postprocessing cleanup:
authorMichael Niedermayer <michaelni@gmx.at>
Mon, 28 Oct 2002 19:31:04 +0000 (19:31 +0000)
committerMichael Niedermayer <michaelni@gmx.at>
Mon, 28 Oct 2002 19:31:04 +0000 (19:31 +0000)
remove opendivx #ifdefs
remove rk1 filter
remove unused / obsolete stuff
add -1,4,2,4,-1 deinterlacing filter (ffmpeg uses that)
threadsafe / no more non-const globals
some optimizations
different strides for Y,U,V possible
remove ebx usage (someone really should fix gcc, this is really lame)
change the dering filter slightly (tell me if its worse for any files)

Originally committed as revision 7947 to svn://svn.mplayerhq.hu/mplayer/trunk/postproc

postproc/Makefile
postproc/postprocess.c
postproc/postprocess.h
postproc/postprocess_template.c

index cf2af02811e583df8d1a2df7c9b64e76f6cc093e..1c2d244ce0f6512c774a3b38580b4a026cbfd5d2 100644 (file)
@@ -1,10 +1,15 @@
 
 include ../config.mak
 
-LIBNAME = libpostproc.a
+SWSLIB = libswscale.a
+SPPLIB = libpostproc.so
+SPPVERSION = 0.0.1
+PPLIB = libpostproc.a
 
-SRCS=postprocess.c swscale.c rgb2rgb.c yuv2rgb.c
-OBJS=$(SRCS:.c=.o)
+SWSSRCS=swscale.c rgb2rgb.c yuv2rgb.c
+SWSOBJS=$(SWSSRCS:.c=.o)
+PPOBJS=postprocess.o
+SPPOBJS=postprocess_pic.o
 CS_TEST_OBJS=cs_test.o rgb2rgb.o ../cpudetect.o ../mp_msg.o
 
 CFLAGS  = $(OPTFLAGS) $(MLIB_INC) -I. -I.. $(EXTRA_INC)
@@ -17,16 +22,16 @@ CFLAGS  = $(OPTFLAGS) $(MLIB_INC) -I. -I.. $(EXTRA_INC)
 .c.o:
        $(CC) -c $(CFLAGS) -o $@ $<
 
-$(LIBNAME):     $(OBJS)
-       $(AR) r $(LIBNAME) $(OBJS)
+all:    $(SWSLIB) $(PPLIB) $(SPPLIB)
 
-all:    $(LIBNAME)
+$(SWSLIB):     $(SWSOBJS)
+       $(AR) r $(SWSLIB) $(SWSOBJS)
 
 clean:
-       rm -f *.o *.a *~
+       rm -f *.o *.a *~ *.so
 
 distclean:
-       rm -f Makefile.bak *.o *.a *~ .depend
+       rm -f Makefile.bak *.o *.a *~ *.so .depend
 
 dep:    depend
 
@@ -36,6 +41,27 @@ depend:
 cs_test: $(CS_TEST_OBJS)
        $(CC) $(CS_TEST_OBJS) -o cs_test
 
+postprocess_pic.o: postprocess.c
+       $(CC) -c $(CFLAGS) -fPIC -DPIC -o $@ $<
+
+$(SPPLIB): $(SPPOBJS)
+       $(CC) -shared -Wl,-soname,$(SPPLIB).0 \
+       -o $(SPPLIB) $(SPPOBJS)
+
+$(PPLIB): $(PPOBJS)
+       $(AR) r $(PPLIB) $(PPOBJS)
+
+install: all
+ifeq ($(SHARED_PP),yes)
+       install -d $(prefix)/lib
+       install -s -m 755 $(SPPLIB) $(prefix)/lib/$(SPPLIB).$(SPPVERSION)
+       ln -sf $(SPPLIB).$(SPPVERSION) $(prefix)/lib/$(SPPLIB)
+       ldconfig || true
+       mkdir -p $(prefix)/include/postproc
+       install -m 644 postprocess.h $(prefix)/include/postproc/postprocess.h
+endif
+
+       
 #
 # include dependency files if they exist
 #
index 15399d48a88111bec917be8e45e16cbc50ada85c..a732d400a41fe055fa2d7f8a73803e458d6b77d2 100644 (file)
@@ -49,7 +49,6 @@ c = checked against the other implementations (-vo md5)
 TODO:
 remove global/static vars
 reduce the time wasted on the mem transfer
-implement everything in C at least (done at the moment but ...)
 unroll stuff if instructions depend too much on the prior one
 we use 8x8 blocks for the horizontal filters, opendivx seems to use 8x4?
 move YScale thing to the end instead of fixing QP
@@ -58,13 +57,9 @@ make the mainloop more flexible (variable number of blocks at once
        (the if/else stuff per block is slowing things down)
 compare the quality & speed of all filters
 split this huge file
-border remover
 optimize c versions
 try to unroll inner for(x=0 ... loop to avoid these damn if(x ... checks
-smart blur
-commandline option for   the deblock / dering thresholds
 put fastmemcpy back
-dont use #ifdef ARCH_X86 for the asm stuff ... cross compilers? (note cpudetect uses ARCH_X86)
 ...
 */
 
@@ -84,7 +79,7 @@ dont use #ifdef ARCH_X86 for the asm stuff ... cross compilers? (note cpudetect
 //#undef HAVE_MMX
 //#undef ARCH_X86
 //#define DEBUG_BRIGHTNESS
-//#include "../libvo/fastmemcpy.h"
+#include "../libvo/fastmemcpy.h"
 #include "postprocess.h"
 #include "../cpudetect.h"
 #include "../mangle.h"
@@ -96,79 +91,83 @@ dont use #ifdef ARCH_X86 for the asm stuff ... cross compilers? (note cpudetect
 
 #define GET_MODE_BUFFER_SIZE 500
 #define OPTIONS_ARRAY_SIZE 10
+#define BLOCK_SIZE 8
+#define TEMP_STRIDE 8
+//#define NUM_BLOCKS_AT_ONCE 16 //not used yet
 
 #ifdef ARCH_X86
-#define CAN_COMPILE_X86_ASM
-#endif
-
-#ifdef CAN_COMPILE_X86_ASM
-static volatile uint64_t __attribute__((aligned(8))) packedYOffset=    0x0000000000000000LL;
-static volatile uint64_t __attribute__((aligned(8))) packedYScale=     0x0100010001000100LL;
 static uint64_t __attribute__((aligned(8))) w05=               0x0005000500050005LL;
 static uint64_t __attribute__((aligned(8))) w20=               0x0020002000200020LL;
-static uint64_t __attribute__((aligned(8))) w1400=             0x1400140014001400LL;
-static uint64_t __attribute__((aligned(8))) bm00000001=                0x00000000000000FFLL;
-static uint64_t __attribute__((aligned(8))) bm00010000=                0x000000FF00000000LL;
-static uint64_t __attribute__((aligned(8))) bm00001000=                0x00000000FF000000LL;
-static uint64_t __attribute__((aligned(8))) bm10000000=                0xFF00000000000000LL;
-static uint64_t __attribute__((aligned(8))) bm10000001=                0xFF000000000000FFLL;
-static uint64_t __attribute__((aligned(8))) bm11000011=                0xFFFF00000000FFFFLL;
-static uint64_t __attribute__((aligned(8))) bm00000011=                0x000000000000FFFFLL;
-static uint64_t __attribute__((aligned(8))) bm11111110=                0xFFFFFFFFFFFFFF00LL;
-static uint64_t __attribute__((aligned(8))) bm11000000=                0xFFFF000000000000LL;
-static uint64_t __attribute__((aligned(8))) bm00011000=                0x000000FFFF000000LL;
-static uint64_t __attribute__((aligned(8))) bm00110011=                0x0000FFFF0000FFFFLL;
-static uint64_t __attribute__((aligned(8))) bm11001100=                0xFFFF0000FFFF0000LL;
 static uint64_t __attribute__((aligned(8))) b00=               0x0000000000000000LL;
 static uint64_t __attribute__((aligned(8))) b01=               0x0101010101010101LL;
 static uint64_t __attribute__((aligned(8))) b02=               0x0202020202020202LL;
-static uint64_t __attribute__((aligned(8))) b0F=               0x0F0F0F0F0F0F0F0FLL;
-static uint64_t __attribute__((aligned(8))) b04=               0x0404040404040404LL;
 static uint64_t __attribute__((aligned(8))) b08=               0x0808080808080808LL;
-static uint64_t __attribute__((aligned(8))) bFF=               0xFFFFFFFFFFFFFFFFLL;
-static uint64_t __attribute__((aligned(8))) b20=               0x2020202020202020LL;
 static uint64_t __attribute__((aligned(8))) b80=               0x8080808080808080LL;
-static uint64_t __attribute__((aligned(8))) mmxDCOffset=       0x7E7E7E7E7E7E7E7ELL;
-static uint64_t __attribute__((aligned(8))) mmxDCThreshold=    0x7C7C7C7C7C7C7C7CLL;
-static uint64_t __attribute__((aligned(8))) b3F=               0x3F3F3F3F3F3F3F3FLL;
-static uint64_t __attribute__((aligned(8))) temp0=0;
-static uint64_t __attribute__((aligned(8))) temp1=0;
-static uint64_t __attribute__((aligned(8))) temp2=0;
-static uint64_t __attribute__((aligned(8))) temp3=0;
-static uint64_t __attribute__((aligned(8))) temp4=0;
-static uint64_t __attribute__((aligned(8))) temp5=0;
-static uint64_t __attribute__((aligned(8))) pQPb=0;
-static uint64_t __attribute__((aligned(8))) pQPb2=0;
-static uint8_t __attribute__((aligned(8))) tempBlocks[8*16*2]; //used for the horizontal code
-static uint32_t __attribute__((aligned(4))) maxTmpNoise[4];
-#else
-static uint64_t packedYOffset= 0x0000000000000000LL;
-static uint64_t packedYScale=  0x0100010001000100LL;
 #endif
 
 extern int divx_quality;
-int newPPFlag=0; //is set if -npp is used
-struct PPMode gPPMode[GET_PP_QUALITY_MAX+1];
-static int firstTime = 0, firstTime2 = 0;
+static int firstTime2 = 0;
 
 extern int verbose;
 
-int hFlatnessThreshold= 56 - 16;
-int vFlatnessThreshold= 56 - 16;
 int deringThreshold= 20;
 
-static int dcOffset;
-static int dcThreshold;
-
 //amount of "black" u r willing to loose to get a brightness corrected picture
 double maxClippedThreshold= 0.01;
 
+struct PPFilter{
+       char *shortName;
+       char *longName;
+       int chromDefault;       // is chrominance filtering on by default if this filter is manually activated
+       int minLumQuality;      // minimum quality to turn luminance filtering on
+       int minChromQuality;    // minimum quality to turn chrominance filtering on
+       int mask;               // Bitmask to turn this filter on
+};
+
+typedef struct PPContext{
+       uint8_t *tempBlocks; //used for the horizontal code
+
+       /* we need 64bit here otherwise we´ll going to have a problem
+          after watching a black picture for 5 hours*/
+       uint64_t *yHistogram;
+
+       uint64_t __attribute__((aligned(8))) packedYOffset;
+       uint64_t __attribute__((aligned(8))) packedYScale;
+
+       /* Temporal noise reducing buffers */
+       uint8_t *tempBlured[3];
+       int32_t *tempBluredPast[3];
+
+       /* Temporary buffers for handling the last row(s) */
+       uint8_t *tempDst;
+       uint8_t *tempSrc;
+
+       /* Temporary buffers for handling the last block */
+       uint8_t *tempDstBlock;
+       uint8_t *tempSrcBlock;
+       uint8_t *deintTemp;
+
+       uint64_t __attribute__((aligned(8))) pQPb;
+       uint64_t __attribute__((aligned(8))) pQPb2;
+       
+       uint64_t __attribute__((aligned(8))) mmxDcOffset;
+       uint64_t __attribute__((aligned(8))) mmxDcThreshold;
+
+       int QP;
+       int dcOffset;
+       int dcThreshold;
+
+       int frameNum;
+
+       PPMode ppMode;
+} PPContext;
+
 static struct PPFilter filters[]=
 {
        {"hb", "hdeblock",              1, 1, 3, H_DEBLOCK},
        {"vb", "vdeblock",              1, 2, 4, V_DEBLOCK},
-       {"hr", "rkhdeblock",            1, 1, 3, H_RK1_FILTER},
-       {"vr", "rkvdeblock",            1, 2, 4, V_RK1_FILTER},
+/*     {"hr", "rkhdeblock",            1, 1, 3, H_RK1_FILTER},
+       {"vr", "rkvdeblock",            1, 2, 4, V_RK1_FILTER},*/
        {"h1", "x1hdeblock",            1, 1, 3, H_X1_FILTER},
        {"v1", "x1vdeblock",            1, 2, 4, V_X1_FILTER},
        {"dr", "dering",                1, 5, 6, DERING},
@@ -177,6 +176,7 @@ static struct PPFilter filters[]=
        {"li", "linipoldeint",          1, 1, 4, LINEAR_IPOL_DEINT_FILTER},
        {"ci", "cubicipoldeint",        1, 1, 4, CUBIC_IPOL_DEINT_FILTER},
        {"md", "mediandeint",           1, 1, 4, MEDIAN_DEINT_FILTER},
+       {"fd", "ffmpegdeint",           1, 1, 4, FFMPEG_DEINT_FILTER},
        {"tn", "tmpnoise",              1, 7, 8, TEMP_NOISE_FILTER},
        {"fq", "forcequant",            1, 0, 0, FORCE_QUANT},
        {NULL, NULL,0,0,0,0} //End Marker
@@ -191,19 +191,13 @@ static char *replaceTable[]=
        NULL //End Marker
 };
 
-#ifdef CAN_COMPILE_X86_ASM
+#ifdef ARCH_X86
 static inline void unusedVariableWarningFixer()
 {
-if(
- packedYOffset + packedYScale + w05 + w20 + w1400 + bm00000001 + bm00010000
- + bm00001000 + bm10000000 + bm10000001 + bm11000011 + bm00000011 + bm11111110
- + bm11000000 + bm00011000 + bm00110011 + bm11001100 + b00 + b01 + b02 + b0F
- + bFF + b20 + b04+ b08 + pQPb2 + b80 + mmxDCOffset + mmxDCThreshold + b3F + temp0 + temp1 + temp2 + temp3 + temp4
- + temp5 + pQPb== 0) b00=0;
+       if(w05 + w20 + b00 + b01 + b02 + b08 + b80 == 0) b00=0;
 }
 #endif
 
-#ifdef TIMING
 static inline long long rdtsc()
 {
        long long l;
@@ -213,9 +207,8 @@ static inline long long rdtsc()
 //     printf("%d\n", int(l/1000));
        return l;
 }
-#endif
 
-#ifdef CAN_COMPILE_X86_ASM
+#ifdef ARCH_X86
 static inline void prefetchnta(void *p)
 {
        asm volatile(   "prefetchnta (%0)\n\t"
@@ -250,22 +243,48 @@ static inline void prefetcht2(void *p)
 /**
  * Check if the given 8x8 Block is mostly "flat"
  */
-static inline int isHorizDC(uint8_t src[], int stride)
+static inline int isHorizDC(uint8_t src[], int stride, PPContext *c)
 {
        int numEq= 0;
        int y;
+       const int dcOffset= c->dcOffset;
+       const int dcThreshold= c->dcThreshold;
        for(y=0; y<BLOCK_SIZE; y++)
        {
-               if(((src[0] - src[1] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[1] - src[2] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[2] - src[3] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[3] - src[4] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[4] - src[5] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[5] - src[6] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
-               if(((src[6] - src[7] + dcOffset) & 0xFFFF) < dcThreshold) numEq++;
+               if(((unsigned)(src[0] - src[1] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[1] - src[2] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[2] - src[3] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[3] - src[4] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[4] - src[5] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[5] - src[6] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[6] - src[7] + dcOffset)) < dcThreshold) numEq++;
+               src+= stride;
+       }
+       return numEq > c->ppMode.flatnessThreshold;
+}
+
+/**
+ * Check if the middle 8x8 Block in the given 8x16 block is flat
+ */
+static inline int isVertDC_C(uint8_t src[], int stride, PPContext *c){
+       int numEq= 0;
+       int y;
+       const int dcOffset= c->dcOffset;
+       const int dcThreshold= c->dcThreshold;
+       src+= stride*4; // src points to begin of the 8x8 Block
+       for(y=0; y<BLOCK_SIZE-1; y++)
+       {
+               if(((unsigned)(src[0] - src[0+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[1] - src[1+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[2] - src[2+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[3] - src[3+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[4] - src[4+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[5] - src[5+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[6] - src[6+stride] + dcOffset)) < dcThreshold) numEq++;
+               if(((unsigned)(src[7] - src[7+stride] + dcOffset)) < dcThreshold) numEq++;
                src+= stride;
        }
-       return numEq > hFlatnessThreshold;
+       return numEq > c->ppMode.flatnessThreshold;
 }
 
 static inline int isHorizMinMaxOk(uint8_t src[], int stride, int QP)
@@ -422,7 +441,7 @@ static inline void horizX1Filter(uint8_t *src, int stride, int QP)
 #define COMPILE_C
 #endif
 
-#ifdef CAN_COMPILE_X86_ASM
+#ifdef ARCH_X86
 
 #if (defined (HAVE_MMX) && !defined (HAVE_3DNOW) && !defined (HAVE_MMX2)) || defined (RUNTIME_CPUDETECT)
 #define COMPILE_MMX
@@ -435,7 +454,7 @@ static inline void horizX1Filter(uint8_t *src, int stride, int QP)
 #if (defined (HAVE_3DNOW) && !defined (HAVE_MMX2)) || defined (RUNTIME_CPUDETECT)
 #define COMPILE_3DNOW
 #endif
-#endif //CAN_COMPILE_X86_ASM
+#endif //ARCH_X86
 
 #undef HAVE_MMX
 #undef HAVE_MMX2
@@ -487,50 +506,47 @@ static inline void horizX1Filter(uint8_t *src, int stride, int QP)
 // minor note: the HAVE_xyz is messed up after that line so dont use it
 
 static inline void postProcess(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height,
-       QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode)
+       QP_STORE_T QPs[], int QPStride, int isColor, PPMode *ppMode, void *vc)
 {
+       PPContext *c= (PPContext *)vc;
+       c->ppMode= *ppMode; //FIXME
+
        // useing ifs here as they are faster than function pointers allthough the
        // difference wouldnt be messureable here but its much better because
        // someone might exchange the cpu whithout restarting mplayer ;)
 #ifdef RUNTIME_CPUDETECT
-#ifdef CAN_COMPILE_X86_ASM
+#ifdef ARCH_X86
        // ordered per speed fasterst first
        if(gCpuCaps.hasMMX2)
-               postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
        else if(gCpuCaps.has3DNow)
-               postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
        else if(gCpuCaps.hasMMX)
-               postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
        else
-               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #else
-               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #endif
 #else //RUNTIME_CPUDETECT
 #ifdef HAVE_MMX2
-               postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_MMX2(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #elif defined (HAVE_3DNOW)
-               postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_3DNow(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #elif defined (HAVE_MMX)
-               postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_MMX(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #else
-               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, ppMode);
+               postProcess_C(src, srcStride, dst, dstStride, width, height, QPs, QPStride, isColor, c);
 #endif
 #endif //!RUNTIME_CPUDETECT
 }
 
-#ifdef HAVE_ODIVX_POSTPROCESS
-#include "../opendivx/postprocess.h"
-int use_old_pp=0;
-#endif
-
 //static void postProcess(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height,
 //     QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode);
 
 /* -pp Command line Help
-NOTE/FIXME: put this at an appropriate place (--help, html docs, man mplayer)?
 */
-char *help=
+char *postproc_help=
 "-npp <filterName>[:<option>[:<option>...]][,[-]<filterName>[:<option>...]]...\n"
 "long form example:\n"
 "-npp vdeblock:autoq,hdeblock:autoq,linblenddeint      -npp default,-vdeblock\n"
@@ -549,8 +565,6 @@ char *help=
 "                      the h & v deblocking filters share these\n"
 "                      so u cant set different thresholds for h / v\n"
 "vb    vdeblock        (2 Threshold)           vertical deblocking filter\n"
-"hr    rkhdeblock\n"
-"vr    rkvdeblock\n"
 "h1    x1hdeblock                              Experimental h deblock filter 1\n"
 "v1    x1vdeblock                              Experimental v deblock filter 1\n"
 "dr    dering                                  Deringing filter\n"
@@ -560,6 +574,7 @@ char *help=
 "li    linipoldeint                            linear interpolating deinterlace\n"
 "ci    cubicipoldeint                          cubic interpolating deinterlacer\n"
 "md    mediandeint                             median deinterlacer\n"
+"fd    ffmpegdeint                             ffmpeg deinterlacer\n"
 "de    default                                 hb:a,vb:a,dr:a,al\n"
 "fa    fast                                    h1:a,v1:a,dr:a,al\n"
 "tn    tmpnoise        (3 Thresholds)          Temporal Noise Reducer\n"
@@ -576,11 +591,13 @@ struct PPMode getPPModeByNameAndQuality(char *name, int quality)
 {
        char temp[GET_MODE_BUFFER_SIZE];
        char *p= temp;
-       char *filterDelimiters= ",";
+       char *filterDelimiters= ",/";
        char *optionDelimiters= ":";
-       struct PPMode ppMode= {0,0,0,0,0,0,{150,200,400}};
+       struct PPMode ppMode= {0,0,0,0,0,{150,200,400}};
        char *filterToken;
 
+       ppMode.flatnessThreshold= 56-16;
+       
        strncpy(temp, name, GET_MODE_BUFFER_SIZE);
 
        if(verbose>1) printf("pp: %s\n", name);
@@ -720,8 +737,7 @@ struct PPMode getPPModeByNameAndQuality(char *name, int quality)
 
                                                numOfUnknownOptions--;
                                                if(o==0) ppMode.maxDcDiff= val;
-                                               else hFlatnessThreshold=
-                                                    vFlatnessThreshold= val;
+                                               else ppMode.flatnessThreshold= val;
                                        }
                                }
                                else if(filters[i].mask == FORCE_QUANT)
@@ -745,46 +761,59 @@ struct PPMode getPPModeByNameAndQuality(char *name, int quality)
                ppMode.error += numOfUnknownOptions;
        }
 
-#ifdef HAVE_ODIVX_POSTPROCESS
-       if(ppMode.lumMode & H_DEBLOCK) ppMode.oldMode |= PP_DEBLOCK_Y_H;
-       if(ppMode.lumMode & V_DEBLOCK) ppMode.oldMode |= PP_DEBLOCK_Y_V;
-       if(ppMode.chromMode & H_DEBLOCK) ppMode.oldMode |= PP_DEBLOCK_C_H;
-       if(ppMode.chromMode & V_DEBLOCK) ppMode.oldMode |= PP_DEBLOCK_C_V;
-       if(ppMode.lumMode & DERING) ppMode.oldMode |= PP_DERING_Y;
-       if(ppMode.chromMode & DERING) ppMode.oldMode |= PP_DERING_C;
-#endif
-
        if(verbose>1) printf("pp: lumMode=%X, chromMode=%X\n", ppMode.lumMode, ppMode.chromMode);
        return ppMode;
 }
 
-/**
- * Check and load the -npp part of the cmd line
- */
-int readNPPOpt(void *conf, char *arg)
-{
-       int quality;
-       
-       if(!strcmp("help", arg))
+void *getPPContext(int width, int height){
+       PPContext *c= memalign(32, sizeof(PPContext));
+       int i;
+
+       c->tempBlocks= (uint8_t*)memalign(8, 2*16*8);
+       c->yHistogram= (uint64_t*)memalign(8, 256*sizeof(uint64_t));
+       for(i=0; i<256; i++)
+               c->yHistogram[i]= width*height/64*15/256;
+
+       for(i=0; i<3; i++)
        {
-               printf("%s", help);
-               exit(1);
+               //Note:the +17*1024 is just there so i dont have to worry about r/w over te end
+               c->tempBlured[i]= (uint8_t*)memalign(8, ((width+7)&(~7))*2*((height+7)&(~7)) + 17*1024); //FIXME dstStride instead of width
+               c->tempBluredPast[i]= (uint32_t*)memalign(8, 256*((height+7)&(~7))/2 + 17*1024);
+
+               memset(c->tempBlured[i], 0, ((width+7)&(~7))*2*((height+7)&(~7)) + 17*1024);
+               memset(c->tempBluredPast[i], 0, 256*((height+7)&(~7))/2 + 17*1024);
        }
        
-       for(quality=0; quality<GET_PP_QUALITY_MAX+1; quality++)
-       {
-               gPPMode[quality]= getPPModeByNameAndQuality(arg, quality);
+       c->tempDst= (uint8_t*)memalign(8, 1024*24);
+       c->tempSrc= (uint8_t*)memalign(8, 1024*24);
+       c->tempDstBlock= (uint8_t*)memalign(8, 1024*24);
+       c->tempSrcBlock= (uint8_t*)memalign(8, 1024*24);
+       c->deintTemp= (uint8_t*)memalign(8, width+16);
 
-               if(gPPMode[quality].error) return -1;
-       }
-       newPPFlag=1;
+       c->frameNum=-1;
 
-//divx_quality is passed to postprocess if autoq if off
-       divx_quality= GET_PP_QUALITY_MAX;
-       firstTime = firstTime2 = 1;
-       return 1;
+       return c;
 }
 
+void freePPContext(void *vc){
+       PPContext *c = (PPContext*)vc;
+       int i;
+       
+       for(i=0; i<3; i++) free(c->tempBlured[i]);
+       for(i=0; i<3; i++) free(c->tempBluredPast[i]);
+       
+       free(c->tempBlocks);
+       free(c->yHistogram);
+       free(c->tempDst);
+       free(c->tempSrc);
+       free(c->tempDstBlock);
+       free(c->tempSrcBlock);
+       free(c->deintTemp);
+       
+       free(c);
+}
+
+//FIXME move this shit away from here
 int readPPOpt(void *conf, char *arg)
 {
   int val;
@@ -799,123 +828,31 @@ int readPPOpt(void *conf, char *arg)
     return -3; // ERR_OUT_OF_RANGE
 
   divx_quality = val;
-  firstTime = firstTime2 = 1;
+  firstTime2 = 1;
 
   return 1;
 }
   
 void revertPPOpt(void *conf, char* opt) 
 {
-  newPPFlag=0;
   divx_quality=0;
 }
 
 
-/**
- * Obsolete, dont use it, use postprocess2() instead
- * this will check newPPFlag automatically and use postprocess2 if it is set
- * mode = quality if newPPFlag
- */
-void  postprocess(unsigned char * src[], int src_stride,
-                 unsigned char * dst[], int dst_stride,
-                 int horizontal_size,   int vertical_size,
-                 QP_STORE_T *QP_store,  int QP_stride,
-                                         int mode)
-{
-       struct PPMode ppMode;
-       static QP_STORE_T zeroArray[2048/8];
-
-       if(newPPFlag)
-       {
-               ppMode= gPPMode[mode];
-//             printf("%d \n",QP_store[5]);
-               postprocess2(src, src_stride, dst, dst_stride,
-                       horizontal_size, vertical_size, QP_store, QP_stride, &ppMode);
-
-               return;
-       }
-
-       if(firstTime && verbose)
-       {
-               printf("using pp filters 0x%X\n", mode);
-               firstTime=0;
-       }
-
-       if(QP_store==NULL)
-       {
-               QP_store= zeroArray;
-               QP_stride= 0;
-       }
-
-       ppMode.lumMode= mode;
-       mode= ((mode&0xFF)>>4) | (mode&0xFFFFFF00);
-       ppMode.chromMode= mode;
-       ppMode.maxTmpNoise[0]= 700;
-       ppMode.maxTmpNoise[1]= 1500;
-       ppMode.maxTmpNoise[2]= 3000;
-       ppMode.maxAllowedY= 234;
-       ppMode.minAllowedY= 16;
-       ppMode.maxDcDiff= 1;
-
-#ifdef HAVE_ODIVX_POSTPROCESS
-// Note: I could make this shit outside of this file, but it would mean one
-// more function call...
-       if(use_old_pp){
-           odivx_postprocess(src,src_stride,dst,dst_stride,horizontal_size,vertical_size,QP_store,QP_stride,mode);
-           return;
-       }
-#endif
-
-       postProcess(src[0], src_stride, dst[0], dst_stride,
-               horizontal_size, vertical_size, QP_store, QP_stride, 0, &ppMode);
-
-       horizontal_size >>= 1;
-       vertical_size   >>= 1;
-       src_stride      >>= 1;
-       dst_stride      >>= 1;
-
-       if(ppMode.chromMode)
-       {
-               postProcess(src[1], src_stride, dst[1], dst_stride,
-                       horizontal_size, vertical_size, QP_store, QP_stride, 1, &ppMode);
-               postProcess(src[2], src_stride, dst[2], dst_stride,
-                       horizontal_size, vertical_size, QP_store, QP_stride, 2, &ppMode);
-       }
-       else if(src_stride == dst_stride)
-       {
-               memcpy(dst[1], src[1], src_stride*vertical_size);
-               memcpy(dst[2], src[2], src_stride*vertical_size);
-       }
-       else
-       {
-               int y;
-               for(y=0; y<vertical_size; y++)
-               {
-                       memcpy(&(dst[1][y*dst_stride]), &(src[1][y*src_stride]), horizontal_size);
-                       memcpy(&(dst[2][y*dst_stride]), &(src[2][y*src_stride]), horizontal_size);
-               }
-       }
-
-#if 0
-               memset(dst[1], 128, dst_stride*vertical_size);
-               memset(dst[2], 128, dst_stride*vertical_size);
-#endif
-}
-
-void  postprocess2(unsigned char * src[], int src_stride,
-                 unsigned char * dst[], int dst_stride,
-                 int horizontal_size,   int vertical_size,
-                 QP_STORE_T *QP_store,  int QP_stride,
-                struct PPMode *mode)
+void  postprocess(uint8_t * src[3], int srcStride[3],
+                 uint8_t * dst[3], int dstStride[3],
+                 int horizontalSize, int verticalSize,
+                 QP_STORE_T *QP_store,  int QPStride,
+                PPMode *mode,  void *c)
 {
 
        QP_STORE_T quantArray[2048/8];
-       
+
        if(QP_store==NULL || (mode->lumMode & FORCE_QUANT)) 
        {
                int i;
                QP_store= quantArray;
-               QP_stride= 0;
+               QPStride= 0;
                if(mode->lumMode & FORCE_QUANT)
                        for(i=0; i<2048/8; i++) quantArray[i]= mode->forcedQuant;
                else
@@ -928,89 +865,32 @@ void  postprocess2(unsigned char * src[], int src_stride,
                firstTime2=0;
        }
 
-#ifdef HAVE_ODIVX_POSTPROCESS
-// Note: I could make this shit outside of this file, but it would mean one
-// more function call...
-       if(use_old_pp){
-           odivx_postprocess(src,src_stride,dst,dst_stride,horizontal_size,vertical_size,QP_store,QP_stride,
-           mode->oldMode);
-           return;
-       }
-#endif
-
-       postProcess(src[0], src_stride, dst[0], dst_stride,
-               horizontal_size, vertical_size, QP_store, QP_stride, 0, mode);
+       postProcess(src[0], srcStride[0], dst[0], dstStride[0],
+               horizontalSize, verticalSize, QP_store, QPStride, 0, mode, c);
 
-       horizontal_size >>= 1;
-       vertical_size   >>= 1;
-       src_stride      >>= 1;
-       dst_stride      >>= 1;
+       horizontalSize = (horizontalSize+1)>> 1;
+       verticalSize   = (verticalSize+1)>>1;
 
        if(mode->chromMode)
        {
-               postProcess(src[1], src_stride, dst[1], dst_stride,
-                       horizontal_size, vertical_size, QP_store, QP_stride, 1, mode);
-               postProcess(src[2], src_stride, dst[2], dst_stride,
-                       horizontal_size, vertical_size, QP_store, QP_stride, 2, mode);
+               postProcess(src[1], srcStride[1], dst[1], dstStride[1],
+                       horizontalSize, verticalSize, QP_store, QPStride, 1, mode, c);
+               postProcess(src[2], srcStride[2], dst[2], dstStride[2],
+                       horizontalSize, verticalSize, QP_store, QPStride, 2, mode, c);
        }
-       else if(src_stride == dst_stride)
+       else if(srcStride[1] == dstStride[1] && srcStride[2] == dstStride[2])
        {
-               memcpy(dst[1], src[1], src_stride*vertical_size);
-               memcpy(dst[2], src[2], src_stride*vertical_size);
+               memcpy(dst[1], src[1], srcStride[1]*verticalSize);
+               memcpy(dst[2], src[2], srcStride[2]*verticalSize);
        }
        else
        {
                int y;
-               for(y=0; y<vertical_size; y++)
+               for(y=0; y<verticalSize; y++)
                {
-                       memcpy(&(dst[1][y*dst_stride]), &(src[1][y*src_stride]), horizontal_size);
-                       memcpy(&(dst[2][y*dst_stride]), &(src[2][y*src_stride]), horizontal_size);
+                       memcpy(&(dst[1][y*dstStride[1]]), &(src[1][y*srcStride[1]]), horizontalSize);
+                       memcpy(&(dst[2][y*dstStride[2]]), &(src[2][y*srcStride[2]]), horizontalSize);
                }
        }
 }
 
-
-/**
- * gets the mode flags for a given quality (larger values mean slower but better postprocessing)
- * with -npp it simply returns quality 
- * 0 <= quality <= 6
- */
-int getPpModeForQuality(int quality){
-       int modes[1+GET_PP_QUALITY_MAX]= {
-               0,
-#if 1
-               // horizontal filters first
-               LUM_H_DEBLOCK,
-               LUM_H_DEBLOCK | LUM_V_DEBLOCK,
-               LUM_H_DEBLOCK | LUM_V_DEBLOCK | CHROM_H_DEBLOCK,
-               LUM_H_DEBLOCK | LUM_V_DEBLOCK | CHROM_H_DEBLOCK | CHROM_V_DEBLOCK,
-               LUM_H_DEBLOCK | LUM_V_DEBLOCK | CHROM_H_DEBLOCK | CHROM_V_DEBLOCK | LUM_DERING,
-               LUM_H_DEBLOCK | LUM_V_DEBLOCK | CHROM_H_DEBLOCK | CHROM_V_DEBLOCK | LUM_DERING | CHROM_DERING
-#else
-               // vertical filters first
-               LUM_V_DEBLOCK,
-               LUM_V_DEBLOCK | LUM_H_DEBLOCK,
-               LUM_V_DEBLOCK | LUM_H_DEBLOCK | CHROM_V_DEBLOCK,
-               LUM_V_DEBLOCK | LUM_H_DEBLOCK | CHROM_V_DEBLOCK | CHROM_H_DEBLOCK,
-               LUM_V_DEBLOCK | LUM_H_DEBLOCK | CHROM_V_DEBLOCK | CHROM_H_DEBLOCK | LUM_DERING,
-               LUM_V_DEBLOCK | LUM_H_DEBLOCK | CHROM_V_DEBLOCK | CHROM_H_DEBLOCK | LUM_DERING | CHROM_DERING
-#endif
-       };
-
-#ifdef HAVE_ODIVX_POSTPROCESS
-       int odivx_modes[1+GET_PP_QUALITY_MAX]= {
-               0,
-               PP_DEBLOCK_Y_H,
-               PP_DEBLOCK_Y_H|PP_DEBLOCK_Y_V,
-               PP_DEBLOCK_Y_H|PP_DEBLOCK_Y_V|PP_DEBLOCK_C_H,
-               PP_DEBLOCK_Y_H|PP_DEBLOCK_Y_V|PP_DEBLOCK_C_H|PP_DEBLOCK_C_V,
-               PP_DEBLOCK_Y_H|PP_DEBLOCK_Y_V|PP_DEBLOCK_C_H|PP_DEBLOCK_C_V|PP_DERING_Y,
-               PP_DEBLOCK_Y_H|PP_DEBLOCK_Y_V|PP_DEBLOCK_C_H|PP_DEBLOCK_C_V|PP_DERING_Y|PP_DERING_C
-       };
-       if(use_old_pp) return odivx_modes[quality];
-#endif
-       if(newPPFlag)   return quality;
-       else            return modes[quality];
-}
-
-
index b2ad963ccbf2f96e8460fcf646cf36e5a949b896..66cccd1131f67181a061fc8e4dfd28707b9b5558 100644 (file)
 #ifndef NEWPOSTPROCESS_H
 #define NEWPOSTPROCESS_H
 
-#define BLOCK_SIZE 8
-#define TEMP_STRIDE 8
-//#define NUM_BLOCKS_AT_ONCE 16 //not used yet
-
 #define V_DEBLOCK      0x01
 #define H_DEBLOCK      0x02
 #define DERING         0x04
 #define CHROM_LEVEL_FIX        (LEVEL_FIX<<4)          // 128 (not implemented yet)
 
 // Experimental vertical filters
-#define V_RK1_FILTER   0x0100                  // 256
 #define V_X1_FILTER    0x0200                  // 512
 
 // Experimental horizontal filters
-#define H_RK1_FILTER   0x1000                  // 4096
 #define H_X1_FILTER    0x2000                  // 8192
 
 // select between full y range (255-0) or standart one (234-16)
@@ -55,6 +49,7 @@
 #define        CUBIC_BLEND_DEINT_FILTER        0x8000  // (not implemented yet)
 #define        CUBIC_IPOL_DEINT_FILTER         0x40000 // 262144
 #define        MEDIAN_DEINT_FILTER             0x80000 // 524288
+#define        FFMPEG_DEINT_FILTER             0x400000
 
 #define TEMP_NOISE_FILTER              0x100000
 #define FORCE_QUANT                    0x200000
 
 #define GET_PP_QUALITY_MAX 6
 
-//must be defined if stride%8 != 0
-//#define PP_FUNNY_STRIDE
-
-//#define TIMING
-//#define MORE_TIMING
-
 //use if u want a faster postprocessing code
 //cant differentiate between chroma & luma filters (both on or both off)
 //obviosly the -pp option at the commandline has no effect except turning the here selected
 
 #define QP_STORE_T int
 
-struct PPMode{
+char *postproc_help;
+
+typedef struct PPMode{
        int lumMode; //acivates filters for luminance
        int chromMode; //acivates filters for chrominance
-       int oldMode; // will be passed to odivx
        int error; // non zero on error
 
        int minAllowedY; // for brigtness correction
@@ -88,35 +78,22 @@ struct PPMode{
        int maxTmpNoise[3]; // for Temporal Noise Reducing filter (Maximal sum of abs differences)
        
        int maxDcDiff; // max abs diff between pixels to be considered flat
-       int forcedQuant; // quantizer if FORCE_QUANT is used
-};
-
-struct PPFilter{
-       char *shortName;
-       char *longName;
-       int chromDefault;       // is chrominance filtering on by default if this filter is manually activated
-       int minLumQuality;      // minimum quality to turn luminance filtering on
-       int minChromQuality;    // minimum quality to turn chrominance filtering on
-       int mask;               // Bitmask to turn this filter on
-};
-
-/* Obsolete, dont use it, use postprocess2() instead */
-void postprocess(unsigned char * src[], int src_stride,
-                 unsigned char * dst[], int dst_stride,
-                 int horizontal_size,   int vertical_size,
-                 QP_STORE_T *QP_store,  int QP_stride, int mode);
-
-void postprocess2(unsigned char * src[], int src_stride,
-                 unsigned char * dst[], int dst_stride,
-                 int horizontal_size,   int vertical_size,
-                 QP_STORE_T *QP_store,  int QP_stride, struct PPMode *mode);
+       int flatnessThreshold;
 
+       int forcedQuant; // quantizer if FORCE_QUANT is used
+} PPMode;
 
-/* Obsolete, dont use it, use getPpModeByNameAndQuality() instead */
-int getPpModeForQuality(int quality);
+void  postprocess(uint8_t * src[3], int srcStride[3],
+                 uint8_t * dst[3], int dstStride[3],
+                 int horizontalSize, int verticalSize,
+                 QP_STORE_T *QP_store,  int QP_stride,
+                PPMode *mode, void *ppContext);
 
 // name is the stuff after "-pp" on the command line
-struct PPMode getPPModeByNameAndQuality(char *name, int quality);
+PPMode getPPModeByNameAndQuality(char *name, int quality);
+
+void *getPPContext(int width, int height);
+void freePPContext(void *ppContext);
 
 int readPPOpt(void *conf, char *arg);
 
index dbb6dd5f428aeeac8599f34696cb2b6fef8eb08d..6cd35b0752db03fead9592cf8e53651b96eefe4b 100644 (file)
 
 
 //FIXME? |255-0| = 1 (shouldnt be a problem ...)
+#ifdef HAVE_MMX
 /**
  * Check if the middle 8x8 Block in the given 8x16 block is flat
  */
-static inline int RENAME(isVertDC)(uint8_t src[], int stride){
+static inline int RENAME(isVertDC)(uint8_t src[], int stride, PPContext *c){
        int numEq= 0;
-#ifndef HAVE_MMX
-       int y;
-#endif
        src+= stride*4; // src points to begin of the 8x8 Block
-#ifdef HAVE_MMX
 asm volatile(
                "leal (%1, %2), %%eax                           \n\t"
-               "leal (%%eax, %2, 4), %%ebx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %1      eax     eax+%2  eax+2%2 %1+4%2  ebx     ebx+%2  ebx+2%2 %1+8%2  ebx+4%2
-               "movq "MANGLE(mmxDCOffset)", %%mm7              \n\t" // mm7 = 0x7F
-               "movq "MANGLE(mmxDCThreshold)", %%mm6           \n\t" // mm6 = 0x7D
+//     %1      eax     eax+%2  eax+2%2 %1+4%2  ecx     ecx+%2  ecx+2%2 %1+8%2  ecx+4%2
+               "movq %3, %%mm7                                 \n\t" // mm7 = 0x7F
+               "movq %4, %%mm6                                 \n\t" // mm6 = 0x7D
                "movq (%1), %%mm0                               \n\t"
                "movq (%%eax), %%mm1                            \n\t"
                "psubb %%mm1, %%mm0                             \n\t" // mm0 = differnece
@@ -79,6 +75,8 @@ asm volatile(
                "paddb %%mm7, %%mm2                             \n\t"
                "pcmpgtb %%mm6, %%mm2                           \n\t"
                "paddb %%mm2, %%mm0                             \n\t"
+               
+               "leal (%%eax, %2, 4), %%eax                     \n\t"
 
                "movq (%1, %2, 4), %%mm2                        \n\t"
                "psubb %%mm2, %%mm1                             \n\t"
@@ -86,19 +84,19 @@ asm volatile(
                "pcmpgtb %%mm6, %%mm1                           \n\t"
                "paddb %%mm1, %%mm0                             \n\t"
 
-               "movq (%%ebx), %%mm1                            \n\t"
+               "movq (%%eax), %%mm1                            \n\t"
                "psubb %%mm1, %%mm2                             \n\t"
                "paddb %%mm7, %%mm2                             \n\t"
                "pcmpgtb %%mm6, %%mm2                           \n\t"
                "paddb %%mm2, %%mm0                             \n\t"
 
-               "movq (%%ebx, %2), %%mm2                        \n\t"
+               "movq (%%eax, %2), %%mm2                        \n\t"
                "psubb %%mm2, %%mm1                             \n\t"
                "paddb %%mm7, %%mm1                             \n\t"
                "pcmpgtb %%mm6, %%mm1                           \n\t"
                "paddb %%mm1, %%mm0                             \n\t"
 
-               "movq (%%ebx, %2, 2), %%mm1                     \n\t"
+               "movq (%%eax, %2, 2), %%mm1                     \n\t"
                "psubb %%mm1, %%mm2                             \n\t"
                "paddb %%mm7, %%mm2                             \n\t"
                "pcmpgtb %%mm6, %%mm2                           \n\t"
@@ -121,49 +119,20 @@ asm volatile(
 #endif
                "movd %%mm0, %0                                 \n\t"
                : "=r" (numEq)
-               : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "r" (src), "r" (stride), "m" (c->mmxDcOffset), "m" (c->mmxDcThreshold)
+               : "%eax"
                );
        numEq= (-numEq) &0xFF;
-
-#else
-       for(y=0; y<BLOCK_SIZE-1; y++)
-       {
-               if(((src[0] - src[0+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[1] - src[1+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[2] - src[2+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[3] - src[3+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[4] - src[4+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[5] - src[5+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[6] - src[6+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               if(((src[7] - src[7+stride] + dcOffset)&0xFFFF) < dcThreshold) numEq++;
-               src+= stride;
-       }
-#endif
-/*     if(abs(numEq - asmEq) > 0)
-       {
-               printf("\nasm:%d  c:%d\n", asmEq, numEq);
-               for(int y=0; y<8; y++)
-               {
-                       for(int x=0; x<8; x++)
-                       {
-                               printf("%d ", temp[x + y*stride]);
-                       }
-                       printf("\n");
-               }
-       }
-*/
-//     for(int i=0; i<numEq/8; i++) src[i]=255;
-       return (numEq > vFlatnessThreshold) ? 1 : 0;
+       return numEq > c->ppMode.flatnessThreshold;
 }
+#endif
 
-static inline int RENAME(isVertMinMaxOk)(uint8_t src[], int stride, int QP)
+static inline int RENAME(isVertMinMaxOk)(uint8_t src[], int stride, PPContext *c)
 {
 #ifdef HAVE_MMX
        int isOk;
        src+= stride*3;
        asm volatile(
-//             "int $3 \n\t"
                "movq (%1, %2), %%mm0                           \n\t"
                "movq (%1, %2, 8), %%mm1                        \n\t"
                "movq %%mm0, %%mm2                              \n\t"
@@ -171,55 +140,39 @@ static inline int RENAME(isVertMinMaxOk)(uint8_t src[], int stride, int QP)
                "psubusb %%mm2, %%mm1                           \n\t"
                "por %%mm1, %%mm0                               \n\t" // ABS Diff
 
-               "movq "MANGLE(pQPb)", %%mm7                     \n\t" // QP,..., QP
+               "movq %3, %%mm7                                 \n\t" // QP,..., QP
                "paddusb %%mm7, %%mm7                           \n\t" // 2QP ... 2QP
                "psubusb %%mm7, %%mm0                           \n\t" // Diff <= 2QP -> 0
-               "pcmpeqd "MANGLE(b00)", %%mm0                   \n\t"
-               "psrlq $16, %%mm0                               \n\t"
-               "pcmpeqd "MANGLE(bFF)", %%mm0                   \n\t"
-//             "movd %%mm0, (%1, %2, 4)\n\t"
+               "packssdw %%mm0, %%mm0                          \n\t"
                "movd %%mm0, %0                                 \n\t"
                : "=r" (isOk)
-               : "r" (src), "r" (stride)
+               : "r" (src), "r" (stride), "m" (c->pQPb)
                );
-       return isOk;
+       return isOk==0;
 #else
-
-       int isOk2= 1;
        int x;
+       const int QP= c->QP;
        src+= stride*3;
        for(x=0; x<BLOCK_SIZE; x++)
        {
-               if(abs((int)src[x + stride] - (int)src[x + (stride<<3)]) > 2*QP) isOk2=0;
+               if((unsigned)(src[x + stride] - src[x + (stride<<3)] + 2*QP) > 4*QP) return 0;
        }
-/*     if(isOk && !isOk2 || !isOk && isOk2)
-       {
-               printf("\nasm:%d  c:%d QP:%d\n", isOk, isOk2, QP);
-               for(int y=0; y<9; y++)
-               {
-                       for(int x=0; x<8; x++)
-                       {
-                               printf("%d ", src[x + y*stride]);
-                       }
-                       printf("\n");
-               }
-       } */
 
-       return isOk2;
+       return 1;
 #endif
-
 }
 
 /**
  * Do a vertical low pass filter on the 8x16 block (only write to the 8x8 block in the middle)
  * using the 9-Tap Filter (1,1,2,2,4,2,2,1,1)/16
  */
-static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
+static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, PPContext *c)
 {
 #if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
        src+= stride*3;
        asm volatile(   //"movv %0 %1 %2\n\t"
-               "movq "MANGLE(pQPb)", %%mm0                     \n\t"  // QP,..., QP
+               "movq %2, %%mm0                 \n\t"  // QP,..., QP
+               "pxor %%mm4, %%mm4                              \n\t"
 
                "movq (%0), %%mm6                               \n\t"
                "movq (%0, %1), %%mm5                           \n\t"
@@ -229,7 +182,7 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                "psubusb %%mm1, %%mm2                           \n\t"
                "por %%mm5, %%mm2                               \n\t" // ABS Diff of lines
                "psubusb %%mm0, %%mm2                           \n\t" // diff <= QP -> 0
-               "pcmpeqb "MANGLE(b00)", %%mm2                   \n\t" // diff <= QP -> FF
+               "pcmpeqb %%mm4, %%mm2                   \n\t" // diff <= QP -> FF
 
                "pand %%mm2, %%mm6                              \n\t"
                "pandn %%mm1, %%mm2                             \n\t"
@@ -237,8 +190,8 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
 
                "movq (%0, %1, 8), %%mm5                        \n\t"
                "leal (%0, %1, 4), %%eax                        \n\t"
-               "leal (%0, %1, 8), %%ebx                        \n\t"
-               "subl %1, %%ebx                                 \n\t"
+               "leal (%0, %1, 8), %%ecx                        \n\t"
+               "subl %1, %%ecx                                 \n\t"
                "addl %1, %0                                    \n\t" // %0 points to line 1 not 0
                "movq (%0, %1, 8), %%mm7                        \n\t"
                "movq %%mm5, %%mm1                              \n\t"
@@ -247,7 +200,7 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                "psubusb %%mm1, %%mm2                           \n\t"
                "por %%mm5, %%mm2                               \n\t" // ABS Diff of lines
                "psubusb %%mm0, %%mm2                           \n\t" // diff <= QP -> 0
-               "pcmpeqb "MANGLE(b00)", %%mm2                   \n\t" // diff <= QP -> FF
+               "pcmpeqb %%mm4, %%mm2                   \n\t" // diff <= QP -> FF
 
                "pand %%mm2, %%mm7                              \n\t"
                "pandn %%mm1, %%mm2                             \n\t"
@@ -255,7 +208,7 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
 
 
                //      1       2       3       4       5       6       7       8
-               //      %0      %0+%1   %0+2%1  eax     %0+4%1  eax+2%1 ebx     eax+4%1
+               //      %0      %0+%1   %0+2%1  eax     %0+4%1  eax+2%1 ecx     eax+4%1
                // 6 4 2 2 1 1
                // 6 4 4 2
                // 6 8 2
@@ -286,7 +239,7 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                "movq %%mm3, (%0,%1)                            \n\t" //  X
                // mm1=2 mm2=3(211) mm4=1 mm5=4(211) mm6=0 mm7=9
                PAVGB(%%mm4, %%mm6)                                   //11      /2
-               "movq (%%ebx), %%mm0                            \n\t" //       1
+               "movq (%%ecx), %%mm0                            \n\t" //       1
                PAVGB((%%eax, %1, 2), %%mm0)                          //      11/2
                "movq %%mm0, %%mm3                              \n\t" //      11/2
                PAVGB(%%mm1, %%mm0)                                   //  2   11/4
@@ -296,7 +249,7 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                "movq %%mm0, (%0, %1, 2)                        \n\t" //   X
                // mm1=2 mm2=3 mm3=6(11) mm4=1 mm5=4(211) mm6=0(11) mm7=9
                "movq (%%eax, %1, 4), %%mm0                     \n\t" //        1
-               PAVGB((%%ebx), %%mm0)                                 //       11       /2
+               PAVGB((%%ecx), %%mm0)                                 //       11       /2
                PAVGB(%%mm0, %%mm6)                                   //11     11       /4
                PAVGB(%%mm1, %%mm4)                                   // 11             /2
                PAVGB(%%mm2, %%mm1)                                   //  11            /2
@@ -323,12 +276,12 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                PAVGB(%%mm0, %%mm1)                                   //  11224222      /16
                "movq %%mm1, (%%eax, %1, 2)                     \n\t" //      X
                // mm2=3(112) mm3=6(11) mm4=5 mm5=4(11) mm6=6 mm7=9
-               PAVGB((%%ebx), %%mm2)                                 //   112 4        /8
+               PAVGB((%%ecx), %%mm2)                                 //   112 4        /8
                "movq (%%eax, %1, 4), %%mm0                     \n\t" //        1
                PAVGB(%%mm0, %%mm6)                                   //      1 1       /2
                PAVGB(%%mm7, %%mm6)                                   //      1 12      /4
                PAVGB(%%mm2, %%mm6)                                   //   1122424      /4
-               "movq %%mm6, (%%ebx)                            \n\t" //       X
+               "movq %%mm6, (%%ecx)                            \n\t" //       X
                // mm0=8 mm3=6(11) mm4=5 mm5=4(11) mm7=9
                PAVGB(%%mm7, %%mm5)                                   //    11   2      /4
                PAVGB(%%mm7, %%mm5)                                   //    11   6      /8
@@ -339,8 +292,8 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
                "subl %1, %0                                    \n\t"
 
                :
-               : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "r" (src), "r" (stride), "m" (c->pQPb)
+               : "%eax", "%ecx"
        );
 #else
        const int l1= stride;
@@ -356,8 +309,8 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
        src+= stride*3;
        for(x=0; x<BLOCK_SIZE; x++)
        {
-               const int first= ABS(src[0] - src[l1]) < QP ? src[0] : src[l1];
-               const int last= ABS(src[l8] - src[l9]) < QP ? src[l9] : src[l8];
+               const int first= ABS(src[0] - src[l1]) < c->QP ? src[0] : src[l1];
+               const int last= ABS(src[l8] - src[l9]) < c->QP ? src[l9] : src[l8];
 
                int sums[9];
                sums[0] = first + src[l1];
@@ -381,10 +334,10 @@ static inline void RENAME(doVertLowPass)(uint8_t *src, int stride, int QP)
 
                src++;
        }
-
 #endif
 }
 
+#if 0
 /**
  * Experimental implementation of the filter (Algorithm 1) described in a paper from Ramkishor & Karandikar
  * values are correctly clipped (MMX2)
@@ -405,9 +358,9 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
                "pxor %%mm7, %%mm7                              \n\t" // 0
                "movq "MANGLE(b80)", %%mm6                      \n\t" // MIN_SIGNED_BYTE
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%ecx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  ecx     ecx+%1  ecx+2%1 %0+8%1  ecx+4%1
                "movq "MANGLE(pQPb)", %%mm0                     \n\t" // QP,..., QP
                "movq %%mm0, %%mm1                              \n\t" // QP,..., QP
                "paddusb "MANGLE(b02)", %%mm0                   \n\t"
@@ -415,7 +368,7 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
                "pand "MANGLE(b3F)", %%mm0                      \n\t" // QP/4,..., QP/4
                "paddusb %%mm1, %%mm0                           \n\t" // QP*1.25 ...
                "movq (%0, %1, 4), %%mm2                        \n\t" // line 4
-               "movq (%%ebx), %%mm3                            \n\t" // line 5
+               "movq (%%ecx), %%mm3                            \n\t" // line 5
                "movq %%mm2, %%mm4                              \n\t" // line 4
                "pcmpeqb %%mm5, %%mm5                           \n\t" // -1
                "pxor %%mm2, %%mm5                              \n\t" // -line 4 - 1
@@ -433,11 +386,11 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
 //             "psubb %%mm6, %%mm2                             \n\t"
                "movq %%mm2, (%0,%1, 4)                         \n\t"
 
-               "movq (%%ebx), %%mm2                            \n\t"
+               "movq (%%ecx), %%mm2                            \n\t"
 //             "paddb %%mm6, %%mm2                             \n\t" // line 5 + 0x80
                "psubb %%mm5, %%mm2                             \n\t"
 //             "psubb %%mm6, %%mm2                             \n\t"
-               "movq %%mm2, (%%ebx)                            \n\t"
+               "movq %%mm2, (%%ecx)                            \n\t"
 
                "paddb %%mm6, %%mm5                             \n\t"
                "psrlw $2, %%mm5                                \n\t"
@@ -450,15 +403,15 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
                "psubb %%mm6, %%mm2                             \n\t"
                "movq %%mm2, (%%eax, %1, 2)                     \n\t"
 
-               "movq (%%ebx, %1), %%mm2                        \n\t"
+               "movq (%%ecx, %1), %%mm2                        \n\t"
                "paddb %%mm6, %%mm2                             \n\t" // line 6 + 0x80
                "psubsb %%mm5, %%mm2                            \n\t"
                "psubb %%mm6, %%mm2                             \n\t"
-               "movq %%mm2, (%%ebx, %1)                        \n\t"
+               "movq %%mm2, (%%ecx, %1)                        \n\t"
 
                :
                : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "%eax", "%ecx"
        );
 #else
        const int l1= stride;
@@ -488,6 +441,7 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
 
 #endif
 }
+#endif
 
 /**
  * Experimental Filter 1
@@ -496,7 +450,7 @@ static inline void RENAME(vertRK1Filter)(uint8_t *src, int stride, int QP)
  * can only smooth blocks at the expected locations (it cant smooth them if they did move)
  * MMX2 version does correct clipping C version doesnt
  */
-static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
+static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, PPContext *co)
 {
 #if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
        src+= stride*3;
@@ -504,17 +458,17 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
        asm volatile(
                "pxor %%mm7, %%mm7                              \n\t" // 0
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%ecx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  ecx     ecx+%1  ecx+2%1 %0+8%1  ecx+4%1
                "movq (%%eax, %1, 2), %%mm0                     \n\t" // line 3
                "movq (%0, %1, 4), %%mm1                        \n\t" // line 4
                "movq %%mm1, %%mm2                              \n\t" // line 4
                "psubusb %%mm0, %%mm1                           \n\t"
                "psubusb %%mm2, %%mm0                           \n\t"
                "por %%mm1, %%mm0                               \n\t" // |l2 - l3|
-               "movq (%%ebx), %%mm3                            \n\t" // line 5
-               "movq (%%ebx, %1), %%mm4                        \n\t" // line 6
+               "movq (%%ecx), %%mm3                            \n\t" // line 5
+               "movq (%%ecx, %1), %%mm4                        \n\t" // line 6
                "movq %%mm3, %%mm5                              \n\t" // line 5
                "psubusb %%mm4, %%mm3                           \n\t"
                "psubusb %%mm5, %%mm4                           \n\t"
@@ -528,7 +482,7 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                "por %%mm5, %%mm4                               \n\t" // |l4 - l5|
                "psubusb %%mm0, %%mm4           \n\t" //d = MAX(0, |l4-l5| - (|l2-l3| + |l5-l6|)/2)
                "movq %%mm4, %%mm3                              \n\t" // d
-               "movq "MANGLE(pQPb)", %%mm0                     \n\t"
+               "movq %2, %%mm0                 \n\t"
                 "paddusb %%mm0, %%mm0                          \n\t"
                "psubusb %%mm0, %%mm4                           \n\t"
                "pcmpeqb %%mm7, %%mm4                           \n\t" // d <= QP ? -1 : 0
@@ -546,11 +500,11 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                "pxor %%mm2, %%mm0                              \n\t"
                "movq %%mm0, (%0, %1, 4)                        \n\t" // line 4
 
-               "movq (%%ebx), %%mm0                            \n\t" // line 5
+               "movq (%%ecx), %%mm0                            \n\t" // line 5
                "pxor %%mm2, %%mm0                              \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5
                "paddusb %%mm3, %%mm0                           \n\t"
                "pxor %%mm2, %%mm0                              \n\t"
-               "movq %%mm0, (%%ebx)                            \n\t" // line 5
+               "movq %%mm0, (%%ecx)                            \n\t" // line 5
 
                PAVGB(%%mm7, %%mm1)                                   // d/4
 
@@ -560,11 +514,11 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                "pxor %%mm2, %%mm0                              \n\t"
                "movq %%mm0, (%%eax, %1, 2)                     \n\t" // line 3
 
-               "movq (%%ebx, %1), %%mm0                        \n\t" // line 6
+               "movq (%%ecx, %1), %%mm0                        \n\t" // line 6
                "pxor %%mm2, %%mm0                              \n\t" //(l4 - l5) <= 0 ? -l5-1 : l5
                "paddusb %%mm1, %%mm0                           \n\t"
                "pxor %%mm2, %%mm0                              \n\t"
-               "movq %%mm0, (%%ebx, %1)                        \n\t" // line 6
+               "movq %%mm0, (%%ecx, %1)                        \n\t" // line 6
 
                PAVGB(%%mm7, %%mm1)                                   // d/8
 
@@ -574,15 +528,15 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                "pxor %%mm2, %%mm0                              \n\t"
                "movq %%mm0, (%%eax, %1)                        \n\t" // line 2
 
-               "movq (%%ebx, %1, 2), %%mm0                     \n\t" // line 7
+               "movq (%%ecx, %1, 2), %%mm0                     \n\t" // line 7
                "pxor %%mm2, %%mm0                              \n\t" //(l4 - l5) <= 0 ? -l7-1 : l7
                "paddusb %%mm1, %%mm0                           \n\t"
                "pxor %%mm2, %%mm0                              \n\t"
-               "movq %%mm0, (%%ebx, %1, 2)                     \n\t" // line 7
+               "movq %%mm0, (%%ecx, %1, 2)                     \n\t" // line 7
 
                :
-               : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "r" (src), "r" (stride), "m" (co->pQPb)
+               : "%eax", "%ecx"
        );
 #else
 
@@ -607,7 +561,7 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                int d= ABS(b) - ((ABS(a) + ABS(c))>>1);
                d= MAX(d, 0);
 
-               if(d < QP*2)
+               if(d < co->QP*2)
                {
                        int v = d * SIGN(-b);
 
@@ -621,39 +575,10 @@ static inline void RENAME(vertX1Filter)(uint8_t *src, int stride, int QP)
                }
                src++;
        }
-       /*
-       const int l1= stride;
-       const int l2= stride + l1;
-       const int l3= stride + l2;
-       const int l4= stride + l3;
-       const int l5= stride + l4;
-       const int l6= stride + l5;
-       const int l7= stride + l6;
-       const int l8= stride + l7;
-       const int l9= stride + l8;
-       for(int x=0; x<BLOCK_SIZE; x++)
-       {
-               int v2= src[l2];
-               int v3= src[l3];
-               int v4= src[l4];
-               int v5= src[l5];
-               int v6= src[l6];
-               int v7= src[l7];
-
-               if(ABS(v4-v5)<QP &&  ABS(v4-v5) - (ABS(v3-v4) + ABS(v5-v6))>0 )
-               {
-                       src[l3] = (6*v2 + 4*v3 + 3*v4 + 2*v5 + v6         )/16;
-                       src[l4] = (3*v2 + 3*v3 + 4*v4 + 3*v5 + 2*v6 + v7  )/16;
-                       src[l5] = (1*v2 + 2*v3 + 3*v4 + 4*v5 + 3*v6 + 3*v7)/16;
-                       src[l6] = (       1*v3 + 2*v4 + 3*v5 + 4*v6 + 6*v7)/16;
-               }
-               src++;
-       }
-*/
 #endif
 }
 
-static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
+static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, PPContext *c)
 {
 #if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
 /*
@@ -676,10 +601,10 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
 #if 0 //sligtly more accurate and slightly slower
                "pxor %%mm7, %%mm7                              \n\t" // 0
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%ecx                     \n\t"
 //     0       1       2       3       4       5       6       7
-//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 ebx+%1  ebx+2%1
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1
+//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 ecx+%1  ecx+2%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  ecx     ecx+%1  ecx+2%1
 
 
                "movq (%0, %1, 2), %%mm0                        \n\t" // l2
@@ -708,7 +633,7 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                PAVGB(%%mm2, %%mm0)                                   // ~(l4 + 2l2)/4
                PAVGB(%%mm4, %%mm0)                                   // ~(5l4 + 2l2)/8
 
-               "movq (%%ebx), %%mm2                            \n\t" // l5
+               "movq (%%ecx), %%mm2                            \n\t" // l5
                "movq %%mm3, %%mm5                              \n\t" // l3
                PAVGB(%%mm7, %%mm3)                                   // ~l3/2
                PAVGB(%%mm2, %%mm3)                                   // ~(l3 + 2l5)/4
@@ -721,13 +646,13 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                "pcmpeqb %%mm7, %%mm0                           \n\t" // SIGN(2l2 - 5l3 + 5l4 - 2l5)
 // mm0= SIGN(menergy), mm1= |lenergy|, mm2= l5, mm3= |menergy|, mm4=l4, mm5= l3, mm7=0
 
-               "movq (%%ebx, %1), %%mm6                        \n\t" // l6
+               "movq (%%ecx, %1), %%mm6                        \n\t" // l6
                "movq %%mm6, %%mm5                              \n\t" // l6
                PAVGB(%%mm7, %%mm6)                                   // ~l6/2
                PAVGB(%%mm4, %%mm6)                                   // ~(l6 + 2l4)/4
                PAVGB(%%mm5, %%mm6)                                   // ~(5l6 + 2l4)/8
 
-               "movq (%%ebx, %1, 2), %%mm5                     \n\t" // l7
+               "movq (%%ecx, %1, 2), %%mm5                     \n\t" // l7
                "movq %%mm2, %%mm4                              \n\t" // l5
                PAVGB(%%mm7, %%mm2)                                   // ~l5/2
                PAVGB(%%mm5, %%mm2)                                   // ~(l5 + 2l7)/4
@@ -741,7 +666,7 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
 
 
                PMINUB(%%mm2, %%mm1, %%mm4)                           // MIN(|lenergy|,|renergy|)/8
-               "movq "MANGLE(pQPb)", %%mm4                     \n\t" // QP //FIXME QP+1 ?
+               "movq %2, %%mm4                                 \n\t" // QP //FIXME QP+1 ?
                "paddusb "MANGLE(b01)", %%mm4                   \n\t"
                "pcmpgtb %%mm3, %%mm4                           \n\t" // |menergy|/8 < QP
                "psubusb %%mm1, %%mm3                           \n\t" // d=|menergy|/8-MIN(|lenergy|,|renergy|)/8
@@ -783,8 +708,8 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                "leal (%0, %1), %%eax                           \n\t"
                "pcmpeqb %%mm6, %%mm6                           \n\t" // -1
 //     0       1       2       3       4       5       6       7
-//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 ebx+%1  ebx+2%1
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1
+//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 ecx+%1  ecx+2%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  ecx     ecx+%1  ecx+2%1
 
 
                "movq (%%eax, %1, 2), %%mm1                     \n\t" // l3
@@ -798,7 +723,7 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                "pxor %%mm6, %%mm2                              \n\t" // -l5-1
                "movq %%mm2, %%mm5                              \n\t" // -l5-1
                "movq "MANGLE(b80)", %%mm4                      \n\t" // 128
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%ecx                     \n\t"
                PAVGB(%%mm3, %%mm2)                                   // (l2-l5+256)/2
                PAVGB(%%mm0, %%mm4)                                   // ~(l4-l3)/4 + 128
                PAVGB(%%mm2, %%mm4)                                   // ~(l2-l5)/4 +(l4-l3)/8 + 128
@@ -815,8 +740,8 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                PAVGB(%%mm2, %%mm3)                                   // ~(l0-l3)/8 +5(l2-l1)/16 + 128
 // mm0=128-q, mm3=lenergy/16 + 128, mm4= menergy/16 + 128, mm5= -l5-1
 
-               PAVGB((%%ebx, %1), %%mm5)                             // (l6-l5+256)/2
-               "movq (%%ebx, %1, 2), %%mm1                     \n\t" // l7
+               PAVGB((%%ecx, %1), %%mm5)                             // (l6-l5+256)/2
+               "movq (%%ecx, %1, 2), %%mm1                     \n\t" // l7
                "pxor %%mm6, %%mm1                              \n\t" // -l7-1
                PAVGB((%0, %1, 4), %%mm1)                             // (l4-l7+256)/2
                "movq "MANGLE(b80)", %%mm2                      \n\t" // 128
@@ -836,7 +761,7 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
 // mm0=128-q, mm3=128 + MIN(|lenergy|,|renergy|)/16, mm4= menergy/16 + 128
 
                "movq "MANGLE(b00)", %%mm7                      \n\t" // 0
-               "movq "MANGLE(pQPb)", %%mm2                     \n\t" // QP
+               "movq %2, %%mm2                                 \n\t" // QP
                PAVGB(%%mm6, %%mm2)                                   // 128 + QP/2
                "psubb %%mm6, %%mm2                             \n\t"
 
@@ -877,8 +802,8 @@ static inline void RENAME(doVertDefFilter)(uint8_t src[], int stride, int QP)
                "movq %%mm2, (%0, %1, 4)                        \n\t"
 
                :
-               : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "r" (src), "r" (stride), "m" (c->pQPb)
+               : "%eax", "%ecx"
        );
 
 /*
@@ -951,10 +876,12 @@ src-=8;
        asm volatile(
                "pxor %%mm7, %%mm7                              \n\t"
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
+               "leal -40(%%esp), %%ecx                         \n\t" // make space for 4 8-byte vars
+               "andl $0xFFFFFFF8, %%ecx                        \n\t" // align
 //     0       1       2       3       4       5       6       7
-//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 ebx+%1  ebx+2%1
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1
+//     %0      %0+%1   %0+2%1  eax+2%1 %0+4%1  eax+4%1 edx+%1  edx+2%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1
 
                "movq (%0), %%mm0                               \n\t"
                "movq %%mm0, %%mm1                              \n\t"
@@ -992,8 +919,8 @@ src-=8;
                "psubw %%mm3, %%mm1                             \n\t" // 2H0 - 5H1 + 5H2 - H3
                "psubw %%mm2, %%mm0                             \n\t" // 2L0 - 5L1 + 5L2 - 2L3
                "psubw %%mm3, %%mm1                             \n\t" // 2H0 - 5H1 + 5H2 - 2H3
-               "movq %%mm0, "MANGLE(temp0)"                    \n\t" // 2L0 - 5L1 + 5L2 - 2L3
-               "movq %%mm1, "MANGLE(temp1)"                    \n\t" // 2H0 - 5H1 + 5H2 - 2H3
+               "movq %%mm0, (%%ecx)                            \n\t" // 2L0 - 5L1 + 5L2 - 2L3
+               "movq %%mm1, 8(%%ecx)                           \n\t" // 2H0 - 5H1 + 5H2 - 2H3
 
                "movq (%0, %1, 4), %%mm0                        \n\t"
                "movq %%mm0, %%mm1                              \n\t"
@@ -1002,8 +929,8 @@ src-=8;
 
                "psubw %%mm0, %%mm2                             \n\t" // L3 - L4
                "psubw %%mm1, %%mm3                             \n\t" // H3 - H4
-               "movq %%mm2, "MANGLE(temp2)"                    \n\t" // L3 - L4
-               "movq %%mm3, "MANGLE(temp3)"                    \n\t" // H3 - H4
+               "movq %%mm2, 16(%%ecx)                          \n\t" // L3 - L4
+               "movq %%mm3, 24(%%ecx)                          \n\t" // H3 - H4
                "paddw %%mm4, %%mm4                             \n\t" // 2L2
                "paddw %%mm5, %%mm5                             \n\t" // 2H2
                "psubw %%mm2, %%mm4                             \n\t" // 2L2 - L3 + L4
@@ -1014,7 +941,7 @@ src-=8;
                "psubw %%mm2, %%mm4                             \n\t" // 2L2 - 5L3 + 5L4
                "psubw %%mm3, %%mm5                             \n\t" // 2H2 - 5H3 + 5H4
 //50 opcodes so far
-               "movq (%%ebx), %%mm2                            \n\t"
+               "movq (%%edx), %%mm2                            \n\t"
                "movq %%mm2, %%mm3                              \n\t"
                "punpcklbw %%mm7, %%mm2                         \n\t" // L5
                "punpckhbw %%mm7, %%mm3                         \n\t" // H5
@@ -1023,10 +950,10 @@ src-=8;
                "psubw %%mm2, %%mm4                             \n\t" // 2L2 - 5L3 + 5L4 - 2L5
                "psubw %%mm3, %%mm5                             \n\t" // 2H2 - 5H3 + 5H4 - 2H5
 
-               "movq (%%ebx, %1), %%mm6                        \n\t"
+               "movq (%%edx, %1), %%mm6                        \n\t"
                "punpcklbw %%mm7, %%mm6                         \n\t" // L6
                "psubw %%mm6, %%mm2                             \n\t" // L5 - L6
-               "movq (%%ebx, %1), %%mm6                        \n\t"
+               "movq (%%edx, %1), %%mm6                        \n\t"
                "punpckhbw %%mm7, %%mm6                         \n\t" // H6
                "psubw %%mm6, %%mm3                             \n\t" // H5 - H6
 
@@ -1040,7 +967,7 @@ src-=8;
                "psubw %%mm2, %%mm0                             \n\t" // 2L4 - 5L5 + 5L6
                "psubw %%mm3, %%mm1                             \n\t" // 2H4 - 5H5 + 5H6
 
-               "movq (%%ebx, %1, 2), %%mm2                     \n\t"
+               "movq (%%edx, %1, 2), %%mm2                     \n\t"
                "movq %%mm2, %%mm3                              \n\t"
                "punpcklbw %%mm7, %%mm2                         \n\t" // L7
                "punpckhbw %%mm7, %%mm3                         \n\t" // H7
@@ -1050,8 +977,8 @@ src-=8;
                "psubw %%mm2, %%mm0                             \n\t" // 2L4 - 5L5 + 5L6 - 2L7
                "psubw %%mm3, %%mm1                             \n\t" // 2H4 - 5H5 + 5H6 - 2H7
 
-               "movq "MANGLE(temp0)", %%mm2                    \n\t" // 2L0 - 5L1 + 5L2 - 2L3
-               "movq "MANGLE(temp1)", %%mm3                    \n\t" // 2H0 - 5H1 + 5H2 - 2H3
+               "movq (%%ecx), %%mm2                            \n\t" // 2L0 - 5L1 + 5L2 - 2L3
+               "movq 8(%%ecx), %%mm3                           \n\t" // 2H0 - 5H1 + 5H2 - 2H3
 
 #ifdef HAVE_MMX2
                "movq %%mm7, %%mm6                              \n\t" // 0
@@ -1106,8 +1033,6 @@ src-=8;
                "psubw %%mm7, %%mm5                             \n\t" // |2H2 - 5H3 + 5H4 - 2H5|
 // 100 opcodes
                "movd %2, %%mm2                                 \n\t" // QP
-               "punpcklwd %%mm2, %%mm2                         \n\t"
-               "punpcklwd %%mm2, %%mm2                         \n\t"
                "psllw $3, %%mm2                                \n\t" // 8QP
                "movq %%mm2, %%mm3                              \n\t" // 8QP
                "pcmpgtw %%mm4, %%mm2                           \n\t"
@@ -1129,18 +1054,8 @@ src-=8;
                "psrlw $6, %%mm4                                \n\t"
                "psrlw $6, %%mm5                                \n\t"
 
-/*
-               "movq w06, %%mm2                                \n\t" // 6
-               "paddw %%mm2, %%mm4                             \n\t"
-               "paddw %%mm2, %%mm5                             \n\t"
-               "movq w1400, %%mm2                              \n\t" // 1400h = 5120 = 5/64*2^16
-//FIXME if *5/64 is supposed to be /13 then we should use 5041 instead of 5120
-               "pmulhw %%mm2, %%mm4                            \n\t" // hd/13
-               "pmulhw %%mm2, %%mm5                            \n\t" // ld/13
-*/
-
-               "movq "MANGLE(temp2)", %%mm0                    \n\t" // L3 - L4
-               "movq "MANGLE(temp3)", %%mm1                    \n\t" // H3 - H4
+               "movq 16(%%ecx), %%mm0                          \n\t" // L3 - L4
+               "movq 24(%%ecx), %%mm1                          \n\t" // H3 - H4
 
                "pxor %%mm2, %%mm2                              \n\t"
                "pxor %%mm3, %%mm3                              \n\t"
@@ -1183,8 +1098,8 @@ src-=8;
                "movq %%mm0, (%0, %1, 4)                        \n\t"
 
                :
-               : "r" (src), "r" (stride), "r" (QP)
-               : "%eax", "%ebx"
+               : "r" (src), "r" (stride), "m" (c->pQPb)
+               : "%eax", "%edx", "%ecx"
        );
 #else
        const int l1= stride;
@@ -1201,7 +1116,7 @@ src-=8;
        for(x=0; x<BLOCK_SIZE; x++)
        {
                const int middleEnergy= 5*(src[l5] - src[l4]) + 2*(src[l3] - src[l6]);
-               if(ABS(middleEnergy) < 8*QP)
+               if(ABS(middleEnergy) < 8*c->QP)
                {
                        const int q=(src[l4] - src[l5])/2;
                        const int leftEnergy=  5*(src[l3] - src[l2]) + 2*(src[l1] - src[l4]);
@@ -1232,21 +1147,25 @@ src-=8;
 #endif
 }
 
-static inline void RENAME(dering)(uint8_t src[], int stride, int QP)
+static inline void RENAME(dering)(uint8_t src[], int stride, PPContext *c)
 {
 #if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
        asm volatile(
-               "movq "MANGLE(pQPb)", %%mm0                     \n\t"
-               "paddusb %%mm0, %%mm0                           \n\t"
-               "movq %%mm0, "MANGLE(pQPb2)"                    \n\t"
+               "pxor %%mm6, %%mm6                              \n\t"
+               "pcmpeqb %%mm7, %%mm7                           \n\t"
+               "movq %2, %%mm0                                 \n\t"
+               "punpcklbw %%mm6, %%mm0                         \n\t"
+               "psrlw $1, %%mm0                                \n\t"
+               "psubw %%mm7, %%mm0                             \n\t"
+               "packuswb %%mm0, %%mm0                          \n\t"
+               "movq %%mm0, %3                                 \n\t"
 
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
+               
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
 
-               "pcmpeqb %%mm7, %%mm7                           \n\t"
-               "pxor %%mm6, %%mm6                              \n\t"
 #undef FIND_MIN_MAX
 #ifdef HAVE_MMX2
 #define FIND_MIN_MAX(addr)\
@@ -1267,9 +1186,9 @@ FIND_MIN_MAX((%%eax))
 FIND_MIN_MAX((%%eax, %1))
 FIND_MIN_MAX((%%eax, %1, 2))
 FIND_MIN_MAX((%0, %1, 4))
-FIND_MIN_MAX((%%ebx))
-FIND_MIN_MAX((%%ebx, %1))
-FIND_MIN_MAX((%%ebx, %1, 2))
+FIND_MIN_MAX((%%edx))
+FIND_MIN_MAX((%%edx, %1))
+FIND_MIN_MAX((%%edx, %1, 2))
 FIND_MIN_MAX((%0, %1, 8))
 
                "movq %%mm7, %%mm4                              \n\t"
@@ -1322,11 +1241,13 @@ FIND_MIN_MAX((%0, %1, 8))
                "movd %%mm6, %%ecx                              \n\t"
                "cmpb "MANGLE(deringThreshold)", %%cl           \n\t"
                " jb 1f                                         \n\t"
+               "leal -24(%%esp), %%ecx                         \n\t"
+               "andl $0xFFFFFFF8, %%ecx                        \n\t" 
                PAVGB(%%mm0, %%mm7)                                   // a=(max + min)/2
                "punpcklbw %%mm7, %%mm7                         \n\t"
                "punpcklbw %%mm7, %%mm7                         \n\t"
                "punpcklbw %%mm7, %%mm7                         \n\t"
-               "movq %%mm7, "MANGLE(temp0)"                    \n\t"
+               "movq %%mm7, (%%ecx)                            \n\t"
 
                "movq (%0), %%mm0                               \n\t" // L10
                "movq %%mm0, %%mm1                              \n\t" // L10
@@ -1390,8 +1311,8 @@ FIND_MIN_MAX((%0, %1, 8))
                PAVGB(t0, lx)                                         /* (src[-1] + src[+1])/2 */\
                PAVGB(sx, lx)                                 /* (src[-1] + 2src[0] + src[+1])/4 */\
                PAVGB(lx, pplx)                                      \
-               "movq " #lx ", "MANGLE(temp1)"                  \n\t"\
-               "movq "MANGLE(temp0)", " #lx "                  \n\t"\
+               "movq " #lx ", 8(%%ecx)                         \n\t"\
+               "movq (%%ecx), " #lx "                          \n\t"\
                "psubusb " #lx ", " #t1 "                       \n\t"\
                "psubusb " #lx ", " #t0 "                       \n\t"\
                "psubusb " #lx ", " #sx "                       \n\t"\
@@ -1405,8 +1326,8 @@ FIND_MIN_MAX((%0, %1, 8))
                PAVGB(plx, pplx)                                      /* filtered */\
                "movq " #dst ", " #t0 "                         \n\t" /* dst */\
                "movq " #t0 ", " #t1 "                          \n\t" /* dst */\
-               "psubusb "MANGLE(pQPb2)", " #t0 "               \n\t"\
-               "paddusb "MANGLE(pQPb2)", " #t1 "               \n\t"\
+               "psubusb %3, " #t0 "                            \n\t"\
+               "paddusb %3, " #t1 "                            \n\t"\
                PMAXUB(t0, pplx)\
                PMINUB(t1, pplx, t0)\
                "paddb " #sx ", " #ppsx "                       \n\t"\
@@ -1418,7 +1339,7 @@ FIND_MIN_MAX((%0, %1, 8))
                "pandn " #dst ", " #ppsx "                      \n\t"\
                "por " #pplx ", " #ppsx "                       \n\t"\
                "movq " #ppsx ", " #dst "                       \n\t"\
-               "movq "MANGLE(temp1)", " #lx "                  \n\t"
+               "movq 8(%%ecx), " #lx "                         \n\t"
 
 /*
 0000000
@@ -1439,15 +1360,15 @@ FIND_MIN_MAX((%0, %1, 8))
 DERING_CORE((%%eax),(%%eax, %1)        ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7)
 DERING_CORE((%%eax, %1),(%%eax, %1, 2) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7)
 DERING_CORE((%%eax, %1, 2),(%0, %1, 4) ,%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7)
-DERING_CORE((%0, %1, 4),(%%ebx)        ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7)
-DERING_CORE((%%ebx),(%%ebx, %1)        ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7)
-DERING_CORE((%%ebx, %1), (%%ebx, %1, 2),%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7)
-DERING_CORE((%%ebx, %1, 2),(%0, %1, 8) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7)
-DERING_CORE((%0, %1, 8),(%%ebx, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7)
+DERING_CORE((%0, %1, 4),(%%edx)        ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7)
+DERING_CORE((%%edx),(%%edx, %1)        ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7)
+DERING_CORE((%%edx, %1), (%%edx, %1, 2),%%mm4,%%mm0,%%mm2,%%mm5,%%mm1,%%mm3,%%mm6,%%mm7)
+DERING_CORE((%%edx, %1, 2),(%0, %1, 8) ,%%mm0,%%mm2,%%mm4,%%mm1,%%mm3,%%mm5,%%mm6,%%mm7)
+DERING_CORE((%0, %1, 8),(%%edx, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm6,%%mm7)
 
                "1:                     \n\t"
-               : : "r" (src), "r" (stride), "r" (QP)
-               : "%eax", "%ebx", "%ecx"
+               : : "r" (src), "r" (stride), "m" (c->pQPb), "m"(c->pQPb2)
+               : "%eax", "%edx", "%ecx"
        );
 #else
        int y;
@@ -1456,6 +1377,7 @@ DERING_CORE((%0, %1, 8),(%%ebx, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm
        int avg;
        uint8_t *p;
        int s[10];
+       const int QP2= c->QP/2 + 1;
 
        for(y=1; y<9; y++)
        {
@@ -1468,30 +1390,41 @@ DERING_CORE((%0, %1, 8),(%%ebx, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm
                        if(*p < min) min= *p;
                }
        }
-       avg= (min + max + 1)/2;
+       avg= (min + max + 1)>>1;
 
        if(max - min <deringThreshold) return;
 
        for(y=0; y<10; y++)
        {
-               int x;
                int t = 0;
-               p= src + stride*y;
-               for(x=0; x<10; x++)
-               {
-                       if(*p > avg) t |= (1<<x);
-                       p++;
-               }
+
+               if(src[stride*y + 0] > avg) t+= 1;
+               if(src[stride*y + 1] > avg) t+= 2;
+               if(src[stride*y + 2] > avg) t+= 4;
+               if(src[stride*y + 3] > avg) t+= 8;
+               if(src[stride*y + 4] > avg) t+= 16;
+               if(src[stride*y + 5] > avg) t+= 32;
+               if(src[stride*y + 6] > avg) t+= 64;
+               if(src[stride*y + 7] > avg) t+= 128;
+               if(src[stride*y + 8] > avg) t+= 256;
+               if(src[stride*y + 9] > avg) t+= 512;
+               
                t |= (~t)<<16;
                t &= (t<<1) & (t>>1);
                s[y] = t;
        }
-
+       
        for(y=1; y<9; y++)
        {
-               int x;
                int t = s[y-1] & s[y] & s[y+1];
                t|= t>>16;
+               s[y-1]= t;
+       }
+
+       for(y=1; y<9; y++)
+       {
+               int x;
+               int t = s[y-1];
 
                p= src + stride*y;
                for(x=1; x<9; x++)
@@ -1544,8 +1477,8 @@ DERING_CORE((%0, %1, 8),(%%ebx, %1, 4) ,%%mm2,%%mm4,%%mm0,%%mm3,%%mm5,%%mm1,%%mm
                                }
                                }
 #endif
-                               if     (*p + 2*QP < f) *p= *p + 2*QP;
-                               else if(*p - 2*QP > f) *p= *p - 2*QP;
+                               if     (*p + QP2 < f) *p= *p + QP2;
+                               else if(*p - QP2 > f) *p= *p - QP2;
                                else *p=f;
                        }
                }
@@ -1582,9 +1515,9 @@ static inline void RENAME(deInterlaceInterpolateLinear)(uint8_t src[], int strid
        src+= 4*stride;
        asm volatile(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%ecx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  ecx     ecx+%1  ecx+2%1 %0+8%1  ecx+4%1
 
                "movq (%0), %%mm0                               \n\t"
                "movq (%%eax, %1), %%mm1                        \n\t"
@@ -1593,15 +1526,15 @@ static inline void RENAME(deInterlaceInterpolateLinear)(uint8_t src[], int strid
                "movq (%0, %1, 4), %%mm0                        \n\t"
                PAVGB(%%mm0, %%mm1)
                "movq %%mm1, (%%eax, %1, 2)                     \n\t"
-               "movq (%%ebx, %1), %%mm1                        \n\t"
+               "movq (%%ecx, %1), %%mm1                        \n\t"
                PAVGB(%%mm1, %%mm0)
-               "movq %%mm0, (%%ebx)                            \n\t"
+               "movq %%mm0, (%%ecx)                            \n\t"
                "movq (%0, %1, 8), %%mm0                        \n\t"
                PAVGB(%%mm0, %%mm1)
-               "movq %%mm1, (%%ebx, %1, 2)                     \n\t"
+               "movq %%mm1, (%%ecx, %1, 2)                     \n\t"
 
                : : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "%eax", "%ecx"
        );
 #else
        int x;
@@ -1631,12 +1564,12 @@ static inline void RENAME(deInterlaceInterpolateCubic)(uint8_t src[], int stride
        src+= stride*3;
        asm volatile(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
-               "leal (%%ebx, %1, 4), %%ecx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
+               "leal (%%edx, %1, 4), %%ecx                     \n\t"
                "addl %1, %%ecx                                 \n\t"
                "pxor %%mm7, %%mm7                              \n\t"
 //     0       1       2       3       4       5       6       7       8       9       10
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1 ecx
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1 ecx
 
 #define DEINT_CUBIC(a,b,c,d,e)\
                "movq " #a ", %%mm0                             \n\t"\
@@ -1660,13 +1593,13 @@ static inline void RENAME(deInterlaceInterpolateCubic)(uint8_t src[], int stride
                "packuswb %%mm3, %%mm1                          \n\t"\
                "movq %%mm1, " #c "                             \n\t"
 
-DEINT_CUBIC((%0), (%%eax, %1), (%%eax, %1, 2), (%0, %1, 4), (%%ebx, %1))
-DEINT_CUBIC((%%eax, %1), (%0, %1, 4), (%%ebx), (%%ebx, %1), (%0, %1, 8))
-DEINT_CUBIC((%0, %1, 4), (%%ebx, %1), (%%ebx, %1, 2), (%0, %1, 8), (%%ecx))
-DEINT_CUBIC((%%ebx, %1), (%0, %1, 8), (%%ebx, %1, 4), (%%ecx), (%%ecx, %1, 2))
+DEINT_CUBIC((%0), (%%eax, %1), (%%eax, %1, 2), (%0, %1, 4), (%%edx, %1))
+DEINT_CUBIC((%%eax, %1), (%0, %1, 4), (%%edx), (%%edx, %1), (%0, %1, 8))
+DEINT_CUBIC((%0, %1, 4), (%%edx, %1), (%%edx, %1, 2), (%0, %1, 8), (%%ecx))
+DEINT_CUBIC((%%edx, %1), (%0, %1, 8), (%%edx, %1, 4), (%%ecx), (%%ecx, %1, 2))
 
                : : "r" (src), "r" (stride)
-               : "%eax", "%ebx", "ecx"
+               : "%eax", "%edx", "ecx"
        );
 #else
        int x;
@@ -1682,6 +1615,85 @@ DEINT_CUBIC((%%ebx, %1), (%0, %1, 8), (%%ebx, %1, 4), (%%ecx), (%%ecx, %1, 2))
 #endif
 }
 
+/**
+ * Deinterlaces the given block
+ * will be called for every 8x8 block and can read & write from line 4-15
+ * lines 0-3 have been passed through the deblock / dering filters allready, but can be read too
+ * lines 4-12 will be read into the deblocking filter and should be deinterlaced
+ * this filter will read lines 4-13 and write 5-11
+ * no cliping in C version
+ */
+static inline void RENAME(deInterlaceFF)(uint8_t src[], int stride, uint8_t *tmp)
+{
+#if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
+       src+= stride*4;
+       asm volatile(
+               "leal (%0, %1), %%eax                           \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
+               "pxor %%mm7, %%mm7                              \n\t"
+               "movq (%2), %%mm0                               \n\t"
+//     0       1       2       3       4       5       6       7       8       9       10
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1 ecx
+
+#define DEINT_FF(a,b,c,d)\
+               "movq " #a ", %%mm1                             \n\t"\
+               "movq " #b ", %%mm2                             \n\t"\
+               "movq " #c ", %%mm3                             \n\t"\
+               "movq " #d ", %%mm4                             \n\t"\
+               PAVGB(%%mm3, %%mm1)                                     \
+               PAVGB(%%mm4, %%mm0)                                     \
+               "movq %%mm0, %%mm3                              \n\t"\
+               "punpcklbw %%mm7, %%mm0                         \n\t"\
+               "punpckhbw %%mm7, %%mm3                         \n\t"\
+               "movq %%mm1, %%mm4                              \n\t"\
+               "punpcklbw %%mm7, %%mm1                         \n\t"\
+               "punpckhbw %%mm7, %%mm4                         \n\t"\
+               "psllw $2, %%mm1                                \n\t"\
+               "psllw $2, %%mm4                                \n\t"\
+               "psubw %%mm0, %%mm1                             \n\t"\
+               "psubw %%mm3, %%mm4                             \n\t"\
+               "movq %%mm2, %%mm5                              \n\t"\
+               "movq %%mm2, %%mm0                              \n\t"\
+               "punpcklbw %%mm7, %%mm2                         \n\t"\
+               "punpckhbw %%mm7, %%mm5                         \n\t"\
+               "paddw %%mm2, %%mm1                             \n\t"\
+               "paddw %%mm5, %%mm4                             \n\t"\
+               "psraw $2, %%mm1                                \n\t"\
+               "psraw $2, %%mm4                                \n\t"\
+               "packuswb %%mm4, %%mm1                          \n\t"\
+               "movq %%mm1, " #b "                             \n\t"\
+
+DEINT_FF((%0)       , (%%eax)       , (%%eax, %1), (%%eax, %1, 2))
+DEINT_FF((%%eax, %1), (%%eax, %1, 2), (%0, %1, 4), (%%edx)       )
+DEINT_FF((%0, %1, 4), (%%edx)       , (%%edx, %1), (%%edx, %1, 2))
+DEINT_FF((%%edx, %1), (%%edx, %1, 2), (%0, %1, 8), (%%edx, %1, 4))
+
+               "movq %%mm0, (%2)                               \n\t"
+               : : "r" (src), "r" (stride), "r"(tmp)
+               : "%eax", "%edx"
+       );
+#else
+       int x;
+       src+= stride*4;
+       for(x=0; x<8; x++)
+       {
+               int t1= tmp[x];
+               int t2= src[stride*1];
+
+               src[stride*1]= (-t1 + 4*src[stride*0] + 2*t2 + 4*src[stride*2] - src[stride*3] + 4)>>3;
+               t1= src[stride*4];
+               src[stride*3]= (-t2 + 4*src[stride*2] + 2*t1 + 4*src[stride*4] - src[stride*5] + 4)>>3;
+               t2= src[stride*6];
+               src[stride*5]= (-t1 + 4*src[stride*4] + 2*t2 + 4*src[stride*6] - src[stride*7] + 4)>>3;
+               t1= src[stride*8];
+               src[stride*7]= (-t2 + 4*src[stride*6] + 2*t1 + 4*src[stride*8] - src[stride*9] + 4)>>3;
+               tmp[x]= t1;
+
+               src++;
+       }
+#endif
+}
+
 /**
  * Deinterlaces the given block
  * will be called for every 8x8 block and can read & write from line 4-15
@@ -1696,9 +1708,9 @@ static inline void RENAME(deInterlaceBlendLinear)(uint8_t src[], int stride)
        src+= 4*stride;
        asm volatile(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
 
                "movq (%0), %%mm0                               \n\t" // L0
                "movq (%%eax, %1), %%mm1                        \n\t" // L2
@@ -1714,30 +1726,30 @@ static inline void RENAME(deInterlaceBlendLinear)(uint8_t src[], int stride)
                PAVGB(%%mm2, %%mm1)                                   // L2+L4
                PAVGB(%%mm0, %%mm1)                                   // 2L3 + L2 + L4
                "movq %%mm1, (%%eax, %1)                        \n\t"
-               "movq (%%ebx), %%mm1                            \n\t" // L5
+               "movq (%%edx), %%mm1                            \n\t" // L5
                PAVGB(%%mm1, %%mm0)                                   // L3+L5
                PAVGB(%%mm2, %%mm0)                                   // 2L4 + L3 + L5
                "movq %%mm0, (%%eax, %1, 2)                     \n\t"
-               "movq (%%ebx, %1), %%mm0                        \n\t" // L6
+               "movq (%%edx, %1), %%mm0                        \n\t" // L6
                PAVGB(%%mm0, %%mm2)                                   // L4+L6
                PAVGB(%%mm1, %%mm2)                                   // 2L5 + L4 + L6
                "movq %%mm2, (%0, %1, 4)                        \n\t"
-               "movq (%%ebx, %1, 2), %%mm2                     \n\t" // L7
+               "movq (%%edx, %1, 2), %%mm2                     \n\t" // L7
                PAVGB(%%mm2, %%mm1)                                   // L5+L7
                PAVGB(%%mm0, %%mm1)                                   // 2L6 + L5 + L7
-               "movq %%mm1, (%%ebx)                            \n\t"
+               "movq %%mm1, (%%edx)                            \n\t"
                "movq (%0, %1, 8), %%mm1                        \n\t" // L8
                PAVGB(%%mm1, %%mm0)                                   // L6+L8
                PAVGB(%%mm2, %%mm0)                                   // 2L7 + L6 + L8
-               "movq %%mm0, (%%ebx, %1)                        \n\t"
-               "movq (%%ebx, %1, 4), %%mm0                     \n\t" // L9
+               "movq %%mm0, (%%edx, %1)                        \n\t"
+               "movq (%%edx, %1, 4), %%mm0                     \n\t" // L9
                PAVGB(%%mm0, %%mm2)                                   // L7+L9
                PAVGB(%%mm1, %%mm2)                                   // 2L8 + L7 + L9
-               "movq %%mm2, (%%ebx, %1, 2)                     \n\t"
+               "movq %%mm2, (%%edx, %1, 2)                     \n\t"
 
 
                : : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "%eax", "%edx"
        );
 #else
        int x;
@@ -1770,9 +1782,9 @@ static inline void RENAME(deInterlaceMedian)(uint8_t src[], int stride)
 #ifdef HAVE_MMX2
        asm volatile(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
 
                "movq (%0), %%mm0                               \n\t" //
                "movq (%%eax, %1), %%mm2                        \n\t" //
@@ -1793,35 +1805,35 @@ static inline void RENAME(deInterlaceMedian)(uint8_t src[], int stride)
                "pminub %%mm1, %%mm2                            \n\t"
                "movq %%mm2, (%%eax, %1, 2)                     \n\t"
 
-               "movq (%%ebx), %%mm2                            \n\t" //
-               "movq (%%ebx, %1), %%mm1                        \n\t" //
+               "movq (%%edx), %%mm2                            \n\t" //
+               "movq (%%edx, %1), %%mm1                        \n\t" //
                "movq %%mm2, %%mm3                              \n\t"
                "pmaxub %%mm0, %%mm2                            \n\t" //
                "pminub %%mm3, %%mm0                            \n\t" //
                "pmaxub %%mm1, %%mm0                            \n\t" //
                "pminub %%mm0, %%mm2                            \n\t"
-               "movq %%mm2, (%%ebx)                            \n\t"
+               "movq %%mm2, (%%edx)                            \n\t"
 
-               "movq (%%ebx, %1, 2), %%mm2                     \n\t" //
+               "movq (%%edx, %1, 2), %%mm2                     \n\t" //
                "movq (%0, %1, 8), %%mm0                        \n\t" //
                "movq %%mm2, %%mm3                              \n\t"
                "pmaxub %%mm0, %%mm2                            \n\t" //
                "pminub %%mm3, %%mm0                            \n\t" //
                "pmaxub %%mm1, %%mm0                            \n\t" //
                "pminub %%mm0, %%mm2                            \n\t"
-               "movq %%mm2, (%%ebx, %1, 2)                     \n\t"
+               "movq %%mm2, (%%edx, %1, 2)                     \n\t"
 
 
                : : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "%eax", "%edx"
        );
 
 #else // MMX without MMX2
        asm volatile(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
                "pxor %%mm7, %%mm7                              \n\t"
 
 #define MEDIAN(a,b,c)\
@@ -1850,11 +1862,11 @@ static inline void RENAME(deInterlaceMedian)(uint8_t src[], int stride)
 
 MEDIAN((%0), (%%eax), (%%eax, %1))
 MEDIAN((%%eax, %1), (%%eax, %1, 2), (%0, %1, 4))
-MEDIAN((%0, %1, 4), (%%ebx), (%%ebx, %1))
-MEDIAN((%%ebx, %1), (%%ebx, %1, 2), (%0, %1, 8))
+MEDIAN((%0, %1, 4), (%%edx), (%%edx, %1))
+MEDIAN((%%edx, %1), (%%edx, %1, 2), (%0, %1, 8))
 
                : : "r" (src), "r" (stride)
-               : "%eax", "%ebx"
+               : "%eax", "%edx"
        );
 #endif // MMX
 #else
@@ -1884,9 +1896,9 @@ static inline void RENAME(transpose1)(uint8_t *dst1, uint8_t *dst2, uint8_t *src
 {
        asm(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
                "movq (%0), %%mm0               \n\t" // 12345678
                "movq (%%eax), %%mm1            \n\t" // abcdefgh
                "movq %%mm0, %%mm2              \n\t" // 12345678
@@ -1922,13 +1934,13 @@ static inline void RENAME(transpose1)(uint8_t *dst1, uint8_t *dst2, uint8_t *src
                "movd %%mm1, 112(%3)            \n\t"
 
                "movq (%0, %1, 4), %%mm0        \n\t" // 12345678
-               "movq (%%ebx), %%mm1            \n\t" // abcdefgh
+               "movq (%%edx), %%mm1            \n\t" // abcdefgh
                "movq %%mm0, %%mm2              \n\t" // 12345678
                "punpcklbw %%mm1, %%mm0         \n\t" // 1a2b3c4d
                "punpckhbw %%mm1, %%mm2         \n\t" // 5e6f7g8h
 
-               "movq (%%ebx, %1), %%mm1        \n\t"
-               "movq (%%ebx, %1, 2), %%mm3     \n\t"
+               "movq (%%edx, %1), %%mm1        \n\t"
+               "movq (%%edx, %1, 2), %%mm3     \n\t"
                "movq %%mm1, %%mm4              \n\t"
                "punpcklbw %%mm3, %%mm1         \n\t"
                "punpckhbw %%mm3, %%mm4         \n\t"
@@ -1957,7 +1969,7 @@ static inline void RENAME(transpose1)(uint8_t *dst1, uint8_t *dst2, uint8_t *src
 
 
        :: "r" (src), "r" (srcStride), "r" (dst1), "r" (dst2)
-       : "%eax", "%ebx"
+       : "%eax", "%edx"
        );
 }
 
@@ -1968,9 +1980,9 @@ static inline void RENAME(transpose2)(uint8_t *dst, int dstStride, uint8_t *src)
 {
        asm(
                "leal (%0, %1), %%eax                           \n\t"
-               "leal (%%eax, %1, 4), %%ebx                     \n\t"
+               "leal (%%eax, %1, 4), %%edx                     \n\t"
 //     0       1       2       3       4       5       6       7       8       9
-//     %0      eax     eax+%1  eax+2%1 %0+4%1  ebx     ebx+%1  ebx+2%1 %0+8%1  ebx+4%1
+//     %0      eax     eax+%1  eax+2%1 %0+4%1  edx     edx+%1  edx+2%1 %0+8%1  edx+4%1
                "movq (%2), %%mm0               \n\t" // 12345678
                "movq 16(%2), %%mm1             \n\t" // abcdefgh
                "movq %%mm0, %%mm2              \n\t" // 12345678
@@ -1998,10 +2010,10 @@ static inline void RENAME(transpose2)(uint8_t *dst, int dstStride, uint8_t *src)
                "movd %%mm3, (%%eax, %1, 2)     \n\t"
                "movd %%mm2, (%0, %1, 4)        \n\t"
                "psrlq $32, %%mm2               \n\t"
-               "movd %%mm2, (%%ebx)            \n\t"
-               "movd %%mm1, (%%ebx, %1)        \n\t"
+               "movd %%mm2, (%%edx)            \n\t"
+               "movd %%mm1, (%%edx, %1)        \n\t"
                "psrlq $32, %%mm1               \n\t"
-               "movd %%mm1, (%%ebx, %1, 2)     \n\t"
+               "movd %%mm1, (%%edx, %1, 2)     \n\t"
 
 
                "movq 64(%2), %%mm0             \n\t" // 12345678
@@ -2031,13 +2043,13 @@ static inline void RENAME(transpose2)(uint8_t *dst, int dstStride, uint8_t *src)
                "movd %%mm3, 4(%%eax, %1, 2)    \n\t"
                "movd %%mm2, 4(%0, %1, 4)       \n\t"
                "psrlq $32, %%mm2               \n\t"
-               "movd %%mm2, 4(%%ebx)           \n\t"
-               "movd %%mm1, 4(%%ebx, %1)       \n\t"
+               "movd %%mm2, 4(%%edx)           \n\t"
+               "movd %%mm1, 4(%%edx, %1)       \n\t"
                "psrlq $32, %%mm1               \n\t"
-               "movd %%mm1, 4(%%ebx, %1, 2)    \n\t"
+               "movd %%mm1, 4(%%edx, %1, 2)    \n\t"
 
        :: "r" (dst), "r" (dstStride), "r" (src)
-       : "%eax", "%ebx"
+       : "%eax", "%edx"
        );
 }
 #endif
@@ -2046,15 +2058,20 @@ static inline void RENAME(transpose2)(uint8_t *dst, int dstStride, uint8_t *src)
 static void inline RENAME(tempNoiseReducer)(uint8_t *src, int stride,
                                    uint8_t *tempBlured, uint32_t *tempBluredPast, int *maxNoise)
 {
+       // to save a register (FIXME do this outside of the loops)
+       tempBluredPast[127]= maxNoise[0];
+       tempBluredPast[128]= maxNoise[1];
+       tempBluredPast[129]= maxNoise[2];
+        
 #define FAST_L2_DIFF
 //#define L1_DIFF //u should change the thresholds too if u try that one
 #if defined (HAVE_MMX2) || defined (HAVE_3DNOW)
        asm volatile(
                "leal (%2, %2, 2), %%eax                        \n\t" // 3*stride
-               "leal (%2, %2, 4), %%ebx                        \n\t" // 5*stride
-               "leal (%%ebx, %2, 2), %%ecx                     \n\t" // 7*stride
+               "leal (%2, %2, 4), %%edx                        \n\t" // 5*stride
+               "leal (%%edx, %2, 2), %%ecx                     \n\t" // 7*stride
 //     0       1       2       3       4       5       6       7       8       9
-//     %x      %x+%2   %x+2%2  %x+eax  %x+4%2  %x+ebx  %x+2eax %x+ecx  %x+8%2
+//     %x      %x+%2   %x+2%2  %x+eax  %x+4%2  %x+edx  %x+2eax %x+ecx  %x+8%2
 //FIXME reorder?
 #ifdef L1_DIFF //needs mmx2
                "movq (%0), %%mm0                               \n\t" // L0
@@ -2069,9 +2086,9 @@ static void inline RENAME(tempNoiseReducer)(uint8_t *src, int stride,
                "movq (%0, %2, 4), %%mm4                        \n\t" // L4
                "paddw %%mm1, %%mm0                             \n\t"
                "psadbw (%1, %2, 4), %%mm4                      \n\t" // |L4-R4|
-               "movq (%0, %%ebx), %%mm5                        \n\t" // L5
+               "movq (%0, %%edx), %%mm5                        \n\t" // L5
                "paddw %%mm2, %%mm0                             \n\t"
-               "psadbw (%1, %%ebx), %%mm5                      \n\t" // |L5-R5|
+               "psadbw (%1, %%edx), %%mm5                      \n\t" // |L5-R5|
                "movq (%0, %%eax, 2), %%mm6                     \n\t" // L6
                "paddw %%mm3, %%mm0                             \n\t"
                "psadbw (%1, %%eax, 2), %%mm6                   \n\t" // |L6-R6|
@@ -2104,7 +2121,7 @@ L2_DIFF_CORE((%0, %2), (%1, %2))
 L2_DIFF_CORE((%0, %2, 2), (%1, %2, 2))
 L2_DIFF_CORE((%0, %%eax), (%1, %%eax))
 L2_DIFF_CORE((%0, %2, 4), (%1, %2, 4))
-L2_DIFF_CORE((%0, %%ebx), (%1, %%ebx))
+L2_DIFF_CORE((%0, %%edx), (%1, %%edx))
 L2_DIFF_CORE((%0, %%eax,2), (%1, %%eax,2))
 L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
 
@@ -2132,7 +2149,7 @@ L2_DIFF_CORE((%0, %2), (%1, %2))
 L2_DIFF_CORE((%0, %2, 2), (%1, %2, 2))
 L2_DIFF_CORE((%0, %%eax), (%1, %%eax))
 L2_DIFF_CORE((%0, %2, 4), (%1, %2, 4))
-L2_DIFF_CORE((%0, %%ebx), (%1, %%ebx))
+L2_DIFF_CORE((%0, %%edx), (%1, %%edx))
 L2_DIFF_CORE((%0, %%eax,2), (%1, %%eax,2))
 L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
 
@@ -2143,31 +2160,31 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "paddd %%mm0, %%mm4                             \n\t"
                "movd %%mm4, %%ecx                              \n\t"
                "shll $2, %%ecx                                 \n\t"
-               "movl %3, %%ebx                                 \n\t"
-               "addl -4(%%ebx), %%ecx                          \n\t"
-               "addl 4(%%ebx), %%ecx                           \n\t"
-               "addl -1024(%%ebx), %%ecx                       \n\t"
+               "movl %3, %%edx                                 \n\t"
+               "addl -4(%%edx), %%ecx                          \n\t"
+               "addl 4(%%edx), %%ecx                           \n\t"
+               "addl -1024(%%edx), %%ecx                       \n\t"
                "addl $4, %%ecx                                 \n\t"
-               "addl 1024(%%ebx), %%ecx                        \n\t"
+               "addl 1024(%%edx), %%ecx                        \n\t"
                "shrl $3, %%ecx                                 \n\t"
-               "movl %%ecx, (%%ebx)                            \n\t"
-               "leal (%%eax, %2, 2), %%ebx                     \n\t" // 5*stride
+               "movl %%ecx, (%%edx)                            \n\t"
 
 //             "movl %3, %%ecx                                 \n\t"
 //             "movl %%ecx, test                               \n\t"
 //             "jmp 4f \n\t"
-               "cmpl 4+"MANGLE(maxTmpNoise)", %%ecx            \n\t"
+               "cmpl 512(%%edx), %%ecx                         \n\t"
                " jb 2f                                         \n\t"
-               "cmpl 8+"MANGLE(maxTmpNoise)", %%ecx            \n\t"
+               "cmpl 516(%%edx), %%ecx                         \n\t"
                " jb 1f                                         \n\t"
 
-               "leal (%%ebx, %2, 2), %%ecx                     \n\t" // 7*stride
+               "leal (%%eax, %2, 2), %%edx                     \n\t" // 5*stride
+               "leal (%%edx, %2, 2), %%ecx                     \n\t" // 7*stride
                "movq (%0), %%mm0                               \n\t" // L0
                "movq (%0, %2), %%mm1                           \n\t" // L1
                "movq (%0, %2, 2), %%mm2                        \n\t" // L2
                "movq (%0, %%eax), %%mm3                        \n\t" // L3
                "movq (%0, %2, 4), %%mm4                        \n\t" // L4
-               "movq (%0, %%ebx), %%mm5                        \n\t" // L5
+               "movq (%0, %%edx), %%mm5                        \n\t" // L5
                "movq (%0, %%eax, 2), %%mm6                     \n\t" // L6
                "movq (%0, %%ecx), %%mm7                        \n\t" // L7
                "movq %%mm0, (%1)                               \n\t" // L0
@@ -2175,13 +2192,14 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "movq %%mm2, (%1, %2, 2)                        \n\t" // L2
                "movq %%mm3, (%1, %%eax)                        \n\t" // L3
                "movq %%mm4, (%1, %2, 4)                        \n\t" // L4
-               "movq %%mm5, (%1, %%ebx)                        \n\t" // L5
+               "movq %%mm5, (%1, %%edx)                        \n\t" // L5
                "movq %%mm6, (%1, %%eax, 2)                     \n\t" // L6
                "movq %%mm7, (%1, %%ecx)                        \n\t" // L7
                "jmp 4f                                         \n\t"
 
                "1:                                             \n\t"
-               "leal (%%ebx, %2, 2), %%ecx                     \n\t" // 7*stride
+               "leal (%%eax, %2, 2), %%edx                     \n\t" // 5*stride
+               "leal (%%edx, %2, 2), %%ecx                     \n\t" // 7*stride
                "movq (%0), %%mm0                               \n\t" // L0
                PAVGB((%1), %%mm0)                                    // L0
                "movq (%0, %2), %%mm1                           \n\t" // L1
@@ -2192,8 +2210,8 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                PAVGB((%1, %%eax), %%mm3)                             // L3
                "movq (%0, %2, 4), %%mm4                        \n\t" // L4
                PAVGB((%1, %2, 4), %%mm4)                             // L4
-               "movq (%0, %%ebx), %%mm5                        \n\t" // L5
-               PAVGB((%1, %%ebx), %%mm5)                             // L5
+               "movq (%0, %%edx), %%mm5                        \n\t" // L5
+               PAVGB((%1, %%edx), %%mm5)                             // L5
                "movq (%0, %%eax, 2), %%mm6                     \n\t" // L6
                PAVGB((%1, %%eax, 2), %%mm6)                          // L6
                "movq (%0, %%ecx), %%mm7                        \n\t" // L7
@@ -2203,7 +2221,7 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "movq %%mm2, (%1, %2, 2)                        \n\t" // R2
                "movq %%mm3, (%1, %%eax)                        \n\t" // R3
                "movq %%mm4, (%1, %2, 4)                        \n\t" // R4
-               "movq %%mm5, (%1, %%ebx)                        \n\t" // R5
+               "movq %%mm5, (%1, %%edx)                        \n\t" // R5
                "movq %%mm6, (%1, %%eax, 2)                     \n\t" // R6
                "movq %%mm7, (%1, %%ecx)                        \n\t" // R7
                "movq %%mm0, (%0)                               \n\t" // L0
@@ -2211,16 +2229,17 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "movq %%mm2, (%0, %2, 2)                        \n\t" // L2
                "movq %%mm3, (%0, %%eax)                        \n\t" // L3
                "movq %%mm4, (%0, %2, 4)                        \n\t" // L4
-               "movq %%mm5, (%0, %%ebx)                        \n\t" // L5
+               "movq %%mm5, (%0, %%edx)                        \n\t" // L5
                "movq %%mm6, (%0, %%eax, 2)                     \n\t" // L6
                "movq %%mm7, (%0, %%ecx)                        \n\t" // L7
                "jmp 4f                                         \n\t"
 
                "2:                                             \n\t"
-               "cmpl "MANGLE(maxTmpNoise)", %%ecx              \n\t"
+               "cmpl 508(%%edx), %%ecx                         \n\t"
                " jb 3f                                         \n\t"
 
-               "leal (%%ebx, %2, 2), %%ecx                     \n\t" // 7*stride
+               "leal (%%eax, %2, 2), %%edx                     \n\t" // 5*stride
+               "leal (%%edx, %2, 2), %%ecx                     \n\t" // 7*stride
                "movq (%0), %%mm0                               \n\t" // L0
                "movq (%0, %2), %%mm1                           \n\t" // L1
                "movq (%0, %2, 2), %%mm2                        \n\t" // L2
@@ -2247,11 +2266,11 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "movq %%mm3, (%0, %%eax)                        \n\t" // L3
 
                "movq (%0, %2, 4), %%mm0                        \n\t" // L4
-               "movq (%0, %%ebx), %%mm1                        \n\t" // L5
+               "movq (%0, %%edx), %%mm1                        \n\t" // L5
                "movq (%0, %%eax, 2), %%mm2                     \n\t" // L6
                "movq (%0, %%ecx), %%mm3                        \n\t" // L7
                "movq (%1, %2, 4), %%mm4                        \n\t" // R4
-               "movq (%1, %%ebx), %%mm5                        \n\t" // R5
+               "movq (%1, %%edx), %%mm5                        \n\t" // R5
                "movq (%1, %%eax, 2), %%mm6                     \n\t" // R6
                "movq (%1, %%ecx), %%mm7                        \n\t" // R7
                PAVGB(%%mm4, %%mm0)
@@ -2263,17 +2282,18 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                PAVGB(%%mm6, %%mm2)
                PAVGB(%%mm7, %%mm3)
                "movq %%mm0, (%1, %2, 4)                        \n\t" // R4
-               "movq %%mm1, (%1, %%ebx)                        \n\t" // R5
+               "movq %%mm1, (%1, %%edx)                        \n\t" // R5
                "movq %%mm2, (%1, %%eax, 2)                     \n\t" // R6
                "movq %%mm3, (%1, %%ecx)                        \n\t" // R7
                "movq %%mm0, (%0, %2, 4)                        \n\t" // L4
-               "movq %%mm1, (%0, %%ebx)                        \n\t" // L5
+               "movq %%mm1, (%0, %%edx)                        \n\t" // L5
                "movq %%mm2, (%0, %%eax, 2)                     \n\t" // L6
                "movq %%mm3, (%0, %%ecx)                        \n\t" // L7
                "jmp 4f                                         \n\t"
 
                "3:                                             \n\t"
-               "leal (%%ebx, %2, 2), %%ecx                     \n\t" // 7*stride
+               "leal (%%eax, %2, 2), %%edx                     \n\t" // 5*stride
+               "leal (%%edx, %2, 2), %%ecx                     \n\t" // 7*stride
                "movq (%0), %%mm0                               \n\t" // L0
                "movq (%0, %2), %%mm1                           \n\t" // L1
                "movq (%0, %2, 2), %%mm2                        \n\t" // L2
@@ -2304,11 +2324,11 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                "movq %%mm3, (%0, %%eax)                        \n\t" // L3
 
                "movq (%0, %2, 4), %%mm0                        \n\t" // L4
-               "movq (%0, %%ebx), %%mm1                        \n\t" // L5
+               "movq (%0, %%edx), %%mm1                        \n\t" // L5
                "movq (%0, %%eax, 2), %%mm2                     \n\t" // L6
                "movq (%0, %%ecx), %%mm3                        \n\t" // L7
                "movq (%1, %2, 4), %%mm4                        \n\t" // R4
-               "movq (%1, %%ebx), %%mm5                        \n\t" // R5
+               "movq (%1, %%edx), %%mm5                        \n\t" // R5
                "movq (%1, %%eax, 2), %%mm6                     \n\t" // R6
                "movq (%1, %%ecx), %%mm7                        \n\t" // R7
                PAVGB(%%mm4, %%mm0)
@@ -2324,18 +2344,18 @@ L2_DIFF_CORE((%0, %%ecx), (%1, %%ecx))
                PAVGB(%%mm6, %%mm2)
                PAVGB(%%mm7, %%mm3)
                "movq %%mm0, (%1, %2, 4)                        \n\t" // R4
-               "movq %%mm1, (%1, %%ebx)                        \n\t" // R5
+               "movq %%mm1, (%1, %%edx)                        \n\t" // R5
                "movq %%mm2, (%1, %%eax, 2)                     \n\t" // R6
                "movq %%mm3, (%1, %%ecx)                        \n\t" // R7
                "movq %%mm0, (%0, %2, 4)                        \n\t" // L4
-               "movq %%mm1, (%0, %%ebx)                        \n\t" // L5
+               "movq %%mm1, (%0, %%edx)                        \n\t" // L5
                "movq %%mm2, (%0, %%eax, 2)                     \n\t" // L6
                "movq %%mm3, (%0, %%ecx)                        \n\t" // L7
 
                "4:                                             \n\t"
 
                :: "r" (src), "r" (tempBlured), "r"(stride), "m" (tempBluredPast)
-               : "%eax", "%ebx", "%ecx", "memory"
+               : "%eax", "%edx", "%ecx", "memory"
                );
 //printf("%d\n", test);
 #else
@@ -2443,7 +2463,7 @@ Switch between
 }
 
 static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height,
-       QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode);
+       QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c);
 
 /**
  * Copies a block from src to dst and fixes the blacklevel
@@ -2452,7 +2472,7 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
 #undef SCALED_CPY
 
 static inline void RENAME(blockCopy)(uint8_t dst[], int dstStride, uint8_t src[], int srcStride,
-       int levelFix)
+       int levelFix, int64_t *packedOffsetAndScale)
 {
 #ifndef HAVE_MMX
        int i;
@@ -2461,10 +2481,10 @@ static inline void RENAME(blockCopy)(uint8_t dst[], int dstStride, uint8_t src[]
        {
 #ifdef HAVE_MMX
                                        asm volatile(
-                                               "leal (%0,%2), %%eax    \n\t"
-                                               "leal (%1,%3), %%ebx    \n\t"
-                                               "movq "MANGLE(packedYOffset)", %%mm2\n\t"
-                                               "movq "MANGLE(packedYScale)", %%mm3\n\t"
+                                               "movq (%%eax), %%mm2    \n\t" // packedYOffset
+                                               "movq 8(%%eax), %%mm3   \n\t" // packedYScale
+                                               "leal (%2,%4), %%eax    \n\t"
+                                               "leal (%3,%5), %%edx    \n\t"
                                                "pxor %%mm4, %%mm4      \n\t"
 #ifdef HAVE_MMX2
 #define SCALED_CPY(src1, src2, dst1, dst2)                                     \
@@ -2518,19 +2538,21 @@ static inline void RENAME(blockCopy)(uint8_t dst[], int dstStride, uint8_t src[]
 
 #endif //!HAVE_MMX2
 
-SCALED_CPY((%0)       , (%0, %2)      , (%1)       , (%1, %3))
-SCALED_CPY((%0, %2, 2), (%%eax, %2, 2), (%1, %3, 2), (%%ebx, %3, 2))
-SCALED_CPY((%0, %2, 4), (%%eax, %2, 4), (%1, %3, 4), (%%ebx, %3, 4))
-                                               "leal (%%eax,%2,4), %%eax       \n\t"
-                                               "leal (%%ebx,%3,4), %%ebx       \n\t"
-SCALED_CPY((%%eax, %2), (%%eax, %2, 2), (%%ebx, %3), (%%ebx, %3, 2))
+SCALED_CPY((%2)       , (%2, %4)      , (%3)       , (%3, %5))
+SCALED_CPY((%2, %4, 2), (%%eax, %4, 2), (%3, %5, 2), (%%edx, %5, 2))
+SCALED_CPY((%2, %4, 4), (%%eax, %4, 4), (%3, %5, 4), (%%edx, %5, 4))
+                                               "leal (%%eax,%4,4), %%eax       \n\t"
+                                               "leal (%%edx,%5,4), %%edx       \n\t"
+SCALED_CPY((%%eax, %4), (%%eax, %4, 2), (%%edx, %5), (%%edx, %5, 2))
 
 
-                                               : : "r"(src),
+                                               : "=&a" (packedOffsetAndScale)
+                                               : "0" (packedOffsetAndScale),
+                                               "r"(src),
                                                "r"(dst),
                                                "r" (srcStride),
                                                "r" (dstStride)
-                                               : "%eax", "%ebx"
+                                               : "%edx"
                                        );
 #else
                                for(i=0; i<8; i++)
@@ -2543,7 +2565,7 @@ SCALED_CPY((%%eax, %2), (%%eax, %2, 2), (%%ebx, %3), (%%ebx, %3, 2))
 #ifdef HAVE_MMX
                                        asm volatile(
                                                "leal (%0,%2), %%eax    \n\t"
-                                               "leal (%1,%3), %%ebx    \n\t"
+                                               "leal (%1,%3), %%edx    \n\t"
 
 #define SIMPLE_CPY(src1, src2, dst1, dst2)                             \
                                                "movq " #src1 ", %%mm0  \n\t"\
@@ -2552,17 +2574,17 @@ SCALED_CPY((%%eax, %2), (%%eax, %2, 2), (%%ebx, %3), (%%ebx, %3, 2))
                                                "movq %%mm1, " #dst2 "  \n\t"\
 
 SIMPLE_CPY((%0)       , (%0, %2)      , (%1)       , (%1, %3))
-SIMPLE_CPY((%0, %2, 2), (%%eax, %2, 2), (%1, %3, 2), (%%ebx, %3, 2))
-SIMPLE_CPY((%0, %2, 4), (%%eax, %2, 4), (%1, %3, 4), (%%ebx, %3, 4))
+SIMPLE_CPY((%0, %2, 2), (%%eax, %2, 2), (%1, %3, 2), (%%edx, %3, 2))
+SIMPLE_CPY((%0, %2, 4), (%%eax, %2, 4), (%1, %3, 4), (%%edx, %3, 4))
                                                "leal (%%eax,%2,4), %%eax       \n\t"
-                                               "leal (%%ebx,%3,4), %%ebx       \n\t"
-SIMPLE_CPY((%%eax, %2), (%%eax, %2, 2), (%%ebx, %3), (%%ebx, %3, 2))
+                                               "leal (%%edx,%3,4), %%edx       \n\t"
+SIMPLE_CPY((%%eax, %2), (%%eax, %2, 2), (%%edx, %3), (%%edx, %3, 2))
 
                                                : : "r" (src),
                                                "r" (dst),
                                                "r" (srcStride),
                                                "r" (dstStride)
-                                               : "%eax", "%ebx"
+                                               : "%eax", "%edx"
                                        );
 #else
                                for(i=0; i<8; i++)
@@ -2602,117 +2624,60 @@ static inline void RENAME(duplicate)(uint8_t src[], int stride)
  * Filters array of bytes (Y or U or V values)
  */
 static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int dstStride, int width, int height,
-       QP_STORE_T QPs[], int QPStride, int isColor, struct PPMode *ppMode)
+       QP_STORE_T QPs[], int QPStride, int isColor, PPContext *c2)
 {
+       PPContext __attribute__((aligned(8))) c= *c2; //copy to stack for faster access
        int x,y;
 #ifdef COMPILE_TIME_MODE
        const int mode= COMPILE_TIME_MODE;
 #else
-       const int mode= isColor ? ppMode->chromMode : ppMode->lumMode;
+       const int mode= isColor ? c.ppMode.chromMode : c.ppMode.lumMode;
 #endif
-       /* we need 64bit here otherwise we´ll going to have a problem
-          after watching a black picture for 5 hours*/
-       static uint64_t *yHistogram= NULL;
        int black=0, white=255; // blackest black and whitest white in the picture
        int QPCorrecture= 256*256;
 
-       /* Temporary buffers for handling the last row(s) */
-       static uint8_t *tempDst= NULL;
-       static uint8_t *tempSrc= NULL;
-
-       /* Temporary buffers for handling the last block */
-       static uint8_t *tempDstBlock= NULL;
-       static uint8_t *tempSrcBlock= NULL;
-
-       /* Temporal noise reducing buffers */
-       static uint8_t *tempBlured[3]= {NULL,NULL,NULL};
-       static uint32_t *tempBluredPast[3]= {NULL,NULL,NULL};
-
        int copyAhead;
 
-#ifdef PP_FUNNY_STRIDE
-       uint8_t *dstBlockPtrBackup;
-       uint8_t *srcBlockPtrBackup;
-#endif
+       //FIXME remove
+       uint64_t * const yHistogram= c.yHistogram;
+       uint8_t * const tempSrc= c.tempSrc;
+       uint8_t * const tempDst= c.tempDst;
 
-#ifdef MORE_TIMING
-       long long T0, T1, diffTime=0;
-#endif
-#ifdef TIMING
-       long long memcpyTime=0, vertTime=0, horizTime=0, sumTime;
-       sumTime= rdtsc();
-#endif
-       dcOffset= ppMode->maxDcDiff;
-       dcThreshold= ppMode->maxDcDiff*2 + 1;
+       c.dcOffset= c.ppMode.maxDcDiff;
+       c.dcThreshold= c.ppMode.maxDcDiff*2 + 1;
 
 #ifdef HAVE_MMX
-       maxTmpNoise[0]= ppMode->maxTmpNoise[0];
-       maxTmpNoise[1]= ppMode->maxTmpNoise[1];
-       maxTmpNoise[2]= ppMode->maxTmpNoise[2];
-       
-       mmxDCOffset= 0x7F - dcOffset;
-       mmxDCThreshold= 0x7F - dcThreshold;
+       c.mmxDcOffset= 0x7F - c.dcOffset;
+       c.mmxDcThreshold= 0x7F - c.dcThreshold;
 
-       mmxDCOffset*= 0x0101010101010101LL;
-       mmxDCThreshold*= 0x0101010101010101LL;
+       c.mmxDcOffset*= 0x0101010101010101LL;
+       c.mmxDcThreshold*= 0x0101010101010101LL;
 #endif
 
        if(mode & CUBIC_IPOL_DEINT_FILTER) copyAhead=16;
-       else if(mode & LINEAR_BLEND_DEINT_FILTER) copyAhead=14;
+       else if(   (mode & LINEAR_BLEND_DEINT_FILTER)
+               || (mode & FFMPEG_DEINT_FILTER)) copyAhead=14;
        else if(   (mode & V_DEBLOCK)
                || (mode & LINEAR_IPOL_DEINT_FILTER)
                || (mode & MEDIAN_DEINT_FILTER)) copyAhead=13;
        else if(mode & V_X1_FILTER) copyAhead=11;
-       else if(mode & V_RK1_FILTER) copyAhead=10;
+//     else if(mode & V_RK1_FILTER) copyAhead=10;
        else if(mode & DERING) copyAhead=9;
        else copyAhead=8;
 
        copyAhead-= 8;
 
-       if(tempDst==NULL)
-       {
-               tempDst= (uint8_t*)memalign(8, 1024*24);
-               tempSrc= (uint8_t*)memalign(8, 1024*24);
-               tempDstBlock= (uint8_t*)memalign(8, 1024*24);
-               tempSrcBlock= (uint8_t*)memalign(8, 1024*24);
-       }
-
-       if(tempBlured[isColor]==NULL && (mode & TEMP_NOISE_FILTER))
-       {
-//             printf("%d %d %d\n", isColor, dstStride, height);
-               //FIXME works only as long as the size doesnt increase
-               //Note:the +17*1024 is just there so i dont have to worry about r/w over te end
-               tempBlured[isColor]= (uint8_t*)memalign(8, dstStride*((height+7)&(~7)) + 17*1024);
-               tempBluredPast[isColor]= (uint32_t*)memalign(8, 256*((height+7)&(~7))/2 + 17*1024);
-
-               memset(tempBlured[isColor], 0, dstStride*((height+7)&(~7)) + 17*1024);
-               memset(tempBluredPast[isColor], 0, 256*((height+7)&(~7))/2 + 17*1024);
-       }
-
-       if(!yHistogram)
-       {
-               int i;
-               yHistogram= (uint64_t*)malloc(8*256);
-               for(i=0; i<256; i++) yHistogram[i]= width*height/64*15/256;
-
-               if(mode & FULL_Y_RANGE)
-               {
-                       ppMode->maxAllowedY=255;
-                       ppMode->minAllowedY=0;
-               }
-       }
-
        if(!isColor)
        {
                uint64_t sum= 0;
                int i;
-               static int framenum= -1;
                uint64_t maxClipped;
                uint64_t clipped;
                double scale;
 
-               framenum++;
-               if(framenum == 1) yHistogram[0]= width*height/64*15/256;
+               c.frameNum++;
+               // first frame is fscked so we ignore it
+               if(c.frameNum == 1) yHistogram[0]= width*height/64*15/256;
 
                for(i=0; i<256; i++)
                {
@@ -2738,29 +2703,29 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                        clipped-= yHistogram[white];
                }
 
-               scale= (double)(ppMode->maxAllowedY - ppMode->minAllowedY) / (double)(white-black);
+               scale= (double)(c.ppMode.maxAllowedY - c.ppMode.minAllowedY) / (double)(white-black);
 
 #ifdef HAVE_MMX2
-               packedYScale= (uint16_t)(scale*256.0 + 0.5);
-               packedYOffset= (((black*packedYScale)>>8) - ppMode->minAllowedY) & 0xFFFF;
+               c.packedYScale= (uint16_t)(scale*256.0 + 0.5);
+               c.packedYOffset= (((black*c.packedYScale)>>8) - c.ppMode.minAllowedY) & 0xFFFF;
 #else
-               packedYScale= (uint16_t)(scale*1024.0 + 0.5);
-               packedYOffset= (black - ppMode->minAllowedY) & 0xFFFF;
+               c.packedYScale= (uint16_t)(scale*1024.0 + 0.5);
+               c.packedYOffset= (black - c.ppMode.minAllowedY) & 0xFFFF;
 #endif
 
-               packedYOffset|= packedYOffset<<32;
-               packedYOffset|= packedYOffset<<16;
+               c.packedYOffset|= c.packedYOffset<<32;
+               c.packedYOffset|= c.packedYOffset<<16;
 
-               packedYScale|= packedYScale<<32;
-               packedYScale|= packedYScale<<16;
+               c.packedYScale|= c.packedYScale<<32;
+               c.packedYScale|= c.packedYScale<<16;
                
                if(mode & LEVEL_FIX)    QPCorrecture= (int)(scale*256*256 + 0.5);
                else                    QPCorrecture= 256*256;
        }
        else
        {
-               packedYScale= 0x0100010001000100LL;
-               packedYOffset= 0;
+               c.packedYScale= 0x0100010001000100LL;
+               c.packedYOffset= 0;
                QPCorrecture= 256*256;
        }
 
@@ -2789,18 +2754,18 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                "shrl $2, %%eax                 \n\t"
                                "andl $6, %%eax                 \n\t"
                                "addl %5, %%eax                 \n\t"
-                               "movl %%eax, %%ebx              \n\t"
+                               "movl %%eax, %%edx              \n\t"
                                "imul %1, %%eax                 \n\t"
-                               "imul %3, %%ebx                 \n\t"
+                               "imul %3, %%edx                 \n\t"
                                "prefetchnta 32(%%eax, %0)      \n\t"
-                               "prefetcht0 32(%%ebx, %2)       \n\t"
+                               "prefetcht0 32(%%edx, %2)       \n\t"
                                "addl %1, %%eax                 \n\t"
-                               "addl %3, %%ebx                 \n\t"
+                               "addl %3, %%edx                 \n\t"
                                "prefetchnta 32(%%eax, %0)      \n\t"
-                               "prefetcht0 32(%%ebx, %2)       \n\t"
+                               "prefetcht0 32(%%edx, %2)       \n\t"
                        :: "r" (srcBlock), "r" (srcStride), "r" (dstBlock), "r" (dstStride),
                        "m" (x), "m" (copyAhead)
-                       : "%eax", "%ebx"
+                       : "%eax", "%edx"
                        );
 
 #elif defined(HAVE_3DNOW)
@@ -2813,7 +2778,7 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
 #endif
 
                        RENAME(blockCopy)(dstBlock + dstStride*8, dstStride,
-                               srcBlock + srcStride*8, srcStride, mode & LEVEL_FIX);
+                               srcBlock + srcStride*8, srcStride, mode & LEVEL_FIX, &c.packedYOffset);
 
                        RENAME(duplicate)(dstBlock + dstStride*8, dstStride);
 
@@ -2825,6 +2790,8 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                RENAME(deInterlaceMedian)(dstBlock, dstStride);
                        else if(mode & CUBIC_IPOL_DEINT_FILTER)
                                RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride);
+                       else if(mode & FFMPEG_DEINT_FILTER)
+                               RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x);
 /*                     else if(mode & CUBIC_BLEND_DEINT_FILTER)
                                RENAME(deInterlaceBlendCubic)(dstBlock, dstStride);
 */
@@ -2834,14 +2801,15 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                memcpy(dst, tempDst + 9*dstStride, copyAhead*dstStride );
        }
 
+//printf("\n");
        for(y=0; y<height; y+=BLOCK_SIZE)
        {
                //1% speedup if these are here instead of the inner loop
                uint8_t *srcBlock= &(src[y*srcStride]);
                uint8_t *dstBlock= &(dst[y*dstStride]);
 #ifdef HAVE_MMX
-               uint8_t *tempBlock1= tempBlocks;
-               uint8_t *tempBlock2= tempBlocks + 8;
+               uint8_t *tempBlock1= c.tempBlocks;
+               uint8_t *tempBlock2= c.tempBlocks + 8;
 #endif
 #ifdef ARCH_X86
                int *QPptr= isColor ? &QPs[(y>>3)*QPStride] :&QPs[(y>>4)*QPStride];
@@ -2873,6 +2841,7 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                        dstBlock= tempDst + dstStride;
                        srcBlock= tempSrc;
                }
+//printf("\n");
 
                // From this point on it is guranteed that we can read and write 16 lines downward
                // finish 1 block before the next otherwise we´ll might have a problem
@@ -2904,20 +2873,20 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                QP= (QP* QPCorrecture + 256*128)>>16;
                                yHistogram[ srcBlock[srcStride*12 + 4] ]++;
                        }
+//printf("%d ", QP);
+                       c.QP= QP;
 #ifdef HAVE_MMX
                        asm volatile(
-                               "movd %0, %%mm7                                 \n\t"
+                               "movd %1, %%mm7                                 \n\t"
                                "packuswb %%mm7, %%mm7                          \n\t" // 0, 0, 0, QP, 0, 0, 0, QP
                                "packuswb %%mm7, %%mm7                          \n\t" // 0,QP, 0, QP, 0,QP, 0, QP
                                "packuswb %%mm7, %%mm7                          \n\t" // QP,..., QP
-                               "movq %%mm7, "MANGLE(pQPb)"                     \n\t"
-                               : : "r" (QP)
+                               "movq %%mm7, %0                 \n\t"
+                               : "=m" (c.pQPb) 
+                               : "r" (QP)
                        );
 #endif
 
-#ifdef MORE_TIMING
-                       T0= rdtsc();
-#endif
 
 #ifdef HAVE_MMX2
 /*
@@ -2932,18 +2901,18 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                "shrl $2, %%eax                 \n\t"
                                "andl $6, %%eax                 \n\t"
                                "addl %5, %%eax                 \n\t"
-                               "movl %%eax, %%ebx              \n\t"
+                               "movl %%eax, %%edx              \n\t"
                                "imul %1, %%eax                 \n\t"
-                               "imul %3, %%ebx                 \n\t"
+                               "imul %3, %%edx                 \n\t"
                                "prefetchnta 32(%%eax, %0)      \n\t"
-                               "prefetcht0 32(%%ebx, %2)       \n\t"
+                               "prefetcht0 32(%%edx, %2)       \n\t"
                                "addl %1, %%eax                 \n\t"
-                               "addl %3, %%ebx                 \n\t"
+                               "addl %3, %%edx                 \n\t"
                                "prefetchnta 32(%%eax, %0)      \n\t"
-                               "prefetcht0 32(%%ebx, %2)       \n\t"
+                               "prefetcht0 32(%%edx, %2)       \n\t"
                        :: "r" (srcBlock), "r" (srcStride), "r" (dstBlock), "r" (dstStride),
                        "m" (x), "m" (copyAhead)
-                       : "%eax", "%ebx"
+                       : "%eax", "%edx"
                        );
 
 #elif defined(HAVE_3DNOW)
@@ -2955,27 +2924,8 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
 */
 #endif
 
-#ifdef PP_FUNNY_STRIDE
-                       //can we mess with a 8x16 block, if not use a temp buffer, yes again
-                       if(x+7 >= width)
-                       {
-                               int i;
-                               dstBlockPtrBackup= dstBlock;
-                               srcBlockPtrBackup= srcBlock;
-
-                               for(i=0;i<BLOCK_SIZE*2; i++)
-                               {
-                                       memcpy(tempSrcBlock+i*srcStride, srcBlock+i*srcStride, width-x);
-                                       memcpy(tempDstBlock+i*dstStride, dstBlock+i*dstStride, width-x);
-                               }
-
-                               dstBlock= tempDstBlock;
-                               srcBlock= tempSrcBlock;
-                       }
-#endif
-
                        RENAME(blockCopy)(dstBlock + dstStride*copyAhead, dstStride,
-                               srcBlock + srcStride*copyAhead, srcStride, mode & LEVEL_FIX);
+                               srcBlock + srcStride*copyAhead, srcStride, mode & LEVEL_FIX, &c.packedYOffset);
 
                        if(mode & LINEAR_IPOL_DEINT_FILTER)
                                RENAME(deInterlaceInterpolateLinear)(dstBlock, dstStride);
@@ -2985,6 +2935,8 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                RENAME(deInterlaceMedian)(dstBlock, dstStride);
                        else if(mode & CUBIC_IPOL_DEINT_FILTER)
                                RENAME(deInterlaceInterpolateCubic)(dstBlock, dstStride);
+                       else if(mode & FFMPEG_DEINT_FILTER)
+                               RENAME(deInterlaceFF)(dstBlock, dstStride, c.deintTemp + x);
 /*                     else if(mode & CUBIC_BLEND_DEINT_FILTER)
                                RENAME(deInterlaceBlendCubic)(dstBlock, dstStride);
 */
@@ -2992,30 +2944,18 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                        /* only deblock if we have 2 blocks */
                        if(y + 8 < height)
                        {
-#ifdef MORE_TIMING
-                               T1= rdtsc();
-                               memcpyTime+= T1-T0;
-                               T0=T1;
-#endif
-                               if(mode & V_RK1_FILTER)
-                                       RENAME(vertRK1Filter)(dstBlock, stride, QP);
-                               else if(mode & V_X1_FILTER)
-                                       RENAME(vertX1Filter)(dstBlock, stride, QP);
+                               if(mode & V_X1_FILTER)
+                                       RENAME(vertX1Filter)(dstBlock, stride, &c);
                                else if(mode & V_DEBLOCK)
                                {
-                                       if( RENAME(isVertDC)(dstBlock, stride))
+                                       if( RENAME(isVertDC)(dstBlock, stride, &c))
                                        {
-                                               if(RENAME(isVertMinMaxOk)(dstBlock, stride, QP))
-                                                       RENAME(doVertLowPass)(dstBlock, stride, QP);
+                                               if(RENAME(isVertMinMaxOk)(dstBlock, stride, &c))
+                                                       RENAME(doVertLowPass)(dstBlock, stride, &c);
                                        }
                                        else
-                                               RENAME(doVertDefFilter)(dstBlock, stride, QP);
+                                               RENAME(doVertDefFilter)(dstBlock, stride, &c);
                                }
-#ifdef MORE_TIMING
-                               T1= rdtsc();
-                               vertTime+= T1-T0;
-                               T0=T1;
-#endif
                        }
 
 #ifdef HAVE_MMX
@@ -3024,23 +2964,18 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                        /* check if we have a previous block to deblock it with dstBlock */
                        if(x - 8 >= 0)
                        {
-#ifdef MORE_TIMING
-                               T0= rdtsc();
-#endif
 #ifdef HAVE_MMX
-                               if(mode & H_RK1_FILTER)
-                                       RENAME(vertRK1Filter)(tempBlock1, 16, QP);
-                               else if(mode & H_X1_FILTER)
-                                       RENAME(vertX1Filter)(tempBlock1, 16, QP);
+                               if(mode & H_X1_FILTER)
+                                       RENAME(vertX1Filter)(tempBlock1, 16, &c);
                                else if(mode & H_DEBLOCK)
                                {
-                                       if( RENAME(isVertDC)(tempBlock1, 16)
+                                       if( RENAME(isVertDC)(tempBlock1, 16, &c))
                                        {
-                                               if(RENAME(isVertMinMaxOk)(tempBlock1, 16, QP))
-                                                       RENAME(doVertLowPass)(tempBlock1, 16, QP);
+                                               if(RENAME(isVertMinMaxOk)(tempBlock1, 16, &c))
+                                                       RENAME(doVertLowPass)(tempBlock1, 16, &c);
                                        }
                                        else
-                                               RENAME(doVertDefFilter)(tempBlock1, 16, QP);
+                                               RENAME(doVertDefFilter)(tempBlock1, 16, &c);
                                }
 
                                RENAME(transpose2)(dstBlock-4, dstStride, tempBlock1 + 4*16);
@@ -3050,7 +2985,7 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                        horizX1Filter(dstBlock-4, stride, QP);
                                else if(mode & H_DEBLOCK)
                                {
-                                       if( isHorizDC(dstBlock-4, stride))
+                                       if( isHorizDC(dstBlock-4, stride, &c))
                                        {
                                                if(isHorizMinMaxOk(dstBlock-4, stride, QP))
                                                        doHorizLowPass(dstBlock-4, stride, QP);
@@ -3058,42 +2993,22 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
                                        else
                                                doHorizDefFilter(dstBlock-4, stride, QP);
                                }
-#endif
-#ifdef MORE_TIMING
-                               T1= rdtsc();
-                               horizTime+= T1-T0;
-                               T0=T1;
 #endif
                                if(mode & DERING)
                                {
                                //FIXME filter first line
-                                       if(y>0) RENAME(dering)(dstBlock - stride - 8, stride, QP);
+                                       if(y>0) RENAME(dering)(dstBlock - stride - 8, stride, &c);
                                }
 
                                if(mode & TEMP_NOISE_FILTER)
                                {
                                        RENAME(tempNoiseReducer)(dstBlock-8, stride,
-                                               tempBlured[isColor] + y*dstStride + x,
-                                               tempBluredPast[isColor] + (y>>3)*256 + (x>>3),
-                                               ppMode->maxTmpNoise);
+                                               c.tempBlured[isColor] + y*dstStride + x,
+                                               c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3),
+                                               c.ppMode.maxTmpNoise);
                                }
                        }
 
-#ifdef PP_FUNNY_STRIDE
-                       /* did we use a tmp-block buffer */
-                       if(x+7 >= width)
-                       {
-                               int i;
-                               dstBlock= dstBlockPtrBackup;
-                               srcBlock= srcBlockPtrBackup;
-
-                               for(i=0;i<BLOCK_SIZE*2; i++)
-                               {
-                                       memcpy(dstBlock+i*dstStride, tempDstBlock+i*dstStride, width-x);
-                               }
-                       }
-#endif
-
                        dstBlock+=8;
                        srcBlock+=8;
 
@@ -3106,15 +3021,15 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
 
                if(mode & DERING)
                {
-                               if(y > 0) RENAME(dering)(dstBlock - dstStride - 8, dstStride, QP);
+                               if(y > 0) RENAME(dering)(dstBlock - dstStride - 8, dstStride, &c);
                }
 
                if((mode & TEMP_NOISE_FILTER))
                {
                        RENAME(tempNoiseReducer)(dstBlock-8, dstStride,
-                               tempBlured[isColor] + y*dstStride + x,
-                               tempBluredPast[isColor] + (y>>3)*256 + (x>>3),
-                               ppMode->maxTmpNoise);
+                               c.tempBlured[isColor] + y*dstStride + x,
+                               c.tempBluredPast[isColor] + (y>>3)*256 + (x>>3),
+                               c.ppMode.maxTmpNoise);
                }
 
                /* did we use a tmp buffer for the last lines*/
@@ -3140,15 +3055,6 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
        asm volatile("emms");
 #endif
 
-#ifdef TIMING
-       // FIXME diff is mostly the time spent for rdtsc (should subtract that but ...)
-       sumTime= rdtsc() - sumTime;
-       if(!isColor)
-               printf("cpy:%4dk, vert:%4dk, horiz:%4dk, sum:%4dk, diff:%4dk, color: %d/%d    \r",
-                       (int)(memcpyTime/1000), (int)(vertTime/1000), (int)(horizTime/1000),
-                       (int)(sumTime/1000), (int)((sumTime-memcpyTime-vertTime-horizTime)/1000)
-                       , black, white);
-#endif
 #ifdef DEBUG_BRIGHTNESS
        if(!isColor)
        {
@@ -3176,4 +3082,6 @@ static void RENAME(postProcess)(uint8_t src[], int srcStride, uint8_t dst[], int
        }
 #endif
 
+       *c2= c; //copy local context back
+
 }