From 402ccf9d4c69ecb1b59127cd34d4303c030f33da Mon Sep 17 00:00:00 2001
From: diego <diego@b3059339-0415-0410-9bf9-f77b7e298cf2>
Date: Sun, 29 Apr 2007 13:39:27 +0000
Subject: cosmetics attack, part II: Remove all tabs and prettyprint/reindent
 the code.

git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@23173 b3059339-0415-0410-9bf9-f77b7e298cf2
---
 libswscale/swscale-example.c  |  306 +--
 libswscale/swscale.c          | 4357 ++++++++++++++++++++---------------------
 libswscale/swscale.h          |   90 +-
 libswscale/swscale_internal.h |  242 +--
 4 files changed, 2497 insertions(+), 2498 deletions(-)

diff --git a/libswscale/swscale-example.c b/libswscale/swscale-example.c
index 803ffa0459..e62faf6d3c 100644
--- a/libswscale/swscale-example.c
+++ b/libswscale/swscale-example.c
@@ -31,129 +31,129 @@
 #include "rgb2rgb.h"
 
 static uint64_t getSSD(uint8_t *src1, uint8_t *src2, int stride1, int stride2, int w, int h){
-	int x,y;
-	uint64_t ssd=0;
+    int x,y;
+    uint64_t ssd=0;
 
 //printf("%d %d\n", w, h);
 
-	for(y=0; y<h; y++){
-		for(x=0; x<w; x++){
-			int d= src1[x + y*stride1] - src2[x + y*stride2];
-			ssd+= d*d;
+    for (y=0; y<h; y++){
+        for (x=0; x<w; x++){
+            int d= src1[x + y*stride1] - src2[x + y*stride2];
+            ssd+= d*d;
 //printf("%d", abs(src1[x + y*stride1] - src2[x + y*stride2])/26 );
-		}
+        }
 //printf("\n");
-	}
-	return ssd;
+    }
+    return ssd;
 }
 
 // test by ref -> src -> dst -> out & compare out against ref
 // ref & out are YV12
 static int doTest(uint8_t *ref[3], int refStride[3], int w, int h, int srcFormat, int dstFormat,
-                   int srcW, int srcH, int dstW, int dstH, int flags){
-	uint8_t *src[3];
-	uint8_t *dst[3];
-	uint8_t *out[3];
-	int srcStride[3], dstStride[3];
-	int i;
-	uint64_t ssdY, ssdU, ssdV;
-	struct SwsContext *srcContext, *dstContext, *outContext;
-	int res;
-
-	res = 0;
-	for(i=0; i<3; i++){
-		// avoid stride % bpp != 0
-		if(srcFormat==PIX_FMT_RGB24 || srcFormat==PIX_FMT_BGR24)
-			srcStride[i]= srcW*3;
-		else
-			srcStride[i]= srcW*4;
-
-		if(dstFormat==PIX_FMT_RGB24 || dstFormat==PIX_FMT_BGR24)
-			dstStride[i]= dstW*3;
-		else
-			dstStride[i]= dstW*4;
-
-		src[i]= (uint8_t*) malloc(srcStride[i]*srcH);
-		dst[i]= (uint8_t*) malloc(dstStride[i]*dstH);
-		out[i]= (uint8_t*) malloc(refStride[i]*h);
-		if ((src[i] == NULL) || (dst[i] == NULL) || (out[i] == NULL)) {
-			perror("Malloc");
-			res = -1;
-
-			goto end;
-		}
-	}
-
-	dstContext = outContext = NULL;
-	srcContext= sws_getContext(w, h, PIX_FMT_YUV420P, srcW, srcH, srcFormat, flags, NULL, NULL, NULL);
-	if (srcContext == NULL) {
-		fprintf(stderr, "Failed to get %s ---> %s\n",
-				sws_format_name(PIX_FMT_YUV420P),
-				sws_format_name(srcFormat));
-		res = -1;
-
-		goto end;
-	}
-	dstContext= sws_getContext(srcW, srcH, srcFormat, dstW, dstH, dstFormat, flags, NULL, NULL, NULL);
-	if (dstContext == NULL) {
-		fprintf(stderr, "Failed to get %s ---> %s\n",
-				sws_format_name(srcFormat),
-				sws_format_name(dstFormat));
-		res = -1;
-
-		goto end;
-	}
-	outContext= sws_getContext(dstW, dstH, dstFormat, w, h, PIX_FMT_YUV420P, flags, NULL, NULL, NULL);
-	if (outContext == NULL) {
-		fprintf(stderr, "Failed to get %s ---> %s\n",
-				sws_format_name(dstFormat),
-				sws_format_name(PIX_FMT_YUV420P));
-		res = -1;
-
-		goto end;
-	}
-//	printf("test %X %X %X -> %X %X %X\n", (int)ref[0], (int)ref[1], (int)ref[2],
-//		(int)src[0], (int)src[1], (int)src[2]);
-
-	sws_scale(srcContext, ref, refStride, 0, h   , src, srcStride);
-	sws_scale(dstContext, src, srcStride, 0, srcH, dst, dstStride);
-	sws_scale(outContext, dst, dstStride, 0, dstH, out, refStride);
+                  int srcW, int srcH, int dstW, int dstH, int flags){
+    uint8_t *src[3];
+    uint8_t *dst[3];
+    uint8_t *out[3];
+    int srcStride[3], dstStride[3];
+    int i;
+    uint64_t ssdY, ssdU, ssdV;
+    struct SwsContext *srcContext, *dstContext, *outContext;
+    int res;
+
+    res = 0;
+    for (i=0; i<3; i++){
+        // avoid stride % bpp != 0
+        if (srcFormat==PIX_FMT_RGB24 || srcFormat==PIX_FMT_BGR24)
+            srcStride[i]= srcW*3;
+        else
+            srcStride[i]= srcW*4;
+
+        if (dstFormat==PIX_FMT_RGB24 || dstFormat==PIX_FMT_BGR24)
+            dstStride[i]= dstW*3;
+        else
+            dstStride[i]= dstW*4;
+
+        src[i]= (uint8_t*) malloc(srcStride[i]*srcH);
+        dst[i]= (uint8_t*) malloc(dstStride[i]*dstH);
+        out[i]= (uint8_t*) malloc(refStride[i]*h);
+        if ((src[i] == NULL) || (dst[i] == NULL) || (out[i] == NULL)) {
+            perror("Malloc");
+            res = -1;
+
+            goto end;
+        }
+    }
+
+    dstContext = outContext = NULL;
+    srcContext= sws_getContext(w, h, PIX_FMT_YUV420P, srcW, srcH, srcFormat, flags, NULL, NULL, NULL);
+    if (srcContext == NULL) {
+        fprintf(stderr, "Failed to get %s ---> %s\n",
+                sws_format_name(PIX_FMT_YUV420P),
+                sws_format_name(srcFormat));
+        res = -1;
+
+        goto end;
+    }
+    dstContext= sws_getContext(srcW, srcH, srcFormat, dstW, dstH, dstFormat, flags, NULL, NULL, NULL);
+    if (dstContext == NULL) {
+        fprintf(stderr, "Failed to get %s ---> %s\n",
+                sws_format_name(srcFormat),
+                sws_format_name(dstFormat));
+        res = -1;
+
+        goto end;
+    }
+    outContext= sws_getContext(dstW, dstH, dstFormat, w, h, PIX_FMT_YUV420P, flags, NULL, NULL, NULL);
+    if (outContext == NULL) {
+        fprintf(stderr, "Failed to get %s ---> %s\n",
+                sws_format_name(dstFormat),
+                sws_format_name(PIX_FMT_YUV420P));
+        res = -1;
+
+        goto end;
+    }
+//    printf("test %X %X %X -> %X %X %X\n", (int)ref[0], (int)ref[1], (int)ref[2],
+//        (int)src[0], (int)src[1], (int)src[2]);
+
+    sws_scale(srcContext, ref, refStride, 0, h   , src, srcStride);
+    sws_scale(dstContext, src, srcStride, 0, srcH, dst, dstStride);
+    sws_scale(outContext, dst, dstStride, 0, dstH, out, refStride);
 
 #if defined(ARCH_X86)
-	asm volatile ("emms\n\t");
+    asm volatile ("emms\n\t");
 #endif
 
-	ssdY= getSSD(ref[0], out[0], refStride[0], refStride[0], w, h);
-	ssdU= getSSD(ref[1], out[1], refStride[1], refStride[1], (w+1)>>1, (h+1)>>1);
-	ssdV= getSSD(ref[2], out[2], refStride[2], refStride[2], (w+1)>>1, (h+1)>>1);
+    ssdY= getSSD(ref[0], out[0], refStride[0], refStride[0], w, h);
+    ssdU= getSSD(ref[1], out[1], refStride[1], refStride[1], (w+1)>>1, (h+1)>>1);
+    ssdV= getSSD(ref[2], out[2], refStride[2], refStride[2], (w+1)>>1, (h+1)>>1);
 
-	if(srcFormat == PIX_FMT_GRAY8 || dstFormat==PIX_FMT_GRAY8) ssdU=ssdV=0; //FIXME check that output is really gray
+    if (srcFormat == PIX_FMT_GRAY8 || dstFormat==PIX_FMT_GRAY8) ssdU=ssdV=0; //FIXME check that output is really gray
 
-	ssdY/= w*h;
-	ssdU/= w*h/4;
-	ssdV/= w*h/4;
+    ssdY/= w*h;
+    ssdU/= w*h/4;
+    ssdV/= w*h/4;
 
-	if(ssdY>100 || ssdU>100 || ssdV>100){
-		printf(" %s %dx%d -> %s %4dx%4d flags=%2d SSD=%5lld,%5lld,%5lld\n",
-			sws_format_name(srcFormat), srcW, srcH,
-			sws_format_name(dstFormat), dstW, dstH,
-			flags,
-			ssdY, ssdU, ssdV);
-	}
+    if (ssdY>100 || ssdU>100 || ssdV>100){
+        printf(" %s %dx%d -> %s %4dx%4d flags=%2d SSD=%5lld,%5lld,%5lld\n",
+               sws_format_name(srcFormat), srcW, srcH,
+               sws_format_name(dstFormat), dstW, dstH,
+               flags,
+               ssdY, ssdU, ssdV);
+    }
 
-	end:
+    end:
 
-	sws_freeContext(srcContext);
-	sws_freeContext(dstContext);
-	sws_freeContext(outContext);
+    sws_freeContext(srcContext);
+    sws_freeContext(dstContext);
+    sws_freeContext(outContext);
 
-	for(i=0; i<3; i++){
-		free(src[i]);
-		free(dst[i]);
-		free(out[i]);
-	}
+    for (i=0; i<3; i++){
+        free(src[i]);
+        free(dst[i]);
+        free(out[i]);
+    }
 
-	return res;
+    return res;
 }
 
 void fast_memcpy(void *a, void *b, int s){ //FIXME
@@ -161,69 +161,69 @@ void fast_memcpy(void *a, void *b, int s){ //FIXME
 }
 
 static void selfTest(uint8_t *src[3], int stride[3], int w, int h){
-	enum PixelFormat srcFormat, dstFormat;
-	int srcW, srcH, dstW, dstH;
-	int flags;
-
-	for(srcFormat = 0; srcFormat < PIX_FMT_NB; srcFormat++) {
-		for(dstFormat = 0; dstFormat < PIX_FMT_NB; dstFormat++) {
-			printf("%s -> %s\n",
-					sws_format_name(srcFormat),
-					sws_format_name(dstFormat));
-
-			srcW= w;
-			srcH= h;
-			for(dstW=w - w/3; dstW<= 4*w/3; dstW+= w/3){
-				for(dstH=h - h/3; dstH<= 4*h/3; dstH+= h/3){
-					for(flags=1; flags<33; flags*=2) {
-						int res;
-
-						res = doTest(src, stride, w, h, srcFormat, dstFormat,
-							srcW, srcH, dstW, dstH, flags);
-						if (res < 0) {
-							dstW = 4 * w / 3;
-							dstH = 4 * h / 3;
-							flags = 33;
-						}
-					}
-				}
-			}
-		}
-	}
+    enum PixelFormat srcFormat, dstFormat;
+    int srcW, srcH, dstW, dstH;
+    int flags;
+
+    for (srcFormat = 0; srcFormat < PIX_FMT_NB; srcFormat++) {
+        for (dstFormat = 0; dstFormat < PIX_FMT_NB; dstFormat++) {
+            printf("%s -> %s\n",
+                   sws_format_name(srcFormat),
+                   sws_format_name(dstFormat));
+
+            srcW= w;
+            srcH= h;
+            for (dstW=w - w/3; dstW<= 4*w/3; dstW+= w/3){
+                for (dstH=h - h/3; dstH<= 4*h/3; dstH+= h/3){
+                    for (flags=1; flags<33; flags*=2) {
+                        int res;
+
+                        res = doTest(src, stride, w, h, srcFormat, dstFormat,
+                                     srcW, srcH, dstW, dstH, flags);
+                        if (res < 0) {
+                            dstW = 4 * w / 3;
+                            dstH = 4 * h / 3;
+                            flags = 33;
+                        }
+                    }
+                }
+            }
+        }
+    }
 }
 
 #define W 96
 #define H 96
 
 int main(int argc, char **argv){
-	uint8_t *rgb_data = malloc (W*H*4);
-	uint8_t *rgb_src[3]= {rgb_data, NULL, NULL};
-	int rgb_stride[3]={4*W, 0, 0};
-	uint8_t *data = malloc (3*W*H);
-	uint8_t *src[3]= {data, data+W*H, data+W*H*2};
-	int stride[3]={W, W, W};
-	int x, y;
-	struct SwsContext *sws;
-
-	sws= sws_getContext(W/12, H/12, PIX_FMT_RGB32, W, H, PIX_FMT_YUV420P, 2, NULL, NULL, NULL);
-
-	for(y=0; y<H; y++){
-		for(x=0; x<W*4; x++){
-			rgb_data[ x + y*4*W]= random();
-		}
-	}
+    uint8_t *rgb_data = malloc (W*H*4);
+    uint8_t *rgb_src[3]= {rgb_data, NULL, NULL};
+    int rgb_stride[3]={4*W, 0, 0};
+    uint8_t *data = malloc (3*W*H);
+    uint8_t *src[3]= {data, data+W*H, data+W*H*2};
+    int stride[3]={W, W, W};
+    int x, y;
+    struct SwsContext *sws;
+
+    sws= sws_getContext(W/12, H/12, PIX_FMT_RGB32, W, H, PIX_FMT_YUV420P, 2, NULL, NULL, NULL);
+
+    for (y=0; y<H; y++){
+        for (x=0; x<W*4; x++){
+            rgb_data[ x + y*4*W]= random();
+        }
+    }
 #if defined(ARCH_X86)
-	sws_rgb2rgb_init(SWS_CPU_CAPS_MMX*0);
+    sws_rgb2rgb_init(SWS_CPU_CAPS_MMX*0);
 #else
-	sws_rgb2rgb_init(0);
+    sws_rgb2rgb_init(0);
 #endif
-	sws_scale(sws, rgb_src, rgb_stride, 0, H   , src, stride);
+    sws_scale(sws, rgb_src, rgb_stride, 0, H   , src, stride);
 
 #if defined(ARCH_X86)
-	asm volatile ("emms\n\t");
+    asm volatile ("emms\n\t");
 #endif
 
-	selfTest(src,  stride, W, H);
+    selfTest(src,  stride, W, H);
 
-        return 123;
+    return 123;
 }
diff --git a/libswscale/swscale.c b/libswscale/swscale.c
index af690e9a5e..901e6c6149 100644
--- a/libswscale/swscale.c
+++ b/libswscale/swscale.c
@@ -102,21 +102,21 @@ untested special converters
 #define PI 3.14159265358979323846
 #endif
 
-#define isSupportedIn(x)  ((x)==PIX_FMT_YUV420P || (x)==PIX_FMT_YUYV422 || (x)==PIX_FMT_UYVY422\
-			|| (x)==PIX_FMT_RGB32|| (x)==PIX_FMT_BGR24|| (x)==PIX_FMT_BGR565|| (x)==PIX_FMT_BGR555\
-			|| (x)==PIX_FMT_BGR32|| (x)==PIX_FMT_RGB24|| (x)==PIX_FMT_RGB565|| (x)==PIX_FMT_RGB555\
-			|| (x)==PIX_FMT_GRAY8 || (x)==PIX_FMT_YUV410P\
-			|| (x)==PIX_FMT_GRAY16BE || (x)==PIX_FMT_GRAY16LE\
-			|| (x)==PIX_FMT_YUV444P || (x)==PIX_FMT_YUV422P || (x)==PIX_FMT_YUV411P\
-			|| (x)==PIX_FMT_PAL8 || (x)==PIX_FMT_BGR8 || (x)==PIX_FMT_RGB8\
-                        || (x)==PIX_FMT_BGR4_BYTE  || (x)==PIX_FMT_RGB4_BYTE)
-#define isSupportedOut(x) ((x)==PIX_FMT_YUV420P || (x)==PIX_FMT_YUYV422 || (x)==PIX_FMT_UYVY422\
-			|| (x)==PIX_FMT_YUV444P || (x)==PIX_FMT_YUV422P || (x)==PIX_FMT_YUV411P\
-			|| isRGB(x) || isBGR(x)\
-			|| (x)==PIX_FMT_NV12 || (x)==PIX_FMT_NV21\
-			|| (x)==PIX_FMT_GRAY16BE || (x)==PIX_FMT_GRAY16LE\
-			|| (x)==PIX_FMT_GRAY8 || (x)==PIX_FMT_YUV410P)
-#define isPacked(x)    ((x)==PIX_FMT_PAL8 || (x)==PIX_FMT_YUYV422 ||\
+#define isSupportedIn(x)  ((x)==PIX_FMT_YUV420P   || (x)==PIX_FMT_YUYV422 || (x)==PIX_FMT_UYVY422 \
+                        || (x)==PIX_FMT_RGB32     || (x)==PIX_FMT_BGR24   || (x)==PIX_FMT_BGR565 || (x)==PIX_FMT_BGR555 \
+                        || (x)==PIX_FMT_BGR32     || (x)==PIX_FMT_RGB24   || (x)==PIX_FMT_RGB565 || (x)==PIX_FMT_RGB555 \
+                        || (x)==PIX_FMT_GRAY8     || (x)==PIX_FMT_YUV410P  \
+                        || (x)==PIX_FMT_GRAY16BE  || (x)==PIX_FMT_GRAY16LE \
+                        || (x)==PIX_FMT_YUV444P   || (x)==PIX_FMT_YUV422P || (x)==PIX_FMT_YUV411P \
+                        || (x)==PIX_FMT_PAL8      || (x)==PIX_FMT_BGR8    || (x)==PIX_FMT_RGB8 \
+                        || (x)==PIX_FMT_BGR4_BYTE || (x)==PIX_FMT_RGB4_BYTE)
+#define isSupportedOut(x) ((x)==PIX_FMT_YUV420P  || (x)==PIX_FMT_YUYV422 || (x)==PIX_FMT_UYVY422 \
+                        || (x)==PIX_FMT_YUV444P  || (x)==PIX_FMT_YUV422P || (x)==PIX_FMT_YUV411P \
+                        || isRGB(x) || isBGR(x) \
+                        || (x)==PIX_FMT_NV12     || (x)==PIX_FMT_NV21 \
+                        || (x)==PIX_FMT_GRAY16BE || (x)==PIX_FMT_GRAY16LE \
+                        || (x)==PIX_FMT_GRAY8    || (x)==PIX_FMT_YUV410P)
+#define isPacked(x)    ((x)==PIX_FMT_PAL8    || (x)==PIX_FMT_YUYV422 || \
                         (x)==PIX_FMT_UYVY422 || isRGB(x) || isBGR(x))
 
 #define RGB2YUV_SHIFT 16
@@ -150,7 +150,7 @@ write special BGR->BGR scaler
 #if defined(ARCH_X86) && defined (CONFIG_GPL)
 static uint64_t attribute_used __attribute__((aligned(8))) bF8=       0xF8F8F8F8F8F8F8F8LL;
 static uint64_t attribute_used __attribute__((aligned(8))) bFC=       0xFCFCFCFCFCFCFCFCLL;
-static uint64_t __attribute__((aligned(8))) w10=       0x0010001000100010LL;
+static uint64_t                __attribute__((aligned(8))) w10=       0x0010001000100010LL;
 static uint64_t attribute_used __attribute__((aligned(8))) w02=       0x0002000200020002LL;
 static uint64_t attribute_used __attribute__((aligned(8))) bm00001111=0x00000000FFFFFFFFLL;
 static uint64_t attribute_used __attribute__((aligned(8))) bm00000111=0x0000000000FFFFFFLL;
@@ -163,36 +163,36 @@ static volatile uint64_t attribute_used __attribute__((aligned(8))) g6Dither;
 static volatile uint64_t attribute_used __attribute__((aligned(8))) r5Dither;
 
 static uint64_t __attribute__((aligned(8))) dither4[2]={
-	0x0103010301030103LL,
-	0x0200020002000200LL,};
+        0x0103010301030103LL,
+        0x0200020002000200LL,};
 
 static uint64_t __attribute__((aligned(8))) dither8[2]={
-	0x0602060206020602LL,
-	0x0004000400040004LL,};
+        0x0602060206020602LL,
+        0x0004000400040004LL,};
 
-static uint64_t __attribute__((aligned(8))) b16Mask=   0x001F001F001F001FLL;
+static uint64_t                __attribute__((aligned(8))) b16Mask=   0x001F001F001F001FLL;
 static uint64_t attribute_used __attribute__((aligned(8))) g16Mask=   0x07E007E007E007E0LL;
 static uint64_t attribute_used __attribute__((aligned(8))) r16Mask=   0xF800F800F800F800LL;
-static uint64_t __attribute__((aligned(8))) b15Mask=   0x001F001F001F001FLL;
+static uint64_t                __attribute__((aligned(8))) b15Mask=   0x001F001F001F001FLL;
 static uint64_t attribute_used __attribute__((aligned(8))) g15Mask=   0x03E003E003E003E0LL;
 static uint64_t attribute_used __attribute__((aligned(8))) r15Mask=   0x7C007C007C007C00LL;
 
-static uint64_t attribute_used __attribute__((aligned(8))) M24A=   0x00FF0000FF0000FFLL;
-static uint64_t attribute_used __attribute__((aligned(8))) M24B=   0xFF0000FF0000FF00LL;
-static uint64_t attribute_used __attribute__((aligned(8))) M24C=   0x0000FF0000FF0000LL;
+static uint64_t attribute_used __attribute__((aligned(8))) M24A=      0x00FF0000FF0000FFLL;
+static uint64_t attribute_used __attribute__((aligned(8))) M24B=      0xFF0000FF0000FF00LL;
+static uint64_t attribute_used __attribute__((aligned(8))) M24C=      0x0000FF0000FF0000LL;
 
 #ifdef FAST_BGR2YV12
-static const uint64_t bgr2YCoeff  attribute_used __attribute__((aligned(8))) = 0x000000210041000DULL;
-static const uint64_t bgr2UCoeff  attribute_used __attribute__((aligned(8))) = 0x0000FFEEFFDC0038ULL;
-static const uint64_t bgr2VCoeff  attribute_used __attribute__((aligned(8))) = 0x00000038FFD2FFF8ULL;
+static const uint64_t bgr2YCoeff   attribute_used __attribute__((aligned(8))) = 0x000000210041000DULL;
+static const uint64_t bgr2UCoeff   attribute_used __attribute__((aligned(8))) = 0x0000FFEEFFDC0038ULL;
+static const uint64_t bgr2VCoeff   attribute_used __attribute__((aligned(8))) = 0x00000038FFD2FFF8ULL;
 #else
-static const uint64_t bgr2YCoeff  attribute_used __attribute__((aligned(8))) = 0x000020E540830C8BULL;
-static const uint64_t bgr2UCoeff  attribute_used __attribute__((aligned(8))) = 0x0000ED0FDAC23831ULL;
-static const uint64_t bgr2VCoeff  attribute_used __attribute__((aligned(8))) = 0x00003831D0E6F6EAULL;
+static const uint64_t bgr2YCoeff   attribute_used __attribute__((aligned(8))) = 0x000020E540830C8BULL;
+static const uint64_t bgr2UCoeff   attribute_used __attribute__((aligned(8))) = 0x0000ED0FDAC23831ULL;
+static const uint64_t bgr2VCoeff   attribute_used __attribute__((aligned(8))) = 0x00003831D0E6F6EAULL;
 #endif /* FAST_BGR2YV12 */
-static const uint64_t bgr2YOffset attribute_used __attribute__((aligned(8))) = 0x1010101010101010ULL;
-static const uint64_t bgr2UVOffset attribute_used __attribute__((aligned(8)))= 0x8080808080808080ULL;
-static const uint64_t w1111       attribute_used __attribute__((aligned(8))) = 0x0001000100010001ULL;
+static const uint64_t bgr2YOffset  attribute_used __attribute__((aligned(8))) = 0x1010101010101010ULL;
+static const uint64_t bgr2UVOffset attribute_used __attribute__((aligned(8))) = 0x8080808080808080ULL;
+static const uint64_t w1111        attribute_used __attribute__((aligned(8))) = 0x0001000100010001ULL;
 #endif /* defined(ARCH_X86) */
 
 // clipping helper table for C implementations:
@@ -297,511 +297,511 @@ char *sws_format_name(enum PixelFormat format)
 #if defined(ARCH_X86) && defined (CONFIG_GPL)
 void in_asm_used_var_warning_killer()
 {
- volatile int i= bF8+bFC+w10+
- bm00001111+bm00000111+bm11111000+b16Mask+g16Mask+r16Mask+b15Mask+g15Mask+r15Mask+
- M24A+M24B+M24C+w02 + b5Dither+g5Dither+r5Dither+g6Dither+dither4[0]+dither8[0]+bm01010101;
- if(i) i=0;
+    volatile int i= bF8+bFC+w10+
+    bm00001111+bm00000111+bm11111000+b16Mask+g16Mask+r16Mask+b15Mask+g15Mask+r15Mask+
+    M24A+M24B+M24C+w02 + b5Dither+g5Dither+r5Dither+g6Dither+dither4[0]+dither8[0]+bm01010101;
+    if (i) i=0;
 }
 #endif
 
 static inline void yuv2yuvXinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
-				    int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
-				    uint8_t *dest, uint8_t *uDest, uint8_t *vDest, int dstW, int chrDstW)
+                               int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
+                               uint8_t *dest, uint8_t *uDest, uint8_t *vDest, int dstW, int chrDstW)
 {
-	//FIXME Optimize (just quickly writen not opti..)
-	int i;
-	for(i=0; i<dstW; i++)
-	{
-		int val=1<<18;
-		int j;
-		for(j=0; j<lumFilterSize; j++)
-			val += lumSrc[j][i] * lumFilter[j];
-
-		dest[i]= av_clip_uint8(val>>19);
-	}
-
-	if(uDest != NULL)
-		for(i=0; i<chrDstW; i++)
-		{
-			int u=1<<18;
-			int v=1<<18;
-			int j;
-			for(j=0; j<chrFilterSize; j++)
-			{
-				u += chrSrc[j][i] * chrFilter[j];
-				v += chrSrc[j][i + 2048] * chrFilter[j];
-			}
-
-			uDest[i]= av_clip_uint8(u>>19);
-			vDest[i]= av_clip_uint8(v>>19);
-		}
+    //FIXME Optimize (just quickly writen not opti..)
+    int i;
+    for (i=0; i<dstW; i++)
+    {
+        int val=1<<18;
+        int j;
+        for (j=0; j<lumFilterSize; j++)
+            val += lumSrc[j][i] * lumFilter[j];
+
+        dest[i]= av_clip_uint8(val>>19);
+    }
+
+    if (uDest != NULL)
+        for (i=0; i<chrDstW; i++)
+        {
+            int u=1<<18;
+            int v=1<<18;
+            int j;
+            for (j=0; j<chrFilterSize; j++)
+            {
+                u += chrSrc[j][i] * chrFilter[j];
+                v += chrSrc[j][i + 2048] * chrFilter[j];
+            }
+
+            uDest[i]= av_clip_uint8(u>>19);
+            vDest[i]= av_clip_uint8(v>>19);
+        }
 }
 
 static inline void yuv2nv12XinC(int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
-				int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
-				uint8_t *dest, uint8_t *uDest, int dstW, int chrDstW, int dstFormat)
+                                int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
+                                uint8_t *dest, uint8_t *uDest, int dstW, int chrDstW, int dstFormat)
 {
-	//FIXME Optimize (just quickly writen not opti..)
-	int i;
-	for(i=0; i<dstW; i++)
-	{
-		int val=1<<18;
-		int j;
-		for(j=0; j<lumFilterSize; j++)
-			val += lumSrc[j][i] * lumFilter[j];
-
-		dest[i]= av_clip_uint8(val>>19);
-	}
-
-	if(uDest == NULL)
-		return;
-
-	if(dstFormat == PIX_FMT_NV12)
-		for(i=0; i<chrDstW; i++)
-		{
-			int u=1<<18;
-			int v=1<<18;
-			int j;
-			for(j=0; j<chrFilterSize; j++)
-			{
-				u += chrSrc[j][i] * chrFilter[j];
-				v += chrSrc[j][i + 2048] * chrFilter[j];
-			}
-
-			uDest[2*i]= av_clip_uint8(u>>19);
-			uDest[2*i+1]= av_clip_uint8(v>>19);
-		}
-	else
-		for(i=0; i<chrDstW; i++)
-		{
-			int u=1<<18;
-			int v=1<<18;
-			int j;
-			for(j=0; j<chrFilterSize; j++)
-			{
-				u += chrSrc[j][i] * chrFilter[j];
-				v += chrSrc[j][i + 2048] * chrFilter[j];
-			}
-
-			uDest[2*i]= av_clip_uint8(v>>19);
-			uDest[2*i+1]= av_clip_uint8(u>>19);
-		}
+    //FIXME Optimize (just quickly writen not opti..)
+    int i;
+    for (i=0; i<dstW; i++)
+    {
+        int val=1<<18;
+        int j;
+        for (j=0; j<lumFilterSize; j++)
+            val += lumSrc[j][i] * lumFilter[j];
+
+        dest[i]= av_clip_uint8(val>>19);
+    }
+
+    if (uDest == NULL)
+        return;
+
+    if (dstFormat == PIX_FMT_NV12)
+        for (i=0; i<chrDstW; i++)
+        {
+            int u=1<<18;
+            int v=1<<18;
+            int j;
+            for (j=0; j<chrFilterSize; j++)
+            {
+                u += chrSrc[j][i] * chrFilter[j];
+                v += chrSrc[j][i + 2048] * chrFilter[j];
+            }
+
+            uDest[2*i]= av_clip_uint8(u>>19);
+            uDest[2*i+1]= av_clip_uint8(v>>19);
+        }
+    else
+        for (i=0; i<chrDstW; i++)
+        {
+            int u=1<<18;
+            int v=1<<18;
+            int j;
+            for (j=0; j<chrFilterSize; j++)
+            {
+                u += chrSrc[j][i] * chrFilter[j];
+                v += chrSrc[j][i + 2048] * chrFilter[j];
+            }
+
+            uDest[2*i]= av_clip_uint8(v>>19);
+            uDest[2*i+1]= av_clip_uint8(u>>19);
+        }
 }
 
 #define YSCALE_YUV_2_PACKEDX_C(type) \
-		for(i=0; i<(dstW>>1); i++){\
-			int j;\
-			int Y1=1<<18;\
-			int Y2=1<<18;\
-			int U=1<<18;\
-			int V=1<<18;\
-			type attribute_unused *r, *b, *g;\
-			const int i2= 2*i;\
-			\
-			for(j=0; j<lumFilterSize; j++)\
-			{\
-				Y1 += lumSrc[j][i2] * lumFilter[j];\
-				Y2 += lumSrc[j][i2+1] * lumFilter[j];\
-			}\
-			for(j=0; j<chrFilterSize; j++)\
-			{\
-				U += chrSrc[j][i] * chrFilter[j];\
-				V += chrSrc[j][i+2048] * chrFilter[j];\
-			}\
-			Y1>>=19;\
-			Y2>>=19;\
-			U >>=19;\
-			V >>=19;\
-			if((Y1|Y2|U|V)&256)\
-			{\
-				if(Y1>255)   Y1=255;\
-				else if(Y1<0)Y1=0;\
-				if(Y2>255)   Y2=255;\
-				else if(Y2<0)Y2=0;\
-				if(U>255)    U=255;\
-				else if(U<0) U=0;\
-				if(V>255)    V=255;\
-				else if(V<0) V=0;\
-			}
+    for (i=0; i<(dstW>>1); i++){\
+        int j;\
+        int Y1 = 1<<18;\
+        int Y2 = 1<<18;\
+        int U  = 1<<18;\
+        int V  = 1<<18;\
+        type attribute_unused *r, *b, *g;\
+        const int i2= 2*i;\
+        \
+        for (j=0; j<lumFilterSize; j++)\
+        {\
+            Y1 += lumSrc[j][i2] * lumFilter[j];\
+            Y2 += lumSrc[j][i2+1] * lumFilter[j];\
+        }\
+        for (j=0; j<chrFilterSize; j++)\
+        {\
+            U += chrSrc[j][i] * chrFilter[j];\
+            V += chrSrc[j][i+2048] * chrFilter[j];\
+        }\
+        Y1>>=19;\
+        Y2>>=19;\
+        U >>=19;\
+        V >>=19;\
+        if ((Y1|Y2|U|V)&256)\
+        {\
+            if (Y1>255)   Y1=255; \
+            else if (Y1<0)Y1=0;   \
+            if (Y2>255)   Y2=255; \
+            else if (Y2<0)Y2=0;   \
+            if (U>255)    U=255;  \
+            else if (U<0) U=0;    \
+            if (V>255)    V=255;  \
+            else if (V<0) V=0;    \
+        }
 
 #define YSCALE_YUV_2_RGBX_C(type) \
-			YSCALE_YUV_2_PACKEDX_C(type)\
-			r = (type *)c->table_rV[V];\
-			g = (type *)(c->table_gU[U] + c->table_gV[V]);\
-			b = (type *)c->table_bU[U];\
-
-#define YSCALE_YUV_2_PACKED2_C \
-		for(i=0; i<(dstW>>1); i++){\
-			const int i2= 2*i;\
-			int Y1= (buf0[i2  ]*yalpha1+buf1[i2  ]*yalpha)>>19;\
-			int Y2= (buf0[i2+1]*yalpha1+buf1[i2+1]*yalpha)>>19;\
-			int U= (uvbuf0[i     ]*uvalpha1+uvbuf1[i     ]*uvalpha)>>19;\
-			int V= (uvbuf0[i+2048]*uvalpha1+uvbuf1[i+2048]*uvalpha)>>19;\
+    YSCALE_YUV_2_PACKEDX_C(type)  \
+    r = (type *)c->table_rV[V];   \
+    g = (type *)(c->table_gU[U] + c->table_gV[V]); \
+    b = (type *)c->table_bU[U];   \
+
+#define YSCALE_YUV_2_PACKED2_C   \
+    for (i=0; i<(dstW>>1); i++){ \
+        const int i2= 2*i;       \
+        int Y1= (buf0[i2  ]*yalpha1+buf1[i2  ]*yalpha)>>19;           \
+        int Y2= (buf0[i2+1]*yalpha1+buf1[i2+1]*yalpha)>>19;           \
+        int U= (uvbuf0[i     ]*uvalpha1+uvbuf1[i     ]*uvalpha)>>19;  \
+        int V= (uvbuf0[i+2048]*uvalpha1+uvbuf1[i+2048]*uvalpha)>>19;  \
 
 #define YSCALE_YUV_2_RGB2_C(type) \
-			YSCALE_YUV_2_PACKED2_C\
-			type *r, *b, *g;\
-			r = (type *)c->table_rV[V];\
-			g = (type *)(c->table_gU[U] + c->table_gV[V]);\
-			b = (type *)c->table_bU[U];\
+    YSCALE_YUV_2_PACKED2_C\
+    type *r, *b, *g;\
+    r = (type *)c->table_rV[V];\
+    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
+    b = (type *)c->table_bU[U];\
 
 #define YSCALE_YUV_2_PACKED1_C \
-		for(i=0; i<(dstW>>1); i++){\
-			const int i2= 2*i;\
-			int Y1= buf0[i2  ]>>7;\
-			int Y2= buf0[i2+1]>>7;\
-			int U= (uvbuf1[i     ])>>7;\
-			int V= (uvbuf1[i+2048])>>7;\
+    for (i=0; i<(dstW>>1); i++){\
+        const int i2= 2*i;\
+        int Y1= buf0[i2  ]>>7;\
+        int Y2= buf0[i2+1]>>7;\
+        int U= (uvbuf1[i     ])>>7;\
+        int V= (uvbuf1[i+2048])>>7;\
 
 #define YSCALE_YUV_2_RGB1_C(type) \
-			YSCALE_YUV_2_PACKED1_C\
-			type *r, *b, *g;\
-			r = (type *)c->table_rV[V];\
-			g = (type *)(c->table_gU[U] + c->table_gV[V]);\
-			b = (type *)c->table_bU[U];\
+    YSCALE_YUV_2_PACKED1_C\
+    type *r, *b, *g;\
+    r = (type *)c->table_rV[V];\
+    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
+    b = (type *)c->table_bU[U];\
 
 #define YSCALE_YUV_2_PACKED1B_C \
-		for(i=0; i<(dstW>>1); i++){\
-			const int i2= 2*i;\
-			int Y1= buf0[i2  ]>>7;\
-			int Y2= buf0[i2+1]>>7;\
-			int U= (uvbuf0[i     ] + uvbuf1[i     ])>>8;\
-			int V= (uvbuf0[i+2048] + uvbuf1[i+2048])>>8;\
+    for (i=0; i<(dstW>>1); i++){\
+        const int i2= 2*i;\
+        int Y1= buf0[i2  ]>>7;\
+        int Y2= buf0[i2+1]>>7;\
+        int U= (uvbuf0[i     ] + uvbuf1[i     ])>>8;\
+        int V= (uvbuf0[i+2048] + uvbuf1[i+2048])>>8;\
 
 #define YSCALE_YUV_2_RGB1B_C(type) \
-			YSCALE_YUV_2_PACKED1B_C\
-			type *r, *b, *g;\
-			r = (type *)c->table_rV[V];\
-			g = (type *)(c->table_gU[U] + c->table_gV[V]);\
-			b = (type *)c->table_bU[U];\
+    YSCALE_YUV_2_PACKED1B_C\
+    type *r, *b, *g;\
+    r = (type *)c->table_rV[V];\
+    g = (type *)(c->table_gU[U] + c->table_gV[V]);\
+    b = (type *)c->table_bU[U];\
 
 #define YSCALE_YUV_2_ANYRGB_C(func, func2)\
-	switch(c->dstFormat)\
-	{\
-	case PIX_FMT_RGB32:\
-	case PIX_FMT_BGR32:\
-		func(uint32_t)\
-			((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];\
-			((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];\
-		}		\
-		break;\
-	case PIX_FMT_RGB24:\
-		func(uint8_t)\
-			((uint8_t*)dest)[0]= r[Y1];\
-			((uint8_t*)dest)[1]= g[Y1];\
-			((uint8_t*)dest)[2]= b[Y1];\
-			((uint8_t*)dest)[3]= r[Y2];\
-			((uint8_t*)dest)[4]= g[Y2];\
-			((uint8_t*)dest)[5]= b[Y2];\
-			dest+=6;\
-		}\
-		break;\
-	case PIX_FMT_BGR24:\
-		func(uint8_t)\
-			((uint8_t*)dest)[0]= b[Y1];\
-			((uint8_t*)dest)[1]= g[Y1];\
-			((uint8_t*)dest)[2]= r[Y1];\
-			((uint8_t*)dest)[3]= b[Y2];\
-			((uint8_t*)dest)[4]= g[Y2];\
-			((uint8_t*)dest)[5]= r[Y2];\
-			dest+=6;\
-		}\
-		break;\
-	case PIX_FMT_RGB565:\
-	case PIX_FMT_BGR565:\
-		{\
-			const int dr1= dither_2x2_8[y&1    ][0];\
-			const int dg1= dither_2x2_4[y&1    ][0];\
-			const int db1= dither_2x2_8[(y&1)^1][0];\
-			const int dr2= dither_2x2_8[y&1    ][1];\
-			const int dg2= dither_2x2_4[y&1    ][1];\
-			const int db2= dither_2x2_8[(y&1)^1][1];\
-			func(uint16_t)\
-				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
-				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
-			}\
-		}\
-		break;\
-	case PIX_FMT_RGB555:\
-	case PIX_FMT_BGR555:\
-		{\
-			const int dr1= dither_2x2_8[y&1    ][0];\
-			const int dg1= dither_2x2_8[y&1    ][1];\
-			const int db1= dither_2x2_8[(y&1)^1][0];\
-			const int dr2= dither_2x2_8[y&1    ][1];\
-			const int dg2= dither_2x2_8[y&1    ][0];\
-			const int db2= dither_2x2_8[(y&1)^1][1];\
-			func(uint16_t)\
-				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
-				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
-			}\
-		}\
-		break;\
-	case PIX_FMT_RGB8:\
-	case PIX_FMT_BGR8:\
-		{\
-			const uint8_t * const d64= dither_8x8_73[y&7];\
-			const uint8_t * const d32= dither_8x8_32[y&7];\
-			func(uint8_t)\
-				((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];\
-				((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];\
-			}\
-		}\
-		break;\
-	case PIX_FMT_RGB4:\
-	case PIX_FMT_BGR4:\
-		{\
-			const uint8_t * const d64= dither_8x8_73 [y&7];\
-			const uint8_t * const d128=dither_8x8_220[y&7];\
-			func(uint8_t)\
-				((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]\
-				                 + ((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);\
-			}\
-		}\
-		break;\
-	case PIX_FMT_RGB4_BYTE:\
-	case PIX_FMT_BGR4_BYTE:\
-		{\
-			const uint8_t * const d64= dither_8x8_73 [y&7];\
-			const uint8_t * const d128=dither_8x8_220[y&7];\
-			func(uint8_t)\
-				((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];\
-				((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];\
-			}\
-		}\
-		break;\
-	case PIX_FMT_MONOBLACK:\
-		{\
-			const uint8_t * const d128=dither_8x8_220[y&7];\
-			uint8_t *g= c->table_gU[128] + c->table_gV[128];\
-			for(i=0; i<dstW-7; i+=8){\
-				int acc;\
-				acc =       g[((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19) + d128[0]];\
-				acc+= acc + g[((buf0[i+1]*yalpha1+buf1[i+1]*yalpha)>>19) + d128[1]];\
-				acc+= acc + g[((buf0[i+2]*yalpha1+buf1[i+2]*yalpha)>>19) + d128[2]];\
-				acc+= acc + g[((buf0[i+3]*yalpha1+buf1[i+3]*yalpha)>>19) + d128[3]];\
-				acc+= acc + g[((buf0[i+4]*yalpha1+buf1[i+4]*yalpha)>>19) + d128[4]];\
-				acc+= acc + g[((buf0[i+5]*yalpha1+buf1[i+5]*yalpha)>>19) + d128[5]];\
-				acc+= acc + g[((buf0[i+6]*yalpha1+buf1[i+6]*yalpha)>>19) + d128[6]];\
-				acc+= acc + g[((buf0[i+7]*yalpha1+buf1[i+7]*yalpha)>>19) + d128[7]];\
-				((uint8_t*)dest)[0]= acc;\
-				dest++;\
-			}\
+    switch(c->dstFormat)\
+    {\
+    case PIX_FMT_RGB32:\
+    case PIX_FMT_BGR32:\
+        func(uint32_t)\
+            ((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];\
+            ((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];\
+        }                \
+        break;\
+    case PIX_FMT_RGB24:\
+        func(uint8_t)\
+            ((uint8_t*)dest)[0]= r[Y1];\
+            ((uint8_t*)dest)[1]= g[Y1];\
+            ((uint8_t*)dest)[2]= b[Y1];\
+            ((uint8_t*)dest)[3]= r[Y2];\
+            ((uint8_t*)dest)[4]= g[Y2];\
+            ((uint8_t*)dest)[5]= b[Y2];\
+            dest+=6;\
+        }\
+        break;\
+    case PIX_FMT_BGR24:\
+        func(uint8_t)\
+            ((uint8_t*)dest)[0]= b[Y1];\
+            ((uint8_t*)dest)[1]= g[Y1];\
+            ((uint8_t*)dest)[2]= r[Y1];\
+            ((uint8_t*)dest)[3]= b[Y2];\
+            ((uint8_t*)dest)[4]= g[Y2];\
+            ((uint8_t*)dest)[5]= r[Y2];\
+            dest+=6;\
+        }\
+        break;\
+    case PIX_FMT_RGB565:\
+    case PIX_FMT_BGR565:\
+        {\
+            const int dr1= dither_2x2_8[y&1    ][0];\
+            const int dg1= dither_2x2_4[y&1    ][0];\
+            const int db1= dither_2x2_8[(y&1)^1][0];\
+            const int dr2= dither_2x2_8[y&1    ][1];\
+            const int dg2= dither_2x2_4[y&1    ][1];\
+            const int db2= dither_2x2_8[(y&1)^1][1];\
+            func(uint16_t)\
+                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
+                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
+            }\
+        }\
+        break;\
+    case PIX_FMT_RGB555:\
+    case PIX_FMT_BGR555:\
+        {\
+            const int dr1= dither_2x2_8[y&1    ][0];\
+            const int dg1= dither_2x2_8[y&1    ][1];\
+            const int db1= dither_2x2_8[(y&1)^1][0];\
+            const int dr2= dither_2x2_8[y&1    ][1];\
+            const int dg2= dither_2x2_8[y&1    ][0];\
+            const int db2= dither_2x2_8[(y&1)^1][1];\
+            func(uint16_t)\
+                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];\
+                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];\
+            }\
+        }\
+        break;\
+    case PIX_FMT_RGB8:\
+    case PIX_FMT_BGR8:\
+        {\
+            const uint8_t * const d64= dither_8x8_73[y&7];\
+            const uint8_t * const d32= dither_8x8_32[y&7];\
+            func(uint8_t)\
+                ((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];\
+                ((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];\
+            }\
+        }\
+        break;\
+    case PIX_FMT_RGB4:\
+    case PIX_FMT_BGR4:\
+        {\
+            const uint8_t * const d64= dither_8x8_73 [y&7];\
+            const uint8_t * const d128=dither_8x8_220[y&7];\
+            func(uint8_t)\
+                ((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]\
+                                 + ((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);\
+            }\
+        }\
+        break;\
+    case PIX_FMT_RGB4_BYTE:\
+    case PIX_FMT_BGR4_BYTE:\
+        {\
+            const uint8_t * const d64= dither_8x8_73 [y&7];\
+            const uint8_t * const d128=dither_8x8_220[y&7];\
+            func(uint8_t)\
+                ((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];\
+                ((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];\
+            }\
+        }\
+        break;\
+    case PIX_FMT_MONOBLACK:\
+        {\
+            const uint8_t * const d128=dither_8x8_220[y&7];\
+            uint8_t *g= c->table_gU[128] + c->table_gV[128];\
+            for (i=0; i<dstW-7; i+=8){\
+                int acc;\
+                acc =       g[((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19) + d128[0]];\
+                acc+= acc + g[((buf0[i+1]*yalpha1+buf1[i+1]*yalpha)>>19) + d128[1]];\
+                acc+= acc + g[((buf0[i+2]*yalpha1+buf1[i+2]*yalpha)>>19) + d128[2]];\
+                acc+= acc + g[((buf0[i+3]*yalpha1+buf1[i+3]*yalpha)>>19) + d128[3]];\
+                acc+= acc + g[((buf0[i+4]*yalpha1+buf1[i+4]*yalpha)>>19) + d128[4]];\
+                acc+= acc + g[((buf0[i+5]*yalpha1+buf1[i+5]*yalpha)>>19) + d128[5]];\
+                acc+= acc + g[((buf0[i+6]*yalpha1+buf1[i+6]*yalpha)>>19) + d128[6]];\
+                acc+= acc + g[((buf0[i+7]*yalpha1+buf1[i+7]*yalpha)>>19) + d128[7]];\
+                ((uint8_t*)dest)[0]= acc;\
+                dest++;\
+            }\
 \
 /*\
 ((uint8_t*)dest)-= dstW>>4;\
 {\
-			int acc=0;\
-			int left=0;\
-			static int top[1024];\
-			static int last_new[1024][1024];\
-			static int last_in3[1024][1024];\
-			static int drift[1024][1024];\
-			int topLeft=0;\
-			int shift=0;\
-			int count=0;\
-			const uint8_t * const d128=dither_8x8_220[y&7];\
-			int error_new=0;\
-			int error_in3=0;\
-			int f=0;\
-			\
-			for(i=dstW>>1; i<dstW; i++){\
-				int in= ((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19);\
-				int in2 = (76309 * (in - 16) + 32768) >> 16;\
-				int in3 = (in2 < 0) ? 0 : ((in2 > 255) ? 255 : in2);\
-				int old= (left*7 + topLeft + top[i]*5 + top[i+1]*3)/20 + in3\
-					+ (last_new[y][i] - in3)*f/256;\
-				int new= old> 128 ? 255 : 0;\
+            int acc=0;\
+            int left=0;\
+            static int top[1024];\
+            static int last_new[1024][1024];\
+            static int last_in3[1024][1024];\
+            static int drift[1024][1024];\
+            int topLeft=0;\
+            int shift=0;\
+            int count=0;\
+            const uint8_t * const d128=dither_8x8_220[y&7];\
+            int error_new=0;\
+            int error_in3=0;\
+            int f=0;\
+            \
+            for (i=dstW>>1; i<dstW; i++){\
+                int in= ((buf0[i  ]*yalpha1+buf1[i  ]*yalpha)>>19);\
+                int in2 = (76309 * (in - 16) + 32768) >> 16;\
+                int in3 = (in2 < 0) ? 0 : ((in2 > 255) ? 255 : in2);\
+                int old= (left*7 + topLeft + top[i]*5 + top[i+1]*3)/20 + in3\
+                         + (last_new[y][i] - in3)*f/256;\
+                int new= old> 128 ? 255 : 0;\
 \
-				error_new+= FFABS(last_new[y][i] - new);\
-				error_in3+= FFABS(last_in3[y][i] - in3);\
-				f= error_new - error_in3*4;\
-				if(f<0) f=0;\
-				if(f>256) f=256;\
+                error_new+= FFABS(last_new[y][i] - new);\
+                error_in3+= FFABS(last_in3[y][i] - in3);\
+                f= error_new - error_in3*4;\
+                if (f<0) f=0;\
+                if (f>256) f=256;\
 \
-				topLeft= top[i];\
-				left= top[i]= old - new;\
-				last_new[y][i]= new;\
-				last_in3[y][i]= in3;\
+                topLeft= top[i];\
+                left= top[i]= old - new;\
+                last_new[y][i]= new;\
+                last_in3[y][i]= in3;\
 \
-				acc+= acc + (new&1);\
-				if((i&7)==6){\
-					((uint8_t*)dest)[0]= acc;\
-					((uint8_t*)dest)++;\
-				}\
-			}\
+                acc+= acc + (new&1);\
+                if ((i&7)==6){\
+                    ((uint8_t*)dest)[0]= acc;\
+                    ((uint8_t*)dest)++;\
+                }\
+            }\
 }\
 */\
-		}\
-		break;\
-	case PIX_FMT_YUYV422:\
-		func2\
-			((uint8_t*)dest)[2*i2+0]= Y1;\
-			((uint8_t*)dest)[2*i2+1]= U;\
-			((uint8_t*)dest)[2*i2+2]= Y2;\
-			((uint8_t*)dest)[2*i2+3]= V;\
-		}		\
-		break;\
-	case PIX_FMT_UYVY422:\
-		func2\
-			((uint8_t*)dest)[2*i2+0]= U;\
-			((uint8_t*)dest)[2*i2+1]= Y1;\
-			((uint8_t*)dest)[2*i2+2]= V;\
-			((uint8_t*)dest)[2*i2+3]= Y2;\
-		}		\
-		break;\
-	}\
+        }\
+        break;\
+    case PIX_FMT_YUYV422:\
+        func2\
+            ((uint8_t*)dest)[2*i2+0]= Y1;\
+            ((uint8_t*)dest)[2*i2+1]= U;\
+            ((uint8_t*)dest)[2*i2+2]= Y2;\
+            ((uint8_t*)dest)[2*i2+3]= V;\
+        }                \
+        break;\
+    case PIX_FMT_UYVY422:\
+        func2\
+            ((uint8_t*)dest)[2*i2+0]= U;\
+            ((uint8_t*)dest)[2*i2+1]= Y1;\
+            ((uint8_t*)dest)[2*i2+2]= V;\
+            ((uint8_t*)dest)[2*i2+3]= Y2;\
+        }                \
+        break;\
+    }\
 
 
 static inline void yuv2packedXinC(SwsContext *c, int16_t *lumFilter, int16_t **lumSrc, int lumFilterSize,
-				    int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
-				    uint8_t *dest, int dstW, int y)
+                                  int16_t *chrFilter, int16_t **chrSrc, int chrFilterSize,
+                                  uint8_t *dest, int dstW, int y)
 {
-	int i;
-	switch(c->dstFormat)
-	{
-	case PIX_FMT_BGR32:
-	case PIX_FMT_RGB32:
-		YSCALE_YUV_2_RGBX_C(uint32_t)
-			((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];
-			((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];
-		}
-		break;
-	case PIX_FMT_RGB24:
-		YSCALE_YUV_2_RGBX_C(uint8_t)
-			((uint8_t*)dest)[0]= r[Y1];
-			((uint8_t*)dest)[1]= g[Y1];
-			((uint8_t*)dest)[2]= b[Y1];
-			((uint8_t*)dest)[3]= r[Y2];
-			((uint8_t*)dest)[4]= g[Y2];
-			((uint8_t*)dest)[5]= b[Y2];
-			dest+=6;
-		}
-		break;
-	case PIX_FMT_BGR24:
-		YSCALE_YUV_2_RGBX_C(uint8_t)
-			((uint8_t*)dest)[0]= b[Y1];
-			((uint8_t*)dest)[1]= g[Y1];
-			((uint8_t*)dest)[2]= r[Y1];
-			((uint8_t*)dest)[3]= b[Y2];
-			((uint8_t*)dest)[4]= g[Y2];
-			((uint8_t*)dest)[5]= r[Y2];
-			dest+=6;
-		}
-		break;
-	case PIX_FMT_RGB565:
-	case PIX_FMT_BGR565:
-		{
-			const int dr1= dither_2x2_8[y&1    ][0];
-			const int dg1= dither_2x2_4[y&1    ][0];
-			const int db1= dither_2x2_8[(y&1)^1][0];
-			const int dr2= dither_2x2_8[y&1    ][1];
-			const int dg2= dither_2x2_4[y&1    ][1];
-			const int db2= dither_2x2_8[(y&1)^1][1];
-			YSCALE_YUV_2_RGBX_C(uint16_t)
-				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
-				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
-			}
-		}
-		break;
-	case PIX_FMT_RGB555:
-	case PIX_FMT_BGR555:
-		{
-			const int dr1= dither_2x2_8[y&1    ][0];
-			const int dg1= dither_2x2_8[y&1    ][1];
-			const int db1= dither_2x2_8[(y&1)^1][0];
-			const int dr2= dither_2x2_8[y&1    ][1];
-			const int dg2= dither_2x2_8[y&1    ][0];
-			const int db2= dither_2x2_8[(y&1)^1][1];
-			YSCALE_YUV_2_RGBX_C(uint16_t)
-				((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
-				((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
-			}
-		}
-		break;
-	case PIX_FMT_RGB8:
-	case PIX_FMT_BGR8:
-		{
-			const uint8_t * const d64= dither_8x8_73[y&7];
-			const uint8_t * const d32= dither_8x8_32[y&7];
-			YSCALE_YUV_2_RGBX_C(uint8_t)
-				((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];
-				((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];
-			}
-		}
-		break;
-	case PIX_FMT_RGB4:
-	case PIX_FMT_BGR4:
-		{
-			const uint8_t * const d64= dither_8x8_73 [y&7];
-			const uint8_t * const d128=dither_8x8_220[y&7];
-			YSCALE_YUV_2_RGBX_C(uint8_t)
-				((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]
-				                  +((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);
-			}
-		}
-		break;
-	case PIX_FMT_RGB4_BYTE:
-	case PIX_FMT_BGR4_BYTE:
-		{
-			const uint8_t * const d64= dither_8x8_73 [y&7];
-			const uint8_t * const d128=dither_8x8_220[y&7];
-			YSCALE_YUV_2_RGBX_C(uint8_t)
-				((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];
-				((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];
-			}
-		}
-		break;
-	case PIX_FMT_MONOBLACK:
-		{
-			const uint8_t * const d128=dither_8x8_220[y&7];
-			uint8_t *g= c->table_gU[128] + c->table_gV[128];
-			int acc=0;
-			for(i=0; i<dstW-1; i+=2){
-				int j;
-				int Y1=1<<18;
-				int Y2=1<<18;
-
-				for(j=0; j<lumFilterSize; j++)
-				{
-					Y1 += lumSrc[j][i] * lumFilter[j];
-					Y2 += lumSrc[j][i+1] * lumFilter[j];
-				}
-				Y1>>=19;
-				Y2>>=19;
-				if((Y1|Y2)&256)
-				{
-					if(Y1>255)   Y1=255;
-					else if(Y1<0)Y1=0;
-					if(Y2>255)   Y2=255;
-					else if(Y2<0)Y2=0;
-				}
-				acc+= acc + g[Y1+d128[(i+0)&7]];
-				acc+= acc + g[Y2+d128[(i+1)&7]];
-				if((i&7)==6){
-					((uint8_t*)dest)[0]= acc;
-					dest++;
-				}
-			}
-		}
-		break;
-	case PIX_FMT_YUYV422:
-		YSCALE_YUV_2_PACKEDX_C(void)
-			((uint8_t*)dest)[2*i2+0]= Y1;
-			((uint8_t*)dest)[2*i2+1]= U;
-			((uint8_t*)dest)[2*i2+2]= Y2;
-			((uint8_t*)dest)[2*i2+3]= V;
-		}
-                break;
-	case PIX_FMT_UYVY422:
-		YSCALE_YUV_2_PACKEDX_C(void)
-			((uint8_t*)dest)[2*i2+0]= U;
-			((uint8_t*)dest)[2*i2+1]= Y1;
-			((uint8_t*)dest)[2*i2+2]= V;
-			((uint8_t*)dest)[2*i2+3]= Y2;
-		}
-                break;
-	}
+    int i;
+    switch(c->dstFormat)
+    {
+    case PIX_FMT_BGR32:
+    case PIX_FMT_RGB32:
+        YSCALE_YUV_2_RGBX_C(uint32_t)
+            ((uint32_t*)dest)[i2+0]= r[Y1] + g[Y1] + b[Y1];
+            ((uint32_t*)dest)[i2+1]= r[Y2] + g[Y2] + b[Y2];
+        }
+        break;
+    case PIX_FMT_RGB24:
+        YSCALE_YUV_2_RGBX_C(uint8_t)
+            ((uint8_t*)dest)[0]= r[Y1];
+            ((uint8_t*)dest)[1]= g[Y1];
+            ((uint8_t*)dest)[2]= b[Y1];
+            ((uint8_t*)dest)[3]= r[Y2];
+            ((uint8_t*)dest)[4]= g[Y2];
+            ((uint8_t*)dest)[5]= b[Y2];
+            dest+=6;
+        }
+        break;
+    case PIX_FMT_BGR24:
+        YSCALE_YUV_2_RGBX_C(uint8_t)
+            ((uint8_t*)dest)[0]= b[Y1];
+            ((uint8_t*)dest)[1]= g[Y1];
+            ((uint8_t*)dest)[2]= r[Y1];
+            ((uint8_t*)dest)[3]= b[Y2];
+            ((uint8_t*)dest)[4]= g[Y2];
+            ((uint8_t*)dest)[5]= r[Y2];
+            dest+=6;
+        }
+        break;
+    case PIX_FMT_RGB565:
+    case PIX_FMT_BGR565:
+        {
+            const int dr1= dither_2x2_8[y&1    ][0];
+            const int dg1= dither_2x2_4[y&1    ][0];
+            const int db1= dither_2x2_8[(y&1)^1][0];
+            const int dr2= dither_2x2_8[y&1    ][1];
+            const int dg2= dither_2x2_4[y&1    ][1];
+            const int db2= dither_2x2_8[(y&1)^1][1];
+            YSCALE_YUV_2_RGBX_C(uint16_t)
+                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
+                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
+            }
+        }
+        break;
+    case PIX_FMT_RGB555:
+    case PIX_FMT_BGR555:
+        {
+            const int dr1= dither_2x2_8[y&1    ][0];
+            const int dg1= dither_2x2_8[y&1    ][1];
+            const int db1= dither_2x2_8[(y&1)^1][0];
+            const int dr2= dither_2x2_8[y&1    ][1];
+            const int dg2= dither_2x2_8[y&1    ][0];
+            const int db2= dither_2x2_8[(y&1)^1][1];
+            YSCALE_YUV_2_RGBX_C(uint16_t)
+                ((uint16_t*)dest)[i2+0]= r[Y1+dr1] + g[Y1+dg1] + b[Y1+db1];
+                ((uint16_t*)dest)[i2+1]= r[Y2+dr2] + g[Y2+dg2] + b[Y2+db2];
+            }
+        }
+        break;
+    case PIX_FMT_RGB8:
+    case PIX_FMT_BGR8:
+        {
+            const uint8_t * const d64= dither_8x8_73[y&7];
+            const uint8_t * const d32= dither_8x8_32[y&7];
+            YSCALE_YUV_2_RGBX_C(uint8_t)
+                ((uint8_t*)dest)[i2+0]= r[Y1+d32[(i2+0)&7]] + g[Y1+d32[(i2+0)&7]] + b[Y1+d64[(i2+0)&7]];
+                ((uint8_t*)dest)[i2+1]= r[Y2+d32[(i2+1)&7]] + g[Y2+d32[(i2+1)&7]] + b[Y2+d64[(i2+1)&7]];
+            }
+        }
+        break;
+    case PIX_FMT_RGB4:
+    case PIX_FMT_BGR4:
+        {
+            const uint8_t * const d64= dither_8x8_73 [y&7];
+            const uint8_t * const d128=dither_8x8_220[y&7];
+            YSCALE_YUV_2_RGBX_C(uint8_t)
+                ((uint8_t*)dest)[i]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]]
+                                  +((r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]])<<4);
+            }
+        }
+        break;
+    case PIX_FMT_RGB4_BYTE:
+    case PIX_FMT_BGR4_BYTE:
+        {
+            const uint8_t * const d64= dither_8x8_73 [y&7];
+            const uint8_t * const d128=dither_8x8_220[y&7];
+            YSCALE_YUV_2_RGBX_C(uint8_t)
+                ((uint8_t*)dest)[i2+0]= r[Y1+d128[(i2+0)&7]] + g[Y1+d64[(i2+0)&7]] + b[Y1+d128[(i2+0)&7]];
+                ((uint8_t*)dest)[i2+1]= r[Y2+d128[(i2+1)&7]] + g[Y2+d64[(i2+1)&7]] + b[Y2+d128[(i2+1)&7]];
+            }
+        }
+        break;
+    case PIX_FMT_MONOBLACK:
+        {
+            const uint8_t * const d128=dither_8x8_220[y&7];
+            uint8_t *g= c->table_gU[128] + c->table_gV[128];
+            int acc=0;
+            for (i=0; i<dstW-1; i+=2){
+                int j;
+                int Y1=1<<18;
+                int Y2=1<<18;
+
+                for (j=0; j<lumFilterSize; j++)
+                {
+                    Y1 += lumSrc[j][i] * lumFilter[j];
+                    Y2 += lumSrc[j][i+1] * lumFilter[j];
+                }
+                Y1>>=19;
+                Y2>>=19;
+                if ((Y1|Y2)&256)
+                {
+                    if (Y1>255)   Y1=255;
+                    else if (Y1<0)Y1=0;
+                    if (Y2>255)   Y2=255;
+                    else if (Y2<0)Y2=0;
+                }
+                acc+= acc + g[Y1+d128[(i+0)&7]];
+                acc+= acc + g[Y2+d128[(i+1)&7]];
+                if ((i&7)==6){
+                    ((uint8_t*)dest)[0]= acc;
+                    dest++;
+                }
+            }
+        }
+        break;
+    case PIX_FMT_YUYV422:
+        YSCALE_YUV_2_PACKEDX_C(void)
+            ((uint8_t*)dest)[2*i2+0]= Y1;
+            ((uint8_t*)dest)[2*i2+1]= U;
+            ((uint8_t*)dest)[2*i2+2]= Y2;
+            ((uint8_t*)dest)[2*i2+3]= V;
+        }
+        break;
+    case PIX_FMT_UYVY422:
+        YSCALE_YUV_2_PACKEDX_C(void)
+            ((uint8_t*)dest)[2*i2+0]= U;
+            ((uint8_t*)dest)[2*i2+1]= Y1;
+            ((uint8_t*)dest)[2*i2+2]= V;
+            ((uint8_t*)dest)[2*i2+3]= Y2;
+        }
+        break;
+    }
 }
 
 
@@ -902,572 +902,572 @@ static inline void yuv2packedXinC(SwsContext *c, int16_t *lumFilter, int16_t **l
 
 static double getSplineCoeff(double a, double b, double c, double d, double dist)
 {
-//	printf("%f %f %f %f %f\n", a,b,c,d,dist);
-	if(dist<=1.0) 	return ((d*dist + c)*dist + b)*dist +a;
-	else		return getSplineCoeff(	0.0,
-						 b+ 2.0*c + 3.0*d,
-						        c + 3.0*d,
-						-b- 3.0*c - 6.0*d,
-						dist-1.0);
+//    printf("%f %f %f %f %f\n", a,b,c,d,dist);
+    if (dist<=1.0)      return ((d*dist + c)*dist + b)*dist +a;
+    else                return getSplineCoeff(        0.0,
+                                             b+ 2.0*c + 3.0*d,
+                                                    c + 3.0*d,
+                                            -b- 3.0*c - 6.0*d,
+                                            dist-1.0);
 }
 
 static inline int initFilter(int16_t **outFilter, int16_t **filterPos, int *outFilterSize, int xInc,
-			      int srcW, int dstW, int filterAlign, int one, int flags,
-			      SwsVector *srcFilter, SwsVector *dstFilter, double param[2])
+                             int srcW, int dstW, int filterAlign, int one, int flags,
+                             SwsVector *srcFilter, SwsVector *dstFilter, double param[2])
 {
-	int i;
-	int filterSize;
-	int filter2Size;
-	int minFilterSize;
-	double *filter=NULL;
-	double *filter2=NULL;
+    int i;
+    int filterSize;
+    int filter2Size;
+    int minFilterSize;
+    double *filter=NULL;
+    double *filter2=NULL;
 #if defined(ARCH_X86)
-	if(flags & SWS_CPU_CAPS_MMX)
-		asm volatile("emms\n\t"::: "memory"); //FIXME this shouldnt be required but it IS (even for non mmx versions)
+    if (flags & SWS_CPU_CAPS_MMX)
+        asm volatile("emms\n\t"::: "memory"); //FIXME this shouldnt be required but it IS (even for non mmx versions)
 #endif
 
-	// Note the +1 is for the MMXscaler which reads over the end
-	*filterPos = av_malloc((dstW+1)*sizeof(int16_t));
-
-	if(FFABS(xInc - 0x10000) <10) // unscaled
-	{
-		int i;
-		filterSize= 1;
-		filter= av_malloc(dstW*sizeof(double)*filterSize);
-		for(i=0; i<dstW*filterSize; i++) filter[i]=0;
-
-		for(i=0; i<dstW; i++)
-		{
-			filter[i*filterSize]=1;
-			(*filterPos)[i]=i;
-		}
-
-	}
-	else if(flags&SWS_POINT) // lame looking point sampling mode
-	{
-		int i;
-		int xDstInSrc;
-		filterSize= 1;
-		filter= av_malloc(dstW*sizeof(double)*filterSize);
-
-		xDstInSrc= xInc/2 - 0x8000;
-		for(i=0; i<dstW; i++)
-		{
-			int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;
-
-			(*filterPos)[i]= xx;
-			filter[i]= 1.0;
-			xDstInSrc+= xInc;
-		}
-	}
-	else if((xInc <= (1<<16) && (flags&SWS_AREA)) || (flags&SWS_FAST_BILINEAR)) // bilinear upscale
-	{
-		int i;
-		int xDstInSrc;
-		if     (flags&SWS_BICUBIC) filterSize= 4;
-		else if(flags&SWS_X      ) filterSize= 4;
-		else			   filterSize= 2; // SWS_BILINEAR / SWS_AREA
-		filter= av_malloc(dstW*sizeof(double)*filterSize);
-
-		xDstInSrc= xInc/2 - 0x8000;
-		for(i=0; i<dstW; i++)
-		{
-			int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;
-			int j;
-
-			(*filterPos)[i]= xx;
-				//Bilinear upscale / linear interpolate / Area averaging
-				for(j=0; j<filterSize; j++)
-				{
-					double d= FFABS((xx<<16) - xDstInSrc)/(double)(1<<16);
-					double coeff= 1.0 - d;
-					if(coeff<0) coeff=0;
-					filter[i*filterSize + j]= coeff;
-					xx++;
-				}
-			xDstInSrc+= xInc;
-		}
-	}
-	else
-	{
-		double xDstInSrc;
-		double sizeFactor, filterSizeInSrc;
-		const double xInc1= (double)xInc / (double)(1<<16);
-
-		if     (flags&SWS_BICUBIC)	sizeFactor= 4.0;
-		else if(flags&SWS_X)		sizeFactor= 8.0;
-		else if(flags&SWS_AREA)		sizeFactor= 1.0; //downscale only, for upscale it is bilinear
-		else if(flags&SWS_GAUSS)	sizeFactor= 8.0;   // infinite ;)
-		else if(flags&SWS_LANCZOS)	sizeFactor= param[0] != SWS_PARAM_DEFAULT ? 2.0*param[0] : 6.0;
-		else if(flags&SWS_SINC)		sizeFactor= 20.0; // infinite ;)
-		else if(flags&SWS_SPLINE)	sizeFactor= 20.0;  // infinite ;)
-		else if(flags&SWS_BILINEAR)	sizeFactor= 2.0;
-		else {
-			sizeFactor= 0.0; //GCC warning killer
-			ASSERT(0)
-		}
-
-		if(xInc1 <= 1.0)	filterSizeInSrc= sizeFactor; // upscale
-		else			filterSizeInSrc= sizeFactor*srcW / (double)dstW;
-
-		filterSize= (int)ceil(1 + filterSizeInSrc); // will be reduced later if possible
-		if(filterSize > srcW-2) filterSize=srcW-2;
-
-		filter= av_malloc(dstW*sizeof(double)*filterSize);
-
-		xDstInSrc= xInc1 / 2.0 - 0.5;
-		for(i=0; i<dstW; i++)
-		{
-			int xx= (int)(xDstInSrc - (filterSize-1)*0.5 + 0.5);
-			int j;
-			(*filterPos)[i]= xx;
-			for(j=0; j<filterSize; j++)
-			{
-				double d= FFABS(xx - xDstInSrc)/filterSizeInSrc*sizeFactor;
-				double coeff;
-				if(flags & SWS_BICUBIC)
-				{
-					double B= param[0] != SWS_PARAM_DEFAULT ? param[0] : 0.0;
-					double C= param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6;
-
-					if(d<1.0)
-						coeff = (12-9*B-6*C)*d*d*d + (-18+12*B+6*C)*d*d + 6-2*B;
-					else if(d<2.0)
-						coeff = (-B-6*C)*d*d*d + (6*B+30*C)*d*d + (-12*B-48*C)*d +8*B+24*C;
-					else
-						coeff=0.0;
-				}
-/*				else if(flags & SWS_X)
-				{
-					double p= param ? param*0.01 : 0.3;
-					coeff = d ? sin(d*PI)/(d*PI) : 1.0;
-					coeff*= pow(2.0, - p*d*d);
-				}*/
-				else if(flags & SWS_X)
-				{
-					double A= param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
-
-					if(d<1.0)
-						coeff = cos(d*PI);
-					else
-						coeff=-1.0;
-					if(coeff<0.0) 	coeff= -pow(-coeff, A);
-					else		coeff=  pow( coeff, A);
-					coeff= coeff*0.5 + 0.5;
-				}
-				else if(flags & SWS_AREA)
-				{
-					double srcPixelSize= 1.0/xInc1;
-					if(d + srcPixelSize/2 < 0.5) coeff= 1.0;
-					else if(d - srcPixelSize/2 < 0.5) coeff= (0.5-d)/srcPixelSize + 0.5;
-					else coeff=0.0;
-				}
-				else if(flags & SWS_GAUSS)
-				{
-					double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
-					coeff = pow(2.0, - p*d*d);
-				}
-				else if(flags & SWS_SINC)
-				{
-					coeff = d ? sin(d*PI)/(d*PI) : 1.0;
-				}
-				else if(flags & SWS_LANCZOS)
-				{
-					double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
-					coeff = d ? sin(d*PI)*sin(d*PI/p)/(d*d*PI*PI/p) : 1.0;
-					if(d>p) coeff=0;
-				}
-				else if(flags & SWS_BILINEAR)
-				{
-					coeff= 1.0 - d;
-					if(coeff<0) coeff=0;
-				}
-				else if(flags & SWS_SPLINE)
-				{
-					double p=-2.196152422706632;
-					coeff = getSplineCoeff(1.0, 0.0, p, -p-1.0, d);
-				}
-				else {
-					coeff= 0.0; //GCC warning killer
-					ASSERT(0)
-				}
-
-				filter[i*filterSize + j]= coeff;
-				xx++;
-			}
-			xDstInSrc+= xInc1;
-		}
-	}
-
-	/* apply src & dst Filter to filter -> filter2
-	   av_free(filter);
-	*/
-	ASSERT(filterSize>0)
-	filter2Size= filterSize;
-	if(srcFilter) filter2Size+= srcFilter->length - 1;
-	if(dstFilter) filter2Size+= dstFilter->length - 1;
-	ASSERT(filter2Size>0)
-	filter2= av_malloc(filter2Size*dstW*sizeof(double));
-
-	for(i=0; i<dstW; i++)
-	{
-		int j;
-		SwsVector scaleFilter;
-		SwsVector *outVec;
-
-		scaleFilter.coeff= filter + i*filterSize;
-		scaleFilter.length= filterSize;
-
-		if(srcFilter) outVec= sws_getConvVec(srcFilter, &scaleFilter);
-		else	      outVec= &scaleFilter;
-
-		ASSERT(outVec->length == filter2Size)
-		//FIXME dstFilter
-
-		for(j=0; j<outVec->length; j++)
-		{
-			filter2[i*filter2Size + j]= outVec->coeff[j];
-		}
-
-		(*filterPos)[i]+= (filterSize-1)/2 - (filter2Size-1)/2;
-
-		if(outVec != &scaleFilter) sws_freeVec(outVec);
-	}
-	av_free(filter); filter=NULL;
-
-	/* try to reduce the filter-size (step1 find size and shift left) */
-	// Assume its near normalized (*0.5 or *2.0 is ok but * 0.001 is not)
-	minFilterSize= 0;
-	for(i=dstW-1; i>=0; i--)
-	{
-		int min= filter2Size;
-		int j;
-		double cutOff=0.0;
-
-		/* get rid off near zero elements on the left by shifting left */
-		for(j=0; j<filter2Size; j++)
-		{
-			int k;
-			cutOff += FFABS(filter2[i*filter2Size]);
-
-			if(cutOff > SWS_MAX_REDUCE_CUTOFF) break;
-
-			/* preserve Monotonicity because the core can't handle the filter otherwise */
-			if(i<dstW-1 && (*filterPos)[i] >= (*filterPos)[i+1]) break;
-
-			// Move filter coeffs left
-			for(k=1; k<filter2Size; k++)
-				filter2[i*filter2Size + k - 1]= filter2[i*filter2Size + k];
-			filter2[i*filter2Size + k - 1]= 0.0;
-			(*filterPos)[i]++;
-		}
-
-		cutOff=0.0;
-		/* count near zeros on the right */
-		for(j=filter2Size-1; j>0; j--)
-		{
-			cutOff += FFABS(filter2[i*filter2Size + j]);
-
-			if(cutOff > SWS_MAX_REDUCE_CUTOFF) break;
-			min--;
-		}
-
-		if(min>minFilterSize) minFilterSize= min;
-	}
-
-        if (flags & SWS_CPU_CAPS_ALTIVEC) {
-          // we can handle the special case 4,
-          // so we don't want to go to the full 8
-          if (minFilterSize < 5)
+    // Note the +1 is for the MMXscaler which reads over the end
+    *filterPos = av_malloc((dstW+1)*sizeof(int16_t));
+
+    if (FFABS(xInc - 0x10000) <10) // unscaled
+    {
+        int i;
+        filterSize= 1;
+        filter= av_malloc(dstW*sizeof(double)*filterSize);
+        for (i=0; i<dstW*filterSize; i++) filter[i]=0;
+
+        for (i=0; i<dstW; i++)
+        {
+            filter[i*filterSize]=1;
+            (*filterPos)[i]=i;
+        }
+
+    }
+    else if (flags&SWS_POINT) // lame looking point sampling mode
+    {
+        int i;
+        int xDstInSrc;
+        filterSize= 1;
+        filter= av_malloc(dstW*sizeof(double)*filterSize);
+
+        xDstInSrc= xInc/2 - 0x8000;
+        for (i=0; i<dstW; i++)
+        {
+            int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;
+
+            (*filterPos)[i]= xx;
+            filter[i]= 1.0;
+            xDstInSrc+= xInc;
+        }
+    }
+    else if ((xInc <= (1<<16) && (flags&SWS_AREA)) || (flags&SWS_FAST_BILINEAR)) // bilinear upscale
+    {
+        int i;
+        int xDstInSrc;
+        if      (flags&SWS_BICUBIC) filterSize= 4;
+        else if (flags&SWS_X      ) filterSize= 4;
+        else                        filterSize= 2; // SWS_BILINEAR / SWS_AREA
+        filter= av_malloc(dstW*sizeof(double)*filterSize);
+
+        xDstInSrc= xInc/2 - 0x8000;
+        for (i=0; i<dstW; i++)
+        {
+            int xx= (xDstInSrc - ((filterSize-1)<<15) + (1<<15))>>16;
+            int j;
+
+            (*filterPos)[i]= xx;
+                //Bilinear upscale / linear interpolate / Area averaging
+                for (j=0; j<filterSize; j++)
+                {
+                    double d= FFABS((xx<<16) - xDstInSrc)/(double)(1<<16);
+                    double coeff= 1.0 - d;
+                    if (coeff<0) coeff=0;
+                    filter[i*filterSize + j]= coeff;
+                    xx++;
+                }
+            xDstInSrc+= xInc;
+        }
+    }
+    else
+    {
+        double xDstInSrc;
+        double sizeFactor, filterSizeInSrc;
+        const double xInc1= (double)xInc / (double)(1<<16);
+
+        if      (flags&SWS_BICUBIC)      sizeFactor=  4.0;
+        else if (flags&SWS_X)            sizeFactor=  8.0;
+        else if (flags&SWS_AREA)         sizeFactor=  1.0; //downscale only, for upscale it is bilinear
+        else if (flags&SWS_GAUSS)        sizeFactor=  8.0;   // infinite ;)
+        else if (flags&SWS_LANCZOS)      sizeFactor= param[0] != SWS_PARAM_DEFAULT ? 2.0*param[0] : 6.0;
+        else if (flags&SWS_SINC)         sizeFactor= 20.0; // infinite ;)
+        else if (flags&SWS_SPLINE)       sizeFactor= 20.0;  // infinite ;)
+        else if (flags&SWS_BILINEAR)     sizeFactor=  2.0;
+        else {
+            sizeFactor= 0.0; //GCC warning killer
+            ASSERT(0)
+        }
+
+        if (xInc1 <= 1.0)       filterSizeInSrc= sizeFactor; // upscale
+        else                    filterSizeInSrc= sizeFactor*srcW / (double)dstW;
+
+        filterSize= (int)ceil(1 + filterSizeInSrc); // will be reduced later if possible
+        if (filterSize > srcW-2) filterSize=srcW-2;
+
+        filter= av_malloc(dstW*sizeof(double)*filterSize);
+
+        xDstInSrc= xInc1 / 2.0 - 0.5;
+        for (i=0; i<dstW; i++)
+        {
+            int xx= (int)(xDstInSrc - (filterSize-1)*0.5 + 0.5);
+            int j;
+            (*filterPos)[i]= xx;
+            for (j=0; j<filterSize; j++)
+            {
+                double d= FFABS(xx - xDstInSrc)/filterSizeInSrc*sizeFactor;
+                double coeff;
+                if (flags & SWS_BICUBIC)
+                {
+                    double B= param[0] != SWS_PARAM_DEFAULT ? param[0] : 0.0;
+                    double C= param[1] != SWS_PARAM_DEFAULT ? param[1] : 0.6;
+
+                    if (d<1.0)
+                        coeff = (12-9*B-6*C)*d*d*d + (-18+12*B+6*C)*d*d + 6-2*B;
+                    else if (d<2.0)
+                        coeff = (-B-6*C)*d*d*d + (6*B+30*C)*d*d + (-12*B-48*C)*d +8*B+24*C;
+                    else
+                        coeff=0.0;
+                }
+/*                else if (flags & SWS_X)
+                {
+                    double p= param ? param*0.01 : 0.3;
+                    coeff = d ? sin(d*PI)/(d*PI) : 1.0;
+                    coeff*= pow(2.0, - p*d*d);
+                }*/
+                else if (flags & SWS_X)
+                {
+                    double A= param[0] != SWS_PARAM_DEFAULT ? param[0] : 1.0;
+
+                    if (d<1.0)
+                        coeff = cos(d*PI);
+                    else
+                        coeff=-1.0;
+                    if (coeff<0.0)      coeff= -pow(-coeff, A);
+                    else                coeff=  pow( coeff, A);
+                    coeff= coeff*0.5 + 0.5;
+                }
+                else if (flags & SWS_AREA)
+                {
+                    double srcPixelSize= 1.0/xInc1;
+                    if      (d + srcPixelSize/2 < 0.5) coeff= 1.0;
+                    else if (d - srcPixelSize/2 < 0.5) coeff= (0.5-d)/srcPixelSize + 0.5;
+                    else coeff=0.0;
+                }
+                else if (flags & SWS_GAUSS)
+                {
+                    double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
+                    coeff = pow(2.0, - p*d*d);
+                }
+                else if (flags & SWS_SINC)
+                {
+                    coeff = d ? sin(d*PI)/(d*PI) : 1.0;
+                }
+                else if (flags & SWS_LANCZOS)
+                {
+                    double p= param[0] != SWS_PARAM_DEFAULT ? param[0] : 3.0;
+                    coeff = d ? sin(d*PI)*sin(d*PI/p)/(d*d*PI*PI/p) : 1.0;
+                    if (d>p) coeff=0;
+                }
+                else if (flags & SWS_BILINEAR)
+                {
+                    coeff= 1.0 - d;
+                    if (coeff<0) coeff=0;
+                }
+                else if (flags & SWS_SPLINE)
+                {
+                    do