Merge remote-tracking branch 'qatar/master'

* qatar/master:
  rtpdec: Use 4 byte startcodes for H.264
  matroskadec: Mark variable as av_unused.
  Move some conditionally used variables into the block where they are used.
  Drop some completely unnecessary av_unused attributes.
  swscale: Remove unused variable alpMmxFilter.
  Drop unnecessary av_uninit attributes from some variable declarations.
  movenc: Support muxing wmapro in ismv/isma
  mpegtsenc: Add an AVOption for forcing a new PAT/PMT/SDT to be written
  swscale: move YUV2PACKED16WRAPPER() macro down to where it is used.
  swscale: handle gray16 as a "planar" YUV format (Y-only, of course).
  swscale: use yuv2packed1() functions for unscaled chroma also.
  swscale: fix incorrect chroma bias in yuv2rgb48_1_c().
  swscale: fix invalid memory accesses in yuvpacked1() functions.
  Move PS2 MMI code below the mips subdirectory, where it belongs.
  mips: Move MMI function declarations to a header.
  build: Set correct dependencies for rtmp* protocols implemented by librtmp.

Conflicts:
	libavcodec/ac3enc_template.c
	libavformat/mpegtsenc.c
	libswscale/output.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

27 files changed, 132 insertions, 192 deletions

diff --git a/configure b/configure
index bb11b692d2..052d354f9a 100755
--- a/configure
+++ b/configure
@@ -1643,6 +1643,10 @@ https_protocol_select="tls_protocol"
 mmsh_protocol_select="http_protocol"
 mmst_protocol_deps="network"
 rtmp_protocol_select="tcp_protocol"
+rtmpe_protocol_deps="librtmp"
+rtmps_protocol_deps="librtmp"
+rtmpt_protocol_deps="librtmp"
+rtmpte_protocol_deps="librtmp"
 rtp_protocol_select="udp_protocol"
 tcp_protocol_deps="network"
 tls_protocol_deps_any="openssl gnutls"
diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index 7ff6975da1..d14dac6010 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -747,7 +747,7 @@ HOSTPROGS = aac_tablegen aacps_tablegen cbrt_tablegen cos_tablegen      \
             dv_tablegen motionpixels_tablegen mpegaudio_tablegen        \
             pcm_tablegen qdm2_tablegen sinewin_tablegen
 
-DIRS = alpha arm avr32 bfin ppc ps2 sh4 sparc x86
+DIRS = alpha arm avr32 bfin mips ppc sh4 sparc x86
 
 CLEANFILES = *_tables.c *_tables.h *_tablegen$(HOSTEXESUF)
 
diff --git a/libavcodec/ac3enc.c b/libavcodec/ac3enc.c
index 2f1e7c8f67..a5d335091d 100644
--- a/libavcodec/ac3enc.c
+++ b/libavcodec/ac3enc.c
@@ -1383,8 +1383,7 @@ static void ac3_output_frame_header(AC3EncodeContext *s)
  */
 static void output_audio_block(AC3EncodeContext *s, int blk)
 {
-    int ch, i, baie, bnd, got_cpl;
-    int av_uninit(ch0);
+    int ch, i, baie, bnd, got_cpl, ch0;
     AC3Block *block = &s->blocks[blk];
 
     /* block switching */
@@ -2233,8 +2232,7 @@ static av_cold int validate_options(AC3EncodeContext *s)
  */
 static av_cold void set_bandwidth(AC3EncodeContext *s)
 {
-    int blk, ch;
-    int av_uninit(cpl_start);
+    int blk, ch, cpl_start;
 
     if (s->cutoff) {
         /* calculate bandwidth based on user-specified cutoff frequency */
diff --git a/libavcodec/ac3enc_template.c b/libavcodec/ac3enc_template.c
index 60472550d1..554e0b007b 100644
--- a/libavcodec/ac3enc_template.c
+++ b/libavcodec/ac3enc_template.c
@@ -338,7 +338,7 @@ static void compute_rematrixing_strategy(AC3EncodeContext *s)
 {
     int nb_coefs;
     int blk, bnd;
-    AC3Block *block, *av_uninit(block0);
+    AC3Block *block, *block0;
 
     if (s->channel_mode != AC3_CHMODE_STEREO)
         return;
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index b496d72948..3ea7a24534 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -1784,7 +1784,7 @@ typedef struct AVCodecContext {
 #define FF_IDCT_SIMPLE        2
 #define FF_IDCT_SIMPLEMMX     3
 #define FF_IDCT_LIBMPEG2MMX   4
-#define FF_IDCT_PS2           5
+#define FF_IDCT_MMI           5
 #define FF_IDCT_ARM           7
 #define FF_IDCT_ALTIVEC       8
 #define FF_IDCT_SH4           9
diff --git a/libavcodec/eatgv.c b/libavcodec/eatgv.c
index 37f84bf5ef..25f1b7b847 100644
--- a/libavcodec/eatgv.c
+++ b/libavcodec/eatgv.c
@@ -66,7 +66,7 @@ static av_cold int tgv_decode_init(AVCodecContext *avctx){
  */
 static int unpack(const uint8_t *src, const uint8_t *src_end, unsigned char *dst, int width, int height) {
     unsigned char *dst_end = dst + width*height;
-    int size, size1, size2, av_uninit(offset), run;
+    int size, size1, size2, offset, run;
     unsigned char *dst_start = dst;
 
     if (src[0] & 0x01)
diff --git a/libavcodec/flacdec.c b/libavcodec/flacdec.c
index ca0863107a..055431b6d2 100644
--- a/libavcodec/flacdec.c
+++ b/libavcodec/flacdec.c
@@ -288,7 +288,7 @@ static int decode_subframe_fixed(FLACContext *s, int channel, int pred_order)
 {
     const int blocksize = s->blocksize;
     int32_t *decoded = s->decoded[channel];
-    int av_uninit(a), av_uninit(b), av_uninit(c), av_uninit(d), i;
+    int a, b, c, d, i;
 
     /* warm up samples */
     for (i = 0; i < pred_order; i++) {
diff --git a/libavcodec/mips/Makefile b/libavcodec/mips/Makefile
index 3f4da68ec9..006617687e 100644
--- a/libavcodec/mips/Makefile
+++ b/libavcodec/mips/Makefile
@@ -1,3 +1,3 @@
-OBJS-$(HAVE_MMI)                       += ps2/dsputil_mmi.o             \
-                                          ps2/idct_mmi.o                \
-                                          ps2/mpegvideo_mmi.o           \
+OBJS-$(HAVE_MMI) += mips/dsputil_mmi.o                                  \
+                    mips/idct_mmi.o                                     \
+                    mips/mpegvideo_mmi.o                                \
diff --git a/libavcodec/ps2/dsputil_mmi.c b/libavcodec/mips/dsputil_mmi.c
index 26259c8453..71a147fba4 100644
--- a/libavcodec/ps2/dsputil_mmi.c
+++ b/libavcodec/mips/dsputil_mmi.c
@@ -25,10 +25,6 @@
 #include "libavcodec/dsputil.h"
 #include "mmi.h"
 
-void ff_mmi_idct_put(uint8_t *dest, int line_size, DCTELEM *block);
-void ff_mmi_idct_add(uint8_t *dest, int line_size, DCTELEM *block);
-void ff_mmi_idct(DCTELEM *block);
-
 static void clear_blocks_mmi(DCTELEM * blocks)
 {
         __asm__ volatile(
@@ -157,7 +153,7 @@ void dsputil_init_mmi(DSPContext* c, AVCodecContext *avctx)
     }
 
     if (avctx->bits_per_raw_sample <= 8 &&
-        (idct_algo==FF_IDCT_AUTO || idct_algo==FF_IDCT_PS2)) {
+        (idct_algo == FF_IDCT_AUTO || idct_algo == FF_IDCT_MMI)) {
         c->idct_put= ff_mmi_idct_put;
         c->idct_add= ff_mmi_idct_add;
         c->idct    = ff_mmi_idct;
diff --git a/libavcodec/ps2/idct_mmi.c b/libavcodec/mips/idct_mmi.c
index 5a3af173de..5a3af173de 100644
--- a/libavcodec/ps2/idct_mmi.c
+++ b/libavcodec/mips/idct_mmi.c
diff --git a/libavcodec/ps2/mmi.h b/libavcodec/mips/mmi.h
index 0265456108..5ff0033a87 100644
--- a/libavcodec/ps2/mmi.h
+++ b/libavcodec/mips/mmi.h
@@ -18,8 +18,16 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
-#ifndef AVCODEC_PS2_MMI_H
-#define AVCODEC_PS2_MMI_H
+#ifndef AVCODEC_MIPS_MMI_H
+#define AVCODEC_MIPS_MMI_H
+
+#include <stdint.h>
+
+#include "libavcodec/dsputil.h"
+
+void ff_mmi_idct_put(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_mmi_idct_add(uint8_t *dest, int line_size, DCTELEM *block);
+void ff_mmi_idct(DCTELEM *block);
 
 #define align16 __attribute__ ((aligned (16)))
 
@@ -168,4 +176,4 @@
 #define         pextlb(rs, rt, rd) \
         __asm__ volatile ("pextlb  " #rd ", " #rs ", " #rt )
 
-#endif /* AVCODEC_PS2_MMI_H */
+#endif /* AVCODEC_MIPS_MMI_H */
diff --git a/libavcodec/ps2/mpegvideo_mmi.c b/libavcodec/mips/mpegvideo_mmi.c
index 8c8d3a7932..8c8d3a7932 100644
--- a/libavcodec/ps2/mpegvideo_mmi.c
+++ b/libavcodec/mips/mpegvideo_mmi.c
diff --git a/libavcodec/msrledec.c b/libavcodec/msrledec.c
index c0e7682689..9ffc3439f9 100644
--- a/libavcodec/msrledec.c
+++ b/libavcodec/msrledec.c
@@ -134,8 +134,8 @@ static int msrle_decode_8_16_24_32(AVCodecContext *avctx, AVPicture *pic, int de
     uint8_t *output, *output_end;
     const uint8_t* src = data;
     int p1, p2, line=avctx->height - 1, pos=0, i;
-    uint16_t av_uninit(pix16);
-    uint32_t av_uninit(pix32);
+    uint16_t pix16;
+    uint32_t pix32;
     unsigned int width= FFABS(pic->linesize[0]) / (depth >> 3);
 
     output     = pic->data[0] + (avctx->height - 1) * pic->linesize[0];
diff --git a/libavcodec/options.c b/libavcodec/options.c
index b11d6abe6b..3be5861dde 100644
--- a/libavcodec/options.c
+++ b/libavcodec/options.c
@@ -219,7 +219,7 @@ static const AVOption options[]={
 {"simple", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_SIMPLE }, INT_MIN, INT_MAX, V|E|D, "idct"},
 {"simplemmx", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_SIMPLEMMX }, INT_MIN, INT_MAX, V|E|D, "idct"},
 {"libmpeg2mmx", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_LIBMPEG2MMX }, INT_MIN, INT_MAX, V|E|D, "idct"},
-{"ps2", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_PS2 }, INT_MIN, INT_MAX, V|E|D, "idct"},
+{"mmi", NULL, 0, AV_OPT_TYPE_CONST, { .dbl = FF_IDCT_MMI }, INT_MIN, INT_MAX, V|E|D, "idct"},
 {"arm", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_ARM }, INT_MIN, INT_MAX, V|E|D, "idct"},
 {"altivec", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_ALTIVEC }, INT_MIN, INT_MAX, V|E|D, "idct"},
 {"sh4", NULL, 0, AV_OPT_TYPE_CONST, {.dbl = FF_IDCT_SH4 }, INT_MIN, INT_MAX, V|E|D, "idct"},
diff --git a/libavcodec/qtrleenc.c b/libavcodec/qtrleenc.c
index a8963ab78a..9aa2edb4e0 100644
--- a/libavcodec/qtrleenc.c
+++ b/libavcodec/qtrleenc.c
@@ -126,7 +126,7 @@ static void qtrle_encode_line(QtrleEncContext *s, AVFrame *p, int line, uint8_t
     unsigned int skipcount;
     /* This will be the number of consecutive equal pixels in the current
      * frame, starting from the ith one also */
-    unsigned int av_uninit(repeatcount);
+    unsigned int repeatcount;
 
     /* The cost of the three different possibilities */
     int total_bulk_cost;
diff --git a/libavfilter/vf_gradfun.c b/libavfilter/vf_gradfun.c
index 8f8b028e1d..49e525507f 100644
--- a/libavfilter/vf_gradfun.c
+++ b/libavfilter/vf_gradfun.c
@@ -120,7 +120,7 @@ static av_cold int init(AVFilterContext *ctx, const char *args, void *opaque)
     GradFunContext *gf = ctx->priv;
     float thresh = 1.2;
     int radius = 16;
-    av_unused int cpu_flags = av_get_cpu_flags();
+    int cpu_flags = av_get_cpu_flags();
 
     if (args)
         sscanf(args, "%f:%d", &thresh, &radius);
diff --git a/libavfilter/vf_yadif.c b/libavfilter/vf_yadif.c
index 3a9637380b..c9f870539d 100644
--- a/libavfilter/vf_yadif.c
+++ b/libavfilter/vf_yadif.c
@@ -370,7 +370,7 @@ static int query_formats(AVFilterContext *ctx)
 static av_cold int init(AVFilterContext *ctx, const char *args, void *opaque)
 {
     YADIFContext *yadif = ctx->priv;
-    av_unused int cpu_flags = av_get_cpu_flags();
+    int cpu_flags = av_get_cpu_flags();
 
     yadif->mode = 0;
     yadif->parity = -1;
diff --git a/libavformat/allformats.c b/libavformat/allformats.c
index 134839ffa1..8b1c525217 100644
--- a/libavformat/allformats.c
+++ b/libavformat/allformats.c
@@ -271,12 +271,10 @@ void av_register_all(void)
     REGISTER_PROTOCOL (MD5,  md5);
     REGISTER_PROTOCOL (PIPE, pipe);
     REGISTER_PROTOCOL (RTMP, rtmp);
-#if CONFIG_LIBRTMP
-    REGISTER_PROTOCOL (RTMP, rtmpt);
-    REGISTER_PROTOCOL (RTMP, rtmpe);
-    REGISTER_PROTOCOL (RTMP, rtmpte);
-    REGISTER_PROTOCOL (RTMP, rtmps);
-#endif
+    REGISTER_PROTOCOL (RTMPE, rtmpe);
+    REGISTER_PROTOCOL (RTMPS, rtmps);
+    REGISTER_PROTOCOL (RTMPT, rtmpt);
+    REGISTER_PROTOCOL (RTMPTE, rtmpte);
     REGISTER_PROTOCOL (RTP, rtp);
     REGISTER_PROTOCOL (TCP, tcp);
     REGISTER_PROTOCOL (TLS, tls);
diff --git a/libavformat/asfdec.c b/libavformat/asfdec.c
index 2c2bd47802..e2f46dc005 100644
--- a/libavformat/asfdec.c
+++ b/libavformat/asfdec.c
@@ -361,8 +361,8 @@ static int asf_read_stream_properties(AVFormatContext *s, int64_t size)
         /* This code assumes that extradata contains only palette */
         /* This is true for all paletted codecs implemented in libavcodec */
         if (st->codec->extradata_size && (st->codec->bits_per_coded_sample <= 8)) {
-            int av_unused i;
 #if HAVE_BIGENDIAN
+            int i;
             for (i = 0; i < FFMIN(st->codec->extradata_size, AVPALETTE_SIZE)/4; i++)
                 asf_st->palette[i] = av_bswap32(((uint32_t*)st->codec->extradata)[i]);
 #else
diff --git a/libavformat/matroskadec.c b/libavformat/matroskadec.c
index bb7a14e072..ddf60e48b2 100644
--- a/libavformat/matroskadec.c
+++ b/libavformat/matroskadec.c
@@ -967,7 +967,7 @@ static int matroska_decode_buffer(uint8_t** buf, int* buf_size,
     uint8_t* data = *buf;
     int isize = *buf_size;
     uint8_t* pkt_data = NULL;
-    uint8_t* newpktdata;
+    uint8_t av_unused *newpktdata;
     int pkt_size = isize;
     int result = 0;
     int olen;
diff --git a/libavformat/movenc.c b/libavformat/movenc.c
index dcc0d76e33..fe5a55c66f 100644
--- a/libavformat/movenc.c
+++ b/libavformat/movenc.c
@@ -362,6 +362,15 @@ static int mov_write_ms_tag(AVIOContext *pb, MOVTrack *track)
     return update_size(pb, pos);
 }
 
+static int mov_write_wfex_tag(AVIOContext *pb, MOVTrack *track)
+{
+    int64_t pos = avio_tell(pb);
+    avio_wb32(pb, 0);
+    ffio_wfourcc(pb, "wfex");
+    ff_put_wav_header(pb, track->enc);
+    return update_size(pb, pos);
+}
+
 static int mov_write_chan_tag(AVIOContext *pb, MOVTrack *track)
 {
     uint32_t layout_tag, bitmap;
@@ -548,6 +557,8 @@ static int mov_write_audio_tag(AVIOContext *pb, MOVTrack *track)
         mov_write_ac3_tag(pb, track);
     else if(track->enc->codec_id == CODEC_ID_ALAC)
         mov_write_extradata_tag(pb, track);
+    else if (track->enc->codec_id == CODEC_ID_WMAPRO)
+        mov_write_wfex_tag(pb, track);
     else if (track->vos_len > 0)
         mov_write_glbl_tag(pb, track);
 
@@ -797,9 +808,13 @@ static int mov_find_codec_tag(AVFormatContext *s, MOVTrack *track)
 {
     int tag = track->enc->codec_tag;
 
-    if (track->mode == MODE_MP4 || track->mode == MODE_PSP || track->mode == MODE_ISM)
+    if (track->mode == MODE_MP4 || track->mode == MODE_PSP)
+        tag = mp4_get_codec_tag(s, track);
+    else if (track->mode == MODE_ISM) {
         tag = mp4_get_codec_tag(s, track);
-    else if (track->mode == MODE_IPOD)
+        if (!tag && track->enc->codec_id == CODEC_ID_WMAPRO)
+            tag = MKTAG('w', 'm', 'a', ' ');
+    } else if (track->mode == MODE_IPOD)
         tag = ipod_get_codec_tag(s, track);
     else if (track->mode & MODE_3GP)
         tag = ff_codec_get_tag(codec_3gp_tags, track->enc->codec_id);
diff --git a/libavformat/mpegtsenc.c b/libavformat/mpegtsenc.c
index 4b303def4b..b0047a216e 100644
--- a/libavformat/mpegtsenc.c
+++ b/libavformat/mpegtsenc.c
@@ -78,6 +78,8 @@ typedef struct MpegTSWrite {
     int pmt_start_pid;
     int start_pid;
     int m2ts_mode;
+
+    int reemit_pat_pmt;
 } MpegTSWrite;
 
 /* a PES packet header is generated every DEFAULT_PES_HEADER_FREQ packets */
@@ -101,6 +103,8 @@ static const AVOption options[] = {
     { "muxrate", NULL, offsetof(MpegTSWrite, mux_rate), AV_OPT_TYPE_INT, {1}, 0, INT_MAX, AV_OPT_FLAG_ENCODING_PARAM},
     { "pes_payload_size", "Minimum PES packet payload in bytes",
       offsetof(MpegTSWrite, pes_payload_size), AV_OPT_TYPE_INT, {DEFAULT_PES_PAYLOAD_SIZE}, 0, INT_MAX, AV_OPT_FLAG_ENCODING_PARAM},
+    { "resend_headers", "Reemit PAT/PMT before writing the next packet",
+      offsetof(MpegTSWrite, reemit_pat_pmt), AV_OPT_TYPE_INT, {0}, 0, INT_MAX, AV_OPT_FLAG_ENCODING_PARAM},
     { NULL },
 };
 
@@ -978,6 +982,12 @@ static int mpegts_write_packet(AVFormatContext *s, AVPacket *pkt)
     const uint64_t delay = av_rescale(s->max_delay, 90000, AV_TIME_BASE)*2;
     int64_t dts = AV_NOPTS_VALUE, pts = AV_NOPTS_VALUE;
 
+    if (ts->reemit_pat_pmt) {
+        ts->pat_packet_count = ts->pat_packet_period - 1;
+        ts->sdt_packet_count = ts->sdt_packet_period - 1;
+        ts->reemit_pat_pmt = 0;
+    }
+
     if (pkt->pts != AV_NOPTS_VALUE)
         pts = pkt->pts + delay;
     if (pkt->dts != AV_NOPTS_VALUE)
diff --git a/libavformat/nsvdec.c b/libavformat/nsvdec.c
index d581cfb50e..87abce9b21 100644
--- a/libavformat/nsvdec.c
+++ b/libavformat/nsvdec.c
@@ -272,7 +272,7 @@ static int nsv_parse_NSVf_header(AVFormatContext *s)
 {
     NSVContext *nsv = s->priv_data;
     AVIOContext *pb = s->pb;
-    unsigned int av_unused file_size;
+    unsigned int file_size;
     unsigned int size;
     int64_t duration;
     int strings_size;
@@ -565,7 +565,6 @@ static int nsv_read_chunk(AVFormatContext *s, int fill_header)
     uint32_t vsize;
     uint16_t asize;
     uint16_t auxsize;
-    uint32_t av_unused auxtag;
 
     av_dlog(s, "%s(%d)\n", __FUNCTION__, fill_header);
 
@@ -595,6 +594,7 @@ null_chunk_retry:
     av_dlog(s, "NSV CHUNK %d aux, %u bytes video, %d bytes audio\n", auxcount, vsize, asize);
     /* skip aux stuff */
     for (i = 0; i < auxcount; i++) {
+        uint32_t auxtag;
         auxsize = avio_rl16(pb);
         auxtag = avio_rl32(pb);
         av_dlog(s, "NSV aux data: '%c%c%c%c', %d bytes\n",
diff --git a/libavformat/rtpdec_h264.c b/libavformat/rtpdec_h264.c
index 730ed14807..01514c1bef 100644
--- a/libavformat/rtpdec_h264.c
+++ b/libavformat/rtpdec_h264.c
@@ -113,7 +113,7 @@ static int sdp_parse_fmtp_config_h264(AVStream * stream,
             h264_data->level_idc = level_idc;
         }
     } else  if (!strcmp(attr, "sprop-parameter-sets")) {
-        uint8_t start_sequence[]= { 0, 0, 1 };
+        uint8_t start_sequence[] = { 0, 0, 0, 1 };
         codec->extradata_size= 0;
         codec->extradata= NULL;
 
@@ -176,7 +176,7 @@ static int h264_handle_packet(AVFormatContext *ctx,
     uint8_t nal = buf[0];
     uint8_t type = (nal & 0x1f);
     int result= 0;
-    uint8_t start_sequence[]= {0, 0, 1};
+    uint8_t start_sequence[] = { 0, 0, 0, 1 };
 
 #ifdef DEBUG
     assert(data);
diff --git a/libavutil/sha.c b/libavutil/sha.c
index 301d1606b4..7194d43204 100644
--- a/libavutil/sha.c
+++ b/libavutil/sha.c
@@ -181,7 +181,7 @@ static void sha256_transform(uint32_t *state, const uint8_t buffer[64])
 {
     unsigned int i, a, b, c, d, e, f, g, h;
     uint32_t block[64];
-    uint32_t T1, av_unused(T2);
+    uint32_t T1;
 
     a = state[0];
     b = state[1];
@@ -193,6 +193,7 @@ static void sha256_transform(uint32_t *state, const uint8_t buffer[64])
     h = state[7];
 #if CONFIG_SMALL
     for (i = 0; i < 64; i++) {
+        uint32_t T2;
         if (i < 16)
             T1 = blk0(i);
         else
diff --git a/libswscale/output.c b/libswscale/output.c
index d0bd72b350..180927892d 100644
--- a/libswscale/output.c
+++ b/libswscale/output.c
@@ -298,129 +298,6 @@ static void yuv2nv12cX_c(SwsContext *c, const int16_t *chrFilter, int chrFilterS
         }
 }
 
-#define output_pixel(pos, val) \
-        if (target == PIX_FMT_GRAY16BE) { \
-            AV_WB16(pos, val); \
-        } else { \
-            AV_WL16(pos, val); \
-        }
-
-static av_always_inline void
-yuv2gray16_X_c_template(SwsContext *c, const int16_t *lumFilter,
-                        const int32_t **lumSrc, int lumFilterSize,
-                        const int16_t *chrFilter, const int32_t **chrUSrc,
-                        const int32_t **chrVSrc, int chrFilterSize,
-                        const int32_t **alpSrc, uint16_t *dest, int dstW,
-                        int y, enum PixelFormat target)
-{
-    int i;
-
-    for (i = 0; i < (dstW >> 1); i++) {
-        int j;
-        int Y1 = (1 << 14) - 0x40000000;
-        int Y2 = (1 << 14) - 0x40000000;
-
-        for (j = 0; j < lumFilterSize; j++) {
-            Y1 += lumSrc[j][i * 2]     * lumFilter[j];
-            Y2 += lumSrc[j][i * 2 + 1] * lumFilter[j];
-        }
-        Y1 >>= 15;
-        Y2 >>= 15;
-        Y1 = av_clip_int16(Y1);
-        Y2 = av_clip_int16(Y2);
-        output_pixel(&dest[i * 2 + 0], 0x8000 + Y1);
-        output_pixel(&dest[i * 2 + 1], 0x8000 + Y2);
-    }
-}
-
-static av_always_inline void
-yuv2gray16_2_c_template(SwsContext *c, const int32_t *buf[2],
-                        const int32_t *ubuf[2], const int32_t *vbuf[2],
-                        const int32_t *abuf[2], uint16_t *dest, int dstW,
-                        int yalpha, int uvalpha, int y,
-                        enum PixelFormat target)
-{
-    int  yalpha1 = 4095 - yalpha;
-    int i;
-    const int32_t *buf0 = buf[0], *buf1 = buf[1];
-
-    for (i = 0; i < (dstW >> 1); i++) {
-        int Y1 = (buf0[i * 2    ] * yalpha1 + buf1[i * 2    ] * yalpha) >> 15;
-        int Y2 = (buf0[i * 2 + 1] * yalpha1 + buf1[i * 2 + 1] * yalpha) >> 15;
-
-        output_pixel(&dest[i * 2 + 0], Y1);
-        output_pixel(&dest[i * 2 + 1], Y2);
-    }
-}
-
-static av_always_inline void
-yuv2gray16_1_c_template(SwsContext *c, const int32_t *buf0,
-                        const int32_t *ubuf[2], const int32_t *vbuf[2],
-                        const int32_t *abuf0, uint16_t *dest, int dstW,
-                        int uvalpha, int y, enum PixelFormat target)
-{
-    int i;
-
-    for (i = 0; i < (dstW >> 1); i++) {
-        int Y1 = (buf0[i * 2    ]+4)>>3;
-        int Y2 = (buf0[i * 2 + 1]+4)>>3;
-
-        output_pixel(&dest[i * 2 + 0], Y1);
-        output_pixel(&dest[i * 2 + 1], Y2);
-    }
-}
-
-#undef output_pixel
-
-#define YUV2PACKED16WRAPPER(name, base, ext, fmt) \
-static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
-                        const int16_t **_lumSrc, int lumFilterSize, \
-                        const int16_t *chrFilter, const int16_t **_chrUSrc, \
-                        const int16_t **_chrVSrc, int chrFilterSize, \
-                        const int16_t **_alpSrc, uint8_t *_dest, int dstW, \
-                        int y) \
-{ \
-    const int32_t **lumSrc  = (const int32_t **) _lumSrc, \
-                  **chrUSrc = (const int32_t **) _chrUSrc, \
-                  **chrVSrc = (const int32_t **) _chrVSrc, \
-                  **alpSrc  = (const int32_t **) _alpSrc; \
-    uint16_t *dest = (uint16_t *) _dest; \
-    name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
-                          chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
-                          alpSrc, dest, dstW, y, fmt); \
-} \
- \
-static void name ## ext ## _2_c(SwsContext *c, const int16_t *_buf[2], \
-                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
-                        const int16_t *_abuf[2], uint8_t *_dest, int dstW, \
-                        int yalpha, int uvalpha, int y) \
-{ \
-    const int32_t **buf  = (const int32_t **) _buf, \
-                  **ubuf = (const int32_t **) _ubuf, \
-                  **vbuf = (const int32_t **) _vbuf, \
-                  **abuf = (const int32_t **) _abuf; \
-    uint16_t *dest = (uint16_t *) _dest; \
-    name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
-                          dest, dstW, yalpha, uvalpha, y, fmt); \
-} \
- \
-static void name ## ext ## _1_c(SwsContext *c, const int16_t *_buf0, \
-                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
-                        const int16_t *_abuf0, uint8_t *_dest, int dstW, \
-                        int uvalpha, int y) \
-{ \
-    const int32_t *buf0  = (const int32_t *)  _buf0, \
-                 **ubuf  = (const int32_t **) _ubuf, \
-                 **vbuf  = (const int32_t **) _vbuf, \
-                  *abuf0 = (const int32_t *)  _abuf0; \
-    uint16_t *dest = (uint16_t *) _dest; \
-    name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
-                                  dstW, uvalpha, y, fmt); \
-}
-
-YUV2PACKED16WRAPPER(yuv2gray16,, LE, PIX_FMT_GRAY16LE)
-YUV2PACKED16WRAPPER(yuv2gray16,, BE, PIX_FMT_GRAY16BE)
-
 #define output_pixel(pos, acc) \
     if (target == PIX_FMT_MONOBLACK) { \
         pos = acc; \
@@ -632,20 +509,20 @@ yuv2422_1_c_template(SwsContext *c, const int16_t *buf0,
                      const int16_t *abuf0, uint8_t *dest, int dstW,
                      int uvalpha, int y, enum PixelFormat target)
 {
-    const int16_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
-                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
+    const int16_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
     int i;
 
     if (uvalpha < 2048) {
         for (i = 0; i < (dstW >> 1); i++) {
             int Y1 = buf0[i * 2]     >> 7;
             int Y2 = buf0[i * 2 + 1] >> 7;
-            int U  = ubuf1[i]        >> 7;
-            int V  = vbuf1[i]        >> 7;
+            int U  = ubuf0[i]        >> 7;
+            int V  = vbuf0[i]        >> 7;
 
             output_pixels(i * 4, Y1, U, Y2, V);
         }
     } else {
+        const int16_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
         for (i = 0; i < (dstW >> 1); i++) {
             int Y1 =  buf0[i * 2]          >> 7;
             int Y2 =  buf0[i * 2 + 1]      >> 7;
@@ -778,8 +655,7 @@ yuv2rgb48_1_c_template(SwsContext *c, const int32_t *buf0,
                        const int32_t *abuf0, uint16_t *dest, int dstW,
                        int uvalpha, int y, enum PixelFormat target)
 {
-    const int32_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
-                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
+    const int32_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
     int i;
 
     if (uvalpha < 2048) {
@@ -810,6 +686,7 @@ yuv2rgb48_1_c_template(SwsContext *c, const int32_t *buf0,
             dest += 6;
         }
     } else {
+        const int32_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
         for (i = 0; i < (dstW >> 1); i++) {
             int Y1 = (buf0[i * 2]    ) >> 2;
             int Y2 = (buf0[i * 2 + 1]) >> 2;
@@ -843,6 +720,52 @@ yuv2rgb48_1_c_template(SwsContext *c, const int32_t *buf0,
 #undef r_b
 #undef b_r
 
+#define YUV2PACKED16WRAPPER(name, base, ext, fmt) \
+static void name ## ext ## _X_c(SwsContext *c, const int16_t *lumFilter, \
+                        const int16_t **_lumSrc, int lumFilterSize, \
+                        const int16_t *chrFilter, const int16_t **_chrUSrc, \
+                        const int16_t **_chrVSrc, int chrFilterSize, \
+                        const int16_t **_alpSrc, uint8_t *_dest, int dstW, \
+                        int y) \
+{ \
+    const int32_t **lumSrc  = (const int32_t **) _lumSrc, \
+                  **chrUSrc = (const int32_t **) _chrUSrc, \
+                  **chrVSrc = (const int32_t **) _chrVSrc, \
+                  **alpSrc  = (const int32_t **) _alpSrc; \
+    uint16_t *dest = (uint16_t *) _dest; \
+    name ## base ## _X_c_template(c, lumFilter, lumSrc, lumFilterSize, \
+                          chrFilter, chrUSrc, chrVSrc, chrFilterSize, \
+                          alpSrc, dest, dstW, y, fmt); \
+} \
+ \
+static void name ## ext ## _2_c(SwsContext *c, const int16_t *_buf[2], \
+                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
+                        const int16_t *_abuf[2], uint8_t *_dest, int dstW, \
+                        int yalpha, int uvalpha, int y) \
+{ \
+    const int32_t **buf  = (const int32_t **) _buf, \
+                  **ubuf = (const int32_t **) _ubuf, \
+                  **vbuf = (const int32_t **) _vbuf, \
+                  **abuf = (const int32_t **) _abuf; \
+    uint16_t *dest = (uint16_t *) _dest; \
+    name ## base ## _2_c_template(c, buf, ubuf, vbuf, abuf, \
+                          dest, dstW, yalpha, uvalpha, y, fmt); \
+} \
+ \
+static void name ## ext ## _1_c(SwsContext *c, const int16_t *_buf0, \
+                        const int16_t *_ubuf[2], const int16_t *_vbuf[2], \
+                        const int16_t *_abuf0, uint8_t *_dest, int dstW, \
+                        int uvalpha, int y) \
+{ \
+    const int32_t *buf0  = (const int32_t *)  _buf0, \
+                 **ubuf  = (const int32_t **) _ubuf, \
+                 **vbuf  = (const int32_t **) _vbuf, \
+                  *abuf0 = (const int32_t *)  _abuf0; \
+    uint16_t *dest = (uint16_t *) _dest; \
+    name ## base ## _1_c_template(c, buf0, ubuf, vbuf, abuf0, dest, \
+                                  dstW, uvalpha, y, fmt); \
+}
+
 YUV2PACKED16WRAPPER(yuv2, rgb48, rgb48be, PIX_FMT_RGB48BE)
 YUV2PACKED16WRAPPER(yuv2, rgb48, rgb48le, PIX_FMT_RGB48LE)
 YUV2PACKED16WRAPPER(yuv2, rgb48, bgr48be, PIX_FMT_BGR48BE)
@@ -980,12 +903,11 @@ yuv2rgb_X_c_template(SwsContext *c, const int16_t *lumFilter,
     int i;
 
     for (i = 0; i < (dstW >> 1); i++) {
-        int j;
+        int j, A1, A2;
         int Y1 = 1 << 18;
         int Y2 = 1 << 18;
         int U  = 1 << 18;
         int V  = 1 << 18;
-        int av_unused A1, A2;
         const void *r, *g, *b;
 
         for (j = 0; j < lumFilterSize; j++) {
@@ -1067,16 +989,15 @@ yuv2rgb_1_c_template(SwsContext *c, const int16_t *buf0,
                      int uvalpha, int y, enum PixelFormat target,
                      int hasAlpha)
 {
-    const int16_t *ubuf0 = ubuf[0], *ubuf1 = ubuf[1],
-                  *vbuf0 = vbuf[0], *vbuf1 = vbuf[1];
+    const int16_t *ubuf0 = ubuf[0], *vbuf0 = vbuf[0];
     int i;
 
     if (uvalpha < 2048) {
         for (i = 0; i < (dstW >> 1); i++) {
             int Y1 = buf0[i * 2]     >> 7;
             int Y2 = buf0[i * 2 + 1] >> 7;
-            int U  = ubuf1[i]        >> 7;
-            int V  = vbuf1[i]        >> 7;
+            int U  = ubuf0[i]        >> 7;
+            int V  = vbuf0[i]        >> 7;
             int A1, A2;
             const void *r =  c->table_rV[V + YUVRGB_TABLE_HEADROOM],
                        *g = (c->table_gU[U + YUVRGB_TABLE_HEADROOM] + c->table_gV[V + YUVRGB_TABLE_HEADROOM]),
@@ -1091,6 +1012,7 @@ yuv2rgb_1_c_template(SwsContext *c, const int16_t *buf0,
                           r, g, b, y, target, hasAlpha);
         }
     } else {
+        const int16_t *ubuf1 = ubuf[1], *vbuf1 = vbuf[1];
         for (i = 0; i < (dstW >> 1); i++) {
             int Y1 =  buf0[i * 2]          >> 7;
             int Y2 =  buf0[i * 2 + 1]      >> 7;
@@ -1180,8 +1102,7 @@ yuv2rgb_full_X_c_template(SwsContext *c, const int16_t *lumFilter,
         int Y = 1<<9;
         int U = (1<<9)-(128 << 19);
         int V = (1<<9)-(128 << 19);
-        int av_unused A;
-        int R, G, B;
+        int R, G, B, A;
 
         for (j = 0; j < lumFilterSize; j++) {
             Y += lumSrc[j][i] * lumFilter[j];
@@ -1489,16 +1410,6 @@ void ff_sws_init_output_funcs(SwsContext *c,
         }
     }
     switch (dstFormat) {
-    case PIX_FMT_GRAY16BE:
-        *yuv2packed1 = yuv2gray16BE_1_c;
-        *yuv2packed2 = yuv2gray16BE_2_c;
-        *yuv2packedX = yuv2gray16BE_X_c;
-        break;
-    case PIX_FMT_GRAY16LE:
-        *yuv2packed1 = yuv2gray16LE_1_c;
-        *yuv2packed2 = yuv2gray16LE_2_c;
-        *yuv2packedX = yuv2gray16LE_X_c;
-        break;
     case PIX_FMT_MONOWHITE:
         *yuv2packed1 = yuv2monowhite_1_c;
         *yuv2packed2 = yuv2monowhite_2_c;
diff --git a/libswscale/swscale.c b/libswscale/swscale.c
index 12ec19bea0..0578e2dacc 100644
--- a/libswscale/swscale.c
+++ b/libswscale/swscale.c
@@ -334,7 +334,6 @@ static int swScale(SwsContext *c, const uint8_t* src[],
     int16_t *hChrFilter= c->hChrFilter;
     int32_t *lumMmxFilter= c->lumMmxFilter;
     int32_t *chrMmxFilter= c->chrMmxFilter;
-    int32_t av_unused *alpMmxFilter= c->alpMmxFilter;
     const int vLumFilterSize= c->vLumFilterSize;
     const int vChrFilterSize= c->vChrFilterSize;
     const int hLumFilterSize= c->hLumFilterSize;
@@ -539,7 +538,7 @@ static int swScale(SwsContext *c, const uint8_t* src[],
             int16_t *vLumFilter= c->vLumFilter;
             int16_t *vChrFilter= c->vChrFilter;
 
-            if (isPlanarYUV(dstFormat) || dstFormat==PIX_FMT_GRAY8) { //YV12 like
+            if (isPlanarYUV(dstFormat) || (isGray(dstFormat) && !isALPHA(dstFormat))) { //YV12 like
                 const int chrSkipMask= (1<<c->chrDstVSubSample)-1;
 
                 vLumFilter +=    dstY * vLumFilterSize;
@@ -594,8 +593,8 @@ static int swScale(SwsContext *c, const uint8_t* src[],
             } else {
                 assert(lumSrcPtr  + vLumFilterSize - 1 < lumPixBuf  + vLumBufSize*2);
                 assert(chrUSrcPtr + vChrFilterSize - 1 < chrUPixBuf + vChrBufSize*2);
-                if (c->yuv2packed1 && vLumFilterSize == 1 && vChrFilterSize == 2) { //unscaled RGB
-                    int chrAlpha = vChrFilter[2 * dstY + 1];
+                if (c->yuv2packed1 && vLumFilterSize == 1 && vChrFilterSize <= 2) { //unscaled RGB
+                    int chrAlpha = vChrFilterSize == 1 ? 0 : vChrFilter[2 * dstY + 1];
                     yuv2packed1(c, *lumSrcPtr, chrUSrcPtr, chrVSrcPtr,
                                 alpPixBuf ? *alpSrcPtr : NULL,
                                 dest[0], dstW, chrAlpha, dstY);