Merge remote-tracking branch 'qatar/master'

* qatar/master:
  float_dsp: ppc: add a separate header for Altivec function prototypes
  ARM: fix float_dsp breakage from d5a7229
  Add a float DSP framework to libavutil
  PPC: Move types_altivec.h and util_altivec.h from libavcodec to libavutil
  ARM: Move asm.S from libavcodec to libavutil
  vc1dsp: mark put/avg_vc1_mspel_mc() always_inline

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

15 files changed, 23 insertions, 204 deletions

diff --git a/libavcodec/ppc/dsputil_altivec.c b/libavcodec/ppc/dsputil_altivec.c
index de8b85891b..5f14f1dea3 100644
--- a/libavcodec/ppc/dsputil_altivec.c
+++ b/libavcodec/ppc/dsputil_altivec.c
@@ -24,9 +24,9 @@
 #if HAVE_ALTIVEC_H
 #include <altivec.h>
 #endif
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
-#include "util_altivec.h"
-#include "types_altivec.h"
 #include "dsputil_altivec.h"
 
 static int sad16_x2_altivec(void *v, uint8_t *pix1, uint8_t *pix2, int line_size, int h)
diff --git a/libavcodec/ppc/fft_altivec.c b/libavcodec/ppc/fft_altivec.c
index e171665b37..651ee2680f 100644
--- a/libavcodec/ppc/fft_altivec.c
+++ b/libavcodec/ppc/fft_altivec.c
@@ -19,9 +19,10 @@
  * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
+
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/fft.h"
-#include "util_altivec.h"
-#include "types_altivec.h"
 
 /**
  * Do a complex FFT with the parameters defined in ff_fft_init(). The
diff --git a/libavcodec/ppc/float_altivec.c b/libavcodec/ppc/float_altivec.c
index a724a46d5c..e7b04ece6a 100644
--- a/libavcodec/ppc/float_altivec.c
+++ b/libavcodec/ppc/float_altivec.c
@@ -18,25 +18,10 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
 
 #include "dsputil_altivec.h"
-#include "util_altivec.h"
-
-static void vector_fmul_altivec(float *dst, const float *src0, const float *src1, int len)
-{
-    int i;
-    vector float d0, d1, s, zero = (vector float)vec_splat_u32(0);
-    for(i=0; i<len-7; i+=8) {
-        d0 = vec_ld(0, src0+i);
-        s = vec_ld(0, src1+i);
-        d1 = vec_ld(16, src0+i);
-        d0 = vec_madd(d0, s, zero);
-        d1 = vec_madd(d1, vec_ld(16,src1+i), zero);
-        vec_st(d0, 0, dst+i);
-        vec_st(d1, 16, dst+i);
-    }
-}
 
 static void vector_fmul_reverse_altivec(float *dst, const float *src0,
                                         const float *src1, int len)
@@ -124,7 +109,6 @@ static void vector_fmul_window_altivec(float *dst, const float *src0, const floa
 
 void ff_float_init_altivec(DSPContext* c, AVCodecContext *avctx)
 {
-    c->vector_fmul = vector_fmul_altivec;
     c->vector_fmul_reverse = vector_fmul_reverse_altivec;
     c->vector_fmul_add = vector_fmul_add_altivec;
     if(!(avctx->flags & CODEC_FLAG_BITEXACT)) {
diff --git a/libavcodec/ppc/fmtconvert_altivec.c b/libavcodec/ppc/fmtconvert_altivec.c
index 30de0e637f..6f24199023 100644
--- a/libavcodec/ppc/fmtconvert_altivec.c
+++ b/libavcodec/ppc/fmtconvert_altivec.c
@@ -20,8 +20,8 @@
 
 #include "libavcodec/fmtconvert.h"
 
+#include "libavutil/ppc/util_altivec.h"
 #include "dsputil_altivec.h"
-#include "util_altivec.h"
 
 static void int32_to_float_fmul_scalar_altivec(float *dst, const int *src, float mul, int len)
 {
diff --git a/libavcodec/ppc/gmc_altivec.c b/libavcodec/ppc/gmc_altivec.c
index 1341594ee5..fadf498f09 100644
--- a/libavcodec/ppc/gmc_altivec.c
+++ b/libavcodec/ppc/gmc_altivec.c
@@ -20,9 +20,9 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
-#include "util_altivec.h"
-#include "types_altivec.h"
 #include "dsputil_altivec.h"
 
 /*
diff --git a/libavcodec/ppc/h264_altivec.c b/libavcodec/ppc/h264_altivec.c
index 4589c4f96d..3e7edc74df 100644
--- a/libavcodec/ppc/h264_altivec.c
+++ b/libavcodec/ppc/h264_altivec.c
@@ -19,13 +19,13 @@
  */
 
 #include "libavutil/cpu.h"
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
 #include "libavcodec/h264data.h"
 #include "libavcodec/h264dsp.h"
 
 #include "dsputil_altivec.h"
-#include "util_altivec.h"
-#include "types_altivec.h"
 
 #define PUT_OP_U8_ALTIVEC(d, s, dst) d = s
 #define AVG_OP_U8_ALTIVEC(d, s, dst) d = vec_avg(dst, s)
diff --git a/libavcodec/ppc/idct_altivec.c b/libavcodec/ppc/idct_altivec.c
index dbf234536c..9a8b91740e 100644
--- a/libavcodec/ppc/idct_altivec.c
+++ b/libavcodec/ppc/idct_altivec.c
@@ -41,8 +41,8 @@
 #if HAVE_ALTIVEC_H
 #include <altivec.h>
 #endif
+#include "libavutil/ppc/types_altivec.h"
 #include "libavcodec/dsputil.h"
-#include "types_altivec.h"
 #include "dsputil_altivec.h"
 
 #define IDCT_HALF                                       \
diff --git a/libavcodec/ppc/int_altivec.c b/libavcodec/ppc/int_altivec.c
index 57e63da9f0..5a73876ad3 100644
--- a/libavcodec/ppc/int_altivec.c
+++ b/libavcodec/ppc/int_altivec.c
@@ -28,12 +28,11 @@
 #include <altivec.h>
 #endif
 
+#include "libavutil/ppc/types_altivec.h"
 #include "libavcodec/dsputil.h"
 
 #include "dsputil_altivec.h"
 
-#include "types_altivec.h"
-
 static int ssd_int8_vs_int16_altivec(const int8_t *pix1, const int16_t *pix2,
                                      int size) {
     int i, size16;
diff --git a/libavcodec/ppc/mpegaudiodec_altivec.c b/libavcodec/ppc/mpegaudiodec_altivec.c
index 2de5dd133a..7d829b86f4 100644
--- a/libavcodec/ppc/mpegaudiodec_altivec.c
+++ b/libavcodec/ppc/mpegaudiodec_altivec.c
@@ -20,7 +20,7 @@
  */
 
 #include "dsputil_altivec.h"
-#include "util_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
 #include "libavcodec/mpegaudiodsp.h"
 
diff --git a/libavcodec/ppc/mpegvideo_altivec.c b/libavcodec/ppc/mpegvideo_altivec.c
index 62cd8e0adc..bbc102bc21 100644
--- a/libavcodec/ppc/mpegvideo_altivec.c
+++ b/libavcodec/ppc/mpegvideo_altivec.c
@@ -23,12 +23,13 @@
 
 #include <stdlib.h>
 #include <stdio.h>
+
 #include "libavutil/cpu.h"
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
 #include "libavcodec/mpegvideo.h"
 
-#include "util_altivec.h"
-#include "types_altivec.h"
 #include "dsputil_altivec.h"
 
 /* AltiVec version of dct_unquantize_h263
diff --git a/libavcodec/ppc/types_altivec.h b/libavcodec/ppc/types_altivec.h
deleted file mode 100644
index 36b6e1f388..0000000000
--- a/libavcodec/ppc/types_altivec.h
+++ /dev/null
@@ -1,47 +0,0 @@
-/*
- * Copyright (c) 2006 Guillaume Poirier <gpoirier@mplayerhq.hu>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg 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.
- *
- * FFmpeg 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 FFmpeg; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-#ifndef AVCODEC_PPC_TYPES_ALTIVEC_H
-#define AVCODEC_PPC_TYPES_ALTIVEC_H
-
-/***********************************************************************
- * Vector types
- **********************************************************************/
-#define vec_u8  vector unsigned char
-#define vec_s8  vector signed char
-#define vec_u16 vector unsigned short
-#define vec_s16 vector signed short
-#define vec_u32 vector unsigned int
-#define vec_s32 vector signed int
-#define vec_f   vector float
-
-/***********************************************************************
- * Null vector
- **********************************************************************/
-#define LOAD_ZERO const vec_u8 zerov = vec_splat_u8( 0 )
-
-#define zero_u8v  (vec_u8)  zerov
-#define zero_s8v  (vec_s8)  zerov
-#define zero_u16v (vec_u16) zerov
-#define zero_s16v (vec_s16) zerov
-#define zero_u32v (vec_u32) zerov
-#define zero_s32v (vec_s32) zerov
-
-#endif /* AVCODEC_PPC_TYPES_ALTIVEC_H */
diff --git a/libavcodec/ppc/util_altivec.h b/libavcodec/ppc/util_altivec.h
deleted file mode 100644
index 6800348570..0000000000
--- a/libavcodec/ppc/util_altivec.h
+++ /dev/null
@@ -1,118 +0,0 @@
-/*
- * This file is part of FFmpeg.
- *
- * FFmpeg 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.
- *
- * FFmpeg 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 FFmpeg; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-/**
- * @file
- * Contains misc utility macros and inline functions
- */
-
-#ifndef AVCODEC_PPC_UTIL_ALTIVEC_H
-#define AVCODEC_PPC_UTIL_ALTIVEC_H
-
-#include <stdint.h>
-
-#include "config.h"
-
-#if HAVE_ALTIVEC_H
-#include <altivec.h>
-#endif
-
-#include "types_altivec.h"
-
-// used to build registers permutation vectors (vcprm)
-// the 's' are for words in the _s_econd vector
-#define WORD_0 0x00,0x01,0x02,0x03
-#define WORD_1 0x04,0x05,0x06,0x07
-#define WORD_2 0x08,0x09,0x0a,0x0b
-#define WORD_3 0x0c,0x0d,0x0e,0x0f
-#define WORD_s0 0x10,0x11,0x12,0x13
-#define WORD_s1 0x14,0x15,0x16,0x17
-#define WORD_s2 0x18,0x19,0x1a,0x1b
-#define WORD_s3 0x1c,0x1d,0x1e,0x1f
-
-#define vcprm(a,b,c,d) (const vector unsigned char){WORD_ ## a, WORD_ ## b, WORD_ ## c, WORD_ ## d}
-#define vcii(a,b,c,d) (const vector float){FLOAT_ ## a, FLOAT_ ## b, FLOAT_ ## c, FLOAT_ ## d}
-
-// vcprmle is used to keep the same index as in the SSE version.
-// it's the same as vcprm, with the index inversed
-// ('le' is Little Endian)
-#define vcprmle(a,b,c,d) vcprm(d,c,b,a)
-
-// used to build inverse/identity vectors (vcii)
-// n is _n_egative, p is _p_ositive
-#define FLOAT_n -1.
-#define FLOAT_p 1.
-
-
-// Transpose 8x8 matrix of 16-bit elements (in-place)
-#define TRANSPOSE8(a,b,c,d,e,f,g,h) \
-do { \
-    vector signed short A1, B1, C1, D1, E1, F1, G1, H1; \
-    vector signed short A2, B2, C2, D2, E2, F2, G2, H2; \
- \
-    A1 = vec_mergeh (a, e); \
-    B1 = vec_mergel (a, e); \
-    C1 = vec_mergeh (b, f); \
-    D1 = vec_mergel (b, f); \
-    E1 = vec_mergeh (c, g); \
-    F1 = vec_mergel (c, g); \
-    G1 = vec_mergeh (d, h); \
-    H1 = vec_mergel (d, h); \
- \
-    A2 = vec_mergeh (A1, E1); \
-    B2 = vec_mergel (A1, E1); \
-    C2 = vec_mergeh (B1, F1); \
-    D2 = vec_mergel (B1, F1); \
-    E2 = vec_mergeh (C1, G1); \
-    F2 = vec_mergel (C1, G1); \
-    G2 = vec_mergeh (D1, H1); \
-    H2 = vec_mergel (D1, H1); \
- \
-    a = vec_mergeh (A2, E2); \
-    b = vec_mergel (A2, E2); \
-    c = vec_mergeh (B2, F2); \
-    d = vec_mergel (B2, F2); \
-    e = vec_mergeh (C2, G2); \
-    f = vec_mergel (C2, G2); \
-    g = vec_mergeh (D2, H2); \
-    h = vec_mergel (D2, H2); \
-} while (0)
-
-
-/** @brief loads unaligned vector @a *src with offset @a offset
-    and returns it */
-static inline vector unsigned char unaligned_load(int offset, uint8_t *src)
-{
-    register vector unsigned char first = vec_ld(offset, src);
-    register vector unsigned char second = vec_ld(offset+15, src);
-    register vector unsigned char mask = vec_lvsl(offset, src);
-    return vec_perm(first, second, mask);
-}
-
-/**
- * loads vector known misalignment
- * @param perm_vec the align permute vector to combine the two loads from lvsl
- */
-static inline vec_u8 load_with_perm_vec(int offset, uint8_t *src, vec_u8 perm_vec)
-{
-    vec_u8 a = vec_ld(offset, src);
-    vec_u8 b = vec_ld(offset+15, src);
-    return vec_perm(a, b, perm_vec);
-}
-
-#endif /* AVCODEC_PPC_UTIL_ALTIVEC_H */
diff --git a/libavcodec/ppc/vc1dsp_altivec.c b/libavcodec/ppc/vc1dsp_altivec.c
index 69670619da..bf76adb359 100644
--- a/libavcodec/ppc/vc1dsp_altivec.c
+++ b/libavcodec/ppc/vc1dsp_altivec.c
@@ -19,12 +19,11 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
 #include "libavcodec/vc1dsp.h"
 
-#include "util_altivec.h"
-#include "dsputil_altivec.h"
-
 // main steps of 8x8 transform
 #define STEP8(s0, s1, s2, s3, s4, s5, s6, s7, vec_rnd) \
 do { \
diff --git a/libavcodec/ppc/vp3dsp_altivec.c b/libavcodec/ppc/vp3dsp_altivec.c
index b0509d8b64..950e5c7419 100644
--- a/libavcodec/ppc/vp3dsp_altivec.c
+++ b/libavcodec/ppc/vp3dsp_altivec.c
@@ -18,9 +18,9 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/dsputil.h"
-#include "util_altivec.h"
-#include "types_altivec.h"
 #include "dsputil_altivec.h"
 
 static const vec_s16 constants =
diff --git a/libavcodec/ppc/vp8dsp_altivec.c b/libavcodec/ppc/vp8dsp_altivec.c
index 7f7d3bd35c..4c86ca5f74 100644
--- a/libavcodec/ppc/vp8dsp_altivec.c
+++ b/libavcodec/ppc/vp8dsp_altivec.c
@@ -21,10 +21,10 @@
  */
 
 #include "libavutil/cpu.h"
+#include "libavutil/ppc/types_altivec.h"
+#include "libavutil/ppc/util_altivec.h"
 #include "libavcodec/vp8dsp.h"
 #include "dsputil_altivec.h"
-#include "types_altivec.h"
-#include "util_altivec.h"
 
 #define REPT4(...) { __VA_ARGS__, __VA_ARGS__, __VA_ARGS__, __VA_ARGS__ }