* code with new PAVGB for MMX only CPU splited into separate file

  and being compiled in the same way as _avg.h
* PAVG_MMX macros accept also output parameter
* implemented faster put_pixels_xy2, but it has slightly smaller precission.
  But there is not visible difference in the image quality - might be eventualy
  easily switched back (#if 0 #endif)- please check

Originally committed as revision 624 to svn://svn.ffmpeg.org/ffmpeg/trunk

2 files changed, 171 insertions, 218 deletions

diff --git a/libavcodec/i386/dsputil_mmx.c b/libavcodec/i386/dsputil_mmx.c
index 4cedb1fb28..58b0c23bbc 100644
--- a/libavcodec/i386/dsputil_mmx.c
+++ b/libavcodec/i386/dsputil_mmx.c
@@ -91,22 +91,45 @@ static const uint64_t mm_wtwo __attribute__ ((aligned(8))) = 0x0002000200020002U
 // using mm6 as temporary and for the output result
 // first argument is unmodifed and second is trashed
 // mm7 is supposed to contain 0xfefefefefefefefe
-#define PAVG_MMX_NO_RND(rega, regb) \
-    "movq " #rega ", %%mm6	\n\t"\
-    "pand " #regb ", %%mm6	\n\t"\
+#define PAVGB_MMX_NO_RND(rega, regb, regr) \
+    "movq " #rega ", " #regr "	\n\t"\
+    "pand " #regb ", " #regr "	\n\t"\
     "pxor " #rega ", " #regb "	\n\t"\
     "pand %%mm7, " #regb "	\n\t"\
     "psrlq $1, " #regb " 	\n\t"\
-    "paddb " #regb ", %%mm6	\n\t"
+    "paddb " #regb ", " #regr "	\n\t"
 
-#define PAVG_MMX(rega, regb) \
-    "movq " #rega ", %%mm6	\n\t"\
-    "por  " #regb ", %%mm6	\n\t"\
+#define PAVGB_MMX(rega, regb, regr) \
+    "movq " #rega ", " #regr "	\n\t"\
+    "por  " #regb ", " #regr "	\n\t"\
     "pxor " #rega ", " #regb "	\n\t"\
     "pand %%mm7, " #regb "     	\n\t"\
     "psrlq $1, " #regb "	\n\t"\
-    "psubb " #regb ", %%mm6 	\n\t"
+    "psubb " #regb ", " #regr "	\n\t"
 
+/***********************************/
+/* MMX no rounding */
+#define DEF(x, y) x ## _no_rnd_ ## y ##_mmx
+
+#define PAVGB(a, b)	PAVGB_MMX_NO_RND(a, b, %%mm6)
+#define PAVGBR(a, b, c)	PAVGB_MMX_NO_RND(a, b, c)
+#include "dsputil_mmx_rnd.h"
+
+#undef DEF
+#undef PAVGB
+#undef PAVGBR
+/***********************************/
+/* MMX rounding */
+
+#define DEF(x, y) x ## _ ## y ##_mmx
+
+#define PAVGB(a, b)	PAVGB_MMX(a, b, %%mm6)
+#define PAVGBR(a, b, c)	PAVGB_MMX(a, b, c)
+#include "dsputil_mmx_rnd.h"
+
+#undef DEF
+#undef PAVGB
+#undef PAVGBR
 
 /***********************************/
 /* 3Dnow specific */
@@ -316,152 +339,7 @@ static void put_pixels_mmx(UINT8 *block, const UINT8 *pixels, int line_size, int
 	);
 }
 
-// will have to be check if it's better to have bigger
-// unrolled code also on Celerons - for now  yes
-#define LONG_UNROLL 1
-static void put_pixels_x2_mmx(UINT8 *block, const UINT8 *pixels, int line_size, int h)
-{
-#if 0
-  UINT8 *p;
-  const UINT8 *pix;
-  p = block;
-  pix = pixels;
-  MOVQ_ZERO(mm7);
-  MOVQ_WONE(mm4);
-  JUMPALIGN();
-  do {
-      __asm __volatile(
-	"movq	%1, %%mm0\n\t"
-	"movq	1%1, %%mm1\n\t"
-	"movq	%%mm0, %%mm2\n\t"
-	"movq	%%mm1, %%mm3\n\t"
-	"punpcklbw %%mm7, %%mm0\n\t"
-	"punpcklbw %%mm7, %%mm1\n\t"
-	"punpckhbw %%mm7, %%mm2\n\t"
-	"punpckhbw %%mm7, %%mm3\n\t"
-	"paddusw %%mm1, %%mm0\n\t"
-	"paddusw %%mm3, %%mm2\n\t"
-	"paddusw %%mm4, %%mm0\n\t"
-	"paddusw %%mm4, %%mm2\n\t"
-	"psrlw	$1, %%mm0\n\t"
-	"psrlw	$1, %%mm2\n\t"
-	"packuswb  %%mm2, %%mm0\n\t"
-	"movq	%%mm0, %0\n\t"
-	:"=m"(*p)
-	:"m"(*pix)
-		       :"memory");
-   pix += line_size; p += line_size;
-  } while (--h);
-#else
-  __asm __volatile(
-  	MOVQ_BFE(%%mm7)
-	"lea (%3, %3), %%eax	\n\t"
-	".balign 8     		\n\t"
-	"1:			\n\t"
-	"movq (%1), %%mm0	\n\t"
-	"movq (%1, %3), %%mm2	\n\t"
-	"movq 1(%1), %%mm1	\n\t"
-	"movq 1(%1, %3), %%mm3	\n\t"
-	PAVG_MMX(%%mm0, %%mm1)
-	"movq %%mm6, (%2)	\n\t"
-	PAVG_MMX(%%mm2, %%mm3)
-	"movq %%mm6, (%2, %3)	\n\t"
-	"addl %%eax, %1		\n\t"
-	"addl %%eax, %2		\n\t"
-#if LONG_UNROLL
-	"movq (%1), %%mm0	\n\t"
-	"movq (%1, %3), %%mm2	\n\t"
-	"movq 1(%1), %%mm1	\n\t"
-	"movq 1(%1, %3), %%mm3	\n\t"
-	PAVG_MMX(%%mm0, %%mm1)
-	"movq %%mm6, (%2)	\n\t"
-	PAVG_MMX(%%mm2, %%mm3)
-	"movq %%mm6, (%2, %3)	\n\t"
-	"addl %%eax, %1		\n\t"
-	"addl %%eax, %2		\n\t"
-	"subl $4, %0		\n\t"
-#else
-	"subl $2, %0		\n\t"
-#endif
-	"jnz 1b			\n\t"
-	:"+g"(h), "+S"(pixels), "+D"(block)
-	:"r"(line_size)
-	:"eax", "memory");
-#endif
-}
-
-static void put_pixels_y2_mmx(UINT8 *block, const UINT8 *pixels, int line_size, int h)
-{
 #if 0
-  UINT8 *p;
-  const UINT8 *pix;
-  p = block;
-  pix = pixels;
-  MOVQ_ZERO(mm7);
-  MOVQ_WONE(mm4);
-  JUMPALIGN();
-  do {
-    __asm __volatile(
-	"movq	%1, %%mm0\n\t"
-	"movq	%2, %%mm1\n\t"
-	"movq	%%mm0, %%mm2\n\t"
-	"movq	%%mm1, %%mm3\n\t"
-	"punpcklbw %%mm7, %%mm0\n\t"
-	"punpcklbw %%mm7, %%mm1\n\t"
-	"punpckhbw %%mm7, %%mm2\n\t"
-	"punpckhbw %%mm7, %%mm3\n\t"
-	"paddusw %%mm1, %%mm0\n\t"
-	"paddusw %%mm3, %%mm2\n\t"
-	"paddusw %%mm4, %%mm0\n\t"
-	"paddusw %%mm4, %%mm2\n\t"
-	"psrlw	$1, %%mm0\n\t"
-	"psrlw	$1, %%mm2\n\t"
-	"packuswb  %%mm2, %%mm0\n\t"
-	"movq	%%mm0, %0\n\t"
-	:"=m"(*p)
-	:"m"(*pix),
-	 "m"(*(pix+line_size))
-	:"memory");
-   pix += line_size;
-   p += line_size;
-  } while (--h);
-#else
-  __asm __volatile(
-  	MOVQ_BFE(%%mm7)
-	"lea (%3, %3), %%eax	\n\t"
-	"movq (%1), %%mm0	\n\t"
-	".balign 8     		\n\t"
-	"1:			\n\t"
-	"movq (%1, %3), %%mm1	\n\t"
-	"movq (%1, %%eax),%%mm2	\n\t"
-	PAVG_MMX(%%mm1, %%mm0)
-	"movq %%mm6, (%2)	\n\t"
-	PAVG_MMX(%%mm2, %%mm1)
-	"movq %%mm6, (%2, %3)	\n\t"
-	"addl %%eax, %1		\n\t"
-	"addl %%eax, %2		\n\t"
-#ifdef LONG_UNROLL
-	"movq (%1, %3), %%mm1	\n\t"
-	"movq (%1, %%eax),%%mm0	\n\t"
-	PAVG_MMX(%%mm1, %%mm2)
-	"movq %%mm6, (%2)	\n\t"
-	PAVG_MMX(%%mm0, %%mm1)
-	"movq %%mm6, (%2, %3)	\n\t"
-	"addl %%eax, %1		\n\t"
-	"addl %%eax, %2		\n\t"
-	"subl $4, %0		\n\t"
-#else
-	"subl $2, %0		\n\t"
-#endif
-	"jnz 1b			\n\t"
-	:"+g"(h), "+S"(pixels), "+D"(block)
-	:"r"(line_size)
-	:"eax", "memory");
-#endif
-
-
-}
-
 static void put_pixels_xy2_mmx(UINT8 *block, const UINT8 *pixels, int line_size, int h)
 {
   UINT8 *p;
@@ -510,70 +388,6 @@ static void put_pixels_xy2_mmx(UINT8 *block, const UINT8 *pixels, int line_size,
   } while(--h);
 }
 
-static void   put_no_rnd_pixels_x2_mmx( UINT8  *block, const UINT8 *pixels, int line_size, int h)
-{
-  UINT8  *p;
-  const UINT8 *pix;
-  p = block;
-  pix = pixels;
-  MOVQ_ZERO(mm7);
-  do {
-    __asm __volatile(
-	"movq	%1, %%mm0\n\t"
-	"movq	1%1, %%mm1\n\t"
-	"movq	%%mm0, %%mm2\n\t"
-	"movq	%%mm1, %%mm3\n\t"
-	"punpcklbw %%mm7, %%mm0\n\t"
-	"punpcklbw %%mm7, %%mm1\n\t"
-	"punpckhbw %%mm7, %%mm2\n\t"
-	"punpckhbw %%mm7, %%mm3\n\t"
-	"paddusw %%mm1, %%mm0\n\t"
-	"paddusw %%mm3, %%mm2\n\t"
-	"psrlw	$1, %%mm0\n\t"
-	"psrlw	$1, %%mm2\n\t"
-	"packuswb  %%mm2, %%mm0\n\t"
-	"movq	%%mm0, %0\n\t"
-	:"=m"(*p)
-	:"m"(*pix)
-	:"memory");
-   pix += line_size;
-   p +=   line_size;
-  } while (--h);
-}
-
-static void put_no_rnd_pixels_y2_mmx( UINT8  *block, const UINT8 *pixels, int line_size, int h)
-{
-  UINT8  *p;
-  const UINT8 *pix;
-  p = block;
-  pix = pixels;
-  MOVQ_ZERO(mm7);
-  JUMPALIGN();
-  do {
-    __asm __volatile(
-	"movq	%1, %%mm0\n\t"
-	"movq	%2, %%mm1\n\t"
-	"movq	%%mm0, %%mm2\n\t"
-	"movq	%%mm1, %%mm3\n\t"
-	"punpcklbw %%mm7, %%mm0\n\t"
-	"punpcklbw %%mm7, %%mm1\n\t"
-	"punpckhbw %%mm7, %%mm2\n\t"
-	"punpckhbw %%mm7, %%mm3\n\t"
-	"paddusw %%mm1, %%mm0\n\t"
-	"paddusw %%mm3, %%mm2\n\t"
-	"psrlw	$1, %%mm0\n\t"
-	"psrlw	$1, %%mm2\n\t"
-	"packuswb  %%mm2, %%mm0\n\t"
-	"movq	%%mm0, %0\n\t"
-	:"=m"(*p)
-	:"m"(*pix),
-	 "m"(*(pix+line_size))
-	:"memory");
-   pix += line_size;
-   p +=   line_size;
-  } while(--h);
-}
-
 static void   put_no_rnd_pixels_xy2_mmx( UINT8  *block, const UINT8 *pixels, int line_size, int h)
 {
   UINT8  *p;
@@ -621,7 +435,7 @@ static void   put_no_rnd_pixels_xy2_mmx( UINT8  *block, const UINT8 *pixels, int
    p +=   line_size;
   } while(--h);
 }
-
+#endif
 static void avg_pixels_mmx(UINT8 *block, const UINT8 *pixels, int line_size, int h)
 {
   UINT8  *p;
diff --git a/libavcodec/i386/dsputil_mmx_rnd.h b/libavcodec/i386/dsputil_mmx_rnd.h
new file mode 100644
index 0000000000..787f706dc2
--- /dev/null
+++ b/libavcodec/i386/dsputil_mmx_rnd.h
@@ -0,0 +1,139 @@
+/*
+ * DSP utils mmx functions are compiled twice for rnd/no_rnd
+ * Copyright (c) 2000, 2001 Fabrice Bellard.
+ *
+ * This library 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 of the License, or (at your option) any later version.
+ *
+ * This library 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 this library; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ *
+ * MMX optimization by Nick Kurshev <nickols_k@mail.ru>
+ * mostly rewritten by Michael Niedermayer <michaelni@gmx.at>
+ * and improved by Zdenek Kabelac <kabi@users.sf.net>
+ */
+
+// will have to be check if it's better to have bigger
+// unrolled code also on Celerons - for now  yes
+#define LONG_UNROLL 1
+
+// put_pixels
+static void DEF(put, pixels_x2)(UINT8 *block, const UINT8 *pixels, int line_size, int h)
+{
+    __asm __volatile(
+	MOVQ_BFE(%%mm7)
+	"lea	(%3, %3), %%eax		\n\t"
+	".balign 8			\n\t"
+	"1:				\n\t"
+	"movq	(%1), %%mm0		\n\t"
+	"movq	(%1, %3), %%mm2		\n\t"
+	"movq	1(%1), %%mm1		\n\t"
+	"movq	1(%1, %3), %%mm3	\n\t"
+	PAVGB(%%mm0, %%mm1)
+	"movq	%%mm6, (%2)		\n\t"
+	PAVGB(%%mm2, %%mm3)
+	"movq	%%mm6, (%2, %3)		\n\t"
+	"addl	%%eax, %1		\n\t"
+	"addl	%%eax, %2		\n\t"
+#if LONG_UNROLL
+	"movq	(%1), %%mm0		\n\t"
+	"movq	(%1, %3), %%mm2		\n\t"
+	"movq	1(%1), %%mm1		\n\t"
+	"movq	1(%1, %3), %%mm3	\n\t"
+	PAVGB(%%mm0, %%mm1)
+	"movq	%%mm6, (%2)		\n\t"
+	PAVGB(%%mm2, %%mm3)
+	"movq	%%mm6, (%2, %3)		\n\t"
+	"addl	%%eax, %1		\n\t"
+	"addl	%%eax, %2		\n\t"
+	"subl	$4, %0			\n\t"
+#else
+	"subl	$2, %0			\n\t"
+#endif
+	"jnz	1b			\n\t"
+	:"+g"(h), "+S"(pixels), "+D"(block)
+	:"r"(line_size)
+	:"eax", "memory");
+}
+
+static void DEF(put, pixels_y2)(UINT8 *block, const UINT8 *pixels, int line_size, int h)
+{
+  __asm __volatile(
+	MOVQ_BFE(%%mm7)
+	"lea (%3, %3), %%eax		\n\t"
+	"movq (%1), %%mm0		\n\t"
+	".balign 8			\n\t"
+	"1:				\n\t"
+	"movq	(%1, %3), %%mm1		\n\t"
+	"movq	(%1, %%eax),%%mm2	\n\t"
+	PAVGB(%%mm1, %%mm0)
+	"movq	%%mm6, (%2)		\n\t"
+	PAVGB(%%mm2, %%mm1)
+	"movq	%%mm6, (%2, %3)		\n\t"
+	"addl	%%eax, %1		\n\t"
+	"addl	%%eax, %2		\n\t"
+#ifdef LONG_UNROLL
+	"movq	(%1, %3), %%mm1		\n\t"
+	"movq	(%1, %%eax),%%mm0	\n\t"
+	PAVGB(%%mm1, %%mm2)
+	"movq	%%mm6, (%2)		\n\t"
+	PAVGB(%%mm0, %%mm1)
+	"movq	%%mm6, (%2, %3)		\n\t"
+	"addl	%%eax, %1		\n\t"
+	"addl	%%eax, %2		\n\t"
+	"subl	$4, %0			\n\t"
+#else
+	"subl	$2, %0			\n\t"
+#endif
+	"jnz	1b			\n\t"
+	:"+g"(h), "+S"(pixels), "+D"(block)
+	:"r"(line_size)
+	:"eax", "memory");
+}
+
+// ((a + b)/2 + (c + d)/2)/2
+// not sure if this is properly replacing original code
+static void DEF(put, pixels_xy2)(UINT8 *block, const UINT8 *pixels, int line_size, int h)
+{
+    __asm __volatile(
+	MOVQ_BFE(%%mm7)
+	"lea (%3, %3), %%eax		\n\t"
+	"movq (%1), %%mm0		\n\t"
+	"movq	(%1), %%mm0		\n\t"
+	"movq	1(%1), %%mm1		\n\t"
+	".balign 8			\n\t"
+	"1:				\n\t"
+	"movq	(%1, %3), %%mm2		\n\t"
+	"movq	1(%1, %3), %%mm3	\n\t"
+	PAVGBR(%%mm2, %%mm0, %%mm4)
+	PAVGBR(%%mm3, %%mm1, %%mm5)
+	PAVGB(%%mm4, %%mm5)
+	"movq	%%mm6, (%2)		\n\t"
+
+	"movq	(%1, %%eax), %%mm0	\n\t"
+	"movq	1(%1, %%eax), %%mm1	\n\t"
+	PAVGBR(%%mm0, %%mm2, %%mm4)
+	PAVGBR(%%mm1, %%mm3, %%mm5)
+	PAVGB(%%mm4, %%mm5)
+	"movq	%%mm6, (%2, %3)		\n\t"
+	"addl	%%eax, %1		\n\t"
+	"addl	%%eax, %2		\n\t"
+
+	"subl	$2, %0			\n\t"
+
+	"jnz	1b			\n\t"
+	:"+g"(h), "+S"(pixels), "+D"(block)
+	:"r"(line_size)
+	:"eax", "memory");
+}
+
+// avg_pixels
+