0N/A/*
2362N/A * Copyright (c) 2003, Oracle and/or its affiliates. All rights reserved.
0N/A * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
0N/A *
0N/A * This code is free software; you can redistribute it and/or modify it
0N/A * under the terms of the GNU General Public License version 2 only, as
2362N/A * published by the Free Software Foundation. Oracle designates this
0N/A * particular file as subject to the "Classpath" exception as provided
2362N/A * by Oracle in the LICENSE file that accompanied this code.
0N/A *
0N/A * This code is distributed in the hope that it will be useful, but WITHOUT
0N/A * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
0N/A * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
0N/A * version 2 for more details (a copy is included in the LICENSE file that
0N/A * accompanied this code).
0N/A *
0N/A * You should have received a copy of the GNU General Public License version
0N/A * 2 along with this work; if not, write to the Free Software Foundation,
0N/A * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
0N/A *
2362N/A * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
2362N/A * or visit www.oracle.com if you need additional information or have any
2362N/A * questions.
0N/A */
0N/A
0N/A#if !defined(JAVA2D_NO_MLIB) || defined(MLIB_ADD_SUFF)
0N/A
0N/A#include "vis_AlphaMacros.h"
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define ARGB_to_GBGR(x) \
0N/A (x << 16) | (x & 0xff00) | ((x >> 16) & 0xff)
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define ARGB_to_BGR(x) \
0N/A ((x << 16) & 0xff0000) | (x & 0xff00) | ((x >> 16) & 0xff)
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define READ_Bgr(i) \
0N/A (src[3*i] << 16) | (src[3*i + 1] << 8) | src[3*i + 2]
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define ARGB_to_GBGR_FL2(dst, src0, src1) { \
0N/A mlib_d64 t0, t1, t2; \
0N/A t0 = vis_fpmerge(src0, src1); \
0N/A t1 = vis_fpmerge(vis_read_lo(t0), vis_read_hi(t0)); \
0N/A t2 = vis_fpmerge(vis_read_lo(t0), vis_read_lo(t0)); \
0N/A dst = vis_fpmerge(vis_read_hi(t2), vis_read_lo(t1)); \
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define ARGB_to_BGR_FL2(dst, src0, src1) { \
0N/A mlib_d64 t0, t1, t2; \
0N/A t0 = vis_fpmerge(src0, src1); \
0N/A t1 = vis_fpmerge(vis_read_lo(t0), vis_read_hi(t0)); \
0N/A t2 = vis_fpmerge(vis_fzeros(), vis_read_lo(t0)); \
0N/A dst = vis_fpmerge(vis_read_hi(t2), vis_read_lo(t1)); \
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntBgrToIntArgbConvert)(BLIT_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd, amask;
0N/A mlib_s32 i, i0, j, x;
0N/A
0N/A if (dstScan == 4*width && srcScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A amask = vis_to_double_dup(0xFF000000);
0N/A vis_alignaddr(NULL, 7);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u32 *src = srcBase;
0N/A mlib_u32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = src[i];
0N/A dst[i] = 0xff000000 | ARGB_to_GBGR(x);
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A ARGB2ABGR_FL2(dd, ((mlib_f32*)src)[i], ((mlib_f32*)src)[i + 1]);
0N/A *(mlib_d64*)(dst + i) = vis_for(dd, amask);
0N/A }
0N/A
0N/A if (i < width) {
0N/A x = src[i];
0N/A dst[i] = 0xff000000 | ARGB_to_GBGR(x);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntBgrToIntArgbScaleConvert)(SCALE_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd, amask;
0N/A mlib_s32 j, x;
0N/A
0N/A amask = vis_to_double_dup(0xFF000000);
0N/A vis_alignaddr(NULL, 7);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u32 *src = srcBase;
0N/A mlib_u32 *dst = dstBase;
0N/A mlib_u32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = src[tmpsxloc >> shift];
0N/A *dst++ = 0xff000000 | ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A ARGB2ABGR_FL2(dd, ((mlib_f32*)src)[tmpsxloc >> shift],
0N/A ((mlib_f32*)src)[(tmpsxloc + sxinc) >> shift]);
0N/A *(mlib_d64*)dst = vis_for(dd, amask);
0N/A tmpsxloc += 2*sxinc;
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A x = src[tmpsxloc >> shift];
0N/A *dst++ = 0xff000000 | ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntArgbToIntBgrConvert)(BLIT_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 i, i0, j, x;
0N/A
0N/A if (dstScan == 4*width && srcScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u32 *src = srcBase;
0N/A mlib_u32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = src[i];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A ARGB_to_GBGR_FL2(dd, ((mlib_f32*)src)[i], ((mlib_f32*)src)[i + 1]);
0N/A *(mlib_d64*)(dst + i) = dd;
0N/A }
0N/A
0N/A if (i < width) {
0N/A x = src[i];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntArgbToIntBgrScaleConvert)(SCALE_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 j, x;
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u32 *src = srcBase;
0N/A mlib_u32 *dst = dstBase;
0N/A mlib_u32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = src[tmpsxloc >> shift];
0N/A *dst++ = ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A ARGB_to_GBGR_FL2(dd, ((mlib_f32*)src)[tmpsxloc >> shift],
0N/A ((mlib_f32*)src)[(tmpsxloc + sxinc) >> shift]);
0N/A *(mlib_d64*)dst = dd;
0N/A tmpsxloc += 2*sxinc;
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A x = src[tmpsxloc >> shift];
0N/A *dst++ = ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/A#define INSERT_U8_34R { \
0N/A mlib_d64 sda, sdb, sdc, sdd; \
0N/A mlib_d64 sde, sdf, sdg, sdh; \
0N/A mlib_d64 sdi, sdj, sdk, sdl; \
0N/A mlib_d64 sdm; \
0N/A \
0N/A sda = vis_fpmerge(vis_read_hi(sd0), vis_read_lo(sd1)); \
0N/A sdb = vis_fpmerge(vis_read_lo(sd0), vis_read_hi(sd2)); \
0N/A sdc = vis_fpmerge(vis_read_hi(sd1), vis_read_lo(sd2)); \
0N/A sdd = vis_fpmerge(vis_read_hi(sda), vis_read_lo(sdb)); \
0N/A sde = vis_fpmerge(vis_read_lo(sda), vis_read_hi(sdc)); \
0N/A sdf = vis_fpmerge(vis_read_hi(sdb), vis_read_lo(sdc)); \
0N/A sdg = vis_fpmerge(vis_read_hi(sdd), vis_read_lo(sde)); \
0N/A sdh = vis_fpmerge(vis_read_lo(sdd), vis_read_hi(sdf)); \
0N/A sdi = vis_fpmerge(vis_read_hi(sde), vis_read_lo(sdf)); \
0N/A sdj = vis_fpmerge(vis_read_hi(sdg), vis_read_hi(sdi)); \
0N/A sdk = vis_fpmerge(vis_read_lo(sdg), vis_read_lo(sdi)); \
0N/A sdl = vis_fpmerge(vis_read_hi(sFF), vis_read_hi(sdh)); \
0N/A sdm = vis_fpmerge(vis_read_lo(sFF), vis_read_lo(sdh)); \
0N/A dd0 = vis_fpmerge(vis_read_hi(sdl), vis_read_hi(sdj)); \
0N/A dd1 = vis_fpmerge(vis_read_lo(sdl), vis_read_lo(sdj)); \
0N/A dd2 = vis_fpmerge(vis_read_hi(sdm), vis_read_hi(sdk)); \
0N/A dd3 = vis_fpmerge(vis_read_lo(sdm), vis_read_lo(sdk)); \
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ThreeByteBgrToIntBgrConvert)(BLIT_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 *sp;
0N/A mlib_d64 sFF;
0N/A mlib_d64 s0, s1, s2, s3, sd0, sd1, sd2, dd0, dd1, dd2, dd3;
0N/A mlib_s32 i, i0, j;
0N/A
0N/A if (width < 16) {
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_u32 *dst = dstBase;
0N/A
0N/A for (i = 0; i < width; i++) {
0N/A dst[i] = READ_Bgr(i);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A return;
0N/A }
0N/A
0N/A if (srcScan == 3*width && dstScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A sFF = vis_fzero();
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_f32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A ((mlib_s32*)dst)[i] = READ_Bgr(i);
0N/A i0 = 1;
0N/A }
0N/A
0N/A sp = vis_alignaddr(src, 3*i0);
0N/A s3 = *sp++;
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 8; i += 8) {
0N/A s0 = s3;
0N/A s1 = *sp++;
0N/A s2 = *sp++;
0N/A s3 = *sp++;
0N/A sd0 = vis_faligndata(s0, s1);
0N/A sd1 = vis_faligndata(s1, s2);
0N/A sd2 = vis_faligndata(s2, s3);
0N/A
0N/A INSERT_U8_34R
0N/A
0N/A *(mlib_d64*)(dst + i ) = dd0;
0N/A *(mlib_d64*)(dst + i + 2) = dd1;
0N/A *(mlib_d64*)(dst + i + 4) = dd2;
0N/A *(mlib_d64*)(dst + i + 6) = dd3;
0N/A }
0N/A
0N/A for (; i < width; i++) {
0N/A ((mlib_s32*)dst)[i] = READ_Bgr(i);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ThreeByteBgrToIntBgrScaleConvert)(SCALE_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd, dzero;
0N/A mlib_s32 i, i0, i1, j;
0N/A
0N/A if (width < 16) {
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A mlib_s32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A i = tmpsxloc >> shift;
0N/A tmpsxloc += sxinc;
0N/A *(mlib_s32*)dst = READ_Bgr(i);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A return;
0N/A }
0N/A
0N/A dzero = vis_fzero();
0N/A
0N/A vis_alignaddr(NULL, 7);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_f32 *dst = dstBase;
0N/A mlib_f32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A i = tmpsxloc >> shift;
0N/A tmpsxloc += sxinc;
0N/A *(mlib_s32*)dst = READ_Bgr(i);
0N/A dst++;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A i0 = tmpsxloc >> shift;
0N/A i1 = (tmpsxloc + sxinc) >> shift;
0N/A tmpsxloc += 2*sxinc;
0N/A
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i1 + 2), dd);
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i1 + 1), dd);
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i1 ), dd);
0N/A dd = vis_faligndata(dzero, dd);
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i0 + 2), dd);
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i0 + 1), dd);
0N/A dd = vis_faligndata(vis_ld_u8(src + 3*i0 ), dd);
0N/A dd = vis_faligndata(dzero, dd);
0N/A
0N/A *(mlib_d64*)dst = dd;
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A i = tmpsxloc >> shift;
0N/A tmpsxloc += sxinc;
0N/A *(mlib_s32*)dst = READ_Bgr(i);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntArgbBmToIntBgrXparOver)(BLIT_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 i, i0, j, mask, x;
0N/A
0N/A if (dstScan == 4*width && srcScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_s32 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A if (*(mlib_u8*)(src + i)) {
0N/A x = src[i];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A }
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A ARGB_to_GBGR_FL2(dd, ((mlib_f32*)src)[i], ((mlib_f32*)src)[i + 1]);
0N/A mask = (((-*(mlib_u8*)(src + i)) >> 31) & 2) |
0N/A (((-*(mlib_u8*)(src + i + 1)) >> 31) & 1);
0N/A vis_pst_32(dd, dst + i, mask);
0N/A }
0N/A
0N/A if (i < width) {
0N/A if (*(mlib_u8*)(src + i)) {
0N/A x = src[i];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A }
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntArgbBmToIntBgrScaleXparOver)(SCALE_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 j, mask;
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_s32 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A mlib_s32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A mlib_s32 *pp = src + (tmpsxloc >> shift);
0N/A if (*(mlib_u8*)pp) {
0N/A *dst = ARGB_to_GBGR(*pp);
0N/A }
0N/A dst++;
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A mlib_s32 *pp0 = src + (tmpsxloc >> shift);
0N/A mlib_s32 *pp1 = src + ((tmpsxloc + sxinc) >> shift);
0N/A ARGB_to_GBGR_FL2(dd, *(mlib_f32*)pp0, *(mlib_f32*)pp1);
0N/A mask = (((-*(mlib_u8*)pp0) >> 31) & 2) |
0N/A ((mlib_u32)(-*(mlib_u8*)pp1) >> 31);
0N/A vis_pst_32(dd, dst, mask);
0N/A tmpsxloc += 2*sxinc;
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A mlib_s32 *pp = src + (tmpsxloc >> shift);
0N/A if (*(mlib_u8*)pp) {
0N/A *dst = ARGB_to_GBGR(*pp);
0N/A }
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntArgbBmToIntBgrXparBgCopy)(BCOPY_PARAMS)
0N/A{
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd, d_bgpixel;
0N/A mlib_s32 i, i0, j, mask;
0N/A
0N/A if (dstScan == 4*width && srcScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A vis_alignaddr(NULL, 1);
0N/A d_bgpixel = vis_to_double_dup(bgpixel);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_s32 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A if (*(mlib_u8*)(src + i)) {
0N/A dst[i] = ARGB_to_GBGR(src[i]);
0N/A } else {
0N/A dst[i] = bgpixel;
0N/A }
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A ARGB_to_GBGR_FL2(dd, ((mlib_f32*)src)[i], ((mlib_f32*)src)[i + 1]);
0N/A mask = (((-*(mlib_u8*)(src + i)) >> 31) & 2) |
0N/A (((-*(mlib_u8*)(src + i + 1)) >> 31) & 1);
0N/A *(mlib_d64*)(dst + i) = d_bgpixel;
0N/A vis_pst_32(dd, dst + i, mask);
0N/A }
0N/A
0N/A if (i < width) {
0N/A if (*(mlib_u8*)(src + i)) {
0N/A dst[i] = ARGB_to_GBGR(src[i]);
0N/A } else {
0N/A dst[i] = bgpixel;
0N/A }
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ByteIndexedToIntBgrConvert)(BLIT_PARAMS)
0N/A{
0N/A jint *pixLut = pSrcInfo->lutBase;
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 i, i0, j, x;
0N/A
0N/A if (srcScan == width && dstScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = pixLut[src[i]];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A ARGB_to_GBGR_FL2(dd, ((mlib_f32*)pixLut)[src[i]],
0N/A ((mlib_f32*)pixLut)[src[i + 1]]);
0N/A *(mlib_d64*)(dst + i) = dd;
0N/A }
0N/A
0N/A for (; i < width; i++) {
0N/A x = pixLut[src[i]];
0N/A dst[i] = ARGB_to_GBGR(x);
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ByteIndexedToIntBgrScaleConvert)(SCALE_PARAMS)
0N/A{
0N/A jint *pixLut = pSrcInfo->lutBase;
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 j, x;
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A mlib_s32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = pixLut[src[tmpsxloc >> shift]];
0N/A *dst++ = ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A mlib_f32 f0 = ((mlib_f32*)pixLut)[src[tmpsxloc >> shift]];
0N/A mlib_f32 f1 = ((mlib_f32*)pixLut)[src[(tmpsxloc + sxinc) >> shift]];
0N/A ARGB_to_GBGR_FL2(dd, f0, f1);
0N/A *(mlib_d64*)dst = dd;
0N/A tmpsxloc += 2*sxinc;
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A x = pixLut[src[tmpsxloc >> shift]];
0N/A *dst++ = ARGB_to_GBGR(x);
0N/A tmpsxloc += sxinc;
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ByteIndexedBmToIntBgrXparOver)(BLIT_PARAMS)
0N/A{
0N/A jint *pixLut = pSrcInfo->lutBase;
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 i, i0, j, x, mask;
0N/A
0N/A if (srcScan == width && dstScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A
0N/A i = i0 = 0;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = pixLut[src[i]];
0N/A if (x < 0) {
0N/A dst[i] = ARGB_to_BGR(x);
0N/A }
0N/A i0 = 1;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (i = i0; i <= (mlib_s32)width - 2; i += 2) {
0N/A mlib_f32 *pp0 = (mlib_f32*)pixLut + src[i];
0N/A mlib_f32 *pp1 = (mlib_f32*)pixLut + src[i + 1];
0N/A ARGB_to_BGR_FL2(dd, *pp0, *pp1);
0N/A mask = (((*(mlib_u8*)pp0) >> 6) & 2) | ((*(mlib_u8*)pp1) >> 7);
0N/A vis_pst_32(dd, dst + i, mask);
0N/A }
0N/A
0N/A for (; i < width; i++) {
0N/A x = pixLut[src[i]];
0N/A if (x < 0) {
0N/A dst[i] = ARGB_to_BGR(x);
0N/A }
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ByteIndexedBmToIntBgrScaleXparOver)(SCALE_PARAMS)
0N/A{
0N/A jint *pixLut = pSrcInfo->lutBase;
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd;
0N/A mlib_s32 j, x, mask;
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A mlib_s32 *dst_end = dst + width;
0N/A mlib_s32 tmpsxloc = sxloc;
0N/A
0N/A PTR_ADD(src, (syloc >> shift) * srcScan);
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = pixLut[src[tmpsxloc >> shift]];
0N/A tmpsxloc += sxinc;
0N/A if (x < 0) {
0N/A *dst = ARGB_to_BGR(x);
0N/A }
0N/A dst++;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= dst_end - 2; dst += 2) {
0N/A mlib_f32 *p0 = (mlib_f32*)pixLut + src[tmpsxloc >> shift];
0N/A mlib_f32 *p1 = (mlib_f32*)pixLut + src[(tmpsxloc + sxinc) >> shift];
0N/A ARGB_to_BGR_FL2(dd, *p0, *p1);
0N/A mask = (((*(mlib_u8*)p0) >> 6) & 2) | ((*(mlib_u8*)p1) >> 7);
0N/A tmpsxloc += 2*sxinc;
0N/A vis_pst_32(dd, dst, mask);
0N/A }
0N/A
0N/A for (; dst < dst_end; dst++) {
0N/A x = pixLut[src[tmpsxloc >> shift]];
0N/A tmpsxloc += sxinc;
0N/A if (x < 0) {
0N/A *dst = ARGB_to_BGR(x);
0N/A }
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A syloc += syinc;
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(ByteIndexedBmToIntBgrXparBgCopy)(BCOPY_PARAMS)
0N/A{
0N/A jint *pixLut = pSrcInfo->lutBase;
0N/A mlib_s32 dstScan = pDstInfo->scanStride;
0N/A mlib_s32 srcScan = pSrcInfo->scanStride;
0N/A mlib_d64 dd, d_bgpixel;
0N/A mlib_s32 j, x, mask;
0N/A
0N/A if (srcScan == width && dstScan == 4*width) {
0N/A width *= height;
0N/A height = 1;
0N/A }
0N/A
0N/A d_bgpixel = vis_to_double_dup(bgpixel);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = srcBase;
0N/A mlib_s32 *dst = dstBase;
0N/A mlib_s32 *dst_end;
0N/A
0N/A dst_end = dst + width;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A x = pixLut[*src++];
0N/A if (x < 0) {
0N/A *dst = ARGB_to_GBGR(x);
0N/A } else {
0N/A *dst = bgpixel;
0N/A }
0N/A dst++;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= (dst_end - 2); dst += 2) {
0N/A mlib_f32 *pp0 = (mlib_f32*)pixLut + src[0];
0N/A mlib_f32 *pp1 = (mlib_f32*)pixLut + src[1];
0N/A ARGB_to_GBGR_FL2(dd, *pp0, *pp1);
0N/A mask = (((*(mlib_u8*)pp0) >> 6) & 2) | ((*(mlib_u8*)pp1) >> 7);
0N/A *(mlib_d64*)dst = d_bgpixel;
0N/A vis_pst_32(dd, dst, mask);
0N/A src += 2;
0N/A }
0N/A
0N/A while (dst < dst_end) {
0N/A x = pixLut[*src++];
0N/A if (x < 0) {
0N/A *dst = ARGB_to_GBGR(x);
0N/A } else {
0N/A *dst = bgpixel;
0N/A }
0N/A dst++;
0N/A }
0N/A
0N/A PTR_ADD(dstBase, dstScan);
0N/A PTR_ADD(srcBase, srcScan);
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/Avoid ADD_SUFF(IntBgrDrawGlyphListAA)(GLYPH_LIST_PARAMS)
0N/A{
0N/A mlib_s32 glyphCounter;
0N/A mlib_s32 scan = pRasInfo->scanStride;
0N/A mlib_u8 *dstBase;
0N/A mlib_s32 j;
0N/A mlib_d64 dmix0, dmix1, dd, d0, d1, e0, e1, fgpixel_d;
0N/A mlib_d64 done, done16, d_half, maskRGB, dzero;
0N/A mlib_s32 pix, mask, mask_z;
0N/A mlib_f32 srcG_f;
0N/A
0N/A done = vis_to_double_dup(0x7fff7fff);
0N/A done16 = vis_to_double_dup(0x7fff);
0N/A d_half = vis_to_double_dup((1 << (16 + 6)) | (1 << 6));
0N/A
0N/A fgpixel_d = vis_to_double_dup(fgpixel);
0N/A srcG_f = vis_to_float(argbcolor);
0N/A maskRGB = vis_to_double_dup(0xffffff);
0N/A dzero = vis_fzero();
0N/A
0N/A ARGB2ABGR_FL(srcG_f)
0N/A
0N/A vis_write_gsr(0 << 3);
0N/A
0N/A for (glyphCounter = 0; glyphCounter < totalGlyphs; glyphCounter++) {
0N/A const jubyte *pixels;
0N/A unsigned int rowBytes;
0N/A int left, top;
0N/A int width, height;
0N/A int right, bottom;
0N/A
0N/A pixels = (const jubyte *) glyphs[glyphCounter].pixels;
0N/A
0N/A if (!pixels) continue;
0N/A
0N/A left = glyphs[glyphCounter].x;
0N/A top = glyphs[glyphCounter].y;
0N/A width = glyphs[glyphCounter].width;
0N/A height = glyphs[glyphCounter].height;
0N/A rowBytes = width;
0N/A right = left + width;
0N/A bottom = top + height;
0N/A if (left < clipLeft) {
0N/A pixels += clipLeft - left;
0N/A left = clipLeft;
0N/A }
0N/A if (top < clipTop) {
0N/A pixels += (clipTop - top) * rowBytes;
0N/A top = clipTop;
0N/A }
0N/A if (right > clipRight) {
0N/A right = clipRight;
0N/A }
0N/A if (bottom > clipBottom) {
0N/A bottom = clipBottom;
0N/A }
0N/A if (right <= left || bottom <= top) {
0N/A continue;
0N/A }
0N/A width = right - left;
0N/A height = bottom - top;
0N/A
0N/A dstBase = pRasInfo->rasBase;
0N/A PTR_ADD(dstBase, top*scan + 4*left);
0N/A
0N/A for (j = 0; j < height; j++) {
0N/A mlib_u8 *src = (void*)pixels;
0N/A mlib_s32 *dst, *dst_end;
0N/A
0N/A dst = (void*)dstBase;
0N/A dst_end = dst + width;
0N/A
0N/A if ((mlib_s32)dst & 7) {
0N/A pix = *src++;
0N/A if (pix) {
0N/A dd = vis_fpadd16(MUL8_VIS(srcG_f, pix), d_half);
0N/A dd = vis_fpadd16(MUL8_VIS(*(mlib_f32*)dst, 255 - pix), dd);
0N/A *(mlib_f32*)dst = vis_fands(vis_fpack16(dd),
0N/A vis_read_hi(maskRGB));
0N/A if (pix == 255) *(mlib_f32*)dst = vis_read_hi(fgpixel_d);
0N/A }
0N/A dst++;
0N/A }
0N/A
0N/A#pragma pipeloop(0)
0N/A for (; dst <= (dst_end - 2); dst += 2) {
0N/A dmix0 = vis_freg_pair(((mlib_f32 *)vis_mul8s_tbl)[src[0]],
0N/A ((mlib_f32 *)vis_mul8s_tbl)[src[1]]);
0N/A mask = vis_fcmplt32(dmix0, done16);
0N/A mask_z = vis_fcmpne32(dmix0, dzero);
0N/A dmix1 = vis_fpsub16(done, dmix0);
0N/A src += 2;
0N/A
0N/A dd = *(mlib_d64*)dst;
0N/A d0 = vis_fmul8x16al(srcG_f, vis_read_hi(dmix0));
0N/A d1 = vis_fmul8x16al(srcG_f, vis_read_lo(dmix0));
0N/A e0 = vis_fmul8x16al(vis_read_hi(dd), vis_read_hi(dmix1));
0N/A e1 = vis_fmul8x16al(vis_read_lo(dd), vis_read_lo(dmix1));
0N/A d0 = vis_fpadd16(vis_fpadd16(d0, d_half), e0);
0N/A d1 = vis_fpadd16(vis_fpadd16(d1, d_half), e1);
0N/A dd = vis_fpack16_pair(d0, d1);
0N/A dd = vis_fand(dd, maskRGB);
0N/A
0N/A vis_pst_32(fgpixel_d, dst, mask_z);
0N/A vis_pst_32(dd, dst, mask & mask_z);
0N/A }
0N/A
0N/A while (dst < dst_end) {
0N/A pix = *src++;
0N/A if (pix) {
0N/A dd = vis_fpadd16(MUL8_VIS(srcG_f, pix), d_half);
0N/A dd = vis_fpadd16(MUL8_VIS(*(mlib_f32*)dst, 255 - pix), dd);
0N/A *(mlib_f32*)dst = vis_fands(vis_fpack16(dd),
0N/A vis_read_hi(maskRGB));
0N/A if (pix == 255) *(mlib_f32*)dst = vis_read_hi(fgpixel_d);
0N/A }
0N/A dst++;
0N/A }
0N/A
0N/A PTR_ADD(dstBase, scan);
0N/A pixels += rowBytes;
0N/A }
0N/A }
0N/A}
0N/A
0N/A/***************************************************************/
0N/A
0N/A#endif /* JAVA2D_NO_MLIB */