Merge remote-tracking branch 'qatar/master'

* qatar/master:
  Fix compilation of iirfilter-test.
  libx264: handle closed GOP codec flag
  lavf: remove duplicate assignment in avformat_alloc_context.
  lavf: use designated initializers for AVClasses.
  flvdec: clenup debug code
  asfdec: fix possible overread on broken files.
  asfdec: do not fall back to binary/generic search
  asfdec: reindent after previous commit c7bd5ed
  asfdec: fallback to binary search internally
  mpegaudio: add _fixed suffix to some names
  Modify x86util.asm to ease transitioning to 10-bit H.264 assembly.
  dct: build dct32 as separate object files
  qdm2: include correct header for rdft

Conflicts:
	ffpresets/libx264-fast.ffpreset
	ffpresets/libx264-fast_firstpass.ffpreset
	ffpresets/libx264-faster.ffpreset
	ffpresets/libx264-faster_firstpass.ffpreset
	ffpresets/libx264-medium.ffpreset
	ffpresets/libx264-medium_firstpass.ffpreset
	ffpresets/libx264-placebo.ffpreset
	ffpresets/libx264-placebo_firstpass.ffpreset
	ffpresets/libx264-slow.ffpreset
	ffpresets/libx264-slow_firstpass.ffpreset
	ffpresets/libx264-slower.ffpreset
	ffpresets/libx264-slower_firstpass.ffpreset
	ffpresets/libx264-superfast.ffpreset
	ffpresets/libx264-superfast_firstpass.ffpreset
	ffpresets/libx264-ultrafast.ffpreset
	ffpresets/libx264-ultrafast_firstpass.ffpreset
	ffpresets/libx264-veryfast.ffpreset
	ffpresets/libx264-veryfast_firstpass.ffpreset
	ffpresets/libx264-veryslow.ffpreset
	ffpresets/libx264-veryslow_firstpass.ffpreset
	libavformat/flvdec.c

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

configure

@@ -2927,8 +2927,8 @@ enabled libvpx     && {
     enabled libvpx_encoder && { check_lib2 "vpx/vpx_encoder.h vpx/vp8cx.h" "vpx_codec_enc_init_ver VPX_CQ" -lvpx ||
                                 die "ERROR: libvpx encoder version must be >=0.9.6"; } }
 enabled libx264    && require  libx264 x264.h x264_encoder_encode -lx264 &&
-                      { check_cpp_condition x264.h "X264_BUILD >= 99" ||
-                        die "ERROR: libx264 version must be >= 0.99."; }
+                      { check_cpp_condition x264.h "X264_BUILD >= 115" ||
+                        die "ERROR: libx264 version must be >= 0.115."; }
 enabled libxavs    && require  libxavs xavs.h xavs_encoder_encode -lxavs
 enabled libxvid    && require  libxvid xvid.h xvid_global -lxvidcore
 enabled mlib       && require  mediaLib mlib_types.h mlib_VectorSub_S16_U8_Mod -lmlib

ffpresets/libx264-lossless_fast.ffpreset

@@ -1,5 +1,5 @@
 coder=0
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=-parti8x8+parti4x4+partp8x8-partp4x4-partb8x8
 me_method=hex

ffpresets/libx264-lossless_max.ffpreset

@@ -1,5 +1,5 @@
 coder=1
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=+parti8x8+parti4x4+partp8x8+partp4x4-partb8x8
 me_method=esa

ffpresets/libx264-lossless_medium.ffpreset

@@ -1,5 +1,5 @@
 coder=1
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=-parti8x8+parti4x4+partp8x8+partp4x4-partb8x8
 me_method=hex

ffpresets/libx264-lossless_slow.ffpreset

@@ -1,5 +1,5 @@
 coder=1
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=+parti8x8+parti4x4+partp8x8+partp4x4-partb8x8
 me_method=umh

ffpresets/libx264-lossless_slower.ffpreset

@@ -1,5 +1,5 @@
 coder=1
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=+parti8x8+parti4x4+partp8x8+partp4x4-partb8x8
 me_method=umh

ffpresets/libx264-lossless_ultrafast.ffpreset

@@ -1,5 +1,5 @@
 coder=0
-flags=+loop
+flags=+loop+cgop
 cmp=+chroma
 partitions=-parti8x8-parti4x4-partp8x8-partp4x4-partb8x8
 me_method=dia

libavcodec/Makefile

@@ -28,7 +28,7 @@ OBJS-$(CONFIG_AANDCT)                  += aandcttab.o
 OBJS-$(CONFIG_AC3DSP)                  += ac3dsp.o
 OBJS-$(CONFIG_CRYSTALHD)               += crystalhd.o
 OBJS-$(CONFIG_ENCODERS)                += faandct.o jfdctfst.o jfdctint.o
-OBJS-$(CONFIG_DCT)                     += dct.o
+OBJS-$(CONFIG_DCT)                     += dct.o dct32_fixed.o dct32_float.o
 OBJS-$(CONFIG_DWT)                     += dwt.o
 OBJS-$(CONFIG_DXVA2)                   += dxva2.o
 FFT-OBJS-$(CONFIG_HARDCODED_TABLES)    += cos_tables.o cos_fixed_tables.o

libavcodec/dct.c

@@ -30,9 +30,7 @@
 #include <math.h>
 #include "libavutil/mathematics.h"
 #include "dct.h"
-
-#define DCT32_FLOAT
-#include "dct32.c"
+#include "dct32.h"
 
 /* sin((M_PI * x / (2*n)) */
 #define SIN(s,n,x) (s->costab[(n) - (x)])
@@ -210,7 +208,7 @@ av_cold int ff_dct_init(DCTContext *s, int nbits, enum DCTTransformType inverse)
         }
     }
 
-    s->dct32 = dct32;
+    s->dct32 = ff_dct32_float;
     if (HAVE_MMX)     ff_dct_init_mmx(s);
 
     return 0;

libavcodec/dct32.c

@@ -19,10 +19,19 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
-#ifdef DCT32_FLOAT
+#include "dct32.h"
+#include "mathops.h"
+
+#if DCT32_FLOAT
+#   define dct32 ff_dct32_float
 #   define FIXHR(x)       ((float)(x))
 #   define MULH3(x, y, s) ((s)*(y)*(x))
 #   define INTFLOAT float
+#else
+#   define dct32 ff_dct32_fixed
+#   define FIXHR(a)       ((int)((a) * (1LL<<32) + 0.5))
+#   define MULH3(x, y, s) MULH((s)*(x), y)
+#   define INTFLOAT int
 #endif
 
 
@@ -103,7 +112,7 @@
 #define ADD(a, b) val##a += val##b
 
 /* DCT32 without 1/sqrt(2) coef zero scaling. */
-static void dct32(INTFLOAT *out, const INTFLOAT *tab)
+void dct32(INTFLOAT *out, const INTFLOAT *tab)
 {
     INTFLOAT tmp0, tmp1;