--- old/src/java.desktop/share/native/libawt/java2d/loops/LoopMacros.h 2016-03-28 11:16:22.878832931 +0530 +++ new/src/java.desktop/share/native/libawt/java2d/loops/LoopMacros.h 2016-03-28 11:16:22.646832931 +0530 @@ -1668,31 +1668,76 @@ } \ } while (0); +/* + * Antialiased glyph drawing results in artifacts around the character edges + * when text is drawn ontop of translucent background color. The standard + * blending equation for two colors: + * destColor = srcColor * glyphAlpha + destColor * (1 - glyphAlpha) + * works only when srcColor and destColor are opaque. For translucent srcColor + * and destColor, the respective alpha components in each color will influence + * the visibility of the color and the visibility of the color below it. Hence + * the equation for blending is given as: + * resA = srcAlpha + dstAlpha * (1 - srcAlpha) + * resCol = (srcColor * srcAlpha + destColor * destAlpha * (1- srcAlpha))/resA + * In addition, srcAlpha is multiplied with the glyphAlpha- that indicates the + * grayscale mask value of the glyph being drawn. The combined result provides + * smooth antialiased text on the buffer without any artifacts. Since the + * logic is executed for every pixel in a glyph, the implementation is further + * optimized to reduce computation and improve execution time. + */ #define GlyphListAABlend4ByteArgb(DST, GLYPH_PIXELS, PIXEL_INDEX, DST_PTR, \ FG_PIXEL, PREFIX, SRC_PREFIX) \ - do { \ - DeclareAlphaVarFor4ByteArgb(dstA) \ - DeclareCompVarsFor4ByteArgb(dst) \ + do { \ + DeclareAlphaVarFor4ByteArgb(resA) \ + DeclareCompVarsFor4ByteArgb(res) \ jint mixValSrc = GLYPH_PIXELS[PIXEL_INDEX]; \ if (mixValSrc) { \ - if (mixValSrc < 255) { \ - jint mixValDst = 255 - mixValSrc; \ - Load ## DST ## To4ByteArgb(DST_PTR, pix, PIXEL_INDEX, \ - dstA, dstR, dstG, dstB); \ - dstA = MUL8(dstA, mixValDst) + \ - MUL8(SRC_PREFIX ## A, mixValSrc); \ - MultMultAddAndStore4ByteArgbComps(dst, mixValDst, dst, \ - mixValSrc, SRC_PREFIX); \ - if (!(DST ## IsOpaque) && \ - !(DST ## IsPremultiplied) && dstA && dstA < 255) { \ - DivideAndStore4ByteArgbComps(dst, dst, dstA); \ - } \ - Store ## DST ## From4ByteArgbComps(DST_PTR, pix, \ - PIXEL_INDEX, dst); \ + if (mixValSrc != 0xff) { \ + PromoteByteAlphaFor4ByteArgb(mixValSrc); \ + resA = MultiplyAlphaFor4ByteArgb(mixValSrc, SRC_PREFIX ## A); \ + MultiplyAndStore4ByteArgbComps(res, resA, SRC_PREFIX); \ } else { \ - Store ## DST ## PixelData(DST_PTR, PIXEL_INDEX, \ - FG_PIXEL, PREFIX); \ + resA = SRC_PREFIX ## A; \ + MultiplyAndStore4ByteArgbComps(res, \ + SRC_PREFIX ## A, \ + SRC_PREFIX); \ + } \ + if (resA != MaxValFor4ByteArgb) { \ + DeclareAndInvertAlphaVarFor4ByteArgb(dstF, resA) \ + DeclareAndClearAlphaVarFor4ByteArgb(dstA) \ + Declare ## DST ## AlphaLoadData(DstPix) \ + jint pixelOffset = PIXEL_INDEX * (DST ## PixelStride); \ + DST ## DataType *pixelAddress = PtrAddBytes(DST_PTR, \ + pixelOffset); \ + LoadAlphaFrom ## DST ## For4ByteArgb(pixelAddress, \ + DstPix, \ + dst); \ + dstA = MultiplyAlphaFor4ByteArgb(dstF, dstA); \ + if (!(DST ## IsPremultiplied)) { \ + dstF = dstA; \ + } \ + resA += dstA; \ + if (dstF) { \ + DeclareCompVarsFor4ByteArgb(tmp) \ + Postload4ByteArgbFrom ## DST(pixelAddress, \ + DstPix, \ + tmp); \ + if (dstF != MaxValFor4ByteArgb) { \ + MultiplyAndStore4ByteArgbComps(tmp, \ + dstF, \ + tmp); \ + } \ + Store4ByteArgbCompsUsingOp(res, +=, tmp); \ + } \ + } \ + if (!(DST ## IsOpaque) && \ + !(DST ## IsPremultiplied) && resA && \ + resA < MaxValFor4ByteArgb) \ + { \ + DivideAndStore4ByteArgbComps(res, res, resA); \ } \ + Store ## DST ## From4ByteArgbComps(DST_PTR, pix, \ + PIXEL_INDEX, res); \ } \ } while (0);