CharacterConverter.cc

Go to the documentation of this file.

/*
TODO:
- Clean up renderGraphics2, it is currently very hard to understand
  with all the masks and quarters etc.
- Correctly implement vertical scroll in text modes.
  Can be implemented by reordering blitting, but uses a smaller
  wrap than GFX modes: 8 lines instead of 256 lines.
*/
 
#include "CharacterConverter.hh"
 
#include "VDP.hh"
#include "VDPVRAM.hh"
 
#include "ranges.hh"
#include "xrange.hh"
 
#include <bit>
#include <cstdint>
 
#ifdef __SSE2__
#include "emmintrin.h" // SSE2
#endif
 
namespace openmsx {
 
using Pixel = CharacterConverter::Pixel;
 
CharacterConverter::CharacterConverter(
        VDP& vdp_, std::span<const Pixel, 16> palFg_, std::span<const Pixel, 16> palBg_)
        : vdp(vdp_), vram(vdp.getVRAM()), palFg(palFg_), palBg(palBg_)
{
}
 
void CharacterConverter::setDisplayMode(DisplayMode mode)
{
        modeBase = mode.getBase();
        assert(modeBase < 0x0C);
}
 
void CharacterConverter::convertLine(std::span<Pixel> buf, int line) const
{
        // TODO: Support YJK on modes other than Graphic 6/7.
        switch (modeBase) {
        case DisplayMode::GRAPHIC1:   // screen 1
                renderGraphic1(subspan<256>(buf), line);
                break;
        case DisplayMode::TEXT1:      // screen 0, width 40
                renderText1(subspan<256>(buf), line);
                break;
        case DisplayMode::MULTICOLOR: // screen 3
                renderMulti(subspan<256>(buf), line);
                break;
        case DisplayMode::GRAPHIC2:   // screen 2
                renderGraphic2(subspan<256>(buf), line);
                break;
        case DisplayMode::GRAPHIC3:   // screen 4
                renderGraphic2(subspan<256>(buf), line); // graphic3, actually
                break;
        case  DisplayMode::TEXT2:     // screen 0, width 80
                renderText2(subspan<512>(buf), line);
                break;
        case DisplayMode::TEXT1Q:     // TMSxxxx only
                if (vdp.isMSX1VDP()) {
                        renderText1Q(subspan<256>(buf), line);
                } else {
                        renderBlank (subspan<256>(buf));
                }
                break;
        case DisplayMode::MULTIQ:     // TMSxxxx only
                if (vdp.isMSX1VDP()) {
                        renderMultiQ(subspan<256>(buf), line);
                } else {
                        renderBlank (subspan<256>(buf));
                }
                break;
        default: // remaining (non-bitmap) modes
                if (vdp.isMSX1VDP()) {
                        renderBogus(subspan<256>(buf));
                } else {
                        renderBlank(subspan<256>(buf));
                }
        }
}
 
#ifdef __SSE2__
// Copied from Scale2xScaler.cc, TODO move to common location?
static inline __m128i select(__m128i a0, __m128i a1, __m128i mask)
{
        return _mm_xor_si128(_mm_and_si128(_mm_xor_si128(a0, a1), mask), a0);
}
#endif
 
static inline void draw6(
        Pixel* __restrict & pixelPtr, Pixel fg, Pixel bg, byte pattern)
{
        pixelPtr[0] = (pattern & 0x80) ? fg : bg;
        pixelPtr[1] = (pattern & 0x40) ? fg : bg;
        pixelPtr[2] = (pattern & 0x20) ? fg : bg;
        pixelPtr[3] = (pattern & 0x10) ? fg : bg;
        pixelPtr[4] = (pattern & 0x08) ? fg : bg;
        pixelPtr[5] = (pattern & 0x04) ? fg : bg;
        pixelPtr += 6;
}
 
static inline void draw8(
        Pixel* __restrict & pixelPtr, Pixel fg, Pixel bg, byte pattern)
{
#ifdef __SSE2__
        // SSE2 version, 32bpp
        const __m128i m74 = _mm_set_epi32(0x10, 0x20, 0x40, 0x80);
        const __m128i m30 = _mm_set_epi32(0x01, 0x02, 0x04, 0x08);
        const __m128i zero = _mm_setzero_si128();
 
        __m128i fg4 = _mm_set1_epi32(fg);
        __m128i bg4 = _mm_set1_epi32(bg);
        __m128i pat = _mm_set1_epi32(pattern);
 
        __m128i b74 = _mm_cmpeq_epi32(_mm_and_si128(pat, m74), zero);
        __m128i b30 = _mm_cmpeq_epi32(_mm_and_si128(pat, m30), zero);
 
        auto* out = std::bit_cast<__m128i*>(pixelPtr);
        _mm_storeu_si128(out + 0, select(fg4, bg4, b74));
        _mm_storeu_si128(out + 1, select(fg4, bg4, b30));
        pixelPtr += 8;
        return;
#endif
 
        // C++ version
        pixelPtr[0] = (pattern & 0x80) ? fg : bg;
        pixelPtr[1] = (pattern & 0x40) ? fg : bg;
        pixelPtr[2] = (pattern & 0x20) ? fg : bg;
        pixelPtr[3] = (pattern & 0x10) ? fg : bg;
        pixelPtr[4] = (pattern & 0x08) ? fg : bg;
        pixelPtr[5] = (pattern & 0x04) ? fg : bg;
        pixelPtr[6] = (pattern & 0x02) ? fg : bg;
        pixelPtr[7] = (pattern & 0x01) ? fg : bg;
        pixelPtr += 8;
}
 
void CharacterConverter::renderText1(std::span<Pixel, 256> buf, int line) const
{
        Pixel fg = palFg[vdp.getForegroundColor()];
        Pixel bg = palFg[vdp.getBackgroundColor()];
 
        // 8 * 256 is small enough to always be contiguous
        auto patternArea = vram.patternTable.getReadArea<256 * 8>(0);
        auto l = (line + vdp.getVerticalScroll()) & 7;
 
        // Note: Because line width is not a power of two, reading an entire line
        //       from a VRAM pointer returned by readArea will not wrap the index
        //       correctly. Therefore we read one character at a time.
        unsigned nameStart = (line / 8) * 40;
        unsigned nameEnd = nameStart + 40;
        Pixel* __restrict pixelPtr = buf.data();
        for (auto name : xrange(nameStart, nameEnd)) {
                unsigned charCode = vram.nameTable.readNP((name + 0xC00) | (~0u << 12));
                auto pattern = patternArea[l + charCode * 8];
                draw6(pixelPtr, fg, bg, pattern);
        }
}
 
void CharacterConverter::renderText1Q(std::span<Pixel, 256> buf, int line) const
{
        Pixel fg = palFg[vdp.getForegroundColor()];
        Pixel bg = palFg[vdp.getBackgroundColor()];
 
        unsigned patternBaseLine = (~0u << 13) | ((line + vdp.getVerticalScroll()) & 7);
 
        // Note: Because line width is not a power of two, reading an entire line
        //       from a VRAM pointer returned by readArea will not wrap the index
        //       correctly. Therefore we read one character at a time.
        unsigned nameStart = (line / 8) * 40;
        unsigned nameEnd = nameStart + 40;
        unsigned patternQuarter = (line & 0xC0) << 2;
        Pixel* __restrict pixelPtr = buf.data();
        for (auto name : xrange(nameStart, nameEnd)) {
                unsigned charCode = vram.nameTable.readNP((name + 0xC00) | (~0u << 12));
                unsigned patternNr = patternQuarter | charCode;
                auto pattern = vram.patternTable.readNP(
                        patternBaseLine | (patternNr * 8));
                draw6(pixelPtr, fg, bg, pattern);
        }
}
 
void CharacterConverter::renderText2(std::span<Pixel, 512> buf, int line) const
{
        Pixel plainFg = palFg[vdp.getForegroundColor()];
        Pixel plainBg = palFg[vdp.getBackgroundColor()];
        Pixel blinkFg, blinkBg;
        if (vdp.getBlinkState()) {
                int fg = vdp.getBlinkForegroundColor();
                blinkFg = palBg[fg ? fg : vdp.getBlinkBackgroundColor()];
                blinkBg = palBg[vdp.getBlinkBackgroundColor()];
        } else {
                blinkFg = plainFg;
                blinkBg = plainBg;
        }
 
        // 8 * 256 is small enough to always be contiguous
        auto patternArea = vram.patternTable.getReadArea<256 * 8>(0);
        auto l = (line + vdp.getVerticalScroll()) & 7;
 
        unsigned colorStart = (line / 8) * (80 / 8);
        unsigned nameStart  = (line / 8) * 80;
        Pixel* __restrict pixelPtr = buf.data();
        for (auto i : xrange(80 / 8)) {
                unsigned colorPattern = vram.colorTable.readNP(
                        (colorStart + i) | (~0u << 9));
                auto nameArea = vram.nameTable.getReadArea<8>(
                        (nameStart + 8 * i) | (~0u << 12));
                draw6(pixelPtr,
                      (colorPattern & 0x80) ? blinkFg : plainFg,
                      (colorPattern & 0x80) ? blinkBg : plainBg,
                      patternArea[l + nameArea[0] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x40) ? blinkFg : plainFg,
                      (colorPattern & 0x40) ? blinkBg : plainBg,
                      patternArea[l + nameArea[1] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x20) ? blinkFg : plainFg,
                      (colorPattern & 0x20) ? blinkBg : plainBg,
                      patternArea[l + nameArea[2] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x10) ? blinkFg : plainFg,
                      (colorPattern & 0x10) ? blinkBg : plainBg,
                      patternArea[l + nameArea[3] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x08) ? blinkFg : plainFg,
                      (colorPattern & 0x08) ? blinkBg : plainBg,
                      patternArea[l + nameArea[4] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x04) ? blinkFg : plainFg,
                      (colorPattern & 0x04) ? blinkBg : plainBg,
                      patternArea[l + nameArea[5] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x02) ? blinkFg : plainFg,
                      (colorPattern & 0x02) ? blinkBg : plainBg,
                      patternArea[l + nameArea[6] * 8]);
                draw6(pixelPtr,
                      (colorPattern & 0x01) ? blinkFg : plainFg,
                      (colorPattern & 0x01) ? blinkBg : plainBg,
                      patternArea[l + nameArea[7] * 8]);
        }
}
 
std::span<const byte, 32> CharacterConverter::getNamePtr(int line, int scroll) const
{
        // no need to test whether multi-page scrolling is enabled,
        // indexMask in the nameTable already takes care of it
        return vram.nameTable.getReadArea<32>(
                ((line / 8) * 32) | ((scroll & 0x20) ? 0x8000 : 0));
}
void CharacterConverter::renderGraphic1(std::span<Pixel, 256> buf, int line) const
{
        auto patternArea = vram.patternTable.getReadArea<256 * 8>(0);
        auto l = line & 7;
        auto colorArea = vram.colorTable.getReadArea<256 / 8>(0);
 
        int scroll = vdp.getHorizontalScrollHigh();
        auto namePtr = getNamePtr(line, scroll);
        Pixel* __restrict pixelPtr = buf.data();
        repeat(32, [&] {
                auto charCode = namePtr[scroll & 0x1F];
                auto pattern = patternArea[l + charCode * 8];
                auto color = colorArea[charCode / 8];
                Pixel fg = palFg[color >> 4];
                Pixel bg = palFg[color & 0x0F];
                draw8(pixelPtr, fg, bg, pattern);
                if (!(++scroll & 0x1F)) namePtr = getNamePtr(line, scroll);
        });
}
 
void CharacterConverter::renderGraphic2(std::span<Pixel, 256> buf, int line) const
{
        int quarter8 = (((line / 8) * 32) & ~0xFF) * 8;
        int line7 = line & 7;
        int scroll = vdp.getHorizontalScrollHigh();
        auto namePtr = getNamePtr(line, scroll);
 
        Pixel* __restrict pixelPtr = buf.data();
        if (vram.colorTable  .isContinuous((8 * 256) - 1) &&
            vram.patternTable.isContinuous((8 * 256) - 1) &&
            ((scroll & 0x1f) == 0)) {
                // Both color and pattern table can be accessed contiguously
                // (no mirroring) and there's no v9958 horizontal scrolling.
                // This is very common, so make an optimized version for this.
                auto patternArea = vram.patternTable.getReadArea<256 * 8>(quarter8);
                auto colorArea   = vram.colorTable  .getReadArea<256 * 8>(quarter8);
                for (auto n : xrange(32)) {
                        auto charCode8 = namePtr[n] * 8;
                        auto pattern = patternArea[line7 + charCode8];
                        auto color   = colorArea  [line7 + charCode8];
                        Pixel fg = palFg[color >> 4];
                        Pixel bg = palFg[color & 0x0F];
                        draw8(pixelPtr, fg, bg, pattern);
                }
        } else {
                // Slower variant, also works when:
                // - there is mirroring in the color table
                // - there is mirroring in the pattern table (TMS9929)
                // - V9958 horizontal scroll feature is used
                unsigned baseLine = (~0u << 13) | quarter8 | line7;
                repeat(32, [&] {
                        unsigned charCode8 = namePtr[scroll & 0x1F] * 8;
                        unsigned index = charCode8 | baseLine;
                        auto pattern = vram.patternTable.readNP(index);
                        auto color   = vram.colorTable  .readNP(index);
                        Pixel fg = palFg[color >> 4];
                        Pixel bg = palFg[color & 0x0F];
                        draw8(pixelPtr, fg, bg, pattern);
                        if (!(++scroll & 0x1F)) namePtr = getNamePtr(line, scroll);
                });
        }
}
 
void CharacterConverter::renderMultiHelper(
        Pixel* __restrict pixelPtr, int line,
        unsigned mask, unsigned patternQuarter) const
{
        unsigned baseLine = mask | ((line / 4) & 7);
        unsigned scroll = vdp.getHorizontalScrollHigh();
        auto namePtr = getNamePtr(line, scroll);
        repeat(32, [&] {
                unsigned patternNr = patternQuarter | namePtr[scroll & 0x1F];
                unsigned color = vram.patternTable.readNP((patternNr * 8) | baseLine);
                Pixel cl = palFg[color >> 4];
                Pixel cr = palFg[color & 0x0F];
                pixelPtr[0] = cl; pixelPtr[1] = cl;
                pixelPtr[2] = cl; pixelPtr[3] = cl;
                pixelPtr[4] = cr; pixelPtr[5] = cr;
                pixelPtr[6] = cr; pixelPtr[7] = cr;
                pixelPtr += 8;
                if (!(++scroll & 0x1F)) namePtr = getNamePtr(line, scroll);
        });
}
void CharacterConverter::renderMulti(std::span<Pixel, 256> buf, int line) const
{
        unsigned mask = (~0u << 11);
        renderMultiHelper(buf.data(), line, mask, 0);
}
 
void CharacterConverter::renderMultiQ(
        std::span<Pixel, 256> buf, int line) const
{
        unsigned mask = (~0u << 13);
        unsigned patternQuarter = (line * 4) & ~0xFF;  // (line / 8) * 32
        renderMultiHelper(buf.data(), line, mask, patternQuarter);
}
 
void CharacterConverter::renderBogus(std::span<Pixel, 256> buf) const
{
        Pixel* __restrict pixelPtr = buf.data();
        Pixel fg = palFg[vdp.getForegroundColor()];
        Pixel bg = palFg[vdp.getBackgroundColor()];
        auto draw = [&](int n, Pixel col) {
                pixelPtr = std::fill_n(pixelPtr, n, col);
        };
        draw(8, bg);
        repeat(40, [&] {
                draw(4, fg);
                draw(2, bg);
        });
        draw(8, bg);
}
 
void CharacterConverter::renderBlank(std::span<Pixel, 256> buf) const
{
        // when this is in effect, the VRAM is not refreshed anymore, but that
        // is not emulated
        ranges::fill(buf, palFg[15]);
}
 
} // namespace openmsx