0x0202020202020202ULL;\
-        uint64_t h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-                   + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-        uint64_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint64_t a= AV_RN64(pixels  );\
-            uint64_t b= AV_RN64(pixels+1);\
-            l1=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL);\
-            h1= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN64(pixels  );\
-            b= AV_RN64(pixels+1);\
-            l0=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL)\
-               + 0x0202020202020202ULL;\
-            h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-static void OPNAME ## _no_rnd_pixels_xy2_c(uint8_t *block, const uint8_t *pixels, int line_size, int h)\
-{\
-        int i;\
-        const uint64_t a= AV_RN64(pixels  );\
-        const uint64_t b= AV_RN64(pixels+1);\
-        uint64_t l0=  (a&0x0303030303030303ULL)\
-                    + (b&0x0303030303030303ULL)\
-                    + 0x0101010101010101ULL;\
-        uint64_t h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-                   + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-        uint64_t l1,h1;\
-\
-        pixels+=line_size;\
-        for(i=0; i<h; i+=2){\
-            uint64_t a= AV_RN64(pixels  );\
-            uint64_t b= AV_RN64(pixels+1);\
-            l1=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL);\
-            h1= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-            a= AV_RN64(pixels  );\
-            b= AV_RN64(pixels+1);\
-            l0=  (a&0x0303030303030303ULL)\
-               + (b&0x0303030303030303ULL)\
-               + 0x0101010101010101ULL;\
-            h0= ((a&0xFCFCFCFCFCFCFCFCULL)>>2)\
-              + ((b&0xFCFCFCFCFCFCFCFCULL)>>2);\
-            OP(*((uint64_t*)block), h0+h1+(((l0+l1)>>2)&0x0F0F0F0F0F0F0F0FULL));\
-            pixels+=line_size;\
-            block +=line_size;\
-        }\
-}\
-\
-CALL_2X_PIXELS(OPNAME ## _pixels16_c    , OPNAME ## _pixels_c    , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_x2_c , OPNAME ## _pixels_x2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_y2_c , OPNAME ## _pixels_y2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _pixels16_xy2_c, OPNAME ## _pixels_xy2_c, 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_x2_c , OPNAME ## _no_rnd_pixels_x2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_y2_c , OPNAME ## _no_rnd_pixels_y2_c , 8*sizeof(pixel))\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_xy2_c, OPNAME ## _no_rnd_pixels_xy2_c, 8*sizeof(pixel))
-
-#define op_avg(a, b) a = ( ((a)|(b)) - ((((a)^(b))&0xFEFEFEFEFEFEFEFEULL)>>1) )
-#else // 64 bit variant
-
 #define PIXOP2(OPNAME, OP) \
 static void FUNCC(OPNAME ## _pixels2)(uint8_t *block, const uint8_t *pixels, int line_size, int h){\
     int i;\
@@ -749,7 +605,6 @@ CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16_y2) , FUNCC(OPNAME ## _no_rnd_pi
 CALL_2X_PIXELS(FUNCC(OPNAME ## _no_rnd_pixels16_xy2), FUNCC(OPNAME ## _no_rnd_pixels8_xy2), 8*sizeof(pixel))\
 
 #define op_avg(a, b) a = rnd_avg_pixel4(a, b)
-#endif
 #define op_put(a, b) a = b
 
 PIXOP2(avg, op_avg)

libavcodec/lsp.c

@@ -150,11 +150,7 @@ void ff_acelp_lp_decode(int16_t* lp_1st, int16_t* lp_2nd, const int16_t* lsp_2nd
 
     /* LSP values for first subframe (3.2.5 of G.729, Equation 24)*/
     for(i=0; i<lp_order; i++)
-#ifdef G729_BITEXACT
-        lsp_1st[i] = (lsp_2nd[i] >> 1) + (lsp_prev[i] >> 1);
-#else
         lsp_1st[i] = (lsp_2nd[i] + lsp_prev[i]) >> 1;
-#endif
 
     ff_acelp_lsp2lpc(lp_1st, lsp_1st, lp_order >> 1);
 

libavcodec/mpegvideo.c

@@ -318,7 +318,7 @@ int ff_alloc_picture(MpegEncContext *s, Picture *pic, int shared){
     s->prev_pict_types[0]= s->dropable ? AV_PICTURE_TYPE_B : s->pict_type;
     if (pic->f.age < PREV_PICT_TYPES_BUFFER_SIZE && s->prev_pict_types[pic->f.age] == AV_PICTURE_TYPE_B)
         pic->f.age = INT_MAX; // Skipped MBs in B-frames are quite rare in MPEG-1/2 and it is a bit tricky to skip them anyway.
-    pic->owner2 = NULL;
+    pic->owner2 = s;
 
     return 0;
 fail: //for the FF_ALLOCZ_OR_GOTO macro
@@ -1036,14 +1036,16 @@ int MPV_frame_start(MpegEncContext *s, AVCodecContext *avctx)
     /* mark&release old frames */
     if (s->pict_type != AV_PICTURE_TYPE_B && s->last_picture_ptr && s->last_picture_ptr != s->next_picture_ptr && s->last_picture_ptr->f.data[0]) {
       if(s->out_format != FMT_H264 || s->codec_id == CODEC_ID_SVQ3){
-          free_frame_buffer(s, s->last_picture_ptr);
+          if (s->last_picture_ptr->owner2 == s)
+              free_frame_buffer(s, s->last_picture_ptr);
 
         /* release forgotten pictures */
         /* if(mpeg124/h263) */
         if(!s->encoding){
             for(i=0; i<s->picture_count; i++){
-                if (s->picture[i].f.data[0] && &s->picture[i] != s->next_picture_ptr && s->picture[i].f.reference) {
-                    av_log(avctx, AV_LOG_ERROR, "releasing zombie picture\n");
+                if (s->picture[i].owner2 == s && s->picture[i].f.data[0] && &s->picture[i] != s->next_picture_ptr && s->picture[i].f.reference) {
+                    if (!(avctx->active_thread_type & FF_THREAD_FRAME))
+                        av_log(avctx, AV_LOG_ERROR, "releasing zombie picture\n");
                     free_frame_buffer(s, &s->picture[i]);
                 }
             }

libavcodec/ppc/dsputil_ppc.c

@@ -178,7 +178,7 @@ void dsputil_init_ppc(DSPContext* c, AVCodecContext *avctx)
         }
 #endif //CONFIG_ENCODERS
 
-        if (avctx->lowres==0) {
+        if (avctx->lowres == 0 && avctx->bits_per_raw_sample <= 8) {
             if ((avctx->idct_algo == FF_IDCT_AUTO) ||
                 (avctx->idct_algo == FF_IDCT_ALTIVEC)) {
                 c->idct_put = idct_put_altivec;

libavcodec/ps2/dsputil_mmi.c

@@ -156,7 +156,8 @@ void dsputil_init_mmi(DSPContext* c, AVCodecContext *avctx)
 
     c->get_pixels = get_pixels_mmi;
 
-    if(idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_PS2){
+    if (avctx->bits_per_raw_sample <= 8 &&
+        (idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_PS2)) {
         c->idct_put= ff_mmi_idct_put;
         c->idct_add= ff_mmi_idct_add;
         c->idct    = ff_mmi_idct;

libavcodec/sh4/dsputil_sh4.c

@@ -97,7 +97,8 @@ void dsputil_init_sh4(DSPContext* c, AVCodecContext *avctx)
 
         if (!high_bit_depth)
         c->clear_blocks = clear_blocks_sh4;
-        if(idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_SH4){
+        if (avctx->bits_per_raw_sample <= 8 &&
+            (idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_SH4)) {
                 c->idct_put = idct_put;
                 c->idct_add = idct_add;
                c->idct     = idct_sh4;

libavcodec/simple_idct.h

@@ -31,12 +31,17 @@
 #include <stdint.h>
 #include "dsputil.h"
 
-void ff_simple_idct_put(uint8_t *dest, int line_size, DCTELEM *block);
-void ff_simple_idct_add(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_simple_idct_put_8(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_simple_idct_add_8(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_simple_idct_8(DCTELEM *block);
+
+void ff_simple_idct_put_10(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_simple_idct_add_10(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_simple_idct_10(DCTELEM *block);
+
 void ff_simple_idct_mmx(int16_t *block);
 void ff_simple_idct_add_mmx(uint8_t *dest, int line_size, int16_t *block);
 void ff_simple_idct_put_mmx(uint8_t *dest, int line_size, int16_t *block);
-void ff_simple_idct(DCTELEM *block);
 
 void ff_simple_idct248_put(uint8_t *dest, int line_size, DCTELEM *block);
 

libavcodec/simple_idct_template.c

+/*
+ * Simple IDCT
+ *
+ * Copyright (c) 2001 Michael Niedermayer <michaelni@gmx.at>
+ *
+ * This file is part of Libav.
+ *
+ * Libav 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.
+ *
+ * Libav 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 Libav; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/**
+ * @file
+ * simpleidct in C.
+ */
+
+/*
+  based upon some outcommented c code from mpeg2dec (idct_mmx.c
+  written by Aaron Holtzman <aholtzma@ess.engr.uvic.ca>)
+ */
+
+#include "bit_depth_template.c"
+
+#undef W1
+#undef W2
+#undef W3
+#undef W4
+#undef W5
+#undef W6
+#undef W7
+#undef ROW_SHIFT
+#undef COL_SHIFT
+#undef DC_SHIFT
+#undef MUL
+#undef MAC
+
+#if BIT_DEPTH == 8
+
+#define W1  22725  //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W2  21407  //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W3  19266  //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W4  16383  //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W5  12873  //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W6  8867   //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+#define W7  4520   //cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5
+
+#define ROW_SHIFT 11
+#define COL_SHIFT 20
+#define DC_SHIFT 3
+
+#define MUL(a, b)    MUL16(a, b)
+#define MAC(a, b, c) MAC16(a, b, c)
+
+#elif BIT_DEPTH == 10
+
+#define W1 90901
+#define W2 85627
+#define W3 77062
+#define W4 65535
+#define W5 51491
+#define W6 35468
+#define W7 18081
+
+#define ROW_SHIFT 15
+#define COL_SHIFT 20
+#define DC_SHIFT 1
+
+#define MUL(a, b)    ((a) * (b))
+#define MAC(a, b, c) ((a) += (b) * (c))
+
+#else
+
+#error "Unsupported bitdepth"
+
+#endif
+
+static inline void FUNC(idctRowCondDC)(DCTELEM *row)
+{
+    int a0, a1, a2, a3, b0, b1, b2, b3;
+
+#if HAVE_FAST_64BIT
+#define ROW0_MASK (0xffffLL << 48 * HAVE_BIGENDIAN)
+    if (((((uint64_t *)row)[0] & ~ROW0_MASK) | ((uint64_t *)row)[1]) == 0) {
+        uint64_t temp = (row[0] << DC_SHIFT) & 0xffff;
+        temp += temp << 16;
+        temp += temp << 32;
+        ((uint64_t *)row)[0] = temp;
+        ((uint64_t *)row)[1] = temp;
+        return;
+    }
+#else
+    if (!(((uint32_t*)row)[1] |
+          ((uint32_t*)row)[2] |
+          ((uint32_t*)row)[3] |
+          row[1])) {
+        uint32_t temp = (row[0] << DC_SHIFT) & 0xffff;
+        temp += temp << 16;
+        ((uint32_t*)row)[0]=((uint32_t*)row)[1] =
+            ((uint32_t*)row)[2]=((uint32_t*)row)[3] = temp;
+        return;
+    }
+#endif
+
+    a0 = (W4 * row[0]) + (1 << (ROW_SHIFT - 1));
+    a1 = a0;
+    a2 = a0;
+    a3 = a0;
+
+    a0 += W2 * row[2];
+    a1 += W6 * row[2];
+    a2 -= W6 * row[2];
+    a3 -= W2 * row[2];
+
+    b0 = MUL(W1, row[1]);
+    MAC(b0, W3, row[3]);
+    b1 = MUL(W3, row[1]);
+    MAC(b1, -W7, row[3]);
+    b2 = MUL(W5, row[1]);
+    MAC(b2, -W1, row[3]);
+    b3 = MUL(W7, row[1]);
+    MAC(b3, -W5, row[3]);
+
+    if (AV_RN64A(row + 4)) {
+        a0 +=   W4*row[4] + W6*row[6];
+        a1 += - W4*row[4] - W2*row[6];
+        a2 += - W4*row[4] + W2*row[6];
+        a3 +=   W4*row[4] - W6*row[6];
+
+        MAC(b0,  W5, row[5]);
+        MAC(b0,  W7, row[7]);
+
+        MAC(b1, -W1, row[5]);
+        MAC(b1, -W5, row[7]);
+
+        MAC(b2,  W7, row[5]);
+        MAC(b2,  W3, row[7]);
+
+        MAC(b3,  W3, row[5]);
+        MAC(b3, -W1, row[7]);
+    }
+
+    row[0] = (a0 + b0) >> ROW_SHIFT;
+    row[7] = (a0 - b0) >> ROW_SHIFT;
+    row[1] = (a1 + b1) >> ROW_SHIFT;
+    row[6] = (a1 - b1) >> ROW_SHIFT;
+    row[2] = (a2 + b2) >> ROW_SHIFT;
+    row[5] = (a2 - b2) >> ROW_SHIFT;
+    row[3] = (a3 + b3) >> ROW_SHIFT;
+    row[4] = (a3 - b3) >> ROW_SHIFT;
+}
+
+#define IDCT_COLS do {                                  \
+        a0 = W4 * (col[8*0] + ((1<<(COL_SHIFT-1))/W4)); \
+        a1 = a0;                                        \
+        a2 = a0;                                        \
+        a3 = a0;                                        \
+                                                        \
+        a0 +=  W2*col[8*2];                             \
+        a1 +=  W6*col[8*2];                             \
+        a2 += -W6*col[8*2];                             \
+        a3 += -W2*col[8*2];                             \
+                                                        \
+        b0 = MUL(W1, col[8*1]);                         \
+        b1 = MUL(W3, col[8*1]);                         \
+        b2 = MUL(W5, col[8*1]);                         \
+        b3 = MUL(W7, col[8*1]);                         \
+                                                        \
+        MAC(b0,  W3, col[8*3]);                         \
+        MAC(b1, -W7, col[8*3]);                         \
+        MAC(b2, -W1, col[8*3]);                         \
+        MAC(b3, -W5, col[8*3]);                         \
+                                                        \
+        if (col[8*4]) {                                 \
+            a0 +=  W4*col[8*4];                         \
+            a1 += -W4*col[8*4];                         \
+            a2 += -W4*col[8*4];                         \
+            a3 +=  W4*col[8*4];                         \
+        }                                               \
+                                                        \
+        if (col[8*5]) {                                 \
+            MAC(b0,  W5, col[8*5]);                     \
+            MAC(b1, -W1, col[8*5]);                     \
+            MAC(b2,  W7, col[8*5]);                     \
+            MAC(b3,  W3, col[8*5]);                     \
+        }                                               \
+                                                        \
+        if (col[8*6]) {                                 \
+            a0 +=  W6*col[8*6];                         \
+            a1 += -W2*col[8*6];                         \
+            a2 +=  W2*col[8*6];                         \
+            a3 += -W6*col[8*6];                         \
+        }                                               \
+                                                        \
+        if (col[8*7]) {                                 \
+            MAC(b0,  W7, col[8*7]);                     \
+            MAC(b1, -W5, col[8*7]);                     \
+            MAC(b2,  W3, col[8*7]);                     \
+            MAC(b3, -W1, col[8*7]);                     \
+        }                                               \
+    } while (0)
+
+static inline void FUNC(idctSparseColPut)(pixel *dest, int line_size,
+                                          DCTELEM *col)
+{
+    int a0, a1, a2, a3, b0, b1, b2, b3;
+    INIT_CLIP;
+
+    IDCT_COLS;
+
+    dest[0] = CLIP((a0 + b0) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a1 + b1) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a2 + b2) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a3 + b3) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a3 - b3) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a2 - b2) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a1 - b1) >> COL_SHIFT);
+    dest += line_size;
+    dest[0] = CLIP((a0 - b0) >> COL_SHIFT);
+}
+
+static inline void FUNC(idctSparseColAdd)(pixel *dest, int line_size,
+                                          DCTELEM *col)
+{
+    int a0, a1, a2, a3, b0, b1, b2, b3;
+    INIT_CLIP;
+
+    IDCT_COLS;
+
+    dest[0] = CLIP(dest[0] + ((a0 + b0) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a1 + b1) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a2 + b2) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a3 + b3) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a3 - b3) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a2 - b2) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a1 - b1) >> COL_SHIFT));
+    dest += line_size;
+    dest[0] = CLIP(dest[0] + ((a0 - b0) >> COL_SHIFT));
+}
+
+static inline void FUNC(idctSparseCol)(DCTELEM *col)
+{
+    int a0, a1, a2, a3, b0, b1, b2, b3;
+
+    IDCT_COLS;
+
+    col[0 ] = ((a0 + b0) >> COL_SHIFT);
+    col[8 ] = ((a1 + b1) >> COL_SHIFT);
+    col[16] = ((a2 + b2) >> COL_SHIFT);
+    col[24] = ((a3 + b3) >> COL_SHIFT);
+    col[32] = ((a3 - b3) >> COL_SHIFT);
+    col[40] = ((a2 - b2) >> COL_SHIFT);
+    col[48] = ((a1 - b1) >> COL_SHIFT);
+    col[56] = ((a0 - b0) >> COL_SHIFT);
+}
+
+void FUNC(ff_simple_idct_put)(uint8_t *dest_, int line_size, DCTELEM *block)
+{
+    pixel *dest = (pixel *)dest_;
+    int i;
+
+    line_size /= sizeof(pixel);
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctRowCondDC)(block + i*8);
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctSparseColPut)(dest + i, line_size, block + i);
+}
+
+void FUNC(ff_simple_idct_add)(uint8_t *dest_, int line_size, DCTELEM *block)
+{
+    pixel *dest = (pixel *)dest_;
+    int i;
+
+    line_size /= sizeof(pixel);
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctRowCondDC)(block + i*8);
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctSparseColAdd)(dest + i, line_size, block + i);
+}
+
+void FUNC(ff_simple_idct)(DCTELEM *block)
+{
+    int i;
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctRowCondDC)(block + i*8);
+
+    for (i = 0; i < 8; i++)
+        FUNC(idctSparseCol)(block + i);
+}

libavcodec/sparc/dsputil_vis.c

@@ -3956,7 +3956,8 @@ void dsputil_init_vis(DSPContext* c, AVCodecContext *avctx)
   const int high_bit_depth = avctx->codec_id == CODEC_ID_H264 && avctx->bits_per_raw_sample > 8;
 
   if (accel & ACCEL_SPARC_VIS) {
-      if(avctx->idct_algo==FF_IDCT_SIMPLEVIS){
+      if (avctx->bits_per_raw_sample <= 8 &&
+          avctx->idct_algo == FF_IDCT_SIMPLEVIS) {
           c->idct_put = ff_simple_idct_put_vis;
           c->idct_add = ff_simple_idct_add_vis;
           c->idct     = ff_simple_idct_vis;

libavcodec/vc1.c

@@ -337,11 +337,11 @@ int vc1_decode_sequence_header(AVCodecContext *avctx, VC1Context *v, GetBitConte
     v->res_fasttx = get_bits1(gb);
     if (!v->res_fasttx)
     {
-        v->vc1dsp.vc1_inv_trans_8x8 = ff_simple_idct;
+        v->vc1dsp.vc1_inv_trans_8x8 = ff_simple_idct_8;
         v->vc1dsp.vc1_inv_trans_8x4 = ff_simple_idct84_add;
         v->vc1dsp.vc1_inv_trans_4x8 = ff_simple_idct48_add;
         v->vc1dsp.vc1_inv_trans_4x4 = ff_simple_idct44_add;
-        v->vc1dsp.vc1_inv_trans_8x8_dc = ff_simple_idct_add;
+        v->vc1dsp.vc1_inv_trans_8x8_dc = ff_simple_idct_add_8;
         v->vc1dsp.vc1_inv_trans_8x4_dc = ff_simple_idct84_add;
         v->vc1dsp.vc1_inv_trans_4x8_dc = ff_simple_idct48_add;
         v->vc1dsp.vc1_inv_trans_4x4_dc = ff_simple_idct44_add;

libavcodec/x86/dsputil_mmx.c

@@ -2369,7 +2369,7 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
     if (mm_flags & AV_CPU_FLAG_MMX) {
         const int idct_algo= avctx->idct_algo;
 
-        if(avctx->lowres==0){
+        if (avctx->lowres == 0 && avctx->bits_per_raw_sample <= 8) {
             if(idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_SIMPLEMMX){
                 c->idct_put= ff_simple_idct_put_mmx;
                 c->idct_add= ff_simple_idct_add_mmx;

libswscale/utils.c

@@ -182,11 +182,13 @@ int sws_isSupportedInput(enum PixelFormat pix_fmt)
         || (x)==PIX_FMT_YUV420P9LE    \
         || (x)==PIX_FMT_YUV420P10LE   \
         || (x)==PIX_FMT_YUV420P16LE   \
+        || (x)==PIX_FMT_YUV422P10LE   \
         || (x)==PIX_FMT_YUV422P16LE   \
         || (x)==PIX_FMT_YUV444P16LE   \
         || (x)==PIX_FMT_YUV422P10BE   \
         || (x)==PIX_FMT_YUV420P9BE    \
         || (x)==PIX_FMT_YUV420P10BE   \
+        || (x)==PIX_FMT_YUV422P10BE   \
         || (x)==PIX_FMT_YUV420P16BE   \
         || (x)==PIX_FMT_YUV422P16BE   \
         || (x)==PIX_FMT_YUV444P16BE   \