// license:BSD-3-Clause // copyright-holders:smf /* * PlayStation GPU emulator * * Copyright 2003-2014 smf * */ #include "emu.h" #include "cpu/psx/psx.h" #include "video/psx.h" #include "screen.h" #define STOP_ON_ERROR ( 0 ) #define VERBOSE_LEVEL ( 0 ) // device type definition DEFINE_DEVICE_TYPE(CXD8514Q, cxd8514q_device, "cxd8514q", "CXD8514Q GPU") // VRAM DEFINE_DEVICE_TYPE(CXD8538Q, cxd8538q_device, "cxd8538q", "CXD8538Q GPU") // VRAM DEFINE_DEVICE_TYPE(CXD8561Q, cxd8561q_device, "cxd8561q", "CXD8561Q GPU") // SGRAM DEFINE_DEVICE_TYPE(CXD8561BQ, cxd8561bq_device, "cxd8561bq", "CXD8561BQ GPU") // SGRAM DEFINE_DEVICE_TYPE(CXD8561CQ, cxd8561cq_device, "cxd8561cq", "CXD8561CQ GPU") // SGRAM DEFINE_DEVICE_TYPE(CXD8654Q, cxd8654q_device, "cxd8654q", "CXD8654Q GPU") // SGRAM psxgpu_device::psxgpu_device(const machine_config &mconfig, device_type type, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, type, tag, owner, clock) { set_vram_size(vram_size); cpu->gpu_read().set(tag, FUNC(psxgpu_device::read)); cpu->gpu_write().set(tag, FUNC(psxgpu_device::write)); cpu->subdevice("dma")->install_read_handler(2, psxdma_device::read_delegate(&psxgpu_device::dma_read, this)); cpu->subdevice("dma")->install_write_handler(2, psxdma_device::write_delegate(&psxgpu_device::dma_write, this)); vblank_callback().set(*cpu->subdevice("irq"), FUNC(psxirq_device::intin0)); } psxgpu_device::psxgpu_device(const machine_config &mconfig, device_type type, const char *tag, device_t *owner, uint32_t clock) : device_t(mconfig, type, tag, owner, clock) , device_video_interface(mconfig, *this) , device_palette_interface(mconfig, *this) , m_vblank_handler(*this) { } void psxgpu_device::device_start() { m_vblank_handler.resolve_safe(); screen().register_vblank_callback(vblank_state_delegate(&psxgpu_device::vblank, this)); for( int n_colour = 0; n_colour < 0x10000; n_colour++ ) { set_pen_color( n_colour, pal555(n_colour,0, 5, 10) ); } if (type() == CXD8538Q) { psx_gpu_init( 1 ); } else { psx_gpu_init( 2 ); } } void psxgpu_device::device_reset() { gpu_reset(); } cxd8514q_device::cxd8514q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8514Q, tag, owner, clock, vram_size, cpu) { } cxd8514q_device::cxd8514q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8514Q, tag, owner, clock) { } cxd8538q_device::cxd8538q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8538Q, tag, owner, clock, vram_size, cpu) { } cxd8538q_device::cxd8538q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8538Q, tag, owner, clock) { } cxd8561q_device::cxd8561q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8561Q, tag, owner, clock, vram_size, cpu) { } cxd8561q_device::cxd8561q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8561Q, tag, owner, clock) { } cxd8561bq_device::cxd8561bq_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8561BQ, tag, owner, clock, vram_size, cpu) { } cxd8561bq_device::cxd8561bq_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8561BQ, tag, owner, clock) { } cxd8561cq_device::cxd8561cq_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8561CQ, tag, owner, clock, vram_size, cpu) { } cxd8561cq_device::cxd8561cq_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8561CQ, tag, owner, clock) { } cxd8654q_device::cxd8654q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock, uint32_t vram_size, psxcpu_device *cpu) : psxgpu_device(mconfig, CXD8654Q, tag, owner, clock, vram_size, cpu) { } cxd8654q_device::cxd8654q_device(const machine_config &mconfig, const char *tag, device_t *owner, uint32_t clock) : psxgpu_device(mconfig, CXD8654Q, tag, owner, clock) { } static const int m_p_n_nextpointlist4[] = { 1, 3, 0, 2 }; static const int m_p_n_prevpointlist4[] = { 2, 0, 3, 1 }; static const int m_p_n_nextpointlist4b[] = { 0, 3, 1, 2 }; static const int m_p_n_prevpointlist4b[] = { 0, 2, 3, 1 }; static const int m_p_n_nextpointlist3[] = { 1, 2, 0 }; static const int m_p_n_prevpointlist3[] = { 2, 0, 1 }; #define SINT11( x ) ( ( (int32_t)( x ) << 21 ) >> 21 ) #define COORD_X( a ) ( a.sw.l ) #define COORD_Y( a ) ( a.sw.h ) #define SIZE_W( a ) ( a.w.l ) #define SIZE_H( a ) ( a.w.h ) #define BGR_C( a ) ( a.b.h3 ) #define BGR_B( a ) ( a.b.h2 ) #define BGR_G( a ) ( a.b.h ) #define BGR_R( a ) ( a.b.l ) #define TEXTURE_V( a ) ( a.b.h ) #define TEXTURE_U( a ) ( a.b.l ) static inline void ATTR_PRINTF(3,4) verboselog( device_t& device, int n_level, const char *s_fmt, ... ) { if( VERBOSE_LEVEL >= n_level ) { va_list v; char buf[ 32768 ]; va_start( v, s_fmt ); vsprintf( buf, s_fmt, v ); va_end( v ); device.logerror( "%s: %s", device.machine().describe_context(), buf ); } } #if PSXGPU_DEBUG_VIEWER void psxgpu_device::DebugMeshInit() { int width = screen().width(); int height = screen().height(); m_debug.b_mesh = 0; m_debug.b_texture = 0; m_debug.n_interleave = -1; m_debug.b_clear = 1; m_debug.n_coord = 0; m_debug.n_skip = 0; m_debug.mesh = std::make_unique(width, height ); } void psxgpu_device::DebugMesh( int n_coordx, int n_coordy ) { int width = screen().width(); int height = screen().height(); n_coordx += m_n_displaystartx; n_coordy += n_displaystarty; if( m_debug.b_clear ) { m_debug.mesh->fill(0x0000); m_debug.b_clear = 0; } int n_coord; int n_colour = 0x1f; for( n_coord = 0; n_coord < m_debug.n_coord; n_coord++ ) { if( n_coordx != m_debug.n_coordx[ n_coord ] || n_coordy != m_debug.n_coordy[ n_coord ] ) { break; } } if( n_coord == m_debug.n_coord && m_debug.n_coord > 1 ) { n_colour = 0xffff; } for( n_coord = 0; n_coord < m_debug.n_coord; n_coord++ ) { int32_t n_xstart = m_debug.n_coordx[ n_coord ]; int32_t n_xend = n_coordx; int32_t n_xlen; if( n_xend > n_xstart ) { n_xlen = n_xend - n_xstart; } else { n_xlen = n_xstart - n_xend; } int32_t n_ystart = m_debug.n_coordy[ n_coord ]; int32_t n_yend = n_coordy; int32_t n_ylen; if( n_yend > n_ystart ) { n_ylen = n_yend - n_ystart; } else { n_ylen = n_ystart - n_yend; } int32_t n_len; if( n_xlen > n_ylen ) { n_len = n_xlen; } else { n_len = n_ylen; } PAIR n_x; n_x.sw.h = n_xstart; n_x.sw.l = 0; PAIR n_y; n_y.sw.h = n_ystart; n_y.sw.l = 0; if( n_len == 0 ) { n_len = 1; } int32_t n_dx = (int32_t)( ( n_xend << 16 ) - n_x.d ) / n_len; int32_t n_dy = (int32_t)( ( n_yend << 16 ) - n_y.d ) / n_len; while( n_len > 0 ) { if( (int16_t)n_x.w.h >= 0 && (int16_t)n_y.w.h >= 0 && (int16_t)n_x.w.h <= width - 1 && (int16_t)n_y.w.h <= height - 1 ) { if( m_debug.mesh->pix16( n_y.w.h, n_x.w.h ) != 0xffff ) m_debug.mesh->pix16( n_y.w.h, n_x.w.h ) = n_colour; } n_x.d += n_dx; n_y.d += n_dy; n_len--; } } if( m_debug.n_coord < DEBUG_COORDS ) { m_debug.n_coordx[ m_debug.n_coord ] = n_coordx; m_debug.n_coordy[ m_debug.n_coord ] = n_coordy; m_debug.n_coord++; } } void psxgpu_device::DebugMeshEnd() { m_debug.n_coord = 0; } void psxgpu_device::DebugCheckKeys() { if( machine().input().code_pressed_once( KEYCODE_M ) ) { m_debug.b_mesh = !m_debug.b_mesh; updatevisiblearea(); } if( machine().input().code_pressed_once( KEYCODE_V ) ) { m_debug.b_texture = !m_debug.b_texture; updatevisiblearea(); } if( machine().input().code_pressed_once( KEYCODE_I ) ) { if( m_debug.b_texture ) { m_debug.n_interleave++; if( m_debug.n_interleave == 2 ) m_debug.n_interleave = -1; if( m_debug.n_interleave == -1 ) popmessage( "interleave off" ); else if( m_debug.n_interleave == 0 ) popmessage( "4 bit interleave" ); else if( m_debug.n_interleave == 1 ) popmessage( "8 bit interleave" ); } else { m_debug.n_skip++; if( m_debug.n_skip > 15 ) m_debug.n_skip = 0; popmessage( "debug skip %d", m_debug.n_skip ); } } #if 0 if( machine().input().code_pressed_once( KEYCODE_D ) ) { FILE *f = fopen( "dump.txt", "w" ); for( int n_y = 256; n_y < 512; n_y++ ) for( int n_x = 640; n_x < 1024; n_x++ ) fprintf( f, "%04u,%04u = %04x\n", n_y, n_x, p_p_vram[ n_y ][ n_x ] ); fclose( f ); } if( machine().input().code_pressed_once( KEYCODE_S ) ) { popmessage( "saving..." ); FILE *f = fopen( "VRAM.BIN", "wb" ); for( int n_y = 0; n_y < 1024; n_y++ ) fwrite( p_p_vram[ n_y ], 1024 * 2, 1, f ); fclose( f ); } if( machine().input().code_pressed_once( KEYCODE_L ) ) { popmessage( "loading..." ); FILE *f = fopen( "VRAM.BIN", "rb" ); for( int n_y = 0; n_y < 1024; n_y++ ) fread( p_p_vram[ n_y ], 1024 * 2, 1, f ); fclose( f ); } #endif } int psxgpu_device::DebugMeshDisplay( bitmap_rgb32 &bitmap, const rectangle &cliprect ) { if( m_debug.b_mesh ) { for( int y = cliprect.min_y; y <= cliprect.max_y; y++ ) draw_scanline16( bitmap, cliprect.min_x, y, cliprect.max_x + 1 - cliprect.min_x, &m_debug.mesh->pix16( y ), pens() ); } m_debug.b_clear = 1; return m_debug.b_mesh; } int psxgpu_device::DebugTextureDisplay( bitmap_rgb32 &bitmap ) { if( m_debug.b_texture ) { int width = screen().width(); int height = screen().height(); for( int n_y = 0; n_y < height; n_y++ ) { uint16_t p_n_interleave[ 1024 ]; for( int n_x = 0; n_x < width; n_x++ ) { int n_xi; int n_yi; if( m_debug.n_interleave == 0 ) { n_xi = ( n_x & ~0x3c ) + ( ( n_y << 2 ) & 0x3c ); n_yi = ( n_y & ~0xf ) + ( ( n_x >> 2 ) & 0xf ); } else if( m_debug.n_interleave == 1 ) { n_xi = ( n_x & ~0x78 ) + ( ( n_x << 3 ) & 0x40 ) + ( ( n_y << 3 ) & 0x38 ); n_yi = ( n_y & ~0x7 ) + ( ( n_x >> 4 ) & 0x7 ); } else { n_xi = n_x; n_yi = n_y; } p_n_interleave[ n_x ] = p_p_vram[ n_yi ][ n_xi ]; } draw_scanline16( bitmap, 0, n_y, width, p_n_interleave, pens() ); } } return m_debug.b_texture; } #endif void psxgpu_device::updatevisiblearea() { rectangle visarea; float refresh; if( ( n_gpustatus & ( 1 << 0x14 ) ) != 0 ) { /* pal */ refresh = 50; switch( ( n_gpustatus >> 0x13 ) & 1 ) { case 0: n_screenheight = 256; break; case 1: n_screenheight = 512; break; } } else { /* ntsc */ refresh = 60; switch( ( n_gpustatus >> 0x13 ) & 1 ) { case 0: n_screenheight = 240; break; case 1: n_screenheight = 480; break; } } switch( ( n_gpustatus >> 0x11 ) & 3 ) { case 0: switch( ( n_gpustatus >> 0x10 ) & 1 ) { case 0: n_screenwidth = 256; break; case 1: n_screenwidth = 368; break; } break; case 1: switch( ( n_gpustatus >> 0x10 ) & 1 ) { case 0: n_screenwidth = 320; break; case 1: n_screenwidth = 384; break; } break; case 2: n_screenwidth = 512; break; case 3: n_screenwidth = 640; break; } #if PSXGPU_DEBUG_VIEWER if( m_debug.b_mesh || m_debug.b_texture ) { n_screenheight = 1024; n_screenwidth = 1024; } #endif visarea.set(0, n_screenwidth - 1, 0, n_screenheight - 1); screen().configure(n_screenwidth, n_screenheight, visarea, HZ_TO_ATTOSECONDS(refresh)); } void psxgpu_device::psx_gpu_init( int n_gputype ) { int width = 1024; int height = ( vramSize / width ) / sizeof( uint16_t ); m_n_gputype = n_gputype; #if PSXGPU_DEBUG_VIEWER DebugMeshInit(); #endif n_gpustatus = 0x14802000; n_gpuinfo = 0; n_gpu_buffer_offset = 0; n_lightgun_x = 0; n_lightgun_y = 0; b_reverseflag = 0; p_vram = make_unique_clear(width * height ); for( int n_line = 0; n_line < 1024; n_line++ ) { p_p_vram[ n_line ] = &p_vram[ ( n_line % height ) * width ]; } for( int n_level = 0; n_level < MAX_LEVEL; n_level++ ) { for( int n_shade = 0; n_shade < MAX_SHADE; n_shade++ ) { /* shaded */ int n_shaded = ( n_level * n_shade ) / MID_SHADE; if( n_shaded > MAX_LEVEL - 1 ) { n_shaded = MAX_LEVEL - 1; } p_n_redshade[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded; p_n_greenshade[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded << 5; p_n_blueshade[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded << 10; /* 1/4 x transparency */ n_shaded = ( n_level * n_shade ) / MID_SHADE; n_shaded >>= 2; if( n_shaded > MAX_LEVEL - 1 ) { n_shaded = MAX_LEVEL - 1; } p_n_f025[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded; /* 1/2 x transparency */ n_shaded = ( n_level * n_shade ) / MID_SHADE; n_shaded >>= 1; if( n_shaded > MAX_LEVEL - 1 ) { n_shaded = MAX_LEVEL - 1; } p_n_f05[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded; /* 1 x transparency */ n_shaded = ( n_level * n_shade ) / MID_SHADE; if( n_shaded > MAX_LEVEL - 1 ) { n_shaded = MAX_LEVEL - 1; } p_n_f1[ ( n_level * MAX_SHADE ) | n_shade ] = n_shaded; } } for( int n_level = 0; n_level < 0x10000; n_level++ ) { p_n_redlevel[ n_level ] = ( n_level & ( MAX_LEVEL - 1 ) ) * MAX_SHADE; p_n_greenlevel[ n_level ] = ( ( n_level >> 5 ) & ( MAX_LEVEL - 1 ) ) * MAX_SHADE; p_n_bluelevel[ n_level ] = ( ( n_level >> 10 ) & ( MAX_LEVEL - 1 ) ) * MAX_SHADE; /* 0.5 * background */ p_n_redb05[ n_level ] = ( ( n_level & ( MAX_LEVEL - 1 ) ) / 2 ) * MAX_LEVEL; p_n_greenb05[ n_level ] = ( ( ( n_level >> 5 ) & ( MAX_LEVEL - 1 ) ) / 2 ) * MAX_LEVEL; p_n_blueb05[ n_level ] = ( ( ( n_level >> 10 ) & ( MAX_LEVEL - 1 ) ) / 2 ) * MAX_LEVEL; /* 1 * background */ p_n_redb1[ n_level ] = ( n_level & ( MAX_LEVEL - 1 ) ) * MAX_LEVEL; p_n_greenb1[ n_level ] = ( ( n_level >> 5 ) & ( MAX_LEVEL - 1 ) ) * MAX_LEVEL; p_n_blueb1[ n_level ] = ( ( n_level >> 10 ) & ( MAX_LEVEL - 1 ) ) * MAX_LEVEL; /* 24bit color */ p_n_g0r0[ n_level ] = ( ( ( n_level >> 8 ) & 0xff ) << 8 ) | ( ( ( n_level >> 0 ) & 0xff ) << 16 ); p_n_b0[ n_level ] = ( ( n_level >> 0 ) & 0xff ) << 0; p_n_r1[ n_level ] = ( ( n_level >> 8 ) & 0xff ) << 16; p_n_b1g1[ n_level ] = ( ( ( n_level >> 8 ) & 0xff ) << 0 ) | ( ( ( n_level >> 0 ) & 0xff ) << 8 ); } for( int n_level = 0; n_level < MAX_LEVEL; n_level++ ) { for( int n_level2 = 0; n_level2 < MAX_LEVEL; n_level2++ ) { /* add transparency */ int n_shaded = ( n_level + n_level2 ); if( n_shaded > MAX_LEVEL - 1 ) { n_shaded = MAX_LEVEL - 1; } p_n_redaddtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded; p_n_greenaddtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded << 5; p_n_blueaddtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded << 10; /* sub transparency */ n_shaded = ( n_level - n_level2 ); if( n_shaded < 0 ) { n_shaded = 0; } p_n_redsubtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded; p_n_greensubtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded << 5; p_n_bluesubtrans[ ( n_level * MAX_LEVEL ) | n_level2 ] = n_shaded << 10; } } save_pointer(NAME(p_vram), width * height ); save_item(NAME(m_packet.n_entry)); save_item(NAME(n_gpu_buffer_offset)); save_item(NAME(n_vramx)); save_item(NAME(n_vramy)); save_item(NAME(n_twy)); save_item(NAME(n_twx)); save_item(NAME(n_tww)); save_item(NAME(n_drawarea_x1)); save_item(NAME(n_drawarea_y1)); save_item(NAME(n_drawarea_x2)); save_item(NAME(n_drawarea_y2)); save_item(NAME(n_horiz_disstart)); save_item(NAME(n_horiz_disend)); save_item(NAME(n_vert_disstart)); save_item(NAME(n_vert_disend)); save_item(NAME(b_reverseflag)); save_item(NAME(n_drawoffset_x)); save_item(NAME(n_drawoffset_y)); save_item(NAME(m_n_displaystartx)); save_item(NAME(n_displaystarty)); save_item(NAME(n_gpustatus)); save_item(NAME(n_gpuinfo)); save_item(NAME(n_lightgun_x)); save_item(NAME(n_lightgun_y)); save_item(NAME(m_n_tx)); save_item(NAME(m_n_ty)); save_item(NAME(n_abr)); save_item(NAME(n_tp)); save_item(NAME(n_ix)); save_item(NAME(n_iy)); save_item(NAME(n_ti)); save_item(NAME(m_draw_stp)); save_item(NAME(m_check_stp)); } void psxgpu_device::device_post_load() { updatevisiblearea(); } uint32_t psxgpu_device::update_screen(screen_device &screen, bitmap_rgb32 &bitmap, const rectangle &cliprect) { uint32_t n_x; uint32_t n_y; int n_top; int n_line; int n_lines; int n_left; int n_column; int n_columns; int n_displaystartx; int n_overscantop; int n_overscanleft; #if PSXGPU_DEBUG_VIEWER if( DebugMeshDisplay( bitmap, cliprect ) ) { return 0; } if( DebugTextureDisplay( bitmap ) ) { return 0; } #endif if( ( n_gpustatus & ( 1 << 0x17 ) ) != 0 ) { /* todo: only draw to necessary area */ bitmap.fill(0, cliprect); } else { if( b_reverseflag ) { n_displaystartx = ( 1023 - m_n_displaystartx ); /* todo: make this flip the screen, in the meantime.. */ n_displaystartx -= ( n_screenwidth - 1 ); } else { n_displaystartx = m_n_displaystartx; } if( ( n_gpustatus & ( 1 << 0x14 ) ) != 0 ) { /* pal */ n_overscantop = 0x23; n_overscanleft = 0x27e; } else { /* ntsc */ n_overscantop = 0x10; n_overscanleft = 0x260; } n_top = (int32_t)n_vert_disstart - n_overscantop; n_lines = (int32_t)n_vert_disend - (int32_t)n_vert_disstart; if( n_top < 0 ) { n_y = -n_top; n_lines += n_top; } else { n_y = 0; /* draw top border */ rectangle clip(cliprect.left(), cliprect.right(), cliprect.top(), n_top); bitmap.fill(0, clip); } if( ( n_gpustatus & ( 1 << 0x16 ) ) != 0 ) { /* interlaced */ n_lines *= 2; } if( n_lines > n_screenheight - ( n_y + n_top ) ) { n_lines = n_screenheight - ( n_y + n_top ); } else { /* draw bottom border */ rectangle clip(cliprect.left(), cliprect.right(), n_y + n_top + n_lines, cliprect.bottom()); bitmap.fill(0, clip); } n_left = ( ( (int32_t)n_horiz_disstart - n_overscanleft ) * (int32_t)n_screenwidth ) / 2560; n_columns = ( ( ( (int32_t)n_horiz_disend - n_horiz_disstart ) * (int32_t)n_screenwidth ) / 2560 ); if( n_left < 0 ) { n_x = -n_left; n_columns += n_left; } else { n_x = 0; /* draw left border */ rectangle clip(cliprect.left(), n_x + n_left, cliprect.top(), cliprect.bottom()); bitmap.fill(0, clip); } if( n_columns > n_screenwidth - ( n_x + n_left ) ) { n_columns = n_screenwidth - ( n_x + n_left ); } else { /* draw right border */ rectangle clip(n_x + n_left + n_columns, cliprect.right(), cliprect.top(), cliprect.bottom()); bitmap.fill(0, clip); } if( ( n_gpustatus & ( 1 << 0x15 ) ) != 0 ) { /* 24bit */ n_line = n_lines; while( n_line > 0 ) { uint16_t *p_n_src = p_p_vram[ n_y + n_displaystarty ] + 3 * n_x + n_displaystartx; uint32_t *p_n_dest = &bitmap.pix32(n_y + n_top, n_x + n_left); n_column = n_columns; while( n_column > 0 ) { uint32_t n_g0r0 = *( p_n_src++ ); uint32_t n_r1b0 = *( p_n_src++ ); uint32_t n_b1g1 = *( p_n_src++ ); *( p_n_dest++ ) = p_n_g0r0[ n_g0r0 ] | p_n_b0[ n_r1b0 ]; n_column--; if( n_column > 0 ) { *( p_n_dest++ ) = p_n_r1[ n_r1b0 ] | p_n_b1g1[ n_b1g1 ]; n_column--; } } n_y++; n_line--; } } else { /* 15bit */ n_line = n_lines; while( n_line > 0 ) { draw_scanline16( bitmap, n_x + n_left, n_y + n_top, n_columns, p_p_vram[ ( n_y + n_displaystarty ) & 1023 ] + n_x + n_displaystartx, pens() ); n_y++; n_line--; } } } return 0; } #define WRITE_PIXEL( p ) \ { \ if( !m_check_stp || ( *( p_vram ) & 0x8000 ) == 0 ) \ { \ if( m_draw_stp ) \ *( p_vram ) = ( p ) | 0x8000; \ else \ *( p_vram ) = p; \ } \ } /* type 1 f e| d| c b| a 9| 8 7| 6 5| 4| 3 2 1 0 |ti| | tp| abr| ty| | tx */ /* type 2 f e| d c| b| a 9| 8 7| 6 5| 4| 3 2 1 0 |iy|ix|ty| | tp| abr|ty| tx */ void psxgpu_device::decode_tpage( uint32_t tpage ) { if( m_n_gputype == 2 ) { n_gpustatus = ( n_gpustatus & 0xffff7800 ) | ( tpage & 0x7ff ) | ( ( tpage & 0x800 ) << 4 ); m_n_tx = ( tpage & 0x0f ) << 6; m_n_ty = ( ( tpage & 0x10 ) << 4 ) | ( ( tpage & 0x800 ) >> 2 ); n_abr = ( tpage & 0x60 ) >> 5; n_tp = ( tpage & 0x180 ) >> 7; n_ix = ( tpage & 0x1000 ) >> 12; n_iy = ( tpage & 0x2000 ) >> 13; n_ti = 0; if( ( tpage & ~0x39ff ) != 0 ) { verboselog( *this, 1, "not handled: draw mode %08x\n", tpage & ~0x39ff ); } if( n_tp == 3 ) { verboselog( *this, 0, "not handled: tp == 3\n" ); } } else { // TODO: confirm status bits on real type 1 gpu n_gpustatus = ( n_gpustatus & 0xffffe000 ) | ( tpage & 0x1fff ); m_n_tx = ( tpage & 0x0f ) << 6; m_n_ty = ( ( tpage & 0x60 ) << 3 ); n_abr = ( tpage & 0x180 ) >> 7; n_tp = ( tpage & 0x600 ) >> 9; n_ti = ( tpage & 0x2000 ) >> 13; n_ix = 0; n_iy = 0; if( ( tpage & ~0x27ef ) != 0 ) { verboselog( *this, 1, "not handled: draw mode %08x\n", tpage & ~0x27ef ); } if( n_tp == 3 ) { verboselog( *this, 0, "not handled: tp == 3\n" ); } else if( n_tp == 2 && n_ti != 0 ) { verboselog( *this, 0, "not handled: interleaved 15 bit texture\n" ); } } } #define SPRITESETUP \ int n_dv; \ if( n_iy != 0 ) \ { \ n_dv = -1; \ } \ else \ { \ n_dv = 1; \ } \ int n_du; \ if( n_ix != 0 ) \ { \ n_du = -1; \ } \ else \ { \ n_du = 1; \ } #define TRANSPARENCYSETUP \ uint16_t *p_n_f = p_n_f1; \ uint16_t *p_n_redb = p_n_redb1; \ uint16_t *p_n_greenb = p_n_greenb1; \ uint16_t *p_n_blueb = p_n_blueb1; \ uint16_t *p_n_redtrans = p_n_redaddtrans; \ uint16_t *p_n_greentrans = p_n_greenaddtrans; \ uint16_t *p_n_bluetrans = p_n_blueaddtrans; \ \ switch( n_cmd & 0x02 ) \ { \ case 0x02: \ switch( n_abr ) \ { \ case 0x00: \ p_n_f = p_n_f05; \ p_n_redb = p_n_redb05; \ p_n_greenb = p_n_greenb05; \ p_n_blueb = p_n_blueb05; \ p_n_redtrans = p_n_redaddtrans; \ p_n_greentrans = p_n_greenaddtrans; \ p_n_bluetrans = p_n_blueaddtrans; \ verboselog( *this, 2, "Transparency Mode: 0.5*B + 0.5*F\n" ); \ break; \ case 0x01: \ p_n_f = p_n_f1; \ p_n_redb = p_n_redb1; \ p_n_greenb = p_n_greenb1; \ p_n_blueb = p_n_blueb1; \ p_n_redtrans = p_n_redaddtrans; \ p_n_greentrans = p_n_greenaddtrans; \ p_n_bluetrans = p_n_blueaddtrans; \ verboselog( *this, 2, "Transparency Mode: 1.0*B + 1.0*F\n" ); \ break; \ case 0x02: \ p_n_f = p_n_f1; \ p_n_redb = p_n_redb1; \ p_n_greenb = p_n_greenb1; \ p_n_blueb = p_n_blueb1; \ p_n_redtrans = p_n_redsubtrans; \ p_n_greentrans = p_n_greensubtrans; \ p_n_bluetrans = p_n_bluesubtrans; \ verboselog( *this, 2, "Transparency Mode: 1.0*B - 1.0*F\n" ); \ break; \ case 0x03: \ p_n_f = p_n_f025; \ p_n_redb = p_n_redb1; \ p_n_greenb = p_n_greenb1; \ p_n_blueb = p_n_blueb1; \ p_n_redtrans = p_n_redaddtrans; \ p_n_greentrans = p_n_greenaddtrans; \ p_n_bluetrans = p_n_blueaddtrans; \ verboselog( *this, 2, "Transparency Mode: 1.0*B + 0.25*F\n" ); \ break; \ } \ break; \ } #define SOLIDSETUP \ TRANSPARENCYSETUP #define TEXTURESETUP \ int n_tx = m_n_tx; \ int n_ty = m_n_ty; \ uint16_t *p_clut = p_p_vram[ n_cluty ] + n_clutx; \ switch( n_tp ) \ { \ case 0: \ n_tx += n_twx >> 2; \ n_ty += n_twy; \ break; \ case 1: \ n_tx += n_twx >> 1; \ n_ty += n_twy; \ break; \ case 2: \ n_tx += n_twx >> 0; \ n_ty += n_twy; \ break; \ } \ TRANSPARENCYSETUP #define FLATPOLYGONUPDATE #define FLATRECTANGEUPDATE #define GOURAUDPOLYGONUPDATE \ n_r.d += n_dr; \ n_g.d += n_dg; \ n_b.d += n_db; #define SOLIDFILL( PIXELUPDATE ) \ if( n_distance > ( (int32_t)n_drawarea_x2 - drawx ) + 1 ) \ { \ n_distance = ( n_drawarea_x2 - drawx ) + 1; \ } \ uint16_t *p_vram = p_p_vram[ drawy ] + drawx; \ \ switch( n_cmd & 0x02 ) \ { \ case 0x00: \ /* transparency off */ \ while( n_distance > 0 ) \ { \ WRITE_PIXEL( \ p_n_redshade[ MID_LEVEL | n_r.w.h ] | \ p_n_greenshade[ MID_LEVEL | n_g.w.h ] | \ p_n_blueshade[ MID_LEVEL | n_b.w.h ] ) \ p_vram++; \ PIXELUPDATE \ n_distance--; \ } \ break; \ case 0x02: \ /* transparency on */ \ while( n_distance > 0 ) \ { \ WRITE_PIXEL( \ p_n_redtrans[ p_n_f[ MID_LEVEL | n_r.w.h ] | p_n_redb[ *( p_vram ) ] ] | \ p_n_greentrans[ p_n_f[ MID_LEVEL | n_g.w.h ] | p_n_greenb[ *( p_vram ) ] ] | \ p_n_bluetrans[ p_n_f[ MID_LEVEL | n_b.w.h ] | p_n_blueb[ *( p_vram ) ] ] ) \ p_vram++; \ PIXELUPDATE \ n_distance--; \ } \ break; \ } #define FLATTEXTUREDPOLYGONUPDATE \ n_u.d += n_du; \ n_v.d += n_dv; #define GOURAUDTEXTUREDPOLYGONUPDATE \ n_r.d += n_dr; \ n_g.d += n_dg; \ n_b.d += n_db; \ n_u.d += n_du; \ n_v.d += n_dv; #define FLATTEXTUREDRECTANGLEUPDATE \ n_u += n_du; #define TEXTURE_LOOP \ while( n_distance > 0 ) \ { #define TEXTURE_ENDLOOP \ } #define TEXTURE4BIT( TXV, TXU ) \ TEXTURE_LOOP \ uint16_t n_bgr = p_clut[ ( *( p_p_vram[ n_ty + TXV ] + n_tx + ( TXU >> 2 ) ) >> ( ( TXU & 0x03 ) << 2 ) ) & 0x0f ]; #define TEXTURE8BIT( TXV, TXU ) \ TEXTURE_LOOP \ uint16_t n_bgr = p_clut[ ( *( p_p_vram[ n_ty + TXV ] + n_tx + ( TXU >> 1 ) ) >> ( ( TXU & 0x01 ) << 3 ) ) & 0xff ]; #define TEXTURE15BIT( TXV, TXU ) \ TEXTURE_LOOP \ uint16_t n_bgr = *( p_p_vram[ n_ty + TXV ] + n_tx + TXU ); #define TEXTUREWINDOW4BIT( TXV, TXU ) TEXTURE4BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define TEXTUREWINDOW8BIT( TXV, TXU ) TEXTURE8BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define TEXTUREWINDOW15BIT( TXV, TXU ) TEXTURE15BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define TEXTUREINTERLEAVED4BIT( TXV, TXU ) \ TEXTURE_LOOP \ int n_xi = ( ( TXU >> 2 ) & ~0x3c ) + ( ( TXV << 2 ) & 0x3c ); \ int n_yi = ( TXV & ~0xf ) + ( ( TXU >> 4 ) & 0xf ); \ uint16_t n_bgr = p_clut[ ( *( p_p_vram[ n_ty + n_yi ] + n_tx + n_xi ) >> ( ( TXU & 0x03 ) << 2 ) ) & 0x0f ]; #define TEXTUREINTERLEAVED8BIT( TXV, TXU ) \ TEXTURE_LOOP \ int n_xi = ( ( TXU >> 1 ) & ~0x78 ) + ( ( TXU << 2 ) & 0x40 ) + ( ( TXV << 3 ) & 0x38 ); \ int n_yi = ( TXV & ~0x7 ) + ( ( TXU >> 5 ) & 0x7 ); \ uint16_t n_bgr = p_clut[ ( *( p_p_vram[ n_ty + n_yi ] + n_tx + n_xi ) >> ( ( TXU & 0x01 ) << 3 ) ) & 0xff ]; #define TEXTUREINTERLEAVED15BIT( TXV, TXU ) \ TEXTURE_LOOP \ int n_xi = TXU; \ int n_yi = TXV; \ uint16_t n_bgr = *( p_p_vram[ n_ty + n_yi ] + n_tx + n_xi ); #define TEXTUREWINDOWINTERLEAVED4BIT( TXV, TXU ) TEXTUREINTERLEAVED4BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define TEXTUREWINDOWINTERLEAVED8BIT( TXV, TXU ) TEXTUREINTERLEAVED8BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define TEXTUREWINDOWINTERLEAVED15BIT( TXV, TXU ) TEXTUREINTERLEAVED15BIT( ( TXV & n_twh ), ( TXU & n_tww ) ) #define SHADEDPIXEL( PIXELUPDATE ) \ if( n_bgr != 0 ) \ { \ WRITE_PIXEL( \ p_n_redshade[ p_n_redlevel[ n_bgr ] | n_r.w.h ] | \ p_n_greenshade[ p_n_greenlevel[ n_bgr ] | n_g.w.h ] | \ p_n_blueshade[ p_n_bluelevel[ n_bgr ] | n_b.w.h ] | \ ( n_bgr & 0x8000 ) ) \ } \ p_vram++; \ PIXELUPDATE \ n_distance--; \ TEXTURE_ENDLOOP #define TRANSPARENTPIXEL( PIXELUPDATE ) \ if( n_bgr != 0 ) \ { \ if( ( n_bgr & 0x8000 ) != 0 ) \ { \ WRITE_PIXEL( \ p_n_redtrans[ p_n_f[ p_n_redlevel[ n_bgr ] | n_r.w.h ] | p_n_redb[ *( p_vram ) ] ] | \ p_n_greentrans[ p_n_f[ p_n_greenlevel[ n_bgr ] | n_g.w.h ] | p_n_greenb[ *( p_vram ) ] ] | \ p_n_bluetrans[ p_n_f[ p_n_bluelevel[ n_bgr ] | n_b.w.h ] | p_n_blueb[ *( p_vram ) ] ] | \ 0x8000 ) \ } \ else \ { \ WRITE_PIXEL( \ p_n_redshade[ p_n_redlevel[ n_bgr ] | n_r.w.h ] | \ p_n_greenshade[ p_n_greenlevel[ n_bgr ] | n_g.w.h ] | \ p_n_blueshade[ p_n_bluelevel[ n_bgr ] | n_b.w.h ] ) \ } \ } \ p_vram++; \ PIXELUPDATE \ n_distance--; \ TEXTURE_ENDLOOP #define TEXTUREFILL( PIXELUPDATE, TXU, TXV ) \ if( n_distance > ( (int32_t)n_drawarea_x2 - drawx ) + 1 ) \ { \ n_distance = ( n_drawarea_x2 - drawx ) + 1; \ } \ uint16_t *p_vram = p_p_vram[ drawy ] + drawx; \ \ if( n_ti != 0 ) \ { \ /* interleaved texture */ \ if( n_twh != 255 || \ n_tww != 255 || \ n_twx != 0 || \ n_twy != 0 ) \ { \ /* texture window */ \ switch( n_cmd & 0x02 ) \ { \ case 0x00: \ /* shading */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREWINDOWINTERLEAVED4BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREWINDOWINTERLEAVED8BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREWINDOWINTERLEAVED15BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ case 0x02: \ /* semi transparency */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREWINDOWINTERLEAVED4BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREWINDOWINTERLEAVED8BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREWINDOWINTERLEAVED15BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ } \ } \ else \ { \ /* no texture window */ \ switch( n_cmd & 0x02 ) \ { \ case 0x00: \ /* shading */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREINTERLEAVED4BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREINTERLEAVED8BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREINTERLEAVED15BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ case 0x02: \ /* semi transparency */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREINTERLEAVED4BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREINTERLEAVED8BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREINTERLEAVED15BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ } \ } \ } \ else \ { \ /* standard texture */ \ if( n_twh != 255 || \ n_tww != 255 || \ n_twx != 0 || \ n_twy != 0 ) \ { \ /* texture window */ \ switch( n_cmd & 0x02 ) \ { \ case 0x00: \ /* shading */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREWINDOW4BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREWINDOW8BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREWINDOW15BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ case 0x02: \ /* semi transparency */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTUREWINDOW4BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTUREWINDOW8BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTUREWINDOW15BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ } \ } \ else \ { \ /* no texture window */ \ switch( n_cmd & 0x02 ) \ { \ case 0x00: \ /* shading */ \ switch( n_tp ) \ { \ case 0: \ TEXTURE4BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTURE8BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTURE15BIT( TXV, TXU ) \ SHADEDPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ case 0x02: \ /* semi transparency */ \ switch( n_tp ) \ { \ case 0: \ /* 4 bit clut */ \ TEXTURE4BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 1: \ /* 8 bit clut */ \ TEXTURE8BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ case 2: \ /* 15 bit */ \ TEXTURE15BIT( TXV, TXU ) \ TRANSPARENTPIXEL( PIXELUPDATE ) \ break; \ } \ break; \ } \ } \ } #define GET_COORD( a ) \ a.sw.l = SINT11( COORD_X( a ) ); \ a.sw.h = SINT11( COORD_Y( a ) ); static inline int CullVertex( int a, int b ) { int d = a - b; if( d < -1023 || d > 1023 ) { return 1; } return 0; } #define CULLPOINT( PacketType, p1, p2 ) \ ( \ CullVertex( COORD_Y( m_packet.PacketType.vertex[ p1 ].n_coord ), COORD_Y( m_packet.PacketType.vertex[ p2 ].n_coord ) ) || \ CullVertex( COORD_X( m_packet.PacketType.vertex[ p1 ].n_coord ), COORD_X( m_packet.PacketType.vertex[ p2 ].n_coord ) ) \ ) #define CULLTRIANGLE( PacketType, start ) \ ( \ CULLPOINT( PacketType, start, start + 1 ) || CULLPOINT( PacketType, start + 1, start + 2 ) || CULLPOINT( PacketType, start + 2, start ) \ ) #define FINDTOPLEFT( PacketType ) \ for( int n_point = 0; n_point < n_points; n_point++ ) \ { \ GET_COORD( m_packet.PacketType.vertex[ n_point ].n_coord ); \ } \ \ const int *p_n_rightpointlist; \ const int *p_n_leftpointlist; \ int n_leftpoint = 0; \ if( n_points == 4 ) \ { \ if( CULLTRIANGLE( PacketType, 0 ) ) \ { \ if( CULLTRIANGLE( PacketType, 1 ) ) \ { \ return; \ } \ \ p_n_rightpointlist = m_p_n_nextpointlist4b; \ p_n_leftpointlist = m_p_n_prevpointlist4b; \ n_leftpoint++; \ } \ else if( CULLTRIANGLE( PacketType, 1 ) ) \ { \ p_n_rightpointlist = m_p_n_nextpointlist3; \ p_n_leftpointlist = m_p_n_prevpointlist3; \ n_points--; \ } \ else \ { \ p_n_rightpointlist = m_p_n_nextpointlist4; \ p_n_leftpointlist = m_p_n_prevpointlist4; \ } \ } \ else if( CULLTRIANGLE( PacketType, 0 ) ) \ { \ return; \ } \ else \ { \ p_n_rightpointlist = m_p_n_nextpointlist3; \ p_n_leftpointlist = m_p_n_prevpointlist3; \ } \ \ for( int n_point = n_leftpoint + 1; n_point < n_points; n_point++ ) \ { \ if( COORD_Y( m_packet.PacketType.vertex[ n_point ].n_coord ) < COORD_Y( m_packet.PacketType.vertex[ n_leftpoint ].n_coord ) || \ ( COORD_Y( m_packet.PacketType.vertex[ n_point ].n_coord ) == COORD_Y( m_packet.PacketType.vertex[ n_leftpoint ].n_coord ) && \ COORD_X( m_packet.PacketType.vertex[ n_point ].n_coord ) < COORD_X( m_packet.PacketType.vertex[ n_leftpoint ].n_coord ) ) ) \ { \ n_leftpoint = n_point; \ } \ } \ int n_rightpoint = n_leftpoint; void psxgpu_device::FlatPolygon( int n_points ) { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 1 ) { return; } for( int n_point = 0; n_point < n_points; n_point++ ) { DebugMesh( SINT11( COORD_X( m_packet.FlatPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_y ); } DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatPolygon.n_bgr ); PAIR n_cx1; n_cx1.d = 0; PAIR n_cx2; n_cx2.d = 0; SOLIDSETUP PAIR n_r; n_r.w.h = BGR_R( m_packet.FlatPolygon.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = BGR_G( m_packet.FlatPolygon.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = BGR_B( m_packet.FlatPolygon.n_bgr ); n_b.w.l = 0; FINDTOPLEFT( FlatPolygon ) int32_t n_dx1 = 0; int32_t n_dx2 = 0; int16_t n_y = COORD_Y( m_packet.FlatPolygon.vertex[ n_rightpoint ].n_coord ); for( ;; ) { if( n_y == COORD_Y( m_packet.FlatPolygon.vertex[ n_leftpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.FlatPolygon.vertex[ p_n_leftpointlist[ n_leftpoint ] ].n_coord ) ) { n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; if( n_leftpoint == n_rightpoint ) { break; } } n_cx1.sw.h = COORD_X( m_packet.FlatPolygon.vertex[ n_leftpoint ].n_coord ); n_cx1.sw.l = 0; n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; int32_t n_distance = COORD_Y( m_packet.FlatPolygon.vertex[ n_leftpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx1 = (int32_t)( ( COORD_X( m_packet.FlatPolygon.vertex[ n_leftpoint ].n_coord ) << 16 ) - n_cx1.d ) / n_distance; } if( n_y == COORD_Y( m_packet.FlatPolygon.vertex[ n_rightpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.FlatPolygon.vertex[ p_n_rightpointlist[ n_rightpoint ] ].n_coord ) ) { n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; if( n_rightpoint == n_leftpoint ) { break; } } n_cx2.sw.h = COORD_X( m_packet.FlatPolygon.vertex[ n_rightpoint ].n_coord ); n_cx2.sw.l = 0; n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; int32_t n_distance = COORD_Y( m_packet.FlatPolygon.vertex[ n_rightpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx2 = (int32_t)( ( COORD_X( m_packet.FlatPolygon.vertex[ n_rightpoint ].n_coord ) << 16 ) - n_cx2.d ) / n_distance; } int drawy = n_y + n_drawoffset_y; if( (int16_t)n_cx1.sw.h != (int16_t)n_cx2.sw.h && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int16_t n_x; int32_t n_distance; if( (int16_t)n_cx1.sw.h < (int16_t)n_cx2.sw.h ) { n_x = n_cx1.sw.h; n_distance = (int16_t)n_cx2.sw.h - n_x; } else { n_x = n_cx2.sw.h; n_distance = (int16_t)n_cx1.sw.h - n_x; } int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } SOLIDFILL( FLATPOLYGONUPDATE ) } n_cx1.d += n_dx1; n_cx2.d += n_dx2; n_y++; } } void psxgpu_device::FlatTexturedPolygon( int n_points ) { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 2 ) { return; } for( int n_point = 0; n_point < n_points; n_point++ ) { DebugMesh( SINT11( COORD_X( m_packet.FlatTexturedPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_y ); } DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatTexturedPolygon.n_bgr ); uint32_t n_clutx = ( m_packet.FlatTexturedPolygon.vertex[ 0 ].n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.FlatTexturedPolygon.vertex[ 0 ].n_texture.w.h >> 6 ) & 0x3ff; PAIR n_cx1; n_cx1.d = 0; PAIR n_cu1; n_cu1.d = 0; PAIR n_cv1; n_cv1.d = 0; PAIR n_cx2; n_cx2.d = 0; PAIR n_cu2; n_cu2.d = 0; PAIR n_cv2; n_cv2.d = 0; decode_tpage( m_packet.FlatTexturedPolygon.vertex[ 1 ].n_texture.w.h ); TEXTURESETUP PAIR n_r; n_r.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.FlatTexturedPolygon.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.FlatTexturedPolygon.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.FlatTexturedPolygon.n_bgr ); n_b.w.l = 0; FINDTOPLEFT( FlatTexturedPolygon ) int32_t n_dx1 = 0; int32_t n_dx2 = 0; int32_t n_du1 = 0; int32_t n_du2 = 0; int32_t n_dv1 = 0; int32_t n_dv2 = 0; int16_t n_y = COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_coord ); for( ;; ) { if( n_y == COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.FlatTexturedPolygon.vertex[ p_n_leftpointlist[ n_leftpoint ] ].n_coord ) ) { n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; if( n_leftpoint == n_rightpoint ) { break; } } n_cx1.sw.h = COORD_X( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_coord ); n_cx1.sw.l = 0; n_cu1.w.h = TEXTURE_U( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_texture ); n_cu1.w.l = 0; n_cv1.w.h = TEXTURE_V( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_texture ); n_cv1.w.l = 0; n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; int32_t n_distance = COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx1 = (int32_t)( ( COORD_X( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_coord ) << 16 ) - n_cx1.d ) / n_distance; n_du1 = (int32_t)( ( TEXTURE_U( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_texture ) << 16 ) - n_cu1.d ) / n_distance; n_dv1 = (int32_t)( ( TEXTURE_V( m_packet.FlatTexturedPolygon.vertex[ n_leftpoint ].n_texture ) << 16 ) - n_cv1.d ) / n_distance; } if( n_y == COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.FlatTexturedPolygon.vertex[ p_n_rightpointlist[ n_rightpoint ] ].n_coord ) ) { n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; if( n_rightpoint == n_leftpoint ) { break; } } n_cx2.sw.h = COORD_X( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_coord ); n_cx2.sw.l = 0; n_cu2.w.h = TEXTURE_U( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_texture ); n_cu2.w.l = 0; n_cv2.w.h = TEXTURE_V( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_texture ); n_cv2.w.l = 0; n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; int32_t n_distance = COORD_Y( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx2 = (int32_t)( ( COORD_X( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_coord ) << 16 ) - n_cx2.d ) / n_distance; n_du2 = (int32_t)( ( TEXTURE_U( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_texture ) << 16 ) - n_cu2.d ) / n_distance; n_dv2 = (int32_t)( ( TEXTURE_V( m_packet.FlatTexturedPolygon.vertex[ n_rightpoint ].n_texture ) << 16 ) - n_cv2.d ) / n_distance; } int drawy = n_y + n_drawoffset_y; if( (int16_t)n_cx1.sw.h != (int16_t)n_cx2.sw.h && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int16_t n_x; int32_t n_distance; PAIR n_u; PAIR n_v; int32_t n_du; int32_t n_dv; if( (int16_t)n_cx1.sw.h < (int16_t)n_cx2.sw.h ) { n_x = n_cx1.sw.h; n_distance = (int16_t)n_cx2.sw.h - n_x; n_u.d = n_cu1.d; n_v.d = n_cv1.d; n_du = (int32_t)( n_cu2.d - n_cu1.d ) / n_distance; n_dv = (int32_t)( n_cv2.d - n_cv1.d ) / n_distance; } else { n_x = n_cx2.sw.h; n_distance = (int16_t)n_cx1.sw.h - n_x; n_u.d = n_cu2.d; n_v.d = n_cv2.d; n_du = (int32_t)( n_cu1.d - n_cu2.d ) / n_distance; n_dv = (int32_t)( n_cv1.d - n_cv2.d ) / n_distance; } int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_u.d += n_du * ( n_drawarea_x1 - drawx ); n_v.d += n_dv * ( n_drawarea_x1 - drawx ); n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } TEXTUREFILL( FLATTEXTUREDPOLYGONUPDATE, n_u.w.h, n_v.w.h ); } n_cx1.d += n_dx1; n_cu1.d += n_du1; n_cv1.d += n_dv1; n_cx2.d += n_dx2; n_cu2.d += n_du2; n_cv2.d += n_dv2; n_y++; } } void psxgpu_device::GouraudPolygon( int n_points ) { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 3 ) { return; } for( int n_point = 0; n_point < n_points; n_point++ ) { DebugMesh( SINT11( COORD_X( m_packet.GouraudPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.GouraudPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_y ); } DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.GouraudPolygon.vertex[ 0 ].n_bgr ); PAIR n_cx1; n_cx1.d = 0; PAIR n_cr1; n_cr1.d = 0; PAIR n_cg1; n_cg1.d = 0; PAIR n_cb1; n_cb1.d = 0; PAIR n_cx2; n_cx2.d = 0; PAIR n_cr2; n_cr2.d = 0; PAIR n_cg2; n_cg2.d = 0; PAIR n_cb2; n_cb2.d = 0; SOLIDSETUP FINDTOPLEFT( GouraudPolygon ) int32_t n_dx1 = 0; int32_t n_dx2 = 0; int32_t n_dr1 = 0; int32_t n_dr2 = 0; int32_t n_dg1 = 0; int32_t n_dg2 = 0; int32_t n_db1 = 0; int32_t n_db2 = 0; int16_t n_y = COORD_Y( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_coord ); for( ;; ) { if( n_y == COORD_Y( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.GouraudPolygon.vertex[ p_n_leftpointlist[ n_leftpoint ] ].n_coord ) ) { n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; if( n_leftpoint == n_rightpoint ) { break; } } n_cx1.sw.h = COORD_X( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_coord ); n_cx1.sw.l = 0; n_cr1.w.h = BGR_R( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ); n_cr1.w.l = 0; n_cg1.w.h = BGR_G( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ); n_cg1.w.l = 0; n_cb1.w.h = BGR_B( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ); n_cb1.w.l = 0; n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; int32_t n_distance = COORD_Y( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx1 = (int32_t)( ( COORD_X( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_coord ) << 16 ) - n_cx1.d ) / n_distance; n_dr1 = (int32_t)( ( BGR_R( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cr1.d ) / n_distance; n_dg1 = (int32_t)( ( BGR_G( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cg1.d ) / n_distance; n_db1 = (int32_t)( ( BGR_B( m_packet.GouraudPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cb1.d ) / n_distance; } if( n_y == COORD_Y( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.GouraudPolygon.vertex[ p_n_rightpointlist[ n_rightpoint ] ].n_coord ) ) { n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; if( n_rightpoint == n_leftpoint ) { break; } } n_cx2.sw.h = COORD_X( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_coord ); n_cx2.sw.l = 0; n_cr2.w.h = BGR_R( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ); n_cr2.w.l = 0; n_cg2.w.h = BGR_G( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ); n_cg2.w.l = 0; n_cb2.w.h = BGR_B( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ); n_cb2.w.l = 0; n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; int32_t n_distance = COORD_Y( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx2 = (int32_t)( ( COORD_X( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_coord ) << 16 ) - n_cx2.d ) / n_distance; n_dr2 = (int32_t)( ( BGR_R( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cr2.d ) / n_distance; n_dg2 = (int32_t)( ( BGR_G( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cg2.d ) / n_distance; n_db2 = (int32_t)( ( BGR_B( m_packet.GouraudPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cb2.d ) / n_distance; } int drawy = n_y + n_drawoffset_y; if( (int16_t)n_cx1.sw.h != (int16_t)n_cx2.sw.h && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int16_t n_x; int32_t n_distance; PAIR n_r; PAIR n_g; PAIR n_b; int32_t n_dr; int32_t n_dg; int32_t n_db; if( (int16_t)n_cx1.sw.h < (int16_t)n_cx2.sw.h ) { n_x = n_cx1.sw.h; n_distance = (int16_t)n_cx2.sw.h - n_x; n_r.d = n_cr1.d; n_g.d = n_cg1.d; n_b.d = n_cb1.d; n_dr = (int32_t)( n_cr2.d - n_cr1.d ) / n_distance; n_dg = (int32_t)( n_cg2.d - n_cg1.d ) / n_distance; n_db = (int32_t)( n_cb2.d - n_cb1.d ) / n_distance; } else { n_x = n_cx2.sw.h; n_distance = (int16_t)n_cx1.sw.h - n_x; n_r.d = n_cr2.d; n_g.d = n_cg2.d; n_b.d = n_cb2.d; n_dr = (int32_t)( n_cr1.d - n_cr2.d ) / n_distance; n_dg = (int32_t)( n_cg1.d - n_cg2.d ) / n_distance; n_db = (int32_t)( n_cb1.d - n_cb2.d ) / n_distance; } int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_r.d += n_dr * ( n_drawarea_x1 - drawx ); n_g.d += n_dg * ( n_drawarea_x1 - drawx ); n_b.d += n_db * ( n_drawarea_x1 - drawx ); n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } SOLIDFILL( GOURAUDPOLYGONUPDATE ) } n_cx1.d += n_dx1; n_cr1.d += n_dr1; n_cg1.d += n_dg1; n_cb1.d += n_db1; n_cx2.d += n_dx2; n_cr2.d += n_dr2; n_cg2.d += n_dg2; n_cb2.d += n_db2; n_y++; } } void psxgpu_device::GouraudTexturedPolygon( int n_points ) { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 4 ) { return; } for( int n_point = 0; n_point < n_points; n_point++ ) { DebugMesh( SINT11( COORD_X( m_packet.GouraudTexturedPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_point ].n_coord ) ) + n_drawoffset_y ); } DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.GouraudTexturedPolygon.vertex[ 0 ].n_bgr ); uint32_t n_clutx = ( m_packet.GouraudTexturedPolygon.vertex[ 0 ].n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.GouraudTexturedPolygon.vertex[ 0 ].n_texture.w.h >> 6 ) & 0x3ff; PAIR n_cx1; n_cx1.d = 0; PAIR n_cr1; n_cr1.d = 0; PAIR n_cg1; n_cg1.d = 0; PAIR n_cb1; n_cb1.d = 0; PAIR n_cu1; n_cu1.d = 0; PAIR n_cv1; n_cv1.d = 0; PAIR n_cx2; n_cx2.d = 0; PAIR n_cr2; n_cr2.d = 0; PAIR n_cg2; n_cg2.d = 0; PAIR n_cb2; n_cb2.d = 0; PAIR n_cu2; n_cu2.d = 0; PAIR n_cv2; n_cv2.d = 0; decode_tpage( m_packet.GouraudTexturedPolygon.vertex[ 1 ].n_texture.w.h ); TEXTURESETUP FINDTOPLEFT( GouraudTexturedPolygon ) int32_t n_dx1 = 0; int32_t n_dx2 = 0; int32_t n_du1 = 0; int32_t n_du2 = 0; int32_t n_dr1 = 0; int32_t n_dr2 = 0; int32_t n_dg1 = 0; int32_t n_dg2 = 0; int32_t n_db1 = 0; int32_t n_db2 = 0; int32_t n_dv1 = 0; int32_t n_dv2 = 0; int16_t n_y = COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_coord ); for( ;; ) { if( n_y == COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ p_n_leftpointlist[ n_leftpoint ] ].n_coord ) ) { n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; if( n_leftpoint == n_rightpoint ) { break; } } n_cx1.sw.h = COORD_X( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_coord ); n_cx1.sw.l = 0; n_cr1.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ); n_cr1.w.l = 0; n_cg1.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ); n_cg1.w.l = 0; n_cb1.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ); n_cb1.w.l = 0; n_cu1.w.h = TEXTURE_U( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_texture ); n_cu1.w.l = 0; n_cv1.w.h = TEXTURE_V( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_texture ); n_cv1.w.l = 0; n_leftpoint = p_n_leftpointlist[ n_leftpoint ]; int32_t n_distance = COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx1 = (int32_t)( ( COORD_X( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_coord ) << 16 ) - n_cx1.d ) / n_distance; n_dr1 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_R( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cr1.d ) / n_distance; n_dg1 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_G( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cg1.d ) / n_distance; n_db1 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_B( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_bgr ) << 16 ) - n_cb1.d ) / n_distance; n_du1 = (int32_t)( ( TEXTURE_U( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_texture ) << 16 ) - n_cu1.d ) / n_distance; n_dv1 = (int32_t)( ( TEXTURE_V( m_packet.GouraudTexturedPolygon.vertex[ n_leftpoint ].n_texture ) << 16 ) - n_cv1.d ) / n_distance; } if( n_y == COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_coord ) ) { while( n_y == COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ p_n_rightpointlist[ n_rightpoint ] ].n_coord ) ) { n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; if( n_rightpoint == n_leftpoint ) { break; } } n_cx2.sw.h = COORD_X( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_coord ); n_cx2.sw.l = 0; n_cr2.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ); n_cr2.w.l = 0; n_cg2.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ); n_cg2.w.l = 0; n_cb2.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ); n_cb2.w.l = 0; n_cu2.w.h = TEXTURE_U( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_texture ); n_cu2.w.l = 0; n_cv2.w.h = TEXTURE_V( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_texture ); n_cv2.w.l = 0; n_rightpoint = p_n_rightpointlist[ n_rightpoint ]; int32_t n_distance = COORD_Y( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_coord ) - n_y; if( n_distance < 1 ) { break; } n_dx2 = (int32_t)( ( COORD_X( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_coord ) << 16 ) - n_cx2.d ) / n_distance; n_dr2 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_R( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cr2.d ) / n_distance; n_dg2 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_G( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cg2.d ) / n_distance; n_db2 = n_cmd & 0x01 ? 0 : (int32_t)( ( BGR_B( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_bgr ) << 16 ) - n_cb2.d ) / n_distance; n_du2 = (int32_t)( ( TEXTURE_U( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_texture ) << 16 ) - n_cu2.d ) / n_distance; n_dv2 = (int32_t)( ( TEXTURE_V( m_packet.GouraudTexturedPolygon.vertex[ n_rightpoint ].n_texture ) << 16 ) - n_cv2.d ) / n_distance; } int drawy = n_y + n_drawoffset_y; if( (int16_t)n_cx1.sw.h != (int16_t)n_cx2.sw.h && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int16_t n_x; int32_t n_distance; PAIR n_r; PAIR n_g; PAIR n_b; PAIR n_u; PAIR n_v; int32_t n_dr; int32_t n_dg; int32_t n_db; int32_t n_du; int32_t n_dv; if( (int16_t)n_cx1.sw.h < (int16_t)n_cx2.sw.h ) { n_x = n_cx1.sw.h; n_distance = (int16_t)n_cx2.sw.h - n_x; n_r.d = n_cr1.d; n_g.d = n_cg1.d; n_b.d = n_cb1.d; n_u.d = n_cu1.d; n_v.d = n_cv1.d; n_dr = (int32_t)( n_cr2.d - n_cr1.d ) / n_distance; n_dg = (int32_t)( n_cg2.d - n_cg1.d ) / n_distance; n_db = (int32_t)( n_cb2.d - n_cb1.d ) / n_distance; n_du = (int32_t)( n_cu2.d - n_cu1.d ) / n_distance; n_dv = (int32_t)( n_cv2.d - n_cv1.d ) / n_distance; } else { n_x = n_cx2.sw.h; n_distance = (int16_t)n_cx1.sw.h - n_x; n_r.d = n_cr2.d; n_g.d = n_cg2.d; n_b.d = n_cb2.d; n_u.d = n_cu2.d; n_v.d = n_cv2.d; n_dr = (int32_t)( n_cr1.d - n_cr2.d ) / n_distance; n_dg = (int32_t)( n_cg1.d - n_cg2.d ) / n_distance; n_db = (int32_t)( n_cb1.d - n_cb2.d ) / n_distance; n_du = (int32_t)( n_cu1.d - n_cu2.d ) / n_distance; n_dv = (int32_t)( n_cv1.d - n_cv2.d ) / n_distance; } int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_r.d += n_dr * ( n_drawarea_x1 - drawx ); n_g.d += n_dg * ( n_drawarea_x1 - drawx ); n_b.d += n_db * ( n_drawarea_x1 - drawx ); n_u.d += n_du * ( n_drawarea_x1 - drawx ); n_v.d += n_dv * ( n_drawarea_x1 - drawx ); n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } TEXTUREFILL( GOURAUDTEXTUREDPOLYGONUPDATE, n_u.w.h, n_v.w.h ); } n_cx1.d += n_dx1; n_cr1.d += n_dr1; n_cg1.d += n_dg1; n_cb1.d += n_db1; n_cu1.d += n_du1; n_cv1.d += n_dv1; n_cx2.d += n_dx2; n_cr2.d += n_dr2; n_cg2.d += n_dg2; n_cb2.d += n_db2; n_cu2.d += n_du2; n_cv2.d += n_dv2; n_y++; } } void psxgpu_device::MonochromeLine() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 5 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.MonochromeLine.vertex[ 0 ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.MonochromeLine.vertex[ 0 ].n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.MonochromeLine.vertex[ 1 ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.MonochromeLine.vertex[ 1 ].n_coord ) ) + n_drawoffset_y ); DebugMeshEnd(); #endif int32_t n_xstart = SINT11( COORD_X( m_packet.MonochromeLine.vertex[ 0 ].n_coord ) ); int32_t n_xend = SINT11( COORD_X( m_packet.MonochromeLine.vertex[ 1 ].n_coord ) ); int32_t n_ystart = SINT11( COORD_Y( m_packet.MonochromeLine.vertex[ 0 ].n_coord ) ); int32_t n_yend = SINT11( COORD_Y( m_packet.MonochromeLine.vertex[ 1 ].n_coord ) ); uint8_t n_cmd = BGR_C( m_packet.MonochromeLine.n_bgr ); uint8_t n_r = BGR_R( m_packet.MonochromeLine.n_bgr ); uint8_t n_g = BGR_G( m_packet.MonochromeLine.n_bgr ); uint8_t n_b = BGR_B( m_packet.MonochromeLine.n_bgr ); TRANSPARENCYSETUP int32_t n_xlen; if( n_xend > n_xstart ) { n_xlen = n_xend - n_xstart; } else { n_xlen = n_xstart - n_xend; } int32_t n_ylen; if( n_yend > n_ystart ) { n_ylen = n_yend - n_ystart; } else { n_ylen = n_ystart - n_yend; } int32_t n_len; if( n_xlen > n_ylen ) { n_len = n_xlen; } else { n_len = n_ylen; } if( n_len == 0 ) { n_len = 1; } PAIR n_x; n_x.sw.h = n_xstart; n_x.sw.l = 0; PAIR n_y; n_y.sw.h = n_ystart; n_y.sw.l = 0; int32_t n_dx = (int32_t)( ( n_xend << 16 ) - n_x.d ) / n_len; int32_t n_dy = (int32_t)( ( n_yend << 16 ) - n_y.d ) / n_len; while( n_len > 0 ) { int drawx = n_x.sw.h + n_drawoffset_x; int drawy = n_y.sw.h + n_drawoffset_y; if( drawx >= (int32_t)n_drawarea_x1 && drawy >= (int32_t)n_drawarea_y1 && drawx <= (int32_t)n_drawarea_x2 && drawy <= (int32_t)n_drawarea_y2 ) { uint16_t *p_vram = p_p_vram[ drawy ] + drawx; switch( n_cmd & 0x02 ) { case 0x00: /* transparency off */ WRITE_PIXEL( p_n_redshade[ MID_LEVEL | n_r ] | p_n_greenshade[ MID_LEVEL | n_g ] | p_n_blueshade[ MID_LEVEL | n_b ] ) break; case 0x02: /* transparency on */ WRITE_PIXEL( p_n_redtrans[ p_n_f[ MID_LEVEL | n_r ] | p_n_redb[ *( p_vram ) ] ] | p_n_greentrans[ p_n_f[ MID_LEVEL | n_g ] | p_n_greenb[ *( p_vram ) ] ] | p_n_bluetrans[ p_n_f[ MID_LEVEL | n_b ] | p_n_blueb[ *( p_vram ) ] ] ) break; } } n_x.d += n_dx; n_y.d += n_dy; n_len--; } } void psxgpu_device::GouraudLine() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 6 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.GouraudLine.vertex[ 0 ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.GouraudLine.vertex[ 0 ].n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.GouraudLine.vertex[ 1 ].n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.GouraudLine.vertex[ 1 ].n_coord ) ) + n_drawoffset_y ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.GouraudLine.vertex[ 0 ].n_bgr ); TRANSPARENCYSETUP int32_t n_xstart = SINT11( COORD_X( m_packet.GouraudLine.vertex[ 0 ].n_coord ) ); int32_t n_ystart = SINT11( COORD_Y( m_packet.GouraudLine.vertex[ 0 ].n_coord ) ); PAIR n_cr1; n_cr1.w.h = BGR_R( m_packet.GouraudLine.vertex[ 0 ].n_bgr ); n_cr1.w.l = 0; PAIR n_cg1; n_cg1.w.h = BGR_G( m_packet.GouraudLine.vertex[ 0 ].n_bgr ); n_cg1.w.l = 0; PAIR n_cb1; n_cb1.w.h = BGR_B( m_packet.GouraudLine.vertex[ 0 ].n_bgr ); n_cb1.w.l = 0; int32_t n_xend = SINT11( COORD_X( m_packet.GouraudLine.vertex[ 1 ].n_coord ) ); int32_t n_yend = SINT11( COORD_Y( m_packet.GouraudLine.vertex[ 1 ].n_coord ) ); PAIR n_cr2; n_cr2.w.h = BGR_R( m_packet.GouraudLine.vertex[ 1 ].n_bgr ); n_cr2.w.l = 0; PAIR n_cg2; n_cg2.w.h = BGR_G( m_packet.GouraudLine.vertex[ 1 ].n_bgr ); n_cg2.w.l = 0; PAIR n_cb2; n_cb2.w.h = BGR_B( m_packet.GouraudLine.vertex[ 1 ].n_bgr ); n_cb2.w.l = 0; PAIR n_x; n_x.sw.h = n_xstart; n_x.sw.l = 0; PAIR n_y; n_y.sw.h = n_ystart; n_y.sw.l = 0; PAIR n_r; n_r.d = n_cr1.d; PAIR n_g; n_g.d = n_cg1.d; PAIR n_b; n_b.d = n_cb1.d; int32_t n_xlen; if( n_xend > n_xstart ) { n_xlen = n_xend - n_xstart; } else { n_xlen = n_xstart - n_xend; } int32_t n_ylen; if( n_yend > n_ystart ) { n_ylen = n_yend - n_ystart; } else { n_ylen = n_ystart - n_yend; } int32_t n_distance; if( n_xlen > n_ylen ) { n_distance = n_xlen; } else { n_distance = n_ylen; } if( n_distance == 0 ) { n_distance = 1; } int32_t n_dx = (int32_t)( ( n_xend << 16 ) - n_x.sd ) / n_distance; int32_t n_dy = (int32_t)( ( n_yend << 16 ) - n_y.sd ) / n_distance; int32_t n_dr = (int32_t)( n_cr2.d - n_cr1.d ) / n_distance; int32_t n_dg = (int32_t)( n_cg2.d - n_cg1.d ) / n_distance; int32_t n_db = (int32_t)( n_cb2.d - n_cb1.d ) / n_distance; while( n_distance > 0 ) { int drawx = n_x.sw.h + n_drawoffset_x; int drawy = n_y.sw.h + n_drawoffset_y; if( drawx >= (int32_t)n_drawarea_x1 && drawy >= (int32_t)n_drawarea_y1 && drawx <= (int32_t)n_drawarea_x2 && drawy <= (int32_t)n_drawarea_y2 ) { uint16_t *p_vram = p_p_vram[ drawy ] + drawx; switch( n_cmd & 0x02 ) { case 0x00: /* transparency off */ WRITE_PIXEL( p_n_redshade[ MID_LEVEL | n_r.w.h ] | p_n_greenshade[ MID_LEVEL | n_g.w.h ] | p_n_blueshade[ MID_LEVEL | n_b.w.h ] ) break; case 0x02: /* transparency on */ WRITE_PIXEL( p_n_redtrans[ p_n_f[ MID_LEVEL | n_r.w.h ] | p_n_redb[ *( p_vram ) ] ] | p_n_greentrans[ p_n_f[ MID_LEVEL | n_g.w.h ] | p_n_greenb[ *( p_vram ) ] ] | p_n_bluetrans[ p_n_f[ MID_LEVEL | n_b.w.h ] | p_n_blueb[ *( p_vram ) ] ] ) break; } } n_x.sd += n_dx; n_y.sd += n_dy; n_r.d += n_dr; n_g.d += n_dg; n_b.d += n_db; n_distance--; } } void psxgpu_device::FrameBufferRectangleDraw() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 7 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + SIZE_W( m_packet.FlatRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + SIZE_H( m_packet.FlatRectangle.n_size ) ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + SIZE_W( m_packet.FlatRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + SIZE_H( m_packet.FlatRectangle.n_size ) ); DebugMeshEnd(); #endif PAIR n_r; n_r.w.h = BGR_R( m_packet.FlatRectangle.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = BGR_G( m_packet.FlatRectangle.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = BGR_B( m_packet.FlatRectangle.n_bgr ); n_b.w.l = 0; int16_t n_y = COORD_Y( m_packet.FlatRectangle.n_coord ); int32_t n_h = SIZE_H( m_packet.FlatRectangle.n_size ); while( n_h > 0 ) { int16_t n_x = COORD_X( m_packet.FlatRectangle.n_coord ); int32_t n_distance = SIZE_W( m_packet.FlatRectangle.n_size ); while( n_distance > 0 ) { p_p_vram[ n_y & 1023 ][ n_x & 1023 ] = p_n_redshade[ MID_LEVEL | n_r.w.h ] | p_n_greenshade[ MID_LEVEL | n_g.w.h ] | p_n_blueshade[ MID_LEVEL | n_b.w.h ]; n_x++; n_distance--; } n_y++; n_h--; } } void psxgpu_device::FlatRectangle() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 8 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_x + SIZE_W( m_packet.FlatRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_y + SIZE_H( m_packet.FlatRectangle.n_size ) ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_x + SIZE_W( m_packet.FlatRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ) + n_drawoffset_y + SIZE_H( m_packet.FlatRectangle.n_size ) ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatRectangle.n_bgr ); SOLIDSETUP PAIR n_r; n_r.w.h = BGR_R( m_packet.FlatRectangle.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = BGR_G( m_packet.FlatRectangle.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = BGR_B( m_packet.FlatRectangle.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.FlatRectangle.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.FlatRectangle.n_coord ) ); int32_t n_h = SIZE_H( m_packet.FlatRectangle.n_size ); while( n_h > 0 ) { int32_t n_distance = SIZE_W( m_packet.FlatRectangle.n_size ); int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } SOLIDFILL( FLATRECTANGEUPDATE ) } n_y++; n_h--; } } void psxgpu_device::FlatRectangle8x8() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 9 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_x + 8, SINT11( COORD_Y( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_y + 8 ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_x + 8, SINT11( COORD_Y( m_packet.FlatRectangle8x8.n_coord ) ) + n_drawoffset_y + 8 ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatRectangle8x8.n_bgr ); SOLIDSETUP PAIR n_r; n_r.w.h = BGR_R( m_packet.FlatRectangle8x8.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = BGR_G( m_packet.FlatRectangle8x8.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = BGR_B( m_packet.FlatRectangle8x8.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.FlatRectangle8x8.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.FlatRectangle8x8.n_coord ) ); int32_t n_h = 8; while( n_h > 0 ) { int32_t n_distance = 8; int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } SOLIDFILL( FLATRECTANGEUPDATE ) } n_y++; n_h--; } } void psxgpu_device::FlatRectangle16x16() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 10 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_x + 16, SINT11( COORD_Y( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_y + 16 ); DebugMesh( SINT11( COORD_X( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_x + 16, SINT11( COORD_Y( m_packet.FlatRectangle16x16.n_coord ) ) + n_drawoffset_y + 16 ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatRectangle16x16.n_bgr ); SOLIDSETUP PAIR n_r; n_r.w.h = BGR_R( m_packet.FlatRectangle16x16.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = BGR_G( m_packet.FlatRectangle16x16.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = BGR_B( m_packet.FlatRectangle16x16.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.FlatRectangle16x16.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.FlatRectangle16x16.n_coord ) ); int32_t n_h = 16; while( n_h > 0 ) { int32_t n_distance = 16; int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && n_y >= (int32_t)n_drawarea_y1 && n_y <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } SOLIDFILL( FLATRECTANGEUPDATE ) } n_y++; n_h--; } } void psxgpu_device::FlatTexturedRectangle() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 11 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_x + SIZE_W( m_packet.FlatTexturedRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_y + SIZE_H( m_packet.FlatTexturedRectangle.n_size ) ); DebugMesh( SINT11( COORD_X( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_x + SIZE_W( m_packet.FlatTexturedRectangle.n_size ), SINT11( COORD_Y( m_packet.FlatTexturedRectangle.n_coord ) ) + n_drawoffset_y + SIZE_H( m_packet.FlatTexturedRectangle.n_size ) ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.FlatTexturedRectangle.n_bgr ); uint32_t n_clutx = ( m_packet.FlatTexturedRectangle.n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.FlatTexturedRectangle.n_texture.w.h >> 6 ) & 0x3ff; TEXTURESETUP SPRITESETUP PAIR n_r; n_r.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.FlatTexturedRectangle.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.FlatTexturedRectangle.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.FlatTexturedRectangle.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.FlatTexturedRectangle.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.FlatTexturedRectangle.n_coord ) ); uint8_t n_v = TEXTURE_V( m_packet.FlatTexturedRectangle.n_texture ); uint32_t n_h = SIZE_H( m_packet.FlatTexturedRectangle.n_size ); while( n_h > 0 ) { uint8_t n_u = TEXTURE_U( m_packet.FlatTexturedRectangle.n_texture ); int16_t n_distance = SIZE_W( m_packet.FlatTexturedRectangle.n_size ); int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_u += ( n_drawarea_x1 - drawx ) * n_du; n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } TEXTUREFILL( FLATTEXTUREDRECTANGLEUPDATE, n_u, n_v ); } n_v += n_dv; n_y++; n_h--; } } void psxgpu_device::Sprite8x8() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 12 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_x + 7, SINT11( COORD_Y( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_y + 7 ); DebugMesh( SINT11( COORD_X( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_x + 7, SINT11( COORD_Y( m_packet.Sprite8x8.n_coord ) ) + n_drawoffset_y + 7 ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.Sprite8x8.n_bgr ); uint32_t n_clutx = ( m_packet.Sprite8x8.n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.Sprite8x8.n_texture.w.h >> 6 ) & 0x3ff; TEXTURESETUP SPRITESETUP PAIR n_r; n_r.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.Sprite8x8.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.Sprite8x8.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.Sprite8x8.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.Sprite8x8.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.Sprite8x8.n_coord ) ); uint8_t n_v = TEXTURE_V( m_packet.Sprite8x8.n_texture ); uint32_t n_h = 8; while( n_h > 0 ) { uint8_t n_u = TEXTURE_U( m_packet.Sprite8x8.n_texture ); int16_t n_distance = 8; int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_u += ( n_drawarea_x1 - drawx ) * n_du; n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } TEXTUREFILL( FLATTEXTUREDRECTANGLEUPDATE, n_u, n_v ); } n_v += n_dv; n_y++; n_h--; } } void psxgpu_device::Sprite16x16() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 13 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_x + 7, SINT11( COORD_Y( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_y ); DebugMesh( SINT11( COORD_X( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_y + 7 ); DebugMesh( SINT11( COORD_X( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_x + 7, SINT11( COORD_Y( m_packet.Sprite16x16.n_coord ) ) + n_drawoffset_y + 7 ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.Sprite16x16.n_bgr ); uint32_t n_clutx = ( m_packet.Sprite16x16.n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.Sprite16x16.n_texture.w.h >> 6 ) & 0x3ff; TEXTURESETUP SPRITESETUP PAIR n_r; n_r.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.Sprite16x16.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.Sprite16x16.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.Sprite16x16.n_bgr ); n_b.w.l = 0; int16_t n_x = SINT11( COORD_X( m_packet.Sprite16x16.n_coord ) ); int16_t n_y = SINT11( COORD_Y( m_packet.Sprite16x16.n_coord ) ); uint8_t n_v = TEXTURE_V( m_packet.Sprite16x16.n_texture ); uint32_t n_h = 16; while( n_h > 0 ) { uint8_t n_u = TEXTURE_U( m_packet.Sprite16x16.n_texture ); int16_t n_distance = 16; int drawy = n_y + n_drawoffset_y; if( n_distance > 0 && drawy >= (int32_t)n_drawarea_y1 && drawy <= (int32_t)n_drawarea_y2 ) { int drawx = n_x + n_drawoffset_x; if( ( (int32_t)n_drawarea_x1 - drawx ) > 0 ) { n_u += ( n_drawarea_x1 - drawx ) * n_du; n_distance -= ( n_drawarea_x1 - drawx ); drawx = n_drawarea_x1; } TEXTUREFILL( FLATTEXTUREDRECTANGLEUPDATE, n_u, n_v ); } n_v += n_dv; n_y++; n_h--; } } void psxgpu_device::Dot() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 14 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.Dot.vertex.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.Dot.vertex.n_coord ) ) + n_drawoffset_y ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.Dot.n_bgr ); uint8_t n_r = BGR_R( m_packet.Dot.n_bgr ); uint8_t n_g = BGR_G( m_packet.Dot.n_bgr ); uint8_t n_b = BGR_B( m_packet.Dot.n_bgr ); int32_t n_x = SINT11( COORD_X( m_packet.Dot.vertex.n_coord ) ); int32_t n_y = SINT11( COORD_Y( m_packet.Dot.vertex.n_coord ) ); TRANSPARENCYSETUP int drawx = n_x + n_drawoffset_x; int drawy = n_y + n_drawoffset_y; if( drawx >= (int32_t)n_drawarea_x1 && drawy >= (int32_t)n_drawarea_y1 && drawx <= (int32_t)n_drawarea_x2 && drawy <= (int32_t)n_drawarea_y2 ) { uint16_t *p_vram = p_p_vram[ drawy ] + drawx; switch( n_cmd & 0x02 ) { case 0x00: /* transparency off */ WRITE_PIXEL( p_n_redshade[ MID_LEVEL | n_r ] | p_n_greenshade[ MID_LEVEL | n_g ] | p_n_blueshade[ MID_LEVEL | n_b ] ) break; case 0x02: /* transparency on */ WRITE_PIXEL( p_n_redtrans[ p_n_f[ MID_LEVEL | n_r ] | p_n_redb[ *( p_vram ) ] ] | p_n_greentrans[ p_n_f[ MID_LEVEL | n_g ] | p_n_greenb[ *( p_vram ) ] ] | p_n_bluetrans[ p_n_f[ MID_LEVEL | n_b ] | p_n_blueb[ *( p_vram ) ] ] ) break; } } } void psxgpu_device::TexturedDot() { #if PSXGPU_DEBUG_VIEWER if (m_debug.n_skip == 15) { return; } DebugMesh( SINT11( COORD_X( m_packet.TexturedDot.vertex.n_coord ) ) + n_drawoffset_x, SINT11( COORD_Y( m_packet.TexturedDot.vertex.n_coord ) ) + n_drawoffset_y ); DebugMeshEnd(); #endif uint8_t n_cmd = BGR_C( m_packet.TexturedDot.n_bgr ); PAIR n_r; n_r.w.h = n_cmd & 0x01 ? 0x80 : BGR_R( m_packet.TexturedDot.n_bgr ); n_r.w.l = 0; PAIR n_g; n_g.w.h = n_cmd & 0x01 ? 0x80 : BGR_G( m_packet.TexturedDot.n_bgr ); n_g.w.l = 0; PAIR n_b; n_b.w.h = n_cmd & 0x01 ? 0x80 : BGR_B( m_packet.TexturedDot.n_bgr ); n_b.w.l = 0; int32_t n_x = SINT11( COORD_X( m_packet.TexturedDot.vertex.n_coord ) ); int32_t n_y = SINT11( COORD_Y( m_packet.TexturedDot.vertex.n_coord ) ); uint8_t n_u = TEXTURE_U(m_packet.TexturedDot.vertex.n_texture ); uint8_t n_v = TEXTURE_V(m_packet.TexturedDot.vertex.n_texture ); uint32_t n_clutx = ( m_packet.TexturedDot.vertex.n_texture.w.h & 0x3f ) << 4; uint32_t n_cluty = ( m_packet.TexturedDot.vertex.n_texture.w.h >> 6 ) & 0x3ff; TEXTURESETUP int32_t n_distance = 1; int drawx = n_x + n_drawoffset_x; int drawy = n_y + n_drawoffset_y; if( drawx >= (int32_t)n_drawarea_x1 && drawy >= (int32_t)n_drawarea_y1 && drawx <= (int32_t)n_drawarea_x2 && drawy <= (int32_t)n_drawarea_y2 ) { TEXTUREFILL( {}, n_u, n_v ); } } void psxgpu_device::MoveImage() { #if PSXGPU_DEBUG_VIEWER if( m_debug.n_skip == 16 ) { return; } DebugMesh( SINT11( COORD_X( m_packet.MoveImage.vertex[ 1 ].n_coord ) ), SINT11( COORD_Y( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) ); DebugMesh( SINT11( COORD_X( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) + SIZE_W( m_packet.MoveImage.n_size ), SINT11( COORD_Y( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) ); DebugMesh( SINT11( COORD_X( m_packet.MoveImage.vertex[ 1 ].n_coord ) ), SINT11( COORD_Y( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) + SIZE_H( m_packet.MoveImage.n_size ) ); DebugMesh( SINT11( COORD_X( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) + SIZE_W( m_packet.MoveImage.n_size ), SINT11( COORD_Y( m_packet.MoveImage.vertex[ 1 ].n_coord ) ) + SIZE_H( m_packet.MoveImage.n_size ) ); DebugMeshEnd(); #endif int16_t n_srcy = COORD_Y( m_packet.MoveImage.vertex[ 0 ].n_coord ); int16_t n_dsty = COORD_Y( m_packet.MoveImage.vertex[ 1 ].n_coord ); int16_t n_h = SIZE_H( m_packet.MoveImage.n_size ); while( n_h > 0 ) { int16_t n_srcx = COORD_X( m_packet.MoveImage.vertex[ 0 ].n_coord ); int16_t n_dstx = COORD_X( m_packet.MoveImage.vertex[ 1 ].n_coord ); int16_t n_w = SIZE_W( m_packet.MoveImage.n_size ); while( n_w > 0 ) { uint16_t *p_vram = p_p_vram[ n_dsty & 1023 ] + ( n_dstx & 1023 ); WRITE_PIXEL( *( p_p_vram[ n_srcy & 1023 ] + ( n_srcx & 1023 ) ) ) n_srcx++; n_dstx++; n_w--; } n_srcy++; n_dsty++; n_h--; } } void psxgpu_device::dma_write( uint32_t *p_n_psxram, uint32_t n_address, int32_t n_size ) { gpu_write( &p_n_psxram[ n_address / 4 ], n_size ); } void psxgpu_device::gpu_write( uint32_t *p_ram, int32_t n_size ) { while( n_size > 0 ) { uint32_t data = *( p_ram ); verboselog( *this, 2, "PSX Packet #%u %08x\n", n_gpu_buffer_offset, data ); m_packet.n_entry[ n_gpu_buffer_offset ] = data; switch( m_packet.n_entry[ 0 ] >> 24 ) { case 0x00: verboselog( *this, 1, "not handled: GPU Command 0x00: (%08x)\n", data ); break; case 0x01: verboselog( *this, 1, "not handled: clear cache\n" ); break; case 0x02: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: frame buffer rectangle %u,%u %u,%u\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 1 ] & 0xffff, m_packet.n_entry[ 1 ] >> 16, m_packet.n_entry[ 2 ] & 0xffff, m_packet.n_entry[ 2 ] >> 16 ); FrameBufferRectangleDraw(); n_gpu_buffer_offset = 0; } break; case 0x20: case 0x21: case 0x22: case 0x23: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: monochrome 3 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); FlatPolygon( 3 ); n_gpu_buffer_offset = 0; } break; case 0x24: case 0x25: case 0x26: case 0x27: if( n_gpu_buffer_offset < 6 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: textured 3 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); FlatTexturedPolygon( 3 ); n_gpu_buffer_offset = 0; } break; case 0x28: case 0x29: case 0x2a: case 0x2b: if( n_gpu_buffer_offset < 4 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: monochrome 4 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); FlatPolygon( 4 ); n_gpu_buffer_offset = 0; } break; case 0x2c: case 0x2d: case 0x2e: case 0x2f: if( n_gpu_buffer_offset < 8 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: textured 4 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); FlatTexturedPolygon( 4 ); n_gpu_buffer_offset = 0; } break; case 0x30: case 0x31: case 0x32: case 0x33: if( n_gpu_buffer_offset < 5 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud 3 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudPolygon( 3 ); n_gpu_buffer_offset = 0; } break; case 0x34: case 0x35: case 0x36: case 0x37: if( n_gpu_buffer_offset < 8 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud textured 3 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudTexturedPolygon( 3 ); n_gpu_buffer_offset = 0; } break; case 0x38: case 0x39: case 0x3a: case 0x3b: if( n_gpu_buffer_offset < 7 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud 4 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudPolygon( 4 ); n_gpu_buffer_offset = 0; } break; case 0x3c: case 0x3d: case 0x3e: case 0x3f: if( n_gpu_buffer_offset < 11 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud textured 4 point polygon\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudTexturedPolygon( 4 ); n_gpu_buffer_offset = 0; } break; case 0x40: case 0x41: case 0x42: case 0x43: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: monochrome line\n", m_packet.n_entry[ 0 ] >> 24 ); MonochromeLine(); n_gpu_buffer_offset = 0; } break; case 0x48: case 0x4a: case 0x4c: case 0x4e: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: monochrome polyline\n", m_packet.n_entry[ 0 ] >> 24 ); MonochromeLine(); if( ( m_packet.n_entry[ 3 ] & 0xf000f000 ) != 0x50005000 ) { m_packet.n_entry[ 1 ] = m_packet.n_entry[ 2 ]; m_packet.n_entry[ 2 ] = m_packet.n_entry[ 3 ]; n_gpu_buffer_offset = 3; } else { n_gpu_buffer_offset = 0; } } break; case 0x50: case 0x51: case 0x52: case 0x53: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud line\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudLine(); n_gpu_buffer_offset = 0; } break; case 0x58: case 0x5a: case 0x5c: case 0x5e: if( n_gpu_buffer_offset < 5 && ( n_gpu_buffer_offset != 4 || ( m_packet.n_entry[ 4 ] & 0xf000f000 ) != 0x50005000 ) ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: gouraud polyline\n", m_packet.n_entry[ 0 ] >> 24 ); GouraudLine(); if( ( m_packet.n_entry[ 4 ] & 0xf000f000 ) != 0x50005000 ) { m_packet.n_entry[ 0 ] = ( m_packet.n_entry[ 0 ] & 0xff000000 ) | ( m_packet.n_entry[ 2 ] & 0x00ffffff ); m_packet.n_entry[ 1 ] = m_packet.n_entry[ 3 ]; m_packet.n_entry[ 2 ] = m_packet.n_entry[ 4 ]; m_packet.n_entry[ 3 ] = m_packet.n_entry[ 5 ]; n_gpu_buffer_offset = 4; } else { n_gpu_buffer_offset = 0; } } break; case 0x60: case 0x61: case 0x62: case 0x63: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: rectangle %d,%d %d,%d\n", m_packet.n_entry[ 0 ] >> 24, (int16_t)( m_packet.n_entry[ 1 ] & 0xffff ), (int16_t)( m_packet.n_entry[ 1 ] >> 16 ), (int16_t)( m_packet.n_entry[ 2 ] & 0xffff ), (int16_t)( m_packet.n_entry[ 2 ] >> 16 ) ); FlatRectangle(); n_gpu_buffer_offset = 0; } break; case 0x64: case 0x65: case 0x66: case 0x67: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: sprite %d,%d %u,%u %08x, %08x\n", m_packet.n_entry[ 0 ] >> 24, (int16_t)( m_packet.n_entry[ 1 ] & 0xffff ), (int16_t)( m_packet.n_entry[ 1 ] >> 16 ), m_packet.n_entry[ 3 ] & 0xffff, m_packet.n_entry[ 3 ] >> 16, m_packet.n_entry[ 0 ], m_packet.n_entry[ 2 ] ); FlatTexturedRectangle(); n_gpu_buffer_offset = 0; } break; case 0x68: case 0x69: case 0x6a: case 0x6b: if( n_gpu_buffer_offset < 1 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: dot %d,%d %08x\n", m_packet.n_entry[ 0 ] >> 24, (int16_t)( m_packet.n_entry[ 1 ] & 0xffff ), (int16_t)( m_packet.n_entry[ 1 ] >> 16 ), m_packet.n_entry[ 0 ] & 0xffffff ); Dot(); n_gpu_buffer_offset = 0; } break; case 0x6c: case 0x6d: case 0x6e: case 0x6f: if (n_gpu_buffer_offset < 2) { n_gpu_buffer_offset++; } else { verboselog(*this, 1, "%02x: textured dot %d,%d %08x\n", m_packet.n_entry[ 0 ] >> 24, (int16_t)( m_packet.n_entry[ 1 ] & 0xffff ), (int16_t)( m_packet.n_entry[ 1 ] >> 16 ), m_packet.n_entry[ 0 ] & 0xffffff ); TexturedDot(); n_gpu_buffer_offset = 0; } break; case 0x70: case 0x71: case 0x72: case 0x73: /* 8*8 rectangle */ if( n_gpu_buffer_offset < 1 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: 16x16 rectangle %08x %08x\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 0 ], m_packet.n_entry[ 1 ] ); FlatRectangle8x8(); n_gpu_buffer_offset = 0; } break; case 0x74: case 0x75: case 0x76: case 0x77: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: 8x8 sprite %08x %08x %08x\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 0 ], m_packet.n_entry[ 1 ], m_packet.n_entry[ 2 ] ); Sprite8x8(); n_gpu_buffer_offset = 0; } break; case 0x78: case 0x79: case 0x7a: case 0x7b: /* 16*16 rectangle */ if( n_gpu_buffer_offset < 1 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: 16x16 rectangle %08x %08x\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 0 ], m_packet.n_entry[ 1 ] ); FlatRectangle16x16(); n_gpu_buffer_offset = 0; } break; case 0x7c: case 0x7d: case 0x7e: case 0x7f: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: 16x16 sprite %08x %08x %08x\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 0 ], m_packet.n_entry[ 1 ], m_packet.n_entry[ 2 ] ); Sprite16x16(); n_gpu_buffer_offset = 0; } break; case 0x80: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "move image in frame buffer %08x %08x %08x %08x\n", m_packet.n_entry[ 0 ], m_packet.n_entry[ 1 ], m_packet.n_entry[ 2 ], m_packet.n_entry[ 3 ] ); MoveImage(); n_gpu_buffer_offset = 0; } break; case 0xa0: if( n_gpu_buffer_offset < 3 ) { n_gpu_buffer_offset++; } else { for( int n_pixel = 0; n_pixel < 2; n_pixel++ ) { verboselog( *this, 2, "send image to framebuffer ( pixel %u,%u = %u )\n", ( n_vramx + m_packet.n_entry[ 1 ] ) & 1023, ( n_vramy + ( m_packet.n_entry[ 1 ] >> 16 ) ) & 1023, data & 0xffff ); uint16_t *p_vram = p_p_vram[ ( n_vramy + ( m_packet.n_entry[ 1 ] >> 16 ) ) & 1023 ] + ( ( n_vramx + m_packet.n_entry[ 1 ] ) & 1023 ); WRITE_PIXEL( data & 0xffff ) n_vramx++; if( n_vramx >= ( m_packet.n_entry[ 2 ] & 0xffff ) ) { n_vramx = 0; n_vramy++; if( n_vramy >= ( m_packet.n_entry[ 2 ] >> 16 ) ) { verboselog( *this, 1, "%02x: send image to framebuffer %u,%u %u,%u\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 1 ] & 0xffff, ( m_packet.n_entry[ 1 ] >> 16 ), m_packet.n_entry[ 2 ] & 0xffff, ( m_packet.n_entry[ 2 ] >> 16 ) ); n_gpu_buffer_offset = 0; n_vramx = 0; n_vramy = 0; break; } } data >>= 16; } } break; case 0xc0: if( n_gpu_buffer_offset < 2 ) { n_gpu_buffer_offset++; } else { verboselog( *this, 1, "%02x: copy image from frame buffer\n", m_packet.n_entry[ 0 ] >> 24 ); n_gpustatus |= ( 1L << 0x1b ); } break; case 0xe1: verboselog( *this, 1, "%02x: draw mode %06x\n", m_packet.n_entry[ 0 ] >> 24, m_packet.n_entry[ 0 ] & 0xffffff ); decode_tpage( m_packet.n_entry[ 0 ] & 0xffffff ); break; case 0xe2: n_twy = ( ( ( m_packet.n_entry[ 0 ] >> 15 ) & 0x1f ) << 3 ); n_twx = ( ( ( m_packet.n_entry[ 0 ] >> 10 ) & 0x1f ) << 3 ); n_twh = 255 - ( ( ( m_packet.n_entry[ 0 ] >> 5 ) & 0x1f ) << 3 ); n_tww = 255 - ( ( m_packet.n_entry[ 0 ] & 0x1f ) << 3 ); verboselog( *this, 1, "%02x: texture window %u,%u %u,%u\n", m_packet.n_entry[ 0 ] >> 24, n_twx, n_twy, n_tww, n_twh ); break; case 0xe3: n_drawarea_x1 = m_packet.n_entry[ 0 ] & 1023; if( m_n_gputype == 2 ) { n_drawarea_y1 = ( m_packet.n_entry[ 0 ] >> 10 ) & 1023; } else { n_drawarea_y1 = ( m_packet.n_entry[ 0 ] >> 12 ) & 1023; } verboselog( *this, 1, "%02x: drawing area top left %d,%d\n", m_packet.n_entry[ 0 ] >> 24, n_drawarea_x1, n_drawarea_y1 ); break; case 0xe4: n_drawarea_x2 = m_packet.n_entry[ 0 ] & 1023; if( m_n_gputype == 2 ) { n_drawarea_y2 = ( m_packet.n_entry[ 0 ] >> 10 ) & 1023; } else { n_drawarea_y2 = ( m_packet.n_entry[ 0 ] >> 12 ) & 1023; } verboselog( *this, 1, "%02x: drawing area bottom right %d,%d\n", m_packet.n_entry[ 0 ] >> 24, n_drawarea_x2, n_drawarea_y2 ); break; case 0xe5: n_drawoffset_x = SINT11( m_packet.n_entry[ 0 ] & 2047 ); if( m_n_gputype == 2 ) { n_drawoffset_y = SINT11( ( m_packet.n_entry[ 0 ] >> 11 ) & 2047 ); } else { n_drawoffset_y = SINT11( ( m_packet.n_entry[ 0 ] >> 12 ) & 2047 ); } verboselog( *this, 1, "%02x: drawing offset %d,%d\n", m_packet.n_entry[ 0 ] >> 24, n_drawoffset_x, n_drawoffset_y ); break; case 0xe6: m_draw_stp = BIT( data, 0 ); m_check_stp = BIT( data, 1 ); // TODO: confirm status bits on real type 1 gpu n_gpustatus &= ~( 3L << 0xb ); n_gpustatus |= ( data & 0x03 ) << 0xb; verboselog( *this, 1, "mask setting %d\n", m_packet.n_entry[ 0 ] & 3 ); break; default: #if defined( MAME_DEBUG ) popmessage( "unknown GPU packet %08x", m_packet.n_entry[ 0 ] ); #endif verboselog( *this, 0, "unknown GPU packet %08x (%08x)\n", m_packet.n_entry[ 0 ], data ); #if ( STOP_ON_ERROR ) n_gpu_buffer_offset = 1; #endif break; } p_ram++; n_size--; } } WRITE32_MEMBER( psxgpu_device::write ) { switch( offset ) { case 0x00: gpu_write( &data, 1 ); break; case 0x01: switch( data >> 24 ) { case 0x00: gpu_reset(); break; case 0x01: verboselog( *this, 1, "not handled: reset command buffer\n" ); n_gpu_buffer_offset = 0; break; case 0x02: verboselog( *this, 1, "not handled: reset irq\n" ); break; case 0x03: n_gpustatus &= ~( 1L << 0x17 ); n_gpustatus |= ( data & 0x01 ) << 0x17; break; case 0x04: verboselog( *this, 1, "dma setup %d\n", data & 3 ); n_gpustatus &= ~( 3L << 0x1d ); n_gpustatus |= ( data & 0x03 ) << 0x1d; n_gpustatus &= ~( 1L << 0x19 ); if( ( data & 3 ) == 1 || ( data & 3 ) == 2 ) { n_gpustatus |= ( 1L << 0x19 ); } break; case 0x05: m_n_displaystartx = data & 1023; if( m_n_gputype == 2 ) { n_displaystarty = ( data >> 10 ) & 1023; } else { n_displaystarty = ( data >> 12 ) & 1023; } verboselog( *this, 1, "start of display area %d %d\n", m_n_displaystartx, n_displaystarty ); break; case 0x06: n_horiz_disstart = data & 4095; n_horiz_disend = ( data >> 12 ) & 4095; verboselog( *this, 1, "horizontal display range %d %d\n", n_horiz_disstart, n_horiz_disend ); break; case 0x07: n_vert_disstart = data & 1023; n_vert_disend = ( data >> 10 ) & 2047; verboselog( *this, 1, "vertical display range %d %d\n", n_vert_disstart, n_vert_disend ); break; case 0x08: verboselog( *this, 1, "display mode %02x\n", data & 0xff ); n_gpustatus &= ~( 127L << 0x10 ); n_gpustatus |= ( data & 0x3f ) << 0x11; /* width 0 + height + videmode + isrgb24 + isinter */ n_gpustatus |= ( ( data & 0x40 ) >> 0x06 ) << 0x10; /* width 1 */ if( m_n_gputype == 1 ) { b_reverseflag = ( data >> 7 ) & 1; } updatevisiblearea(); break; case 0x09: verboselog( *this, 1, "not handled: GPU Control 0x09: %08x\n", data ); break; case 0x0d: verboselog( *this, 1, "reset lightgun coordinates %08x\n", data ); n_lightgun_x = 0; n_lightgun_y = 0; break; case 0x10: switch( data & 0xff ) { case 0x03: if( m_n_gputype == 2 ) { n_gpuinfo = n_drawarea_x1 | ( n_drawarea_y1 << 10 ); } else { n_gpuinfo = n_drawarea_x1 | ( n_drawarea_y1 << 12 ); } verboselog( *this, 1, "GPU Info - Draw area top left %08x\n", n_gpuinfo ); break; case 0x04: if( m_n_gputype == 2 ) { n_gpuinfo = n_drawarea_x2 | ( n_drawarea_y2 << 10 ); } else { n_gpuinfo = n_drawarea_x2 | ( n_drawarea_y2 << 12 ); } verboselog( *this, 1, "GPU Info - Draw area bottom right %08x\n", n_gpuinfo ); break; case 0x05: if( m_n_gputype == 2 ) { n_gpuinfo = ( n_drawoffset_x & 2047 ) | ( ( n_drawoffset_y & 2047 ) << 11 ); } else { n_gpuinfo = ( n_drawoffset_x & 2047 ) | ( ( n_drawoffset_y & 2047 ) << 12 ); } verboselog( *this, 1, "GPU Info - Draw offset %08x\n", n_gpuinfo ); break; case 0x07: n_gpuinfo = m_n_gputype; verboselog( *this, 1, "GPU Info - GPU Type %08x\n", n_gpuinfo ); break; case 0x08: n_gpuinfo = n_lightgun_x | ( n_lightgun_y << 16 ); verboselog( *this, 1, "GPU Info - lightgun coordinates %08x\n", n_gpuinfo ); break; default: verboselog( *this, 0, "GPU Info - unknown request (%08x)\n", data ); n_gpuinfo = 0; break; } break; case 0x20: verboselog( *this, 1, "not handled: GPU Control 0x20: %08x\n", data ); break; default: #if defined( MAME_DEBUG ) popmessage( "unknown GPU command %08x", data ); #endif verboselog( *this, 0, "gpu_w( %08x ) unknown GPU command\n", data ); break; } break; default: verboselog( *this, 0, "gpu_w( %08x, %08x, %08x ) unknown register\n", offset, data, mem_mask ); break; } } void psxgpu_device::dma_read( uint32_t *p_n_psxram, uint32_t n_address, int32_t n_size ) { gpu_read( &p_n_psxram[ n_address / 4 ], n_size ); } void psxgpu_device::gpu_read( uint32_t *p_ram, int32_t n_size ) { while( n_size > 0 ) { if( ( n_gpustatus & ( 1L << 0x1b ) ) != 0 ) { PAIR data; verboselog( *this, 2, "copy image from frame buffer ( %d, %d )\n", n_vramx, n_vramy ); data.d = 0; for( int n_pixel = 0; n_pixel < 2; n_pixel++ ) { data.w.l = data.w.h; data.w.h = *( p_p_vram[ ( n_vramy + ( m_packet.n_entry[ 1 ] >> 16 ) ) & 0x3ff ] + ( ( n_vramx + ( m_packet.n_entry[ 1 ] & 0xffff ) ) & 0x3ff ) ); n_vramx++; if( n_vramx >= ( m_packet.n_entry[ 2 ] & 0xffff ) ) { n_vramx = 0; n_vramy++; if( n_vramy >= ( m_packet.n_entry[ 2 ] >> 16 ) ) { verboselog( *this, 1, "copy image from frame buffer end\n" ); n_gpustatus &= ~( 1L << 0x1b ); n_gpu_buffer_offset = 0; n_vramx = 0; n_vramy = 0; if( n_pixel == 0 ) { data.w.l = data.w.h; data.w.h = 0; } break; } } } *( p_ram ) = data.d; } else { verboselog( *this, 2, "read GPU info (%08x)\n", n_gpuinfo ); *( p_ram ) = n_gpuinfo; } p_ram++; n_size--; } } READ32_MEMBER( psxgpu_device::read ) { uint32_t data; switch( offset ) { case 0x00: gpu_read( &data, 1 ); break; case 0x01: data = n_gpustatus; verboselog( *this, 1, "read GPU status (%08x)\n", data ); break; default: verboselog( *this, 0, "gpu_r( %08x, %08x ) unknown register\n", offset, mem_mask ); data = 0; break; } return data; } void psxgpu_device::vblank(screen_device &screen, bool vblank_state) { if( vblank_state ) { #if PSXGPU_DEBUG_VIEWER DebugCheckKeys(); #endif n_gpustatus ^= ( 1L << 31 ); m_vblank_handler(1); } } void psxgpu_device::gpu_reset() { verboselog( *this, 1, "reset gpu\n" ); n_gpu_buffer_offset = 0; n_gpustatus = 0x14802000; n_drawarea_x1 = 0; n_drawarea_y1 = 0; n_drawarea_x2 = 1023; n_drawarea_y2 = 1023; n_drawoffset_x = 0; n_drawoffset_y = 0; m_n_displaystartx = 0; n_displaystarty = 0; n_horiz_disstart = 0x260; n_horiz_disend = 0xc60; n_vert_disstart = 0x010; n_vert_disend = 0x100; n_vramx = 0; n_vramy = 0; n_twx = 0; n_twy = 0; n_twh = 255; n_tww = 255; updatevisiblearea(); } void psxgpu_device::lightgun_set( int n_x, int n_y ) { n_lightgun_x = n_x; n_lightgun_y = n_y; } //------------------------------------------------- // device_config_complete - perform any // operations now that the configuration is // complete //------------------------------------------------- void psxgpu_device::device_config_complete() { if (!has_screen()) return; if (!screen().refresh_attoseconds()) { screen().set_refresh_hz(60); screen().set_vblank_time(ATTOSECONDS_IN_USEC(2500) /* not accurate */); screen().set_size(1024, 1024); screen().set_visarea(0, 639, 0, 479); } if (!screen().has_screen_update()) screen().set_screen_update(*this, FUNC(psxgpu_device::update_screen)); }