diff --git a/libjpegtwrp/jdmerge.c b/libjpegtwrp/jdmerge.c
new file mode 100644
index 0000000..171fbd3
--- /dev/null
+++ b/libjpegtwrp/jdmerge.c
@@ -0,0 +1,902 @@
+/*
+ * jdmerge.c
+ *
+ * Copyright (C) 1994-1996, Thomas G. Lane.
+ * Copyright (c) 2010, Code Aurora Forum. All rights reserved.
+ * This file is part of the Independent JPEG Group's software.
+ * For conditions of distribution and use, see the accompanying README file.
+ *
+ * This file contains code for merged upsampling/color conversion.
+ *
+ * This file combines functions from jdsample.c and jdcolor.c;
+ * read those files first to understand what's going on.
+ *
+ * When the chroma components are to be upsampled by simple replication
+ * (ie, box filtering), we can save some work in color conversion by
+ * calculating all the output pixels corresponding to a pair of chroma
+ * samples at one time.  In the conversion equations
+ *	R = Y           + K1 * Cr
+ *	G = Y + K2 * Cb + K3 * Cr
+ *	B = Y + K4 * Cb
+ * only the Y term varies among the group of pixels corresponding to a pair
+ * of chroma samples, so the rest of the terms can be calculated just once.
+ * At typical sampling ratios, this eliminates half or three-quarters of the
+ * multiplications needed for color conversion.
+ *
+ * This file currently provides implementations for the following cases:
+ *	YCbCr => RGB color conversion only.
+ *	Sampling ratios of 2h1v or 2h2v.
+ *	No scaling needed at upsample time.
+ *	Corner-aligned (non-CCIR601) sampling alignment.
+ * Other special cases could be added, but in most applications these are
+ * the only common cases.  (For uncommon cases we fall back on the more
+ * general code in jdsample.c and jdcolor.c.)
+ */
+
+#define JPEG_INTERNALS
+#include "jinclude.h"
+#include "jpeglib.h"
+
+#ifdef UPSAMPLE_MERGING_SUPPORTED
+
+#ifdef ANDROID_RGB
+
+/* Declarations for ordered dithering.
+ * 
+ * We use 4x4 ordered dither array packed into 32 bits. This array is
+ * sufficent for dithering RGB_888 to RGB_565.
+ */
+
+#define DITHER_MASK         0x3
+#define DITHER_ROTATE(x)    (((x)<<24) | (((x)>>8)&0x00FFFFFF))
+static const INT32 dither_matrix[4] = {
+  0x0008020A,
+  0x0C040E06,
+  0x030B0109,
+  0x0F070D05
+};
+
+#endif
+
+/* Private subobject */
+
+typedef struct {
+  struct jpeg_upsampler pub;	/* public fields */
+
+  /* Pointer to routine to do actual upsampling/conversion of one row group */
+  JMETHOD(void, upmethod, (j_decompress_ptr cinfo,
+			   JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+			   JSAMPARRAY output_buf));
+
+  /* Private state for YCC->RGB conversion */
+  int * Cr_r_tab;		/* => table for Cr to R conversion */
+  int * Cb_b_tab;		/* => table for Cb to B conversion */
+  INT32 * Cr_g_tab;		/* => table for Cr to G conversion */
+  INT32 * Cb_g_tab;		/* => table for Cb to G conversion */
+
+  /* For 2:1 vertical sampling, we produce two output rows at a time.
+   * We need a "spare" row buffer to hold the second output row if the
+   * application provides just a one-row buffer; we also use the spare
+   * to discard the dummy last row if the image height is odd.
+   */
+  JSAMPROW spare_row;
+  boolean spare_full;		/* T if spare buffer is occupied */
+
+  JDIMENSION out_row_width;	/* samples per output row */
+  JDIMENSION rows_to_go;	/* counts rows remaining in image */
+} my_upsampler;
+
+typedef my_upsampler * my_upsample_ptr;
+
+#define SCALEBITS	16	/* speediest right-shift on some machines */
+#define ONE_HALF	((INT32) 1 << (SCALEBITS-1))
+#define FIX(x)		((INT32) ((x) * (1L<<SCALEBITS) + 0.5))
+
+
+/*
+ * Initialize tables for YCC->RGB colorspace conversion.
+ * This is taken directly from jdcolor.c; see that file for more info.
+ */
+
+LOCAL(void)
+build_ycc_rgb_table (j_decompress_ptr cinfo)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  int i;
+  INT32 x;
+  SHIFT_TEMPS
+
+  upsample->Cr_r_tab = (int *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+				(MAXJSAMPLE+1) * SIZEOF(int));
+  upsample->Cb_b_tab = (int *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+				(MAXJSAMPLE+1) * SIZEOF(int));
+  upsample->Cr_g_tab = (INT32 *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+				(MAXJSAMPLE+1) * SIZEOF(INT32));
+  upsample->Cb_g_tab = (INT32 *)
+    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+				(MAXJSAMPLE+1) * SIZEOF(INT32));
+
+  for (i = 0, x = -CENTERJSAMPLE; i <= MAXJSAMPLE; i++, x++) {
+    /* i is the actual input pixel value, in the range 0..MAXJSAMPLE */
+    /* The Cb or Cr value we are thinking of is x = i - CENTERJSAMPLE */
+    /* Cr=>R value is nearest int to 1.40200 * x */
+    upsample->Cr_r_tab[i] = (int)
+		    RIGHT_SHIFT(FIX(1.40200) * x + ONE_HALF, SCALEBITS);
+    /* Cb=>B value is nearest int to 1.77200 * x */
+    upsample->Cb_b_tab[i] = (int)
+		    RIGHT_SHIFT(FIX(1.77200) * x + ONE_HALF, SCALEBITS);
+    /* Cr=>G value is scaled-up -0.71414 * x */
+    upsample->Cr_g_tab[i] = (- FIX(0.71414)) * x;
+    /* Cb=>G value is scaled-up -0.34414 * x */
+    /* We also add in ONE_HALF so that need not do it in inner loop */
+    upsample->Cb_g_tab[i] = (- FIX(0.34414)) * x + ONE_HALF;
+  }
+}
+
+
+/*
+ * Initialize for an upsampling pass.
+ */
+
+METHODDEF(void)
+start_pass_merged_upsample (j_decompress_ptr cinfo)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+
+  /* Mark the spare buffer empty */
+  upsample->spare_full = FALSE;
+  /* Initialize total-height counter for detecting bottom of image */
+  upsample->rows_to_go = cinfo->output_height;
+}
+
+
+/*
+ * Control routine to do upsampling (and color conversion).
+ *
+ * The control routine just handles the row buffering considerations.
+ */
+
+METHODDEF(void)
+merged_2v_upsample (j_decompress_ptr cinfo,
+		    JSAMPIMAGE input_buf, JDIMENSION *in_row_group_ctr,
+		    JDIMENSION in_row_groups_avail,
+		    JSAMPARRAY output_buf, JDIMENSION *out_row_ctr,
+		    JDIMENSION out_rows_avail)
+/* 2:1 vertical sampling case: may need a spare row. */
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  JSAMPROW work_ptrs[2];
+  JDIMENSION num_rows;		/* number of rows returned to caller */
+
+  if (upsample->spare_full) {
+    /* If we have a spare row saved from a previous cycle, just return it. */
+      JDIMENSION size = upsample->out_row_width;
+#ifdef ANDROID_RGB
+    if (cinfo->out_color_space == JCS_RGB_565)
+      size = cinfo->output_width*2;
+#endif
+    jcopy_sample_rows(& upsample->spare_row, 0, output_buf + *out_row_ctr, 0,
+		      1, size);
+
+    num_rows = 1;
+    upsample->spare_full = FALSE;
+  } else {
+    /* Figure number of rows to return to caller. */
+    num_rows = 2;
+    /* Not more than the distance to the end of the image. */
+    if (num_rows > upsample->rows_to_go)
+      num_rows = upsample->rows_to_go;
+    /* And not more than what the client can accept: */
+    out_rows_avail -= *out_row_ctr;
+    if (num_rows > out_rows_avail)
+      num_rows = out_rows_avail;
+    /* Create output pointer array for upsampler. */
+    work_ptrs[0] = output_buf[*out_row_ctr];
+    if (num_rows > 1) {
+      work_ptrs[1] = output_buf[*out_row_ctr + 1];
+    } else {
+      work_ptrs[1] = upsample->spare_row;
+      upsample->spare_full = TRUE;
+    }
+    /* Now do the upsampling. */
+    (*upsample->upmethod) (cinfo, input_buf, *in_row_group_ctr, work_ptrs);
+  }
+
+  /* Adjust counts */
+  *out_row_ctr += num_rows;
+  upsample->rows_to_go -= num_rows;
+  /* When the buffer is emptied, declare this input row group consumed */
+  if (! upsample->spare_full)
+    (*in_row_group_ctr)++;
+}
+
+
+METHODDEF(void)
+merged_1v_upsample (j_decompress_ptr cinfo,
+		    JSAMPIMAGE input_buf, JDIMENSION *in_row_group_ctr,
+		    JDIMENSION in_row_groups_avail,
+		    JSAMPARRAY output_buf, JDIMENSION *out_row_ctr,
+		    JDIMENSION out_rows_avail)
+/* 1:1 vertical sampling case: much easier, never need a spare row. */
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+
+  /* Just do the upsampling. */
+  (*upsample->upmethod) (cinfo, input_buf, *in_row_group_ctr,
+			 output_buf + *out_row_ctr);
+  /* Adjust counts */
+  (*out_row_ctr)++;
+  (*in_row_group_ctr)++;
+}
+
+
+/*
+ * These are the routines invoked by the control routines to do
+ * the actual upsampling/conversion.  One row group is processed per call.
+ *
+ * Note: since we may be writing directly into application-supplied buffers,
+ * we have to be honest about the output width; we can't assume the buffer
+ * has been rounded up to an even width.
+ */
+
+
+/*
+ * Upsample and color convert for the case of 2:1 horizontal and 1:1 vertical.
+ */
+
+METHODDEF(void)
+h2v1_merged_upsample (j_decompress_ptr cinfo,
+		      JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+		      JSAMPARRAY output_buf)
+#ifdef ANDROID_JPEG_USE_VENUM
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  JSAMPROW inptr0, inptr1, inptr2;
+  JSAMPROW outptr;
+
+  inptr0 = input_buf[0][in_row_group_ctr];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr = output_buf[0];
+
+#ifdef ANDROID_RGB
+  if (cinfo->out_color_space == JCS_RGBA_8888) {
+    yyvup2abgr8888_venum((UINT8*) inptr0,
+                         (UINT8*) inptr2,
+                         (UINT8*) inptr1,
+                         (UINT8*) outptr,
+                         cinfo->output_width);
+  } else
+#endif
+  {
+    yyvup2bgr888_venum((UINT8*) inptr0,
+                       (UINT8*) inptr2,
+                       (UINT8*) inptr1,
+                       (UINT8*) outptr,
+                       cinfo->output_width);
+  }
+}
+#else
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr;
+  JSAMPROW inptr0, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  JDIMENSION pixelWidth = RGB_PIXELSIZE;
+  SHIFT_TEMPS
+
+#ifdef ANDROID_RGB
+  if ((cinfo->out_color_components == 4) &&
+      (cinfo->out_color_space == JCS_RGBA_8888))
+     pixelWidth = 4;  // RGBA pixel size
+#endif /* ANDROID_RGB */
+
+  inptr0 = input_buf[0][in_row_group_ctr];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr = output_buf[0];
+  /* Loop for each pair of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 2 Y values and emit 2 pixels */
+    y  = GETJSAMPLE(*inptr0++);
+    outptr[RGB_RED] = range_limit[y + cred];
+    outptr[RGB_GREEN] = range_limit[y + cgreen];
+    outptr[RGB_BLUE] = range_limit[y + cblue];
+    outptr += pixelWidth;
+    y  = GETJSAMPLE(*inptr0++);
+    outptr[RGB_RED] = range_limit[y + cred];
+    outptr[RGB_GREEN] = range_limit[y + cgreen];
+    outptr[RGB_BLUE] = range_limit[y + cblue];
+    outptr += pixelWidth;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr0);
+    outptr[RGB_RED] = range_limit[y + cred];
+    outptr[RGB_GREEN] = range_limit[y + cgreen];
+    outptr[RGB_BLUE] = range_limit[y + cblue];
+  }
+}
+#endif /* ANDROID_JPEG_USE_VENUM */
+
+
+#ifdef ANDROID_RGB
+METHODDEF(void)
+h2v1_merged_upsample_565 (j_decompress_ptr cinfo,
+              JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+              JSAMPARRAY output_buf)
+#if defined(ANDROID_JPEG_USE_VENUM)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  JSAMPROW inptr0, inptr1, inptr2;
+  JSAMPROW outptr;
+
+  inptr0 = input_buf[0][in_row_group_ctr];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr = output_buf[0];
+
+  yyvup2rgb565_venum((UINT8*) inptr0,
+                     (UINT8*) inptr2,
+                     (UINT8*) inptr1,
+                     (UINT8*) outptr,
+                     cinfo->output_width);
+}
+#else
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr;
+  JSAMPROW inptr0, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  unsigned int r, g, b;
+  INT32 rgb;
+  SHIFT_TEMPS
+
+  inptr0 = input_buf[0][in_row_group_ctr];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr = output_buf[0];
+  /* Loop for each pair of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 2 Y values and emit 2 pixels */
+    y  = GETJSAMPLE(*inptr0++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr0++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr, rgb);
+    outptr += 4;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr0);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_SHORT_565(r,g,b);
+    *(INT16*)outptr = rgb;
+  }
+}
+#endif /* ANDROID_JPEG_USE_VENUM */
+
+
+METHODDEF(void)
+h2v1_merged_upsample_565D (j_decompress_ptr cinfo,
+              JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+              JSAMPARRAY output_buf)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr;
+  JSAMPROW inptr0, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  JDIMENSION col_index = 0;
+  INT32 d0 = dither_matrix[cinfo->output_scanline & DITHER_MASK];
+  unsigned int r, g, b;
+  INT32 rgb;
+  SHIFT_TEMPS
+
+  inptr0 = input_buf[0][in_row_group_ctr];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr = output_buf[0];
+  /* Loop for each pair of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 2 Y values and emit 2 pixels */
+    y  = GETJSAMPLE(*inptr0++);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    d0 = DITHER_ROTATE(d0);
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr0++);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    d0 = DITHER_ROTATE(d0);
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr, rgb);
+    outptr += 4;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr0);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    rgb = PACK_SHORT_565(r,g,b);
+    *(INT16*)outptr = rgb;
+  }
+}
+
+
+#endif
+
+/*
+ * Upsample and color convert for the case of 2:1 horizontal and 2:1 vertical.
+ */
+
+METHODDEF(void)
+h2v2_merged_upsample (j_decompress_ptr cinfo,
+		      JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+		      JSAMPARRAY output_buf)
+#ifdef ANDROID_JPEG_USE_VENUM
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  JSAMPROW outptr0, outptr1;
+  JSAMPROW inptr00, inptr01, inptr1, inptr2;
+  inptr00 = input_buf[0][in_row_group_ctr*2];
+  inptr01 = input_buf[0][in_row_group_ctr*2 + 1];
+  inptr1  = input_buf[1][in_row_group_ctr];
+  inptr2  = input_buf[2][in_row_group_ctr];
+  outptr0 = output_buf[0];
+  outptr1 = output_buf[1];
+#ifdef ANDROID_RGB
+  if (cinfo->out_color_space == JCS_RGBA_8888) {
+    yyvup2abgr8888_venum((UINT8*) inptr00,
+                         (UINT8*) inptr2,
+                         (UINT8*) inptr1,
+                         (UINT8*) outptr0,
+                         cinfo->output_width);
+    yyvup2abgr8888_venum((UINT8*) inptr01,
+                         (UINT8*) inptr2,
+                         (UINT8*) inptr1,
+                         (UINT8*) outptr1,
+                         cinfo->output_width);
+  } else
+#endif
+  {
+    yyvup2bgr888_venum((UINT8*) inptr00,
+                       (UINT8*) inptr2,
+                       (UINT8*) inptr1,
+                       (UINT8*) outptr0,
+                       cinfo->output_width);
+
+    yyvup2bgr888_venum((UINT8*) inptr01,
+                       (UINT8*) inptr2,
+                       (UINT8*) inptr1,
+                       (UINT8*) outptr1,
+                       cinfo->output_width);
+  }
+}
+#else
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr0, outptr1;
+  JSAMPROW inptr00, inptr01, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  JDIMENSION pixelWidth = RGB_PIXELSIZE;
+  SHIFT_TEMPS
+
+#ifdef ANDROID_RGB
+  if ((cinfo->out_color_components == 4) &&
+      (cinfo->out_color_space == JCS_RGBA_8888))
+     pixelWidth = 4;  // RGBA pixel size
+#endif /* ANDROID_RGB */
+
+  inptr00 = input_buf[0][in_row_group_ctr*2];
+  inptr01 = input_buf[0][in_row_group_ctr*2 + 1];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr0 = output_buf[0];
+  outptr1 = output_buf[1];
+  /* Loop for each group of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 4 Y values and emit 4 pixels */
+    y  = GETJSAMPLE(*inptr00++);
+    outptr0[RGB_RED] = range_limit[y + cred];
+    outptr0[RGB_GREEN] = range_limit[y + cgreen];
+    outptr0[RGB_BLUE] = range_limit[y + cblue];
+    outptr0 += pixelWidth;
+    y  = GETJSAMPLE(*inptr00++);
+    outptr0[RGB_RED] = range_limit[y + cred];
+    outptr0[RGB_GREEN] = range_limit[y + cgreen];
+    outptr0[RGB_BLUE] = range_limit[y + cblue];
+    outptr0 += pixelWidth;
+    y  = GETJSAMPLE(*inptr01++);
+    outptr1[RGB_RED] = range_limit[y + cred];
+    outptr1[RGB_GREEN] = range_limit[y + cgreen];
+    outptr1[RGB_BLUE] = range_limit[y + cblue];
+    outptr1 += pixelWidth;
+    y  = GETJSAMPLE(*inptr01++);
+    outptr1[RGB_RED] = range_limit[y + cred];
+    outptr1[RGB_GREEN] = range_limit[y + cgreen];
+    outptr1[RGB_BLUE] = range_limit[y + cblue];
+    outptr1 += pixelWidth;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr00);
+    outptr0[RGB_RED] = range_limit[y + cred];
+    outptr0[RGB_GREEN] = range_limit[y + cgreen];
+    outptr0[RGB_BLUE] = range_limit[y + cblue];
+    y  = GETJSAMPLE(*inptr01);
+    outptr1[RGB_RED] = range_limit[y + cred];
+    outptr1[RGB_GREEN] = range_limit[y + cgreen];
+    outptr1[RGB_BLUE] = range_limit[y + cblue];
+  }
+}
+#endif /* ANDROID_JPEG_USE_VENUM */
+
+
+#ifdef ANDROID_RGB
+
+METHODDEF(void)
+h2v2_merged_upsample_565 (j_decompress_ptr cinfo,
+              JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+              JSAMPARRAY output_buf)
+#if defined(ANDROID_JPEG_USE_VENUM)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  JSAMPROW outptr0, outptr1;
+  JSAMPROW inptr00, inptr01, inptr1, inptr2;
+  inptr00 = input_buf[0][in_row_group_ctr*2];
+  inptr01 = input_buf[0][in_row_group_ctr*2 + 1];
+  inptr1  = input_buf[1][in_row_group_ctr];
+  inptr2  = input_buf[2][in_row_group_ctr];
+  outptr0 = output_buf[0];
+  outptr1 = output_buf[1];
+
+  yyvup2rgb565_venum((UINT8*) inptr00,
+                     (UINT8*) inptr2,
+                     (UINT8*) inptr1,
+                     (UINT8*) outptr0,
+                     cinfo->output_width);
+
+  yyvup2rgb565_venum((UINT8*) inptr01,
+                     (UINT8*) inptr2,
+                     (UINT8*) inptr1,
+                     (UINT8*) outptr1,
+                     cinfo->output_width);
+}
+#else
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr0, outptr1;
+  JSAMPROW inptr00, inptr01, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  unsigned int r, g, b;
+  INT32 rgb;
+  SHIFT_TEMPS
+
+  inptr00 = input_buf[0][in_row_group_ctr*2];
+  inptr01 = input_buf[0][in_row_group_ctr*2 + 1];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr0 = output_buf[0];
+  outptr1 = output_buf[1];
+  /* Loop for each group of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 4 Y values and emit 4 pixels */
+    y  = GETJSAMPLE(*inptr00++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr00++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr0, rgb);
+    outptr0 += 4;
+    y  = GETJSAMPLE(*inptr01++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr01++);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr1, rgb);
+    outptr1 += 4;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr00);
+    r = range_limit[y + cred];
+    g = range_limit[y + cgreen];
+    b = range_limit[y + cblue];
+    rgb = PACK_SHORT_565(r,g,b);
+    *(INT16*)outptr0 = rgb;
+   y  = GETJSAMPLE(*inptr01);
+   r = range_limit[y + cred];
+   g = range_limit[y + cgreen];
+   b = range_limit[y + cblue];
+   rgb = PACK_SHORT_565(r,g,b);
+   *(INT16*)outptr1 = rgb;
+  }
+}
+#endif /* ANDROID_JPEG_USE_VENUM */
+
+
+METHODDEF(void)
+h2v2_merged_upsample_565D (j_decompress_ptr cinfo,
+              JSAMPIMAGE input_buf, JDIMENSION in_row_group_ctr,
+              JSAMPARRAY output_buf)
+{
+  my_upsample_ptr upsample = (my_upsample_ptr) cinfo->upsample;
+  register int y, cred, cgreen, cblue;
+  int cb, cr;
+  register JSAMPROW outptr0, outptr1;
+  JSAMPROW inptr00, inptr01, inptr1, inptr2;
+  JDIMENSION col;
+  /* copy these pointers into registers if possible */
+  register JSAMPLE * range_limit = cinfo->sample_range_limit;
+  int * Crrtab = upsample->Cr_r_tab;
+  int * Cbbtab = upsample->Cb_b_tab;
+  INT32 * Crgtab = upsample->Cr_g_tab;
+  INT32 * Cbgtab = upsample->Cb_g_tab;
+  JDIMENSION col_index = 0;
+  INT32 d0 = dither_matrix[cinfo->output_scanline & DITHER_MASK];
+  INT32 d1 = dither_matrix[(cinfo->output_scanline+1) & DITHER_MASK];
+  unsigned int r, g, b;
+  INT32 rgb;
+  SHIFT_TEMPS
+
+  inptr00 = input_buf[0][in_row_group_ctr*2];
+  inptr01 = input_buf[0][in_row_group_ctr*2 + 1];
+  inptr1 = input_buf[1][in_row_group_ctr];
+  inptr2 = input_buf[2][in_row_group_ctr];
+  outptr0 = output_buf[0];
+  outptr1 = output_buf[1];
+  /* Loop for each group of output pixels */
+  for (col = cinfo->output_width >> 1; col > 0; col--) {
+    
+    /* Do the chroma part of the calculation */
+    cb = GETJSAMPLE(*inptr1++);
+    cr = GETJSAMPLE(*inptr2++);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    /* Fetch 4 Y values and emit 4 pixels */    
+    y  = GETJSAMPLE(*inptr00++);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    d0 = DITHER_ROTATE(d0);
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr00++);
+    r = range_limit[DITHER_565_R(y + cred, d1)];
+    g = range_limit[DITHER_565_G(y + cgreen, d1)];
+    b = range_limit[DITHER_565_B(y + cblue, d1)];
+    d1 = DITHER_ROTATE(d1);
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr0, rgb);
+    outptr0 += 4;
+    y  = GETJSAMPLE(*inptr01++);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    d0 = DITHER_ROTATE(d0);
+    rgb = PACK_SHORT_565(r,g,b);
+    y  = GETJSAMPLE(*inptr01++);
+    r = range_limit[DITHER_565_R(y + cred, d1)];
+    g = range_limit[DITHER_565_G(y + cgreen, d1)];
+    b = range_limit[DITHER_565_B(y + cblue, d1)];
+    d1 = DITHER_ROTATE(d1);
+    rgb = PACK_TWO_PIXELS(rgb, PACK_SHORT_565(r,g,b));
+    WRITE_TWO_PIXELS(outptr1, rgb);
+    outptr1 += 4;
+  }
+  /* If image width is odd, do the last output column separately */
+  if (cinfo->output_width & 1) {
+    cb = GETJSAMPLE(*inptr1);
+    cr = GETJSAMPLE(*inptr2);
+    cred = Crrtab[cr];
+    cgreen = (int) RIGHT_SHIFT(Cbgtab[cb] + Crgtab[cr], SCALEBITS);
+    cblue = Cbbtab[cb];
+    y  = GETJSAMPLE(*inptr00);
+    r = range_limit[DITHER_565_R(y + cred, d0)];
+    g = range_limit[DITHER_565_G(y + cgreen, d0)];
+    b = range_limit[DITHER_565_B(y + cblue, d0)];
+    rgb = PACK_SHORT_565(r,g,b);
+    *(INT16*)outptr0 = rgb;
+   y  = GETJSAMPLE(*inptr01);
+   r = range_limit[DITHER_565_R(y + cred, d1)];
+   g = range_limit[DITHER_565_G(y + cgreen, d1)];
+   b = range_limit[DITHER_565_B(y + cblue, d1)];
+   rgb = PACK_SHORT_565(r,g,b);
+   *(INT16*)outptr1 = rgb;
+  }
+}
+
+#endif
+
+/*
+ * Module initialization routine for merged upsampling/color conversion.
+ *
+ * NB: this is called under the conditions determined by use_merged_upsample()
+ * in jdmaster.c.  That routine MUST correspond to the actual capabilities
+ * of this module; no safety checks are made here.
+ */
+
+GLOBAL(void)
+jinit_merged_upsampler (j_decompress_ptr cinfo)
+{
+  my_upsample_ptr upsample;
+
+  upsample = (my_upsample_ptr)
+    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+				SIZEOF(my_upsampler));
+  cinfo->upsample = (struct jpeg_upsampler *) upsample;
+  upsample->pub.start_pass = start_pass_merged_upsample;
+  upsample->pub.need_context_rows = FALSE;
+
+  upsample->out_row_width = cinfo->output_width * cinfo->out_color_components;
+  
+  if (cinfo->max_v_samp_factor == 2) {
+    upsample->pub.upsample = merged_2v_upsample;
+    upsample->upmethod = h2v2_merged_upsample;
+#ifdef ANDROID_RGB
+    if (cinfo->out_color_space == JCS_RGB_565) {
+#if !defined(ANDROID_JPEG_USE_VENUM)
+      if (cinfo->dither_mode != JDITHER_NONE) {
+        upsample->upmethod = h2v2_merged_upsample_565D;
+      } else
+#endif
+      {
+        /* If VeNum routines are enabled, use h2v2_merged_upsample_565
+         * function regardless of dither mode. */
+        upsample->upmethod = h2v2_merged_upsample_565;
+      }
+    }
+#endif
+    /* Allocate a spare row buffer */
+    upsample->spare_row = (JSAMPROW)
+      (*cinfo->mem->alloc_large) ((j_common_ptr) cinfo, JPOOL_IMAGE,
+		(size_t) (upsample->out_row_width * SIZEOF(JSAMPLE)));
+  } else {
+    upsample->pub.upsample = merged_1v_upsample;
+    upsample->upmethod = h2v1_merged_upsample;
+#ifdef ANDROID_RGB
+    if (cinfo->out_color_space == JCS_RGB_565) {
+#ifndef ANDROID_JPEG_USE_VENUM
+      if (cinfo->dither_mode != JDITHER_NONE) {
+        upsample->upmethod = h2v1_merged_upsample_565D;
+      } else
+#endif
+      {
+        /* If VeNum routines are enabled, use h2v1_merged_upsample_565
+         * function regardless of dither mode. */
+        upsample->upmethod = h2v1_merged_upsample_565;
+      }
+    }
+#endif
+    /* No spare row needed */
+    upsample->spare_row = NULL;
+  }
+
+#if !defined(ANDROID_JPEG_USE_VENUM)
+    build_ycc_rgb_table(cinfo);
+#endif
+#ifdef NEEDS_ARM_ERRATA_754319_754320
+  asm volatile ( "vmov s0,s0\n" );
+#endif
+}
+
+#endif /* UPSAMPLE_MERGING_SUPPORTED */