Merge remote-tracking branch 'qatar/master'

* qatar/master: (35 commits)
  h264_idct_10bit: port x86 assembly to cpuflags.
  x86inc: clip num_args to 7 on x86-32.
  x86inc: sync to latest version from x264.
  fft: rename "z" to "zc" to prevent name collision.
  wv: return meaningful error codes.
  wv: return AVERROR_EOF on EOF, not EIO.
  mp3dec: forward errors for av_get_packet().
  mp3dec: remove a pointless local variable.
  mp3dec: remove commented out cruft.
  lavfi: bump minor to mark stabilizing the ABI.
  FATE: add tests for yadif.
  FATE: add a test for delogo video filter.
  FATE: add a test for amix audio filter.
  audiogen: allow specifying random seed as a commandline parameter.
  vc1dec: Override invalid macroblock quantizer
  vc1: avoid reading beyond the last line in vc1_draw_sprites()
  vc1dec: check that coded slice positions and interlacing match.
  vc1dec: Do not ignore ff_vc1_parse_frame_header_adv return value
  configure: Move parts that should not be user-selectable to CONFIG_EXTRA
  lavf: remove commented out cruft in avformat_find_stream_info()
  ...

Conflicts:
	Makefile
	configure
	libavcodec/vc1dec.c
	libavcodec/x86/h264_deblock.asm
	libavcodec/x86/h264_deblock_10bit.asm
	libavcodec/x86/h264dsp_mmx.c
	libavfilter/version.h
	libavformat/mp3dec.c
	libavformat/utils.c
	libavformat/wv.c
	libavutil/x86/x86inc.asm

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

12 files changed, 360 insertions, 355 deletions

diff --git a/libavcodec/x86/dsputil_mmx.c b/libavcodec/x86/dsputil_mmx.c
index 1488389572..85753e07f5 100644
--- a/libavcodec/x86/dsputil_mmx.c
+++ b/libavcodec/x86/dsputil_mmx.c
@@ -2137,10 +2137,10 @@ void ff_ ## OP ## _h264_chroma_mc ## NUM ## _ ## DEPTH ## _ ## OPT      \
                                       (uint8_t *dst, uint8_t *src,      \
                                        int stride, int h, int x, int y);
 
-CHROMA_MC(put, 2, 10, mmxext)
-CHROMA_MC(avg, 2, 10, mmxext)
-CHROMA_MC(put, 4, 10, mmxext)
-CHROMA_MC(avg, 4, 10, mmxext)
+CHROMA_MC(put, 2, 10, mmx2)
+CHROMA_MC(avg, 2, 10, mmx2)
+CHROMA_MC(put, 4, 10, mmx2)
+CHROMA_MC(avg, 4, 10, mmx2)
 CHROMA_MC(put, 8, 10, sse2)
 CHROMA_MC(avg, 8, 10, sse2)
 CHROMA_MC(put, 8, 10, avx)
@@ -2841,10 +2841,10 @@ static void dsputil_init_mmx2(DSPContext *c, AVCodecContext *avctx,
         c->put_h264_chroma_pixels_tab[2] = ff_put_h264_chroma_mc2_mmx2;
     }
     if (bit_depth == 10 && CONFIG_H264CHROMA) {
-        c->put_h264_chroma_pixels_tab[2] = ff_put_h264_chroma_mc2_10_mmxext;
-        c->avg_h264_chroma_pixels_tab[2] = ff_avg_h264_chroma_mc2_10_mmxext;
-        c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_10_mmxext;
-        c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_10_mmxext;
+        c->put_h264_chroma_pixels_tab[2] = ff_put_h264_chroma_mc2_10_mmx2;
+        c->avg_h264_chroma_pixels_tab[2] = ff_avg_h264_chroma_mc2_10_mmx2;
+        c->put_h264_chroma_pixels_tab[1] = ff_put_h264_chroma_mc4_10_mmx2;
+        c->avg_h264_chroma_pixels_tab[1] = ff_avg_h264_chroma_mc4_10_mmx2;
     }
 
     c->add_hfyu_median_prediction   = ff_add_hfyu_median_prediction_mmx2;
diff --git a/libavcodec/x86/fft_mmx.asm b/libavcodec/x86/fft_mmx.asm
index 5e5004ca97..a709287f48 100644
--- a/libavcodec/x86/fft_mmx.asm
+++ b/libavcodec/x86/fft_mmx.asm
@@ -517,23 +517,23 @@ INIT_MMX 3dnow
 FFT48_3DN
 
 
-%define Z(x) [zq + o1q*(x&6) + mmsize*(x&1)]
-%define Z2(x) [zq + o3q + mmsize*(x&1)]
-%define ZH(x) [zq + o1q*(x&6) + mmsize*(x&1) + mmsize/2]
-%define Z2H(x) [zq + o3q + mmsize*(x&1) + mmsize/2]
+%define Z(x) [zcq + o1q*(x&6) + mmsize*(x&1)]
+%define Z2(x) [zcq + o3q + mmsize*(x&1)]
+%define ZH(x) [zcq + o1q*(x&6) + mmsize*(x&1) + mmsize/2]
+%define Z2H(x) [zcq + o3q + mmsize*(x&1) + mmsize/2]
 
 %macro DECL_PASS 2+ ; name, payload
 align 16
 %1:
-DEFINE_ARGS z, w, n, o1, o3
+DEFINE_ARGS zc, w, n, o1, o3
     lea o3q, [nq*3]
     lea o1q, [nq*8]
     shl o3q, 4
 .loop:
     %2
-    add zq, mmsize*2
-    add wq, mmsize
-    sub nd, mmsize/8
+    add zcq, mmsize*2
+    add  wq, mmsize
+    sub  nd, mmsize/8
     jg .loop
     rep ret
 %endmacro
@@ -748,7 +748,7 @@ section .text
 
 ; On x86_32, this function does the register saving and restoring for all of fft.
 ; The others pass args in registers and don't spill anything.
-cglobal fft_dispatch%2, 2,5,8, z, nbits
+cglobal fft_dispatch%2, 2,5,8, zc, nbits
     FFT_DISPATCH fullsuffix, nbits
     RET
 %endmacro ; DECL_FFT
diff --git a/libavcodec/x86/h264_chromamc_10bit.asm b/libavcodec/x86/h264_chromamc_10bit.asm
index 3f7c513069..370c7b5a46 100644
--- a/libavcodec/x86/h264_chromamc_10bit.asm
+++ b/libavcodec/x86/h264_chromamc_10bit.asm
@@ -60,10 +60,10 @@ SECTION .text
 ;-----------------------------------------------------------------------------
 ; void put/avg_h264_chroma_mc8(pixel *dst, pixel *src, int stride, int h, int mx, int my)
 ;-----------------------------------------------------------------------------
-%macro CHROMA_MC8 2
+%macro CHROMA_MC8 1
 ; put/avg_h264_chroma_mc8_*(uint8_t *dst /*align 8*/, uint8_t *src /*align 1*/,
 ;                              int stride, int h, int mx, int my)
-cglobal %1_h264_chroma_mc8_10_%2, 6,7,8
+cglobal %1_h264_chroma_mc8_10, 6,7,8
     movsxdifnidn  r2, r2d
     mov          r6d, r5d
     or           r6d, r4d
@@ -173,8 +173,8 @@ cglobal %1_h264_chroma_mc8_10_%2, 6,7,8
     add           r0, r2
 %endmacro
 
-%macro CHROMA_MC4 2
-cglobal %1_h264_chroma_mc4_10_%2, 6,6,7
+%macro CHROMA_MC4 1
+cglobal %1_h264_chroma_mc4_10, 6,6,7
     movsxdifnidn  r2, r2d
     movd          m2, r4m         ; x
     movd          m3, r5m         ; y
@@ -203,8 +203,8 @@ cglobal %1_h264_chroma_mc4_10_%2, 6,6,7
 ;-----------------------------------------------------------------------------
 ; void put/avg_h264_chroma_mc2(pixel *dst, pixel *src, int stride, int h, int mx, int my)
 ;-----------------------------------------------------------------------------
-%macro CHROMA_MC2 2
-cglobal %1_h264_chroma_mc2_10_%2, 6,7
+%macro CHROMA_MC2 1
+cglobal %1_h264_chroma_mc2_10, 6,7
     movsxdifnidn  r2, r2d
     mov          r6d, r4d
     shl          r4d, 16
@@ -250,24 +250,24 @@ cglobal %1_h264_chroma_mc2_10_%2, 6,7
 %endmacro
 
 %define CHROMAMC_AVG  NOTHING
-INIT_XMM
-CHROMA_MC8 put, sse2
+INIT_XMM sse2
+CHROMA_MC8 put
 %if HAVE_AVX
-INIT_AVX
-CHROMA_MC8 put, avx
+INIT_XMM avx
+CHROMA_MC8 put
 %endif
-INIT_MMX
-CHROMA_MC4 put, mmxext
-CHROMA_MC2 put, mmxext
+INIT_MMX mmx2
+CHROMA_MC4 put
+CHROMA_MC2 put
 
 %define CHROMAMC_AVG  AVG
 %define PAVG          pavgw
-INIT_XMM
-CHROMA_MC8 avg, sse2
+INIT_XMM sse2
+CHROMA_MC8 avg
 %if HAVE_AVX
-INIT_AVX
-CHROMA_MC8 avg, avx
+INIT_XMM avx
+CHROMA_MC8 avg
 %endif
-INIT_MMX
-CHROMA_MC4 avg, mmxext
-CHROMA_MC2 avg, mmxext
+INIT_MMX mmx2
+CHROMA_MC4 avg
+CHROMA_MC2 avg
diff --git a/libavcodec/x86/h264_deblock.asm b/libavcodec/x86/h264_deblock.asm
index 2f33f320a1..ea422e3b3a 100644
--- a/libavcodec/x86/h264_deblock.asm
+++ b/libavcodec/x86/h264_deblock.asm
@@ -282,8 +282,8 @@ cextern pb_A1
 ;-----------------------------------------------------------------------------
 ; void deblock_v_luma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-%macro DEBLOCK_LUMA 1
-cglobal deblock_v_luma_8_%1, 5,5,10
+%macro DEBLOCK_LUMA 0
+cglobal deblock_v_luma_8, 5,5,10
     movd    m8, [r4] ; tc0
     lea     r4, [r1*3]
     dec     r2d        ; alpha-1
@@ -327,8 +327,8 @@ cglobal deblock_v_luma_8_%1, 5,5,10
 ;-----------------------------------------------------------------------------
 ; void deblock_h_luma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-INIT_MMX
-cglobal deblock_h_luma_8_%1, 5,9
+INIT_MMX cpuname
+cglobal deblock_h_luma_8, 5,9
     movsxd r7,  r1d
     lea    r8,  [r7+r7*2]
     lea    r6,  [r0-4]
@@ -355,7 +355,7 @@ cglobal deblock_h_luma_8_%1, 5,9
 %if WIN64
     mov    [rsp+0x20], r4
 %endif
-    call   deblock_v_luma_8_%1
+    call   deblock_v_luma_8
 
     ; transpose 16x4 -> original space  (only the middle 4 rows were changed by the filter)
     add    r6, 2
@@ -384,26 +384,26 @@ cglobal deblock_h_luma_8_%1, 5,9
     RET
 %endmacro
 
-INIT_XMM
-DEBLOCK_LUMA sse2
+INIT_XMM sse2
+DEBLOCK_LUMA
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA avx
+INIT_XMM avx
+DEBLOCK_LUMA
 %endif
 
 %else
 
-%macro DEBLOCK_LUMA 3
+%macro DEBLOCK_LUMA 2
 ;-----------------------------------------------------------------------------
 ; void deblock_v8_luma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-cglobal deblock_%2_luma_8_%1, 5,5
+cglobal deblock_%1_luma_8, 5,5
     lea     r4, [r1*3]
     dec     r2     ; alpha-1
     neg     r4
     dec     r3     ; beta-1
     add     r4, r0 ; pix-3*stride
-    %assign pad 2*%3+12-(stack_offset&15)
+    %assign pad 2*%2+12-(stack_offset&15)
     SUB     esp, pad
 
     mova    m0, [r4+r1]   ; p1
@@ -417,7 +417,7 @@ cglobal deblock_%2_luma_8_%1, 5,5
     movd    m4, [r3] ; tc0
     punpcklbw m4, m4
     punpcklbw m4, m4 ; tc = 4x tc0[3], 4x tc0[2], 4x tc0[1], 4x tc0[0]
-    mova   [esp+%3], m4 ; tc
+    mova   [esp+%2], m4 ; tc
     pcmpgtb m4, m3
     mova    m3, [r4] ; p2
     pand    m4, m7
@@ -425,7 +425,7 @@ cglobal deblock_%2_luma_8_%1, 5,5
 
     DIFF_GT2 m1, m3, m5, m6, m7 ; |p2-p0| > beta-1
     pand    m6, m4
-    pand    m4, [esp+%3] ; tc
+    pand    m4, [esp+%2] ; tc
     psubb   m7, m4, m6
     pand    m6, m4
     LUMA_Q1 m0, m3, [r4], [r4+r1], m6, m4
@@ -433,7 +433,7 @@ cglobal deblock_%2_luma_8_%1, 5,5
     mova    m4, [r0+2*r1] ; q2
     DIFF_GT2 m2, m4, m5, m6, m3 ; |q2-q0| > beta-1
     pand    m6, [esp] ; mask
-    mova    m5, [esp+%3] ; tc
+    mova    m5, [esp+%2] ; tc
     psubb   m7, m6
     pand    m5, m6
     mova    m3, [r0+r1]
@@ -448,8 +448,8 @@ cglobal deblock_%2_luma_8_%1, 5,5
 ;-----------------------------------------------------------------------------
 ; void deblock_h_luma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-INIT_MMX
-cglobal deblock_h_luma_8_%1, 0,5
+INIT_MMX cpuname
+cglobal deblock_h_luma_8, 0,5
     mov    r0, r0mp
     mov    r3, r1m
     lea    r4, [r3*3]
@@ -472,11 +472,11 @@ cglobal deblock_h_luma_8_%1, 0,5
     PUSH   dword r2m
     PUSH   dword 16
     PUSH   dword r0
-    call   deblock_%2_luma_8_%1
-%ifidn %2, v8
+    call   deblock_%1_luma_8
+%ifidn %1, v8
     add    dword [esp   ], 8 ; pix_tmp+0x38
     add    dword [esp+16], 2 ; tc0+2
-    call   deblock_%2_luma_8_%1
+    call   deblock_%1_luma_8
 %endif
     ADD    esp, 20
 
@@ -503,13 +503,13 @@ cglobal deblock_h_luma_8_%1, 0,5
     RET
 %endmacro ; DEBLOCK_LUMA
 
-INIT_MMX
-DEBLOCK_LUMA mmxext, v8, 8
-INIT_XMM
-DEBLOCK_LUMA sse2, v, 16
+INIT_MMX mmx2
+DEBLOCK_LUMA v8, 8
+INIT_XMM sse2
+DEBLOCK_LUMA v, 16
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA avx, v, 16
+INIT_XMM avx
+DEBLOCK_LUMA v, 16
 %endif
 
 %endif ; ARCH
@@ -612,7 +612,7 @@ DEBLOCK_LUMA avx, v, 16
     %define mask1p mask1q
 %endmacro
 
-%macro DEBLOCK_LUMA_INTRA 2
+%macro DEBLOCK_LUMA_INTRA 1
     %define p1 m0
     %define p0 m1
     %define q0 m2
@@ -647,7 +647,7 @@ DEBLOCK_LUMA avx, v, 16
 ;-----------------------------------------------------------------------------
 ; void deblock_v_luma_intra( uint8_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_%2_luma_intra_8_%1, 4,6,16
+cglobal deblock_%1_luma_intra_8, 4,6,16
 %if ARCH_X86_64 == 0
     sub     esp, 0x60
 %endif
@@ -704,12 +704,12 @@ cglobal deblock_%2_luma_intra_8_%1, 4,6,16
 %endif
     RET
 
-INIT_MMX
+INIT_MMX cpuname
 %if ARCH_X86_64
 ;-----------------------------------------------------------------------------
 ; void deblock_h_luma_intra( uint8_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_h_luma_intra_8_%1, 4,9
+cglobal deblock_h_luma_intra_8, 4,9
     movsxd r7,  r1d
     lea    r8,  [r7*3]
     lea    r6,  [r0-4]
@@ -725,7 +725,7 @@ cglobal deblock_h_luma_intra_8_%1, 4,9
 
     lea    r0,  [pix_tmp+0x40]
     mov    r1,  0x10
-    call   deblock_v_luma_intra_8_%1
+    call   deblock_v_luma_intra_8
 
     ; transpose 16x6 -> original space (but we can't write only 6 pixels, so really 16x8)
     lea    r5, [r6+r8]
@@ -738,7 +738,7 @@ cglobal deblock_h_luma_intra_8_%1, 4,9
     add    rsp, 0x88
     RET
 %else
-cglobal deblock_h_luma_intra_8_%1, 2,4
+cglobal deblock_h_luma_intra_8, 2,4
     lea    r3,  [r1*3]
     sub    r0,  4
     lea    r2,  [r0+r3]
@@ -757,10 +757,10 @@ cglobal deblock_h_luma_intra_8_%1, 2,4
     PUSH   dword r2m
     PUSH   dword 16
     PUSH   r0
-    call   deblock_%2_luma_intra_8_%1
-%ifidn %2, v8
+    call   deblock_%1_luma_intra_8
+%ifidn %1, v8
     add    dword [rsp], 8 ; pix_tmp+8
-    call   deblock_%2_luma_intra_8_%1
+    call   deblock_%1_luma_intra_8
 %endif
     ADD    esp, 16
 
@@ -779,18 +779,18 @@ cglobal deblock_h_luma_intra_8_%1, 2,4
 %endif ; ARCH_X86_64
 %endmacro ; DEBLOCK_LUMA_INTRA
 
-INIT_XMM
-DEBLOCK_LUMA_INTRA sse2, v
+INIT_XMM sse2
+DEBLOCK_LUMA_INTRA v
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA_INTRA avx , v
+INIT_XMM avx
+DEBLOCK_LUMA_INTRA v
 %endif
 %if ARCH_X86_64 == 0
-INIT_MMX
-DEBLOCK_LUMA_INTRA mmxext, v8
+INIT_MMX mmx2
+DEBLOCK_LUMA_INTRA v8
 %endif
 
-INIT_MMX
+INIT_MMX mmx2
 
 %macro CHROMA_V_START 0
     dec    r2d      ; alpha-1
@@ -815,13 +815,13 @@ INIT_MMX
 ;-----------------------------------------------------------------------------
 ; void ff_deblock_v_chroma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_chroma_8_mmxext, 5,6
+cglobal deblock_v_chroma_8, 5,6
     CHROMA_V_START
     movq  m0, [t5]
     movq  m1, [t5+r1]
     movq  m2, [r0]
     movq  m3, [r0+r1]
-    call ff_chroma_inter_body_mmxext
+    call ff_chroma_inter_body_mmx2
     movq  [t5+r1], m1
     movq  [r0], m2
     RET
@@ -829,7 +829,7 @@ cglobal deblock_v_chroma_8_mmxext, 5,6
 ;-----------------------------------------------------------------------------
 ; void ff_deblock_h_chroma( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-cglobal deblock_h_chroma_8_mmxext, 5,7
+cglobal deblock_h_chroma_8, 5,7
 %if UNIX64
     %define buf0 [rsp-24]
     %define buf1 [rsp-16]
@@ -859,7 +859,7 @@ cglobal deblock_h_chroma_8_mmxext, 5,7
     RET
 
 ALIGN 16
-ff_chroma_inter_body_mmxext:
+ff_chroma_inter_body_mmx2:
     LOAD_MASK  r2d, r3d
     movd       m6, [r4] ; tc0
     punpcklbw  m6, m6
@@ -886,13 +886,13 @@ ff_chroma_inter_body_mmxext:
 ;-----------------------------------------------------------------------------
 ; void ff_deblock_v_chroma_intra( uint8_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_chroma_intra_8_mmxext, 4,5
+cglobal deblock_v_chroma_intra_8, 4,5
     CHROMA_V_START
     movq  m0, [t5]
     movq  m1, [t5+r1]
     movq  m2, [r0]
     movq  m3, [r0+r1]
-    call ff_chroma_intra_body_mmxext
+    call ff_chroma_intra_body_mmx2
     movq  [t5+r1], m1
     movq  [r0], m2
     RET
@@ -900,15 +900,15 @@ cglobal deblock_v_chroma_intra_8_mmxext, 4,5
 ;-----------------------------------------------------------------------------
 ; void ff_deblock_h_chroma_intra( uint8_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_h_chroma_intra_8_mmxext, 4,6
+cglobal deblock_h_chroma_intra_8, 4,6
     CHROMA_H_START
     TRANSPOSE4x8_LOAD  bw, wd, dq, PASS8ROWS(t5, r0, r1, t6)
-    call ff_chroma_intra_body_mmxext
+    call ff_chroma_intra_body_mmx2
     TRANSPOSE8x4B_STORE PASS8ROWS(t5, r0, r1, t6)
     RET
 
 ALIGN 16
-ff_chroma_intra_body_mmxext:
+ff_chroma_intra_body_mmx2:
     LOAD_MASK r2d, r3d
     movq   m5, m1
     movq   m6, m2
diff --git a/libavcodec/x86/h264_deblock_10bit.asm b/libavcodec/x86/h264_deblock_10bit.asm
index caf270163c..d625eee4a4 100644
--- a/libavcodec/x86/h264_deblock_10bit.asm
+++ b/libavcodec/x86/h264_deblock_10bit.asm
@@ -151,11 +151,11 @@ cextern pw_4
 %endif
 %endmacro
 
-%macro DEBLOCK_LUMA 1
+%macro DEBLOCK_LUMA 0
 ;-----------------------------------------------------------------------------
 ; void deblock_v_luma( uint16_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_luma_10_%1, 5,5,8*(mmsize/16)
+cglobal deblock_v_luma_10, 5,5,8*(mmsize/16)
     %assign pad 5*mmsize+12-(stack_offset&15)
     %define tcm [rsp]
     %define ms1 [rsp+mmsize]
@@ -210,7 +210,7 @@ cglobal deblock_v_luma_10_%1, 5,5,8*(mmsize/16)
     ADD         rsp, pad
     RET
 
-cglobal deblock_h_luma_10_%1, 5,6,8*(mmsize/16)
+cglobal deblock_h_luma_10, 5,6,8*(mmsize/16)
     %assign pad 7*mmsize+12-(stack_offset&15)
     %define tcm [rsp]
     %define ms1 [rsp+mmsize]
@@ -301,7 +301,6 @@ cglobal deblock_h_luma_10_%1, 5,6,8*(mmsize/16)
     RET
 %endmacro
 
-INIT_XMM
 %if ARCH_X86_64
 ; in:  m0=p1, m1=p0, m2=q0, m3=q1, m8=p2, m9=q2
 ;      m12=alpha, m13=beta
@@ -339,8 +338,8 @@ INIT_XMM
     SWAP         3, 9
 %endmacro
 
-%macro DEBLOCK_LUMA_64 1
-cglobal deblock_v_luma_10_%1, 5,5,15
+%macro DEBLOCK_LUMA_64 0
+cglobal deblock_v_luma_10, 5,5,15
     %define p2 m8
     %define p1 m0
     %define p0 m1
@@ -377,7 +376,7 @@ cglobal deblock_v_luma_10_%1, 5,5,15
     jg .loop
     REP_RET
 
-cglobal deblock_h_luma_10_%1, 5,7,15
+cglobal deblock_h_luma_10, 5,7,15
     shl        r2d, 2
     shl        r3d, 2
     LOAD_AB    m12, m13, r2d, r3d
@@ -417,11 +416,11 @@ cglobal deblock_h_luma_10_%1, 5,7,15
     REP_RET
 %endmacro
 
-INIT_XMM
-DEBLOCK_LUMA_64 sse2
+INIT_XMM sse2
+DEBLOCK_LUMA_64
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA_64 avx
+INIT_XMM avx
+DEBLOCK_LUMA_64
 %endif
 %endif
 
@@ -604,8 +603,8 @@ DEBLOCK_LUMA_64 avx
 ;-----------------------------------------------------------------------------
 ; void deblock_v_luma_intra( uint16_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-%macro DEBLOCK_LUMA_INTRA_64 1
-cglobal deblock_v_luma_intra_10_%1, 4,7,16
+%macro DEBLOCK_LUMA_INTRA_64 0
+cglobal deblock_v_luma_intra_10, 4,7,16
     %define t0 m1
     %define t1 m2
     %define t2 m4
@@ -655,7 +654,7 @@ cglobal deblock_v_luma_intra_10_%1, 4,7,16
 ;-----------------------------------------------------------------------------
 ; void deblock_h_luma_intra( uint16_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_h_luma_intra_10_%1, 4,7,16
+cglobal deblock_h_luma_intra_10, 4,7,16
     %define t0 m15
     %define t1 m14
     %define t2 m2
@@ -714,20 +713,20 @@ cglobal deblock_h_luma_intra_10_%1, 4,7,16
     RET
 %endmacro
 
-INIT_XMM
-DEBLOCK_LUMA_INTRA_64 sse2
+INIT_XMM sse2
+DEBLOCK_LUMA_INTRA_64
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA_INTRA_64 avx
+INIT_XMM avx
+DEBLOCK_LUMA_INTRA_64
 %endif
 
 %endif
 
-%macro DEBLOCK_LUMA_INTRA 1
+%macro DEBLOCK_LUMA_INTRA 0
 ;-----------------------------------------------------------------------------
 ; void deblock_v_luma_intra( uint16_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_luma_intra_10_%1, 4,7,8*(mmsize/16)
+cglobal deblock_v_luma_intra_10, 4,7,8*(mmsize/16)
     LUMA_INTRA_INIT 3
     lea     r4, [r1*4]
     lea     r5, [r1*3]
@@ -755,7 +754,7 @@ cglobal deblock_v_luma_intra_10_%1, 4,7,8*(mmsize/16)
 ;-----------------------------------------------------------------------------
 ; void deblock_h_luma_intra( uint16_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_h_luma_intra_10_%1, 4,7,8*(mmsize/16)
+cglobal deblock_h_luma_intra_10, 4,7,8*(mmsize/16)
     LUMA_INTRA_INIT 8
 %if mmsize == 8
     lea     r4, [r1*3]
@@ -797,16 +796,16 @@ cglobal deblock_h_luma_intra_10_%1, 4,7,8*(mmsize/16)
 %endmacro
 
 %if ARCH_X86_64 == 0
-INIT_MMX
-DEBLOCK_LUMA mmxext
-DEBLOCK_LUMA_INTRA mmxext
-INIT_XMM
-DEBLOCK_LUMA sse2
-DEBLOCK_LUMA_INTRA sse2
+INIT_MMX mmx2
+DEBLOCK_LUMA
+DEBLOCK_LUMA_INTRA
+INIT_XMM sse2
+DEBLOCK_LUMA
+DEBLOCK_LUMA_INTRA
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_LUMA avx
-DEBLOCK_LUMA_INTRA avx
+INIT_XMM avx
+DEBLOCK_LUMA
+DEBLOCK_LUMA_INTRA
 %endif
 %endif
 
@@ -849,11 +848,11 @@ DEBLOCK_LUMA_INTRA avx
     psraw       %1, 6
 %endmacro
 
-%macro DEBLOCK_CHROMA 1
+%macro DEBLOCK_CHROMA 0
 ;-----------------------------------------------------------------------------
 ; void deblock_v_chroma( uint16_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_chroma_10_%1, 5,7-(mmsize/16),8*(mmsize/16)
+cglobal deblock_v_chroma_10, 5,7-(mmsize/16),8*(mmsize/16)
     mov         r5, r0
     sub         r0, r1
     sub         r0, r1
@@ -887,7 +886,7 @@ cglobal deblock_v_chroma_10_%1, 5,7-(mmsize/16),8*(mmsize/16)
 ;-----------------------------------------------------------------------------
 ; void deblock_v_chroma_intra( uint16_t *pix, int stride, int alpha, int beta )
 ;-----------------------------------------------------------------------------
-cglobal deblock_v_chroma_intra_10_%1, 4,6-(mmsize/16),8*(mmsize/16)
+cglobal deblock_v_chroma_intra_10, 4,6-(mmsize/16),8*(mmsize/16)
     mov         r4, r0
     sub         r0, r1
     sub         r0, r1
@@ -914,12 +913,12 @@ cglobal deblock_v_chroma_intra_10_%1, 4,6-(mmsize/16),8*(mmsize/16)
 %endmacro
 
 %if ARCH_X86_64 == 0
-INIT_MMX
-DEBLOCK_CHROMA mmxext
+INIT_MMX mmx2
+DEBLOCK_CHROMA
 %endif
-INIT_XMM
-DEBLOCK_CHROMA sse2
+INIT_XMM sse2
+DEBLOCK_CHROMA
 %if HAVE_AVX
-INIT_AVX
-DEBLOCK_CHROMA avx
+INIT_XMM avx
+DEBLOCK_CHROMA
 %endif
diff --git a/libavcodec/x86/h264_idct_10bit.asm b/libavcodec/x86/h264_idct_10bit.asm
index a126573347..67bc68ec5c 100644
--- a/libavcodec/x86/h264_idct_10bit.asm
+++ b/libavcodec/x86/h264_idct_10bit.asm
@@ -72,25 +72,25 @@ SECTION .text
     STORE_DIFFx2 m2, m3, m4, m5, %1, %3
 %endmacro
 
-%macro IDCT_ADD_10 1
-cglobal h264_idct_add_10_%1, 3,3
+%macro IDCT_ADD_10 0
+cglobal h264_idct_add_10, 3,3
     IDCT4_ADD_10 r0, r1, r2
     RET
 %endmacro
 
-INIT_XMM
-IDCT_ADD_10 sse2
+INIT_XMM sse2
+IDCT_ADD_10
 %if HAVE_AVX
-INIT_AVX
-IDCT_ADD_10 avx
+INIT_XMM avx
+IDCT_ADD_10
 %endif
 
 ;-----------------------------------------------------------------------------
 ; h264_idct_add16(pixel *dst, const int *block_offset, dctcoef *block, int stride, const uint8_t nnzc[6*8])
 ;-----------------------------------------------------------------------------
 ;;;;;;; NO FATE SAMPLES TRIGGER THIS
-%macro ADD4x4IDCT 1
-add4x4_idct_%1:
+%macro ADD4x4IDCT 0
+add4x4_idct %+ SUFFIX:
     add   r5, r0
     mova  m0, [r2+ 0]
     mova  m1, [r2+16]
@@ -107,52 +107,52 @@ add4x4_idct_%1:
     ret
 %endmacro
 
-INIT_XMM
+INIT_XMM sse2
 ALIGN 16
-ADD4x4IDCT sse2
+ADD4x4IDCT
 %if HAVE_AVX
-INIT_AVX
+INIT_XMM avx
 ALIGN 16
-ADD4x4IDCT avx
+ADD4x4IDCT
 %endif
 
-%macro ADD16_OP 3
-    cmp          byte [r4+%3], 0
-    jz .skipblock%2
-    mov         r5d, [r1+%2*4]
-    call add4x4_idct_%1
-.skipblock%2:
-%if %2<15
+%macro ADD16_OP 2
+    cmp          byte [r4+%2], 0
+    jz .skipblock%1
+    mov         r5d, [r1+%1*4]
+    call add4x4_idct %+ SUFFIX
+.skipblock%1:
+%if %1<15
     add          r2, 64
 %endif
 %endmacro
 
-%macro IDCT_ADD16_10 1
-cglobal h264_idct_add16_10_%1, 5,6
-    ADD16_OP %1, 0, 4+1*8
-    ADD16_OP %1, 1, 5+1*8
-    ADD16_OP %1, 2, 4+2*8
-    ADD16_OP %1, 3, 5+2*8
-    ADD16_OP %1, 4, 6+1*8
-    ADD16_OP %1, 5, 7+1*8
-    ADD16_OP %1, 6, 6+2*8
-    ADD16_OP %1, 7, 7+2*8
-    ADD16_OP %1, 8, 4+3*8
-    ADD16_OP %1, 9, 5+3*8
-    ADD16_OP %1, 10, 4+4*8
-    ADD16_OP %1, 11, 5+4*8
-    ADD16_OP %1, 12, 6+3*8
-    ADD16_OP %1, 13, 7+3*8
-    ADD16_OP %1, 14, 6+4*8
-    ADD16_OP %1, 15, 7+4*8
+%macro IDCT_ADD16_10 0
+cglobal h264_idct_add16_10, 5,6
+    ADD16_OP 0, 4+1*8
+    ADD16_OP 1, 5+1*8
+    ADD16_OP 2, 4+2*8
+    ADD16_OP 3, 5+2*8
+    ADD16_OP 4, 6+1*8
+    ADD16_OP 5, 7+1*8
+    ADD16_OP 6, 6+2*8
+    ADD16_OP 7, 7+2*8
+    ADD16_OP 8, 4+3*8
+    ADD16_OP 9, 5+3*8
+    ADD16_OP 10, 4+4*8
+    ADD16_OP 11, 5+4*8
+    ADD16_OP 12, 6+3*8
+    ADD16_OP 13, 7+3*8
+    ADD16_OP 14, 6+4*8
+    ADD16_OP 15, 7+4*8
     REP_RET
 %endmacro
 
-INIT_XMM
-IDCT_ADD16_10 sse2
+INIT_XMM sse2
+IDCT_ADD16_10
 %if HAVE_AVX
-INIT_AVX
-IDCT_ADD16_10 avx
+INIT_XMM avx
+IDCT_ADD16_10
 %endif
 
 ;-----------------------------------------------------------------------------
@@ -185,8 +185,8 @@ IDCT_ADD16_10 avx
     mova [%1+%3  ], m4
 %endmacro
 
-INIT_MMX
-cglobal h264_idct_dc_add_10_mmx2,3,3
+INIT_MMX mmx2
+cglobal h264_idct_dc_add_10,3,3
     movd      m0, [r1]
     paddd     m0, [pd_32]
     psrad     m0, 6
@@ -199,8 +199,8 @@ cglobal h264_idct_dc_add_10_mmx2,3,3
 ;-----------------------------------------------------------------------------
 ; void h264_idct8_dc_add(pixel *dst, dctcoef *block, int stride)
 ;-----------------------------------------------------------------------------
-%macro IDCT8_DC_ADD 1
-cglobal h264_idct8_dc_add_10_%1,3,3,7
+%macro IDCT8_DC_ADD 0
+cglobal h264_idct8_dc_add_10,3,3,7
     mov      r1d, [r1]
     add       r1, 32
     sar       r1, 6
@@ -214,45 +214,45 @@ cglobal h264_idct8_dc_add_10_%1,3,3,7
     RET
 %endmacro
 
-INIT_XMM
-IDCT8_DC_ADD sse2
+INIT_XMM sse2
+IDCT8_DC_ADD
 %if HAVE_AVX
-INIT_AVX
-IDCT8_DC_ADD avx
+INIT_XMM avx
+IDCT8_DC_ADD
 %endif
 
 ;-----------------------------------------------------------------------------
 ; h264_idct_add16intra(pixel *dst, const int *block_offset, dctcoef *block, int stride, const uint8_t nnzc[6*8])
 ;-----------------------------------------------------------------------------
-%macro AC 2
-.ac%2
-    mov  r5d, [r1+(%2+0)*4]
-    call add4x4_idct_%1
-    mov  r5d, [r1+(%2+1)*4]
+%macro AC 1
+.ac%1
+    mov  r5d, [r1+(%1+0)*4]
+    call add4x4_idct %+ SUFFIX
+    mov  r5d, [r1+(%1+1)*4]
     add  r2, 64
-    call add4x4_idct_%1
+    call add4x4_idct %+ SUFFIX
     add  r2, 64
-    jmp .skipadd%2
+    jmp .skipadd%1
 %endmacro
 
 %assign last_block 16
-%macro ADD16_OP_INTRA 3
-    cmp      word [r4+%3], 0
-    jnz .ac%2
+%macro ADD16_OP_INTRA 2
+    cmp      word [r4+%2], 0
+    jnz .ac%1
     mov      r5d, [r2+ 0]
     or       r5d, [r2+64]
-    jz .skipblock%2
-    mov      r5d, [r1+(%2+0)*4]
-    call idct_dc_add_%1
-.skipblock%2:
-%if %2<last_block-2
+    jz .skipblock%1
+    mov      r5d, [r1+(%1+0)*4]
+    call idct_dc_add %+ SUFFIX
+.skipblock%1:
+%if %1<last_block-2
     add       r2, 128
 %endif
-.skipadd%2:
+.skipadd%1:
 %endmacro
 
-%macro IDCT_ADD16INTRA_10 1
-idct_dc_add_%1:
+%macro IDCT_ADD16INTRA_10 0
+idct_dc_add %+ SUFFIX:
     add       r5, r0
     movq      m0, [r2+ 0]
     movhps    m0, [r2+64]
@@ -265,46 +265,46 @@ idct_dc_add_%1:
     IDCT_DC_ADD_OP_10 r5, r3, r6
     ret
 
-cglobal h264_idct_add16intra_10_%1,5,7,8
-    ADD16_OP_INTRA %1, 0, 4+1*8
-    ADD16_OP_INTRA %1, 2, 4+2*8
-    ADD16_OP_INTRA %1, 4, 6+1*8
-    ADD16_OP_INTRA %1, 6, 6+2*8
-    ADD16_OP_INTRA %1, 8, 4+3*8
-    ADD16_OP_INTRA %1, 10, 4+4*8
-    ADD16_OP_INTRA %1, 12, 6+3*8
-    ADD16_OP_INTRA %1, 14, 6+4*8
+cglobal h264_idct_add16intra_10,5,7,8
+    ADD16_OP_INTRA 0, 4+1*8
+    ADD16_OP_INTRA 2, 4+2*8
+    ADD16_OP_INTRA 4, 6+1*8
+    ADD16_OP_INTRA 6, 6+2*8
+    ADD16_OP_INTRA 8, 4+3*8
+    ADD16_OP_INTRA 10, 4+4*8
+    ADD16_OP_INTRA 12, 6+3*8
+    ADD16_OP_INTRA 14, 6+4*8
     REP_RET
-    AC %1, 8
-    AC %1, 10
-    AC %1, 12
-    AC %1, 14
-    AC %1, 0
-    AC %1, 2
-    AC %1, 4
-    AC %1, 6
+    AC 8
+    AC 10
+    AC 12
+    AC 14
+    AC 0
+    AC 2
+    AC 4
+    AC 6
 %endmacro
 
-INIT_XMM
-IDCT_ADD16INTRA_10 sse2
+INIT_XMM sse2
+IDCT_ADD16INTRA_10
 %if HAVE_AVX
-INIT_AVX
-IDCT_ADD16INTRA_10 avx
+INIT_XMM avx
+IDCT_ADD16INTRA_10
 %endif
 
 %assign last_block 36
 ;-----------------------------------------------------------------------------
 ; h264_idct_add8(pixel **dst, const int *block_offset, dctcoef *block, int stride, const uint8_t nnzc[6*8])
 ;-----------------------------------------------------------------------------
-%macro IDCT_ADD8 1
-cglobal h264_idct_add8_10_%1,5,8,7
+%macro IDCT_ADD8 0
+cglobal h264_idct_add8_10,5,8,7
 %if ARCH_X86_64
     mov      r7, r0
 %endif
     add      r2, 1024
     mov      r0, [r0]
-    ADD16_OP_INTRA %1, 16, 4+ 6*8
-    ADD16_OP_INTRA %1, 18, 4+ 7*8
+    ADD16_OP_INTRA 16, 4+ 6*8
+    ADD16_OP_INTRA 18, 4+ 7*8
     add      r2, 1024-128*2
 %if ARCH_X86_64
     mov      r0, [r7+gprsize]
@@ -312,21 +312,21 @@ cglobal h264_idct_add8_10_%1,5,8,7
     mov      r0, r0m
     mov      r0, [r0+gprsize]
 %endif
-    ADD16_OP_INTRA %1, 32, 4+11*8
-    ADD16_OP_INTRA %1, 34, 4+12*8
+    ADD16_OP_INTRA 32, 4+11*8
+    ADD16_OP_INTRA 34, 4+12*8
     REP_RET
-    AC %1, 16
-    AC %1, 18
-    AC %1, 32
-    AC %1, 34
+    AC 16
+    AC 18
+    AC 32
+    AC 34
 
 %endmacro ; IDCT_ADD8
 
-INIT_XMM
-IDCT_ADD8 sse2
+INIT_XMM sse2
+IDCT_ADD8
 %if HAVE_AVX
-INIT_AVX
-IDCT_ADD8 avx
+INIT_XMM avx
+IDCT_ADD8
 %endif
 
 ;-----------------------------------------------------------------------------
@@ -432,19 +432,19 @@ IDCT_ADD8 avx
     STORE_DIFFx2 m0, m1, m6, m7, %1, %3
 %endmacro
 
-%macro IDCT8_ADD 1
-cglobal h264_idct8_add_10_%1, 3,4,16
+%macro IDCT8_ADD 0
+cglobal h264_idct8_add_10, 3,4,16
 %if UNIX64 == 0
     %assign pad 16-gprsize-(stack_offset&15)
     sub  rsp, pad
-    call h264_idct8_add1_10_%1
+    call h264_idct8_add1_10 %+ SUFFIX
     add  rsp, pad
     RET
 %endif
 
 ALIGN 16
 ; TODO: does not need to use stack
-h264_idct8_add1_10_%1:
+h264_idct8_add1_10 %+ SUFFIX:
 %assign pad 256+16-gprsize
     sub          rsp, pad
     add   dword [r1], 32
@@ -499,31 +499,31 @@ h264_idct8_add1_10_%1:
     ret
 %endmacro
 
-INIT_XMM
-IDCT8_ADD sse2
+INIT_XMM sse2
+IDCT8_ADD
 %if HAVE_AVX
-INIT_AVX
-IDCT8_ADD avx
+INIT_XMM avx
+IDCT8_ADD
 %endif
 
 ;-----------------------------------------------------------------------------
 ; h264_idct8_add4(pixel **dst, const int *block_offset, dctcoef *block, int stride, const uint8_t nnzc[6*8])
 ;-----------------------------------------------------------------------------
 ;;;;;;; NO FATE SAMPLES TRIGGER THIS
-%macro IDCT8_ADD4_OP 3
-    cmp       byte [r4+%3], 0
-    jz .skipblock%2
-    mov      r0d, [r6+%2*4]
+%macro IDCT8_ADD4_OP 2
+    cmp       byte [r4+%2], 0
+    jz .skipblock%1
+    mov      r0d, [r6+%1*4]
     add       r0, r5
-    call h264_idct8_add1_10_%1
-.skipblock%2:
-%if %2<12
+    call h264_idct8_add1_10 %+ SUFFIX
+.skipblock%1:
+%if %1<12
     add       r1, 256
 %endif
 %endmacro
 
-%macro IDCT8_ADD4 1
-cglobal h264_idct8_add4_10_%1, 0,7,16
+%macro IDCT8_ADD4 0
+cglobal h264_idct8_add4_10, 0,7,16
     %assign pad 16-gprsize-(stack_offset&15)
     SUB      rsp, pad
     mov       r5, r0mp
@@ -531,17 +531,17 @@ cglobal h264_idct8_add4_10_%1, 0,7,16
     mov       r1, r2mp
     mov      r2d, r3m
     movifnidn r4, r4mp
-    IDCT8_ADD4_OP %1,  0, 4+1*8
-    IDCT8_ADD4_OP %1,  4, 6+1*8
-    IDCT8_ADD4_OP %1,  8, 4+3*8
-    IDCT8_ADD4_OP %1, 12, 6+3*8
+    IDCT8_ADD4_OP  0, 4+1*8
+    IDCT8_ADD4_OP  4, 6+1*8
+    IDCT8_ADD4_OP  8, 4+3*8
+    IDCT8_ADD4_OP 12, 6+3*8
     ADD       rsp, pad
     RET
 %endmacro ; IDCT8_ADD4
 
-INIT_XMM
-IDCT8_ADD4 sse2
+INIT_XMM sse2
+IDCT8_ADD4
 %if HAVE_AVX
-INIT_AVX
-IDCT8_ADD4 avx
+INIT_XMM avx
+IDCT8_ADD4
 %endif
diff --git a/libavcodec/x86/h264dsp_mmx.c b/libavcodec/x86/h264dsp_mmx.c
index be7224aed4..d767c687ad 100644
--- a/libavcodec/x86/h264dsp_mmx.c
+++ b/libavcodec/x86/h264dsp_mmx.c
@@ -249,12 +249,12 @@ void ff_deblock_ ## DIR ## _ ## TYPE ## _ ## DEPTH ## _ ## OPT (uint8_t *pix, in
                                                                 int alpha, int beta);
 
 #define LF_FUNCS(type, depth)\
-LF_FUNC (h,  chroma,       depth, mmxext)\
-LF_IFUNC(h,  chroma_intra, depth, mmxext)\
-LF_FUNC (v,  chroma,       depth, mmxext)\
-LF_IFUNC(v,  chroma_intra, depth, mmxext)\
-LF_FUNC (h,  luma,         depth, mmxext)\
-LF_IFUNC(h,  luma_intra,   depth, mmxext)\
+LF_FUNC (h,  chroma,       depth, mmx2)\
+LF_IFUNC(h,  chroma_intra, depth, mmx2)\
+LF_FUNC (v,  chroma,       depth, mmx2)\
+LF_IFUNC(v,  chroma_intra, depth, mmx2)\
+LF_FUNC (h,  luma,         depth, mmx2)\
+LF_IFUNC(h,  luma_intra,   depth, mmx2)\
 LF_FUNC (h,  luma,         depth, sse2)\
 LF_IFUNC(h,  luma_intra,   depth, sse2)\
 LF_FUNC (v,  luma,         depth, sse2)\
@@ -276,24 +276,24 @@ LF_FUNCS( uint8_t,  8)
 LF_FUNCS(uint16_t, 10)
 
 #if ARCH_X86_32 && HAVE_YASM
-LF_FUNC (v8, luma,             8, mmxext)
-static void ff_deblock_v_luma_8_mmxext(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
+LF_FUNC (v8, luma,             8, mmx2)
+static void ff_deblock_v_luma_8_mmx2(uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0)
 {
     if((tc0[0] & tc0[1]) >= 0)
-        ff_deblock_v8_luma_8_mmxext(pix+0, stride, alpha, beta, tc0);
+        ff_deblock_v8_luma_8_mmx2(pix+0, stride, alpha, beta, tc0);
     if((tc0[2] & tc0[3]) >= 0)
-        ff_deblock_v8_luma_8_mmxext(pix+8, stride, alpha, beta, tc0+2);
+        ff_deblock_v8_luma_8_mmx2(pix+8, stride, alpha, beta, tc0+2);
 }
-LF_IFUNC(v8, luma_intra,        8, mmxext)
-static void ff_deblock_v_luma_intra_8_mmxext(uint8_t *pix, int stride, int alpha, int beta)
+LF_IFUNC(v8, luma_intra,        8, mmx2)
+static void ff_deblock_v_luma_intra_8_mmx2(uint8_t *pix, int stride, int alpha, int beta)
 {
-    ff_deblock_v8_luma_intra_8_mmxext(pix+0, stride, alpha, beta);
-    ff_deblock_v8_luma_intra_8_mmxext(pix+8, stride, alpha, beta);
+    ff_deblock_v8_luma_intra_8_mmx2(pix+0, stride, alpha, beta);
+    ff_deblock_v8_luma_intra_8_mmx2(pix+8, stride, alpha, beta);
 }
 #endif /* ARCH_X86_32 */
 
-LF_FUNC (v,  luma,            10, mmxext)
-LF_IFUNC(v,  luma_intra,      10, mmxext)
+LF_FUNC (v,  luma,            10, mmx2)
+LF_IFUNC(v,  luma_intra,      10, mmx2)
 
 /***********************************/
 /* weighted prediction */
@@ -373,17 +373,17 @@ void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth, const int chrom
                 c->h264_idct_add8  = ff_h264_idct_add8_8_mmx2;
             c->h264_idct_add16intra= ff_h264_idct_add16intra_8_mmx2;
 
-            c->h264_v_loop_filter_chroma= ff_deblock_v_chroma_8_mmxext;
-            c->h264_v_loop_filter_chroma_intra= ff_deblock_v_chroma_intra_8_mmxext;
+            c->h264_v_loop_filter_chroma= ff_deblock_v_chroma_8_mmx2;
+            c->h264_v_loop_filter_chroma_intra= ff_deblock_v_chroma_intra_8_mmx2;
             if (chroma_format_idc == 1) {
-                c->h264_h_loop_filter_chroma= ff_deblock_h_chroma_8_mmxext;
-                c->h264_h_loop_filter_chroma_intra= ff_deblock_h_chroma_intra_8_mmxext;
+                c->h264_h_loop_filter_chroma= ff_deblock_h_chroma_8_mmx2;
+                c->h264_h_loop_filter_chroma_intra= ff_deblock_h_chroma_intra_8_mmx2;
             }
 #if ARCH_X86_32
-            c->h264_v_loop_filter_luma= ff_deblock_v_luma_8_mmxext;
-            c->h264_h_loop_filter_luma= ff_deblock_h_luma_8_mmxext;
-            c->h264_v_loop_filter_luma_intra = ff_deblock_v_luma_intra_8_mmxext;
-            c->h264_h_loop_filter_luma_intra = ff_deblock_h_luma_intra_8_mmxext;
+            c->h264_v_loop_filter_luma= ff_deblock_v_luma_8_mmx2;
+            c->h264_h_loop_filter_luma= ff_deblock_h_luma_8_mmx2;
+            c->h264_v_loop_filter_luma_intra = ff_deblock_v_luma_intra_8_mmx2;
+            c->h264_h_loop_filter_luma_intra = ff_deblock_h_luma_intra_8_mmx2;
 #endif
             c->weight_h264_pixels_tab[0]= ff_h264_weight_16_mmx2;
             c->weight_h264_pixels_tab[1]= ff_h264_weight_8_mmx2;
@@ -436,12 +436,12 @@ void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth, const int chrom
     if (mm_flags & AV_CPU_FLAG_MMX) {
         if (mm_flags & AV_CPU_FLAG_MMX2) {
 #if ARCH_X86_32
-            c->h264_v_loop_filter_chroma= ff_deblock_v_chroma_10_mmxext;
-            c->h264_v_loop_filter_chroma_intra= ff_deblock_v_chroma_intra_10_mmxext;
-            c->h264_v_loop_filter_luma= ff_deblock_v_luma_10_mmxext;
-            c->h264_h_loop_filter_luma= ff_deblock_h_luma_10_mmxext;
-            c->h264_v_loop_filter_luma_intra = ff_deblock_v_luma_intra_10_mmxext;
-            c->h264_h_loop_filter_luma_intra = ff_deblock_h_luma_intra_10_mmxext;
+            c->h264_v_loop_filter_chroma= ff_deblock_v_chroma_10_mmx2;
+            c->h264_v_loop_filter_chroma_intra= ff_deblock_v_chroma_intra_10_mmx2;
+            c->h264_v_loop_filter_luma= ff_deblock_v_luma_10_mmx2;
+            c->h264_h_loop_filter_luma= ff_deblock_h_luma_10_mmx2;
+            c->h264_v_loop_filter_luma_intra = ff_deblock_v_luma_intra_10_mmx2;
+            c->h264_h_loop_filter_luma_intra = ff_deblock_h_luma_intra_10_mmx2;
 #endif
             c->h264_idct_dc_add= ff_h264_idct_dc_add_10_mmx2;
             if (mm_flags&AV_CPU_FLAG_SSE2) {
diff --git a/libavcodec/x86/rv34dsp.asm b/libavcodec/x86/rv34dsp.asm
index ae04e82c2d..78d8c92b0b 100644
--- a/libavcodec/x86/rv34dsp.asm
+++ b/libavcodec/x86/rv34dsp.asm
@@ -46,7 +46,7 @@ SECTION .text
 %endmacro
 
 %macro rv34_idct 1
-cglobal rv34_idct_%1_mmx2, 1, 2, 0
+cglobal rv34_idct_%1, 1, 2, 0
     movsx   r1, word [r0]
     IDCT_DC r1
     movd    m0, r1d
@@ -58,14 +58,15 @@ cglobal rv34_idct_%1_mmx2, 1, 2, 0
     REP_RET
 %endmacro
 
-INIT_MMX
+INIT_MMX mmx2
 %define IDCT_DC IDCT_DC_ROUND
 rv34_idct dc
 %define IDCT_DC IDCT_DC_NOROUND
 rv34_idct dc_noround
 
 ; ff_rv34_idct_dc_add_mmx(uint8_t *dst, int stride, int dc);
-cglobal rv34_idct_dc_add_mmx, 3, 3
+INIT_MMX mmx
+cglobal rv34_idct_dc_add, 3, 3
     ; calculate DC
     IDCT_DC_ROUND r2
     pxor       m1, m1
@@ -167,8 +168,8 @@ cglobal rv34_idct_add, 3,3,0, d, s, b
     ret
 
 ; ff_rv34_idct_dc_add_sse4(uint8_t *dst, int stride, int dc);
-INIT_XMM
-cglobal rv34_idct_dc_add_sse4, 3, 3, 6
+INIT_XMM sse4
+cglobal rv34_idct_dc_add, 3, 3, 6
     ; load data
     IDCT_DC_ROUND r2
     pxor       m1, m1
diff --git a/libavcodec/x86/vp3dsp.asm b/libavcodec/x86/vp3dsp.asm
index 46bd9d8f86..dcd55f5f8b 100644
--- a/libavcodec/x86/vp3dsp.asm
+++ b/libavcodec/x86/vp3dsp.asm
@@ -102,8 +102,8 @@ SECTION .text
     mov  [r0+r3  -1], r2w
 %endmacro
 
-INIT_MMX
-cglobal vp3_v_loop_filter_mmx2, 3, 4
+INIT_MMX mmx2
+cglobal vp3_v_loop_filter, 3, 4
 %if ARCH_X86_64
     movsxd        r1, r1d
 %endif
@@ -120,7 +120,7 @@ cglobal vp3_v_loop_filter_mmx2, 3, 4
     movq     [r0   ], m3
     RET
 
-cglobal vp3_h_loop_filter_mmx2, 3, 4
+cglobal vp3_h_loop_filter, 3, 4
 %if ARCH_X86_64
     movsxd        r1, r1d
 %endif
@@ -354,38 +354,6 @@ cglobal vp3_h_loop_filter_mmx2, 3, 4
     movq        I(2), m2
 %endmacro
 
-%macro VP3_IDCT_mmx 1
-    ; eax = quantized input
-    ; ebx = dequantizer matrix
-    ; ecx = IDCT constants
-    ;  M(I) = ecx + MaskOffset(0) + I * 8
-    ;  C(I) = ecx + CosineOffset(32) + (I-1) * 8
-    ; edx = output
-    ; r0..r7 = mm0..mm7
-%define OC_8 [pw_8]
-%define C(x) [vp3_idct_data+16*(x-1)]
-
-    ; at this point, function has completed dequantization + dezigzag +
-    ; partial transposition; now do the idct itself
-%define I(x) [%1+16* x     ]
-%define J(x) [%1+16*(x-4)+8]
-    RowIDCT
-    Transpose
-
-%define I(x) [%1+16* x   +64]
-%define J(x) [%1+16*(x-4)+72]
-    RowIDCT
-    Transpose
-
-%define I(x) [%1+16*x]
-%define J(x) [%1+16*x]
-    ColumnIDCT
-
-%define I(x) [%1+16*x+8]
-%define J(x) [%1+16*x+8]
-    ColumnIDCT
-%endmacro
-
 %macro VP3_1D_IDCT_SSE2 0
     movdqa        m2, I(3)      ; xmm2 = i3
     movdqa        m6, C(3)      ; xmm6 = c3
@@ -501,7 +469,8 @@ cglobal vp3_h_loop_filter_mmx2, 3, 4
     movdqa      O(7), m%8
 %endmacro
 
-%macro VP3_IDCT_sse2 1
+%macro VP3_IDCT 1
+%if mmsize == 16
 %define I(x) [%1+16*x]
 %define O(x) [%1+16*x]
 %define C(x) [vp3_idct_data+16*(x-1)]
@@ -519,11 +488,42 @@ cglobal vp3_h_loop_filter_mmx2, 3, 4
 %define ADD(x)   paddsw x, [pw_8]
         VP3_1D_IDCT_SSE2
         PUT_BLOCK 0, 1, 2, 3, 4, 5, 6, 7
+%else ; mmsize == 8
+    ; eax = quantized input
+    ; ebx = dequantizer matrix
+    ; ecx = IDCT constants
+    ;  M(I) = ecx + MaskOffset(0) + I * 8
+    ;  C(I) = ecx + CosineOffset(32) + (I-1) * 8
+    ; edx = output
+    ; r0..r7 = mm0..mm7
+%define OC_8 [pw_8]
+%define C(x) [vp3_idct_data+16*(x-1)]
+
+    ; at this point, function has completed dequantization + dezigzag +
+    ; partial transposition; now do the idct itself
+%define I(x) [%1+16* x     ]
+%define J(x) [%1+16*(x-4)+8]
+    RowIDCT
+    Transpose
+
+%define I(x) [%1+16* x   +64]
+%define J(x) [%1+16*(x-4)+72]
+    RowIDCT
+    Transpose
+
+%define I(x) [%1+16*x]
+%define J(x) [%1+16*x]
+    ColumnIDCT
+
+%define I(x) [%1+16*x+8]
+%define J(x) [%1+16*x+8]
+    ColumnIDCT
+%endif ; mmsize == 16/8
 %endmacro
 
-%macro vp3_idct_funcs 1
-cglobal vp3_idct_put_%1, 3, 4, 9
-    VP3_IDCT_%1   r2
+%macro vp3_idct_funcs 0
+cglobal vp3_idct_put, 3, 4, 9
+    VP3_IDCT      r2
 
     movsxdifnidn  r1, r1d
     mova          m4, [pb_80]
@@ -565,8 +565,8 @@ cglobal vp3_idct_put_%1, 3, 4, 9
 %endrep
     RET
 
-cglobal vp3_idct_add_%1, 3, 4, 9
-    VP3_IDCT_%1   r2
+cglobal vp3_idct_add, 3, 4, 9
+    VP3_IDCT      r2
 
     mov           r3, 4
     pxor          m4, m4
@@ -607,10 +607,13 @@ cglobal vp3_idct_add_%1, 3, 4, 9
     RET
 %endmacro
 
-INIT_MMX
-vp3_idct_funcs mmx
-INIT_XMM
-vp3_idct_funcs sse2
+%if ARCH_X86_32
+INIT_MMX mmx
+vp3_idct_funcs
+%endif
+
+INIT_XMM sse2
+vp3_idct_funcs
 
 %macro DC_ADD 0
     movq          m2, [r0     ]
@@ -631,8 +634,8 @@ vp3_idct_funcs sse2
     movq   [r0+r3  ], m5
 %endmacro
 
-INIT_MMX
-cglobal vp3_idct_dc_add_mmx2, 3, 4
+INIT_MMX mmx2
+cglobal vp3_idct_dc_add, 3, 4
 %if ARCH_X86_64
     movsxd        r1, r1d
 %endif
diff --git a/libavcodec/x86/vp3dsp_init.c b/libavcodec/x86/vp3dsp_init.c
index 3ae2a90e57..e0366d0f06 100644
--- a/libavcodec/x86/vp3dsp_init.c
+++ b/libavcodec/x86/vp3dsp_init.c
@@ -41,11 +41,13 @@ av_cold void ff_vp3dsp_init_x86(VP3DSPContext *c, int flags)
 #if HAVE_YASM
     int cpuflags = av_get_cpu_flags();
 
+#if ARCH_X86_32
     if (HAVE_MMX && cpuflags & AV_CPU_FLAG_MMX) {
         c->idct_put  = ff_vp3_idct_put_mmx;
         c->idct_add  = ff_vp3_idct_add_mmx;
         c->idct_perm = FF_PARTTRANS_IDCT_PERM;
     }
+#endif
 
     if (HAVE_MMX2 && cpuflags & AV_CPU_FLAG_MMX2) {
         c->idct_dc_add = ff_vp3_idct_dc_add_mmx2;
diff --git a/libavcodec/x86/vp56dsp.asm b/libavcodec/x86/vp56dsp.asm
index 2d409bfca1..21ff041d61 100644
--- a/libavcodec/x86/vp56dsp.asm
+++ b/libavcodec/x86/vp56dsp.asm
@@ -27,7 +27,8 @@ cextern pw_64
 
 SECTION .text
 
-%macro DIAG4_MMX 6
+%macro DIAG4 6
+%if mmsize == 8
     movq          m0, [%1+%2]
     movq          m1, [%1+%3]
     movq          m3, m0
@@ -64,9 +65,7 @@ SECTION .text
     psraw         m3, 7
     packuswb      m0, m3
     movq        [%6], m0
-%endmacro
-
-%macro DIAG4_SSE2 6
+%else ; mmsize == 16
     movq          m0, [%1+%2]
     movq          m1, [%1+%3]
     punpcklbw     m0, m7
@@ -86,9 +85,11 @@ SECTION .text
     psraw         m0, 7
     packuswb      m0, m0
     movq        [%6], m0
+%endif ; mmsize == 8/16
 %endmacro
 
-%macro SPLAT4REGS_MMX 0
+%macro SPLAT4REGS 0
+%if mmsize == 8
     movq         m5, m3
     punpcklwd    m3, m3
     movq         m4, m3
@@ -102,9 +103,7 @@ SECTION .text
     movq [rsp+8*12], m4
     movq [rsp+8*13], m5
     movq [rsp+8*14], m2
-%endmacro
-
-%macro SPLAT4REGS_SSE2 0
+%else ; mmsize == 16
     pshuflw      m4, m3, 0x0
     pshuflw      m5, m3, 0x55
     pshuflw      m6, m3, 0xAA
@@ -113,15 +112,16 @@ SECTION .text
     punpcklqdq   m5, m5
     punpcklqdq   m6, m6
     punpcklqdq   m3, m3
+%endif ; mmsize == 8/16
 %endmacro
 
-%macro vp6_filter_diag4 2
+%macro vp6_filter_diag4 0
 ; void ff_vp6_filter_diag4_<opt>(uint8_t *dst, uint8_t *src, int stride,
 ;                                const int16_t h_weight[4], const int16_t v_weights[4])
-cglobal vp6_filter_diag4_%1, 5, 7, %2
+cglobal vp6_filter_diag4, 5, 7, 8
     mov          r5, rsp         ; backup stack pointer
     and         rsp, ~(mmsize-1) ; align stack
-%ifidn %1, sse2
+%if mmsize == 16
     sub         rsp, 8*11
 %else
     sub         rsp, 8*15
@@ -162,12 +162,10 @@ cglobal vp6_filter_diag4_%1, 5, 7, %2
     RET
 %endmacro
 
-INIT_MMX
-%define DIAG4      DIAG4_MMX
-%define SPLAT4REGS SPLAT4REGS_MMX
-vp6_filter_diag4 mmx,  0
+%if ARCH_X86_32
+INIT_MMX mmx
+vp6_filter_diag4
+%endif
 
-INIT_XMM
-%define DIAG4      DIAG4_SSE2
-%define SPLAT4REGS SPLAT4REGS_SSE2
-vp6_filter_diag4 sse2, 8
+INIT_XMM sse2
+vp6_filter_diag4
diff --git a/libavcodec/x86/vp56dsp_init.c b/libavcodec/x86/vp56dsp_init.c
index 87fc935315..9108aa446d 100644
--- a/libavcodec/x86/vp56dsp_init.c
+++ b/libavcodec/x86/vp56dsp_init.c
@@ -36,9 +36,11 @@ av_cold void ff_vp56dsp_init_x86(VP56DSPContext* c, enum CodecID codec)
     int mm_flags = av_get_cpu_flags();
 
     if (CONFIG_VP6_DECODER && codec == CODEC_ID_VP6) {
+#if ARCH_X86_32
         if (mm_flags & AV_CPU_FLAG_MMX) {
             c->vp6_filter_diag4 = ff_vp6_filter_diag4_mmx;
         }
+#endif
 
         if (mm_flags & AV_CPU_FLAG_SSE2) {
             c->vp6_filter_diag4 = ff_vp6_filter_diag4_sse2;