--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/jdk/src/java.desktop/unix/native/libmlib_image/mlib_v_ImageConvIndex3_8_8nw.c Sun Aug 17 15:54:13 2014 +0100
@@ -0,0 +1,1673 @@
+/*
+ * Copyright (c) 2000, 2003, Oracle and/or its affiliates. All rights reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation. Oracle designates this
+ * particular file as subject to the "Classpath" exception as provided
+ * by Oracle in the LICENSE file that accompanied this code.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+
+
+/*
+ * FUNCTION
+ * Internal functions for mlib_ImageConv* on U8 type
+ * and MLIB_EDGE_DST_NO_WRITE mask
+ *
+ */
+
+/***************************************************************/
+
+#include <vis_proto.h>
+#include <mlib_image.h>
+#include <mlib_ImageCheck.h>
+#include <mlib_ImageColormap.h>
+
+/*
+ This defines switches between functions in
+ files: mlib_v_ImageConv_8nw.c,
+ mlib_v_ImageConvIndex3_8_8nw.c,
+ mlib_v_ImageConvIndex4_8_8nw.c,
+ mlib_v_ImageConvIndex3_8_16nw.c,
+ mlib_v_ImageConvIndex4_8_16nw.c
+*/
+
+#define CONV_INDEX
+
+#define DTYPE mlib_u8
+#define LTYPE mlib_u8
+
+/***************************************************************/
+
+#ifdef CONV_INDEX
+
+#define CONV_FUNC(KERN) \
+ mlib_conv##KERN##_Index3_8_8nw(mlib_image *dst, \
+ mlib_image *src, \
+ mlib_s32 *kern, \
+ mlib_s32 scale, \
+ void *colormap)
+
+#else
+
+#define CONV_FUNC(KERN) \
+ mlib_conv##KERN##_8nw_f(mlib_image *dst, \
+ mlib_image *src, \
+ mlib_s32 *kern, \
+ mlib_s32 scale)
+
+#endif
+
+/***************************************************************/
+
+#ifdef CONV_INDEX
+
+#define NCHAN 3
+
+#else
+
+#define NCHAN nchan
+
+#endif
+
+/***************************************************************/
+
+#define DEF_VARS \
+ DTYPE *sl, *sp, *dl; \
+ mlib_s32 hgt = mlib_ImageGetHeight(src); \
+ mlib_s32 wid = mlib_ImageGetWidth(src); \
+ mlib_s32 sll = mlib_ImageGetStride(src) / sizeof(DTYPE); \
+ mlib_s32 dll = mlib_ImageGetStride(dst) / sizeof(DTYPE); \
+ DTYPE *adr_src = (DTYPE *)mlib_ImageGetData(src); \
+ DTYPE *adr_dst = (DTYPE *)mlib_ImageGetData(dst); \
+ mlib_s32 ssize, xsize, dsize, esize, emask, buff_ind = 0; \
+ mlib_d64 *pbuff, *dp; \
+ mlib_f32 *karr = (mlib_f32 *)kern; \
+ mlib_s32 gsr_scale = (31 - scale) << 3; \
+ mlib_d64 drnd = vis_to_double_dup(mlib_round_8[31 - scale]); \
+ mlib_s32 i, j, l
+
+/***************************************************************/
+
+#ifdef CONV_INDEX
+
+#define DEF_EXTRA_VARS \
+ int offset = mlib_ImageGetLutOffset(colormap); \
+ LTYPE **lut_table = (LTYPE**)mlib_ImageGetLutData(colormap); \
+ LTYPE *ltbl0 = lut_table[0] - offset; \
+ LTYPE *ltbl1 = lut_table[1] - offset; \
+ LTYPE *ltbl2 = lut_table[2] - offset; \
+ LTYPE *ltbl3 = (NCHAN > 3) ? lut_table[3] - offset : ltbl2
+
+#else
+
+#define DEF_EXTRA_VARS \
+ mlib_s32 nchan = mlib_ImageGetChannels(dst)
+
+#endif
+
+/***************************************************************/
+
+#if NCHAN == 3
+
+#define LOAD_SRC() { \
+ mlib_s32 s0 = sp[0], s1 = sp[1], s2 = sp[2], s3 = sp[3]; \
+ mlib_s32 s4 = sp[4], s5 = sp[5], s6 = sp[6], s7 = sp[7]; \
+ mlib_d64 t0, t1, t2; \
+ \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl2, s7), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl1, s7), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl0, s7), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl2, s6), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl1, s6), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl0, s6), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl2, s5), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl1, s5), t2); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl0, s5), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl2, s4), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl1, s4), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl0, s4), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl2, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl1, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl0, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl2, s2), t1); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl1, s2), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl0, s2), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl2, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl1, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl0, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl2, s0), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl1, s0), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl0, s0), t0); \
+ \
+ buffn[i] = t0; \
+ buffn[i + 1] = t1; \
+ buffn[i + 2] = t2; \
+ \
+ sp += 8; \
+ }
+
+#else
+
+#define LOAD_SRC() { \
+ mlib_s32 s0 = sp[0], s1 = sp[1], s2 = sp[2], s3 = sp[3]; \
+ mlib_s32 s4 = sp[4], s5 = sp[5], s6 = sp[6], s7 = sp[7]; \
+ mlib_d64 t0, t1, t2; \
+ \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl3, s5), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl2, s5), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl1, s5), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl0, s5), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl3, s4), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl2, s4), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl1, s4), t2); \
+ t2 = vis_faligndata(vis_ld_u8_i(ltbl0, s4), t2); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl3, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl2, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl1, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl0, s3), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl3, s2), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl2, s2), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl1, s2), t1); \
+ t1 = vis_faligndata(vis_ld_u8_i(ltbl0, s2), t1); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl3, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl2, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl1, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl0, s1), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl3, s0), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl2, s0), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl1, s0), t0); \
+ t0 = vis_faligndata(vis_ld_u8_i(ltbl0, s0), t0); \
+ \
+ buffn[i] = t0; \
+ buffn[i + 1] = t1; \
+ buffn[i + 2] = t2; \
+ \
+ sp += 6; \
+ }
+
+#endif
+
+/***************************************************************/
+
+static mlib_s32 mlib_round_8[16] = { 0x00400040, 0x00200020, 0x00100010, 0x00080008,
+ 0x00040004, 0x00020002, 0x00010001, 0x00000000,
+ 0x00000000, 0x00000000, 0x00000000, 0x00000000,
+ 0x00000000, 0x00000000, 0x00000000, 0x00000000 };
+
+/***************************************************************/
+
+void mlib_ImageCopy_na(mlib_u8 *sa, mlib_u8 *da, int size);
+
+/***************************************************************/
+
+#define KSIZE 2
+
+mlib_status CONV_FUNC(2x2)
+{
+ mlib_d64 *buffs[2*(KSIZE + 1)];
+ mlib_d64 *buff0, *buff1, *buffn, *buffd, *buffe;
+ mlib_d64 s00, s01, s10, s11, s0, s1;
+ mlib_d64 d0, d1, d00, d01, d10, d11;
+ DEF_VARS;
+ DEF_EXTRA_VARS;
+
+ sl = adr_src;
+ dl = adr_dst;
+
+ ssize = NCHAN*wid;
+ dsize = (ssize + 7)/8;
+ esize = dsize + 4;
+ pbuff = mlib_malloc((KSIZE + 4)*esize*sizeof(mlib_d64));
+ if (pbuff == NULL) return MLIB_FAILURE;
+
+ for (i = 0; i < (KSIZE + 1); i++) buffs[i] = pbuff + i*esize;
+ for (i = 0; i < (KSIZE + 1); i++) buffs[(KSIZE + 1) + i] = buffs[i];
+ buffd = buffs[KSIZE] + esize;
+ buffe = buffd + 2*esize;
+
+ wid -= (KSIZE - 1);
+ hgt -= (KSIZE - 1);
+ xsize = ssize - NCHAN*(KSIZE - 1);
+ emask = (0xFF00 >> (xsize & 7)) & 0xFF;
+
+ vis_write_gsr(gsr_scale + 7);
+
+ for (l = 0; l < KSIZE; l++) {
+ mlib_d64 *buffn = buffs[l];
+ sp = sl + l*sll;
+
+#ifndef CONV_INDEX
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+
+#else
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ LOAD_SRC();
+ }
+#endif /* CONV_INDEX */
+ }
+
+ for (j = 0; j < hgt; j++) {
+ mlib_d64 **buffc = buffs + buff_ind;
+ mlib_f32 *pk = karr, k0, k1;
+ sp = sl + KSIZE*sll;
+
+ buff0 = buffc[0];
+ buff1 = buffc[1];
+ buffn = buffc[KSIZE];
+
+#ifndef CONV_INDEX
+ if ((((mlib_addr)(sl )) & 7) == 0) buff0 = (mlib_d64*)sl;
+ if ((((mlib_addr)(sl + sll)) & 7) == 0) buff1 = (mlib_d64*)(sl + sll);
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+#endif
+
+ k0 = pk[1];
+ k1 = pk[3];
+ vis_write_gsr(gsr_scale + NCHAN);
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = vis_fpadd16(d00, d10);
+ d1 = vis_fpadd16(d01, d11);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ k0 = pk[0];
+ k1 = pk[2];
+#ifndef CONV_INDEX
+ dp = ((mlib_addr)dl & 7) ? buffe : (mlib_d64*)dl;
+
+#pragma pipeloop(0)
+ for (i = 0; i < xsize/8; i++) {
+ s0 = buff0[i];
+ s1 = buff1[i];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d0, drnd);
+ d0 = vis_fpadd16(d0, d00);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d1, drnd);
+ d1 = vis_fpadd16(d1, d01);
+ dp[i] = vis_fpack16_pair(d0, d1);
+ }
+
+ if (emask) {
+ s0 = buff0[i];
+ s1 = buff1[i];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d0, drnd);
+ d0 = vis_fpadd16(d0, d00);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d1, drnd);
+ d1 = vis_fpadd16(d1, d01);
+
+ d0 = vis_fpack16_pair(d0, d1);
+ vis_pst_8(d0, dp + i, emask);
+ }
+
+ if ((mlib_u8*)dp != dl) mlib_ImageCopy_na((void*)buffe, dl, xsize);
+
+#else
+ vis_write_gsr(gsr_scale + 7);
+
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ mlib_d64 d00, d01, d02, d03, d04, d05;
+ mlib_d64 d10, d11, d12, d13, d14, d15;
+ mlib_d64 d0, d1, d2, d3, d4, d5;
+ mlib_d64 s00 = buff0[i];
+ mlib_d64 s01 = buff0[i + 1];
+ mlib_d64 s02 = buff0[i + 2];
+ mlib_d64 s10 = buff1[i];
+ mlib_d64 s11 = buff1[i + 1];
+ mlib_d64 s12 = buff1[i + 2];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d02 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d03 = vis_fmul8x16au(vis_read_lo(s01), k0);
+ d04 = vis_fmul8x16au(vis_read_hi(s02), k0);
+ d05 = vis_fmul8x16au(vis_read_lo(s02), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d12 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d13 = vis_fmul8x16au(vis_read_lo(s11), k1);
+ d14 = vis_fmul8x16au(vis_read_hi(s12), k1);
+ d15 = vis_fmul8x16au(vis_read_lo(s12), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d2 = buffd[2*i + 2];
+ d3 = buffd[2*i + 3];
+ d4 = buffd[2*i + 4];
+ d5 = buffd[2*i + 5];
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d0, drnd);
+ d0 = vis_fpadd16(d0, d00);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d1, drnd);
+ d1 = vis_fpadd16(d1, d01);
+ d02 = vis_fpadd16(d02, d12);
+ d2 = vis_fpadd16(d2, drnd);
+ d2 = vis_fpadd16(d2, d02);
+ d03 = vis_fpadd16(d03, d13);
+ d3 = vis_fpadd16(d3, drnd);
+ d3 = vis_fpadd16(d3, d03);
+ d04 = vis_fpadd16(d04, d14);
+ d4 = vis_fpadd16(d4, drnd);
+ d4 = vis_fpadd16(d4, d04);
+ d05 = vis_fpadd16(d05, d15);
+ d5 = vis_fpadd16(d5, drnd);
+ d5 = vis_fpadd16(d5, d05);
+
+ buffe[i ] = vis_fpack16_pair(d0, d1);
+ buffe[i + 1] = vis_fpack16_pair(d2, d3);
+ buffe[i + 2] = vis_fpack16_pair(d4, d5);
+
+ LOAD_SRC();
+ }
+
+ mlib_ImageColorTrue2IndexLine_U8_U8_3((void*)buffe, dl, wid, colormap);
+#endif /* CONV_INDEX */
+
+ sl += sll;
+ dl += dll;
+
+ buff_ind++;
+ if (buff_ind >= (KSIZE + 1)) buff_ind = 0;
+ }
+
+ mlib_free(pbuff);
+
+ return MLIB_SUCCESS;
+}
+
+/***************************************************************/
+
+#undef KSIZE
+#define KSIZE 3
+
+mlib_status CONV_FUNC(3x3)
+{
+ mlib_d64 *buffs[2*(KSIZE + 1)];
+ mlib_d64 *buff0, *buff1, *buff2, *buffn, *buffd, *buffe;
+ mlib_d64 s00, s01, s10, s11, s20, s21, s0, s1, s2;
+ mlib_d64 dd, d0, d1, d00, d01, d10, d11, d20, d21;
+ mlib_s32 ik, ik_last, off, doff;
+ DEF_VARS;
+ DEF_EXTRA_VARS;
+
+ sl = adr_src;
+#ifdef CONV_INDEX
+ dl = adr_dst + ((KSIZE - 1)/2)*(dll + 1);
+#else
+ dl = adr_dst + ((KSIZE - 1)/2)*(dll + NCHAN);
+#endif
+
+ ssize = NCHAN*wid;
+ dsize = (ssize + 7)/8;
+ esize = dsize + 4;
+ pbuff = mlib_malloc((KSIZE + 4)*esize*sizeof(mlib_d64));
+ if (pbuff == NULL) return MLIB_FAILURE;
+
+ for (i = 0; i < (KSIZE + 1); i++) buffs[i] = pbuff + i*esize;
+ for (i = 0; i < (KSIZE + 1); i++) buffs[(KSIZE + 1) + i] = buffs[i];
+ buffd = buffs[KSIZE] + esize;
+ buffe = buffd + 2*esize;
+
+ wid -= (KSIZE - 1);
+ hgt -= (KSIZE - 1);
+ xsize = ssize - NCHAN*(KSIZE - 1);
+ emask = (0xFF00 >> (xsize & 7)) & 0xFF;
+
+ vis_write_gsr(gsr_scale + 7);
+
+ for (l = 0; l < KSIZE; l++) {
+ mlib_d64 *buffn = buffs[l];
+ sp = sl + l*sll;
+
+#ifndef CONV_INDEX
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+#else
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ LOAD_SRC();
+ }
+#endif /* CONV_INDEX */
+ }
+
+ /* init buffer */
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ for (j = 0; j < hgt; j++) {
+ mlib_d64 **buffc = buffs + buff_ind, *pbuff0, *pbuff1, *pbuff2;
+ mlib_f32 *pk = karr, k0, k1, k2;
+ sp = sl + KSIZE*sll;
+
+ pbuff0 = buffc[0];
+ pbuff1 = buffc[1];
+ pbuff2 = buffc[2];
+ buffn = buffc[KSIZE];
+
+#ifndef CONV_INDEX
+ if ((((mlib_addr)(sl )) & 7) == 0) pbuff0 = (mlib_d64*)sl;
+ if ((((mlib_addr)(sl + sll)) & 7) == 0) pbuff1 = (mlib_d64*)(sl + sll);
+ if ((((mlib_addr)(sl + 2*sll)) & 7) == 0) pbuff2 = (mlib_d64*)(sl + 2*sll);
+
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+#endif
+
+#ifdef CONV_INDEX
+ ik_last = 0;
+#else
+ ik_last = (KSIZE - 1);
+#endif
+
+ for (ik = 0; ik < KSIZE; ik++) {
+ k0 = pk[ik];
+ k1 = pk[ik + KSIZE];
+ k2 = pk[ik + 2*KSIZE];
+
+ off = ik*NCHAN;
+ doff = off/8;
+ off &= 7;
+ buff0 = pbuff0 + doff;
+ buff1 = pbuff1 + doff;
+ buff2 = pbuff2 + doff;
+ vis_write_gsr(gsr_scale + off);
+
+ if (ik == ik_last) continue;
+ /*if (!ik_last) {
+ if ((off & 3) || (ik == (KSIZE - 1))) {
+ ik_last = ik;
+ continue;
+ }
+ }*/
+
+ if (off == 0) {
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ s0 = buff0[i];
+ s1 = buff1[i];
+ s2 = buff2[i];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d00, d0);
+ d0 = vis_fpadd16(d10, d0);
+ d0 = vis_fpadd16(d20, d0);
+ d1 = vis_fpadd16(d01, d1);
+ d1 = vis_fpadd16(d11, d1);
+ d1 = vis_fpadd16(d21, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else if (off == 4) {
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+
+ d00 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d10 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d20 = vis_fmul8x16au(vis_read_lo(s20), k2);
+ d21 = vis_fmul8x16au(vis_read_hi(s21), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d00, d0);
+ d0 = vis_fpadd16(d10, d0);
+ d0 = vis_fpadd16(d20, d0);
+ d1 = vis_fpadd16(d01, d1);
+ d1 = vis_fpadd16(d11, d1);
+ d1 = vis_fpadd16(d21, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else {
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d00, d0);
+ d0 = vis_fpadd16(d10, d0);
+ d0 = vis_fpadd16(d20, d0);
+ d1 = vis_fpadd16(d01, d1);
+ d1 = vis_fpadd16(d11, d1);
+ d1 = vis_fpadd16(d21, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+ }
+ }
+
+ k0 = pk[ik_last];
+ k1 = pk[ik_last + KSIZE];
+ k2 = pk[ik_last + 2*KSIZE];
+
+ off = ik_last*NCHAN;
+ doff = off/8;
+ off &= 7;
+ buff0 = pbuff0 + doff;
+ buff1 = pbuff1 + doff;
+ buff2 = pbuff2 + doff;
+ vis_write_gsr(gsr_scale + off);
+
+#ifndef CONV_INDEX
+ dp = ((mlib_addr)dl & 7) ? buffe : (mlib_d64*)dl;
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < xsize/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+
+ dd = vis_fpack16_pair(d0, d1);
+ dp[i] = dd;
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if (emask) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+
+ dd = vis_fpack16_pair(d0, d1);
+ vis_pst_8(dd, dp + i, emask);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if ((mlib_u8*)dp != dl) mlib_ImageCopy_na((void*)buffe, dl, xsize);
+
+#else
+ vis_write_gsr(gsr_scale + 7);
+
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ mlib_d64 d00, d01, d02, d03, d04, d05;
+ mlib_d64 d10, d11, d12, d13, d14, d15;
+ mlib_d64 d20, d21, d22, d23, d24, d25;
+ mlib_d64 d0, d1, d2, d3, d4, d5;
+ mlib_d64 s00 = buff0[i];
+ mlib_d64 s01 = buff0[i + 1];
+ mlib_d64 s02 = buff0[i + 2];
+ mlib_d64 s10 = buff1[i];
+ mlib_d64 s11 = buff1[i + 1];
+ mlib_d64 s12 = buff1[i + 2];
+ mlib_d64 s20 = buff2[i];
+ mlib_d64 s21 = buff2[i + 1];
+ mlib_d64 s22 = buff2[i + 2];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d02 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d03 = vis_fmul8x16au(vis_read_lo(s01), k0);
+ d04 = vis_fmul8x16au(vis_read_hi(s02), k0);
+ d05 = vis_fmul8x16au(vis_read_lo(s02), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d12 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d13 = vis_fmul8x16au(vis_read_lo(s11), k1);
+ d14 = vis_fmul8x16au(vis_read_hi(s12), k1);
+ d15 = vis_fmul8x16au(vis_read_lo(s12), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s20), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s20), k2);
+ d22 = vis_fmul8x16au(vis_read_hi(s21), k2);
+ d23 = vis_fmul8x16au(vis_read_lo(s21), k2);
+ d24 = vis_fmul8x16au(vis_read_hi(s22), k2);
+ d25 = vis_fmul8x16au(vis_read_lo(s22), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d2 = buffd[2*i + 2];
+ d3 = buffd[2*i + 3];
+ d4 = buffd[2*i + 4];
+ d5 = buffd[2*i + 5];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+ d2 = vis_fpadd16(d2, d02);
+ d2 = vis_fpadd16(d2, d12);
+ d2 = vis_fpadd16(d2, d22);
+ d3 = vis_fpadd16(d3, d03);
+ d3 = vis_fpadd16(d3, d13);
+ d3 = vis_fpadd16(d3, d23);
+ d4 = vis_fpadd16(d4, d04);
+ d4 = vis_fpadd16(d4, d14);
+ d4 = vis_fpadd16(d4, d24);
+ d5 = vis_fpadd16(d5, d05);
+ d5 = vis_fpadd16(d5, d15);
+ d5 = vis_fpadd16(d5, d25);
+
+ buffe[i ] = vis_fpack16_pair(d0, d1);
+ buffe[i + 1] = vis_fpack16_pair(d2, d3);
+ buffe[i + 2] = vis_fpack16_pair(d4, d5);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ buffd[2*i + 2] = drnd;
+ buffd[2*i + 3] = drnd;
+ buffd[2*i + 4] = drnd;
+ buffd[2*i + 5] = drnd;
+
+ LOAD_SRC();
+ }
+
+ mlib_ImageColorTrue2IndexLine_U8_U8_3((void*)buffe, dl, wid, colormap);
+#endif /* CONV_INDEX */
+
+ sl += sll;
+ dl += dll;
+
+ buff_ind++;
+ if (buff_ind >= (KSIZE + 1)) buff_ind = 0;
+ }
+
+ mlib_free(pbuff);
+
+ return MLIB_SUCCESS;
+}
+
+/***************************************************************/
+
+#undef KSIZE
+#define MAX_N 11
+
+#ifdef CONV_INDEX
+
+mlib_status mlib_convMxN_Index3_8_8nw(mlib_image *dst,
+ mlib_image *src,
+ mlib_s32 m,
+ mlib_s32 n,
+ mlib_s32 dm,
+ mlib_s32 dn,
+ mlib_s32 *kern,
+ mlib_s32 scale,
+ void *colormap)
+
+#else
+
+mlib_status mlib_convMxN_8nw_f(mlib_image *dst,
+ mlib_image *src,
+ mlib_s32 m,
+ mlib_s32 n,
+ mlib_s32 dm,
+ mlib_s32 dn,
+ mlib_s32 *kern,
+ mlib_s32 scale)
+
+#endif
+{
+ mlib_d64 *buffs_local[3*(MAX_N + 1)], **buffs = buffs_local, **buff;
+ mlib_d64 *buff0, *buff1, *buff2, *buff3, *buffn, *buffd, *buffe;
+ mlib_d64 s00, s01, s10, s11, s20, s21, s30, s31, s0, s1, s2, s3;
+ mlib_d64 d00, d01, d10, d11, d20, d21, d30, d31;
+ mlib_d64 dd, d0, d1;
+ mlib_s32 ik, jk, ik_last, jk_size, coff, off, doff;
+ DEF_VARS;
+ DEF_EXTRA_VARS;
+
+ if (n > MAX_N) {
+ buffs = mlib_malloc(3*(n + 1)*sizeof(mlib_d64*));
+ if (buffs == NULL) return MLIB_FAILURE;
+ }
+
+ buff = buffs + 2*(n + 1);
+
+ sl = adr_src;
+#ifdef CONV_INDEX
+ dl = adr_dst + dn*dll + dm;
+#else
+ dl = adr_dst + dn*dll + dm*NCHAN;
+#endif
+
+ ssize = NCHAN*wid;
+ dsize = (ssize + 7)/8;
+ esize = dsize + 4;
+ pbuff = mlib_malloc((n + 4)*esize*sizeof(mlib_d64));
+ if (pbuff == NULL) {
+ if (buffs != buffs_local) mlib_free(buffs);
+ return MLIB_FAILURE;
+ }
+
+ for (i = 0; i < (n + 1); i++) buffs[i] = pbuff + i*esize;
+ for (i = 0; i < (n + 1); i++) buffs[(n + 1) + i] = buffs[i];
+ buffd = buffs[n] + esize;
+ buffe = buffd + 2*esize;
+
+ wid -= (m - 1);
+ hgt -= (n - 1);
+ xsize = ssize - NCHAN*(m - 1);
+ emask = (0xFF00 >> (xsize & 7)) & 0xFF;
+
+ vis_write_gsr(gsr_scale + 7);
+
+ for (l = 0; l < n; l++) {
+ mlib_d64 *buffn = buffs[l];
+ sp = sl + l*sll;
+
+#ifndef CONV_INDEX
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+#else
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ LOAD_SRC();
+ }
+#endif /* CONV_INDEX */
+ }
+
+ /* init buffer */
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ for (j = 0; j < hgt; j++) {
+ mlib_d64 **buffc = buffs + buff_ind;
+ mlib_f32 *pk = karr, k0, k1, k2, k3;
+ sp = sl + n*sll;
+
+ for (l = 0; l < n; l++) {
+ buff[l] = buffc[l];
+ }
+ buffn = buffc[n];
+
+#ifndef CONV_INDEX
+ for (l = 0; l < n; l++) {
+ if ((((mlib_addr)(sl + l*sll)) & 7) == 0) buff[l] = (mlib_d64*)(sl + l*sll);
+ }
+ if ((mlib_addr)sp & 7) mlib_ImageCopy_na((void*)sp, (void*)buffn, ssize);
+#endif
+
+#ifdef CONV_INDEX
+ ik_last = 0;
+#else
+ ik_last = (m - 1);
+#endif
+
+ for (jk = 0; jk < n; jk += jk_size) {
+ jk_size = n - jk;
+#ifdef CONV_INDEX
+ if (jk_size >= 5) jk_size = 3;
+ if (jk_size == 4) jk_size = 2;
+#else
+ if (jk_size >= 6) jk_size = 4;
+ if (jk_size == 5) jk_size = 3;
+#endif
+ coff = 0;
+
+ if (jk_size == 2) {
+
+ for (ik = 0; ik < m; ik++, coff += NCHAN) {
+ if (!jk && ik == ik_last) continue;
+
+ k0 = pk[ik];
+ k1 = pk[ik + m];
+
+ doff = coff/8;
+ buff0 = buff[jk ] + doff;
+ buff1 = buff[jk + 1] + doff;
+
+ off = coff & 7;
+ vis_write_gsr(gsr_scale + off);
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d00, d0);
+ d0 = vis_fpadd16(d10, d0);
+ d1 = vis_fpadd16(d01, d1);
+ d1 = vis_fpadd16(d11, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ }
+
+ pk += 2*m;
+
+ } else if (jk_size == 3) {
+
+ for (ik = 0; ik < m; ik++, coff += NCHAN) {
+ if (!jk && ik == ik_last) continue;
+
+ k0 = pk[ik];
+ k1 = pk[ik + m];
+ k2 = pk[ik + 2*m];
+
+ doff = coff/8;
+ buff0 = buff[jk ] + doff;
+ buff1 = buff[jk + 1] + doff;
+ buff2 = buff[jk + 2] + doff;
+
+ off = coff & 7;
+ vis_write_gsr(gsr_scale + off);
+
+ if (off == 0) {
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s0 = buff0[i];
+ s1 = buff1[i];
+ s2 = buff2[i];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d20, d0);
+ d0 = vis_fpadd16(d00, d0);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d21, d1);
+ d1 = vis_fpadd16(d01, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else if (off == 4) {
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+
+ d00 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d10 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d20 = vis_fmul8x16au(vis_read_lo(s20), k2);
+ d21 = vis_fmul8x16au(vis_read_hi(s21), k2);
+
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d20, d0);
+ d0 = vis_fpadd16(d00, d0);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d21, d1);
+ d1 = vis_fpadd16(d01, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else {
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d00 = vis_fpadd16(d00, d10);
+ d0 = vis_fpadd16(d20, d0);
+ d0 = vis_fpadd16(d00, d0);
+ d01 = vis_fpadd16(d01, d11);
+ d1 = vis_fpadd16(d21, d1);
+ d1 = vis_fpadd16(d01, d1);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+ }
+ }
+
+ pk += 3*m;
+
+ } else { /* jk_size == 4 */
+
+ for (ik = 0; ik < m; ik++, coff += NCHAN) {
+ if (!jk && ik == ik_last) continue;
+
+ k0 = pk[ik];
+ k1 = pk[ik + m];
+ k2 = pk[ik + 2*m];
+ k3 = pk[ik + 3*m];
+
+ doff = coff/8;
+ buff0 = buff[jk ] + doff;
+ buff1 = buff[jk + 1] + doff;
+ buff2 = buff[jk + 2] + doff;
+ buff3 = buff[jk + 3] + doff;
+
+ off = coff & 7;
+ vis_write_gsr(gsr_scale + off);
+
+ if (off == 0) {
+
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s0 = buff0[i];
+ s1 = buff1[i];
+ s2 = buff2[i];
+ s3 = buff3[i];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+ d30 = vis_fmul8x16au(vis_read_hi(s3), k3);
+ d31 = vis_fmul8x16au(vis_read_lo(s3), k3);
+
+ d00 = vis_fpadd16(d00, d10);
+ d20 = vis_fpadd16(d20, d30);
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d20);
+ d01 = vis_fpadd16(d01, d11);
+ d21 = vis_fpadd16(d21, d31);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d21);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else if (off == 4) {
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+ s31 = buff3[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s30 = s31;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s31 = buff3[i + 1];
+
+ d00 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d10 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d20 = vis_fmul8x16au(vis_read_lo(s20), k2);
+ d21 = vis_fmul8x16au(vis_read_hi(s21), k2);
+ d30 = vis_fmul8x16au(vis_read_lo(s30), k3);
+ d31 = vis_fmul8x16au(vis_read_hi(s31), k3);
+
+ d00 = vis_fpadd16(d00, d10);
+ d20 = vis_fpadd16(d20, d30);
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d20);
+ d01 = vis_fpadd16(d01, d11);
+ d21 = vis_fpadd16(d21, d31);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d21);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+
+ } else {
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+ s31 = buff3[0];
+#pragma pipeloop(0)
+ for (i = 0; i < (xsize + 7)/8; i++) {
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s30 = s31;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s31 = buff3[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+ s3 = vis_faligndata(s30, s31);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+ d30 = vis_fmul8x16au(vis_read_hi(s3), k3);
+ d31 = vis_fmul8x16au(vis_read_lo(s3), k3);
+
+ d00 = vis_fpadd16(d00, d10);
+ d20 = vis_fpadd16(d20, d30);
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d20);
+ d01 = vis_fpadd16(d01, d11);
+ d21 = vis_fpadd16(d21, d31);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d21);
+ buffd[2*i] = d0;
+ buffd[2*i + 1] = d1;
+ }
+ }
+ }
+
+ pk += 4*m;
+ }
+ }
+
+ /*****************************************
+ *****************************************
+ ** Final iteration **
+ *****************************************
+ *****************************************/
+
+ jk_size = n;
+#ifdef CONV_INDEX
+ if (jk_size >= 5) jk_size = 3;
+ if (jk_size == 4) jk_size = 2;
+#else
+ if (jk_size >= 6) jk_size = 4;
+ if (jk_size == 5) jk_size = 3;
+#endif
+
+ k0 = karr[ik_last];
+ k1 = karr[ik_last + m];
+ k2 = karr[ik_last + 2*m];
+ k3 = karr[ik_last + 3*m];
+
+ off = ik_last*NCHAN;
+ doff = off/8;
+ off &= 7;
+ buff0 = buff[0] + doff;
+ buff1 = buff[1] + doff;
+ buff2 = buff[2] + doff;
+ buff3 = buff[3] + doff;
+ vis_write_gsr(gsr_scale + off);
+
+#ifndef CONV_INDEX
+ if (jk_size == 2) {
+ dp = ((mlib_addr)dl & 7) ? buffe : (mlib_d64*)dl;
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+#pragma pipeloop(0)
+ for (i = 0; i < xsize/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+
+ dd = vis_fpack16_pair(d0, d1);
+ dp[i] = dd;
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if (emask) {
+ s00 = s01;
+ s10 = s11;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+
+ dd = vis_fpack16_pair(d0, d1);
+ vis_pst_8(dd, dp + i, emask);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if ((mlib_u8*)dp != dl) mlib_ImageCopy_na((void*)buffe, dl, xsize);
+
+ } else if (jk_size == 3) {
+
+ dp = ((mlib_addr)dl & 7) ? buffe : (mlib_d64*)dl;
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+#pragma pipeloop(0)
+ for (i = 0; i < xsize/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+
+ dd = vis_fpack16_pair(d0, d1);
+ dp[i] = dd;
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if (emask) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+
+ dd = vis_fpack16_pair(d0, d1);
+ vis_pst_8(dd, dp + i, emask);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if ((mlib_u8*)dp != dl) mlib_ImageCopy_na((void*)buffe, dl, xsize);
+
+ } else /* if (jk_size == 4) */ {
+
+ dp = ((mlib_addr)dl & 7) ? buffe : (mlib_d64*)dl;
+
+ s01 = buff0[0];
+ s11 = buff1[0];
+ s21 = buff2[0];
+ s31 = buff3[0];
+#pragma pipeloop(0)
+ for (i = 0; i < xsize/8; i++) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s30 = s31;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s31 = buff3[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+ s3 = vis_faligndata(s30, s31);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+ d30 = vis_fmul8x16au(vis_read_hi(s3), k3);
+ d31 = vis_fmul8x16au(vis_read_lo(s3), k3);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d0 = vis_fpadd16(d0, d30);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+ d1 = vis_fpadd16(d1, d31);
+
+ dd = vis_fpack16_pair(d0, d1);
+ dp[i] = dd;
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if (emask) {
+ s00 = s01;
+ s10 = s11;
+ s20 = s21;
+ s30 = s31;
+ s01 = buff0[i + 1];
+ s11 = buff1[i + 1];
+ s21 = buff2[i + 1];
+ s31 = buff3[i + 1];
+ s0 = vis_faligndata(s00, s01);
+ s1 = vis_faligndata(s10, s11);
+ s2 = vis_faligndata(s20, s21);
+ s3 = vis_faligndata(s30, s31);
+
+ d00 = vis_fmul8x16au(vis_read_hi(s0), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s0), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s1), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s1), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s2), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s2), k2);
+ d30 = vis_fmul8x16au(vis_read_hi(s3), k3);
+ d31 = vis_fmul8x16au(vis_read_lo(s3), k3);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d0 = vis_fpadd16(d0, d30);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+ d1 = vis_fpadd16(d1, d31);
+
+ dd = vis_fpack16_pair(d0, d1);
+ vis_pst_8(dd, dp + i, emask);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ }
+
+ if ((mlib_u8*)dp != dl) mlib_ImageCopy_na((void*)buffe, dl, xsize);
+ }
+
+#else /* CONV_INDEX */
+
+ if (jk_size == 2) {
+ vis_write_gsr(gsr_scale + 7);
+
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ mlib_d64 d00, d01, d02, d03, d04, d05;
+ mlib_d64 d10, d11, d12, d13, d14, d15;
+ mlib_d64 d0, d1, d2, d3, d4, d5;
+ mlib_d64 s00 = buff0[i];
+ mlib_d64 s01 = buff0[i + 1];
+ mlib_d64 s02 = buff0[i + 2];
+ mlib_d64 s10 = buff1[i];
+ mlib_d64 s11 = buff1[i + 1];
+ mlib_d64 s12 = buff1[i + 2];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d02 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d03 = vis_fmul8x16au(vis_read_lo(s01), k0);
+ d04 = vis_fmul8x16au(vis_read_hi(s02), k0);
+ d05 = vis_fmul8x16au(vis_read_lo(s02), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d12 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d13 = vis_fmul8x16au(vis_read_lo(s11), k1);
+ d14 = vis_fmul8x16au(vis_read_hi(s12), k1);
+ d15 = vis_fmul8x16au(vis_read_lo(s12), k1);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d2 = buffd[2*i + 2];
+ d3 = buffd[2*i + 3];
+ d4 = buffd[2*i + 4];
+ d5 = buffd[2*i + 5];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d2 = vis_fpadd16(d2, d02);
+ d2 = vis_fpadd16(d2, d12);
+ d3 = vis_fpadd16(d3, d03);
+ d3 = vis_fpadd16(d3, d13);
+ d4 = vis_fpadd16(d4, d04);
+ d4 = vis_fpadd16(d4, d14);
+ d5 = vis_fpadd16(d5, d05);
+ d5 = vis_fpadd16(d5, d15);
+
+ buffe[i ] = vis_fpack16_pair(d0, d1);
+ buffe[i + 1] = vis_fpack16_pair(d2, d3);
+ buffe[i + 2] = vis_fpack16_pair(d4, d5);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ buffd[2*i + 2] = drnd;
+ buffd[2*i + 3] = drnd;
+ buffd[2*i + 4] = drnd;
+ buffd[2*i + 5] = drnd;
+
+ LOAD_SRC();
+ }
+
+ } else /* if (jk_size == 3) */ {
+ vis_write_gsr(gsr_scale + 7);
+
+#pragma pipeloop(0)
+ for (i = 0; i < dsize; i += 3) {
+ mlib_d64 d00, d01, d02, d03, d04, d05;
+ mlib_d64 d10, d11, d12, d13, d14, d15;
+ mlib_d64 d20, d21, d22, d23, d24, d25;
+ mlib_d64 d0, d1, d2, d3, d4, d5;
+ mlib_d64 s00 = buff0[i];
+ mlib_d64 s01 = buff0[i + 1];
+ mlib_d64 s02 = buff0[i + 2];
+ mlib_d64 s10 = buff1[i];
+ mlib_d64 s11 = buff1[i + 1];
+ mlib_d64 s12 = buff1[i + 2];
+ mlib_d64 s20 = buff2[i];
+ mlib_d64 s21 = buff2[i + 1];
+ mlib_d64 s22 = buff2[i + 2];
+
+ d00 = vis_fmul8x16au(vis_read_hi(s00), k0);
+ d01 = vis_fmul8x16au(vis_read_lo(s00), k0);
+ d02 = vis_fmul8x16au(vis_read_hi(s01), k0);
+ d03 = vis_fmul8x16au(vis_read_lo(s01), k0);
+ d04 = vis_fmul8x16au(vis_read_hi(s02), k0);
+ d05 = vis_fmul8x16au(vis_read_lo(s02), k0);
+ d10 = vis_fmul8x16au(vis_read_hi(s10), k1);
+ d11 = vis_fmul8x16au(vis_read_lo(s10), k1);
+ d12 = vis_fmul8x16au(vis_read_hi(s11), k1);
+ d13 = vis_fmul8x16au(vis_read_lo(s11), k1);
+ d14 = vis_fmul8x16au(vis_read_hi(s12), k1);
+ d15 = vis_fmul8x16au(vis_read_lo(s12), k1);
+ d20 = vis_fmul8x16au(vis_read_hi(s20), k2);
+ d21 = vis_fmul8x16au(vis_read_lo(s20), k2);
+ d22 = vis_fmul8x16au(vis_read_hi(s21), k2);
+ d23 = vis_fmul8x16au(vis_read_lo(s21), k2);
+ d24 = vis_fmul8x16au(vis_read_hi(s22), k2);
+ d25 = vis_fmul8x16au(vis_read_lo(s22), k2);
+
+ d0 = buffd[2*i];
+ d1 = buffd[2*i + 1];
+ d2 = buffd[2*i + 2];
+ d3 = buffd[2*i + 3];
+ d4 = buffd[2*i + 4];
+ d5 = buffd[2*i + 5];
+ d0 = vis_fpadd16(d0, d00);
+ d0 = vis_fpadd16(d0, d10);
+ d0 = vis_fpadd16(d0, d20);
+ d1 = vis_fpadd16(d1, d01);
+ d1 = vis_fpadd16(d1, d11);
+ d1 = vis_fpadd16(d1, d21);
+ d2 = vis_fpadd16(d2, d02);
+ d2 = vis_fpadd16(d2, d12);
+ d2 = vis_fpadd16(d2, d22);
+ d3 = vis_fpadd16(d3, d03);
+ d3 = vis_fpadd16(d3, d13);
+ d3 = vis_fpadd16(d3, d23);
+ d4 = vis_fpadd16(d4, d04);
+ d4 = vis_fpadd16(d4, d14);
+ d4 = vis_fpadd16(d4, d24);
+ d5 = vis_fpadd16(d5, d05);
+ d5 = vis_fpadd16(d5, d15);
+ d5 = vis_fpadd16(d5, d25);
+
+ buffe[i ] = vis_fpack16_pair(d0, d1);
+ buffe[i + 1] = vis_fpack16_pair(d2, d3);
+ buffe[i + 2] = vis_fpack16_pair(d4, d5);
+
+ buffd[2*i ] = drnd;
+ buffd[2*i + 1] = drnd;
+ buffd[2*i + 2] = drnd;
+ buffd[2*i + 3] = drnd;
+ buffd[2*i + 4] = drnd;
+ buffd[2*i + 5] = drnd;
+
+ LOAD_SRC();
+ }
+ }
+#endif /* CONV_INDEX */
+
+#ifdef CONV_INDEX
+ mlib_ImageColorTrue2IndexLine_U8_U8_3((void*)buffe, dl, wid, colormap);
+#endif /* CONV_INDEX */
+
+ sl += sll;
+ dl += dll;
+
+ buff_ind++;
+ if (buff_ind >= (n + 1)) buff_ind = 0;
+ }
+
+ mlib_free(pbuff);
+ if (buffs != buffs_local) mlib_free(buffs);
+
+ return MLIB_SUCCESS;
+}
+
+/***************************************************************/