dsputil: Move pix_sum, pix_norm1, shrink function pointers to mpegvideoenc

5 files changed, 106 insertions, 74 deletions

diff --git a/libavcodec/x86/Makefile b/libavcodec/x86/Makefile
index bfc13737d1..701aad7916 100644
--- a/libavcodec/x86/Makefile
+++ b/libavcodec/x86/Makefile
@@ -92,6 +92,7 @@ YASM-OBJS-$(CONFIG_HPELDSP)            += x86/fpel.o                    \
                                           x86/hpeldsp.o
 YASM-OBJS-$(CONFIG_HUFFYUVDSP)         += x86/huffyuvdsp.o
 YASM-OBJS-$(CONFIG_MPEGAUDIODSP)       += x86/imdct36.o
+YASM-OBJS-$(CONFIG_MPEGVIDEOENC)       += x86/mpegvideoencdsp.o
 YASM-OBJS-$(CONFIG_QPELDSP)            += x86/qpeldsp.o                 \
                                           x86/fpel.o                    \
                                           x86/qpel.o
diff --git a/libavcodec/x86/dsputilenc.asm b/libavcodec/x86/dsputilenc.asm
index f676ba63fd..d996962eac 100644
--- a/libavcodec/x86/dsputilenc.asm
+++ b/libavcodec/x86/dsputilenc.asm
@@ -418,72 +418,3 @@ cglobal diff_pixels, 4,5
     add          r4, 16
     jne .loop
     REP_RET
-
-INIT_MMX mmx
-; int ff_pix_sum16_mmx(uint8_t *pix, int line_size)
-cglobal pix_sum16, 2, 3
-    movsxdifnidn r1, r1d
-    mov          r2, r1
-    neg          r2
-    shl          r2, 4
-    sub          r0, r2
-    pxor         m7, m7
-    pxor         m6, m6
-.loop:
-    mova         m0, [r0+r2+0]
-    mova         m1, [r0+r2+0]
-    mova         m2, [r0+r2+8]
-    mova         m3, [r0+r2+8]
-    punpcklbw    m0, m7
-    punpckhbw    m1, m7
-    punpcklbw    m2, m7
-    punpckhbw    m3, m7
-    paddw        m1, m0
-    paddw        m3, m2
-    paddw        m3, m1
-    paddw        m6, m3
-    add          r2, r1
-    js .loop
-    mova         m5, m6
-    psrlq        m6, 32
-    paddw        m6, m5
-    mova         m5, m6
-    psrlq        m6, 16
-    paddw        m6, m5
-    movd        eax, m6
-    and         eax, 0xffff
-    RET
-
-INIT_MMX mmx
-; int ff_pix_norm1_mmx(uint8_t *pix, int line_size)
-cglobal pix_norm1, 2, 4
-    movsxdifnidn r1, r1d
-    mov          r2, 16
-    pxor         m0, m0
-    pxor         m7, m7
-.loop:
-    mova         m2, [r0+0]
-    mova         m3, [r0+8]
-    mova         m1, m2
-    punpckhbw    m1, m0
-    punpcklbw    m2, m0
-    mova         m4, m3
-    punpckhbw    m3, m0
-    punpcklbw    m4, m0
-    pmaddwd      m1, m1
-    pmaddwd      m2, m2
-    pmaddwd      m3, m3
-    pmaddwd      m4, m4
-    paddd        m2, m1
-    paddd        m4, m3
-    paddd        m7, m2
-    add          r0, r1
-    paddd        m7, m4
-    dec r2
-    jne .loop
-    mova         m1, m7
-    psrlq        m7, 32
-    paddd        m1, m7
-    movd        eax, m1
-    RET
-
diff --git a/libavcodec/x86/dsputilenc_mmx.c b/libavcodec/x86/dsputilenc_mmx.c
index 7af091365f..563543ee7d 100644
--- a/libavcodec/x86/dsputilenc_mmx.c
+++ b/libavcodec/x86/dsputilenc_mmx.c
@@ -35,8 +35,6 @@ void ff_get_pixels_mmx(int16_t *block, const uint8_t *pixels, int line_size);
 void ff_get_pixels_sse2(int16_t *block, const uint8_t *pixels, int line_size);
 void ff_diff_pixels_mmx(int16_t *block, const uint8_t *s1, const uint8_t *s2,
                         int stride);
-int ff_pix_sum16_mmx(uint8_t *pix, int line_size);
-int ff_pix_norm1_mmx(uint8_t *pix, int line_size);
 
 #if HAVE_INLINE_ASM
 
@@ -831,8 +829,6 @@ av_cold void ff_dsputilenc_init_mmx(DSPContext *c, AVCodecContext *avctx,
         if (!high_bit_depth)
             c->get_pixels = ff_get_pixels_mmx;
         c->diff_pixels = ff_diff_pixels_mmx;
-        c->pix_sum     = ff_pix_sum16_mmx;
-        c->pix_norm1   = ff_pix_norm1_mmx;
     }
 
     if (EXTERNAL_SSE2(cpu_flags))
diff --git a/libavcodec/x86/mpegvideoencdsp.asm b/libavcodec/x86/mpegvideoencdsp.asm
new file mode 100644
index 0000000000..9326ee776d
--- /dev/null
+++ b/libavcodec/x86/mpegvideoencdsp.asm
@@ -0,0 +1,95 @@
+;*****************************************************************************
+;* SIMD-optimized MPEG encoding functions
+;*****************************************************************************
+;* Copyright (c) 2000, 2001 Fabrice Bellard
+;* Copyright (c) 2002-2004 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
+;*****************************************************************************
+
+%include "libavutil/x86/x86util.asm"
+
+SECTION .text
+
+INIT_MMX mmx
+; int ff_pix_sum16_mmx(uint8_t *pix, int line_size)
+cglobal pix_sum16, 2, 3
+    movsxdifnidn r1, r1d
+    mov          r2, r1
+    neg          r2
+    shl          r2, 4
+    sub          r0, r2
+    pxor         m7, m7
+    pxor         m6, m6
+.loop:
+    mova         m0, [r0+r2+0]
+    mova         m1, [r0+r2+0]
+    mova         m2, [r0+r2+8]
+    mova         m3, [r0+r2+8]
+    punpcklbw    m0, m7
+    punpckhbw    m1, m7
+    punpcklbw    m2, m7
+    punpckhbw    m3, m7
+    paddw        m1, m0
+    paddw        m3, m2
+    paddw        m3, m1
+    paddw        m6, m3
+    add          r2, r1
+    js .loop
+    mova         m5, m6
+    psrlq        m6, 32
+    paddw        m6, m5
+    mova         m5, m6
+    psrlq        m6, 16
+    paddw        m6, m5
+    movd        eax, m6
+    and         eax, 0xffff
+    RET
+
+INIT_MMX mmx
+; int ff_pix_norm1_mmx(uint8_t *pix, int line_size)
+cglobal pix_norm1, 2, 4
+    movsxdifnidn r1, r1d
+    mov          r2, 16
+    pxor         m0, m0
+    pxor         m7, m7
+.loop:
+    mova         m2, [r0+0]
+    mova         m3, [r0+8]
+    mova         m1, m2
+    punpckhbw    m1, m0
+    punpcklbw    m2, m0
+    mova         m4, m3
+    punpckhbw    m3, m0
+    punpcklbw    m4, m0
+    pmaddwd      m1, m1
+    pmaddwd      m2, m2
+    pmaddwd      m3, m3
+    pmaddwd      m4, m4
+    paddd        m2, m1
+    paddd        m4, m3
+    paddd        m7, m2
+    add          r0, r1
+    paddd        m7, m4
+    dec r2
+    jne .loop
+    mova         m1, m7
+    psrlq        m7, 32
+    paddd        m1, m7
+    movd        eax, m1
+    RET
+
diff --git a/libavcodec/x86/mpegvideoencdsp_init.c b/libavcodec/x86/mpegvideoencdsp_init.c
index db2c37f43e..4ef2f34961 100644
--- a/libavcodec/x86/mpegvideoencdsp_init.c
+++ b/libavcodec/x86/mpegvideoencdsp_init.c
@@ -22,6 +22,9 @@
 #include "libavcodec/avcodec.h"
 #include "libavcodec/mpegvideoencdsp.h"
 
+int ff_pix_sum16_mmx(uint8_t *pix, int line_size);
+int ff_pix_norm1_mmx(uint8_t *pix, int line_size);
+
 #if HAVE_INLINE_ASM
 
 #define PHADDD(a, t)                            \
@@ -95,9 +98,15 @@
 av_cold void ff_mpegvideoencdsp_init_x86(MpegvideoEncDSPContext *c,
                                          AVCodecContext *avctx)
 {
-#if HAVE_INLINE_ASM
     int cpu_flags = av_get_cpu_flags();
 
+    if (EXTERNAL_MMX(cpu_flags)) {
+        c->pix_sum   = ff_pix_sum16_mmx;
+        c->pix_norm1 = ff_pix_norm1_mmx;
+    }
+
+#if HAVE_INLINE_ASM
+
     if (INLINE_MMX(cpu_flags)) {
         if (!(avctx->flags & CODEC_FLAG_BITEXACT)) {
             c->try_8x8basis = try_8x8basis_mmx;