Merge remote-tracking branch 'qatar/master'

* qatar/master:
  proresdsp: port x86 assembly to cpuflags.
  lavr: x86: improve non-SSE4 version of S16_TO_S32_SX macro
  lavfi: better channel layout negotiation
  alac: check for truncated packets
  alac: reverse lpc coeff order, simplify filter
  lavr: add x86-optimized mixing functions
  x86: add support for fmaddps fma4 instruction with abstraction to avx/sse
  tscc2: fix typo in array index
  build: use COMPILE template for HOSTOBJS
  build: do full flag handling for all compiler-type tools
  eval: fix printing of NaN in eval fate test.
  build: Rename aandct component to more descriptive aandcttables
  mpegaudio: bury inline asm under HAVE_INLINE_ASM.
  x86inc: automatically insert vzeroupper for YMM functions.
  rtmp: Check the buffer length of ping packets
  rtmp: Allow having more unknown data at the end of a chunk size packet without failing
  rtmp: Prevent reading outside of an allocate buffer when receiving server bandwidth packets

Conflicts:
	Makefile
	configure
	libavcodec/x86/proresdsp.asm
	libavutil/eval.c

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

4 files changed, 30 insertions, 32 deletions

diff --git a/libavcodec/x86/dsputil_yasm.asm b/libavcodec/x86/dsputil_yasm.asm
index d2e5439e61..5a6c3d1eae 100644
--- a/libavcodec/x86/dsputil_yasm.asm
+++ b/libavcodec/x86/dsputil_yasm.asm
@@ -1158,12 +1158,7 @@ ALIGN 16
     add     src1q, 2*mmsize
     sub     lenq,  2*mmsize
     jge     .loop
-%if mmsize == 32
-    vzeroupper
-    RET
-%else
     REP_RET
-%endif
 %endmacro
 
 INIT_XMM sse
@@ -1193,12 +1188,7 @@ ALIGN 16
 
     sub     lenq,   2*mmsize
     jge     .loop
-%if mmsize == 32
-    vzeroupper
-    RET
-%else
     REP_RET
-%endif
 %endmacro
 
 INIT_XMM sse
@@ -1243,10 +1233,6 @@ cglobal butterflies_float_interleave, 4,4,3, dst, src0, src1, len
 %endif
     add       lenq, mmsize
     jl .loop
-%if mmsize == 32
-    vzeroupper
-    RET
-%endif
 .end:
     REP_RET
 %endmacro
diff --git a/libavcodec/x86/fft_mmx.asm b/libavcodec/x86/fft_mmx.asm
index 5bf0e2f259..5e5004ca97 100644
--- a/libavcodec/x86/fft_mmx.asm
+++ b/libavcodec/x86/fft_mmx.asm
@@ -750,9 +750,6 @@ section .text
 ; The others pass args in registers and don't spill anything.
 cglobal fft_dispatch%2, 2,5,8, z, nbits
     FFT_DISPATCH fullsuffix, nbits
-%if mmsize == 32
-    vzeroupper
-%endif
     RET
 %endmacro ; DECL_FFT
 
@@ -958,9 +955,6 @@ cglobal imdct_half, 3,12,8; FFTContext *s, FFTSample *output, const FFTSample *i
 %if ARCH_X86_64 == 0
     add esp, 12
 %endif
-%if mmsize == 32
-    vzeroupper
-%endif
     RET
 %endmacro
 
diff --git a/libavcodec/x86/mpegaudiodec_mmx.c b/libavcodec/x86/mpegaudiodec_mmx.c
index 939b441277..0d6cc08305 100644
--- a/libavcodec/x86/mpegaudiodec_mmx.c
+++ b/libavcodec/x86/mpegaudiodec_mmx.c
@@ -36,6 +36,8 @@ void ff_four_imdct36_float_avx(float *out, float *buf, float *in, float *win,
 
 DECLARE_ALIGNED(16, static float, mdct_win_sse)[2][4][4*40];
 
+#if HAVE_INLINE_ASM
+
 #define MACS(rt, ra, rb) rt+=(ra)*(rb)
 #define MLSS(rt, ra, rb) rt-=(ra)*(rb)
 
@@ -178,6 +180,7 @@ static void apply_window_mp3(float *in, float *win, int *unused, float *out,
     *out = sum;
 }
 
+#endif /* HAVE_INLINE_ASM */
 
 #define DECL_IMDCT_BLOCKS(CPU1, CPU2)                                       \
 static void imdct36_blocks_ ## CPU1(float *out, float *buf, float *in,      \
@@ -241,9 +244,11 @@ void ff_mpadsp_init_mmx(MPADSPContext *s)
         }
     }
 
+#if HAVE_INLINE_ASM
     if (mm_flags & AV_CPU_FLAG_SSE2) {
         s->apply_window_float = apply_window_mp3;
     }
+#endif /* HAVE_INLINE_ASM */
 #if HAVE_YASM
     if (0) {
 #if HAVE_AVX
diff --git a/libavcodec/x86/proresdsp.asm b/libavcodec/x86/proresdsp.asm
index a09d871fda..01e5deec93 100644
--- a/libavcodec/x86/proresdsp.asm
+++ b/libavcodec/x86/proresdsp.asm
@@ -83,8 +83,7 @@ section .text align=16
 
 ; %1 = row or col (for rounding variable)
 ; %2 = number of bits to shift at the end
-; %3 = optimization
-%macro IDCT_1D 3
+%macro IDCT_1D 2
     ; a0 = (W4 * row[0]) + (1 << (15 - 1));
     ; a1 = a0;
     ; a2 = a0;
@@ -235,8 +234,8 @@ section .text align=16
 
 ; void prores_idct_put_10_<opt>(uint8_t *pixels, int stride,
 ;                               DCTELEM *block, const int16_t *qmat);
-%macro idct_put_fn 2
-cglobal prores_idct_put_10_%1, 4, 4, %2
+%macro idct_put_fn 1
+cglobal prores_idct_put_10, 4, 4, %1
     movsxd      r1,  r1d
     pxor        m15, m15           ; zero
 
@@ -252,7 +251,7 @@ cglobal prores_idct_put_10_%1, 4, 4, %2
     pmullw      m13,[r3+64]
     pmullw      m12,[r3+96]
 
-    IDCT_1D     row, 15,  %1
+    IDCT_1D     row, 15
 
     ; transpose for second part of IDCT
     TRANSPOSE8x8W 8, 0, 1, 2, 4, 11, 9, 10, 3
@@ -267,7 +266,7 @@ cglobal prores_idct_put_10_%1, 4, 4, %2
 
     ; for (i = 0; i < 8; i++)
     ;     idctSparseColAdd(dest + i, line_size, block + i);
-    IDCT_1D     col, 18,  %1
+    IDCT_1D     col, 18
 
     ; clip/store
     mova        m3, [pw_4]
@@ -302,13 +301,27 @@ cglobal prores_idct_put_10_%1, 4, 4, %2
     RET
 %endmacro
 
-INIT_XMM
-idct_put_fn sse2, 16
-INIT_XMM
-idct_put_fn sse4, 16
+%macro SIGNEXTEND 2-3 ; dstlow, dsthigh, tmp
+%if cpuflag(sse4)
+    movhlps     %2,  %1
+    pmovsxwd    %1,  %1
+    pmovsxwd    %2,  %2
+%else ; sse2
+    pxor        %3,  %3
+    pcmpgtw     %3,  %1
+    mova        %2,  %1
+    punpcklwd   %1,  %3
+    punpckhwd   %2,  %3
+%endif
+%endmacro
+
+INIT_XMM sse2
+idct_put_fn 16
+INIT_XMM sse4
+idct_put_fn 16
 %if HAVE_AVX
-INIT_AVX
-idct_put_fn avx,  16
+INIT_XMM avx
+idct_put_fn 16
 %endif
 
 %endif