diff --git a/README.md b/README.md index b4b4142be..ce38632e6 100644 --- a/README.md +++ b/README.md @@ -53,10 +53,14 @@ We operate an IRC channel and a Discord server for discussing 86Box, its develop [![Visit our Discord server](https://discordapp.com/api/guilds/262614059009048590/embed.png)](https://discord.gg/QXK9XTv) Contributions ---------- +------------- We welcome all contributions to the project, as long as the [contribution guidelines](CONTRIBUTING.md) are followed. +Building +--------- +For instructions on how to build 86Box from source, see the [build guide](https://86box.readthedocs.io/en/latest/dev/buildguide.html). + Licensing --------- diff --git a/scripts/86Box-install-roms.sh b/scripts/86Box-install-roms.sh deleted file mode 100755 index b13d93acc..000000000 --- a/scripts/86Box-install-roms.sh +++ /dev/null @@ -1,73 +0,0 @@ -#!/bin/sh - -URL="https://github.com/86Box/roms/archive/refs/tags/v4.2.zip" -TMP_FILE="/tmp/86Box-ROMS.zip" -EXTRACT_DIR="/tmp/86Box-ROMS-extracted" -DEFAULT_TARGET_DIR="$HOME/.local/share/86Box/roms/" -TARGET_DIR=${TARGET_DIR:-$DEFAULT_TARGET_DIR} - -install_roms() { - if [ -d "$TARGET_DIR" ] && [ "$(ls -A $TARGET_DIR)" ]; then - echo "ROMS already installed in $TARGET_DIR" - echo "To (re)install, please first remove ROMS with -r parameter" - exit 1 - fi - fetch -o "$TMP_FILE" "$URL" - - if [ $? -ne 0 ]; then - echo "Failed to download the file from $URL" - exit 1 - fi - - mkdir -p "$EXTRACT_DIR" - unzip "$TMP_FILE" -d "$EXTRACT_DIR" - - if [ $? -ne 0 ]; then - echo "Failed to decompress the file" - rm "$TMP_FILE" - exit 1 - fi - - mkdir -p "$TARGET_DIR" - cd "$EXTRACT_DIR" - TOP_LEVEL_DIR=$(find . -mindepth 1 -maxdepth 1 -type d) - - if [ -d "$TOP_LEVEL_DIR" ]; then - mv "$TOP_LEVEL_DIR"/* "$TARGET_DIR" - fi - - rm -rf "$TMP_FILE" "$EXTRACT_DIR" - echo "ROMS installed successfully in $TARGET_DIR" -} - -remove_roms() { - if [ -d "$TARGET_DIR" ]; then - rm -rf "$TARGET_DIR" - echo "ROMS removed successfully from $TARGET_DIR" - else - echo "No ROMS directory found in $TARGET_DIR" - fi -} - -help() { - echo "" - echo "$0 [-h|-i|-r]" - echo " -h : this help" - echo " -i : install (this parameter can be omitted)" - echo " -r : remove the ROMS" - echo "" -} - -case "$1" in - -h) - help - ;; - -r) - remove_roms - ;; - -i|*) - install_roms - ;; -esac - -exit 0 diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 0841fabdd..ad339040f 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -27,7 +27,9 @@ if(CMAKE_SYSTEM_NAME MATCHES "Linux") add_compile_definitions(_FILE_OFFSET_BITS=64 _LARGEFILE_SOURCE=1 _LARGEFILE64_SOURCE=1) endif() -if(CPPTHREADS) +if(WIN32) + target_sources(86Box PRIVATE qt/win_thread.c) +else() target_sources(86Box PRIVATE thread.cpp) endif() diff --git a/src/device/cartridge.c b/src/device/cartridge.c index edabd3ed0..a40c29c9b 100644 --- a/src/device/cartridge.c +++ b/src/device/cartridge.c @@ -169,6 +169,7 @@ cart_close(int drive) cart_image_close(drive); cart_fns[drive][0] = 0; ui_sb_update_icon_state(SB_CARTRIDGE | drive, 1); + resetx86(); } void diff --git a/src/floppy/fdc.c b/src/floppy/fdc.c index 3fd492bd2..6baef8ab8 100644 --- a/src/floppy/fdc.c +++ b/src/floppy/fdc.c @@ -1635,6 +1635,8 @@ fdc_callback(void *priv) return; } if (fdd_get_head(real_drive(fdc, fdc->drive)) == 0) { + fdc->sector = 1; + fdc->head |= 1; fdd_set_head(real_drive(fdc, fdc->drive), 1); if (!fdd_is_double_sided(real_drive(fdc, fdc->drive))) { fdc_noidam(fdc); diff --git a/src/machine/m_pcjr.c b/src/machine/m_pcjr.c index 9ac4ff631..e569dc674 100644 --- a/src/machine/m_pcjr.c +++ b/src/machine/m_pcjr.c @@ -443,16 +443,16 @@ vid_poll(void *priv) pcjr->ma++; buffer32->line[l][ef_x] = buffer32->line[l][ef_x + 1] = buffer32->line[l + 1][ef_x] = buffer32->line[l + 1][ef_x + 1] = - pcjr->array[((dat >> 12) & pcjr->array[1]) + 16] + 16; + pcjr->array[((dat >> 12) & pcjr->array[1] & 0x0f) + 16] + 16; buffer32->line[l][ef_x + 2] = buffer32->line[l][ef_x + 3] = buffer32->line[l + 1][ef_x + 2] = buffer32->line[l + 1][ef_x + 3] = - pcjr->array[((dat >> 8) & pcjr->array[1]) + 16] + 16; + pcjr->array[((dat >> 8) & pcjr->array[1] & 0x0f) + 16] + 16; buffer32->line[l][ef_x + 4] = buffer32->line[l][ef_x + 5] = buffer32->line[l + 1][ef_x + 4] = buffer32->line[l + 1][ef_x + 5] = - pcjr->array[((dat >> 4) & pcjr->array[1]) + 16] + 16; + pcjr->array[((dat >> 4) & pcjr->array[1] & 0x0f) + 16] + 16; buffer32->line[l][ef_x + 6] = buffer32->line[l][ef_x + 7] = buffer32->line[l + 1][ef_x + 6] = buffer32->line[l + 1][ef_x + 7] = - pcjr->array[(dat & pcjr->array[1]) + 16] + 16; + pcjr->array[(dat & pcjr->array[1] & 0x0f) + 16] + 16; } break; case 0x12: /*160x200x16*/ @@ -493,7 +493,7 @@ vid_poll(void *priv) chr = (dat >> 7) & 1; chr |= ((dat >> 14) & 2); buffer32->line[l][ef_x + c] = buffer32->line[l + 1][ef_x + c] = - pcjr->array[(chr & pcjr->array[1]) + 16] + 16; + pcjr->array[(chr & pcjr->array[1] & 0x0f) + 16] + 16; dat <<= 1; } } @@ -505,13 +505,13 @@ vid_poll(void *priv) attr = pcjr->vram[((pcjr->ma << 1) & mask) + offset + 1]; drawcursor = ((pcjr->ma == ca) && pcjr->con && pcjr->cursoron); if (pcjr->array[3] & 4) { - cols[1] = pcjr->array[((attr & 15) & pcjr->array[1]) + 16] + 16; - cols[0] = pcjr->array[(((attr >> 4) & 7) & pcjr->array[1]) + 16] + 16; + cols[1] = pcjr->array[((attr & 15) & pcjr->array[1] & 0x0f) + 16] + 16; + cols[0] = pcjr->array[(((attr >> 4) & 7) & pcjr->array[1] & 0x0f) + 16] + 16; if ((pcjr->blink & 16) && (attr & 0x80) && !drawcursor) cols[1] = cols[0]; } else { - cols[1] = pcjr->array[((attr & 15) & pcjr->array[1]) + 16] + 16; - cols[0] = pcjr->array[((attr >> 4) & pcjr->array[1]) + 16] + 16; + cols[1] = pcjr->array[((attr & 15) & pcjr->array[1] & 0x0f) + 16] + 16; + cols[0] = pcjr->array[((attr >> 4) & pcjr->array[1] & 0x0f) + 16] + 16; } if (pcjr->sc & 8) for (uint8_t c = 0; c < 8; c++) @@ -537,13 +537,13 @@ vid_poll(void *priv) attr = pcjr->vram[((pcjr->ma << 1) & mask) + offset + 1]; drawcursor = ((pcjr->ma == ca) && pcjr->con && pcjr->cursoron); if (pcjr->array[3] & 4) { - cols[1] = pcjr->array[((attr & 15) & pcjr->array[1]) + 16] + 16; - cols[0] = pcjr->array[(((attr >> 4) & 7) & pcjr->array[1]) + 16] + 16; + cols[1] = pcjr->array[((attr & 15) & pcjr->array[1] & 0x0f) + 16] + 16; + cols[0] = pcjr->array[(((attr >> 4) & 7) & pcjr->array[1] & 0x0f) + 16] + 16; if ((pcjr->blink & 16) && (attr & 0x80) && !drawcursor) cols[1] = cols[0]; } else { - cols[1] = pcjr->array[((attr & 15) & pcjr->array[1]) + 16] + 16; - cols[0] = pcjr->array[((attr >> 4) & pcjr->array[1]) + 16] + 16; + cols[1] = pcjr->array[((attr & 15) & pcjr->array[1] & 0x0f) + 16] + 16; + cols[0] = pcjr->array[((attr >> 4) & pcjr->array[1] & 0x0f) + 16] + 16; } pcjr->ma++; if (pcjr->sc & 8) diff --git a/src/qt/win_thread.c b/src/qt/win_thread.c new file mode 100644 index 000000000..9a4f046b3 --- /dev/null +++ b/src/qt/win_thread.c @@ -0,0 +1,179 @@ +/* + * 86Box A hypervisor and IBM PC system emulator that specializes in + * running old operating systems and software designed for IBM + * PC systems and compatibles from 1981 through fairly recent + * system designs based on the PCI bus. + * + * This file is part of the 86Box distribution. + * + * Implement threads and mutexes for the Win32 platform. + * + * + * + * Authors: Sarah Walker, + * Fred N. van Kempen, + * + * Copyright 2008-2018 Sarah Walker. + * Copyright 2017-2018 Fred N. van Kempen. + */ +#define UNICODE +#define BITMAP WINDOWS_BITMAP +#include +#include +#include +#undef BITMAP +#include +#include +#include +#include +#include +#include <86box/86box.h> +#include <86box/plat.h> +#include <86box/thread.h> + +typedef struct { + HANDLE handle; +} win_event_t; + +/* For compatibility with thread.h, but Win32 does not allow named threads. */ +thread_t * +thread_create_named(void (*func)(void *param), void *param, UNUSED(const char *name)) +{ + uintptr_t bt = _beginthread(func, 0, param); + return ((thread_t *) bt); +} + +int +thread_test_mutex(thread_t *arg) +{ + if (arg == NULL) + return (0); + + return (WaitForSingleObject(arg, 0) == WAIT_OBJECT_0) ? 1 : 0; +} + +int +thread_wait(thread_t *arg) +{ + if (arg == NULL) + return (0); + + if (WaitForSingleObject(arg, INFINITE)) + return (1); + + return (0); +} + +event_t * +thread_create_event(void) +{ + win_event_t *ev = malloc(sizeof(win_event_t)); + + ev->handle = CreateEvent(NULL, FALSE, FALSE, NULL); + + return ((event_t *) ev); +} + +void +thread_set_event(event_t *arg) +{ + win_event_t *ev = (win_event_t *) arg; + + if (arg == NULL) + return; + + SetEvent(ev->handle); +} + +void +thread_reset_event(event_t *arg) +{ + win_event_t *ev = (win_event_t *) arg; + + if (arg == NULL) + return; + + ResetEvent(ev->handle); +} + +int +thread_wait_event(event_t *arg, int timeout) +{ + win_event_t *ev = (win_event_t *) arg; + + if (arg == NULL) + return (0); + + if (ev->handle == NULL) + return (0); + + if (timeout == -1) + timeout = INFINITE; + + if (WaitForSingleObject(ev->handle, timeout)) + return (1); + + return (0); +} + +void +thread_destroy_event(event_t *arg) +{ + win_event_t *ev = (win_event_t *) arg; + + if (arg == NULL) + return; + + CloseHandle(ev->handle); + + free(ev); +} + +mutex_t * +thread_create_mutex(void) +{ + mutex_t *mutex = malloc(sizeof(CRITICAL_SECTION)); + + InitializeCriticalSection(mutex); + + return mutex; +} + +int +thread_wait_mutex(mutex_t *mutex) +{ + if (mutex == NULL) + return (0); + + LPCRITICAL_SECTION critsec = (LPCRITICAL_SECTION) mutex; + + EnterCriticalSection(critsec); + + return 1; +} + +int +thread_release_mutex(mutex_t *mutex) +{ + if (mutex == NULL) + return (0); + + LPCRITICAL_SECTION critsec = (LPCRITICAL_SECTION) mutex; + + LeaveCriticalSection(critsec); + + return 1; +} + +void +thread_close_mutex(mutex_t *mutex) +{ + if (mutex == NULL) + return; + + LPCRITICAL_SECTION critsec = (LPCRITICAL_SECTION) mutex; + + DeleteCriticalSection(critsec); + + free(critsec); +} diff --git a/src/video/vid_s3.c b/src/video/vid_s3.c index d2b5a87fe..8a9bedf02 100644 --- a/src/video/vid_s3.c +++ b/src/video/vid_s3.c @@ -9930,8 +9930,14 @@ s3_init(const device_t *info) s3->width = 1024; svga->ramdac = device_add(&sc11483_ramdac_device); - svga->clock_gen = device_add(&av9194_device); - svga->getclock = av9194_getclock; + if (s3->card_type == S3_ORCHID_86C911) { + svga->clock_gen = device_add(&av9194_device); + svga->getclock = av9194_getclock; + } else { + /* DCS2824-0 = Diamond ICD2061A-compatible. */ + svga->clock_gen = device_add(&icd2061_device); + svga->getclock = icd2061_getclock; + } break; case S3_AMI_86C924: diff --git a/src/video/vid_s3_virge.c b/src/video/vid_s3_virge.c index 700722253..a5448fa66 100644 --- a/src/video/vid_s3_virge.c +++ b/src/video/vid_s3_virge.c @@ -68,7 +68,7 @@ static int dither[4][4] = { #define ROM_DIAMOND_STEALTH3D_2000PRO "roms/video/s3virge/virgedxdiamond.vbi" #define ROM_VIRGE_GX "roms/video/s3virge/86c375_4.bin" #define ROM_VIRGE_GX2 "roms/video/s3virge/flagpoint.VBI" -#define ROM_DIAMOND_STEALTH3D_4000 "roms/video/s3virge/86c357.bin" +#define ROM_DIAMOND_STEALTH3D_4000 "roms/video/s3virge/DS3D4K v1.03 Brightness bug fix.bin" #define ROM_TRIO3D2X "roms/video/s3virge/TRIO3D2X_8mbsdr.VBI" #define RB_SIZE 256 @@ -288,9 +288,9 @@ typedef struct virge_t { s3d_t s3d_tri; s3d_t s3d_buffer[RB_SIZE]; - int s3d_read_idx; - int s3d_write_idx; - int s3d_busy; + atomic_int s3d_read_idx; + atomic_int s3d_write_idx; + atomic_int s3d_busy; struct { uint32_t pri_ctrl; @@ -330,15 +330,14 @@ typedef struct virge_t { } streams; fifo_entry_t fifo[FIFO_SIZE]; - volatile int fifo_read_idx, fifo_write_idx; - - volatile int fifo_thread_run, render_thread_run; + atomic_int fifo_read_idx, fifo_write_idx; + atomic_int fifo_thread_run, render_thread_run; thread_t * fifo_thread; event_t *wake_fifo_thread; event_t * fifo_not_full_event; - int virge_busy; + atomic_int virge_busy; uint8_t subsys_stat; uint8_t subsys_cntl; @@ -363,7 +362,16 @@ typedef struct virge_t { uint8_t cmd_dma; uint32_t cmd_dma_base; - uint32_t dma_ptr; + uint32_t cmd_dma_buf_size; + uint32_t cmd_dma_buf_size_mask; + uint32_t cmd_base_addr; + uint32_t cmd_dma_write_ptr_reg; + uint32_t cmd_dma_write_ptr_update; + uint32_t cmd_dma_read_ptr_reg; + uint32_t dma_val; + uint32_t dma_dbl_words; + uint32_t dma_mmio_addr; + uint32_t dma_data_type; int pci; int is_agp; @@ -396,6 +404,8 @@ static void s3_virge_mmio_write(uint32_t addr, uint8_t val, void *priv); static void s3_virge_mmio_write_w(uint32_t addr, uint16_t val, void *priv); static void s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv); +static void s3_virge_queue(virge_t *virge, uint32_t addr, uint32_t val, uint32_t type); + enum { CMD_SET_AE = 1, CMD_SET_HC = (1 << 1), @@ -442,6 +452,8 @@ enum { #define INT_S3D_DONE (1 << 1) #define INT_FIFO_OVF (1 << 2) #define INT_FIFO_EMP (1 << 3) +#define INT_HOST_DONE (1 << 4) +#define INT_CMD_DONE (1 << 5) #define INT_3DF_EMP (1 << 6) #define INT_MASK 0xff @@ -722,12 +734,12 @@ s3_virge_in(uint16_t addr, void *priv) { case 0x31: ret = (svga->crtc[0x31] & 0xcf) | ((virge->ma_ext & 3) << 4); break; + case 0x33: + ret = (svga->crtc[0x33] | 0x04); + break; case 0x35: ret = (svga->crtc[0x35] & 0xf0) | (virge->bank & 0xf); break; - case 0x36: - ret = (svga->crtc[0x36] & 0xfc) | 2; - break; /*PCI bus*/ case 0x45: virge->hwc_col_stack_pos = 0; ret = svga->crtc[0x45]; @@ -782,7 +794,8 @@ s3_virge_in(uint16_t addr, void *priv) { } static void -s3_virge_recalctimings(svga_t *svga) { +s3_virge_recalctimings(svga_t *svga) +{ virge_t *virge = (virge_t *) svga->priv; svga->hdisp = svga->hdisp_old; @@ -854,7 +867,7 @@ s3_virge_recalctimings(svga_t *svga) { video_force_resize_set_monitor(1, svga->monitor_index); } - if ((svga->crtc[0x67] & 0xc) != 0xc) { /*VGA mode*/ + if (((svga->crtc[0x67] & 0xc) != 0xc) || (virge->chip >= S3_VIRGEGX2)) { /*VGA mode*/ svga->ma_latch |= (virge->ma_ext << 16); if (svga->crtc[0x51] & 0x30) svga->rowoffset |= (svga->crtc[0x51] & 0x30) << 4; @@ -891,8 +904,8 @@ s3_virge_recalctimings(svga_t *svga) { svga->render = svga_render_32bpp_highres; break; - default: - break; + default: + break; } svga->vram_display_mask = (!(svga->crtc[0x31] & 0x08) && @@ -1034,7 +1047,7 @@ s3_virge_updatemapping(virge_t *virge) { virge->linear_base = (svga->crtc[0x5a] << 16) | (svga->crtc[0x59] << 24); if ((svga->crtc[0x58] & 0x10) || (virge->advfunc_cntl & 0x10)) { /*Linear framebuffer*/ - switch (svga->crtc[0x58] & 3) { + switch (svga->crtc[0x58] & 7) { case 0: /*64k*/ virge->linear_size = 0x10000; break; @@ -1103,7 +1116,8 @@ s3_virge_wait_fifo_idle(virge_t *virge) { } static uint8_t -s3_virge_mmio_read(uint32_t addr, void *priv) { +s3_virge_mmio_read(uint32_t addr, void *priv) +{ virge_t *virge = (virge_t *) priv; uint8_t ret; @@ -1120,6 +1134,7 @@ s3_virge_mmio_read(uint32_t addr, void *priv) { ret |= 0x10; else ret |= 0x30; + if (!virge->virge_busy) wake_fifo_thread(virge); return ret; @@ -1152,7 +1167,8 @@ s3_virge_mmio_read(uint32_t addr, void *priv) { } static uint16_t -s3_virge_mmio_read_w(uint32_t addr, void *priv) { +s3_virge_mmio_read_w(uint32_t addr, void *priv) +{ virge_t *virge = (virge_t *) priv; uint16_t ret; @@ -1186,7 +1202,6 @@ s3_virge_mmio_read_w(uint32_t addr, void *priv) { static uint32_t s3_virge_mmio_read_l(uint32_t addr, void *priv) { virge_t *virge = (virge_t *) priv; - svga_t *svga = &virge->svga; uint32_t ret = 0xffffffff; switch (addr & 0xfffc) { @@ -1207,46 +1222,30 @@ s3_virge_mmio_read_l(uint32_t addr, void *priv) { break; case 0x81a0: ret = virge->streams.blend_ctrl; - svga_recalctimings(svga); break; case 0x81c0: ret = virge->streams.pri_fb0; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81c4: ret = virge->streams.pri_fb1; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81c8: ret = virge->streams.pri_stride; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81cc: ret = virge->streams.buffer_ctrl; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81d0: ret = virge->streams.sec_fb0; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81d4: ret = virge->streams.sec_fb1; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81d8: ret = virge->streams.sec_stride; - s3_virge_update_buffer(virge); - svga->fullchange = changeframecount; break; case 0x81dc: ret = virge->streams.overlay_ctrl; - svga->fullchange = changeframecount; break; case 0x81e0: ret = virge->streams.k1_vert_scale; @@ -1279,6 +1278,7 @@ s3_virge_mmio_read_l(uint32_t addr, void *priv) { ret |= 0x00001000; else ret |= 0x00003000; + ret |= virge->subsys_stat; if (!virge->virge_busy) wake_fifo_thread(virge); @@ -1293,12 +1293,15 @@ s3_virge_mmio_read_l(uint32_t addr, void *priv) { ret = virge->cmd_dma_base; break; case 0x8594: + ret = virge->cmd_dma_write_ptr_reg; break; case 0x8598: - ret = virge->dma_ptr; + ret = virge->cmd_dma_read_ptr_reg; + if (ret > virge->cmd_dma_write_ptr_reg) + ret = virge->cmd_dma_write_ptr_reg; break; case 0x859c: - ret = virge->cmd_dma; + ret = 0; /*To prevent DMA overflows.*/ break; case 0xa4d4: @@ -1425,13 +1428,322 @@ s3_virge_mmio_read_l(uint32_t addr, void *priv) { default: ret = s3_virge_mmio_read_w(addr, priv) | (s3_virge_mmio_read_w(addr + 2, priv) << 16); + break; } + //pclog("MMIO ReadL=%04x, ret=%08x.\n", addr & 0xfffc, ret); return ret; } static void -fifo_thread(void *param) { +s3_virge_mmio_write_fifo_l(virge_t *virge, uint32_t addr, uint32_t val) +{ + switch (addr) { + case 0xa000 ... 0xa1fc: + { + int x = addr & 4; + int y = (addr >> 3) & 7; + int color; + int byte; + uint32_t newaddr = addr; + virge->s3d.pattern_8[y * 8 + x] = val & 0xff; + virge->s3d.pattern_8[y * 8 + x + 1] = val >> 8; + virge->s3d.pattern_8[y * 8 + x + 2] = val >> 16; + virge->s3d.pattern_8[y * 8 + x + 3] = val >> 24; + + x = (addr >> 1) & 6; + y = (addr >> 4) & 7; + virge->s3d.pattern_16[y * 8 + x] = val & 0xffff; + virge->s3d.pattern_16[y * 8 + x + 1] = val >> 16; + + newaddr &= 0x00ff; + for (uint8_t xx = 0; xx < 4; xx++) { + x = ((newaddr + xx) / 3) % 8; + y = ((newaddr + xx) / 24) % 8; + color = ((newaddr + xx) % 3) << 3; + byte = (xx << 3); + virge->s3d.pattern_24[y * 8 + x] &= ~(0xff << color); + virge->s3d.pattern_24[y * 8 + x] |= ((val >> byte) & 0xff) << color; + } + + x = (addr >> 2) & 7; + y = (addr >> 5) & 7; + virge->s3d.pattern_32[y * 8 + x] = val & 0xffffff; + } break; + + case 0xa4d4: + case 0xa8d4: + case 0xacd4: + virge->s3d.src_base = val & ((virge->memory_size == 8) ? + (val & 0x7ffff8) : (val & 0x3ffff8)); + break; + case 0xa4d8: + case 0xa8d8: + case 0xacd8: + virge->s3d.dest_base = val & ((virge->memory_size == 8) ? + (val & 0x7ffff8) : (val & 0x3ffff8)); + break; + case 0xa4dc: + case 0xa8dc: + case 0xacdc: + virge->s3d.clip_l = (val >> 16) & 0x7ff; + virge->s3d.clip_r = val & 0x7ff; + break; + case 0xa4e0: + case 0xa8e0: + case 0xace0: + virge->s3d.clip_t = (val >> 16) & 0x7ff; + virge->s3d.clip_b = val & 0x7ff; + break; + case 0xa4e4: + case 0xa8e4: + case 0xace4: + virge->s3d.dest_str = (val >> 16) & 0xff8; + virge->s3d.src_str = val & 0xff8; + break; + case 0xa4e8: + case 0xace8: + virge->s3d.mono_pat_0 = val; + break; + case 0xa4ec: + case 0xacec: + virge->s3d.mono_pat_1 = val; + break; + case 0xa4f0: + case 0xacf0: + virge->s3d.pat_bg_clr = val; + break; + case 0xa4f4: + case 0xa8f4: + case 0xacf4: + virge->s3d.pat_fg_clr = val; + break; + case 0xa4f8: + virge->s3d.src_bg_clr = val; + break; + case 0xa4fc: + virge->s3d.src_fg_clr = val; + break; + case 0xa500: + case 0xa900: + case 0xad00: + virge->s3d.cmd_set = val; + if (!(val & CMD_SET_AE)) + s3_virge_bitblt(virge, -1, 0); + break; + case 0xa504: + virge->s3d.r_width = (val >> 16) & 0x7ff; + virge->s3d.r_height = val & 0x7ff; + break; + case 0xa508: + virge->s3d.rsrc_x = (val >> 16) & 0x7ff; + virge->s3d.rsrc_y = val & 0x7ff; + break; + case 0xa50c: + virge->s3d.rdest_x = (val >> 16) & 0x7ff; + virge->s3d.rdest_y = val & 0x7ff; + if (virge->s3d.cmd_set & CMD_SET_AE) + s3_virge_bitblt(virge, -1, 0); + break; + case 0xa96c: + virge->s3d.lxend0 = (val >> 16) & 0x7ff; + virge->s3d.lxend1 = val & 0x7ff; + break; + case 0xa970: + virge->s3d.ldx = (int32_t)val; + break; + case 0xa974: + virge->s3d.lxstart = val; + break; + case 0xa978: + virge->s3d.lystart = val & 0x7ff; + break; + case 0xa97c: + virge->s3d.lycnt = val & 0x7ff; + virge->s3d.line_dir = val >> 31; + if (virge->s3d.cmd_set & CMD_SET_AE) + s3_virge_bitblt(virge, -1, 0); + break; + + case 0xad68: + virge->s3d.prdx = val; + break; + case 0xad6c: + virge->s3d.prxstart = val; + break; + case 0xad70: + virge->s3d.pldx = val; + break; + case 0xad74: + virge->s3d.plxstart = val; + break; + case 0xad78: + virge->s3d.pystart = val & 0x7ff; + break; + case 0xad7c: + virge->s3d.pycnt = val & 0x300007ff; + if (virge->s3d.cmd_set & CMD_SET_AE) + s3_virge_bitblt(virge, -1, 0); + break; + + case 0xb0d4: + case 0xb4d4: + virge->s3d_tri.z_base = val & ((virge->memory_size == 8) ? + (val & 0x7ffff8) : (val & 0x3ffff8)); + break; + case 0xb0d8: + case 0xb4d8: + virge->s3d_tri.dest_base = val & ((virge->memory_size == 8) ? + (val & 0x7ffff8) : (val & 0x3ffff8)); + break; + case 0xb0dc: + case 0xb4dc: + virge->s3d_tri.clip_l = (val >> 16) & 0x7ff; + virge->s3d_tri.clip_r = val & 0x7ff; + break; + case 0xb0e0: + case 0xb4e0: + virge->s3d_tri.clip_t = (val >> 16) & 0x7ff; + virge->s3d_tri.clip_b = val & 0x7ff; + break; + case 0xb0e4: + case 0xb4e4: + virge->s3d_tri.dest_str = (val >> 16) & 0xff8; + virge->s3d.src_str = val & 0xff8; + break; + case 0xb0e8: + case 0xb4e8: + virge->s3d_tri.z_str = val & 0xff8; + break; + case 0xb4ec: + virge->s3d_tri.tex_base = val & ((virge->memory_size == 8) ? + (val & 0x7ffff8) : (val & 0x3ffff8)); + break; + case 0xb4f0: + virge->s3d_tri.tex_bdr_clr = val & 0xffffff; + break; + case 0xb0f4: + case 0xb4f4: + virge->s3d_tri.fog_b = val & 0xff; + virge->s3d_tri.fog_g = (val >> 8) & 0xff; + virge->s3d_tri.fog_r = (val >> 16) & 0xff; + break; + case 0xb100: + case 0xb500: + virge->s3d_tri.cmd_set = val; + if (!(val & CMD_SET_AE)) + queue_triangle(virge); + break; + case 0xb504: + virge->s3d_tri.tbv = val & 0xfffff; + break; + case 0xb508: + virge->s3d_tri.tbu = val & 0xfffff; + break; + case 0xb50c: + virge->s3d_tri.TdWdX = val; + break; + case 0xb510: + virge->s3d_tri.TdWdY = val; + break; + case 0xb514: + virge->s3d_tri.tws = val; + break; + case 0xb518: + virge->s3d_tri.TdDdX = val; + break; + case 0xb51c: + virge->s3d_tri.TdVdX = val; + break; + case 0xb520: + virge->s3d_tri.TdUdX = val; + break; + case 0xb524: + virge->s3d_tri.TdDdY = val; + break; + case 0xb528: + virge->s3d_tri.TdVdY = val; + break; + case 0xb52c: + virge->s3d_tri.TdUdY = val; + break; + case 0xb530: + virge->s3d_tri.tds = val; + break; + case 0xb534: + virge->s3d_tri.tvs = val; + break; + case 0xb538: + virge->s3d_tri.tus = val; + break; + case 0xb53c: + virge->s3d_tri.TdGdX = val >> 16; + virge->s3d_tri.TdBdX = val & 0xffff; + break; + case 0xb540: + virge->s3d_tri.TdAdX = val >> 16; + virge->s3d_tri.TdRdX = val & 0xffff; + break; + case 0xb544: + virge->s3d_tri.TdGdY = val >> 16; + virge->s3d_tri.TdBdY = val & 0xffff; + break; + case 0xb548: + virge->s3d_tri.TdAdY = val >> 16; + virge->s3d_tri.TdRdY = val & 0xffff; + break; + case 0xb54c: + virge->s3d_tri.tgs = (val >> 16) & 0xffff; + virge->s3d_tri.tbs = val & 0xffff; + break; + case 0xb550: + virge->s3d_tri.tas = (val >> 16) & 0xffff; + virge->s3d_tri.trs = val & 0xffff; + break; + + case 0xb554: + virge->s3d_tri.TdZdX = val; + break; + case 0xb558: + virge->s3d_tri.TdZdY = val; + break; + case 0xb55c: + virge->s3d_tri.tzs = val; + break; + case 0xb560: + virge->s3d_tri.TdXdY12 = val; + break; + case 0xb564: + virge->s3d_tri.txend12 = val; + break; + case 0xb568: + virge->s3d_tri.TdXdY01 = val; + break; + case 0xb56c: + virge->s3d_tri.txend01 = val; + break; + case 0xb570: + virge->s3d_tri.TdXdY02 = val; + break; + case 0xb574: + virge->s3d_tri.txs = val; + break; + case 0xb578: + virge->s3d_tri.tys = val; + break; + case 0xb57c: + virge->s3d_tri.ty01 = (val >> 16) & 0x7ff; + virge->s3d_tri.ty12 = val & 0x7ff; + virge->s3d_tri.tlr = val >> 31; + if (virge->s3d_tri.cmd_set & CMD_SET_AE) + queue_triangle(virge); + break; + } +} + +static void +fifo_thread(void *param) +{ virge_t *virge = (virge_t *)param; while (virge->fifo_thread_run) { @@ -1447,19 +1759,16 @@ fifo_thread(void *param) { switch (fifo->addr_type & FIFO_TYPE) { case FIFO_WRITE_BYTE: - if (((fifo->addr_type & FIFO_ADDR) & 0xfffc) < 0x8000) + if (((fifo->addr_type & FIFO_ADDR) & 0xffff) < 0x8000) s3_virge_bitblt(virge, 8, val); - else if (((fifo->addr_type & FIFO_ADDR) & 0xffff) == 0x859c) - virge->cmd_dma = val; break; case FIFO_WRITE_WORD: - if (((fifo->addr_type & FIFO_ADDR) & 0xfffc) < 0x8000) { + if (((fifo->addr_type & FIFO_ADDR) & 0xfffe) < 0x8000) { if (virge->s3d.cmd_set & CMD_SET_MS) s3_virge_bitblt(virge, 16, ((val >> 8) | (val << 8)) << 16); else - s3_virge_bitblt(virge, 16, val); - } else if (((fifo->addr_type & FIFO_ADDR) & 0xfffe) == 0x859c) - virge->cmd_dma = val; + s3_virge_bitblt(virge, 16, val); + } break; case FIFO_WRITE_DWORD: if (((fifo->addr_type & FIFO_ADDR) & 0xfffc) < 0x8000) { @@ -1469,324 +1778,8 @@ fifo_thread(void *param) { ((val & 0x0000ff00) << 8) | ((val & 0x000000ff) << 24)); else s3_virge_bitblt(virge, 32, val); - } else { - switch ((fifo->addr_type & FIFO_ADDR) & 0xfffc) { - case 0x8590: - virge->cmd_dma_base = val; - break; - - case 0x8594: - virge->dma_ptr = val; - break; - - case 0x8598: - break; - - case 0x859c: - virge->cmd_dma = val; - break; - - case 0xa000 ... 0xa1fc: - { - int x = (fifo->addr_type & FIFO_ADDR) & 4; - int y = ((fifo->addr_type & FIFO_ADDR) >> 3) & 7; - int color; - int byte; - uint32_t addr = (fifo->addr_type & FIFO_ADDR); - virge->s3d.pattern_8[y * 8 + x] = val & 0xff; - virge->s3d.pattern_8[y * 8 + x + 1] = val >> 8; - virge->s3d.pattern_8[y * 8 + x + 2] = val >> 16; - virge->s3d.pattern_8[y * 8 + x + 3] = val >> 24; - - x = ((fifo->addr_type & FIFO_ADDR) >> 1) & 6; - y = ((fifo->addr_type & FIFO_ADDR) >> 4) & 7; - virge->s3d.pattern_16[y * 8 + x] = val & 0xffff; - virge->s3d.pattern_16[y * 8 + x + 1] = val >> 16; - - addr &= 0x00ff; - for (uint8_t xx = 0; xx < 4; xx++) { - x = ((addr + xx) / 3) % 8; - y = ((addr + xx) / 24) % 8; - color = ((addr + xx) % 3) << 3; - byte = (xx << 3); - virge->s3d.pattern_24[y * 8 + x] &= ~(0xff << color); - virge->s3d.pattern_24[y * 8 + x] |= ((val >> byte) & 0xff) << color; - } - - x = ((fifo->addr_type & FIFO_ADDR) >> 2) & 7; - y = ((fifo->addr_type & FIFO_ADDR) >> 5) & 7; - virge->s3d.pattern_32[y * 8 + x] = val & 0xffffff; - } break; - - case 0xa4d4: - case 0xa8d4: - case 0xacd4: - virge->s3d.src_base = val & ((virge->memory_size == 8) ? - (val & 0x7ffff8) : (val & 0x3ffff8)); - break; - case 0xa4d8: - case 0xa8d8: - case 0xacd8: - virge->s3d.dest_base = val & ((virge->memory_size == 8) ? - (val & 0x7ffff8) : (val & 0x3ffff8)); - break; - case 0xa4dc: - case 0xa8dc: - case 0xacdc: - virge->s3d.clip_l = (val >> 16) & 0x7ff; - virge->s3d.clip_r = val & 0x7ff; - break; - case 0xa4e0: - case 0xa8e0: - case 0xace0: - virge->s3d.clip_t = (val >> 16) & 0x7ff; - virge->s3d.clip_b = val & 0x7ff; - break; - case 0xa4e4: - case 0xa8e4: - case 0xace4: - virge->s3d.dest_str = (val >> 16) & 0xff8; - virge->s3d.src_str = val & 0xff8; - break; - case 0xa4e8: - case 0xace8: - virge->s3d.mono_pat_0 = val; - break; - case 0xa4ec: - case 0xacec: - virge->s3d.mono_pat_1 = val; - break; - case 0xa4f0: - case 0xacf0: - virge->s3d.pat_bg_clr = val; - break; - case 0xa4f4: - case 0xa8f4: - case 0xacf4: - virge->s3d.pat_fg_clr = val; - break; - case 0xa4f8: - virge->s3d.src_bg_clr = val; - break; - case 0xa4fc: - virge->s3d.src_fg_clr = val; - break; - case 0xa500: - case 0xa900: - case 0xad00: - virge->s3d.cmd_set = val; - if (!(val & CMD_SET_AE)) - s3_virge_bitblt(virge, -1, 0); - break; - case 0xa504: - virge->s3d.r_width = (val >> 16) & 0x7ff; - virge->s3d.r_height = val & 0x7ff; - break; - case 0xa508: - virge->s3d.rsrc_x = (val >> 16) & 0x7ff; - virge->s3d.rsrc_y = val & 0x7ff; - break; - case 0xa50c: - virge->s3d.rdest_x = (val >> 16) & 0x7ff; - virge->s3d.rdest_y = val & 0x7ff; - if (virge->s3d.cmd_set & CMD_SET_AE) - s3_virge_bitblt(virge, -1, 0); - break; - case 0xa96c: - virge->s3d.lxend0 = (val >> 16) & 0x7ff; - virge->s3d.lxend1 = val & 0x7ff; - break; - case 0xa970: - virge->s3d.ldx = (int32_t)val; - break; - case 0xa974: - virge->s3d.lxstart = val; - break; - case 0xa978: - virge->s3d.lystart = val & 0x7ff; - break; - case 0xa97c: - virge->s3d.lycnt = val & 0x7ff; - virge->s3d.line_dir = val >> 31; - if (virge->s3d.cmd_set & CMD_SET_AE) - s3_virge_bitblt(virge, -1, 0); - break; - - case 0xad68: - virge->s3d.prdx = val; - break; - case 0xad6c: - virge->s3d.prxstart = val; - break; - case 0xad70: - virge->s3d.pldx = val; - break; - case 0xad74: - virge->s3d.plxstart = val; - break; - case 0xad78: - virge->s3d.pystart = val & 0x7ff; - break; - case 0xad7c: - virge->s3d.pycnt = val & 0x300007ff; - if (virge->s3d.cmd_set & CMD_SET_AE) - s3_virge_bitblt(virge, -1, 0); - break; - - case 0xb0d4: - case 0xb4d4: - virge->s3d_tri.z_base = val & ((virge->memory_size == 8) ? - (val & 0x7ffff8) : (val & 0x3ffff8)); - break; - case 0xb0d8: - case 0xb4d8: - virge->s3d_tri.dest_base = val & ((virge->memory_size == 8) ? - (val & 0x7ffff8) : (val & 0x3ffff8)); - break; - case 0xb0dc: - case 0xb4dc: - virge->s3d_tri.clip_l = (val >> 16) & 0x7ff; - virge->s3d_tri.clip_r = val & 0x7ff; - break; - case 0xb0e0: - case 0xb4e0: - virge->s3d_tri.clip_t = (val >> 16) & 0x7ff; - virge->s3d_tri.clip_b = val & 0x7ff; - break; - case 0xb0e4: - case 0xb4e4: - virge->s3d_tri.dest_str = (val >> 16) & 0xff8; - virge->s3d.src_str = val & 0xff8; - break; - case 0xb0e8: - case 0xb4e8: - virge->s3d_tri.z_str = val & 0xff8; - break; - case 0xb4ec: - virge->s3d_tri.tex_base = val & ((virge->memory_size == 8) ? - (val & 0x7ffff8) : (val & 0x3ffff8)); - break; - case 0xb4f0: - virge->s3d_tri.tex_bdr_clr = val & 0xffffff; - break; - case 0xb0f4: - case 0xb4f4: - virge->s3d_tri.fog_b = val & 0xff; - virge->s3d_tri.fog_g = (val >> 8) & 0xff; - virge->s3d_tri.fog_r = (val >> 16) & 0xff; - break; - case 0xb100: - case 0xb500: - virge->s3d_tri.cmd_set = val; - if (!(val & CMD_SET_AE)) - queue_triangle(virge); - break; - case 0xb504: - virge->s3d_tri.tbv = val & 0xfffff; - break; - case 0xb508: - virge->s3d_tri.tbu = val & 0xfffff; - break; - case 0xb50c: - virge->s3d_tri.TdWdX = val; - break; - case 0xb510: - virge->s3d_tri.TdWdY = val; - break; - case 0xb514: - virge->s3d_tri.tws = val; - break; - case 0xb518: - virge->s3d_tri.TdDdX = val; - break; - case 0xb51c: - virge->s3d_tri.TdVdX = val; - break; - case 0xb520: - virge->s3d_tri.TdUdX = val; - break; - case 0xb524: - virge->s3d_tri.TdDdY = val; - break; - case 0xb528: - virge->s3d_tri.TdVdY = val; - break; - case 0xb52c: - virge->s3d_tri.TdUdY = val; - break; - case 0xb530: - virge->s3d_tri.tds = val; - break; - case 0xb534: - virge->s3d_tri.tvs = val; - break; - case 0xb538: - virge->s3d_tri.tus = val; - break; - case 0xb53c: - virge->s3d_tri.TdGdX = val >> 16; - virge->s3d_tri.TdBdX = val & 0xffff; - break; - case 0xb540: - virge->s3d_tri.TdAdX = val >> 16; - virge->s3d_tri.TdRdX = val & 0xffff; - break; - case 0xb544: - virge->s3d_tri.TdGdY = val >> 16; - virge->s3d_tri.TdBdY = val & 0xffff; - break; - case 0xb548: - virge->s3d_tri.TdAdY = val >> 16; - virge->s3d_tri.TdRdY = val & 0xffff; - break; - case 0xb54c: - virge->s3d_tri.tgs = (val >> 16) & 0xffff; - virge->s3d_tri.tbs = val & 0xffff; - break; - case 0xb550: - virge->s3d_tri.tas = (val >> 16) & 0xffff; - virge->s3d_tri.trs = val & 0xffff; - break; - - case 0xb554: - virge->s3d_tri.TdZdX = val; - break; - case 0xb558: - virge->s3d_tri.TdZdY = val; - break; - case 0xb55c: - virge->s3d_tri.tzs = val; - break; - case 0xb560: - virge->s3d_tri.TdXdY12 = val; - break; - case 0xb564: - virge->s3d_tri.txend12 = val; - break; - case 0xb568: - virge->s3d_tri.TdXdY01 = val; - break; - case 0xb56c: - virge->s3d_tri.txend01 = val; - break; - case 0xb570: - virge->s3d_tri.TdXdY02 = val; - break; - case 0xb574: - virge->s3d_tri.txs = val; - break; - case 0xb578: - virge->s3d_tri.tys = val; - break; - case 0xb57c: - virge->s3d_tri.ty01 = (val >> 16) & 0x7ff; - virge->s3d_tri.ty12 = val & 0x7ff; - virge->s3d_tri.tlr = val >> 31; - if (virge->s3d_tri.cmd_set & CMD_SET_AE) - queue_triangle(virge); - break; - } - } + } else + s3_virge_mmio_write_fifo_l(virge, (fifo->addr_type & FIFO_ADDR) & 0xfffc, val); break; } @@ -1801,18 +1794,42 @@ fifo_thread(void *param) { } virge->virge_busy = 0; virge->subsys_stat |= INT_FIFO_EMP | INT_3DF_EMP; + if (virge->cmd_dma) + virge->subsys_stat |= INT_HOST_DONE | INT_CMD_DONE; + s3_virge_update_irqs(virge); } } static void -s3_virge_queue(virge_t *virge, uint32_t addr, uint32_t val, uint32_t type) { +s3_virge_queue(virge_t *virge, uint32_t addr, uint32_t val, uint32_t type) +{ fifo_entry_t *fifo = &virge->fifo[virge->fifo_write_idx & FIFO_MASK]; + int limit = 0; - if (FIFO_FULL) { - thread_reset_event(virge->fifo_not_full_event); - if (FIFO_FULL) - thread_wait_event(virge->fifo_not_full_event, -1); /*Wait for room in ringbuffer*/ + if (type == FIFO_WRITE_DWORD) { + switch (addr & 0xfffc) { + case 0xa500: + if (val & CMD_SET_AE) + limit = 1; + break; + default: + break; + } + } + + if (limit) { + if (FIFO_ENTRIES >= 16) { + thread_reset_event(virge->fifo_not_full_event); + if (FIFO_ENTRIES >= 16) + thread_wait_event(virge->fifo_not_full_event, -1); /*Wait for room in ringbuffer*/ + } + } else { + if (FIFO_FULL) { + thread_reset_event(virge->fifo_not_full_event); + if (FIFO_FULL) + thread_wait_event(virge->fifo_not_full_event, -1); /*Wait for room in ringbuffer*/ + } } fifo->val = val; @@ -1830,7 +1847,7 @@ static void s3_virge_mmio_write(uint32_t addr, uint8_t val, void *priv) { virge_t *virge = (virge_t *) priv; - if ((addr & 0xfffc) < 0x8000) + if ((addr & 0xffff) < 0x8000) s3_virge_queue(virge, addr, val, FIFO_WRITE_BYTE); else switch (addr & 0xffff) { case 0x83b0 ... 0x83df: @@ -1848,9 +1865,10 @@ static void s3_virge_mmio_write_w(uint32_t addr, uint16_t val, void *priv) { virge_t *virge = (virge_t *) priv; - if ((addr & 0xfffc) < 0x8000) + if ((addr & 0xfffe) < 0x8000) s3_virge_queue(virge, addr, val, FIFO_WRITE_WORD); else switch (addr & 0xfffe) { + default: case 0x83d4: s3_virge_mmio_write(addr, val, priv); s3_virge_mmio_write(addr + 1, val >> 8, priv); @@ -1865,144 +1883,192 @@ s3_virge_mmio_write_w(uint32_t addr, uint16_t val, void *priv) { static void s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv) { virge_t *virge = (virge_t *) priv; - svga_t * svga = &virge->svga; + svga_t *svga = &virge->svga; if ((addr & 0xfffc) < 0x8000) s3_virge_queue(virge, addr, val, FIFO_WRITE_DWORD); else if ((addr & 0xe000) == 0xa000) s3_virge_queue(virge, addr, val, FIFO_WRITE_DWORD); - else switch (addr & 0xfffc) { - case 0x8180: - virge->streams.pri_ctrl = val; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x8184: - virge->streams.chroma_ctrl = val; - break; - case 0x8190: - virge->streams.sec_ctrl = val; - virge->streams.dda_horiz_accumulator = val & 0xfff; - if (val & (1 << 11)) - virge->streams.dda_horiz_accumulator |= 0xfffff800; - virge->streams.sdif = (val >> 24) & 7; - break; - case 0x8194: - virge->streams.chroma_upper_bound = val; - break; - case 0x8198: - virge->streams.sec_filter = val; - virge->streams.k1_horiz_scale = val & 0x7ff; - if (val & (1 << 10)) - virge->streams.k1_horiz_scale |= 0xfffff800; - virge->streams.k2_horiz_scale = (val >> 16) & 0x7ff; - if ((val >> 16) & (1 << 10)) - virge->streams.k2_horiz_scale |= 0xfffff800; - break; - case 0x81a0: - virge->streams.blend_ctrl = val; - break; - case 0x81c0: - virge->streams.pri_fb0 = val & ((virge->memory_size == 8) ? - (val & 0x7fffff) : (val & 0x3fffff)); - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81c4: - virge->streams.pri_fb1 = ((virge->memory_size == 8) ? - (val & 0x7fffff) : (val & 0x3fffff)); - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81c8: - virge->streams.pri_stride = val & 0xfff; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81cc: - virge->streams.buffer_ctrl = val; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81d0: - virge->streams.sec_fb0 = val; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81d4: - virge->streams.sec_fb1 = val; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81d8: - virge->streams.sec_stride = val; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81dc: - virge->streams.overlay_ctrl = val; - break; - case 0x81e0: - virge->streams.k1_vert_scale = val & 0x7ff; - if (val & (1 << 10)) - virge->streams.k1_vert_scale |= 0xfffff800; - break; - case 0x81e4: - virge->streams.k2_vert_scale = val & 0x7ff; - if (val & (1 << 10)) - virge->streams.k2_vert_scale |= 0xfffff800; - break; - case 0x81e8: - virge->streams.dda_vert_accumulator = val & 0xfff; - if (val & (1 << 11)) - virge->streams.dda_vert_accumulator |= 0xfffff800; - break; - case 0x81ec: - virge->streams.fifo_ctrl = val; - break; - case 0x81f0: - virge->streams.pri_start = val; - virge->streams.pri_x = (val >> 16) & 0x7ff; - virge->streams.pri_y = val & 0x7ff; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81f4: - virge->streams.pri_size = val; - virge->streams.pri_w = (val >> 16) & 0x7ff; - virge->streams.pri_h = val & 0x7ff; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81f8: - virge->streams.sec_start = val; - virge->streams.sec_x = (val >> 16) & 0x7ff; - virge->streams.sec_y = val & 0x7ff; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; - case 0x81fc: - virge->streams.sec_size = val; - virge->streams.sec_w = (val >> 16) & 0x7ff; - virge->streams.sec_h = val & 0x7ff; - svga_recalctimings(svga); - svga->fullchange = changeframecount; - break; + else { + switch (addr & 0xfffc) { + case 0x8180: + virge->streams.pri_ctrl = val; + svga_recalctimings(svga); + svga->fullchange = changeframecount; + break; + case 0x8184: + virge->streams.chroma_ctrl = val; + break; + case 0x8190: + virge->streams.sec_ctrl = val; + virge->streams.dda_horiz_accumulator = val & 0xfff; + if (val & (1 << 11)) + virge->streams.dda_horiz_accumulator |= 0xfffff800; + virge->streams.sdif = (val >> 24) & 7; + break; + case 0x8194: + virge->streams.chroma_upper_bound = val; + break; + case 0x8198: + virge->streams.sec_filter = val; + virge->streams.k1_horiz_scale = val & 0x7ff; + if (val & (1 << 10)) + virge->streams.k1_horiz_scale |= 0xfffff800; + virge->streams.k2_horiz_scale = (val >> 16) & 0x7ff; + if ((val >> 16) & (1 << 10)) + virge->streams.k2_horiz_scale |= 0xfffff800; + break; + case 0x81a0: + virge->streams.blend_ctrl = val; + break; + case 0x81c0: + virge->streams.pri_fb0 = val & ((virge->memory_size == 8) ? + (val & 0x7fffff) : (val & 0x3fffff)); + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81c4: + virge->streams.pri_fb1 = ((virge->memory_size == 8) ? + (val & 0x7fffff) : (val & 0x3fffff)); + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81c8: + virge->streams.pri_stride = val & 0xfff; + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81cc: + virge->streams.buffer_ctrl = val; + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81d0: + virge->streams.sec_fb0 = val; + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81d4: + virge->streams.sec_fb1 = val; + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81d8: + virge->streams.sec_stride = val; + s3_virge_update_buffer(virge); + svga->fullchange = changeframecount; + break; + case 0x81dc: + virge->streams.overlay_ctrl = val; + break; + case 0x81e0: + virge->streams.k1_vert_scale = val & 0x7ff; + if (val & (1 << 10)) + virge->streams.k1_vert_scale |= 0xfffff800; + break; + case 0x81e4: + virge->streams.k2_vert_scale = val & 0x7ff; + if (val & (1 << 10)) + virge->streams.k2_vert_scale |= 0xfffff800; + break; + case 0x81e8: + virge->streams.dda_vert_accumulator = val & 0xfff; + if (val & (1 << 11)) + virge->streams.dda_vert_accumulator |= 0xfffff800; + break; + case 0x81ec: + virge->streams.fifo_ctrl = val; + break; + case 0x81f0: + virge->streams.pri_start = val; + virge->streams.pri_x = (val >> 16) & 0x7ff; + virge->streams.pri_y = val & 0x7ff; + svga_recalctimings(svga); + svga->fullchange = changeframecount; + break; + case 0x81f4: + virge->streams.pri_size = val; + virge->streams.pri_w = (val >> 16) & 0x7ff; + virge->streams.pri_h = val & 0x7ff; + svga_recalctimings(svga); + svga->fullchange = changeframecount; + break; + case 0x81f8: + virge->streams.sec_start = val; + virge->streams.sec_x = (val >> 16) & 0x7ff; + virge->streams.sec_y = val & 0x7ff; + svga_recalctimings(svga); + svga->fullchange = changeframecount; + break; + case 0x81fc: + virge->streams.sec_size = val; + virge->streams.sec_w = (val >> 16) & 0x7ff; + virge->streams.sec_h = val & 0x7ff; + svga_recalctimings(svga); + svga->fullchange = changeframecount; + break; - case 0x8504: - virge->subsys_stat &= ~(val & 0xff); - virge->subsys_cntl = (val >> 8); - s3_virge_update_irqs(virge); - break; + case 0x8504: + virge->subsys_stat &= ~(val & 0xff); + virge->subsys_cntl = (val >> 8); + s3_virge_update_irqs(virge); + break; - case 0x850c: - virge->advfunc_cntl = val & 0xff; - s3_virge_updatemapping(virge); - break; + case 0x850c: + virge->advfunc_cntl = val & 0xff; + s3_virge_updatemapping(virge); + break; - case 0xff20: - s3_virge_mmio_write(addr, val, priv); - break; + case 0x8590: + virge->cmd_dma_base = val; + virge->cmd_dma_buf_size = (val & 2) ? 0x10000 : 0x1000; + virge->cmd_dma_buf_size_mask = virge->cmd_dma_buf_size - 1; + virge->cmd_base_addr = (val & 2) ? (val & 0xffff0000) : (val & 0xfffff000); + break; + + case 0x8594: + virge->cmd_dma_write_ptr_update = val & (1 << 16); + if (virge->cmd_dma_write_ptr_update) { + virge->cmd_dma_write_ptr_reg = (virge->cmd_dma_buf_size == 0x10000) ? (val & 0xffff) : (val & 0xfff); + virge->dma_dbl_words = 0; + virge->dma_data_type = 0; + virge->dma_val = 0; + if (virge->cmd_dma) { + while (virge->cmd_dma_read_ptr_reg != virge->cmd_dma_write_ptr_reg) { + virge->cmd_dma_write_ptr_update = 0; + dma_bm_read(virge->cmd_base_addr + virge->cmd_dma_read_ptr_reg, (uint8_t *)&virge->dma_val, 4, 4); + if (!virge->dma_dbl_words) { + virge->dma_dbl_words = (virge->dma_val & 0xffff); + virge->dma_data_type = !!(virge->dma_val & (1 << 31)); + if (virge->dma_data_type) + virge->dma_mmio_addr = 0; + else + virge->dma_mmio_addr = ((virge->dma_val >> 16) << 2) & 0xfffc; + } else { + s3_virge_mmio_write_l(virge->dma_mmio_addr, virge->dma_val, virge); + virge->dma_dbl_words--; + virge->dma_mmio_addr = (virge->dma_mmio_addr + 4) & 0xfffc; + } + virge->cmd_dma_read_ptr_reg = (virge->cmd_dma_read_ptr_reg + 4) & (virge->cmd_dma_buf_size_mask - 3); + } + } + } + break; + + case 0x8598: + virge->cmd_dma_read_ptr_reg = (virge->cmd_dma_buf_size == 0x10000) ? (val & 0xffff) : (val & 0xfff); + break; + + case 0x859c: + virge->cmd_dma = val & 1; + virge->cmd_dma_write_ptr_reg = 0; + virge->cmd_dma_read_ptr_reg = 0; + break; + + case 0xff20: + s3_virge_mmio_write(addr, val, priv); + break; + } } } @@ -2010,22 +2076,22 @@ s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv) { do { \ switch (bpp) { \ case 0: /*8 bpp*/ \ - val = vram[addr & svga->vram_mask]; \ + val = vram[addr & virge->vram_mask]; \ break; \ case 1: /*16 bpp*/ \ - val = *(uint16_t *)&vram[addr & svga->vram_mask]; \ + val = *(uint16_t *)&vram[addr & virge->vram_mask]; \ break; \ case 2: /*24 bpp*/ \ - val = (*(uint32_t *)&vram[addr & svga->vram_mask]) & 0xffffff; \ + val = (*(uint32_t *)&vram[addr & virge->vram_mask]) & 0xffffff; \ break; \ } \ } while (0) -#define Z_READ(addr) *(uint16_t *)&vram[addr & svga->vram_mask] +#define Z_READ(addr) *(uint16_t *)&vram[addr & virge->vram_mask] #define Z_WRITE(addr, val) \ if (!(s3d_tri->cmd_set & CMD_SET_ZB_MODE)) \ - *(uint16_t *)&vram[addr & svga->vram_mask] = val + *(uint16_t *)&vram[addr & virge->vram_mask] = val #define CLIP(x, y) \ do { \ @@ -2040,7 +2106,7 @@ s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv) { if ((s3d_tri->cmd_set & CMD_SET_HC) && (x < s3d_tri->clip_l || \ x > s3d_tri->clip_r || y < s3d_tri->clip_t || \ y > s3d_tri->clip_b)) \ - update = 0; \ + update = 0; \ } while (0) #define Z_CLIP(Zzb, Zs) \ @@ -2111,19 +2177,19 @@ s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv) { do { \ switch (bpp) { \ case 0: /*8 bpp*/ \ - vram[addr & svga->vram_mask] = val; \ - virge->svga.changedvram[(addr & svga->vram_mask) >> 12] = \ + vram[addr & virge->vram_mask] = val; \ + virge->svga.changedvram[(addr & virge->vram_mask) >> 12] = \ changeframecount; \ break; \ case 1: /*16 bpp*/ \ - *(uint16_t *)&vram[addr & svga->vram_mask] = val; \ - virge->svga.changedvram[(addr & svga->vram_mask) >> 12] = \ + *(uint16_t *)&vram[addr & virge->vram_mask] = val; \ + virge->svga.changedvram[(addr & virge->vram_mask) >> 12] = \ changeframecount; \ break; \ case 2: /*24 bpp*/ \ - *(uint32_t *)&vram[addr & svga->vram_mask] = (val & 0xffffff) |\ - (vram[(addr + 3) & svga->vram_mask] << 24); \ - virge->svga.changedvram[(addr & svga->vram_mask) >> 12] = \ + *(uint32_t *)&vram[addr & virge->vram_mask] = (val & 0xffffff) |\ + (vram[(addr + 3) & virge->vram_mask] << 24); \ + virge->svga.changedvram[(addr & virge->vram_mask) >> 12] = \ changeframecount; \ break; \ } \ @@ -2131,7 +2197,6 @@ s3_virge_mmio_write_l(uint32_t addr, uint32_t val, void *priv) { static void s3_virge_bitblt(virge_t *virge, int count, uint32_t cpu_dat) { - svga_t *svga = &virge->svga; uint8_t *vram = virge->svga.vram; uint32_t mono_pattern[64]; int count_mask; @@ -2192,13 +2257,13 @@ s3_virge_bitblt(virge_t *virge, int count, uint32_t cpu_dat) { for (y = 0; y < 4; y++) { for (x = 0; x < 8; x++) { if (virge->s3d.mono_pat_0 & (1 << (x + y * 8))) - mono_pattern[y * 8 + x] = virge->s3d.pat_fg_clr; + mono_pattern[y * 8 + (7 - x)] = virge->s3d.pat_fg_clr; else - mono_pattern[y * 8 + x] = virge->s3d.pat_bg_clr; + mono_pattern[y * 8 + (7 - x)] = virge->s3d.pat_bg_clr; if (virge->s3d.mono_pat_1 & (1 << (x + y * 8))) - mono_pattern[(y + 4) * 8 + x] = virge->s3d.pat_fg_clr; + mono_pattern[(y + 4) * 8 + (7 - x)] = virge->s3d.pat_fg_clr; else - mono_pattern[(y + 4) * 8 + x] = virge->s3d.pat_bg_clr; + mono_pattern[(y + 4) * 8 + (7 - x)] = virge->s3d.pat_bg_clr; } } } @@ -2228,7 +2293,7 @@ s3_virge_bitblt(virge_t *virge, int count, uint32_t cpu_dat) { uint32_t dest_addr = virge->s3d.dest_base + (virge->s3d.dest_x * x_mul) + (virge->s3d.dest_y * virge->s3d.dest_str); uint32_t source = 0; - uint32_t dest; + uint32_t dest = 0; uint32_t pattern; uint32_t out = 0; int update = 1; @@ -2333,7 +2398,7 @@ s3_virge_bitblt(virge_t *virge, int count, uint32_t cpu_dat) { uint32_t dest_addr = virge->s3d.dest_base + (virge->s3d.dest_x * x_mul) + (virge->s3d.dest_y * virge->s3d.dest_str); uint32_t source = 0; - uint32_t dest; + uint32_t dest = 0; uint32_t pattern = virge->s3d.pat_fg_clr; uint32_t out = 0; int update = 1; @@ -2398,7 +2463,7 @@ s3_virge_bitblt(virge_t *virge, int count, uint32_t cpu_dat) { uint32_t dest_addr = virge->s3d.dest_base + (x * x_mul) + (virge->s3d.dest_y * virge->s3d.dest_str); uint32_t source = 0; - uint32_t dest; + uint32_t dest = 0; uint32_t pattern; uint32_t out = 0; int update = 1; @@ -2454,7 +2519,7 @@ skip_line: do { uint32_t dest_addr = virge->s3d.dest_base + (x * x_mul) + (y * virge->s3d.dest_str); uint32_t source = 0; - uint32_t dest; + uint32_t dest = 0; uint32_t pattern; uint32_t out = 0; int update = 1; @@ -3163,7 +3228,6 @@ dest_pixel_lit_texture_modulate(s3d_state_t *state) { static void tri(virge_t *virge, s3d_t *s3d_tri, s3d_state_t *state, int yc, int32_t dx1, int32_t dx2) { - svga_t *svga = &virge->svga; uint8_t *vram = virge->svga.vram; int x_dir = s3d_tri->tlr ? 1 : -1; int use_z = !(s3d_tri->cmd_set & CMD_SET_ZB_MODE); @@ -3283,7 +3347,7 @@ tri(virge_t *virge, s3d_t *s3d_tri, s3d_state_t *state, int yc, int32_t dx1, int } } - virge->svga.changedvram[(dest_offset & svga->vram_mask) >> 12] = changeframecount; + virge->svga.changedvram[(dest_offset & virge->vram_mask) >> 12] = changeframecount; dest_addr = dest_offset + (x * (bpp + 1)); z_addr = z_offset + (x << 1); @@ -3314,23 +3378,24 @@ tri(virge_t *virge, s3d_t *s3d_tri, s3d_state_t *state, int yc, int32_t dx1, int state->dest_rgba.g = ((state->dest_rgba.g * a) + (s3d_tri->fog_g * (255 - a))) / 255; state->dest_rgba.b = ((state->dest_rgba.b * a) + (s3d_tri->fog_b * (255 - a))) / 255; } - + if (s3d_tri->cmd_set & CMD_SET_ABC_ENABLE) { uint32_t src_col; int src_r = 0; uint32_t src_g = 0; uint32_t src_b = 0; + switch (bpp) { case 0: /*8 bpp*/ /*Not implemented yet*/ break; case 1: /*16 bpp*/ - src_col = *(uint16_t *)&vram[dest_addr & svga->vram_mask]; + src_col = *(uint16_t *)&vram[dest_addr & virge->vram_mask]; RGB15_TO_24(src_col, src_r, src_g, src_b); break; case 2: /*24 bpp*/ - src_col = (*(uint32_t *)&vram[dest_addr & svga->vram_mask]) & 0xffffff; + src_col = (*(uint32_t *)&vram[dest_addr & virge->vram_mask]) & 0xffffff; RGB24_TO_24(src_col, src_r, src_g, src_b); break; } @@ -3377,7 +3442,7 @@ tri(virge_t *virge, s3d_t *s3d_tri, s3d_state_t *state, int yc, int32_t dx1, int virge->pixel_count++; } } - + tri_skip_line: state->x1 += dx1; state->x2 += dx2; @@ -3543,7 +3608,8 @@ s3_virge_triangle(virge_t *virge, s3d_t *s3d_tri) { } static void -render_thread(void *param) { +render_thread(void *param) +{ virge_t *virge = (virge_t *)param; while (virge->render_thread_run) { @@ -3554,7 +3620,7 @@ render_thread(void *param) { s3_virge_triangle(virge, &virge->s3d_buffer[virge->s3d_read_idx & RB_MASK]); virge->s3d_read_idx++; - if (RB_ENTRIES == (RB_SIZE - 1)) + if (RB_ENTRIES == RB_MASK) thread_set_event(virge->not_full_event); } virge->s3d_busy = 0; @@ -3564,7 +3630,8 @@ render_thread(void *param) { } static void -queue_triangle(virge_t *virge) { +queue_triangle(virge_t *virge) +{ if (RB_FULL) { thread_reset_event(virge->not_full_event); if (RB_FULL) @@ -4342,7 +4409,7 @@ s3_virge_init(const device_t *info) io_sethandler(0x03c0, 0x0020, s3_virge_in, NULL, NULL, s3_virge_out, NULL, NULL, virge); - virge->pci_regs[PCI_REG_COMMAND] = 3; + virge->pci_regs[PCI_REG_COMMAND] = 7; virge->pci_regs[0x05] = 0; virge->pci_regs[0x06] = 0; virge->pci_regs[0x07] = 2; diff --git a/src/video/vid_svga_render.c b/src/video/vid_svga_render.c index 74d9a4d0c..c05d308fa 100644 --- a/src/video/vid_svga_render.c +++ b/src/video/vid_svga_render.c @@ -178,14 +178,14 @@ svga_render_text_40(svga_t *svga) charaddr = svga->charseta + (chr * 128); if (drawcursor) { - bg = svga->pallook[svga->egapal[attr & 15]]; - fg = svga->pallook[svga->egapal[attr >> 4]]; + bg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + fg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; } else { - fg = svga->pallook[svga->egapal[attr & 15]]; - bg = svga->pallook[svga->egapal[attr >> 4]]; + fg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + bg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; if (attr & 0x80 && svga->attrregs[0x10] & 8) { - bg = svga->pallook[svga->egapal[(attr >> 4) & 7]]; + bg = svga->pallook[svga->egapal[(attr >> 4) & 7] & svga->dac_mask]; if (svga->blink & 16) fg = bg; } @@ -256,13 +256,13 @@ svga_render_text_80(svga_t *svga) charaddr = svga->charseta + (chr * 128); if (drawcursor) { - bg = svga->pallook[svga->egapal[attr & 15]]; - fg = svga->pallook[svga->egapal[attr >> 4]]; + bg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + fg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; } else { - fg = svga->pallook[svga->egapal[attr & 15]]; - bg = svga->pallook[svga->egapal[attr >> 4]]; + fg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + bg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; if (attr & 0x80 && svga->attrregs[0x10] & 8) { - bg = svga->pallook[svga->egapal[(attr >> 4) & 7]]; + bg = svga->pallook[svga->egapal[(attr >> 4) & 7] & svga->dac_mask]; if (svga->blink & 16) fg = bg; } @@ -323,13 +323,13 @@ svga_render_text_80_ksc5601(svga_t *svga) attr = svga->vram[addr + 1]; if (drawcursor) { - bg = svga->pallook[svga->egapal[attr & 15]]; - fg = svga->pallook[svga->egapal[attr >> 4]]; + bg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + fg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; } else { - fg = svga->pallook[svga->egapal[attr & 15]]; - bg = svga->pallook[svga->egapal[attr >> 4]]; + fg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + bg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; if (attr & 0x80 && svga->attrregs[0x10] & 8) { - bg = svga->pallook[svga->egapal[(attr >> 4) & 7]]; + bg = svga->pallook[svga->egapal[(attr >> 4) & 7] & svga->dac_mask]; if (svga->blink & 16) fg = bg; } @@ -378,13 +378,13 @@ svga_render_text_80_ksc5601(svga_t *svga) attr = svga->vram[((svga->ma << 1) + 1) & svga->vram_display_mask]; if (drawcursor) { - bg = svga->pallook[svga->egapal[attr & 15]]; - fg = svga->pallook[svga->egapal[attr >> 4]]; + bg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + fg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; } else { - fg = svga->pallook[svga->egapal[attr & 15]]; - bg = svga->pallook[svga->egapal[attr >> 4]]; + fg = svga->pallook[svga->egapal[attr & 15] & svga->dac_mask]; + bg = svga->pallook[svga->egapal[attr >> 4] & svga->dac_mask]; if (attr & 0x80 && svga->attrregs[0x10] & 8) { - bg = svga->pallook[svga->egapal[(attr >> 4) & 7]]; + bg = svga->pallook[svga->egapal[(attr >> 4) & 7] & svga->dac_mask]; if (svga->blink & 16) fg = bg; } @@ -468,14 +468,14 @@ svga_render_2bpp_s3_lowres(svga_t *svga) else svga->ma += 4; svga->ma &= svga->vram_mask; - p[0] = p[1] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3]]; - p[2] = p[3] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3]]; - p[4] = p[5] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3]]; - p[6] = p[7] = svga->pallook[svga->egapal[dat[0] & 3]]; - p[8] = p[9] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3]]; - p[10] = p[11] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3]]; - p[12] = p[13] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3]]; - p[14] = p[15] = svga->pallook[svga->egapal[dat[1] & 3]]; + p[0] = p[1] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3] & svga->dac_mask]; + p[2] = p[3] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3] & svga->dac_mask]; + p[4] = p[5] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3] & svga->dac_mask]; + p[6] = p[7] = svga->pallook[svga->egapal[dat[0] & 3] & svga->dac_mask]; + p[8] = p[9] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3] & svga->dac_mask]; + p[10] = p[11] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3] & svga->dac_mask]; + p[12] = p[13] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3] & svga->dac_mask]; + p[14] = p[15] = svga->pallook[svga->egapal[dat[1] & 3] & svga->dac_mask]; p += 16; } } @@ -501,14 +501,14 @@ svga_render_2bpp_s3_lowres(svga_t *svga) svga->ma &= svga->vram_mask; - p[0] = p[1] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3]]; - p[2] = p[3] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3]]; - p[4] = p[5] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3]]; - p[6] = p[7] = svga->pallook[svga->egapal[dat[0] & 3]]; - p[8] = p[9] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3]]; - p[10] = p[11] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3]]; - p[12] = p[13] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3]]; - p[14] = p[15] = svga->pallook[svga->egapal[dat[1] & 3]]; + p[0] = p[1] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3] & svga->dac_mask]; + p[2] = p[3] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3] & svga->dac_mask]; + p[4] = p[5] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3] & svga->dac_mask]; + p[6] = p[7] = svga->pallook[svga->egapal[dat[0] & 3] & svga->dac_mask]; + p[8] = p[9] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3] & svga->dac_mask]; + p[10] = p[11] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3] & svga->dac_mask]; + p[12] = p[13] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3] & svga->dac_mask]; + p[14] = p[15] = svga->pallook[svga->egapal[dat[1] & 3] & svga->dac_mask]; p += 16; } @@ -566,14 +566,14 @@ svga_render_2bpp_s3_highres(svga_t *svga) else svga->ma += 4; svga->ma &= svga->vram_mask; - p[0] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3]]; - p[1] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3]]; - p[2] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3]]; - p[3] = svga->pallook[svga->egapal[dat[0] & 3]]; - p[4] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3]]; - p[5] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3]]; - p[6] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3]]; - p[7] = svga->pallook[svga->egapal[dat[1] & 3]]; + p[0] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3] & svga->dac_mask]; + p[1] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3] & svga->dac_mask]; + p[2] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3] & svga->dac_mask]; + p[3] = svga->pallook[svga->egapal[dat[0] & 3] & svga->dac_mask]; + p[4] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3] & svga->dac_mask]; + p[5] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3] & svga->dac_mask]; + p[6] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3] & svga->dac_mask]; + p[7] = svga->pallook[svga->egapal[dat[1] & 3] & svga->dac_mask]; p += 8; } } @@ -599,14 +599,14 @@ svga_render_2bpp_s3_highres(svga_t *svga) svga->ma &= svga->vram_mask; - p[0] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3]]; - p[1] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3]]; - p[2] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3]]; - p[3] = svga->pallook[svga->egapal[dat[0] & 3]]; - p[4] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3]]; - p[5] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3]]; - p[6] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3]]; - p[7] = svga->pallook[svga->egapal[dat[1] & 3]]; + p[0] = svga->pallook[svga->egapal[(dat[0] >> 6) & 3] & svga->dac_mask]; + p[1] = svga->pallook[svga->egapal[(dat[0] >> 4) & 3] & svga->dac_mask]; + p[2] = svga->pallook[svga->egapal[(dat[0] >> 2) & 3] & svga->dac_mask]; + p[3] = svga->pallook[svga->egapal[dat[0] & 3] & svga->dac_mask]; + p[4] = svga->pallook[svga->egapal[(dat[1] >> 6) & 3] & svga->dac_mask]; + p[5] = svga->pallook[svga->egapal[(dat[1] >> 4) & 3] & svga->dac_mask]; + p[6] = svga->pallook[svga->egapal[(dat[1] >> 2) & 3] & svga->dac_mask]; + p[7] = svga->pallook[svga->egapal[dat[1] & 3] & svga->dac_mask]; p += 8; } @@ -652,17 +652,17 @@ svga_render_2bpp_headland_highres(svga_t *svga) svga->ma &= svga->vram_mask; dat = edatlookup[edat[0] >> 6][edat[1] >> 6] | (edatlookup[edat[2] >> 6][edat[3] >> 6] << 2); - p[0] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask]]; - p[1] = svga->pallook[svga->egapal[dat & svga->plane_mask]]; + p[0] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask] & svga->dac_mask]; + p[1] = svga->pallook[svga->egapal[dat & svga->plane_mask] & svga->dac_mask]; dat = edatlookup[(edat[0] >> 4) & 3][(edat[1] >> 4) & 3] | (edatlookup[(edat[2] >> 4) & 3][(edat[3] >> 4) & 3] << 2); - p[2] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask]]; - p[3] = svga->pallook[svga->egapal[dat & svga->plane_mask]]; + p[2] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask] & svga->dac_mask]; + p[3] = svga->pallook[svga->egapal[dat & svga->plane_mask] & svga->dac_mask]; dat = edatlookup[(edat[0] >> 2) & 3][(edat[1] >> 2) & 3] | (edatlookup[(edat[2] >> 2) & 3][(edat[3] >> 2) & 3] << 2); - p[4] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask]]; - p[5] = svga->pallook[svga->egapal[dat & svga->plane_mask]]; + p[4] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask] & svga->dac_mask]; + p[5] = svga->pallook[svga->egapal[dat & svga->plane_mask] & svga->dac_mask]; dat = edatlookup[edat[0] & 3][edat[1] & 3] | (edatlookup[edat[2] & 3][edat[3] & 3] << 2); - p[6] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask]]; - p[7] = svga->pallook[svga->egapal[dat & svga->plane_mask]]; + p[6] = svga->pallook[svga->egapal[(dat >> 4) & svga->plane_mask] & svga->dac_mask]; + p[7] = svga->pallook[svga->egapal[dat & svga->plane_mask] & svga->dac_mask]; p += 8; } @@ -868,8 +868,8 @@ svga_render_indexed_gfx(svga_t *svga, bool highres, bool combine8bits) } } else if (combine8bits) { if (svga->packed_4bpp) { - uint32_t p0 = svga->map8[c0]; - uint32_t p1 = svga->map8[c1]; + uint32_t p0 = svga->map8[c0 & svga->dac_mask]; + uint32_t p1 = svga->map8[c1 & svga->dac_mask]; const int outoffs = i << dwshift; for (int subx = 0; subx < dotwidth; subx++) p[outoffs + subx] = p0; @@ -877,14 +877,14 @@ svga_render_indexed_gfx(svga_t *svga, bool highres, bool combine8bits) p[outoffs + subx + dotwidth] = p1; } else { uint32_t ccombined = (c0 << 4) | c1; - uint32_t p0 = svga->map8[ccombined]; + uint32_t p0 = svga->map8[ccombined & svga->dac_mask]; const int outoffs = (i >> 1) << dwshift; for (int subx = 0; subx < dotwidth; subx++) p[outoffs + subx] = p0; } } else { - uint32_t p0 = svga->pallook[svga->egapal[c0]]; - uint32_t p1 = svga->pallook[svga->egapal[c1]]; + uint32_t p0 = svga->pallook[svga->egapal[c0] & svga->dac_mask]; + uint32_t p1 = svga->pallook[svga->egapal[c1] & svga->dac_mask]; const int outoffs = i << dwshift; for (int subx = 0; subx < dotwidth; subx++) p[outoffs + subx] = p0; @@ -934,16 +934,16 @@ svga_render_8bpp_clone_highres(svga_t *svga) for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 8) { dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; dat = *(uint32_t *) (&svga->vram[(svga->ma + 4) & svga->vram_display_mask]); - p[4] = svga->map8[dat & 0xff]; - p[5] = svga->map8[(dat >> 8) & 0xff]; - p[6] = svga->map8[(dat >> 16) & 0xff]; - p[7] = svga->map8[(dat >> 24) & 0xff]; + p[4] = svga->map8[dat & svga->dac_mask & 0xff]; + p[5] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[6] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 8; p += 8; @@ -963,16 +963,16 @@ svga_render_8bpp_clone_highres(svga_t *svga) if (!svga->remap_required) { for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 8) { dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; dat = *(uint32_t *) (&svga->vram[(svga->ma + 4) & svga->vram_display_mask]); - p[4] = svga->map8[dat & 0xff]; - p[5] = svga->map8[(dat >> 8) & 0xff]; - p[6] = svga->map8[(dat >> 16) & 0xff]; - p[7] = svga->map8[(dat >> 24) & 0xff]; + p[4] = svga->map8[dat & svga->dac_mask & 0xff]; + p[5] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[6] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 8; p += 8; @@ -981,10 +981,10 @@ svga_render_8bpp_clone_highres(svga_t *svga) for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 4) { addr = svga->remap_func(svga, svga->ma); dat = *(uint32_t *) (&svga->vram[addr & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 4; p += 4; @@ -1043,10 +1043,10 @@ svga_render_8bpp_lowres(svga_t *svga) if (!svga->remap_required) { for (x = 0; x <= (svga->hdisp + svga->scrollcache); x += 8) { dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); - p[0] = p[1] = svga->map8[dat & 0xff]; - p[2] = p[3] = svga->map8[(dat >> 8) & 0xff]; - p[4] = p[5] = svga->map8[(dat >> 16) & 0xff]; - p[6] = p[7] = svga->map8[(dat >> 24) & 0xff]; + p[0] = p[1] = svga->map8[dat & svga->dac_mask & 0xff]; + p[2] = p[3] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[4] = p[5] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[6] = p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 4; p += 8; @@ -1055,10 +1055,10 @@ svga_render_8bpp_lowres(svga_t *svga) for (x = 0; x <= (svga->hdisp + svga->scrollcache); x += 8) { addr = svga->remap_func(svga, svga->ma); dat = *(uint32_t *) (&svga->vram[addr & svga->vram_display_mask]); - p[0] = p[1] = svga->map8[dat & 0xff]; - p[2] = p[3] = svga->map8[(dat >> 8) & 0xff]; - p[4] = p[5] = svga->map8[(dat >> 16) & 0xff]; - p[6] = p[7] = svga->map8[(dat >> 24) & 0xff]; + p[0] = p[1] = svga->map8[dat & svga->dac_mask & 0xff]; + p[2] = p[3] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[4] = p[5] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[6] = p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 4; p += 8; @@ -1091,16 +1091,16 @@ svga_render_8bpp_highres(svga_t *svga) for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 8) { dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; dat = *(uint32_t *) (&svga->vram[(svga->ma + 4) & svga->vram_display_mask]); - p[4] = svga->map8[dat & 0xff]; - p[5] = svga->map8[(dat >> 8) & 0xff]; - p[6] = svga->map8[(dat >> 16) & 0xff]; - p[7] = svga->map8[(dat >> 24) & 0xff]; + p[4] = svga->map8[dat & svga->dac_mask & 0xff]; + p[5] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[6] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 8; p += 8; @@ -1120,16 +1120,16 @@ svga_render_8bpp_highres(svga_t *svga) if (!svga->remap_required) { for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 8) { dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; dat = *(uint32_t *) (&svga->vram[(svga->ma + 4) & svga->vram_display_mask]); - p[4] = svga->map8[dat & 0xff]; - p[5] = svga->map8[(dat >> 8) & 0xff]; - p[6] = svga->map8[(dat >> 16) & 0xff]; - p[7] = svga->map8[(dat >> 24) & 0xff]; + p[4] = svga->map8[dat & svga->dac_mask & 0xff]; + p[5] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[6] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[7] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 8; p += 8; @@ -1138,10 +1138,10 @@ svga_render_8bpp_highres(svga_t *svga) for (x = 0; x <= (svga->hdisp /* + svga->scrollcache*/); x += 4) { addr = svga->remap_func(svga, svga->ma); dat = *(uint32_t *) (&svga->vram[addr & svga->vram_display_mask]); - p[0] = svga->map8[dat & 0xff]; - p[1] = svga->map8[(dat >> 8) & 0xff]; - p[2] = svga->map8[(dat >> 16) & 0xff]; - p[3] = svga->map8[(dat >> 24) & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; + p[1] = svga->map8[(dat >> 8) & svga->dac_mask & 0xff]; + p[2] = svga->map8[(dat >> 16) & svga->dac_mask & 0xff]; + p[3] = svga->map8[(dat >> 24) & svga->dac_mask & 0xff]; svga->ma += 4; p += 4; @@ -1173,19 +1173,19 @@ svga_render_8bpp_tseng_lowres(svga_t *svga) dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[0] = p[1] = svga->map8[dat & 0xff]; + p[0] = p[1] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[2] = p[3] = svga->map8[dat & 0xff]; + p[2] = p[3] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[4] = p[5] = svga->map8[dat & 0xff]; + p[4] = p[5] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[6] = p[7] = svga->map8[dat & 0xff]; + p[6] = p[7] = svga->map8[dat & svga->dac_mask & 0xff]; svga->ma += 4; p += 8; @@ -1214,36 +1214,36 @@ svga_render_8bpp_tseng_highres(svga_t *svga) dat = *(uint32_t *) (&svga->vram[svga->ma & svga->vram_display_mask]); if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[0] = svga->map8[dat & 0xff]; + p[0] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[1] = svga->map8[dat & 0xff]; + p[1] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[2] = svga->map8[dat & 0xff]; + p[2] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[3] = svga->map8[dat & 0xff]; + p[3] = svga->map8[dat & svga->dac_mask & 0xff]; dat = *(uint32_t *) (&svga->vram[(svga->ma + 4) & svga->vram_display_mask]); if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[4] = svga->map8[dat & 0xff]; + p[4] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[5] = svga->map8[dat & 0xff]; + p[5] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[6] = svga->map8[dat & 0xff]; + p[6] = svga->map8[dat & svga->dac_mask & 0xff]; dat >>= 8; if (svga->attrregs[0x10] & 0x80) dat = (dat & ~0xf0) | ((svga->attrregs[0x14] & 0x0f) << 4); - p[7] = svga->map8[dat & 0xff]; + p[7] = svga->map8[dat & svga->dac_mask & 0xff]; svga->ma += 8; p += 8;